RHEUMATOID ARTHRITIS THE MANAGEMENT AND TREATMENT OF RHEUMATOID by jizhen1947

VIEWS: 80 PAGES: 307

									     DRAFT FOR CONSULTATION


1                                      The National Collaborating Centre
2                                                for Chronic Conditions
3
4                                               Funded to produce guidelines for the NHS by NICE
5
6




7              RHEUMATOID ARTHRITIS: THE
8                   MANAGEMENT AND TREATMENT
9                    OF RHEUMATOID ARTHRITIS IN
10                                        ADULTS
11
12
13                National clinical guideline for rheumatoid arthritis
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29   Published by the Royal College of Physicians




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                       1 of 307
     DRAFT FOR CONSULTATION


1    Contents
2
 3   Guideline Development Group members.........................................................4
 4   Preface ............................................................................................................7
 5   1 Introduction ...............................................................................................9
 6   2 Methodology ...........................................................................................14
 7     About this guideline ....................................................................................14
 8     2.1    Aim ..................................................................................................14
 9     2.2    Scope ..............................................................................................14
10     2.3    Audience..........................................................................................14
11     2.4    Involvement of people with rheumatoid arthritis...............................15
12     2.5    Guideline limitations ........................................................................15
13     2.6    Other work relevant to the guideline ................................................15
14     2.7    Background .....................................................................................16
15     2.8    The process of guideline development ............................................18
16     2.9    Disclaimer........................................................................................22
17     2.10 Funding............................................................................................22
18   3 Key messages of the guideline ...............................................................23
19     3.1    Priorities for implementation ............................................................23
20     3.2    Algorithms........................................................................................24
21     THE GUIDELINE........................................................................................25
22   4 Identification and diagnosis ....................................................................26
23     4.2    Investigations...................................................................................36
24     4.3    Presenting symptoms and signs......................................................43
25   5 The patient..............................................................................................57
26     5.1    Patient perceptions and beliefs........................................................57
27     5.2    Patient education.............................................................................68
28   6 The multidisciplinary team ......................................................................77
29     6.1    The multidisciplinary team ...............................................................77
30     6.2    Physiotherapy..................................................................................82
31     6.3    Occupational therapy.......................................................................94
32     6.4    Podiatry .........................................................................................102
33   7 Pharmacological management .............................................................106
34     7.1    Symptom control............................................................................106
35     7.2    Glucocorticoids ..............................................................................126
36     7.3    Disease modifying and biological drugs: when to introduce, and
37     optimal sequencing ..................................................................................134
38     7.4    Disease modifying and biological drugs: relative merits ................162
39     7.5    Disease modifying and biological drugs: when to withdraw them ..187
40   8 Monitoring the course of RA .................................................................195
41     8.1    Monitoring disease ........................................................................195
42     8.2    Content and frequency of review ...................................................200
43     8.3    Timing and referral for surgery ......................................................203
44   9 Other aspects and treatment ................................................................211
45     9.1    Diet ................................................................................................211
46     9.2    Complementary and alternative interventions................................217
47   10    Areas for future research ..................................................................220
48   11    Abbreviations ....................................................................................221
49   12    Glossary ............................................................................................224


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                                2 of 307
     DRAFT FOR CONSULTATION


 1   Appendix A ..................................................................................................228
 2   Appendix B: SCOPE ....................................................................................235
 3   Appendix C ..................................................................................................239
 4     A.1. Introduction .......................................................................................240
 5     A.2. Background.......................................................................................240
 6     A.3. Comparator Treatments ....................................................................241
 7     A.4. Model Structure.................................................................................244
 8     A.5. Model Inputs .....................................................................................247
 9     A.6. Results ..............................................................................................256
10     A.7. Discussion.........................................................................................260
11     A.8. Conclusions ......................................................................................265
12




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                            3 of 307
     DRAFT FOR CONSULTATION



1    Guideline Development Group members
2
 3   Dr Michael Rudolf
 4   NCC-CC Chair/Respiratory Physician, Ealing Hospital NHS Trust
 5
 6   Dr Chris Deighton
 7   NCC-CC Clinical Advisor/ Consultant Rheumatologist, Derby Hospitals NHS
 8   Foundation Trust
 9
10   Ms Jaim Sutton
11   NCC-CC Project Manager - until March 2008
12
13   Ms Claire Turner
14   NCC-CC Senior Project Manager - from April 2008
15
16   Dr Rachel O’ Mahony
17   NCC-CC Research Fellow
18
19   Mr Jonathan Tosh
20   ScHARR Health Economist, University of Sheffield
21
22   Mrs Alison Richards
23   NCC-CC Information Scientist
24
25   Mrs Isabel Raiman
26   Community Based Nurse, Brighton and Hove City Teaching PCT
27
28   Ms Sheena Hennell
29   Specialist Nurse, Royal Liverpool University Hospital
30
31   Dr Louise Warburton
32   General Practitioner (specialist in RA), Shrewsbury
33
34   Professor David L Scott
35   Consultant Rheumatologist, Kings College Hospital
36
37   Dr Alison Hammond
38   Reader, University of Salford
39
40   Mrs Enid Quest
41   Patient and Carer Representative, Bristol
42
43   Mrs Ailsa Bosworth
44   Patient and Carer Representative, Maidenhead
45
46   Dr David Morgan
47   General Practitioner (non RA specialist), Birmingham


     Rheumatoid arthritis: full guideline DRAFT (August 2008)         4 of 307
     DRAFT FOR CONSULTATION


 1
 2   Dr Jane Hall
 3   Senior Clinical Research Physiotherapist and Hon. Sen. Lecturer
 4   RACE/Physiotherapy Royal National Hospital for Rheumatic Diseases and
 5   University of Bath
 6
 7   Dr Patrick Kiely
 8   Consultant Rheumatologist, St George's NHS Healthcare Trust, London.
 9
10   Dr Raashid Luqmani
11   Consultant Rheumatologist, Nuffield Orthopaedic Centre and University of
12   Oxford
13
14   Ms Zara Bingham
15   Patient and carer representative (acted as a deputy for Alisa Bosworth at a
16   GDG meeting), Manchester
17
18   Dr Paul D’Orso
19   General Practitioner (non RA specialist), Birmingham (acted as a deputy for
20   Dr Morgan)
21
22   Dr Joanna Sheldon
23   Immunologist, St George’s Hospital, London (invited expert attended one
24   GDG meeting)
25
26   Dr Colin Howie
27   Consultant Orthopaedic and Trauma Surgeon, Edinburgh New Royal
28   Infirmary (invited expert attended one GDG meeting)
29
30   Dr Andrew Robinson
31   Consultant Foot and Ankle Surgeon, Cambridge University Hospital NHS
32   Trust (invited expert attended one GDG meeting)
33
34   Dr Anthony Redmond
35   Podiatrist & Senior Lecturer, University of Leeds (invited expert attended one
36   GDG meeting)
37
38
39
40
41
42
43
44




     Rheumatoid arthritis: full guideline DRAFT (August 2008)               5 of 307
     DRAFT FOR CONSULTATION


1    Acknowledgements
2
3    The Guideline Development Group is grateful to the following people for their
4    valuable contributions to the development of this guideline:
5
6         •   Dr Bernard Higgins, Director, NCC-CC
7         •   Ms Jane Ingham, Assistant Director, NCC-CC
8         •   Ms Susan Tann, Administrator, NCC-CC
9         •   Ms Jill Parnham, Assistant Director Operations,, NCC-CC
10        •   Dr Allan Wailoo, Senior Health Economist, ScHARR
11        •   Dr Alan Brennan,ScHARR
12        •   Dr Joanne Lord, Health economics advisor, NICE
13        •   Mr Nick Latimer, Health Economist, NCC-CC
14        •   Dr Anthony Redmond, Podiatrist
15        •   Dr Colin Howie, Surgeon
16        •   Dr Andrew Robinson, Surgeon
17        •   Dr Joanne Sheldon, Immunologist
18        •   Mr Rob Grant, Senior Technical Advisor
19
20        •   Thanks also to the following NCC-CC research fellows Anupam
21            Garrib, Jose Diaz, Emily Crow and Sharon Swain.
22
23




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              6 of 307
    DRAFT FOR CONSULTATION


1   Preface
2
3
4   To be inserted after consultation




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   7 of 307
    DRAFT FOR CONSULTATION




1




2         DEVELOPMENT OF THE GUIDELINE




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   8 of 307
     DRAFT FOR CONSULTATION


1    1      Introduction
 2   Background
 3   Rheumatoid arthritis (RA) is an inflammatory disease that exerts its greatest
 4   impact on the joints of the body which are lined with a specialised tissue
 5   responsible for maintaining the nutrition and lubrication of the joint called
 6   synovium (synovial joints). The distribution of joints affected is characteristic.
 7   It typically affects the small joints of the hands and the feet, and usually both
 8   sides equally in a symmetrical distribution, though any synovial joint can be
 9   affected. In patients with established and aggressive disease, most joints
10   unfortunately will be affected over time.
11
12   The initial trigger for RA is unknown, but there is evidence to suggest
13   abnormalities in components of the immune system that lead to the body
14   developing immune and inflammatory reactions particularly in joints. These
15   changes may precede the symptomatic onset of RA by many years. Whatever
16   sets the pathology in motion results in a large increase in blood flow to the
17   joint (giving heat and sometimes redness), proliferation of the synovial
18   membrane with an increase in synovial fluid (swelling), and most important to
19   the patient, pain (due to stretching of pain receptors in the soft tissues around,
20   and the bone on either side, of the joint). These features result in rapid loss of
21   muscle around an affected joint, and this, along with pain and swelling,
22   understandably lead to loss of joint function. If the inflammation of the synovial
23   membrane cannot be suppressed it will result in increasing damage to the
24   joint, due to the release of protein-degrading enzymes from inflammatory and
25   other cells, and a conversion of parts of the synovial membrane into an
26   inflammatory tissue called pannus which can invade the bone and cartilage at
27   the margins of the joint. The degree of progressive damage is related to the
28   intensity and duration of the inflammation. Damage to joints results in
29   progressive deformity, disability and handicap. Other structures have synovial
30   linings, such as tendon sheaths, and inflammation of these can result in
31   tendon rupture. Consequently, suppression of inflammation in the early stages
32   of the disease can result in substantial improvements in long-term outcomes
33   for joints and other components of the musculoskeletal system.
34
35   Compounding this widespread inflammatory arthritis is the fact that RA is a
36   misnomer for many patients with the disease, because it affects much more
37   than joints, and is a systemic disease. In all patients the release of large
38   concentrations of proteins that drive inflammatory processes (such as tumour
39   necrosis factor-α (TNF-α)), results in symptoms of profound fatigue, with a
40   feeling of ongoing influenza-like symptoms, and even fever, sweats and
41   weight loss. Furthermore, other body organ systems may be affected by the
42   inflammatory process, with dryness of the eyes and mouth (Sjögren’s
43   syndrome), and nodules (hard lumps particularly over extensor surfaces like
44   the backs of elbows) affecting up to a third of patients. More significant
45   inflammatory manifestations may lead to serious pathology, such as fibrosis in
46   the lungs, inflammation affecting the lining of the heart and lungs (pleural and
47   pericardial effusions), or vasculitis in which inflammation of the inner lining of
48   the blood vessels may lead to potentially devastating effects for whichever

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                   9 of 307
     DRAFT FOR CONSULTATION


 1   organ is supplied by the affected blood vessels (scleritis of the eye, a painful
 2   and potentially sight-threatening vasculitis, or peripheral neuropathy (where
 3   nerves are irreversibly damaged, leading to weakness or sensory
 4   abnormalities). Inflammation of the joints can also be life threatening when it
 5   affects the neck, causing potentially unstable articulations between the bones,
 6   and inflammatory pannus. This combination of bone deformity and swollen
 7   inflammatory tissue can press on the spinal cord leading to ischaemia and
 8   widespread neurological consequences affecting all four limbs, bowel and
 9   bladder function, or the muscles and brain stem controlling centres of
10   respiration, potentially resulting in death.
11
12   Thankfully, these life-threatening inflammatory manifestations of disease are
13   uncommon, and may be becoming rarer. However, it has become increasingly
14   evident that the ongoing inflammation and loss of mobility can have other
15   unforeseen circumstances for people with RA. Heart conditions such as
16   ischaemic heart disease and cardiac failure have been shown to be more
17   common in RA, and result in premature death for many patients.
18   Atherosclerosis (where the inner lining of arteries become progressively
19   thickened and impair blood supply to whichever organ is being served) is
20   driven in part by ongoing inflammation, so that the patients with the most
21   active RA have the greatest risk of heart disease. Osteoporosis is more
22   common, due to reduced mobility, inflammation, and sometimes the drugs the
23   patients are on (particularly steroids). Patients with RA are more prone to
24   infections than the rest of the population, probably due abnormalities in the
25   immune system, and sometimes contributed to by medication.
26   Clearly RA has the potential for not only widespread joint and soft tissue
27   damage, but inflammatory processes that can directly or indirectly affect most
28   organ systems in the body, and result in premature death. Appropriate
29   management therefore needs to address not only the impact on joints, but
30   also focus on the whole body, the person suffering from the disease, their
31   families and carers, and where appropriate their employers.
32
33   Definition
34   The most commonly used classification criteria for RA were drawn together in
35   1987 by a committee of the American College of Rheumatology2 The
36   American Rheumatism Association 1987 revised criteria for the classification
37   of rheumatoid arthritis (Arthritis Rheum 1988;31:315-24) (see table 1 for a
38   summary). It is important to note that these are classification criteria and not
39   diagnostic criteria. In other words, they were designed to facilitate
40   communication between researchers, and ensure more robust research
41   through reliable case definition, so that irrespective of where a study was
42   taking place, everybody means the same thing by the term “RA”. They are not
43   diagnostic criteria (although commonly misnamed as such), because there
44   are no diagnostic tests for RA that differentiate it from normality or from other
45   types of arthritis. The diagnosis is therefore largely a clinical one, relying
46   particularly in the early stages on the history and examination, with tests
47   (blood or imaging) sometimes helping to confirm the most likely diagnosis.
48
49   The 1987 ACR criteria were not designed for clinical practice, although they
50   do influence the way that clinicians think of RA, and do appear in national

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               10 of 307
     DRAFT FOR CONSULTATION


 1   clinical guidelines (e.g. to be eligible to go onto anti-TNF in the UK an RA
 2   patient needs to fulfill the 1987 ACR criteria. However, they are not useful in
 3   discriminating between RA and self-limiting disease in early synovitis that
 4   might evolve eventually into disease that fulfils the criteria. They also exclude
 5   some individuals with a polyarthritis that does not quite fulfill the criteria, but
 6   whose disease closely resembles RA, has a similar course and response to
 7   treatments.
 8
 9   For the purposes of this guideline, the guideline development group (GDG)
10   accepted a clinical diagnosis of RA as being more important than the 1987
11   classification criteria for RA2, because an early persistent synovitis where
12   other pathologies have been ruled out, needs to treated as if it is RA to try to
13   prevent damage to joints. Identification of persistent synovitis, and appropriate
14   early management is more important than whether the disease satisfies
15   classification criteria3. The GDG were therefore keen to take a pragmatic
16   approach and include consideration of all patients with an early or established
17   inflammatory arthritis where other underlying pathologies had been excluded.
18
     ACR 1987 Classification criteria for rheumatoid arthritis
     Patients must have four of the seven criteria:
          • Morning stiffness lasting at least 1 hour*
          •   Swelling in 3 or more joints*
          •   Swelling in hand joints*
          •   Symmetric joint swelling*
          •   Erosions or decalcification on X-ray of hand
          •   Rheumatoid Nodules
          •   Abnormal serum rheumatoid factor
     *Must be present at least 6 weeks
19
20   Table 1. A summary of the 1987 ACR Classification criteria for Rheumatoid
21   Arthritis2
22
23   Prevalence and Incidence
24   The first study to address prevalence of RA in the UK was published in 1961,
25   where Lawrence estimated that 1.1% of the population of Leigh and
26   Wensleydale had RA4. Using the 1987 ACR criteria, Symmons and
27   colleagues came up with a similar figure (0.8%) in a Norfolk population study5.
28   This equates to approximately 400,000 people with this condition in the UK.
29
30   The incidence of the disease is low with a study from Norfolk showing a rate
31   of 1.5 males per 10,000 population per year, and a figure of 3.6 for females6.
32   This translates into approximately 12,000 people being diagnosed with new
33   RA each year in the UK. These figures also illustrate another feature of RA in
34   that the overall occurrence of RA is 2-4 times greater in women than men.
35   The peak age of incidence in the UK for both genders is in the 70’s, but with a
36   long tail on either side, illustrating that all ages can develop the disease.6

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  11 of 307
     DRAFT FOR CONSULTATION


 1   The GDG categorised RA into 2 categories of disease duration; ‘recent onset’
 2   (mean disease duration of ≤2 years) and ‘established disease’ (mean disease
 3   duration of > 2 years) The British Rheumatology Society also took this
 4   approach when developing their guidelines.7
 5
 6   Management principles
 7   The drug management of RA can be considered under two headings. The first
 8   is the relief of symptoms, with pain being the number one priority for patients.
 9   The second is modification of the disease process so that radiological
10   progression, which is closely correlated with progressive functional
11   impairment, can be retarded or stopped.
12
13   Analgesics and non-steroidal anti-inflammatory drugs can be helpful in
14   relieving pain, where they are not contra-indicated. Within the group of non-
15   steroidal anti-inflammatory drugs is a group of drugs that were originally
16   developed with a safer gastrointestinal profile in mind. These are referred to
17   as cox II-inhibitors. The GDG was asked to update the existing technology
18   appraisals on these drugs but were aware that the recently published NICE
19   osteoarthritis management clinical guidelines7,8 addressed this issue, and
20   their recommendations might be relevant to RA management also.
21
22   Conventional disease modifying anti-rheumatic drugs (DMARDs) include
23   methotrexate, sulphasalazine, hydroxychloroquine, leflunomide and gold
24   injections can be helpful in slowing down the damaging component of the
25   disease process, but the precise modes of action for these drugs is still
26   subject to research. The optimal sequencing of DMARDs remains a source of
27   debate, and whether patients should be started on combinations of therapies
28   or single DMARDs is also contentious. The GDG felt that this was such an
29   important area that health economic analyses should focus on this area to
30   attempt to determine which DMARD strategy was most cost-effective.
31
32   A group of drugs, labelled “biological” (because they consist of monoclonal
33   antibodies and soluble receptors that specifically block key drivers of the
34   disease process) have been developed, based on an increasing
35   understanding of the disease pathology. These key drivers include cytokines
36   such as tumour necrosis factor α (TNF) and interleukin-1 (IL-1). An IL1
37   receptor antagonist called anakinra had been appraised by NICE and rejected
38   for use in the NHS as not being cost-effective.9 The GDG were asked to
39   update this technology appraisal.. The GDG were also aware that during their
40   time of developing this management guideline that other multiple and single
41   technology appraisals were taking place on biological therapies for RA:
42   NICE were re-appraising the use of anti-TNF in RA (an update of Guidance
43   on the use of etanercept and infliximab for the treatment of rheumatoid
44   arthritis. NICE technology appraisal guidance 36 (2002). Available from:
45   www.nice.org.uk/TA036). This is due to be published in September 2008.
46   Rituximab (a B-lymphocyte depletory) for the treatment of rheumatoid arthritis.
47   This was subsequently published7,10
48
49   Abatacept (a co-stimulation inhibitor preventing T-lymphocyte activation) for
50   the treatment of rheumatoid arthritis. This was subsequently published7,11

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              12 of 307
     DRAFT FOR CONSULTATION


1    Certolizumab pegol (a TNF inhibitor) for rheumatoid arthritis (TBC).
2
3    Health and resource burden
4    Rheumatoid arthritis can result in a wide range of complications for the
5    individual patient, their carers, the NHS and society in general. The economic
6    impact of this disease includes:
7
8         •   direct cost to the NHS and associated healthcare support services
 9        •   indirect costs to the economy, including the effects of early mortality
10            and lost productivity
11        •   personal impact of RA and subsequent complications on patients and
12            their families
13   Although the course of RA is heterogeneous and variable, within 2 years of
14   diagnosis, patients usually experience moderate disability, and after 10 years
15   30% are severely disabled. Life expectancy in patients with RA is also
16   reduced. For example, a 50-year-old woman with RA is expected to die
17   4 years earlier than a 50-year-old woman without RA.
18
19   Approximately one third of patients cease work because of the disease within
20   two years of onset, and this prevalence increases thereafter. The total costs
21   of RA in the UK, including indirect costs and work related disability, have been
22   estimated at between £3.8 billion and £4.75 billion per year12. A survey by the
23   National Rheumatoid Arthritis Society estimated that when a patient stops
24   work due to RA, it represents an average loss of productivity equivalent to
25   £287, 544.7,13 Clearly this disease represents a huge cost to the UK economy,
26   and an enormous cost to individual patients.
27
28   Living with RA
29   Getting a diagnosis of a painful, life-long disease like RA, can come as a
30   shock and people often get extremely worried about ‘ending up in a
31   wheelchair’. Fortunately if diagnosed and treated early today, that is very
32   unlikely to happen. However, due sometimes to the difficulty of diagnosing
33   RA, it may take several months before someone is diagnosed and during this
34   time, irreversible joint damage is happening.
35
36   If people with RA are asked to single out one symptom which causes the most
37   difficulty, it is pain first, second and third. Pain can be difficult for other people
38   to see, particularly in the early stages of disease when you may not look any
39   different and can therefore cause feelings of isolation and depression. RA
40   impacts on absolutely every area of a person’s life. It changes family
41   dynamics and relationships, can impact at a relatively early stage in the
42   disease on your ability to remain in work and in charge of your life and
43   therefore affect you financially as well. It can be very difficult for people to
44   come to terms with having a disease like RA as you begin to face the fact that
45   you may not be able to do all you wanted to do in life. We all have dreams
46   and hopes for our lives and having to let go of some of those means that you
47   go through a bereavement process which can take time and varies between
48   individuals. It is vital that we do all we can to ensure that people are aware of


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                   13 of 307
     DRAFT FOR CONSULTATION


1    the symptoms of RA so that early diagnosis and treatment gives the best long
2    term outcomes for people.
3

4    2         Methodology

5    About this guideline

6    2.1       Aim
 7   The aim of the National Collaborating Centre for Chronic Conditions (NCC-
 8   CC) is to provide a user-friendly, clinical, evidence-based guideline for the
 9   National Health Service (NHS) in England and Wales that:
10        • offers best clinical advice for the management and treatment of
11            rheumatoid arthritis (RA) in adults in primary and secondary care
12         •    is based on best published clinical and economics evidence,
13              alongside expert consensus
14         •    takes into account patient choice and informed decision-making
15         •    defines the major components of NHS care provision for RA
16         •    details areas of uncertainty or controversy requiring further research
17         •    provides a choice of guideline versions for different audiences.

18   2.2       Scope
19   The guideline was developed in accordance with a scope, which detailed the
20   remit of the guideline originating from the Department of Health and specified
21   those aspects of RA care to be included and excluded.
22
23   Prior to the commencement of the guideline development, the scope was
24   subjected to stakeholder consultation in accordance with processes
25   established by NICE.1,7 The full scope is shown in Appendix B.

26   2.3       Audience
27   The guideline is intended for use by the following people or organisations:
28        • all healthcare professionals
29         •    people with RA and their carers
30         •    patient support groups
31         •    commissioning organisations
32         •    service providers.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                14 of 307
     DRAFT FOR CONSULTATION


1    2.4       Involvement of people with rheumatoid arthritis
2    The NCC-CC was keen to ensure the views and preferences of people with
3    RA and their carers informed all stages of the guideline. This was achieved
4    by:
5         • having 2 people with RA as patient representatives on the guideline
6            development group
7          •    consulting the Patient and Public Involvement Programme (PPIP)
8               housed within NICE during the pre-development (scoping) and final
9               validation stages of the guideline project.
10         •    The inclusion of patient groups as registered stakeholders for the
11              guideline

12   2.5       Guideline limitations
13   These include:
14       • NICE clinical guidelines usually do not cover issues of service
15           delivery, organisation or provision (unless specified in the remit from
16           the Department of Health).
17         •    NICE is primarily concerned with Health Services and so
18              recommendations are not provided for Social Services and the
19              voluntary sector. However, the guideline may address important
20              issues in how NHS clinicians interface with these sectors.
21         •    Generally, the guideline does not cover rare, complex, complicated or
22              unusual conditions.
23         •    It is not possible in the development of a clinical guideline to
24              complete extensive systematic literature review of all
25              pharmacological toxicity. NICE expect the guidelines to be read
26              alongside the Summaries of Product Characteristics.

27   2.6       Other work relevant to the guideline
28   Related NICE Technology Appraisals.
29        • Adalimumab, etanercept and infliximab for the treatment of
30           rheumatoid arthritis. NICE technology appraisal guidance 130 (2007).
31         •    Rituximab for the treatment of rheumatoid arthritis. NICE technology
32              appraisal guidance 126 (2007).
33         •    Abatacept for the treatment of rheumatoid arthritis. NICE technology
34              appraisal guidance 141 (2008).
35         •    Sequential use of adalimumab, etanercept and infliximab for the
36              treatment of rheumatoid arthritis. (anticipated publication date
37              September 2008).
38         •    Certolizumab pegol for rheumatoid arthritis (anticipated publication
39              date November 2009).
40


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                15 of 307
     DRAFT FOR CONSULTATION


1    Related NICE clinical guidelines:
2         • Osteoarthritis: the care and management of adults with osteoarthritis.
3            NICE clinical guideline 59. (2008).

4    2.7   Background
 5   The development of this evidence-based clinical guideline draws upon the
 6   methods described by the NICE Guideline Development Methods manual
 7   specifically developed by the NCC-CC for each chronic condition guideline
 8   (see http://www.rcplondon.ac.uk/college/ceeu/ncccc_index.htm). The
 9   developers’ role and remit is summarised in Table 1.4 below.
10




     Rheumatoid arthritis: full guideline DRAFT (August 2008)            16 of 307
     DRAFT FOR CONSULTATION


 1   Table 1.4 Role and remit of the developers
 2
     National                       The NCC-CC was set up in 2001 and is housed within the
 3
     Collaborating                  Royal College of Physicians (RCP). The NCC-CC
 4                                  undertakes commissions received from the National Institute
 5   Centre for Chronic
     It
     Conditions                     for Health and Clinical Excellence (NICE).
 6
 7   (NCC-CC)
                                    A multiprofessional partners’ board inclusive of patient
 8                                  groups and NHS management governs the NCC-CC.
 9
10
11   NCC-CC Technical               The technical team met approximately two weeks before each
     Team                           Guideline Development Group (GDG) meeting and comprised the
12                                  following members:
13                                                  GDG chair
14                                                  GDG clinical advisor
15                                                  Information scientist
16                                                  Research fellow
                                                    Health economist
17                                                  Project manager.
18
19
20
21
     Guideline                      The GDG met monthly (June 2007 to July 2008) and comprised a
22                                  multi disciplinary team of health professionals and people with
     Development
23                                  Rheumatoid Arthritis, who were supported by the technical team.
     Group
24
25                                  The GDG membership details including patient representation and
                                    professional groups are detailed in the GDG membership table at
26
                                    the front of this guideline.
27
28
29
30
31                                  The PE was involved in overseeing all phases of the guideline. It
     Guideline Project
32                                  also reviewed the quality of the guideline and compliance with the
     Executive (PE)
33                                  DH remit and NICE scope.
34
35                                  The PE comprised of:
                                                  NCC-CC Director
36                                                NCC-CC Assistant Director
37                                                NCC-CC Manager
38                                                NICE Commissioning Manager
39                                                NCC-CC Technical Team.
40
41
42
     Formal consensus               At the end of the guideline development process the GDG met to
43                                  review and agree the guideline recommendations.
44
45
46
47   Members of the GDG declared any interests in accordance with the NICE technical manual1 A
48   register is given in Appendix E




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                         17 of 307
     DRAFT FOR CONSULTATION



1    2.8         The process of guideline development
 2   The basic steps in the process of producing a guideline are:
 3
 4      1.       Developing clinical questions
 5      2.       Systematically searching for the evidence
 6      3.       Critically appraising the evidence
 7      4.       Incorporating health economics evidence
 8      5.       Distilling and synthesising the evidence and writing recommendations
 9      6.       Grading the evidence statements
10      7.       Agreeing the recommendations
11      8.       Structuring and writing the guideline
12      9.       Updating the guideline
13
14   1. Developing evidence based questions
15   The technical team drafted a series of clinical questions that covered the
16   guideline scope. The GDG and Project Executive refined and approved these
17   questions, which are shown in Appendix A.
18
19   2. Searching for the evidence
20   The information scientist developed a search strategy for each question. Key
21   words for the search were identified by the GDG. In addition, the health
22   economist searched for additional papers providing economics evidence or to
23   inform detailed health economics work (for example, modelling). Papers that
24   were published or accepted for publication in peer-reviewed journals were
25   considered as evidence by the GDG. Conference paper abstracts and non-
26   English language papers were excluded from the searches.
27
28   Each clinical question dictated the appropriate study design that was
29   prioritised in the search strategy but the strategy was not limited solely to
30   these study types. The research fellow or health economist identified relevant
31   titles and abstracts from the search results for each clinical question and full
32   papers were obtained. Exclusion lists were generated for each question
33   together with the rationale for the exclusion. The exclusion lists were
34   presented to the GDG. See Appendix A for literature search details.
35
36   3. Appraising the evidence
37   The research fellow or health economist, as appropriate, critically appraised
38   the full papers. In general, no formal contact was made with authors however
39   there were ad hoc occasions when this was required in order to clarify specific
40   details. Critical appraisal checklists were compiled for each full paper. One
41   research fellow undertook the critical appraisal and data extraction. The
42   evidence was considered carefully by the GDG for accuracy and
43   completeness.
44   All procedures are fully compliant with the:
45
46           •    NICE methodology as detailed in the ‘Guideline Development
47                Methods – Information for National Collaborating Centres and
48                Guideline Developers’ Manual1

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                18 of 307
     DRAFT FOR CONSULTATION


1          •   NCC-CC Quality assurance document and systematic review chart
2              available at:
3              http://www.rcplondon.ac.uk/college/ceeu/ncccc_index.htm
 4
 5   4. Health economics evidence
 6   Published economics evaluations were retrieved, assessed and reviewed for
 7   every guideline question. Full economics evaluations were included – that is
 8   those studies that compare the overall health outcomes of different
 9   interventions as well as their cost. Cost analyses and cost-consequences
10   analysis, which do not evaluate overall health gain, were not included.
11
12   Evaluations conducted in the context of non-OECD countries were also
13   excluded, since costs and care pathways are unlikely to be transferable to the
14   UK NHS.
15
16   Areas for health economics modelling were agreed by the GDG after the
17   formation of the clinical questions. The health economist reviewed the clinical
18   questions to consider the potential application of health economics modelling,
19   and these priorities were agreed with the GDG.
20
21   The health economist performed supplemental literature searches to obtain
22   additional data for modelling. Assumptions, data and structures of the models
23   were explained to and agreed by the GDG members during meetings, and
24   they commented on subsequent revisions.
25
26   5. Distilling and synthesising the evidence and developing recommendations
27   The evidence from each full paper was distilled into an evidence table and
28   synthesised into evidence statements before being presented to the GDG.
29   This evidence was then reviewed by the GDG and used as a basis upon
30   which to formulate recommendations. The criteria for grading evidence are
31   shown in Table 1.5.
32
33   Evidence tables are available on-line at (to be completed upon publication)
34
35   6. Grading the evidence statements
36
37   Table 1.5 Levels of evidence for intervention studies 1:
38
     Level of      Type of evidence
     evidence
     1++           High-quality meta-analyses, systematic reviews of RCTs, or
                   RCTs with a very low risk of bias
     1+            Well-conducted meta-analyses, systematic reviews of RCTs, or
                   RCTs with a low risk of bias
     1–            Meta-analyses, systematic reviews of RCTs, or RCTs with a
                   high risk of bias*
     2++           High-quality systematic reviews of case–control or cohort
                   studies


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              19 of 307
     DRAFT FOR CONSULTATION


                   High-quality case–control or cohort studies with a very low risk
                   of confounding, bias or chance and a high probability that the
                   relationship is causal
      +
     2             Well-conducted case–control or cohort studies with a low risk of
                   confounding, bias or chance and a moderate probability that the
                   relationship is causal
      –
     2             Case–control or cohort studies with a high risk of confounding,
                   bias, or chance and a significant risk that the relationship is not
                   causal*
     3             Non-analytic studies (for example, case reports, case series)
     4             Expert opinion, formal consensus
     *Studies with a level of evidence ‘–‘ should not be used as a basis for making
     a recommendation (see section 7.4)
1
2    7. Agreeing the recommendations
3
4    The GDG employed formal consensus techniques to:
5         •   ensure that the recommendations reflected the evidence-base
6         •   approve recommendations based on lesser evidence or
7             extrapolations from other situations
8         •   reach consensus recommendations where the evidence was
9             inadequate
10        •   debate areas of disagreement and finalise recommendations.
11   The GDG also reached agreement on the following:
12        •   recommendations as key priorities for implementation
13        •   five key research recommendations
14        •   algorithms.
15   In prioritising key recommendations for implementation, the GDG took into
16   account the following criteria:
17        •   high clinical impact
18        •   high impact on reducing variation in practice
19        •   more efficient use of NHS resources
20        •   allowing the patient to reach critical points in the care pathway more
21            quickly.
22        •   Audit criteria for this guideline will be produced for NICE following
23            publication in order to provide suggestions of areas for audit in line
24            with the key recommendations for implementation.
25
26   8. Structuring and writing the guideline
27   The guideline is divided into sections for ease of reading. For each section the
28   layout is similar and contains:

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                20 of 307
     DRAFT FOR CONSULTATION


1
2         •   Clinical introduction
3    Sets a succinct background and describes the current clinical context
4         • Methodological introduction
 5   Describes any issues or limitations that were apparent when reading the
 6   evidence base. Point estimates (PE) and confidence intervals (CI) are
 7   provided for all outcomes in the evidence tables available at (to be completed
 8   upon publication). In addition within the guideline PE and CI are cited in
 9   summary tables. In the absence of a summary table PE and CI are
10   provided in the narrative text when the outcome adds something to the text
11   and to make a particular point. These may be primary or secondary outcomes
12   that were of particular importance to the GDG when discussing the
13   recommendations. The rationale for not citing all statistical outcomes in the
14   text is to try to provide a 'user friendly' readable guideline balanced with
15   statistical evidence where this is thought to be of interest to the reader.
16          • Evidence statements
17   Provides a synthesis of the evidence-base and usually describes what the
18   evidence showed in relation to the outcomes of interest. Where the evidence
19   statements are considerable the GDG have attempted to summarise these
20   into a useful summary. The full evidence statements are not published as part
21   of the full guideline, however they will be made available in the appendices
22   which are published separately on-line at (to be completed upon publication)
23         • Health economics
24   Presents, where appropriate, an overview of the cost effectiveness evidence-
25   base, or any economics modelling
26        • From evidence to recommendations
27   This section sets out the Guideline Development Group (GDG) decision-
28   making rationale providing a clear and explicit audit trail from the evidence to
29   the evolution of the recommendations.
30        • Recommendations
31   Provides stand alone, action orientated recommendations.
32        • Evidence tables
33   The evidence tables are not published as part of the full guideline but are
34   available on-line at (to be completed upon publication). These describe
35   comprehensive details of the primary evidence that was considered during the
36   writing of each section.
37
38   9. Writing the guideline
39   The first draft version of the guideline was drawn up by the technical team in
40   accordance with the decisions of the GDG, incorporating contributions from
41   individual GDG members in their expert areas and edited for consistency of
42   style and terminology. The guideline was then submitted for a formal public
43   and stakeholder consultation prior to publication. The registered stakeholders
44   for this guideline are detailed on the NICE website www.nice.org.uk. Editorial
45   responsibility for the full guideline rests with the GDG.
46
47   The following versions of the guideline are available:

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               21 of 307
     DRAFT FOR CONSULTATION


1    Table 1.6
2
           Full version:                 Details the recommendations, the supporting
                                         evidence base and the expert considerations of
                                         the GDG. Published by the NCC-CC. Available
                                         at (to be completed upon publication)

           NICE version:                 Documents the recommendations without any
                                         supporting evidence.
                                         Available at (to be completed upon publication)
           "Quick reference guide":      An abridged version.
                                         Available online upon publication

           "Understanding NICE           A lay version of the guideline recommendations
           guidance":                    Available online upon publication
3
 4   10. Updating the guideline
 5   Literature searches were repeated for all of the evidence based questions at
 6   the end of the GDG development process allowing any relevant papers
 7   published up until 13th June 2008 to be considered. Future guideline updates
 8   will consider evidence published after this cut-off date.
 9
10   Following publication and in accordance with the technical manual, NICE will
11   ask a National Collaborating Centre to determine whether the evidence base
12   has progressed significantly to alter the guideline recommendations and
13   warrant an update.

14   2.9    Disclaimer
15   Healthcare providers need to use clinical judgement, knowledge and expertise
16   when deciding whether it is appropriate to apply guidelines. The
17   recommendations cited here are a guide and may not be appropriate for use
18   in all situations. The decision to adopt any of the recommendations cited here
19   must be made by the practitioner in light of individual patient circumstances,
20   the wishes of the patient, clinical expertise and resources.
21   The NCC-CC disclaims any responsibility for damages arising out of the use
22   or non-use of these guidelines and the literature used in support of these
23   guidelines.

24   2.10 Funding
25   The National Collaborating Centre for Chronic Conditions was commissioned
26   by the National Institute for Health and Clinical Excellence to undertake the
27   work on this guideline.
28
29
30
31
32


     Rheumatoid arthritis: full guideline DRAFT (August 2008)            22 of 307
     DRAFT FOR CONSULTATION


1    3      Key messages of the guideline

2    3.1    Priorities for implementation
3    A referral for a specialist opinion should be made in any person who is
4    suspected of having a persistent synovitis of undetermined cause and this
5    should be urgent if any of the following apply:
6       • The small joints of the hands or feet are affected
7       • More than one joint is involved
8       • There has already been a delay of 3 months or longer between onset
9           of symptoms and seeking medical advice.

10   In people with recent onset of active rheumatoid arthritis, treatment with a
11   combination of disease modifying drugs (including methotrexate and
12   glucocorticoids) should be offered as first line treatment as soon as possible,
13   ideally within three months of the onset of persistent symptoms.

14   In all patients with early active disease, key constituent parts of disease
15   activity and CRP should be measured monthly (using a composite score such
16   as DAS28) until an agreed level of disease control has been achieved by
17   appropriate treatment

18   In people with a recent onset of RA where combination therapy is not
19   appropriate * and DMARD monotherapy is started, greater emphasis should
20   be placed on rapid escalation to a clinically effective dose, rather than the
21   choice of DMARD.

22   In people with recent onset of rheumatoid arthritis receiving combination
23   DMARD therapy and in whom sustained and satisfactory levels of disease
24   control have been achieved, cautious attempts should be made to reduce
25   drug doses to levels which still maintain disease control.

26   People with suspected rheumatoid arthritis who are seronegative for
27   rheumatoid factor should have their anti-CCP antibodies tested.

28   In order to ensure the optimal use of the multidisciplinary team, all people with
29   RA should have access to a named member of the team (e.g. the specialist
30   nurse) who has been designated as the person responsible for coordinating
31   the various aspects of their care




     Rheumatoid arthritis: full guideline DRAFT (August 2008)               23 of 307
    DRAFT FOR CONSULTATION


1   3.2   Algorithms
2   Please see separate folder on NICE website for the 3 algorithms on early
3   inflammatory, recent onset and established Rheumatoid Arthritis.




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           24 of 307
    DRAFT FOR CONSULTATION


1

2                               THE GUIDELINE
3




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   25 of 307
     DRAFT FOR CONSULTATION



1    4      Identification and diagnosis

2    4.1.1 Clinical introduction
 3   The key point to the identification of early rheumatoid arthritis is the
 4   identification of synovitis. This is inflammation of the membrane that lines the
 5   inside of synovial joints (most of the joints in the body). Inflammation
 6   manifests itself in pain, swelling, heat and loss of function of the affected joint.
 7   The joints will also be stiff, as if trying to move them through resistance,
 8   especially in a morning, and sometimes in an evening also. Occasionally the
 9   joints can also be red, but this is more unusual with RA, and a single red hot
10   swollen joint should always be treated as septic until proven otherwise.
11   Widespread synovitis may also lead to systemic symptoms of inflammation,
12   with malaise, fever, sweats, fatigue and weight loss. All of these symptoms
13   and signs may be present to a lesser or greater extent in a person with recent
14   onset RA.
15
16   The initial trigger for rheumatoid arthritis is not known. However, whatever
17   initiates the inflammation leads to a big increase in blood flow to the joint with
18   resultant heat (and occasionally overlying redness). Subsequently the number
19   of blood vessels in the synovial-lining proliferate markedly, perpetuating the
20   heat of the joint. The inner lining of the joint is composed of cells called
21   synoviocytes, and these are normally only two to three cells deep. Half of
22   these cells produce the lubricating and nutrition-giving synovial fluid, which
23   forms a thin viscous layer in a normal joint. In inflammation, the synoviocytes
24   increase in size and number and production of synovial fluid increases
25   markedly. This combination of increase in synovial fluid and proliferation of
26   the synovial membrane causes the joint to become swollen, so that it has a
27   “boggy” feel on palpation. A number of pain receptors exist in the soft tissues
28   and bone around the joint, and their stimulation by the swelling, and other
29   nerve irritants leads to pain. The combination of pain and swelling results in
30   surrounding muscle weakness and loss of function.
31
32   The main priority for non-specialists is to recognise synovitis as soon as
33   possible. This means a low threshold for detecting heat, swelling, pain, loss of
34   function, morning stiffness, and systemic features of inflammation. In the
35   context of rheumatoid arthritis, if some of these signs and symptoms are
36   present in small joints, then the threshold for considering RA needs to be
37   lowered considerably. Some authorities have suggested if squeezing
38   metacarpophalangeal or metatarsophalangeal joints is tender, this is sufficient
39   to trigger concern.7,14
40
41   Colleagues in primary care face a number of challenges in recognising early
42   RA, especially when it can initially present in such a variety of ways. Many
43   patients present with musculoskeletal aches and pains, and even evidence of
44   synovitis, but only a small minority will progress to rheumatoid arthritis. With
45   the incidence of idiopathic inflammatory arthritis being around 2 per 10,000
46   population5,7, a GP with a list size of 2000 patients will see one new case of


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  26 of 307
     DRAFT FOR CONSULTATION


 1   RA approximately every two years. Other challenges for general practitioners
 2   include:
 3
 4        •   They are being asked to identify relatively rare persistent synovitis in
 5            patients that need prompt interventions, from all of the other very
 6            common causes of musculoskeletal aches and pains. The ARA
 7            classification criteria for RA do not help, and were designed to
 8            differentiate established RA from other established musculoskeletal
 9            conditions.2,7 These criteria do not perform well in people with recent
10            onset of disease that goes onto develop into RA.
11        •   Where patients develop a polyarthritis very rapidly, the clinical
12            diagnosis and abnormal results, as well as the frightening pain and
13            disability, will usually result in a prompt referral to secondary care.
14            These onsets are in the minority, with most patients having a much
15            more insidious onset, or a palindromic onset where the disease might
16            wax and wane in attacks before becoming established. The majority
17            of patients therefore present a diagnostic dilemma in the early stages
18            of disease.
19        •   Laboratory results and x-rays may all be normal in the early stages,
20            even in patients with definite RA, particularly if this is just affecting
21            small joints. Much of the time the key skills for identification are the
22            clinical skills of detecting synovitis, and this is not always straight-
23            forward, even for specialists.
24        •   Some patients with a recent onset of persistent synovitis may gain
25            useful benefit from NSAIDs or analgesics that lulls both patient and
26            general practitioner into a false sense of security with regard to the
27            impact the drugs are having on the disease.
28   A further challenge in managing patients with recent onsets of synovitis is the
29   delay in the person with the symptoms presenting to their GP in the first place.
30   Recent studies have suggested that this accounts for a greater part of the
31   delay in people with RA seeing a specialist for the first time, when compared
32   with delays once the GP has seen the person.15 People with synovitis may
33   have accumulated damage in their joints before presenting to any member of
34   the medical profession, and it may be that a public health awareness
35   campaign is required to address this. However, GPs may still delay referral,
36   because symptoms may be vague, or signs of synovitis difficult to identify, or
37   positive responses to analgesics and NSAIDs lead to a false sense of
38   security. Elsewhere in these guidelines, reference will be made to the need to
39   initiate DMARDs in the 3-4 month window of opportunity that makes a huge
40   difference to long-term outcomes.(See section 7.3.13). This highlights the
41   need for:
42          • the general public to be aware of the symptoms and signs of
43             synovitis so that they can attend their GP early, and
44        •   ongoing attempts to reinforce to GPs the need for prompt referral.
45   These challenges lead to the following questions:




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  27 of 307
     DRAFT FOR CONSULTATION


1         •   Are there any clinical patterns of symptoms or signs of inflammatory
2             arthritis where prognostic concerns need to register and the patient
3             be referred to secondary services promptly?
4         •   If such signs and symptoms are evident, is there any evidence to
5             suggest a time-frame for such a referral?

6    4.1.2 Clinical methodological introduction
 7   We looked for studies that investigated the clinical features a non-specialist
 8   should recognise in order to refer to specialist services and how quickly the
 9   referral should be made with respect to impact on symptoms, joint damage,
10   function and quality of life in patients with a recent onset of undifferentiated
11   inflammatory arthritis. All types of study with a UK-relevant population were
12   selected.
13
14   Clinical features
15
16   1 cohort study16 and 12 case-series’ (prospective)17-27 were found which
17   fulfilled the criteria for which clinical features a non-specialist should recognise
18   in order to refer. All trials were methodologically sound.
19
20   The cohort study16 looked at the clinical features of N=474 patients with
21   arthritic symptoms who attended an early arthritis clinic vs a routine clinic and
22   these patients were followed for 1 year.
23
24   The 12 case-series’ looked at the clinical features of patients with early
25   inflammatory arthritis and in some studies, patients were followed up to look
26   at the features of those that went on to develop RA. Studies differed with
27   respect to:
28         • Sample size (range: N=41 to N=903)
29        •   Trial length (range: time not mentioned, 1 year to 8 years)
30
31   Timing of referral
32
33   Three retrospective case-series’15,28,29 were found which looked at the
34   optimum timing of referral in patients with RA. The first case-series28looked at
35   the effects of early vs late referral in N=200 patients. The second case-
36   series29 looked at the effects of delay in referral and starting DMARD therapy
37   in N=198 patients. The third case-series15 looked at delay in referral times and
38   reasons for delay in N=169 patients.

39   4.1.3 Health economic methodological introduction
40   No health economic papers were identified

41   4.1.4 Clinical evidence statements
42   Clinical features



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 28 of 307
    DRAFT FOR CONSULTATION



1
2   Symptoms
3
    Study          Patient group       Use                     Clinical features
    1 case-        Early IA who        Distinguish RA from     Higher number of painful and swollen joints
    series18       developed RA        other disorders
    Level 3
    1 case-        Early IA            Referral for suspected any joint swelling (significant association)
    series19                           RA
    Level 3                            predictive for referral Morning stiffness (not predictive)
                                       and RA diagnosis
    1 case-        Early IA            Features present        Clinical synovitis (14%)
    series20                                                   RA diagnostic criteria (56%)
    Level 3
    1 case-        Very early          Common features         Pain in the hand joint and long duration of morning stiffness
    series23       persistent IA       present
    Level 3
    1 case-        UPA who             Features present        Higher baseline joint counts
    series24       developed RA        Predictors of           pain/tenderness in small joints (SS Year 1); swelling count in small
    Level 3                            developing RA           joints ( SS year 1 and 5, NS year 3).
    1 case-        early IA            Distinguish IA from     Distinguish: currently had or had a history of significant stiffness in the
    series25                           non-IA conditions       morning or after rest; objective joint swelling
    Level 3                            (features GPs and
                                       RNs deemed most
                                       important)              Not distinguish: joint pain, joint swelling, joint tenderness, redness,
                                                               heat
    1 case-        suspected           Predictors of           morning stiffness and tender and swollen joint counts (SS)
    series27       arthritis           developing RA
    Level 3



    Rheumatoid arthritis: full guideline DRAFT (August 2008)            29 of 307
    DRAFT FOR CONSULTATION



    1 case-        UA who              Distinguish from those Higher % of patients had compatible with RA features: MRI synovitis
    series26       developed RA        not develop RA         (100% vs 40%); MRI synovitis OR MRI erosion (100% vs 50%); MRI
    Level 3                                                   synovitis AND MRI erosion (64% vs 13%); MRI synovitis AND MRI
                                                              erosion AND scintigraphy (45% vs 0%)


1
2   Pattern and site of arthritis
3
    Study          Patient group       Use                       Clinical features
    1 cohort       early arthritis     Common features           Atypical presentation (asymmetrical arthritis), monoarthritis or
    study16                            present                   oligoarthritis
    Level 2+
    1 case-        Early IA who        Distinguish RA from       Useful (higher number): Involvement of hands (pain or swelling of
    series18       developed RA        other disorders           wrists or finger joints)
    Level 3                                                      Not useful: Pain (VAS) and Pain or swelling of MTP joints
    1 case-        Early IA            Features present          swollen joints predominantly in the hands or knee
    series19
    Level 3                            Predictive for referral   Not predictive: restriction of swollen joints to hands or fingers
                                       and RA diagnosis
    1 case-        Early IA who        Features present at       17% - knee involvement
    series20       developed RA        first visit /             100% - symmetrical synovitis of the small joints of the hands and feet
    Level 3                            presentation
    1 case-        Very early          Common features           symmetrical arthritis
    series23       persistent IA       present
    Level 3




    Rheumatoid arthritis: full guideline DRAFT (August 2008)              30 of 307
    DRAFT FOR CONSULTATION



    1 case-        Early IA            Distinguish IA from      Not distinguish: signs of MCP / MTP joint involvement
    series25                           non-IA conditions
    Level 3                            (features GPs and
                                       RNs deemed most
                                       important)
    1 case-        Suspected           Predictors of            joint symptoms in the small/large joints, asymmetric localisation of the
    series27       arthritis           developing RA            affected joints and localisation of affected joints in both upper and lower
    Level 3                                                     extremities

1
2   Function
3
    Study       Patient group     Use                Clinical features
    1 case-     Early IA          Predictive for     Predictive: limitations when clenching the hands completely to a fist as well as
    series19                      referral and RA    limitations of finger flexion
    Level 3                       diagnosis          Not predictive: general questions on every day function
    1 case-     Very early        Common features    Positive squeeze test of the MCP joints
    series23    persistent IA     present
    Level 3
    1 case-     Early IA          Distinguish IA     Not distinguish: loss of function and reduced range of movement
    series25                      from non-IA
    Level 3                       conditions
                                  (features GPs
                                  and RNs deemed
                                  most important)
4
5   Joint damage
6
    Study         Patient group      Use                     Clinical features




    Rheumatoid arthritis: full guideline DRAFT (August 2008)             31 of 307
    DRAFT FOR CONSULTATION



    1 cohort      early arthritis   Common features         Erosions at presentation; acute arthritis
    study16                         present
    Level 2+
    1 case-       Early IA who      Distinguish RA from     13% Erosions; 21% signs of nonerosive joint involvement (mainly soft
    series18      developed RA      other disorders         tissue swelling); Larsen score mean 3.5; acute onset of symptoms (most
    Level 3                                                 patients)


    1 case-       UA who            Distinguish from        Higher % of patients had compatible with RA features: Radiographic
    series26      developed RA      those not develop       Larsen score grade 1 (36% vs 3%); MRI erosions (64% vs 23%);
    Level 3                         RA                      Scintigraphy (64% vs 26%)


                                                            Not distinguish: RF+ (similar in both groups; 36% vs 33%)
1
2   Biochemical markers
3
    Study         Patient group       Use                      Clinical features
    1 case-       Early IA who        Distinguish RA from      47% were RF+
    series18      developed RA        other disorders
    Level 3                                                    Not useful: ESR and CRP
    1 case-       Early IA            Predictive for RA        1 or more lab parameters: ESR, CRP or RF was (SS association).
    series19                          diagnosis
    Level 3
    1 case-       Very early          Predictors of            Not predictors: inflammatory markers
    series23      persistent IA       persistent IA
    Level 3
    1 case-       UPA who             Predictors of            Predictor: antinuclear antibodies (year 1) and 5; ESR (year 5)
    series24      developed RA        developing RA
    Level 3                                                    Not predictor: RF; ESR (year 3)



    Rheumatoid arthritis: full guideline DRAFT (August 2008)             32 of 307
    DRAFT FOR CONSULTATION



    1 case-       Suspected            Predictors of            CRP level, RF+, anti-CCP+.
    series27      arthritis            developing RA
    Level 3
    1 case-       UA who               Distinguish from         Not distinguish: RF+ (similar in both groups; 36% vs 33%)
    series26      developed RA         those not develop
    Level 3                            RA
1
2   ARA and NY criteria
3
    Study         Patient group        Use                      Clinical features
    1 case-       Arthritic            Predictors of            Best predictors: symmetrical swelling or tenderness in PIP or MCP or
    series17      symptoms who         developing RA            MTP joints + Swelling in 1 joint and swelling or tenderness in another 4
    Level 3       developed RA                                  joints
                                                                Not predictors: NY and ARA criteria (except 8th ARA criterion)
    1 case-       Early IA who         Distinguish RA from      Not distinguish: ACR criteria
    series18      developed RA         other disorders
    Level 3
    1 case-       Early IA who         Features present at      47% had RA or fulfilled ACR criteria at follow-up
    series20      developed RA         first visit /
    Level 3                            presentation
    1 case-       Early IA             Identify patients with   Low sensitivity and specificity: ARA criteria (list and tree format)
    series21                           RA diagnosis
    Level 3                            (physician diagnosis)
    1 case-       Recent               Detect RA                Best ARA and NY classification criteria: RF, symmetrical polyarthritis
    series22      inflammatory joint                            (especially the NY clinical criterion), morning stiffness and X-ray
    Level 3       disease                                       changes




    Rheumatoid arthritis: full guideline DRAFT (August 2008)              33 of 307
     DRAFT FOR CONSULTATION

 1
 2
 3   Timing of referral
 4
 5   1. Function
 6   1 case-series28 found that patients referred early were significantly better than
 7   those referred late for: NHP physical function scores and mean HAQ scores.
 8
 9   1 case- series29 found that late referral was the most powerful predictor of
10   functional disability (MAP physical function score), however it was not a
11   significant predictor of HAQ score. The same case-series29found that patients
12   referred early were significantly better than those referred late for: delay to
13   starting DMARD therapy. However the 2 groups were similar for: rate of RF+
14   and for erosive changes (until the delay to referral was >1 year).
15
16   1 case-series15 found that patient-dependent factors (such as delay from the
17   onset of symptoms to the assessment in primary care), leading to a delay in
18   consulting primary care physicians, are the principal reasons for delay in
19   patients with RA being seen by rheumatologists.

20   4.1.5 Summary of evidence statements
21        •   The ARA criteria identify patients who are likely to have persistent
22            synovitis and a poor prognosis, but do not perform well as diagnostic
23            criteria in recent onset RA 17,20 18,21,22. The key clinical features to
24            facilitate identification of patients who are likely to have persistent
25            synovitis and a poor prognosis include:
26                 •   number of joints affected (the more joints the worse the
27                     prognosis) 18,22,
28                 • presence of both swelling and tenderness in affected joints
29                     (particularly small joints) 18,19
30                 • a positive MCP squeeze test23,
31                 • involvement of PIPs and MCPs23, and symmetry of joints
32                     affected.
33        •   Inability to make a fist or flex the fingers was associated with an
34            ability to identify RA from other diagnoses in one study 19. Ever
35            having prolonged morning stiffness is more helpful than currently
36            having morning stiffness for early RA23.
37        •   Rheumatoid factor detects less than half of eventual RA patients at
38            presentation 18
39        •   Acute phase markers are no different in patients with an inflammatory
40            arthritis that evolves in RA than those that evolve into non-RA 23
41        •   Delays in referral are associated with worse function at presentation
              28
42             , and if delayed by 1 year an increase in erosive changes on x-ray
              29
43

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 34 of 307
     DRAFT FOR CONSULTATION


1         •   The most significant factor in delaying start of DMARD was delay in
2             referral to a rheumatologist 29
3         •   There is evidence that the greatest delay in patients presenting to
4             specialist care is in the patients attending their GP with symptoms in
5             the first place, rather than the delay in patients being referred by the
6             GP to specialist care15.

7    4.1.6 From evidence to recommendations
 8   The ARA criteria were considered to be better prognostic guides than
 9   diagnostic guides in early inflammatory arthritis. In other words, if a person
10   presents with small joint disease, then this carries a poor prognosis, and
11   should lead to urgent referral to specialist care. The same considerations
12   apply to the number of joints affected, where evidence suggests that the
13   greater the number, the worse the prognosis, and the greater need for early
14   identification and referral. The most important feature of early RA is the
15   clinical detection of synovitis. This is a clinical skill, and in early RA all blood
16   tests may be normal despite significant disabling disease. It was therefore
17   considered important that this message was emphasised, and that normal
18   investigations should not put a GP off referring patients urgently to secondary
19   care. It was acknowledged that there is evidence that much of the delay in
20   referral to specialist care is often beyond the control of the GP, and relates to
21   the person themselves not presenting promptly at the start of their symptoms.
22   Given the evidence that this delay increases the risk of damage to joints, and
23   delays introduction of DMARDs, people who present to their GP with well
24   established disease should be referred urgently to specialist care to try to
25   minimise any further damage, especially if symptoms have already been
26   present for more than three months.

27   4.1.7 Recommendations
28   R 1 A referral for a specialist opinion should be made in any person who is
29   suspected of having a persistent synovitis of undetermined cause and this
30   should be urgent if any of the following apply:
31        • the small joints of the hands or feet are affected
32        •   more than one joint is involved
33        •   there has already been a delay of 3 months or longer between onset
34            of symptoms and seeking medical advice.
35   R 2 In any person who is suspected of having a persistent synovitis of
36   undetermined cause, blood tests demonstrating a normal acute phase
37   response or negative rheumatoid factor should not preclude the need for
38   urgent referral.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  35 of 307
     DRAFT FOR CONSULTATION


1    4.2    Investigations

2    4.2.1 Clinical introduction
 3   The identification of persistent synovitis is largely a clinical skill. However,
 4   there are investigations that can help to demonstrate that there are
 5   abnormalities that require intervention. Some of these may be non-specific,
 6   such as evidence of inflammation taking place (e.g. elevated C-reactive
 7   protein, anaemia of chronic disease), whereas others may be more helpful in
 8   pointing towards a diagnosis (e.g. rheumatoid factor present in high titres,
 9   erosive change on x-rays). Some investigations help to rule out other causes
10   of polyarthritis or polyarthralgias, such as thyroid function tests. Other
11   investigations are useful baselines that help in management, such as renal
12   and liver function tests prior to commencing NSAIDs or DMARDs. In this
13   section the focus is on those tests that help with early recognition of the
14   disease, but it needs to be borne in mind that these do not replace the need
15   for a careful history and examination, and even if all tests are normal, this
16   should not prevent appropriate interventions from taking place in a patient
17   with persistent synovitis. Antibodies against cyclic citrullinated peptides (Anti-
18   CCP) have emerged in recent years as being as sensitive, but more specific
19   than rheumatoid factors in the diagnosis of RA. These are covered in section
20   4.3 presenting signs and symptoms. There is evidence to suggest that
21   ultrasound and magnetic resonance imaging scans are superior to clinical
22   examination in the detection of synovitis, and that they are more sensitive to
23   the presence of erosions and other early inflammatory and damage signs than
24   conventional x-rays.30,31 However, the long-term significance of their findings,
25   and the limited availability to many clinicians, restricts their utility, leaving the
26   detection of synovitis on clinician examination as the gold standard.

27   4.2.2 Clinical methodological introduction
28   We looked for studies that assessed the ability of investigative procedures to
29   identify patients with undifferentiated inflammatory arthritis who would go on
30   to develop RA. Due to the large volume of evidence, trials were selected
31   which were of a UK-relevant population, patients were pre-RA / had
32   undifferentiated arthritis (UA), if the population was mixed arthritis there had to
33   be >75% RA or RA subgroup analysis, and had a sample size of N>50
34   (except for MRI or Ultrasound studies).
35
36   Two MA32,33, 3 case-control studies34-37 and 12 case-series’ 27,38-48 were found
37   that fulfilled the criteria. One of the case-control studies was published as two
38   separate papers35,36 reporting different outcomes and so the trial has only
39   been counted once. However results from both papers are reported and
40   referenced here. No studies were found on ultrasound. All trials were
41   methodologically sound.
42
43   The first MA32 focused on all trials looking at anti-CCP tests for the diagnosis
44   of RA and included N=68 trials with data. Of these, N=14 trials looked at
45   investigative procedures predicting the development of RA (N=11 used UA
46   patients and N=3 used RA patients who had given blood before developing

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  36 of 307
     DRAFT FOR CONSULTATION


1    RA). The MA itself was fairly well conducted, however no tests for
2    heterogeneity were performed. Studies included in the analysis all used the
3    same method for detecting anti-CCP antibodies (ELISA), however they
4    differed with respect to:
5
6         •   Type of investigative test used (N=5 trials used anti-CCP1, N=10
7             trials used anti-CCP2).
8         •   Cut-off for anti-CCP+ (anti-CCP1 range 21.4 IU to 1000 IU, anti-
9             CCP2 3.8IU to 50 IU).
10        •   Study size (UA patients for anti-CCP1: N=1327; UA patients for anti-
11            CCP2: N=2017; RA patients given blood before RA development for
12            anti-CCP1: N=79 and for anti-CCP2: N=142)
13        •   Study duration – length of follow-up (UA patients: range 5-36 months;
14            RA patients given blood before RA development: range <1.5 years to
15            9 years)
16   The second MA33 looked at the diagnostic accuracy of anti-CCP tests and RF
17   tests in patients with a recent onset of RA (< 1 year duration) and included
18   N=86 trials. Of these, N=37 studies looked at anti-CCP tests and N=50 at RF
19   tests. Trials differed in terms of:
20         • Study size (range not mentioned)
21        •   Study design (Prospective in N=18/37 anti-CCP; N=25/50 RF)
22        •   Study quality – max score of 5 (N=1 very good quality; N=22, 30%
23            reasonable quality; N=9, 10% poorer quality)
24        •   Study duration – length of follow-up (range not mentioned)
25        •   Comparison group (mainly patients with UA; healthy patients; other
26            diseases; other rheumatic diseases)
27        •   Intervention - type of anti-CCP test (anti-CCP1 N=8; anti-CCP2,
28            N=29)
29        •   Intervention – type of RF test (IgM, IgA, IgG)
30
31   The 3 case-control studies34-37 all looked at investigative tests which could be
32   used to predict the development of RA in pre-RA / UA patients. The first two
33   case-control studies34-36 looked at investigative tests (RF and AFA; CRP, anti-
34   CCP, RF and collagen types respectively) in people who were at risk of
35   developing RA (both N=19, 072). The trials compared cases (those who went
36   on to develop RA) with matched controls (who did not develop RA) and had
37   follow-up times of 22 years or 12-16 years respectively. The third case-control
38   study37 looked at investigative tests (RF and AFA) in N=330 patients with UA.
39   The trial compared cases (those who went on to develop RA) with controls
40   (patients who already had RA) and had a 1 year follow-up time.
41
42   The 12 case-series’ 27,38-48 which investigative tests and procedures could be
43   used predict the development of RA in pre-RA / UA patients. Studies differed
44   with respect to:
45         • Sample size (range: N=30 to N=1003)

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              37 of 307
     DRAFT FOR CONSULTATION


1         •     Trial length (range: 1 year to 6.9 years mean)
2         •     Investigative procedure / test used (RF, anti-CCP, ACPA, CRP, ANA,
3               APF, ESR, MRI, symptoms, radiographs, histopathology, ACR
4               criteria and others)

5    4.2.3 Health economic methodological introduction
6    No health economic papers were identified.

7    4.2.4 Evidence Statements
8    Rheumatoid Factor (RF)
9
     Study          Patient group    Use                   Results – baseline predictors
     1 case-                         RA development vs     RF latex test (high sensitivity,
     series38                        no development        low specificity)
     Level 3
     1 case-                         Risk of RA            NS: RF status
     control                         development
     study35,36                      (cases vs controls)
     Level 2+
     1 case-                         RA development vs     RF+ (more patients; 38% vs
     series39                        no development        11%)
     Level 3
     1 case-        UA patients      Predictor of RA       Positivity of 2 of the 3 following
     series43                        development and       tests: RF; antiperinuclear
     Level 3                         distinguishing from   factors and the HLA DR4
                                     other diagnoses       antigen
     1 case-                         RA development vs     NS: RF+
     series44                        no development
     Level 3
     1 case-                         RA development        RF+ (SS more cases)
     control                         (cases vs controls)
     study37
     Level 2+
     1 case-        patients with    RA development vs     GAL0 (SS higher levels)
     series47       synovitis        no development
     Level 3                         Predictor of RA       ARA clinical criteria (68%
                                     development           patients), RF+ (83%), GAL0
                                                           levels 78%
     1 case-                         Predictor of RA       RF+ (Univariate and
     series27                        development           multivariate)
10
11
12   Anti-CCP
13
14   1 MA/SR32 reported 3 studies which looked at patients with RA who had
15   donated blood samples before development of RA.
16



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                38 of 307
     DRAFT FOR CONSULTATION


 1   1 study found that anti-CCP2 predicted RA development with low sensitivity
 2   (4%, 25% and 52% at 9 years, <1.5 years and <1.5 years before symptoms)
 3   and had a high specificity (98%). OR 28 (95% CI 8 to 95). Level 1+
 4
 5   1 study did further analysis of the same patients and found that anti-CCP2
 6   had highest predictive value compared to RF; OR 15.9 for anti-CCP2 and 6.8
 7   for RF. Level 1+
 8
 9   1 study found that 5 years before symptom onset, anti-CCP1 had a low
10   sensitivity and high specificity for predicting RA (29% and 99.5% respectively;
11   OR 64.5 (95% CI 8.5 to 489). Level 1+
12
     Study     Patient group       Use                   Results – baseline predictors
     1 MA/SR32 UA patients         Predictor of RA       anti-CCP+ (anti-CCP1:OR 20,
     Level 1+                      development           95% CI 14 to 31; anti-CCP2:
                                                         OR 25, 95% CI 18 to 35)
     1 MA33                        RA diagnosis          anti-CCP better than RF
     Level 1+                                            (similar sensitivity, higher
                                                         specificity); IgM RF+ plus anti-
                                                         CCP+ (even better)
     1 case-                       RA development        Abs to CCPs (higher levels)
     control                       (cases vs controls)
     study35,36
     Level 2+
     1 case-      UA patients      RA development        NS: anti-CCP
     control                       (cases vs controls)
     study37
     Level 2+
     1 case-      UA patients      RA development vs     IgA, IgM, IgG2 and IgG3 anti-
     series45                      no development        CCP (Higher frequencies and
     Level 3                                             levels); number of isotypes of
                                                         anti-CCP response (higher and
                                                         higher levels)
                                   Risk of RA            IgA anti-CCP+ (RR 1.3, 95% CI
                                   development within    1.0-1.7), IgM anti-CCP (RR 1.4,
                                   1 year of follow-up   95% CI 1.1 to 1.8) or IgG anti-
                                                         CCP (RR 1.4, 95% CI 1.1 to
                                                         1.8).
     1 case-                       Predictors of RA      anti-CCP+ (univariate and
     series27                      development           multivariate)
     Level 3
13
14   AFA
15
     Study        Patient group    Use                   Results – baseline predictors
     1 case-                       Risk of developing    AFA (SS for RF+ RA –
     control                       RA (RF+ or RF-)       increased OR; NS for RF- RA)
     study34                                             NS: interaction RF and AFA



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              39 of 307
    DRAFT FOR CONSULTATION


    Level 2+                     <5 years and 5-10     elevated AFA (higher risk of
                                 years to disease      RF+ RA); weak association >10
                                 onset                 years
                 RA patients                           Baseline RF and AFA
                 and control                           associated to same extent
                 patients
    1 case-                      RA development vs     AFA (moderate sensitivity and
    series38                     no development        specificity)
    Level 3
1
2   APF
3
    Study        Patient group   Use                   Results – baseline predictors
    1 case-                      RA development vs     APF (moderate sensitivity and
    series38                     no development        specificity)
    Level 3
    1 case-      UA patients     Predictor of RA       APF (fairly high sensitivity and
    series42                     development vs no     specificity; 77% and 75%)
    Level 3                      development
4
5   CRP
6
    Study        Patient group   Use                   Results – baseline predictors
    1 case-                      Risk of RA            NS: CRP
    control                      development
    study35,36                   (cases vs controls)
    Level 2+
    1 case-      polyarthritis   RA development vs     NS: CRP
    series44                     no development
    Level 3
    1 case-                      Risk of RA            NS: CRP
    control                      development
    study37                      (cases vs controls)
    Level 2+
    1 case-                      Predictor of RA       SS: CRP (Univariate and
    series27                     development           multivariate)
    Level 3
7
8   Radiographs and MRI
9
    Study        Patient group   Use                   Results – baseline predictors
    1 case-      UA patients     Predictor of RA       Erosions typical of RA; hand
    series40                     development           radiographs (low sensitivity and
    Level 3                                            fairly high specificity - 23% and
                                                       88%; NPV 66% and PPV 50%)
                                 Predictor of RA       hand radiographs (low
                                 diagnosis 2 years     sensitivity, fairly high specificity
                                 later                 - 30% and 85%; NPV 60% and
                                                       PPV 58%)


    Rheumatoid arthritis: full guideline DRAFT (August 2008)              40 of 307
    DRAFT FOR CONSULTATION


    1 case-     polyarthritis    Predictor of RA       MRI - OMERACT score for
    series44                     development           erosions in the MCP joints and
    Level 3                                            the second and third MCP
                                                       joints (fairly high specificity and
                                                       sensitivity - 70% and 64%)
                                 Distinguish RA        NS: Radiographs (carpus
                                 development and       erosions and MCP erosions);
                                 other diseases        MRI - OMERACT score
                                                       (synovitis and tenosynovitis;
                                                       carpus erosions)
    1 case-     UA patients      RA development        NS: erosive disease (hands
    control                      (cases vs controls)   and feet) and SHS score
    study37
    Level 2+
1
2   Other
3
    Study       Patient group    Use                   Results – baseline predictors
    1 case-                      RA development vs     AKA (low sensitivity, high
    series38                     no development        specificity)
    Level 3
    1 case-                      RA development vs     Most patients: histological
    series39                     no development        changes considered to be RA+,
    Level 3                                            distinctive IgM staining (RA+,
                                                       Immunofluorescence); high
                                                       white cell counts
    1 case-     Newly referred   RA development vs     Most patients: at least 1
    series41    patients         no development        swollen joint; a-pepA Abs (best
    Level 3                                            sensitivity, h specificity); RF
                                                       test + ACPA test (increased
                                                       PPV); one serologic marker +
                                                       swollen joints (increased PPV)
    1 case-     Patients with    RA development vs     SS: GAL0 (higher levels)
    series47    synovitis        no development
    Level 3                      Predictor of RA    ARA clinical criteria (68%
                                 development        patients); RF+ (83%) and GAL0
                                                    levels (78%); RF+/GAL0 (high
                                                    predictive ability - 91% of
                                                    patients; 90% sensitivity, 95%
                                                    specificity and 94% PPV)
    1 case-     UA patients      Predictor of RA    No extra value: ACPA testing +
    series48                     development 1 year HLA shared epitope
    Level 3                      later
                                                    RF testing + ACPA testing
                                                    better than RF alone
                                                    (especially for patients with at
                                                    least 1 swollen joint)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)              41 of 307
     DRAFT FOR CONSULTATION


1    4.2.5 Summary of evidence statements
2         •   RF in most studies is a useful predictor of RA development 38,39,43 37
              47 27
3
4         •   Anti-CCP positivity is a useful predictor of RA development27,32,35,36,45,
5             and in comparison to RF appears to have a higher specificity, but
6             similar sensitivity32,33
7         •   Baseline CRP is a poor predictor of who will go onto develop RA35,36
              37,44
8
 9        •   Baseline erosions on hand x-rays show high specificity but low
10            sensitivity for the development of RA40
11        •   Erosions on MCP MRI scans have a fairly high sensitivity and
12            specificity for the development of RA 44

13   4.2.6 From evidence to recommendations
14   Although anti-CCP antibodies are more specific than rheumatoid factor, this
15   difference is not great, and sensitivities seem very similar. Recommendations
16   on anti-CCP also need to be informed by health economic analysis to
17   determine whether the extra cost and increased specificity render this test
18   cost-effective, either for all early inflammatory arthritis, or for sub-groups (e.g.
19   RA suspected, but rheumatoid factor negative). Rheumatoid factor remains a
20   relatively cheap and useful test in undifferentiated synovitis that is helpful both
21   diagnostically and prognostically.
22
23   After much deliberation, it was decided that x-rays of the hands and feet in
24   early synovitis are worthwhile, because although this is a blunt instrument in
25   detecting joint inflammation, there are occasions when erosive damage will be
26   detected when all other tests are normal, and it also acts as a readily
27   accessible base-line for future determinations of disease progression. As
28   ultrasound and small joint MRI become more widely available, the long-term
29   significance of some of the early inflammatory and erosive changes that have
30   been described using these imaging modalities should become apparent.

31   4.2.7 Recommendations
32   R 3 A blood test for rheumatoid factor should be carried out in all people
33   found to have synovitis on clinical examination, in whom rheumatoid arthritis
34   is suspected.
35
36   R 4 Radiographs of the hands and feet should be performed early in the
37   course of the disease in all people with persistent synovitis affecting these
38   joints.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  42 of 307
     DRAFT FOR CONSULTATION


1    4.3       Presenting symptoms and signs

2    4.3.1 Clinical introduction
 3   The term RA covers a very broad spectrum of disease. For early RA some
 4   patients will have rapid onset of disease, and quick evolution to a polyarthritis.
 5   Others will have an insidious onset and perhaps months of a mono- or
 6   oligoarthritis before gradually evolving into a symmetrical peripheral pattern.
 7   Patients with early and with established disease may also have extra-articular
 8   features that impact on the severity and disability of the disease, such as
 9   interstitial lung disease and vasculitis. Others may have no obvious extra-
10   articular disease, or relatively mild but nevertheless irritating symptoms such
11   as dry eyes. It would be useful to be able to identify patients who are likely to
12   have a poor prognosis early in the course of the disease. These patients
13   could then be monitored more closely so that a lower threshold could operate
14   for intensive intervention to modify the course of their aggressive disease.
15   Conversely, in patients lacking poor prognostic markers, or possessing good
16   prognostic markers, a less intensive follow-up and treatment strategy might be
17   pursued. Are there any markers that might help to make management more
18   targeted, depending on disease prognosis?

19   4.3.2 Clinical methodological introduction
20   We looked for studies that investigated which clinical features of RA patients
21   (recent onset and established disease) can be used to identify those with a
22   good or poor prognosis. Due to the large volume of evidence on prognostic
23   features, trials were selected which were of a UK-relevant population, if the
24   population was mixed arthritis there had to be >75% RA or RA subgroup
25   analysis, had a sample size N>200. For the second part of the question,
26   looking at different treatments for patients with a poor prognosis, no limits
27   were set for selection criteria except for UK-relevant population.
28
29   Thirty-one case-series’ were found that fulfilled the inclusion criteria. One of
30   these was published as two separate papers49,50 reporting different follow-up
31   times and so the trial has only been counted once. However, results from both
32   papers are reported and referenced here. All trials were methodologically
33   sound and assessed the clinical features of RA patients who had either a
34   good or poor prognostic outcome (patients were either followed prospectively
35   or data was gathered retrospectively).
36
37   Recent onset RA
38   23 case-series’51-6346,64-73 were found which fulfilled the criteria. They differed
39   with respect to:
40         • Sample size (range: N=211 to N=1387)
41         •    Trial length (range: 1 year to 43 years)
42   Established RA
43   10 case-series’74-83 were found which fulfilled the criteria. They differed with
44   respect to:
45        • Sample size (range: N=263 to N=2448)

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  43 of 307
     DRAFT FOR CONSULTATION


1           •    Trial length (range: 6 months to 50 years)
 2
 3   Treatment of poor prognosis patients
 4   3 RCTs84-86 and 1 cohort study49,50 were found which fulfilled the criteria for
 5   which was the best treatment for RA patients with a poor prognosis. All trials
 6   were performed on patients with UA or a recent onset of RA. The cohort
 7   study49,50 was published as 2 separated papers with different follow-up times
 8   and so the trial has only been counted once. However, results from both
 9   papers are reported and referenced here. All trials were methodologically
10   sound.
11
12   The 3 RCTs84-86 were parallel group studies. The first two RCTs84,85 looked at
13   DMARD treatment (single vs combination) in patients with a recent onset of
14   RA who had a poor prognosis (N=82 and N=20 respectively). The first RCT84
15   compared 2 different treatment arms: SSZ 500 mg/day vs CSA 1.5 mg/kg/day
16   + MTX 7.5 mg/week + CS methylprednisolone in a 48 week treatment phase.
17   The second RCT85 compared 2 different treatment arms: MTX 7.5 mg/week +
18   placebo vs MTX 7.5 mg/week + IFX 3 mg/kg/day in a 1 year treatment phase
19   with follow-up at 1 year post-treatment. The third RCT86 looked at DMARD
20   treatment in N=110 patients with UA of which N=51 went on to develop RA.
21   The trial compared 2 different treatment arms: MTX vs placebo in a 30 month
22   treatment phase and performed a subgroup analysis of the outcome of
23   patients in each group who had a poor prognosis (anti-CCP+ or RF+).
24
25   The cohort study49,50 looked at which was the best treatment in N=206
26   patients with a recent onset of RA who had a poor prognosis. The trial
27   compared 2 different treatment arms: early treatment (DMARDs + NSAIDs) vs
28   delayed treatment (NSAIDs then DMARDs) in a 4 year treatment phase.

29   4.3.3 Health economic methodological introduction
30   No health economic papers were identified

31   4.3.4 Clinical evidence statements
32   Recent onset RA (all Level 3 studies)
33
34      •       Radiological damage / disease severity
35
     Study          Prognosis                 Clinical features / predictors (at
                                              presentation)
     1 case-        Severity of RA at         Combination of RF, Hb and platelet level.
     series69       follow-up (mean 6         Best single indicator: RF titre at onset, but
                    years)                    greater accuracy when combined with other
                                              variables
     1 case-        Radiological damage       RF, erosion score and nodules, and 1 year
     series53       (presence of erosions     ESR
                    or not by Larsen score)   Multivariate: combination of: RF and ESR
                    at 3 year follow up       (PPV 68%), 1st year erosion score and ESR
                                              (PPV 84%).

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                44 of 307
DRAFT FOR CONSULTATION


                                     Not predictors (multivariate): erosion score,
                                     swollen joint count and nodules (82%, PPV
                                     77%).
1 case-    Radiographic              Predictors: HAQ score and grip strength
series60   progression (modified
           Sharp score) at 10
           years
1 case-    Greater radiological      Predictors: Anti-CP+ and RF+
series63   progression (greater
           change in Larsen score
           at 2 years)
1 case-    Radiological damage     Predictors (univariate): Joint damage (Sharp
series56   (Sharp van-der Heijde   van-der Heijde score), high ESR, high HAQ
           score) at 3 years       score and poor physician global assessment.
           follow-up               Predictors (multivariate): Joint damage
                                   (Sharp van-der Heijde score), RF+, high
                                   ESR, shorter time from diagnosis, worse
                                   overall patient estimation of health
                                   Not predictor (Univariate): RF+
1 case-    Radiographic            Predictors (univariate): ESR, joint count and
series67   progression (Sharp      grip strength
           scores) at 19 years     Predictors (multivariate): ESR, RF+, joint
           follow-up               count, disease duration and grip strength
                                   Not predictor (Univariate): Age and gender
1 case-    Radiographic            Predictors (bivariate): RF+, ESR and
series65   progression (modified   radiographic damage (modified sharp score)
           sharp score ≤30 or      Predictors (linear regression): ESR and CRP
           >30) at 5 years follow- Not predictors (bivariate): HAQ and CRP
           up                      Not predictors (linear regression): Physical
                                   function and RF+
1 case-    Radiological damage at Predictors: IgM-RF status and radiologic
series58   both 3 and 6 years      score;
                                   Anti-CCP+ (6 years but not 3 years)
                                   Age and IgM RF+ (3 years and 6 years)
                                   DAS (3 years)
1 case-    Radiological            Predictors: CRP levels and number of
series61   progression (change in damaged joints (which was worse in higher
           Larsen score) and new CRP groups)
           joint involvement at 5
           years follow-up
1 case-    Radiological            Predictors (Univariate): Larsen score (best
series55   progression (change in predictor), anti-CCP+, RF+, high ESR and
           Larsen score) at 2      high CRP. Other predictors were greater age,
           years                   smoking and male gender
                                   Predictors (multivariate): Larsen score, anti-
                                   CCP and ESR
                                   Not predictors (Univariate): Pain (VAS) and
                                   HAQ



Rheumatoid arthritis: full guideline DRAFT (August 2008)             45 of 307
DRAFT FOR CONSULTATION


1 case-    Radiographic severity      Predictors (Univariate): High CRP level
series52   (Increase in Larsen        (patients in the top third), high-titer RF,
           score) at 2 year follow-   presence of nodules and being in the upper
           up                         third of number of swollen joints.
                                      Predictors (multivariate): top third CRP level
                                      and high RF titre
           Radiographic severity      Predictors (Univariate): Larsen score at 2
           (Increase in Larsen        years (strongest), CRP was (less strong) and
           score) at 5 year follow-   similar to high RF titre and presence of
           up                         nodules
                                      Predictors (multivariate): CRP and RF
           Radiographic severity      Predictor in patients already with erosions:
           (Increase in Larsen        only top third CRP (2 years); RF at high titre
           score) at 2 and 5 years    (5 years – univariate and multivariate)
           follow-up
1 case-    Radiographic               foot involvement. RF+ correlated with the
series62   progression at 1 year      presence of foot erosions and worse outcome
           follow-up
1 case-    Severe disease and         Predictors (Univariate): RF+ and anti-CCP,
series59   worse radiological joint   more swollen joints and arthritis of the
           damage at 3 years          shoulders, elbows, proximal interphalangeal
           follow-up                  joints, knees and ankles
                                      Predictors (regression): presence of a
                                      swollen knee, total number of swollen joints
                                      (predictor of damage at 1 year follow-up) and
                                      swelling of the knee (predictors of damage at
                                      1, 2 and 3 year follow-up),
                                      Not predictors (Univariate): Prevalence of
                                      swollen MCP and MTP joints
           Joint destruction at 1     Predictors: total number of swollen joints,
           year follow-up             anti-CCP+, CRP level and symptom duration
                                      Not predictors: Presence of knee arthritis
1 case-    Radiographic               Univariate: anti-CCP+, anti-MCV+, RF+ and
series70   progression at follow-     ESR; therapeutic response at 6, 12 or 24
           up (2 years)               months (predicted less radiological
                                      progression).
                                      Multivariate (therapeutic response): Anti-
                                      CCP, anti-MCV and RF (predictors for
                                      radiological progression)

1 case-    Radiographic             Predictors (multivariate): Anti-CCP+
series72   progression (SHS) after (strongest predictor), Female gender, High
           10 years                 ESR and IgM RF+.
                                    Not predictors (multivariate): IgA RF and CRP
1 case-    More rapid decrease in APF- patients; RF+APF- (vs RF+APF+)
series57   joint score; worse joint
           score at 2-3 years
           follow-up
           Joint involvement        Predictor: APF better than RF


Rheumatoid arthritis: full guideline DRAFT (August 2008)             46 of 307
    DRAFT FOR CONSULTATION


                More involvement of         APF+, IIF+ or anti-CCP
                the large joints and
                small joints;
                Large joints (more          APF- (vs APF+); NS for RF+ vs RF-
                rapid decrease)
                Small joints (more          RF+ (vs RF-) but NS
                affected)
                Worse radiological          APF+ (if RF+ or RF-); RF+, anti-CCP+ or
                damage scores at 2-3        APFIIF+ (vs negatives); RF+APF+ (worst),
                years follow-up             RF+APF- (intermediate), RF-APF- (low); NS
                                            (RF+APF- vs RF-APF+); NS (RF-APF+ vs
                                            RF+APF+); RF+APF+ worse (vs RF-APF-)
                Obvious radiological        NS (RF+ vs RF-); APF+ worse (vs APF-);
                damage in the wrist
                More often had              RF+ and APF+ (vs negatives)
                damage in small hand
                and foot joints
    1 case-     Odds of radiographic        Increase of 1 U/ml anti-CCP (0.8% increase);
    series72    progression                 Increase of 50 U/ml (49% increase).

                More likely to develop      low to moderate levels (25 to 200 U/ml) and
                radiographic                high levels(>200 U/ml) worse [vs anti-CCP-
                progression                 patients (<25 U/ml)]; High levels worse (vs
                                            low to moderate)
    1 case-     Higher radiographic         RF+ or ACPA+ patients treated with:
    series of   progression scores          sequential monotherapy, step-up combination
    BeSt                                    therapy or initial combination therapy with IFX
    study73                                 (Groups 1,2 and 4) (vs RF- or ACPA-patients)


                                            NS difference between RF+ or ACPA+ (vs
                                            RF- or ACPA-) for patients treated with initial
                                            combination therapy with CS (group 3)
                Progressive disease         Predictors: RF and ACPA in patients treated
                                            with sequential monotherapy, but not the
                                            other treatment groups
1
2      •    Symptoms/ Function / disability
3
    Study        Prognosis                   Clinical features / predictors (at
                                             presentation)
    1 case-      functional status at        Predictors: combination of RF, Hb and platelet
    series69     follow-up (mean 6           count, which achieved success in 62% of
                 years)                      patients
                                             Not predictor: RF on its own
    1 case-      functional disability       Predictor (bivariate): HAQ score, ESR and
    series65     (HAQ) at 5 years follow-    radiolographic damage (modified Sharp score).
                 up                          Predictor (regression): HAQ and age



    Rheumatoid arthritis: full guideline DRAFT (August 2008)                47 of 307
    DRAFT FOR CONSULTATION


                                             Not predictor (bivariate): RF+ and CRP
                                             Not predictor (regression): Radiographic
                                             damage
    1 case-      HAQ functional disability   Gender, disease activity, IgM RF+, and age
    series58     at both 3 and 6 years

    1 case-      Higher disease activity     anti-MCV+ (vs anti-MCV-); anti-CCP+ (vs Anti-
    series71     (Physician’s assessment     MCV+/anti-CCP- for Physician’s assessment)
                 and DAS28 score) and
                 more swollen and tender
                 joints
                 CRP, ESR, Physicians’       NS: Anti-MCV+/anti-CCP- vs anti-MCV-/anti-
                 assessment of disease       CCP- (but Anti-MCV+/anti-CCP- worse for
                 activity, Number of         HAQ)
                 tender and swollen
                 joints, DAS28 score,
                 Global VAS score, Pain
                 (VAS) score); HAQ at 3
                 years
1
2      •    Mortality / development of serious disease
3
    Study        Prognosis           Clinical features / predictors (at presentation)
    1 case-      Mortality at        Best predictor: combination of ExRA Malmo and RF+
    series64     follow-up (mean     ExRA - Malmo criteria (strongest single predictor): RR
                 16 years)           4.3, CI 2.9 to 6.3), Presence of subcutaneous
                                     rheumatoid nodules and presence of RF (moderate
                                     predictors: RR 1.5 and 1.9 respectively).
    1 case-      Increased risk of   Predictors (Univariate): male gender, higher age at
    series66     mortality and CV    disease onset, earlier progression of erosions, higher
                 events at follow-   ESR, CS treatment given early in disease. DMARD
                 up (mean range      treatment (>2 drugs) was associated with decreased risk
                 17-21 years)        Predictors (multivariate – increased risk): male gender,
                                     higher age at disease onset and last value ESR
                                     Not predictors (Univariate): prolonged / extensive CS
                                     treatment
    1 case-      Mortality at        Predictors (Univariate): Age (strongest predictor); HAQ
    series68     follow-up (up to    disability (most important, better predictor in men than
                 23 years)           women), followed by Global disease severity, pain,
                                     depression, anxiety and grip strength. Lab variables
                                     were less important
                                     Predictors (multivariate): only radiographic progression
                                     Weak predictors (Univariate): RF+, nodules and
                                     radiographic progression rates
4
5      •    Remission
6
    Study        Prognosis               Clinical features / predictors (at presentation)



    Rheumatoid arthritis: full guideline DRAFT (August 2008)             48 of 307
    DRAFT FOR CONSULTATION


    1 case-       Remission at follow-     Predictors (Univariate): RF+ (latex test)
    series46      up (up to 23 years)      Predictors (regression): ACR criteria, latex test RF-
                                           and lower duration of disease
    1 case-       Remission at follow-     Predictors (Univariate): gender, duration of disease,
    series54      up (3 months, 6          anti-CCP, RF, DAS28, HAQ
                  months, 1 year, 18       Predictors (regression): Male gender (major
                  months, 2 years and      predictor), short disease duration, low DAS28, low
                  5 years)                 HAQ, RF-
                                           Not predictors (Univariate): SOFI
1
2   Established RA (All Level 3 studies)
3
4      •    Radiological damage / disease severity
5
    Study       Prognosis              Clinical features / predictors (at
                                       presentation)
    1 case-     TJA at follow-up       ESR, WBC count, haemoglobin level, HAQ
    series46    (up to 23 years)       disability score, global severity score, BMI,
                                       disease duration and smoking (past or
                                       current).
    1 case-     Radiological           Predictor: RF status (in patients with
    series74    progression            disease duration greater than 12 years
                (change in Larsen      only)
                score) at 5 years      Not predictor: AFA status
                follow-up
6
7      •    Function / disability
8
    Study         Prognosis              Clinical features / predictors (at presentation)
    1 case-       HAQ score at 8         HAQ score, number of work hours and global health
    series76      years follow-up        status

    1 case-       HAQ score at 5         Predictors (univariate): RAI and CRP levels
    series77      years follow-up        Predictors (multivariate): RAI, pain (VAS), early
                                         morning stiffness and radiographic progression
                                         (modified Larsen score)
    1 case-       Worse disability       older age, female gender, duration of illness,
    series78      score at follow-up     radiologic variable, initial disability, elevated ESR and
                  (mean range 1.7 to     latex titres
                  12 years)
    1 case-       More functional        Predictors (univariate): higher age, longer disease
    series80      disability at 13       duration; higher ESR scores over time; higher RAI
                  years follow-up        scores, more pain and distress and more disability
                                         over the preceding years
                                         Predictors (multivariate): disease duration; disability,
                                         ESR and pain and distress over the preceding years

                                         Not predictor (multivariate): RAI over the preceding
                                         years


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 49 of 307
     DRAFT FOR CONSULTATION


                 More functional         Predictors (univariate): female gender, longer disease
                 disability at 21        duration; higher RAI scores, more pain and distress,
                 years follow-up         less social companionship and more disability over the
                                         preceding years
                                         Predictors (multivariate): gender, disease duration;
                                         RAI, disability and pain over the preceding years
                                         Not predictor (multivariate): social companionship,
                                         distress and ESR over the preceding years
     1 case-     Low disability          NS difference (RF+ or RF- patients)
     series83    score (HAQ <0.5)
                 at 6 months follow-
                 up
1
2       •   Mortality / development of serious disease
3
     Study       Prognosis             Clinical features / predictors (at presentation)
     1 case-     Mortality at 8        age, followed by prednisone use and HAQ score; Male
     series76    years                 gender

     1 case-     Mortality at 2        Predictors (univariate): RF+, high levels of anti-CCP (but
     series79    years                 not anti-CCP+ >25U)
                                       Predictors (multivariate): age, gender, disease duration,
                                       RF+, High IgA and IgM RF
                                       Not predictors (univariate): RF+ and/or anti-CCP+,
                                       pANCA+ and high ANCA titres
                                       Not predictors (multivariate): IgG RF, anti-CCP levels,
                                       pANCA and high ANCA titres. If HAQ or subcutaneous
                                       nodules were added into the model, high anti-CCP level
                                       or RF+ did not predict mortality
     1 case-     CVD and new           presence of ExRA (adjusted for age, sex and smoking).
     series81    onset coronary
                 artery disease at
                 follow-up (mean
                 16 years)
4
5       •   Remission
6
     Study      Prognosis            Clinical features / predictors (at presentation)
     1 case-    Remission at 6       Male gender, age of patient <53 years, HAQ score
     series83   months               <1.63
                (patients
                receiving anti-
                TNFs)
 7
 8   Treatment of poor prognosis patients
 9
10      •   Symptoms
11


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              50 of 307
    DRAFT FOR CONSULTATION


    Study      Treatment     Outcomes                          Follow-     Result -
                                                               up          best
                                                                           treatment
    1 RCT85    IFX + MTX     Reduction in synovitis (MRI)      14 and      Arm 1
    Level      vs MTX                                          54          (p<0.05)
    1++                                                        weeks

                             DAS8 score                        14          Arm 1
                                                               weeks       (p<0.05)
                             ACR20, ACR50 and ACR70                        NS

                             reduction in bone oedema          54          Arm 1
                             (MRI); ACR50 and ACR70            weeks       (p<0.05)

                             ACR20; DAS8 score                             NS

                             ACR20, 50 and 70; DAS8            2 year      NS
                             score                             follow-up

    1 RCT84    CSA + MTX Swollen joint count                   24 and      Arm 1
    Level      +                                               48
    1+         methylpredni Pain (VAS), DAS28 score            weeks       NS
               solone vs
               SSZ          Tender joint count                 24          Arm 1
                                                               weeks

                             ACR20 and ACR50, Tender           48          NS
                             joint count                       weeks

    1 RCT86    MTX vs        Number of patients                30          MTX (poor
    Level      placebo       developing RA and DAS             months      prognosis
    1+                       score                                         patients)

                                                                           NS (good
                                                                           prognosis
                                                                           patients)
1      •    Function
2
    Study      Treatment     Outcomes                          Follow-     Result -
                                                               up          best
                                                                           treatment
    1 RCT85    IFX + MTX     HAQ score                         14 and      Arm 1
    Level      vs MTX                                          54          (p<0.05)
    1++                                                        weeks
                                                               and 2
                                                               year
                                                               follow-up




    Rheumatoid arthritis: full guideline DRAFT (August 2008)               51 of 307
    DRAFT FOR CONSULTATION


    1 RCT84    CSA + MTX HAQ score                             24 and     NS
    Level      +                                               48
    1+         methylpredni                                    weeks
               solone vs
               SSZ

1
2      •    Joint damage
3
    Study      Treatment      Outcomes                          Follow-    Result -
                                                                up         best
                                                                           treatment
    1 RCT85    IFX + MTX      New erosions (MRI)                14 and     Arm 1
    Level      vs MTX                                           54         (p<0.05)
    1++                                                         weeks

                              Radiographic progression          14         NS
                              (SHS)                             weeks

    1 RCT84    CSA + MTX      Radiographic progression -     48            NS
    Level      +              SHS (total score, erosions and weeks
    1+         methylpredni   JSN)
               solone vs
               SSZ


    1 Cohort early            Change in joint damage in         from 0-2 Early
    study49,5 treatment vs    patients with Sharp score >0      and from
    0
              delayed                                           0-4 years
    Level     treatment
    2+



                                                                1-4 years NS

                              Change in joint damage in         from 0-2 NS
                              patients with Sharp score 0       and from
                                                                0-4 years
                                                                and 1-4
                                                                years

    1 RCT86    MTX vs         Slowing radiographic              30         Arm 1
    Level      placebo        progression (SHS score) in        months     (p<0.001 or
    1+                        poor prognosis patients (anti-               p=0.036)
                              CCP+ or RF+)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)              52 of 307
     DRAFT FOR CONSULTATION


                                  Slowing radiographic                          NS
                                  progression (SHS score) in
                                  good prognosis patients (anti-
                                  CCP- or RF-)

1
2       •    Global assessment
3
     Study        Treatment       Outcomes                         Follow-      Result -
                                                                   up           best
                                                                                treatment
     1 RCT84      CSA + MTX       Patient’s global assessment      24 and       NS
     Level        +                                                48
     1+           methylpredni                                     weeks
                  solone vs
                  SSZ

4
5       •    QoL
6
     Study        Treatment       Outcomes                          Follow-      Result - best
                                                                    up           treatment
     1 RCT85      IFX + MTX       RAQoL score                       14 and       Arm 1
     Level        vs MTX                                            54 weeks     (p<0.05)
     1++                                                            and 2
                                                                    year
                                                                    follow-up

7
8       • Biochemical markers
     Study    Treatment       Outcomes                       Follow-up               Result -
                                                                                     best
                                                                                     treatment
     1 RCT85       IFX + MTX vs      CRP levels (AUC)        Over the 54             Arm 1
     Level         MTX                                       weeks                   (p<0.05)
     1++
                                                             54 weeks to 2           NS
                                                             year follow-up

     1 RCT84       CSA + MTX +    CRP and ESR                24 and 48 weeks         NS
     Level 1+      methylpredniso
                   lone vs SSZ

     1 Cohort      early treatment   change in joint         2 years                 Early
     study49,50    vs delayed        damage (RF+ and
     Level 2+      treatment         RF- patients)

 9
10
11

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 53 of 307
     DRAFT FOR CONSULTATION


1       •       Remission
2
     Study        Treatment     Outcomes                        Follow- Result -
                                                                up      best
                                                                        treatment
     1            IFX + MTX      remission time and             Over    Arm 1
     RCT85        vs MTX         remission rates                the 2   (p<0.05)
     Level                                                      years
     1++
     1            CSA +          Remissions (% patients,        48         NS
     RCT84        MTX +          ACR)                           weeks
     Level        methylpred
     1+           nisolone vs
                  SSZ

3
4       •       AEs
5
     Study        Treatment     Outcomes                        Follow- Result -
                                                                up      best
                                                                        treatment
     1            IFX + MTX      AEs                            Over    Similar
     RCT85        vs MTX                                        the 2
     Level                                                      years
     1++
6
7       •       Withdrawals
8
     Study        Treatment     Outcomes                        Follow- Result -
                                                                up      best
                                                                        treatment
     1            IFX + MTX      Withdrawals                    Over    Similar
     RCT85        vs MTX                                        the 2
     Level                                                      years
     1++
     1            CSA +          withdrawals due to lack of     48         Arm 1
     RCT84        MTX +          efficacy                       weeks
     Level        methylpred
     1+           nisolone vs    withdrawals due to AEs                    NS
                  SSZ


9    4.3.5 Summary of evidence statements
10          •    In early disease most of the studies looked at radiological outcomes
11               over variable follow-up periods. The themes that emerge are:
12
13          •    RF titre stands out as repeatedly being a good predictor of prognosis
14               (both for radiology and function) in most studies 52-55,58,59,62,63,65,67,69


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                      54 of 307
     DRAFT FOR CONSULTATION


1         •   For those studies in which it was done, anti-CCP positivity is a
2             powerful predictor of prognosis55,59,63, with interesting interactions
3             with RF (worst prognosis RF+ACCP+, best prognosis RF-ACCP-,
4             with single positivities for one or the other showing intermediate
5             prognoses) 57
6         •   Other variables appearing to predict prognosis in more than one
7             study include:
8               • baseline radiological score53,55,56,58,65, nodules52,53
9               •   acute phase markers52,53,56,59,61,65,67
10              •   HAQ score54,56,60
11              • grip strength60,67
12              • swollen joint count52,59,67.
13        •   In established disease studies looked at a mixture of functional and
14            radiological outcomes. Recurring themes are:
15              • Disability is predicted by:
16                  − baseline disability score76,78,82
17                  − older age 78-80 and,
18                  − longer disease duration78-80,82.
                                                          78,79
19              • Women tend to do worse than men
20              • No study examined whether poor prognosis patients should be
21                  treated differently.

22   4.3.6 From evidence to recommendations
23   The GDG noted that currently there is not universal availability for testing of
24   anti-CCP antibodies. Nevertheless, it was felt that for prognosis there was
25   now sufficient evidence to warrant widespread access to this test. Anti-CCP
26   appears to add information above and beyond testing for rheumatoid factor,
27   and there is evidence of synergism between the two, with being positive for a
28   combination of both rheumatoid factor and anti-CCP associating with a
29   particularly poor prognosis. The principle strength of anti-CCP antibodies
30   appears to be in people who are seronegative for rheumatoid factor where an
31   anti-CCP test can be helpful both for diagnosis and prognosis. Discussion in
32   the GDG also suggested that an anti-CCP test was useful in people who are
33   seropositive for rheumatoid factor, but where the clinical picture is not
34   suggestive of RA. A positive anti-CCP test under these circumstances would
35   suggest that this individual is at risk of developing RA subsequently, and
36   should be followed up closely. However, there is currently only limited data
37   available in this group of patients. In people who are understandably reluctant
38   to go onto intensive therapy for a recent onset of persistent synovitis, being
39   positive for anti-CCP might be used to help to inform their decision about
40   taking such medication, because of the prognostic value of this test.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 55 of 307
    DRAFT FOR CONSULTATION


1   4.3.7 Recommendations
2        •   R 5 People with suspected rheumatoid arthritis who are seronegative
3            for rheumatoid factor should have their anti-CCP antibodies tested.
4        •   R 6 Anti –CCP antibody status should be measured in people with
5            suspected or definite RA when more information about prognosis is
6            required




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           56 of 307
     DRAFT FOR CONSULTATION


1    5         The patient

2    5.1       Patient perceptions and beliefs

3    5.1.1 Clinical introduction
 4   Guidelines are ultimately produced for the benefit of patients and it is
 5   appropriate that this guideline includes a section on the beliefs and
 6   perceptions that patients have about their disease. A paternalistic approach in
 7   which the healthcare professionals know best, and the patient is a passive
 8   recipient of care, should be a distant memory of how medicine might once
 9   have been practised. Many patients understandably desire to be active
10   participants in their own care, and the associated decision-making process.
11   However, even when professionals strive for patient involvement in disease
12   management, there is still scope for disconnection between patients, their
13   carers, and their healthcare professionals. This applies to the perception of
14   the disease and therapeutic interventions, and the priorities for their
15   healthcare needs.
16
17   Some pre-conceptions about RA in newly diagnosed patients may need to be
18   addressed early in the disease course. Patients who attend their first clinic
19   appointment may already have received a disease label diagnosis from their
20   GP, and this may have led to a number of myths and rumours about the
21   disease. These may have been alluded to by well-meaning but ill-informed
22   friends and relatives, and sometimes, unfortunately, other health
23   professionals. Some pre-conceptions about the disease may have to be
24   addressed before appropriate perceptions and beliefs can be fostered.
25
26   For people with established RA it will be important to appreciate that their
27   perceptions of RA and their priorities in its management might differ from
28   those of the multidisciplinary team managing them.
29
30   Although this is a very broad field, the two key areas that need to be
31   addressed are:
32        • What patient experiences, perceptions and beliefs exert positive
33           impacts on symptoms, joint damage, function and quality of life, and
34           which of these can be identified, fostered and encouraged by the
35           patient themselves and those seeking to help them?
36         •    Conversely which patient experiences, perceptions and beliefs exert
37              a negative impact and need to be minimised or avoided by the patient
38              themselves and those seeking to help them?

39   5.1.2 Clinical methodological introduction
40   We looked for studies that investigated patients’ experiences of RA and its
41   treatments. Due to the large volume of evidence, studies were only included if
42   they were published within the last 10 years (1997onwards), for non-
43   qualitative studies had a sample size of N>100, had a UK-relevant population

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               57 of 307
     DRAFT FOR CONSULTATION

1    and if the population was mixed arthritis there had to be >75% RA or RA
2    subgroup analysis.
3
4    Recent onset of RA:
5    8 studies were found 65,87-93 that fulfilled the criteria. Studies differed with
6    respect to the following:
7         • Sample size (range: N=68 to N=573)
8         •   Trial length – follow-up (range: immediate to 5 years)
 9        •   Trial design (N=6 observational-correlation studies; N=1
10            observational-longitudinal study; N=1 cross-sectional study; N=1
11            retrospective observational study of an RCT)
12   Established RA:
13   25 studies were found94-117 that fulfilled the criteria. Two papers123124 reported
14   the same trial but different outcomes. Results from both papers have been
15   reported here but this trial has been counted once. Studies differed with
16   respect to the following:
17        • Sample size (range: N=6 to N=7702)
18        •   Trial length – follow-up (range: immediate to 5 years)
19        •   Trial design (N=13 observational-correlation studies; N=7 qualitative
20            studies; N=4 cross-sectional studies; N=1 case-control study; N=1
21            observational study)
22   Disease duration mixed (recent onset and established RA) or not mentioned
23   5 studies were found118-122 that fulfilled the criteria. Studies differed with
24   respect to the following:
25        • Sample size (range: N=10 to N=190)
26        •   Trial length – follow-up (N=4 trials immediate, N=1 trial 1 year)
27        •   Trial design (N=2 observational-correlation studies; N=2 qualitative
28            studies; N=1 cross-sectional study)

29   5.1.3 Health economic methodological introduction
30   No health economic papers were identified for this question

31   5.1.4 Clinical evidence statements
32   All trials were evidence grade 3 except for the qualitative studies which were
33   given a 3+.
34
35   Recent onset of RA:
36   There were several main themes that emerged from the studies.
37
38   1. Areas of life worst affected and how the disease impacts patients (4
39   studies87,88,90,118)
40
41   Observational studies found that pain, ability to control pain and
42   dissatisfaction with abilities affected psychological well-being, self-esteem and
43   adjustment to disease. High levels of anxiety and depression were associated
44   with fatigue, pain and low acceptance. Additionally, patients who experienced
     Rheumatoid arthritis: full guideline DRAFT (August 2008)                    58 of 307
     DRAFT FOR CONSULTATION

 1   more fatigue were more at risk of pain, were more disabled, felt more
 2   depressed, had lower self-esteem, were less satisfied with the support
 3   provided to them, showed more reduction in leisure activities, felt less
 4   independent and adjusted, and appraised their health as markedly less well.
 5
 6   2. Care – areas patients found important, areas for improvement -
 7   including information provision (1 study118)
 8
 9   Observational studies found that patient knowledge and the need for
10   information was the same for patients with a recent onset of disease and
11   those with established RA.
12
13   3. Correlations between demographics, disease characteristics and
14   disease measures (4 studies65,89,91,92)
15
16   Observational studies found that:
17       • Women were significantly worse than men for many measures of
18          symptoms, function and QoL, but not for pain, hand x-ray
19          abnormalities and CRP level.
20        •   Quality of life was weakly associated with clinical and laboratory
21            variables and patients’ global assessment was associated with pain,
22            depression, disability and tender joints.
23        •   Physical health status outcomes were predicted by baseline values of
24            HAQ, AIMS physical, high age and AIMS psychological health status.
25            Baseline HAQ and AIMS physical were associated with physical
26            disability. Psychological health status was predicted by AIMS
27            psychological dimension.
28        •   Older, less anxious patients (STAI-SF) were significantly more likely
29            to discontinue to take their initial DMARDs within the first year.
30            Continuing to take DMARDs was associated with HAQ, RHD
31            (relationship with hospital doctors), BMQ (Beliefs about medication
32            questionnaire), and SOS (Significant others scale).
33   4. Attitudes to treatment (including preferences and effectiveness); 1
34   study93 of the BeSt trial
35
36        •   There was NS difference between the 4 treatment groups for: much-
37            to very much Improvement of general health since start of treatment;
38            current state of health with the medication they had to take (however
39            group 3 were less satisfied). Groups 1 and 2 had significantly less
40            rapid relief of symptoms than groups 3 and 4 but had NS difference
41            from each other.
42        •   Patients’ preference for a particular group before start of study: No
43            preference (44%); Only group 3 had an effect of group allocation –
44            22% of patients who actually received this treatment had hoped not to
45            be assigned to group 3, whereas this percentage was much higher
46            (>40%) in the other groups.
47        •   Treatment patients would prefer if diagnosed with RA today:
48            Treatment with 1 well-known anti-rheumatic drug (21%), Combination

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              59 of 307
     DRAFT FOR CONSULTATION

1             without prednisone (19%); Combination with prednisone (12%);
2             Combination with the newest IV drug (IFX at the time – 44%).
3         •   Patients feelings about taking prednisone: 50% of patients assigned
4             to Group 3 disliked taking prednisone (15%, 20% and 9% in groups 1,
5             2 and 4 respectively).
6         •   Patients feelings about going to hospital for IV treatment: 8% of
7             Group 4 patients disliked having to go to hospital for IV treatment.
8             (2%, 3% and 2% in groups 1, 2 and 3 respectively).
 9   Established RA:
10
11   There were several main themes that emerged from the studies.
12
13   1. Areas of life that patients most wanted improvement (2 studies97,115)
14
15   Recent onset of RA:
16
17   There were several main themes that emerged from the studies.
18
19   1. Areas of life worst affected and how the disease impacts patients (4
20   studies87,88,90,118)
21
22   Observational studies found that pain, ability to control pain and
23   dissatisfaction with abilities affected psychological well-being, self-esteem and
24   adjustment to disease. High levels of anxiety and depression were associated
25   with fatigue, pain and low acceptance. Additionally, patients who experienced
26   more fatigue were more at risk of pain, were more disabled, felt more
27   depressed, had lower self-esteem, were less satisfied with the support
28   provided to them, showed more reduction in leisure activities, felt less
29   independent and adjusted, and appraised their health as markedly less well.
30
31   2. Care – areas patients found important, areas for improvement -
32   including information provision (1 study118)
33
34   Observational studies found that patient knowledge and the need for
35   information was the same for patients with a recent onset of disease and
36   those with established RA.
37   3. Correlations between demographics, disease characteristics and
38   disease measures (4 studies65,89,91,92)
39
40   Observational studies found that:
41       • Women were significantly worse than men for many measures of
42          symptoms, function and QoL, but not for pain, hand x-ray
43          abnormalities and CRP level.
44        •   Quality of life was weakly associated with clinical and laboratory
45            variables and patients’ global assessment was associated with pain,
46            depression, disability and tender joints.
47        •   Physical health status outcomes were predicted by baseline values of
48            HAQ, AIMS physical, high age and AIMS psychological health status.
49            Baseline HAQ and AIMS physical were associated with physical
     Rheumatoid arthritis: full guideline DRAFT (August 2008)               60 of 307
     DRAFT FOR CONSULTATION

1             disability. Psychological health status was predicted by AIMS
2             psychological dimension.
3         •   Older, less anxious patients (STAI-SF) were significantly more likely
4             to discontinue to take their initial DMARDs within the first year.
5             Continuing to take DMARDs was associated with HAQ, RHD
6             (relationship with hospital doctors), BMQ (Beliefs about medication
7             questionnaire), and SOS (Significant others scale).
 8   4. Attitudes to treatment (including preferences and effectiveness); 1
 9   study93 of the BeSt trial
10
11        •   There was NS difference between the 4 treatment groups for: much-
12            to very much Improvement of general health since start of treatment;
13            current state of health with the medication they had to take (however
14            group 3 were less satisfied). Groups 1 and 2 had significantly less
15            rapid relief of symptoms than groups 3 and 4 but had NS difference
16            from each other.
17        •   Patients’ preference for a particular group before start of study: No
18            preference (44%); Only group 3 had an effect of group allocation –
19            22% of patients who actually received this treatment had hoped not to
20            be assigned to group 3, whereas this percentage was much higher
21            (>40%) in the other groups.
22        •   Treatment patients would prefer if diagnosed with RA today:
23            Treatment with 1 well-known anti-rheumatic drug (21%), Combination
24            without prednisone (19%); Combination with prednisone (12%);
25            Combination with the newest IV drug (IFX at the time – 44%).
26        •   Patients feelings about taking prednisone: 50% of patients assigned
27            to Group 3 disliked taking prednisone (15%, 20% and 9% in groups 1,
28            2 and 4 respectively).
29        •   Patients feelings about going to hospital for IV treatment: 8% of
30            Group 4 patients disliked having to go to hospital for IV treatment.
31            (2%, 3% and 2% in groups 1, 2 and 3 respectively).
32   Established RA:
33
34   There were several main themes that emerged from the studies.
35
36   1. Areas of life that patients most wanted improvement (2 studies97,115)
37
38        •   Observational and qualitative studies found that patients most
39            wanted improvement in pain, function (hand and fingers) and walking
40            and bending. Patients wanting improvements in pain had a lower pain
41            self-efficacy, greater fatigue, worse global health and used more
42            analgesic drugs. Older people most wanted improvement in function,
43            whereas younger patients priorities were pain, work and mental
44            conditions. When footwear was discussed, women wanted more
45            information and time on which to base their choice, they wanted to
46            voice their opinions, know they were being listened to and
47            acknowledged that they understood their disease. A feeling of trust in
48            the practitioner was seen as important factor in the consultation. Men
     Rheumatoid arthritis: full guideline DRAFT (August 2008)               61 of 307
     DRAFT FOR CONSULTATION

1             did not mention any aspect of their experience that needed
2             improving.
 3   2. Areas of life worst affected and how the disease impacts patients (10
 4   studies94,95,98,100,104,107-109,111,114,115,123)
 5
 6   Observational, qualitative and case-control studies found that:
 7       • Areas that caused major problems for patients were: their identity
 8          (private and public sphere), physical ADLs which they valued, QoL,
 9          Pain, physical and mental fatigue, decreased activity, depression,
10          loss of confidence and motivation, frustration, self-consciousness or
11          embarrassment at deformities, effect on relationships and family life.
12        •   Patients were concerned about: the future (increases in pain and
13            disability); the inefficacy of treatments, inabilities (to carry out ADLs,
14            to be sexually active, to work and its financial implications, to go out
15            due to pain).
16        •   Ongoing emotions expressed by patients were: grieving and anger
17            due to the loss of ability to do things while making necessary changes
18            in lifestyle, courage to confront daily pain and apparent losses, fear of
19            the future and of medication side-effects, frustration and depression.
20        •   A number of patients found they had to persuade themselves and
21            others of the authenticity of their RA (the disease was ‘invisible’ in the
22            early stages and people don’t understand the disease). Because of
23            negative reactions, some patients pretended to be well when they
24            were not. Patients wanted to feel valued by society (have their
25            difficulties appreciated and understood). They also felt an essential
26            form of support was to receive validation and understanding from
27            family. Family/friends often tended to overestimate the severity and
28            characteristics of pain and to underestimate negative effects of RA on
29            the patient’s life. Physicians, on the contrary, tended to underestimate
30            pain severity and characteristics. It is important for husbands to
31            understand wives’ views on their control over RA and its cyclic
32            nature, have optimistic views and not to underestimate the
33            consequences of RA.
34        •   Patients had to mastermind new ‘lifeways’ (find methods of disease
35            management, adapt to changes, develop new skills and reconcile lost
36            abilities). Accommodations were widely used to perform daily
37            activities. These included: limits and more time for more activities,
38            assistance and devices.
39        •   Patients’ experiences of pain were described as: variable (80%),
40            unpredictable (68%), caused major interference with paid work or
41            domestic chores (67%), underestimation of pain by the spouse (23%
42            patients) and by the physician (14% patients), by other family
43            members or friends (38%).
44        •   Women were worse than men for QoL. QoL declined with age. The
45            impact of RA on mental health was lower in patients <50 years old
46            compared to other age groups



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  62 of 307
     DRAFT FOR CONSULTATION

1         •   Women were concerned about the look of their feet and their
2             footwear whereas none of the men talked about appearance. Men
3             were positive about footwear and both felt it helped their mobility and
4             reduced pain.
5         •   RA affected patients’ physical identity, social role and self-image and
6             many changed their physical appearance to accommodate
7             restrictions or tried to hide physical deformities.
 8        •   The importance of disease outcomes changes with time and depends
 9            on circumstances (eg. at different stages of disease and during flare-
10            ups).
11   3. Care – areas patients found important, areas for improvement -
12   including information provision (10 studies94,95,98,99,101,102,113,115,117,118,124)
13
14   Observational and qualitative studies found that:
15       • Areas of care most important to patients were: knowledge of RA (no
16          difference between those with recent onset and established RA);
17          information about concomitant medication and about medication
18          efficacy and side-effects; communication (wanted it to be clear and
19          effective and valued positive relationships with practitioners); to be
20          able to access practitioners between scheduled appointments;
21          access to other departments; familiarity with the staff.
22        •   Areas of care deemed inadequate were: MDT care (lack of modified
23            toilet in the practice; having access to their file; choice of care
24            provider; information on RA, course of symptoms, aids and home
25            adjustments; good care co-ordination and being open to questions);
26            limited contact with providers; lack of continuity of care; Social
27            support (unhappy having to rely on partners or family members); to
28            be more involved in medical decisions (although some wanted the
29            doctor to take more control as the disease gets worse).
30        •   Information - patients wanted more (especially women; no difference
31            in recent onset and established RA patients) and the content to be
32            about: diagnosis, pathogenesis and medication; exercise; daily
33            activities; the disease as they get sicker; purpose of lab tests; other
34            ways to treat a problem. Many felt there was a lack of clear and
35            unambiguous information throughout their treatment and lack of
36            general advice on services available and claiming financial benefits.
37            They wanted both oral and written information.
38        •   Satisfaction with care: somewhat satisfied or very satisfied (68%);
39            experienced unmet healthcare needs (particularly physical
40            symptoms, consequences of disease - body structures and function,
41            quality of care and healthcare services, 27%). Unmet healthcare
42            needs were associated with worse health, more comorbidities and
43            dissatisfaction; there were concerns about communication (with
44            healthcare professionals). Women trusted the practitioners’ skills
45            regarding footwear, but felt negatively about the way they were
46            treated by the assessors who often dismissed their concerns and
47            needs. Men felt differently – had some camaraderie with and trusted
48            the skills of the practitioners.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                   63 of 307
     DRAFT FOR CONSULTATION

1         •   Decisions about treatment efficacy were based upon symptom
2             reduction, ‘forgetting you have RA’, change in priorities for outcomes
3             over time, magnitude of improvement/change varies with disease
4             duration.
 5        •   Patients wanted to feel in control of their condition and tended to
 6            refuse interventions as a way of gaining control; they felt hope when
 7            medical staff searched for new treatment options. Treatments gave
 8            them improvement in symptoms, helped them get back to normal,
 9            and gave them better sleep. Biologics particularly had positive
10            physical and emotional effects. Satisfaction with medication was
11            significantly greater among patients taking biologics than those not
12            taking them
13        •   Most patients would not want to change their therapy (including those
14            on biologics) as long as their condition didn’t get worse. Many felt
15            their physician thought they did not need to change or there were no
16            better medications than those they were currently taking and some
17            did not want IV administration or injection.
18        •   Complementary therapies: many had tried one or more for their pain
19            including acupuncture and massage. One patient mentioned they
20            “can’t do without…acupuncture and massage…and heat really
21            helped”. Many were told they had no choice but to take toxic drugs to
22            slow deterioration or alleviate their symptoms and were concerned
23            about side-effects. Nearly all patients hoped that new research would
24            find a cure
25        •   Many patients wanted better feedback from secondary care. Primary
26            care was described in both complimentary and critical ways (delays in
27            diagnosis and early care).
28        •   Many patients presented themselves to healthcare staff as a ‘coper’
29            or tried to please staff ‘by not being a nuisance’. They were more at
30            ease with nurses than doctors, and only half of them were receiving
31            treatment from other MDT members. The presence of medical or
32            nursing students and seeing a male doctor by female patients made a
33            number of patients feel uncomfortable, especially when talking about
34            personal issues (gynaecological/emotional).
35   4. Correlations between demographics, disease characteristics and
36   disease measures (12 studies96,100-103,106,110,112,116-118)
37   Observational studies found that:
38        • Predictors of poorer outcomes (including disability, pain and QoL)
39           included: Pain, fatigue, RA duration disease activity, depression
40           radiological damage, female, older age, less favourable
41           socioeconomic status, had paid work less often and were less often
42           married/co-habiting, higher disease activity, more
43           somatic/psychological comorbidity, higher number of activities
44           affected by RA,
45        •   Low involvement in decisions was associated with: younger age,
46            greater satisfaction with care, living with a partner, still working,
47            longer time in formal education, less comorbidities, lower disease
48            activity, lower fatigue and lower levels of pain. High involvement in
     Rheumatoid arthritis: full guideline DRAFT (August 2008)               64 of 307
     DRAFT FOR CONSULTATION

1             decisions was associated with: younger age, high levels of formal
2             education, high levels of received patient information and high levels
3             of patient satisfaction.
4         •   The need for information was associated with: age and education (in
5             women) and with fatigue, number of DMARDs taken and experience
6             of AEs (in men). Decision-making preferences was associated with:
7             age, education, number of DMARDs and RA knowledge (in women)
8             and RA knowledge (in men).
 9        •   Poor function, greater impairment and disability in valued activities
10            was associated with lower satisfaction with physical abilities, which
11            was in turn associated with greater depression.
12        •   High levels of anxiety and depression were predicted by higher pain,
13            fatigue and lower acceptance.
14        •   Passive coping was a psychological predictor of both pain and
15            depression, and a mediator of the impact of physical disability on both
16            pain and depression.
17        •   Patients with RA performed significantly less non-vocational activities
18            compared to 1 an 10 years ago but were positive about their ability to
19            perform more activities in the future.
20        •   Performing a large number of activities was correlated with a good
21            mental health status or psychological well-being, and the reverse was
22            true for a low amount of activities performed.
23        •   SF-36 role physical correlated with number of activities patients
24            performed in past (1 year but not 10 years), at present and number
25            planned to pursue in the future. Whereas SF-36 physical function
26            only correlated with number planned to pursue in the future
27        •   The patient’s disability was a stressor for patients but not for
28            partners. Stressors for partners and not patients were: negative
29            transactions and marital quality and patient’s disability was linked to
30            the partner’s burden. The effect of marital quality on patient’s distress
31            depended on partners’ burden.
32        •   Many patients with poor HAQ or PAS scores (ie. poor function or
33            disease activity levels), were satisfied with their RA control, while
34            others with ‘good’ scores were dissatisfied.
35        •   Purpose In Life was associated with (multivariate): younger age,
36            participation in leisure/social activities, better mental health and an
37            optimistic coping style were associated with.
38
39
40   Disease duration mixed (recent onset and established RA) or not
41   mentioned:
42
43   1. Areas of life worst affected and how the disease impacts patients (2
44   studies120,121)
45
46   Observational and qualitative studies found that:

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 65 of 307
     DRAFT FOR CONSULTATION

1         •   patients either regarded RA as a challenge for mastery in their lives
2             and were actively trying to master it, or they adapted to their disease
3             and felt it was ‘something to get used to’ and ‘made the best out of a
4             bad situation’
5         •   Employed patients had significantly lower depression than those not
6             employed.
 7
 8   2. Treatment – preferences and decision-making (2 studies119,122)
 9
10   Observational and qualitative studies found that:
11       • Who should choose medicine (views of patients who had not used
12          anti-TNFs): Rheumatologists (41%); Decide themselves (33%);
13          unsure (18%); joint decision (7%). Men were significantly more likely
14          to want rheumatologists and there was NS difference between young
15          and old patients. Some patients did not feel confident to make
16          decisions without further support and discussion with healthcare staff
17          and one patient felt internet information had ‘death was quoted an
18          awful lot’
19        •   Who should choose medicine (views of patients who had used anti-
20            TNFs): All patients wanted to be involved and those who had been
21            involved found shared decision-making positive and beneficial and
22            made them want to choose their treatment.
23        •   4 ways (themes) in which treatment decisions had been arrived at:
24            • 1. Relinquished decision “leave it in the hands of the doctor” as
25                the “doctor knows best”;
26            • 2. Forced/informed choice the doctor’s preference maybe
27                because he had more success with a particular drug so he
28                “pushed it…whereas the other drug might be the one that you
29                really want”;
30            • 3. Shared decision “allowing you to come back to another
31                consultation…go away and thinking. You have to be sure it’s the
32                one you want”;
33            • 4. Patient choice: patients choose for themselves, “information
34                should be provided in such a way as not pushed into it”
35        • The majority of patients using splints for hand/wrist RA indicated that
36          their splint use was dependent on the seriousness of the symptoms.
37          They wore them in order to reduce symptoms and to support and
38          immobilize the wrist. They stopped wearing the splint when they had
39          reduced functional abilities and when performing dirty or wet
40          activities.
41
42   3. Correlations between demographics, disease characteristics and
43   disease measures (2 studies118,121)
44   Observational studies found that:
45        • RA knowledge and the need for information was NS different
46           between patients with long (>10 years) and short (<1 year) disease
47           duration.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               66 of 307
     DRAFT FOR CONSULTATION

1         •   High fatigue, pain, greater functional disability and low acceptance
2             were predictors of high anxiety and depression and lower life
3             satisfaction. Psychological well-being was correlated with optimism,
4             pessimism and perceived stress but not with healthcare social
5             support, active cognitive and behavioural coping, disease duration,
6             inflammation, antidepressant use and presence of comorbidity.
7             General social support significantly correlated with lower depression
8             and greater life satisfaction.
 9        •   Among patients with lower stress, there was little effect of active
10            behavioural coping on depression and life satisfaction. However,
11            among those with higher stress, engaging in active behavioural
12            coping was related to lower depression and greater life satisfaction.
13

14   5.1.5 Summary of evidence statements
15   Observational and qualitative studies found that patients with RA:
16     • often experienced a lack of information (especially AEs of medication,
17         other options for treatment, aids and devices and home
18         adjustments);94,95,99,101,102,wanted more involvement in decisions about
19         their managemen101,102 .
20     • many patients still had unmet healthcare needs and wanted
21         improvements within each area of the multidisciplinary team97, 98,99,101,
22     • areas of their lives most affected and where patients wanted
23         improvement were: pain, function, ADLs and hobbies,94,97, 100
24     • QoL and disease status measures were often worse in women than in
25         men89. 97,
26     • priorities for younger patients were pain, work (please see
27         Occupational Therapy section in Chapter 6) and anxiety and
28         depression. 97,
29     • The main priority for older patients was function. 97,
       •
30         pain, fatigue, depression, loss of function, loss of valued activities,
31         interference with ADLs and inability to work were problems for many
32         patients with RA87,88,90, 94,95,98, 97,103
       •
33         poor psychological status was often associated with poorer outcomes
34         of symptoms, function and QoL65, 96, 104,106, 118
35

36   5.1.6 From evidence to recommendations
37   The GDG noted that the evidence highlighted a number of problems faced by
38   people with RA that are often not acknowledged satisfactorily by healthcare
39   professionals. These include pain and fatigue, depression, mobility, inability to
40   work or undertake leisure and social activities, and impact on sexual
41   relationships. Many of these aspects and complications of their disease would
42   best be discussed with appropriate members of the multidisciplinary team
43   (see Section 6.1), and there should also be an opportunity to raise any of
44   these issues at annual review (see Section 8.2). However, the GDG felt that it
45   was pragmatic to have a specific recommendation for patients ensuring that
46   periodic assessment of their disease encompassed these important factors,

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               67 of 307
     DRAFT FOR CONSULTATION

 1   and this should form part of a recommendation relating to the multidisciplinary
 2   team (see Chapter 6 on multidisciplinary teams)
 3
 4   The GDG also noted the evidence that patients want more involvement in the
 5   management of their disease. Because of the long-term nature of rheumatoid
 6   arthritis and (from the patient’s point of view) changing priorities over time, the
 7   GDG felt that regular discussions with patients about all aspects of their
 8   treatment (including the advantages and disadvantages of different
 9   therapeutic options) was advisable. This view was strongly advocated by
10   patient representatives who also emphasised the crucial need to allow
11   sufficient time for these discussions with healthcare professionals in ways that
12   could be easily understood and acknowledged the importance of patient
13   autonomy in decision making.
14

15   5.1.7 Recommendations
16   R 7 People with RA should be offered the opportunity to have all aspects of
17   their care explained, discussed and agreed with them throughout the course
18   of their disease, communicating the risks and benefits of various treatment
19   options in ways that can be easily understood, and respecting the decisions
20   that they make.

21   5.2    Patient education

22   5.2.1 Clinical introduction
23   In helping people with RA participate fully in decision making about treatment,
24   it is highly appropriate to provide information and education about their
25   disease, treatment and how they might help themselves in managing their
26   disease. Quality information giving is required by all health professionals, as
27   part of their codes of conduct, to enable people to understand their condition
28   and make decisions about treatment. This is usually provided informally on a
29   one-to-one basis in clinic, with educational materials (eg leaflets, books,
30   DVDs, arthritis websites). Patient education goes beyond improving
31   knowledge and is “a planned, organised learning experience designed to
32   facilitate voluntary adoption of behaviours and/ or beliefs conducive to
33   health”.125 It additionally focuses on enabling people to effectively self-
34   manage, i.e. “monitor one’s condition and effect the cognitive, behavioural and
35   emotional responses necessary to maintain a satisfactory quality of life”.126
36   through use of educational, motivational and behavioural techniques. This can
37   be provided 1:1, through self-study or computer based interventions or in
38   formal organised group sessions led by rheumatology health professionals or
39   trained lay leaders with arthritis or other chronic conditions. Different formats
40   may be used: an educational approach of lecture/ discussion sessions to
41   increase knowledge, satisfaction and reduce concerns; or a behavioural, also
42   termed psychoeducational, approach, including regular skills practice, goal
43   setting, contracting and use of home programmes to facilitate behavioural
44   change
45


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 68 of 307
     DRAFT FOR CONSULTATION

 1   Information giving and patient education programmes can be time consuming
 2   and it is therefore important to consider the benefits of such interventions. A
 3   balance needs to be struck between efforts invested by people with RA and
 4   health professionals, and the benefits gained. What is the evidence that
 5   education exerts an impact on symptoms, disease progression, function and
 6   quality of life? In particular:
 7
 8      1. Having been provided with information, do people with RA retain the
 9         information provided and find this of benefit?
10      2. Having undergone a patient education programme (1:1 or group) does
11         this lead to sustained benefits or is improvement only short term?
12      3. Are there stages during a person’s disease in which different
13         educational interventions are more likely to be successful? Is patient
14         education more likely to be helpful if it is provided in early than
15         established disease?
16      4. Are there any methods of delivering patient education that are more
17         successful than others?
18
19   It was decided to exclude formal cognitive behavioural therapy as being
20   beyond the scope of this question, but educational programmes using a
21   behavioural approach are considered. Please note that NICE has published a
22   technology appraisal on “Computerised cognitive behaviour therapy for
23   depression and anxiety”.127

24   5.2.2 Clinical methodological introduction
25   We looked for studies that investigated the benefits and harms of different
26   patient information provision and/or educational methods and/or different
27   patient education or self-management programmes, with respect to
28   symptoms, joint damage, function and quality of life in patients with a recent
29   onset of RA or in patients with established RA. Due to the large volume of
30   evidence, only RCTs were selected which were published from 1997 onwards,
31   had a sample size of N≥50, had a UK-relevant population and if this was
32   mixed arthritis there had to be >75% RA or RA subgroup analysis. Trials were
33   also selected which compared the following:
34          a) education/self-management vs other education/self-management
35            methods
36          b) education/self-management vs usual care
37
38   NOTE: Due to the inclusion/exclusion criteria for studies, some useful
39   evidence was not included in this review, namely the long-term UK Arthritis
40   Self-management Programme128,129.
41
42   Mixed population (recent onset and established RA):
43
44   It was decided that for this question, due to the large amount of evidence on
45   education and self-management programmes, it would be useful to include a
46   MA even though it did not separate the RA population into recent onset and
47   established disease.
48
49   One Cochrane SR/MA130 was found that fulfilled the criteria and focused on
50   RCTs which compared patient education interventions (that included an
     Rheumatoid arthritis: full guideline DRAFT (August 2008)              69 of 307
     DRAFT FOR CONSULTATION

1    instructional component) vs a no intervention control group in patients with
2    recent onset and established disease. The ‘no intervention’ control group did
3    receive some intervention as they interacted with the healthcare professional
4    making their assessments. The MA itself was well conducted, however, the 31
5    RCTs it included were of varying quality and differed with respect to the
6    following:
7
8         •   Blinding (N=7 RCTs double blind; N=20 RCTs single blind; N=23
9             RCTs no blinding)
10        •   Study size (range N=18 to N=1140)
11        •   Study quality – max score of 8 [N=21 studies reasonable to good
12            quality (1+ and 1++); N=29 poor quality(1-)]
13        •   Study duration – length of intervention (range: 7 hours to 15 months)
14        •   Study duration – length of follow-up (range: 8 days to 18 months)
15
16   NOTE: Because the Cochrane MA pooled together papers which had patients
17   of both early and established RA, it was decided that duplicate papers (ie.
18   papers picked up in our search which were already included in the MA) would
19   still be included as evidence in this section for either ‘recent onset’ or
20   ‘established’ RA in order to tease out any effects on these sub-populations.
21   Results reported here from the MA, were pooled from the N=17 high quality
22   RCTs only (data from low quality studies was excluded).
23
24   Recent onset of RA:
25
26   Four RCTs131-135 were found that fulfilled the criteria. One of these RCTs was
27   published as two separate papers133,134 reporting different outcomes and so
28   the trial has only been counted once, however results from both papers are
29   reported and referenced here. All trials were methodologically sound.
30
31   All four RCTs were single-blind, parallel group studies, but they differed with
32   respect to the following:
33
34        •   Sample size (range: N=64 to N=326)
35        •   Trial length (range from 3 months to 4 years post-intervention)
36        •   Treatment (1 RCT131 Education programme + DMARDs vs Education
37            leaflet + DMARDs; 1 RCT132 Standard education programme vs
38            Cognitive-behavioural education programme; 1 RCT133,134 standard
39            education programme vs joint protection education programme; 1
40            RCT135self-management OT programme vs usual care)
41
42   Established RA:
43
44   9 RCTs136-144 were found that fulfilled the criteria. Three of these
45   RCTs136,139,141 were excluded as evidence due to methodological limitations.
46   The rest of the trials were methodologically sound.
47

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               70 of 307
     DRAFT FOR CONSULTATION

1    The 6 included RCTs137,138,140,142-144 were all parallel group studies, but they
2    differed with respect to the following:
3
4         •   Sample size (range: N=59 to N=363)
5         •   Blinding (4 RCTs single blind, 2 RCTs unblinded/blinding not
6             mentioned)
7         •   Trial length (range: 4 weeks to 1 year intervention; follow-up ranged
8             from immediate to 9 months post-intervention)
 9        •   Treatment (1 RCT137 standard education programme vs standard
10            care; 1 RCT138 Education programme + leaflet vs standard care +
11            leaflet; 1 RCT140 Group education with significant other vs patient
12            only group education vs control self-help guide; 1 RCT142 spouse-
13            included self-management programme + usual care vs patient only
14            education programme + usual care; 1 RCT143 ACR RA leaflets + mind
15            map vs ARC RA leaflets; 1 RCT144 Education programme +
16            information leaflet vs usual medical care + information leaflet)

17   5.2.3 Health economic methodological introduction
18   No health economic studies were appraised

19   5.2.4 Evidence Statements
20   1. Mixed population (recent onset and established RA)
21
     Study      Treatmen      Follow-       Outcomes                        Result -
                t             up                                            best
                                                                            treatment
     Cochra     Patient       1st and       Joint counts, Anxiety, ESR, CRP NS (all
     ne         education     final                                         p<0.05)
     MA130      vs no         follow-up
     Level      interventio   1st follow-   Disability, Patient global         Education
     1++        n             up            assessment, Psychological
                                            status and depression
                              Final         Disability, Patient global         NS
                              follow-up     assessment, Psychological
                                            status and depression
22
23   2. Recent onset of RA
24
25   a) Education/self-management vs usual care
               135
26   • 1 RCT       found that the self-management/OT programme was significantly
27      better than the control (no intervention) groups for: Use of some self-
28      management methods (particularly hand and arm exercises, joint
29      protection and rest); Receipt of a working splint and a resting splint;
30      Owning and use of assistive devices. However there was NS difference
31      for: DAS28, HAQ, AIMS2 scores and for self-efficacy (ASES) score. Level
32      1++
33

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 71 of 307
    DRAFT FOR CONSULTATION

1   b) Special education (CBT or joint protection) vs standard education
2   programmes
3
    Study      Treatme     Outcomes                             Follow-up      Result -
               nt                                                              best
                                                                               treatment
    1          CBT         morning stiffness; Pain (VAS);       3 months       NS
    RCT132     education   tender and swollen joints; ESR       post
    Level      program                                          interventio
    1+         me vs                                            n
               standard    AIMS2 physical function and          3 months       CBT
               education   affect; RAI arthritis helplessness   post-          (p=0.009
               program     subscale;                                           and
               me                                                              p=0.01;
                                                                6 months       p=0.003)
                                                                post-
                                                                interventio NS
                                                                n
                          AIMS2 subscales of current            6 months    NS
                          health and symptoms); RAI             post-
                          arthritis internality subscale;       interventio
                          Total self-efficacy scale             n
    1          Joint      Pain (VAS); Tender and swollen        12 months NS
    RCT133,1   protection joints; Joint damage (numbers         and 4
    34
               education of deformities); Grip strength;        years-post-
    Level      program    Hand joint alignment and              interventio
    1+         me vs      motion; AIMS2 scores; ASE             n
               standard dimensions; Number of patients
               education taking RA medication and
               program    participating in OT
               me         Early morning stiffness; AIMS2                       CBT
                          dimension of ADLs; Joint                             (p=0.01
                          Protection behaviour                                 and
                          assessment; Number of visits to                      p<0.05;
                          doctor in previous 6 months                          both:
                                                                               p=0.04;
                                                                               p=0.001;
                                                                               p<0.01)
                           Hand Pain (VAS); Assessor’s          12 months      CBT
                           and Patient’s global disease                        (p=0.02;
                           status; number of disease flare-                    p=0.003
                           ups; Number of patients                             and
                           participating in physiotherapy       4 years-       p=0.03;p=0
                                                                post-          .004;
                                                                interventio    p=0.005)
                                                                n
                                                                               NS
4
5   c) Education programme vs education leaflet
             131
6   • 1 RCT      found that there was NS difference between education
7      programme + DMARD and the Education leaflet + DMARD for DAS score,
    Rheumatoid arthritis: full guideline DRAFT (August 2008)                  72 of 307
    DRAFT FOR CONSULTATION

1      M-HAQ score and Range of Motion - shoulders, elbows and knees, AIMS
2      subscales, CRP, compliance with activities and joint protection (but
3      Education significantly better for compliance with energy conservation).
4      Level 1+
5
6   3. Established RA
7
8   a) Education programme vs standard care
9
    Study     Treatment        Follow-    Outcomes                           Result -
                               up                                            best
                                                                             treatment
    1         Education        24 weeks   Total number of withdrawals        Similar
    RCT138    programme                   Pain scores; Joint                 NS
    Level     vs standard                 inflammation (Articular Index);
    1+        care                        Morning stiffness; CRP
    1         Education        1 year     Morning stiffness; DAS28        NS
    RCT144    programme +
    Level     leaflet and
    1++       usual medical
              care + leaflet
    1         education        8 months   Pain (VAS) and HAQ; AIMS 2         Education
    RCT145    programme                   dimensions of physical,            (p<0.001;
    Level     (joint                      symptoms and social                p<0.05)
    1+        protection) +               interaction
              drug                        RAI, AIMS 2 dimensions of          NS
              treatment                   psychological and work
              (IFX) vs drug
              treatment
              (IFX)
    1         Education        1 year     PKQ - patient knowledge,           Education
    RCT137    programme                   p=0.0002
    Level     vs standard                 Improvements in Larsen             NS
    1+        care                        scores, HAQ, SF-36
                                          dimensions of social
                                          functioning and general
                                          health perception, RAI,
                                          compliance
    1         Education        1 year     Coping; QoL (EMIR)                 Education
    RCT144    programme +                 symptomatic dimension;             (p=0.03;
    Level     leaflet and                 Patient satisfaction;              p<0.0001;
    1++       the Usual                   Knowledge                          p=0.02)
              medical care                Physical activity (Baecke          NS
              + leaflet                   questionnaire – sports activity
                                          and hobbies), Behavioural
                                          changes, Nocturnal
                                          awakening, HAQ (QoL),
                                          HADS anxiety and
                                          depression, QoL (EMIR
                                          dimensions); Fatigue (FACIT-
                                          F)
    Rheumatoid arthritis: full guideline DRAFT (August 2008)                73 of 307
     DRAFT FOR CONSULTATION

1
2         •   1 RCT145 found that 75% of patients found the education programme
3             very useful and only 8% found it not useful at all for ADLs.
4    b) Education / self-management programme (patient + significant other)
5    vs Education programme (patient only)
6
     Study     Treatment     Follow-up     Outcomes                              Result -
                                                                                 best
                                                                                 treatme
                                                                                 nt
     1         Spouse        2 weeks       Increased Communication               Spouse
     RCT142    Included      and 6                                               + patient
     Level     self-         months                                              educatio
     1+        manageme      post-                                               n
               nt            interventio                                         (p<0.00
               programm      n                                                   1)
               e + usual                   Disease activity; DAS28 score;        NS or
               treatment                   DAS score; IRGL dimensions            similar
               vs the                      (mobility, dexterity and pain);
               Patient                     Psychological functioning (IRGL
               only                        dimensions); Disease stressors:
               education                   Pain, limitations and dependence
               programm                    (CORS); Coping - decreasing
               e + usual                   activity; Marital satisfaction
               treatment                   (MMQ); Social support (IRGL
                                           dimensions); Spousal Criticism;
                                           total number of withdrawals
     1         Patient +     12 months     Fatigue                               Educatio
     RCT140    significant   (3 months                                           n for
     Level     other         post-                                               patients
     1++       education     interventio                                         only
               programm      n)                                                  (p=0.00
               e vs                                                              1)
               patient                     Disease activity; DAS28 score;        NS
               only                        DAS score; All Self-efficacy and
               education                   health behaviour measures;
               programm                    Effects on social interactions; Use
               e                           of self-management activities and
                                           active coping strategies
 7
 8   c) Education programme vs education leaflet
               140
 9   • 1 RCT       found that there was NS difference between patient + significant
10      other education programme and education leaflet for DAS28 score, Self-
11      efficacy measures, all Health behaviour measures; Use of self-
12      management activities; degree to which people use active coping
13      strategies - Dutch Coping with Rheumatoid Stressors). However education
14      programme with significant other programme was significantly worse for
15      fatigue p=0.04). Level 1++
16
17   d) ARC booklet + mind map vs ARC booklet

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              74 of 307
     DRAFT FOR CONSULTATION

1    •   1 RCT143 found that there was NS difference between the ARC booklet +
2        mind map vs ARC booklet groups for: Increase in knowledge.
3    •   The same RCT143 found that poor reading ability leads to poor knowledge
4        which is associated with more anxiety and depression. Level 1+

5    5.2.5 Summary of evidence statements
 6   The Cochrane meta-analysis 130 showed patient education had small, short-
 7   term effects on disability, joint counts, patient global assessment,
 8   psychological status and depression. There was no evidence of long-term
 9   benefits. Education groups generally do not show benefit compared with
10   controls. One RCT showed no difference between an education programme
11   and education leaflets for a range of outcomes 131. Another RCT showed that
12   a one-to-one education programme compared with standard care only
13   resulted in improved patient knowledge and adherence to treatment, but no
14   difference for a range of other outcomes 138 137. There was some evidence of
15   short-term benefits in hand pain, number of disease flare-ups and global
16   assessments, joint protection behaviour and a long-term (4 years) decrease in
17   morning stiffness, and decreased visits to the GP, though there was no
18   change in use of conventional drugs 133,134. The same studies showed less
19   development of some hand deformities, and although there was no functional
20   hand improvement 133,134, some short and long-term improvements in aspects
21   of activities of daily living were seen 132 133,134. One RCT135 found that the self-
22   management/OT programme was significantly better than the control (no
23   intervention) groups for use of some self-management methods and for
24   receipt and use of splints and other assistive devices.
25
26   Cognitive behavioural approaches seemed to have short-term benefits for
27   mood and decreased a sense of helplessness 132, but other trials showed no
28   improvement 133,134. Attending a programme with a carer or “significant other”
29   did not improve outcomes on the whole 142 140

30   5.2.6 From evidence to recommendations
31   The GDG noted that there was an enormous amount of information from
32   which it was difficult to tease out precisely what was, and what was not,
33   effective. There was no doubt that patients want to be given clear
34   explanations during the course of their disease, and that they wish to receive
35   this in both written and verbal forms. Although there was some evidence that
36   in established disease written information can improve some outcomes, no
37   studies were identified which evaluated written information in early disease.
38   There have been very few evaluations of structured 1:1 patient education
39   programmes, and the GDG felt that this was one of a number of areas where
40   further studies were needed. Nevertheless, in view of the clear wishes of
41   patients to receive both written and verbal information about their condition,
42   the GDG felt that it was appropriate to make a recommendation to this effect.
43
44   A recurring feature of many studies was the demonstration of short-term, but
45   not long-term, benefits. The GDG felt that this was an area where it would be
46   sensible to make a research recommendation about the provision of refresher
47   courses which might overcome this problem; these would also enable patients

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  75 of 307
     DRAFT FOR CONSULTATION

 1   who do not wish to receive information early on, the opportunity to obtain this
 2   at a later date.
 3
 4   The GDG noted that the provision of programmes for patients that encourage
 5   self-management is now considered a fundamental aspect of care for all long-
 6   term conditions, and that there was evidence that group education, led by
 7   health care professionals using a behavioural approach, was effective in this
 8   regard for people with RA. Most studies of group programmes led by trained
 9   lay people using a behavioural approach, such as the Arthritis Self
10   Management Programme and Chronic Disease Self Management Programme
11   have been with community-based volunteers with a range of rheumatological
12   diagnoses. There was therefore currently insufficient evidence to justify these
13   programmes specifically for RA (although an appropriate research
14   recommendation could be made), but the generic concept of attending lay-led
15   self management programmes was supported.
16
17   In summary, the GDG noted the lack of clearcut evidence in many areas, the
18   need for further research, and the clear desire of patients to have access to a
19   wide range of educational activities. It was accordingly felt appropriate to
20   make a rather general recommendation supporting a range of activities that
21   might already be available until such time as further studies (including cost
22   effectiveness data) have identified the most appropriate educational methods.

23   5.2.7 Recommendations
24   R 8 Accurate and understandable written and verbal information should be
25   offered to all people with early and established rheumatoid arthritis in order to:
26         • enhance understanding of their disease and its management
27        •   counter any misconceptions that they may have.
28
29   R 9 All people with RA who wish to know more about their disease and its
30   management should be offered the opportunity to engage in any existing
31   educational activities, including self-management programmes, until further
32   research has identified the most effective methods of acquiring this
33   information
34




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                76 of 307
     DRAFT FOR CONSULTATION


1    6      The multidisciplinary team

2    6.1    The multidisciplinary team

3    6.1.1 Clinical introduction
 4   A multidisciplinary team approach (incorporating various healthcare
 5   professions such as specialist nurses, physiotherapists, occupational
 6   therapists and podiatrists) is often used in the management of patients with
 7   rheumatoid arthritis. The composition of the team in any individual centre will
 8   vary, but emphasis should be placed on the tasks required to care for the
 9   individual patient's needs, with the aim of minimising the impact of the
10   disease. This combined approach brings together the skills and knowledge of
11   all team members, for both the assessment and management of disease but
12   requires a high level of communication and cooperation. The patient can often
13   be an active member of the team, in order to address and manage all aspects
14   of care. Although patients may approach any member of the team in the
15   knowledge that their problem will be appropriately dealt with, in practice one
16   health care professional usually coordinates patient care to ensure all needs
17   are addressed and managed with optimum use of the team expertise.
18
19   This chapter of the guideline assesses the evidence for the effectiveness of all
20   of the interventions traditionally delivered by individual members of the
21   multidisciplinary team. Details of the specific interventions offered by
22   physiotherapists, occupational therapists, and podiatrists are summarised in
23   the relevant sections. The role of the rheumatology nurse specialist is
24   summarised below.
25
26   The Rheumatology Nurse Specialist
27   The rheumatology nurse specialist is an integral part of the multi disciplinary
28   team caring for patients with rheumatoid arthritis. Although the role may vary
29   depending on the needs of the service, most nurses will have clinical,
30   educational and advisory aspects as part of their work. Rheumatology nurse
31   specialists may also be involved in research, management and strategic
32   development. However the key feature of these posts is to ensure the
33   physical, emotional and social wellbeing of people with rheumatoid arthritis
34   using a patient centred approach.
35
36   Clinical
37   Most rheumatology nurse specialists will run nurse-led clinics where they
38   assess and treat patients working in close collaboration with the consultant
39   rheumatologist. This will usually involve examination of the joints, assessment
40   of the patients overall condition and wellbeing, monitoring blood results of
41   patients on immunosuppressive drug treatments and requesting and
42   reviewing investigations. Some rheumatology nurse specialists will perform
43   joint injections, prescribe and alter treatments and carry out other therapies
44   such as infusions. The rheumatology nurse specialist is usually the
45   professional who is the main co-ordinator of care within the multi disciplinary
46   team, referring to team members and other agencies as required.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               77 of 307
     DRAFT FOR CONSULTATION

 1
 2   Education, support and advice
 3   The diagnosis of rheumatoid arthritis can be devastating and in conjunction
 4   with the team members the rheumatology nurse specialist will be able to offer
 5   support and advice to help reduce the fear and anxiety such a diagnosis can
 6   bring. Detailed explanations of diagnosis, drug therapies and other managerial
 7   strategies and therapies can be provided by the rheumatology nurse specialist
 8   at the onset and throughout the course of the condition. Due to the impact of
 9   the disease process and concerns over therapies instigated, a key element of
10   the rheumatology nurse specialist role is to provide an easily accessible point
11   of contact for patients and their carers. Most rheumatology nurse specialists
12   will provide this service through a telephone helpline. In collaboration with the
13   multi disciplinary team many rheumatology nurse specialists have developed
14   or assist with patient self management groups in order to empower patients
15   and help them develop coping strategies.
16
17   Education is an important element of the role and most rheumatology nurse
18   specialists act as a resource for their colleagues and the wider healthcare
19   community. Some will be involved in the development and running of
20   rheumatology educational programmes for healthcare professionals. These
21   may range from study days to formal diploma, degree and MSc courses in
22   association with a university in order to increase knowledge and expertise in
23   caring for patients with rheumatoid arthritis.
24
25   This section specifically addresses the role of the multidisciplinary team as a
26   whole.

27   6.1.2 Clinical methodological introduction
28   We looked for studies that investigated the benefits and harms of
29   multidisciplinary teams on patients with RA (recent onset and established
30   disease). Selected trials were of a UK-relevant population and if this was of
31   mixed arthritis there had to be >75% RA or RA subgroup analysis.
32
33   Five RCTs146-151 and three case-series’152-154 were found that fulfilled the
34   inclusion criteria. One of the RCTs was published as two separate
35   papers147,148 reporting different follow-up times and so the trial has only been
36   counted once. However, results from both papers are reported and referenced
37   here. Two of the RCTs150,151 were excluded as evidence due to
38   methodological limitations and all other trials were methodologically sound.
39
40   Recent onset RA
41   2 case-series’153,154 of N=110 and N=70 patients respectively, evaluated the
42   effects of a multidisciplinary team (MDT) care programme on patients with a
43   recent-onset of RA. Prier et al153 had a 3 month treatment phase whilst
44   Nordmark et al154 had a 2 yr treatment phase
45
46   Established RA
47   3 RCTs146-149 and one case-series’152 were found for patients with established
48   RA.
49

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               78 of 307
     DRAFT FOR CONSULTATION

1    The 3 RCTs compared MDT care to either non team care (N=59) (Ahlmen)146,
2    routine out-patient care (N=68) (Vliet Vlieland, Vljeland)147,148 or waiting list
3    control (N=68) (Scholten)149. All had a one-year treatment phase with follow
4    up at two years (Vlijeland) and 5 years (Scholten).
5
6    The case-series152 looked at the effects of a MDT care programme on N=92
7    patients and had a 3 week treatment phase with follow-up at 3 months.

8    6.1.3 Health economic methodological introduction
 9   One study was identified and appraised. Van den Hout et al155 is Dutch cost-
10   utility study comparing Clinical nurse specialist care, inpatient multidisciplinary
11   team care and day patient multidisciplinary team care. Two papers156,157 were
12   excluded for not being specific to a multidisciplinary team.

13   6.1.4 Clinical evidence statements
14   Recent onset RA
15
16   Symptoms, Function and QoL
                     153
17   • 1 case-series     found that MDT care programme led to a NS change in
18     QoL (AIMS) at 3 months follow-up. Level 3
                     154
19   • 1 case-series     found that after 2 years of MDT care, all patients
20     experienced significant decreases in pain except for those who stopped
21     working and were receiving sickness benefit. Level 3

22   Knowledge and satisfaction
                     153
23   • 1 case-series     found that patients who attended the MDT care
24     programme had a significant increase in knowledge of RA at the 3 month
25     follow-up. Level 3
26
27   Use of RA treatments
                     154
28   • 1 case-series     found that patients who attended the MDT care
29     programme had no change in the number of patients receiving DMARDs;
30     the number of patients receiving MTX or combination therapy with MTX
31     increased from, 8% to 41% at 2 years follow-up (end of team treatment
32     programme) as did the number receiving corticosteroids (increase from
33     11% to 15%). Level 3
34
35   Established RA
36
37   Symptoms
                    146-148
38   • Two RCTs             found that there was NS difference for RAI at 12
39            146
       weeks , 52 weeks and 104 weeks147,148 . Also NS were self-rated
40     physical discomforts (Body Symptoms Scale, BSS), and LAI (Lansbury
41     Articular Index – joints painful on pressure or motion) at 12 weeks 146 and
42     pain (VAS), number of swollen joints, fatigue and grip strength at 12
43     weeks, 52 weeks, 104 weeks147,148; ACR20 at 104 weeks147,148. Level 1+
              147,148
44   • 1 RCT          found that in-patient MDT care was significantly better than
45     routine out-patient care for: morning stiffness at 12 weeks and ACR20 at
46     12 weeks and 52 weeks (all p<0.05). Level 1+

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 79 of 307
     DRAFT FOR CONSULTATION

1         •   1 case-series152 found that patients who attended the MDT care
2             programme had significant improvements in: DAS, pain (VAS),
3             swollen joints, and RAI at 3 months follow-up and 26% and 52% of
4             patients fulfilled the ACR20 and EULAR response criteria. Level 3
5
 6   Function
                147,148
 7   • 1 RCT            found that there was NS difference between in-patient MDT
 8      and routine out-patient care for: HAQ at 12 weeks, 52 weeks and 104
 9      weeks. Level 1+
                 149
10   • 1 RCT         found that MDT care programme significantly improved HAQ
11      disability score at 52 weeks and at 5 years (both p<0.0001). Level 1+
                         152
12   • 1 case-series         found that patients who attended the MDT care
13      programme had significant improvements in: HAQ and SOFI at 3 months
14      and 26% and 52% of patients fulfilled the ACR20 and EULAR response
15      criteria respectively. Level 3
16
17   Global assessment
               147,148
18   • 1 RCT           found that in-patient MDT treatment was significantly better
19      for patient’s global assessment of disease activity at 4 weeks, 12 weeks
20      and 52 weeks, however there was NS difference at 104 weeks. Level 1+
                        152
21   • 1 case-series        found that patients who attended the multidisciplinary care
22      programme had significant improvements in: patient’s and physician’s
23      global assessment of disease activity at 3 months follow-up. Level 3
24
25   QoL
26   • 1 RCT146 found that the MDT care group was significantly better for:
27     overall health (Sickness Impact Profile – SIP) but NS for MACL (mood)
28     scores at 12 weeks. Level 1+
               147,148
29   • 1 RCT           found that in-patient MDT treatment was significantly better
30     than routine out-patient care for: anxiety at 12 weeks (but NS 52 weeks
31     and 104 weeks); Depression at 12 weeks (but NS at 52 weeks and 104
32     weeks). Level 1+
                149
33   • 1 RCT        found that MDT care programme significantly improved coping
34     with illness (FQCI) (p<0.01) and Beck depression score at 52 weeks
35     (p<0.001). Level 1+
36
37   Biochemical markers
38        • 2 RCTs146 147,148 found that there was NS difference between the two
39          groups for CRP at 12 weeks146 and ESR and CRP at 12 weeks, 52
40          weeks and 104 weeks147,148. Level 1+
41        •   1 case-series152 found that patients who attended the MDT care
42            programme had significant improvements in ESR at 3 months follow-
43            up. Level 3
44
45   Use of medication / treatment
               146
46   • 1 RCT       found that there was NS difference between the two groups for:
47      use of medication (DMARDs, NSAIDs and Corticosteroids) at 12 weeks.
48      Level 1+


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                80 of 307
     DRAFT FOR CONSULTATION

1    •   1 RCT 149 found that MDT care programme significantly improved the use
2        of joint protection devices, regular relaxation exercises and regular
3        remedial gymnastics at 52 weeks. Level 1+

4    6.1.5 Summary of evidence statements
 5   In recent onset RA the roles for the multidisciplinary team include:
 6          Recognising the importance and efficacy of work and social roles and
 7          ensuring that patient can fulfil these154.
 8          Usually being responsible for patient education153
 9          can ensure that patients have access to a wide range of professionals
10          with different knowledge and skills153
11          can monitor the disease and alter DMARDs appropriately to maintain
12          good control154
13
14   In established RA the multidisciplinary team:
                                                                              147,148,152
15   • Can be helpful in monitoring and ensuring good disease control                      ,
                                                                                        146-
16       and improvement in mood and coping with the disease in the short-term
         149
17           , but there is a lack of evidence to show whether or not there is any
18       benefit in the long-term147,148
                                 149,152
19   • May influence function            , though the evidence is contradictory 147,148

20   6.1.6 From evidence to recommendations
21   Despite the lack of the proven long-term benefit the GDG recognised the
22   importance of the multidisciplinary team, and that it could provide a variety of
23   services to the patient with RA using knowledge and skills that complemented
24   those of the rheumatologist, addressing issues above and beyond the purely
25   medical problems in both early and established RA. It was also noted that the
26   opportunity for patients to raise any issues about their disease (a need
27   identified in the patient perceptions and beliefs section – see section 6.3)
28   could be most appropriately addressed by ensuring that patients were offered
29   ongoing access to the MDT.
30
31   The GDG supported the view of the patient perspective that non medically
32   qualified members of the MDT (eg specialist nurses) seemed to have more
33   time with patients in assessing disease activity and monitoring the impact of
34   drugs in treating their disease. There is a concern that in the absence of a
35   multidisciplinary team, individuals may work in isolation and therefore
36   duplication of activities might take place. Therefore the GDG felt it was
37   important to have one nominated member of MDT team nominated as the
38   person responsible for coordinating their care.

39   6.1.7 Recommendations
40   R 10 People with RA should be offered the opportunity for periodic
41   assessments of the effect of their disease on their lives*, by having ongoing



     *
       Such as pain, fatigue, activities of daily living, mobility, inability to work or undertake social or
     leisure activities, quality of life, mood, impact on sexual relationships.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                  81 of 307
     DRAFT FOR CONSULTATION

1    access to a multidisciplinary team to assess the impact, and treat the
2    consequences, of their condition†.
3
4    R 11 In order to ensure the optimal use of the multidisciplinary team, all
5    people with RA should have access to a named member of the team (e.g. the
6    specialist nurse) who has been designated as the person responsible for
7    coordinating the various aspects of their care

8    6.2      Physiotherapy

9    6.2.1 Clinical introduction
10   Despite the pharmacological advances for the treatment of RA many patients
11   still present with functional deficits which physiotherapy seeks to address.
12   Physiotherapy aims to reduce pain and stiffness, prevent deformity and
13   maximise function, independence and quality of life. These aims are
14   supported by a variety of interventions which may be conveniently grouped
15   under active and passive headings. In the former, education and exercise are
16   the key components and until recently, the fear of accelerating joint damage
17   has limited activities to range of movement and isometric exercises. However,
18   the past couple of decades has seen a paradigm shift towards dynamic
19   conditioning exercise, fuelled, in part, by studies which show that patients with
20   RA are inactive158 and at higher risk of cardiovascular disease and
21   osteoporotic fractures than the non-RA population159
22
23   Threefold increased risk of hip fractures with rheumatoid arthritis in Central
24   Finland.160.Consequently, patients with RA have reduced physical capacity
25   which occurs early in the course of disease161 Exercise strategies aim to
26   improve or maintain health-related fitness by enhancing biomechanical
27   efficiency without exacerbating joint damage. The benefits of physical activity
28   and exercise strategies are only apparent if the stimulus is sufficient, regular
29   and sustained, hence issues of concordance are as important as the activity
30   itself.
31
32   Passive treatments, success of which depends less on patient concordance,
33   include manual therapy techniques and the application of electrophysical
34   agents. These are used for specific clinical impairments on a time limited
35   basis with the aim of enhancing the ability to exercise or increase physical
36   activity. Manual therapy is “a clinical approach utilizing skilled, specific hands-
37   on techniques….” which includes joint and soft tissue mobilization and
38   manipulation.162 Electrophysical agents represent both thermotherapy
39   (superficial heat/cold) and electrotherapy and are used to reduce pain and
40   improve function. Electrotherapy includes Transcutaneous Electrical Nerve
41   Stimulation (TENS), ultrasound, pulsed short wave therapy (also called
42   Pulsed Electromagnetic Energy, PEME), interferential therapy (IFT) and laser.
43
44   Physiotherapy management of RA uses a multi-modal or comprehensive
45   approach which consists of a combination of education, exercise and pain
46   relief agents, with the emphasis shifting depending on clinical need and the

     †
         Please see recommendation 38
     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 82 of 307
     DRAFT FOR CONSULTATION

 1   goals agreed between therapist and patient. Encouraging self-management
 2   strategies and self-efficacy for sustained regular physical activity is an
 3   overarching theme and demands supportive and reinforcing health education
 4   and promotion.
 5
 6   In considering the effectiveness of physiotherapy we need to evaluate the
 7   evidence for physiotherapy in minimising the impact of disease on symptoms,
 8   joint damage, function and quality of life across the RA life span. In particular:
 9         1. Which exercise strategies offer the greatest benefit in respect of
10         symptoms, joint damage, function and quality of life?
11         2. Which of the various manual therapy strategies provide superior
12         symptomatic relief and functional gain and for what duration?
13         3. Which of the electrophysical agents offer the greatest therapeutic
14         effects in terms of symptomatic relief and functional gain and for what
15         duration?
16
17   According to the World Confederation of Physical Therapists Physiotherapy is
18   “concerned with identifying and maximizing movement potential, within the
19   spheres of promotion, prevention, treatment and rehabilitation, in partnership
20   with their clients”.163 The components of Physiotherapy interventions include:
21
22   Exercise therapy – on land and in water (hydrotherapy/aquatic physiotherapy)
23   and includes aerobic activities, flexibility and muscle strengthening exercises,
24   core stability exercise, balance rehabilitation, promotion of lifestyle physical
25   activity.
26
27   Patient education and self-management - joint protection strategies, energy
28   conservation/fatigue management, sleep hygiene training, management of
29   flare, pain relief strategies, relaxation training, exercise and physical activity
30   recommendations.
31
32   Thermotherapy - hot/cold packs, paraffin/wax baths, infrared,
33
34   Electro-physical agents - Transcutaneous Electrical Nerve Stimulation
35   (TENS), Ultrasound, Pulsed Electromagnetic Energy (PEME),Interferential
36   therapy (IFT) and Laser.
37
38   Provision and education of use of Assistive Devices – walking aids, splints,
39   orthoses, insoles
40
41   Manual therapy – includes mobilisation, manipulation, myofascial release,
42   trigger point therapy, acupuncture and massage.

43   6.2.2 Clinical methodological introduction
44   We looked for studies that investigated the efficacy of different aspects of
45   Physiotherapy (PT) with respect to symptoms, joint damage, function and
46   quality of life in patients with RA (recent onset and established disease). Due
47   to the large volume of evidence, only MA and RCTs were selected which were
48   of a UK-relevant population and intervention; if the population was mixed
49   arthritis there had to be >75% RA or RA subgroup analysis; and for trials

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  83 of 307
     DRAFT FOR CONSULTATION

 1   looking at TENS and exercise, the sample size had to be N≥50 (and for
 2   strengthening exercises, N>70).
 3
 4   Five SRs/MAs164-168 and 17 RCTs169-174,174-190 were found that fulfilled the
 5   inclusion criteria. Three of these RCTs were published as multiple papers,
 6   reporting different outcomes or time-points and so these trials have only been
 7   counted once, however results from all the papers are reported and
 8   referenced here. 3 RCTs171,184-187,191,192 were excluded as evidence due to
 9   methodological limitations (lack of blinding and no ITT analysis). All other trials
10   were methodologically sound. One of the RCTs189 on exercise was included
11   even though it had a sample size <70, because it was the only trial of exercise
12   looking at patients who had active RA.
13
14   All trials except one190 (patients with a recent onset of RA), were conducted
15   using patients with established RA, and the MAs used trials which were of a
16   mixed population (recent onset and established RA).
17
18   Mixed population (Recent onset and Established RA)
19
20   Five SRs/MAs 164-168 were found that fulfilled the criteria. All MAs were well-
21   conducted, however, the RCTs included in the analysis were of varying quality
22
23   The first SR/MA164 looked at Hot and cold therapy and included 7 RCTs in the
24   analysis, which were all of poor to moderate quality. However, the RCTs
25   differed with respect to:
26          • Patients (N=4 hospitalised, N=7 outpatients)
27          • Disease duration (ranging from duration 5 yrs or less to mean 14
28              yrs)
29          • Intervention (1 RCT each on ice therapy, paraffin bath plus
30              exercise, three different thermotherapy modalities (paraffin wax
31              bath, faradic bath and ultrasound), different temperatures of heat, 2
32              RCTs heat)
33          • Comparison group (Control, exercise, cryotherapy)
34          • Study size (range: N=14 to N=90)
35
36   The second SR/MA165 looked at laser therapy and included 6 RCTs in the
37   analysis which differed with respect to:
38           • Intervention – wavelength (1 RCT 633 nm, 1 RCT 850 nm, 1 RCT
39              820 nm, 1 RCT 830 nm, 1 RCT 820 nm, 1RCT 632.5 nm)
40           • Intervention – Output power (1 RCT 10mW, 1 RCT 940 mW,, 1
41              RCT 40mW, 1 RCT 21 mW, 1 RCT 15 mW, 1RCT 1mW)
42           • Comparison (5 RCTs placebo and 1 RCT contralateral joint)
43           • Study size (range: N=17 to N=72)
44           • Blinding (4 RCTs double blind, 1 RCT triple blind, 1 RCT partial)
45   The third SR/MA166 looked at TENS therapy and included 3 RCTs (which were
46   all of reasonable to good quality) in the analysis but differed with respect to:
47           • Intervention (1 RCT 15 mins of 70 Hz, 1 RCT 20 mins of 100 Hz, 1
48              RCT 5 mins of 70 Hz)
49           • Comparison group (2 RCTs placebo, 1 RCT AL-TENS)
50           • Study size (range: N=19 to N=33)
51           • Blinding (1 RCT double blind, 1 RCT single blind, 1 RCT unblinded)

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                84 of 307
     DRAFT FOR CONSULTATION

 1           •    Follow-up (1 RCT 15 days, 2 RCTs not specified)
 2
 3   The fourth SR/MA167 looked at Ultrasound therapy and included 2 RCTs
 4   (which were of poor to moderate quality) in the analysis and had a 3 week
 5   treatment phase. However, the RCTs differed with respect to:
 6           • Intervention (1 RCT Ultrasound combined with either exercises,
 7              electric current, wax baths or electric current and exercises; 1 RCT
 8              ultrasound alone)
 9           • Comparison group (placebo ultrasound)
10           • Study size (1 RCT N=30, 1 RCT N=50)
11           • Blinding (1 double blind, 1 unblinded)
12   The fifth SR/MA168 looked at Tai-chi exercise therapy and included 4 RCTs
13   (which were of poor quality and unblinded) in the analysis, but differed with
14   respect to:
15           • Intervention (1 RCT health education + ROM Dance and relaxation;
16              1 RCT oral Shan Pi Tong + education + exercise + massage + hot
17              compress; 2 RCTs tai chi exercises)
18           • Comparison group (1 RCT oral Lei Gong; 1 RCT oral Shan Pi Tong;
19              2 RCTs no exercise)
20           • Study size (range N=28 to N=100)
21           • Study duration – length of intervention (range: 8 weeks to 10
22              weeks)
23
24   Established RA
25   The 13 included RCTs differed with respect to the following:
26         • Sample size (range: N=57 to N=310)
27         • Blinding (12 RCTs single blind, 1 RCT unblinded)
28         • Trial length (range: 3 weeks to 2 years; follow-up 1 RCT 6 months)
29         • Treatment (2 RCTs general PT; 11 RCTs exercise)
30
31   Recent onset of RA
32   The 1 RCT190 was a methodologically sound randomised, parallel group study
33   in N=228 patients. The trial compared 2 different treatment arms: Exercise
34   programme (healthy physical activity) vs usual care in a 1 year treatment
35   phase.

36   6.2.3 Health economic methodological introduction
37   One study was identified and appraised. Van den Hout et al193 is a Dutch o-
38   utility analysis comparing a long-term high intensity exercise program (RAPIT
39   program) to usual individual physical therapy.

40   6.2.4 Clinical evidence statements
41   Mixed population (Recent onset and Established RA)
42
43   1. HOT AND COLD THERAPY
44
     Study       Treatment     Follow-up    Outcomes                        Result -
                                                                            best
                                                                            treatment

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             85 of 307
    DRAFT FOR CONSULTATION

    1       Knee: Heat      72 hrs (end Pain measurement (amount            NS
    MA164   therapy (50°F   of          of morphine).
    Level   vs 60°F and     treatment)
    1++     50°F vs 70°F)
    1       Knee: Ice       5 days       Thermographic Index, joint         NS
    MA164   packs vs hot    (end of      circumference, Number of
    Level   packs           treatment)   patients preferring ice,
    1++                                  Number of patients with
                                         improved pain and stiffness
                                         grading
            Shoulder: Ice   3 weeks      McGill pain questionnaire,         NS
            packs vs hot    (end of      Flexion and Abduction ROM
            packs           treatment)

            Hand: wax       4 weeks      Change in: flexion and             Wax bath
            bath vs         (end of      extension of the dominant          (all p=0.04)
            control         treatment)   hand, pinch function, grip
                                         strength, pain on resisted and
                                         non-resisted motion, stiffness
                                         (both hands)
                                         Grip function                      NS
            Hand: Wax       4 weeks      Change in flexion and              Wax bath
            bath +          (end of      extension of the dominant          (all p=0.04)
            exercises vs    treatment)   hand; Change in grip strength
            exercises                    and function; Change in pain
                                         on resisted and non-resisted
                                         motion and for change in
                                         stiffness (both hands);
                                         Pinch function                     NS
            Hand: Wax       4 weeks      Change in: flexion and             Wax bath
            bath vs         (end of      extension of the dominant          (all p=0.04
            exercises       treatment)   hand, grip strength, grip and      except grip
                                         pinch function, stiffness (both    strength
                                         hands) and pain on resisted        p=0.008)
                                         and non-resisted motion
            Hand: wax      End of        Hand grip; PIP circumference;      NS
            therapy vs     treatment:    Articular index; Timed task;
            ultrasound     1, 2 and 3    Activity score; ROM (3 wks
                           weeks         only)
            Wax bath vs    End of        Hand grip, PIP circumference,      NS
            faradic bath + treatment:    Articular Index, timed task;
            ultrasound     1, 2 and 3
                           weeks
                           1 and 2       Activity score                     NS
                           weeks                                            SS
                           3 weeks
            Cryotherapy    End of        Change in post-surgery             NS
            vs control     treatment:    oedema
                           2, 3 and 4
                           days
1
    Rheumatoid arthritis: full guideline DRAFT (August 2008)               86 of 307
     DRAFT FOR CONSULTATION

1    2. LASER
2
     Study   Treatment        Follow-up    Outcomes                        Result -
                                                                           best
                                                                           treatment
     1       Laser vs         End of       Change in Pain (VAS); Pain,     Laser
     MA165   placebo          treatment    0 to 12 scale; Knee ROM (left (most
     Level                    (10 and 20   and overall); Morning stiffness p<0.01)
     1++                      weeks)       duration; Grip strength
                                           (mmHg –10 weeks, NS 20
                                           weeks) and ESR
                                           Pain (McGill); RAI; HAQ;        NS
                                           ROM (PIP, MCP, Right knee
                                           and ankle); Morning stiffness;
                                           RF+; Grip strength (kg);
                                           Swelling (right and left
                                           suprapatellar, MCP and PIP);
                                           Walking speed and CRP
 3
 4   The same MA165 found that there were no significant differences between
 5   laser and placebo according to: Methodological quality; Treatment duration
 6   (pain); Joint compared with nerve application (pain); Wavelength (pain).
 7   However, there was a significant difference in favour of dose: low dose laser
 8   therapy (≤3 J/cm²) compared with placebo but not high dose laser therapy
 9   compared with placebo (for change in pain (VAS) but NS dose effect for grip
10   strength). 1++
11
12   3. ULTRASOUND
13
     Study   Treatment        Follow-up    Outcomes                            Result -
                                                                               best
                                                                               treatment
     1       Ultrasound vs 10 weeks        change in: number of painful        Ultrasound
     MA167   placebo       (end of         articulations and swollen           (p<0.001)
     Level                 treatment)      articulations; dorsal flexion of
     1++                                   the wrist and grip strength
                                           change in: circumference of         NS
                                           PIP joints and duration of
                                           morning stiffness
             Hand:          end of         Hand grip ; PIP                     NS
             Ultrasound vs treatment       circumference; Articular
             wax            (1, 2 and 3    index; Timed task and Activity
                            weeks)         score
             Hand:          end of         Activity score                      Ultrasound
             Ultrasound vs treatment                                           (p<0.05)
             faradic bath + (1, 2 and 3
             ultrasound     weeks)
                            3 weeks        Hand grip; PIP circumference; NS
                                           Articular Index; Time task;
                                           ROM


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 87 of 307
    DRAFT FOR CONSULTATION

              Hand:           end of        Activity score                US +
              Ultrasound +    treatment                                   faradic
              Faradic bath    (1, 2 and 3                                 bath
              vs wax bath     weeks)                                      (p<0.05)
                              3 weeks       Hand grip; PIP circumference; NS
                                            Articular Index; Time task;
                                            ROM
1
2   4. TENS
3
    Study     Treatment       Follow-up     Outcomes                          Result -
                                                                              best
                                                                              treatment
    1         Hand: TENS      3 weeks       Change in resting pain (VAS)      TENS
    MA166     vs placebo      (end of                                         (p<0.00001
    Level                     treatment)                                      )
    1++                                     Change in grip pain               NS
              Hand: C-        end of        Change in joint tenderness        C-TENS
              TENS vs         treatment                                       (p=0.004)
              placebo         (same day) Resting pain (VAS); Grip pain        NS
                                         (VAS) and Tender joints
              Hand: C-        end of     Number of patients improved          NS
              TENS vs AL-     treatment
              TENS            (15 days)
4
5   5. EXERCISE
6
    Study     Treatment      Follow- Outcomes                                 Result -
                             up                                               best
                                                                              treatment
    1         Hand:          End of    change in: flexion and extension of    Exercise
    MA164     Exercise vs    treatme   the dominant hand, grip strength,      (all p=0.04
    Level     control        nt (4     grip and pinch function, stiffness     except
    1++                      weeks)    (both hands), pain on resisted and     flexion/exte
                                       non-resisted motion                    nsion
                                                                              p=0.004)
    1         Tai Chi        Range:    ROM - Ankle plantar flexion; lower     Tai-Chi
    MA168     exercises      8-10      extremity flexion; Withdrawals         (p=0.02;
    Level     vs control     weeks                                            p=0.004;
    1++                                                                       p=0.003)
                                       Functional Assessment; RAI;            NS
                                       Swollen joints; 50-foot walk; Grip
                                       strength; Patient global: number
                                       rated ‘recovery’ at 2 months and 3
                                       months; ROM – shoulder flexion,
                                       shoulder internal and external
                                       rotation and total upper extremity
                                       combined; Self-reported enjoyment
                                       – Benefit and Frequency
7
8   Established RA
    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 88 of 307
    DRAFT FOR CONSULTATION

1
2   1. GENERAL PT
3
    Study    Treatment        Follow- Outcomes                                  Result -
                              up                                                best
                                                                                treatment
    1       PT vs             6         Morning stiffness                       PT
    RCT172, Control           weeks     Pain (VAS), Grip strength, Tender       (p<0.036)
    173
            (waiting list)    (end of   joints, Stanford Self-Efficacy Scale    NS
    Level                     treatme
    1++                       nt)
                              52        Pain (VAS); Morning stiffness; Grip     PT (all
                              weeks     strength; Tender joint count (all       p<0.001
                                        p<0.001) and Stanford Self-             except
                                        Efficacy Scale and ADLs                 ADLs
                                                                                p<0.05)
    1                         3         RAI, Pain (VAS), ROM and ADL            PT (all
    RCT174                    weeks                                             p<0.005)
                                        ESR and Joint size                      NS
4
5   2. EXERCISE
6
7   a) Strengthening/mobilisation
8
    Study    Treatment        Follow- Outcomes                                  Result -
                              up                                                best
                                                                                treatment
    1        Joint        6             Dominant key grip; AIMS2 upper          Joint
    RCT175   protection + months        limb function, was worse for            protect +
    Level    strengtheni                number of drop-outs                     exerc
    1++      ng/mobilisat                                                       (p<0.01)
             ion exercise               AIMS (hand and finger function);        NS
             vs joint                   Jebsen-Taylor function score;
             protection                 Right index finger flexion;
                                        Dominant gross grip; Tender and
                                        swollen joint counts and Patient’s
                                        global assessment of disease
                                        activity
    1        Strengtheni      5         Quadriceps strength                     Exercise
    RCT169   ng exercise      weeks                                             (p<0.05)
    Level    (rehabilitati    (end of
    1+       on) vs           treatme
             Control          nt)
             (waiting list)   5         HAQ score                               NS;
                              weeks;                                            Exercise
                              6                                                 (p<0.05)
                              months
                              follow-
                              up



    Rheumatoid arthritis: full guideline DRAFT (August 2008)                   89 of 307
    DRAFT FOR CONSULTATION

                             5         Morning stiffness, Pain (VAS),            NS
                             weeks     Patient’s and Assessor’s global
                             and 6     assessment and Swollen and
                             months    Tender joints
                             follow-
                             up
    1        Intensive       24        ACR responders; muscle strength           Intensive
    RCT189   exercise        weeks     (isometric extension)                     Exercise
    Level    (rehabilitati                                                       (p<0.05)
    1+       on) vs                    Swollen joints, ESR, Pain (VAS),          NS
             Conservativ               DAS, Joint mobility, HAQ, 50 foot
             e exercises               walk time, Joint mobility (EPM-
                                       ROM).
1
2   b) Hydrotherapy
3
    Study    Treatment       Follow- Outcomes                                    Result -
                             up                                                  best
                                                                                 treatment
    1        Hydrothera      3         self-rated overall effect of              Hydrothera
    RCT176   py vs land      months    treatment                                 py
    Level    exercise                                                            (p<0.001)
    1+                                 EQ-5D utility and VAS; HAQ; Pain          NS
                                       (VAS) and 10 m walk time
    1        Hydrothera      4         RAI (joint tenderness) and AIMS2          Hydrothera
    RCT170   py vs land      weeks     (mood and tension)                        py (all
    Level    exercise vs     (end of                                             p=0.03)
    1+       seated          treatme   Knee and wrist ROM; Morning               NS
             immersion       nt)       stiffness; Grip strength; AIMS 2
             vs                        (physical capacity, pain, social,
             progressive               work and affect) and Pain (McGill)
             relaxation
4
5   c) ROM and Resistance
6
    Study    Treatment       Follow- Outcomes                                    Result -
                             up                                                  best
                                                                                 treatment
    1        ROM             12        Painful joints in the left hand           Exercise
    RCT177   exercise vs     weeks                                               (p<0.05)
    Level    Control                   Painful joints in the right hand;         NS
    1+       (active                   MCP and PIP extension, dexterity
             lifestyle)                and grip strength (all in the left and
                                       right hands)
             Resistance                Dexterity in the left hand                Exercise
             + ROM                                                               (p<0.05)
             exercise vs               Painful joints, MCP and PIP               NS
             Control                   extension and grip strength in the
             (active                   left and right hands; Dexterity in
             lifestyle)                the right hand

    Rheumatoid arthritis: full guideline DRAFT (August 2008)                    90 of 307
    DRAFT FOR CONSULTATION

1
2   d) Aerobic
3
    Study Treatment          Follow- Outcomes                                   Result -
                             up                                                 best
                                                                                treatment
    1        High            Over      Radiographic damage (Larsen              Exercise
    RCT1     intensity       the 2     score for all small joints - hands       (all p<0.05)
    78,179
             aerobic         years     and feet; and for small joints of the
    Leve     exercise                  feet only – all p=0.047) and Bone
    l 1++    (RAPIT) vs                mineral density of the hip
             control                   Radiographic damage (Larsen              NS
             group                     score for small joints of the hands
             (usual care)              only) and Bone mineral density of
                                       the spine
                             1 and 2   MACTAR score and Muscle                  Exercise
                             years     strength                                 (all p<0.05)
                                       HAQ score; DAS4 score and                NS
                                       radiographic damage (Larsen
                                       score for large joints)
    1        Aerobic         2 years   ESR; Number of swollen joints;           NS
    RCT1     exercises                 Pain (VAS); Morning stiffness;           Dropouts:
    82
             (Self-                    HAQ; Larsen score; Functional            Aerobic
    Leve     training vs               score and Isometric Muscle               WORSE
    l1+      self-training             strength of knee extensors
             + PT
             training vs
             group
             training vs
             group
             training and
             pool) vs
             Control (no
             training)
    1        Class           Over      Overall symptoms; Walk time and          Aerobic
    RCT1     aerobic         the 12    grip strength                            (p=0.04;
    81
             exercise vs     weeks                                              both
    Leve     Control                                                            p<0.005)
    l1+      (usual
             exercise)

             Home                      Walk time and grip strength              Aerobic
             aerobic                                                            (both
             exercise vs                                                        p<0.005)
             Control                   Overall symptoms                         NS
             (usual
             exercise)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                   91 of 307
     DRAFT FOR CONSULTATION

      1       Individualis    12         MACTAR score; HAQ score;                NS
      RCT1    ed aerobic      months     DAS28 and QoL (RAQoL and
      83
              exercise vs                RAND-36 mental and physical)
      Leve    General
      l 1+    exercise

 1
 2   Recent onset RA
 3
 4   Physical exercise programme vs control group (usual care)
 5   1 RCT190 found that physical exercise programme was significantly better than
 6   control group (usual care) at 1 year for: EuroQoL (VAS); Timed Stands test
 7   and Grip strength (both p<0.01). However, there was NS difference for: ROM;
 8   Pain (VAS); HAQ-DI; DAS28; Percentages of patients taking different types of
 9   medication and Percentage of patients reaching healthy physical activity.
10   Level 1++

11   6.2.5 Health economic evidence statements
12   The average annual medical cost per patient for the RAPIT program was
13   2,115 euros, compared to 1683euros for usual care. The RAPIT group
14   showed no significant different in QALY using SF-6D, and using EQ-5D and
15   VAS the usual care program showed greater QALY gains. The study reports
16   ICERs for total societal cost, concluding that using EQ-5D and VAS, the UC
17   has better cost-utility, and using the SF-6D the ICER is 67,000 euros per
18   QALY, but with no significant difference in the net benefit. Therefore this
19   paper does not provided convincing cost-effectiveness evidence in support of
20   a high intensity exercise program193.

21   6.2.6 Evidence statements summary
22   In general, exercise provokes a favourable response in terms of physical and
23   psychological benefits (176; 181; 168; }148,170,180; 182). Aerobic (dynamic) exercise
24   programmes improve the components of health-related fitness, enhance
25   psychological status, reduce pain and fatigue and have a positive effect on
26   functional capacity without exacerbating disease activity or accelerating joint
27   damage (179,180; 182; 181). Specific exercise, aimed at enhancing joint range of
28   motion (joint flexibility) or muscle strength (resistance training) results in
29   specific improvements170,170,180. The majority of evidence comes from samples
30   of patients with chronic, stable RA in functional class I/III 180,182,182 Exercise in
31   water provides similar physical benefits to exercise on land but may have
32   additional and important psychological effects (which positively impact on
33   concordance)170; 176. Adoption of regular physical activity and exercise
34   strategies is more successful if personalized contact with a health professional
35   occurs in which the benefits and barriers to exercise are discussed and there
36   is opportunity for group contact181 183. The different modes of exercise and
37   their method of delivery affect outcome directly 183 and indirectly via issues of
38   concordance178,179 181
39
40   There is some evidence that a comprehensive package of care, delivered in
41   the community and addressing patient specific needs through education,
42   exercise and pain relief modalities has long-term benefits on self-efficacy,

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  92 of 307
     DRAFT FOR CONSULTATION

 1   disease management knowledge and some measures of disease activity and
 2   function172 ;173
 3
 4   The data for clear benefit for the use of electrophysical agents in the
 5   management of RA is lacking. The present data set highlights conflicting
 6   results in, mainly poor quality studies in which multiple interventions and
 7   outcomes confuse synthesis; therefore little confidence can be attached to the
 8   findings. However, some agents, eg, TENS, wax baths, suggest that short-
 9   term symptomatic relief may occur166; 164.
10
11   The therapeutic effect of manual therapy cannot be determined for RA due to
12   the dearth of literature in this area.

13   6.2.7 From evidence to recommendations
14   Exercise, whether on land or in water is beneficial for most people with RA.
15   The evidence shows that exercise is associated with improved cardiovascular
16   outcomes and the GDG felt that it was necessary for all people with RA to
17   have access, with periodic review, to physiotherapy for advice about
18   incorporating an appropriately tailored exercise programme aimed at
19   improving both general health related fitness and specific locomotor and
20   balance problems into their lives. To do this the healthcare professional
21   providing the intervention would be required to have specific expertise in
22   rheumatoid arthritis.
23
24   Concordance with exercise was highlighted as a particular problem. The GDG
25   therefore felt it is important that all members of the MDT provide people with
26   RA consistent and supportive messages regarding its benefits.
27
28   In reviewing the evidence for comprehensive physiotherapy the GDG noted
29   that the results, whilst promising, would benefit from additional research
30
31   The GDG noted that the provision of electrophysical agents such as wax
32   baths and TENS were akin to the provision of analgesics, in that although
33   there is no evidence of long term effect on disease progression many patients
34   find that they provide short-term symptomatic relief. The GDG felt it would be
35   premature to discard these treatments which may be favoured by patients for
36   their palliative effect.
37

38   6.2.8 Recommendations
39   R 12 Access to specialist physiotherapy, with periodic review, should be
40   offered to all people with rheumatoid arthritis in order to:
41       • Improve general health-related fitness and encourage regular exercise.
42       • Receive instruction on specific exercises aimed at enhancing joint
43          flexibility, muscle strength and other functional impairments.
44       • Provide information and advice about the short-term pain relief
45          provided by electrophysical agents (e.g. TENS and wax baths)




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             93 of 307
     DRAFT FOR CONSULTATION

1    6.3    Occupational therapy

2    6.3.1 Clinical introduction
 3   Occupational therapy (OT) aims to improve: a person’s ability to perform daily
 4   activities and participate in valued life activities and roles at work, in the home,
 5   leisure and socially; facilitate successful adaptations in lifestyle; and to
 6   prevent or minimise functional and psychological problems. Comprehensive
 7   OT programmes can include a wide range of interventions (see Table below).
 8   Enabling self-management using strategies such as joint protection and
 9   fatigue management is a central part of treatment. Enhancing concordance is
10   essential and so OTs frequently use cognitive-behavioural approaches. There
11   is a particular focus in OT on maintaining hand function in RA. “We use our
12   hands for almost everything we do, so keeping them going is really important
13   for me.” Work rehabilitation is provided both for those still in work at risk of job
14   loss, as well as those unemployed. “Working has helped boost my self
15   esteem. I truly believe that if I wasn’t working, my RA would have progressed
16   faster and I would be in a worse state than I am now.” “The visit [OT work
17   assessment] was a turning point in my life……with the adjustments that
18   followed I could continue to support my family…..and achieve a measure of
19   success in the workplace” (NRAS booklet: “I want to work: a self-help guide
20   for people with RA (2007)13.
21

22   Table 1: Comprehensive Occupational Therapy interventions.

     Occupational                    Therapy include:
     Interventions

     - Activities of daily living (ADL)         - Self-management education (group
     rehabilitation                             and individual)

     - Work rehabilitation: including on-site   - Joint protection training; energy
     work assessment, ergonomics/               conservation/ fatigue management
     adaptations, employer liaison, work        and sleep hygiene training
     environment adaptation, functional
     capacity evaluation and work               - Assistive devices
     hardening

     - Exploring voluntary work & adult         - Orthoses
     education; leisure rehabilitation

     - Stress and pain management               - Hand and upper limb therapy &
                                                exercise

     - Relaxation training                      - Therapeutic activities

     - Communication and assertiveness          - Home assessment, recommending
     training                                   environmental modifications and
                                                housing adaptations

     - Counselling. (May also provide           - Mobility aids prescription (including
     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 94 of 307
     DRAFT FOR CONSULTATION

     cognitive-behavioural therapy with        wheelchair/powered aids)
     post graduate training)

     - Family/carer liaison and support        - Foot care advice

     - Advice on social security benefits      - Exercise for health & well-being
     and community resources

 1
 2   All OTs have dual training in physical and mental health rehabilitation. Many
 3   people with RA discuss the “frustration” caused by multiple functional
 4   difficulties, pain, fatigue and at times feelings of stress or low mood. Thus
 5   psychological interventions are often used to help people in adjusting to living
 6   with their disease.
 7
 8   Data suggests194-198 that from an early stage some 60% of people have
 9   problems that could potentially benefit from OT for hand therapy (joint
10   protection, hand exercises, orthoses) and ADL training. Over a third could
11   benefit from work rehabilitation. At a later disease stage, this figure could be
12   higher. What is the evidence that OT has an impact on symptoms, disease
13   progression, function and quality of life in either early or established RA? In
14   particular:
15         • Does OT help to maintain functional ability and/or slow its
16             deterioration?
17        •   Does hand therapy (ie joint protection, hand exercises, hand splints)
18            have any impact on maintaining hand function?
19        •   Does OT help maintain or improve participation in social roles, in
20            particular work?
21        •   Does OT have any impact on psychological status?

22   6.3.2 Clinical methodological introduction
23   We looked for studies that investigated the efficacy of different aspects of
24   Occupational therapy (OT) with respect to symptoms, joint damage, function
25   and quality of life in patients with RA (recent onset and established disease).
26   Due to the large volume of evidence, only MAs and RCTs were selected
27   which were of a UK-relevant population; if the population was mixed arthritis
28   there had to be >75% RA or RA subgroup analysis and for trials looking at
29   splints, orthoses or CBT the sample size had to be N≥30.
30
31   Two MAs199,200 and 8 RCTs135,201-208 were found that fulfilled the inclusion
32   criteria. One of the RCTs206,207 was published as 2 separate papers and
33   reported the 18-month and 5-year follow-up results of an RCT already
34   included in the MA on psychological interventions199. Two RCTs135,206 were
35   found looking at patients with a recent onset of RA, all the other RCTs were
36   conducted using patients with established RA, and the two MAs used trials
37   which were of a mixed population (recent onset and established RA). All trials
38   were methodologically sound.
39
40   Note: studies on joint protection have been covered in the education section.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                95 of 307
     DRAFT FOR CONSULTATION

1
2    Mixed population (Recent onset and Established RA)
3    Two MAs199,200 were found and focused on RCTs which compared either
4    psychological interventions199 or splints and orthoses200. Both MAs were well-
5    conducted, however the RCTs included in the analysis were of varying quality
6
7    The first SR/MA199 included 25 RCTs in the analysis which differed with
8    respect to:
9         • Study size (range N=8 to N=141)
10        •   Study quality – max score of 10 (some poor and some reasonable-
11            good quality)
12        •   Study duration – length of intervention (range 3 days to 9 months
13            with follow-up range from 2 to 18 months)
14        •   Comparison group (placebo; usual care; waiting list)
15        •   Intervention (N=13 multimodal cognitive-behavioural interventions;
16            N=5 included biofeedback; N=5 more traditional psychotherapeutic
17            interventions; N=2 intervention involved patients expressing difficult
18            emotions or stressful experiences)
19   The second SR/MA200 included 7 RCTs in the analysis, all looking at working
20   wrist splints (the RCTs on foot orthoses have been covered in the podiatry
21   question). However, the RCTs differed with respect to:
22         • Comparison group (3 RCTs no splint, 2 RCTs other splints)
23        •   Study size (range N=10 to N=110 for wrist splints)
24        •   Study quality – max score of 5 (some poor and some reasonable-
25            good quality for wrist splints)
26        •   Study duration – length of intervention (range 1 week to 6 months for
27            wrist splints)
28   Recent-onset RA
29   The two included RCTs135,206 compared OT vs usual care and had sample
30   sizes of N=326 and N=53 respectively. One RCT135 had a treatment phases of
31   6 to 8 weeks with 6 month follow-up and the other RCT206 was an 18 month
32   follow-up of a trial included in the MA199 on psychological interventions.
33
34   Established RA
35   The 7 included RCTs differed with respect to the following:
36        • Sample size (range: N=47 to N=144)
37        •   Blinding (1 RCT double blind, 4 RCTs single blind, 2 RCTs
38            unblinded)
39        •   Trial length (range: 4 weeks to 6 months; follow-up ranged from 6
40            months to 5 years)
41        •   Treatment (4 RCTs CBT vs standard care or attention-placebo; 1
42            RCT customised splint vs two commercial splints; 1 RCT OT vs no
43            treatment; 1 RCT CBT for pain vs mindfulness-based emotion
44            regulation therapy vs education control group)


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               96 of 307
    DRAFT FOR CONSULTATION

1   6.3.3 Health economic methodological introduction
2   Two studies were identified and appraised. Li et al157 is a Canadian cost-utility
3   analysis comparing a primary therapist model (PTM) to a traditional therapy
4   model (TTM). Van den Hout et al193 is a cost-utility analysis of a
5   multidisciplinary job retention and rehabilitation program compared to usual
6   care in patients at risk of job loss.

7   6.3.4 Clinical evidence statements
8   Mixed population (Recent onset and Established RA)
9
    Study    Treatment          Outcomes                Follow-up     Result - best treatment
    1        Psychological      Pain                    Treatment     Psych (13 RCTs, effect
             interventions vs
    MA199    control                                    end           size 0.22, 95% CI 0.07 to
    Level                                                             0.37, p=0.003)
    1++                                                 Follow-up     NS
                                                        end
                                Disability              Treatment     Psych (5 RCTs: effect
                                                        end           size 0.30, 95% CI 0.12 to
                                                                      0.42, p=0.00001).
                                                        Follow-up     NS
                                                        end
                                Self-efficacy           Treatment     Psych (5 RCTs: effect
                                                        end           size 0.35, 95% CI 0.11 to
                                                                      0.59, p=0.017).
                                                        Follow-up     NS
                                                        end
                                Tender joints           Treatment     NS
                                                        end           Psych (5 RCTs: effect
                                                        Follow-up     size 0.30, 95% CI 0.04 to
                                                        end           0.56, p=0.005)
                                psychological status    Treatment     Psych (p=0.007 and
                                and coping              end and       p=0.04)
                                                        follow-up
                                                        end
    1        Working wrist Grip strength of non-        Immediate     Gauntlet (1 RCT, N=38;
    MA200    gauntlet vs   dominant hand (palmar                      p<0.05)
    Level    no splint     splint and elastic with
    1++                    metal stay ready made
                           gauntlet)
                           Grip strength of                           NS
                           dominant hand and
                           non-dominant hand
                           (working splint,
                           plastazote and
                           polythene sheeting
                           custom-made gauntlet)
             Working wrist Passive joint motion         1 week        Gauntlet (1 RCT, N=55;
             gauntlet                                                 p<0.05);


    Rheumatoid arthritis: full guideline DRAFT (August 2008)               97 of 307
     DRAFT FOR CONSULTATION

              (elastic with Work performance and                     NS
              metal insert) pain using screwdriver
              vs no splint  or shears; dexterity;
                            Pain on motion, activity
                            or at rest; grip strength;
                            morning stiffness;
                            Active joint motion,
                            pronation & supination;
                            Pinch grip; Joint and
                            forearm circumference;
                            HAQ
              Futuro wrist  Wrist pain; Tender +         2 weeks     NS
              gauntlets and swollen joints; Passive
              Thermolyn     wrist ROM; Grip
              custom-made strength with & without
              wrist         orthoses
              gauntlets
              Futuro wrist  Dexterity and grip           1 week      NS
              gauntlets,    strength without
              Alimed wrist  orthoses
              gauntlets and
              Roylan wrist
              gauntlets
              resting hand Patient preference            1-6 months Splints (1 RCT, N=78;
              and wrist                                             p<0.001)
              splints were  Grip strength; Swollen
              significantly joints and RAI                           NS
              better than
              no splints
              circumferenti Patient preference           1 month     NS
              al cotton-
              padded splint
              and pan-type
              hard
              thermoplastic
              splint
 1
 2   Recent onset RA
 3        •   1 RCT135 found that OT was significantly better than the control (no
 4            intervention) groups for: Use of some self-management methods
 5            (particularly hand and arm exercises, joint protection and rest);
 6            Receipt of a working splint and a resting splint; Owning and use of
 7            assistive devices. However, there was NS difference for: HAQ;
 8            DAS28 score; AIMS2 scores; self-efficacy (ASES) score. Level 1++
 9
10   Established RA
11
     Study    Treatment       Outcomes                             Follow-up     Result - best
                                                                                 treatment
     1        OT vs       Functional score (AIMS); Pooled          6 weeks       OT (p=0.006;
     RCT201   Control (no index (symptoms and function)                          p=0.04)
     Rheumatoid arthritis: full guideline DRAFT (August 2008)             98 of 307
DRAFT FOR CONSULTATION

Level    treatment)     Pain (VAS), HAQ and Beck                             NS
1++                     Depression scale
1        CBT vs         Physical functioning (functional        6 months     CBT (all
RCT203   Control        disability, pain, fatigue) and          treatment    p<0.05)
Level    (routine       Psychological functioning               and 1 year
1+       care)          (depression, negative mood, anxiety)    follow-up
                        Illness cognition, Coping with stress                Similar; NS
                        and coping with pain; Disease
                        Activity and Social functioning
1        CBT vs         HAD depression and anxiety and for      18 months    CBT (p<0.05)
RCT206   Control        HAQ
Level    (standard      Pain (11-point scale), RAI, ESR and                  NS
1+       care)          Coping strategies Questionnaire
                        (CSQ).
                        Lower use of healthcare resources       5 years      CBT (p<0.05)
                        overall; number of: inpatient nights,
                        physiotherapy referrals, injections
                        and total occasions of care
                        number of: Rheumatology                              NS
                        consultations, psychiatric referrals,
                        patients discharged as improved,
                        orthopaedic referrals and surgeries
1        CBT vs         Coping strategies questionnaire         6 months     CBT
RCT204   Control                                                and 12       (p=0.0017 and
Level    (routine                                               months       p=0.0001)
1+       care) vs
         Control        Pain (VAS and McGill); AIMS, Ways                    NS
         (attention-    of coping scale, AHI and Beck
         placebo)       Depression scale
1        Mindfulnes     Psychological distress (p=0.04) and   2 months       NS
RCT205   s-Based        Well-being                            (end of
Level    Stress                                               treatment)
1+       Reduction      Depressive symptoms and DAS28         2 months       NS
         programm                                             and 6
         e (MBSR)                                             months
         vs Control     Psychological distress and well-being 6 months       MBSR
         (waiting                                                            (p=0.04;
         list)                                                               p=0.03)
1        Customise      Pain (VAS, AHFT - Arthritis hand        4 weeks      NS
RCT202   d leather      function test - all items); MACTAR      (end of
Level    wrist splint   score                                   treatment)
1+       (LWS) and
         the
         commercia
         lly
         available
         wrist
         splints
         (RWS and
         AWS)
         Commerci       Grip                                    4 weeks      RWS (p=0.03)
Rheumatoid arthritis: full guideline DRAFT (August 2008)              99 of 307
     DRAFT FOR CONSULTATION

              ally           Pain (VAS), AHFT (Arthritis hand         (end of      NS
              available      function test - all items) and           treatment)
              wrist          MACTAR score
              splints        Dexterity                                             AWS (p=0.04)
              (RWS vs
              AWS)
     1        Treatment      positive affect, coping efficacy for     30 days      Mindfulness
     RCT208   (Mindfulne     pain and catastrophizing                              and CBT
     Level    ss vs CBT                                                            (p<0.01)
     1++      for pain) vs   change in: Daily pain, negative affect                NS
              Control        and daily depression symptoms
              (education     Pain control                                          Mindfulness
              )
1
2
3         •   The same RCT208 found that for patients with recurrent depression
4             (RD+), mindfulness treatment was significantly better than CBT or
5             education at 30 days for change in: positive affect, negative affect,
6             coping efficacy for pain, catastrophizing and for physicians’ ratings of
7             joint tenderness and swelling (all: p<0.001 except negative affect
8             p<0.01). Level 1++

9    6.3.5 Health economic evidence statements
10   The Li et al study157 showed no statistically significant differences in either
11   QALY or societal cost, but suggests that a primary therapy model has the
12   potential to be a cost-effective alternative to the traditional occupational and
13   physical therapist roles. The trial was only a 6-month study and so long-term
14   analysis of the costs and benefits could help. The van den Houtet al study
15   concluded that there was no significant difference in either costs or QALYs.
16   The variability of costs in this study means that impact of the job retention
17   program are unclear and the lack of any significant difference of costs or
18   QALYs means conclusions on the cost-effectiveness cannot be made. The
19   applicability of this Dutch care program may be limited in a UK setting

20   6.3.6 Summary of evidence statements
21   A programme of comprehensive occupational therapy (see Table 1 for
22   description) improved functional ability in established RA 201. In early RA,
23   comprehensive OT significantly improved use of self-management strategies
     135
24       but not symptoms or functional ability (most participants’ RA was well
25   controlled by DMARDs).
26
27   Hand therapy can impact on hand function. Several studies (reviewed under
28   EDU section) in early RA patients have evaluated joint protection and energy
29   conservation training using a group educational-behavioural approach. One
30   study 133,134 found that the joint protection group had significant reductions in
31   symptoms and increased function compared to controls at 1 year and 4 years.
32
33   Another RCT in patients with established RA receiving anti-TNF therapy 145
34   found that joint protection, energy conservation and hand/upper limb exercise


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              100 of 307
     DRAFT FOR CONSULTATION

 1   training, was significantly better than a control group for reductions in pain,
 2   physical symptoms and improved functional ability at 8-months follow-up.
 3   Most studies of working wrist orthoses200 were generally comparisons of splint
 4   designs and had short-term follow-ups. These identified little difference
 5   between splint designs, indicating splint choice should suit patients’
 6   preferences. In the short–term (ie. immediately or 1-2 weeks) and the longer-
 7   term (only one study in the MA looked at longer term - 6 months; 209) the
 8   studies found that there were no differences between splint and control
 9   groups. However, the splint group had significant improvements in pain, pinch
10   and grip strength when wearing the splint as compared to not-wearing and
11   there was a significant preference for splint use 209.
12
13   Hand exercises are provided by both Occupational and Physiotherapy. There
14   is moderate evidence that a combination of range of movement and resistive/
15   strengthening hand exercises can significantly improve dexterity, key grip and
16   self-reported upper limb function 175,177. This is more effective than range of
17   movement exercise alone.
18
19   There was a lack of evidence that work rehabilitation for people with RA still in
20   work at risk of job loss is effective
21
22   Psychological interventions commonly used by OTs (relaxation, imagery,
23   stress management and teaching cognitive coping skills) were found in a
24   systematic review 199 to significantly reduce pain and improve functional ability
25   and psychological status at the end of treatment and follow-up. However,
26   effectiveness varied depending upon the nature of the intervention and was
27   less in studies with comparison groups of education, attention or placebo
28   control. Stress reduction techniques and CBT were also found to be
29   significantly better at improving some aspects of psychological status (eg
30   coping, psychological distress)204,205. One study206 also identified changes in
31   functional ability. Studies have not identified changes in pain. Outcomes may
32   be dependent on the nature and duration of the therapy.

33   6.3.7 From evidence to recommendations
34   The GDG noted that comprehensive occupational therapy programmes are
35   beneficial to improve functional ability in people with established RA and
36   increase use of self-management in early RA. Additionally it was felt it was
37   important for all people with RA to have access, with periodic review, to
38   occupational therapy for assessment and provision of tailored comprehensive
39   programmes. The GDG considered that the evidence supported joint
40   protection training and hand exercises (strength and flexibility) should be
41   provided for people with hand function difficulties. To be effective this should
42   be provided by health professionals using an educational-behavioural
43   approach and not just simple advice.
44
45   The GDG noted that the limited evidence for resting splints indicated that
46   these do not have a significant effect on symptoms, although it was accepted
47   that patients commonly express a preference for wearing these, particularly
48   during active periods of hand pain and inflammation. Although there is no
49   overall evidence of the long-term benefit from hand splints, the GDG noted

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              101 of 307
     DRAFT FOR CONSULTATION

 1   that the provision of such splints would be similar to the provision of
 2   analgesics in that many patients find them symptomatically beneficial.
 3
 4   The GDG were disappointed with the lack of evidence that work rehabilitation
 5   can minimise the impact of disease on symptoms, joint damage, function and
 6   quality of life, as working people with RA report finding work assessment and
 7   interventions particularly beneficial.
 8
 9   The GDG noted there was good evidence that psychological interventions
10   (e.g. relaxation, stress management and cognitive coping skills) have a
11   beneficial impact on the reduction of pain and improved functional ability in
12   both early and established disease, and that stress reduction techniques and
13   cognitive behavioural therapy could improve some aspects of psychological
14   status. Although the management of depression was beyond the scope of this
15   guideline‡, the GDG felt that it was appropriate to make a recommendation
16   offering these psychological interventions to those patients who needed help
17   in coping with the consequences of their disease. These could be provided
18   either by an occupational therapist or any other member of the MDT who had
19   appropriate training, or by specialist referral if deemed necessary.
20
21   In view of the limited evidence for many specific OT interventions the GDG felt
22   that there should be a research recommendation to asses the effectiveness of
23   both comprehensive and specific OT interventions.

24   6.3.8 Recommendations
25   R 13 Access to specialist occupational therapy, with periodic review, should
26   be offered to all people with RA who:
27         • are experiencing difficulties with any of their usual everyday
28            activities§
29           •    are experiencing problems with hand function
30   R 14 Psychological interventions (e.g. relaxation, stress management and
31   cognitive coping skills) should be offered to help people with RA to adjust to
32   living with their condition.
33

34   6.4         Podiatry

35   6.4.1 Clinical introduction
36   Feet are a part of the musculoskeletal system that most of us take for granted,
37   but the large number of bones and synovially-lined structures mean that the
38   vast majority of people with RA will experience significant problems usually in
39   a symmetrical pattern. For many this will be a part of their presenting
40   complaint, although the increasing prevalence of foot problems is strongly
41   related to disease duration.211
42

     ‡
         For management of depression, see NICE guideline210
     §
         ie at home, work and leisure
     Rheumatoid arthritis: full guideline DRAFT (August 2008)             102 of 307
     DRAFT FOR CONSULTATION

 1   Involvement of he feet, even to a mild degree, is a significant cause of
 2   impaired mobility and functional capacity in RA.212 In about three quarters of
 3   people with RA, the foot contributes to difficulty with walking and is the main
 4   or only cause of walking impairment in one quarter.213
 5
 6   Although foot problems are almost universal for people with RA, this area is
 7   often neglected, and the provision of podiatry services is often inadequate.214
 8   Because feet are complicated yet vital, experts are needed who can assist
 9   people with RA with their foot-related disease and disability. In the UK, foot
10   health services can be provided by a range of professions although
11   podiatry/chiropody is recognised as the profession primarily providing foot
12   care. Podiatrists/chiropodists must be HPC registered and will usually have
13   undertaken pre-registration training at degree level.
14
15   There are five areas of practice in which podiatrists may be involved in
16   providing care for people with RA:
17      1. Education and footwear advice
18      2. Provision of, or assistance with finding orthoses and footwear.
19      3. General foot care, nail cutting, corn and callus reduction, provision of
20          padding.
21      4. Management of the high risk vasculitic or ulcerative foot
22      5. Extended Scope and Practice surgery.

23   6.4.2 Clinical methodological introduction
24   We looked for studies that investigated the efficacy of different aspects of
25   podiatry with respect to symptoms, joint damage, function and quality of life on
26   patients with RA (recent onset and established disease). Trials were selected
27   which were of a UK-relevant population and if the population was mixed
28   arthritis there had to be >75% RA or RA subgroup analysis.
29
30   1 MA,200 3 RCTs215-218 and 1 case-series219 were found that fulfilled the
31   inclusion criteria. One of the RCTs was published as two separate
32   papers216,218 reporting different outcomes and so the trial has only been
33   counted once. However results from both papers are reported and referenced
34   here. One of the RCTs217 was excluded as evidence due to methodological
35   limitations. All other trials were methodologically sound. No papers were found
36   looking at patients with a recent-onset of RA.
37
38   Mixed population (Recent onset and Established RA)
39
40   One Cochrane SR/MA200was found and focused on RCTs which compared all
41   types of orthoses vs placebo, active intervention or regular treatment. Only the
42   results for trials on foot orthoses have been reported here. The MA itself was
43   well conducted however, the 3 RCTs it included were of varying quality.
44   Studies included in the analysis differed with respect to:
45         • Intervention (supporting insoles, extra depth shoes and insoles in
46             extra depth shoes)
47        •   Comparison group (regular footwear, extra depth shoes, placebo
48            insoles)

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              103 of 307
     DRAFT FOR CONSULTATION

1         •   Study size (range N=28 to N=102 for foot orthosis)
2         •   Study quality – max score of 5 (All studies reasonable to good quality
3             for foot orthoses)
4         •   Study duration – length of intervention (2 months to 3 years for foot
5             orthoses)
 6   Established RA
 7
 8   2 RCTs215,216,218 and 1 case-series219 were found which looked at aspects of
 9   podiatry in patients with established RA.
10
11   The 2 RCTs215,216,218 were both single-blind, parallel group studies. The first
12   RCT215 looked at callus treatment in N=38 patients and compared 2 different
13   treatment arms: normal callus treatment vs sham callus treatment, with a 5
14   week follow-up. The second RCT216,218 looked at foot orthoses in N=101
15   patients and compared 2 different treatment arms: rigid foot orthoses vs
16   control (no orthoses) in a 30 month treatment phase.
17
18   The case-series219 looked at the effects of heat-mouldable shoes on N=25
19   patients with RA (disease duration not mentioned) in a 3 month treatment
20   phase.

21   6.4.3 Health economic methodological introduction
22   Three papers were found and two did not meet the inclusion criteria.220,221 The
23   third paper, Clark et al222 was a critical review of foot orthoses in patients with
24   RA. Two papers reviewed216,223 considered the cost of orthoses, but neither
25   contained any form of cost-comparison or health economic evaluation.

26   6.4.4 Clinical evidence statements
27   Mixed population (Recent onset and Established RA)
28
29   Extra depth shoes vs regular footwear (2 months)
30        • 1 MA200 found that extra depth shoes were significantly better than
31           regular footwear at 2 months for: HAQ (1 RCT, N=30; effect size
32           WMD –0.20, 95% CI –0.35 to –0.05; p=0.01); Pain on walking (1
33           RCT, N=30; effect size WMD –18.7, 95% CI –28.5 to –8.9;
34           p=0.0002); Pain on climbing stairs (1 RCT, N=30; effect size WMD –
35           27.0, 95% CI –37.8 to –16.2; p<0.00001)and pain-free walking time
36           (1 RCT, N=30; effect size WMD 18.2, 95% CI 8.2 to 28.2; p=0.0004).
37           However, there was NS difference for: Fatigue and subjective well-
38           being. Level 1++
39   Semi-rigid insoles vs extra-depth shoes (12 weeks)
40       • 1 MA200 found that semi-rigid insoles were significantly better than
41           extra-depth shoes at 12 weeks for: Pain, VAS (1 RCT, N=48; effect
42           size WMD –1.9, 95% CI –3.3 to –0.51; p=0.007). However, there was
43           NS difference for: RB walking, RB stairs, RB stand, Toronto ADL –
44           walking and stairs dimensions, Toronto ADL – stairs dimension,
45           Walking, Lower extremity joint counts and MTP joint count (number of
46           painful joints). Level 1++
     Rheumatoid arthritis: full guideline DRAFT (August 2008)               104 of 307
     DRAFT FOR CONSULTATION

1    Soft insoles vs extra-depth shoes (12 weeks)
2          • 1 MA200 found that there was NS difference between soft insoles and
3            extra depth shoes at 12 weeks for: Pain (VAS); RB Walking, RB
4            stairs, RB stand, Toronto ADL – walking and stairs dimensions, 50
5            foot walk time, Lower extremity joint counts and MTP joint count
6            (number of painful joints). Level 1++
 7   Semi-rigid insoles vs extra-depth shoes (12 weeks)
 8       • 1 MA200 found that supporting insoles (Rohadar posted foot orthoses)
 9           were significantly better than placebo insoles at 3 years for: Hallux
10           abductus angle remained < 21 degrees (1 RCT, N=98; effect size
11           WMD RR 3.6, 95% CI 2.2 to 5.9; p<0.00001). However there was NS
12           difference for: Painful foot joint count, Foot function index and Foot
13           pain. Level 1++
14
15   Established RA
16
17   Normal callus debridement vs sham callus debridement
18       • 1 RCT215 found that there was NS difference between normal callus
19          debridement and sham callus debridement for: forefoot pain (VAS),
20          Plantar pressure measures and spatial temporal gait measures at 5
21          weeks post-intervention. Level 1++
22
23   Heat-mouldable shoes (Level 3)
24        • 1 case-series219 found that for heat mouldable shoes:
25        •   80% wore their shoes all the time during the day and 20% sometime.
26        •   72% wore their custom-made semi-rigid foot orthoses in their shoes
27            and 28% did not
28        •   20% had their shoes modified to control hindfoot vagus
29        •   50% of those who had foot orthoses stated that they always wore
30            their inserts in their shoes.
31        •   80% of patients felt they walked better with the heat-mouldable
32            shoes. 20% were not walking better
33        •   Significantly more patients found that they walked better with the
34            heat-mouldable shoes compared to previous shoes
35        •   Patients found that their heat-mouldable shoes were significantly
36            better than previous shoes and were significantly more comfortable.
37
38   Customised foot orthosis vs control (no orthosis)
39       • 1 RCT216,218 found that the customised foot orthosis was significantly
40         better than the control group (no orthosis) for: Foot function Index
41         (total, pain and disability – all p<0.05) and for
42         Dorsioflexion/plantarflexion motion, Inversion/eversion motion,
43         internal/external AJC rotation and Internal rotation at 30 weeks (all
44         p<0.01). However there was NS difference between customised foot
45         orthosis and the control group (no orthosis) for: foot function Index

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             105 of 307
     DRAFT FOR CONSULTATION

1            (functional limitation), global pain, DAS score, HAQ and Larsen score
2            (hands and feet) at 30 weeks. Level 1++

3    6.4.5 Summary of evidence statements
 4   The following interventions have evidence for efficacy: extra-depth shoes
 5   compared with regular footwear200, semi-rigid insoles compared with extra-
 6   depth shoes (pain and toe deformity200), heat-mouldable shoes (according to
 7   consumer feedback) 219, and customised orthoses compared with no orthoses
 8   (function, pain and disability) 216,218. There was no evidence to support callus
 9   debridement215, and no difference between soft insoles and extra-depth
10   shoes200.
11
12   Most trials were conducted over relatively short time periods (5 weeks to 30
13   weeks) 215 200,216,218 with one exception (3 years for supporting insoles versus
14   placebo insoles) 200.

15   6.4.6 From evidence to recommendations
16   There is evidence that insoles and footwear have a positive impact on
17   symptoms, function and quality of life. There is a hierarchy of strength of
18   evidence, with the most robust evidence being for custom-built shoes, tailored
19   to the patient’s own feet, and the least evidence for soft insoles. The GDG felt
20   that it was necessary for all patients to have access to an expert podiatrist.
21   Some simple assessments and interventions could be conducted by non-
22   specialists, such as mass-produced insoles. However other interventions
23   would require the expertise of a trained podiatrist, such as custom-built
24   footwear and moulded insoles.

25   6.4.7 Recommendations
26   R 15 All people with RA who are identified as having problems with their feet
27   should have access to specialist podiatry assessment with periodic review.

28   7      Pharmacological management

29   7.1    Symptom control
30
31   Analgesics

32   7.1.1 Clinical introduction
33   When people with RA are asked to list their priorities for the management of
34   RA, pain relief is almost always in first place. Pain is a cardinal sign of
35   inflammation, and therefore adequate disease control will result in satisfactory
36   pain relief in many people with RA. However, others will not gain sufficient
37   disease-control to relieve symptoms, or will develop mechanical pain due to
38   damage to joints or surrounding structures. Analgesics will be used by all
39   patients with RA at some time during the course of their disease. Key
40   questions include:
41         • Are analgesics helpful for the pain of RA?
     Rheumatoid arthritis: full guideline DRAFT (August 2008)              106 of 307
     DRAFT FOR CONSULTATION

1         •   Are there any particular side effects of analgesics that are a problem
2             in the RA population?
3         •   Does the addition of analgesics to other pain-relieving medication
4             (particularly NSAIDs) add anything to pain-control?
5         •   Are analgesics better than NSAIDs for some aspects of pain control?
6         •   Is there any evidence to support the use of low dose anti-
7             depressants in RA, in keeping with their use in other diseases
8             associated with chronic pain?

9    7.1.2 Clinical methodological introduction
10   We looked for studies that investigated the efficacy and safety of analgesic
11   drugs with respect to symptoms, function and quality of life in patients with a
12   recent onset of rheumatoid arthritis, and in established disease. Due to the
13   large volume of evidence on analgesics, trials were selected which fulfilled the
14   following criteria: were of a UK-relevant population; if the population was
15   mixed arthritis there had to be >75% RA or RA subgroup analysis; for Opioids
16   or opioid-paracetamol comparisons - sample size >50; for anti-depressant
17   comparisons – all papers were considered apart from those in patients who
18   were already depressed and those which used doses of anti-depressants
19   which were greater than those used in clinical practice for their analgesic
20   effect.
21
22   Eight trials224-231 were found that fulfilled the criteria. Trials were found which
23   looked at opioids, opioids + paracetamol, NSAIDs + paracetamol, anti-
24   depressants, and other analgesics (nefopam). No suitable trials were found
25   which looked at paracetamol alone. All trials were methodologically sound.
26
27   NOTE: all trials found looked at either patients with established RA or duration
28   not mentioned, no studies were found which specifically mentioned patients
29   with a recent onset of RA.
30
31   Opioids and opioid-paracetamol
32
33   1 RCT224 and 2 case-series’ (prospective)225,226 were found which fulfilled the
34   criteria for opioids and opioid-paracetamol analgesics.
35
36   The RCT224 was a double-blind, parallel group study looking at opioids or
37   opioid-paracetamol analgesics in patients with established RA or disease
38   duration not mentioned. It compared 2 different treatment arms (codeine 90
39   mg/day + paracetamol 1500 mg/day + Diclofenac 50 mg once/day vs
40   Diclofenac 100 mg/day) in N=60 patients with RA in a 7 day treatment phase.
41
42   The 2 case-series’ 225 both looked at transdermal fentanyl (TF) and compared
43   treatment outcome with baseline values. The first case-series225 assessed
44   N=104 patients (disease duration not mentioned) in a 28 day treatment phase
45   and the second226 assessed N=226 patients with established RA in both a 30
46   day and 12 month treatment phase.
47
48   NSAIDs + paracetamol

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 107 of 307
     DRAFT FOR CONSULTATION

 1   Two RCTs227,228 (both double blind and N=20 patients) were found which
 2   looked at NSAIDs + paracetamol. The first RCT228 was a parallel group study
 3   comparing 2 different treatment arms (Indomethacin 50 mg/day + paracetamol
 4   4 g/day vs Indomethacin 150 mg/day) in a 4 week treatment phase. The
 5   second RCT227 was a cross-over study comparing 6 different treatment arms
 6   (Naproxen 500 mg/day, 1000 mg/day or 1500 mg /day + paracetamol 4 g/day
 7   vs Naproxen 500 mg/day, 1000 mg/day or 1500 mg /day) in a 2 week
 8   treatment phase.
 9
10   Anti-depressants
11   2 RCTs229,231 (both double-blind and disease duration not mentioned) were
12   found which looked at anti-depressants as analgesics. The first RCT229 was a
13   cross-over trial of N=256 patients comparing 3 arms: amitriptyline (1
14   mg/kg/day for 3 days then 1.5 mg/kg/day thereafter) vs Trazodone 1.5
15   mg/kg/day for 3 days then 3 mg/kg/day thereafter) vs placebo in a 7 week
16   treatment phase. The second RCT231 was a parallel study in N=36 patients
17   comparing 2 arms: amitriptyline (25 mg/day for 1 week then increased to 50
18   mg/day for week 2 then 75 mg/day thereafter) vs placebo in a 12 week
19   treatment phase.
20
21   Other analgesics
22   1 RCT230 was found which looked at other analgesics. The RCT was a double
23   blind, cross-over study comparing nefopam 180 mg/day vs placebo in N=27
24   patients with established RA in a 4 week treatment phase.

25   7.1.3 Health economic methodological introduction
26   No health economic papers were identified

27   7.1.4 Clinical evidence statements
28   Opioids and opioid-paracetamol
29
30   Symptoms, Function, QoL and global assessment
31        •   1 RCT232 found that the opioid Sativex was significantly better than
32            placebo for: DAS, Pain on movement and at rest, Quality of sleep and
33            SF-MPQ. However, there was NS different for morning stiffness and
34            SF-MPQ dimensions of total intensity of pain at present and intensity
35            of pain at present Level 1+
36        •   1 RCT224 found that there was NS difference between codeine +
37            paracetamol + diclofenac vs diclofenac for: Pain (VAS), morning
38            stiffness, Ritchie Index, Patient’s global judgement of efficacy and
39            physician’s global assessment of tolerability and Number of nocturnal
40            awakenings on the disability score. Level 1+
41        •   1 case-series226 found that TF treatment was significantly better than
42            baseline values for: ADLs, social activities, pain intensity and sleep
43            disturbance due to pain, Quality of sleep and satisfaction with pain
44            treatment at 30 days; better for general well being at 30 days and
45            satisfaction, pain intensity, ADLs and social activities remained stable
46            at 12 months and 85% of symptoms disappeared by the end of the
47            study. Level 3

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              108 of 307
     DRAFT FOR CONSULTATION

1         •   1 case-series225 found that TF treatment was significantly better than
2             baseline values for: Pain control, WBPI pain, pain recorded in
3             patients’ diaries, patient’s assessment of treatment, HAQ score and
4             HAQ components of eating, activities and arising. Level 3
5
6    Use of rescue medication, withdrawals and AEs
7         •   1 RCT232 found that the opioid Sativex was better than placebo for:
8             withdrawals due to AEs and SAEs, and was similar for AEs of nausea
9             and arthritic pains. Level 1+
10        •   1 RCT224 found that there was NS difference between codeine +
11            paracetamol + diclofenac vs diclofenac for: AEs and withdrawals due
12            to AEs. Level 1+
13        •   1 case-series225 found that of patients who received TF treatment,
14            27% withdrew due to AEs, 65% had AEs but there were no SAEs.
15            Level 3
16        •   1 case-series226 found that of patients who received TF treatment,
17            17% had AEs, 10% withdrew due to AEs and 40% used rescue
18            medication. Level 3
19   NSAIDs + paracetamol
20       • 1 RCT228 found that there was NS difference between Indomethacin
21          + paracetamol and Indomethacin for: morning Pain (VAS), Night
22          Pain, morning stiffness, Joint movement, patient’s assessment of
23          therapeutic efficacy and Ritchie articular index. The 2 groups were
24          also similar for ESR, CRP and AEs. Level 1+
25        •   1 RCT227 found that Naproxen 500 mg/day + paracetamol was
26            significantly better than Naproxen (500, 1000 and 1500 mg/day) for:
27            Joint index, Joint pain, Morning stiffness and Global assessment of
28            disease activity. However there was NS difference for ADLs. Level
29            1+
30        •   The same RCT227 found that Naproxen 1000 mg/day + paracetamol
31            was significantly better than Naproxen (500, 1000 and 1500 mg/day)
32            for: Number of painful joints (Ritchie); morning stiffness; pain at rest
33            and movement (VAS) and for Global assessment of disease activity.
34            Level 1+
35        •   The same RCT227 found that Naproxen 500 mg/day + paracetamol
36            was significantly better than Naproxen 1000 mg/day) for: Number of
37            AEs. Additionally, AEs were significantly related to dose of naproxen.
38            Level 1+
39   Anti-depressants
40         • 1 RCT229 found that amatriptyline was significantly better than
41           placebo for: present pain intensity, worst pain, number of painful and
42           tender joints, severity rating summary of painful and tender joints but
43           there was NS difference for least pain (VAS). However, there was NS
44           difference for physical incapacitation. Level 1+
45        •The same RCT229 found that trazadone was significantly better than
46         placebo for depression. However, there was NS difference for: least
47         pain (VAS) or physical incapacitation. Level 1+
     Rheumatoid arthritis: full guideline DRAFT (August 2008)       109 of 307
     DRAFT FOR CONSULTATION

1         •   1 RCT231 found that there was NS difference between amitryptiline
2             and placebo at 12 weeks for Pain and joint tenderness, and the
3             groups were similar for: Total number of withdrawals (both N=4),
4             withdrawals due to AEs (N=2 and N=3 respectively) and withdrawals
5             due to lack of efficacy (N=2 and N=1 respectively). Level 1+
 6   Other analgesics
 7        • 1 RCT230 found that nefopam was significantly better than placebo
 8           for: Pain, Morning stiffness, grip strength and joint tenderness.
 9           However it was worse for AEs and there was NS difference for ESR.
10           Level 1+

11   7.1.5 Summary of evidence statements
12        •   There are very few good quality trials of analgesics in RA. Most of the
13            available studies were over very short time periods, and some in
14            small populations
15        •   A variety of analgesics provide symptomatic benefit in RA (e.g.
16            decreased pain, better sleep, improved activities of daily living,
17            improved social activities, satisfaction with medication)225,226,230,232
18        •   Analgesics appear to be well tolerated 225,226,232
19        •   Some studies suggest decreases in more “inflammatory” symptoms
20            such as morning stiffness and grip strength224,230
21        •   Amitryptiline (in quite high doses) has been shown to be successful
22            in reducing joint swelling, and help to lift a low mood and chronic
23            fatigue 229
24        •   There was no evidence to show that low dose amitryptiline was
25            helpful in a study over a three month period231
26        •   100mg of diclofenac did not show any increased benefit over 50mg
27            diclofenac plus paracetamol and codeine224
28        •   150mg of indomethacin did not show any increased benefit over
29            50mg indomethacin plus paracetamol 228
30        •   The addition of paracetamol to naproxen improved pain control 227
31

32   7.1.6 From evidence to recommendations
33   Although the evidence for analgesics being helpful in RA was sparse, the
34   GDG considered that there was nevertheless sufficient data to suggest that
35   they are effective in controlling pain, and that there should accordingly be a
36   recommendation that these drugs should be offered where other approaches
37   have not resulted in satisfactory pain control. In addition, bearing in mind the
38   need to use NSAIDs and COX2 inhibitors in the lowest effective doses for the
39   shortest periods of time (see recommendation 18), it was also felt important to
40   emphasise that analgesics should be considered as a way to decrease
41   reliance on these drugs
42

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 110 of 307
     DRAFT FOR CONSULTATION

1    The GDG also noted that, although low dose antidepressants are commonly
2    used in clinical practice to treat pain and sleep disturbance in RA patients,
3    there was surprisingly no evidence to support this despite the observation that
4    patients and their health care professionals can testify to their benefits. It was
5    agreed that there should be a research recommendation to address this issue.

6    7.1.7 Recommendations
 7   R 16 Analgesics should be offered to people with rheumatoid arthritis whose
 8   pain control is not adequate.
 9
10   R 17 Analgesics should be considered for use in people with RA to potentially
11   reduce their need for long-term treatment with NSAIDs or COX2 inhibitors.
12
13   NSAID
14   7.1.8 Clinical introduction
15   NSAIDs (e.g. ibuprofen, diclofenac, naproxen) and COX2 inhibitor drugs (e.g.
16   rofecoxib, celecoxib and etoricoxib) have anti-inflammatory and analgesic
17   properties, and many people with RA can testify to their effectiveness in
18   controlling their symptoms. Indeed the quality of life of many would be
19   diminished for many people by not allowing continuing treatment with NSAIDs
20   that they may need to take in significant doses over many years. There is no
21   evidence to suggest that NSAIDs modify the course of RA, and they are
22   purely for symptomatic benefit.
23
24   For some decades it has been clear that NSAIDs also have major
25   disadvantages, with evidence of increased toxicity, particularly in the gastro-
26   intestinal tract, but in other organ systems also (such as the propensity for
27   impairing renal function, and aggravating asthma), and interactions with other
28   commonly prescribed medications (such as warfarin, diuretics, ACE
29   inhibitors). The COX2 selective drugs were designed to decrease
30   gastrointestinal morbidity and mortality, and there is some evidence that they
31   may have been successful in this regard. However, a concern over selectively
32   blocking COX2 is that this might lead to an imbalance in the ratio of
33   prostacyclin to thromboxane, and lead to an increased risk of cardiovascular
34   morbidity and mortality. The VIGOR trial showed clear and early divergence
35   between rofecoxib and naproxen for adverse cardiovascular outcomes, and
36   eventually was instrumental in the withdrawal of rofecoxib due to increasing
37   concerns over the balance of efficacy to toxicity. The data for celecoxib is
38   rather mixed, with some trials suggesting increased cardiovascular risk, but
39   other trials, and observational databases, showing no increased risk. For
40   etoricoxib the MEDAL study showed no increased risk for cardiovascular
41   mortality when compared with traditional NSAIDs, but the COX2 inhibitor was
42   associated with more peripheral oedema and hypertension .More recently
43   evidence has emerged to show increased cardiovascular risk for all NSAIDs,
44   irrespective of their COX2 specificity, and regulatory bodies around the world
45   have added to the cautions before using any of these drugs. Therefore all
46   NSAIDs and COX2 inhibitor drugs should be considered to be similar for
47   cardiovascular risk.
48
     Rheumatoid arthritis: full guideline DRAFT (August 2008)              111 of 307
     DRAFT FOR CONSULTATION

 1   Against the background of these concerns, one would be forgiven for trying to
 2   avoid NSAIDs and COX2 drugs altogether, but this runs the risk of denying
 3   patients access to useful symptom controlling drugs. Clearly caution needs to
 4   be exercised in ensuring that people with RA are screened to ensure that
 5   these drugs are not contra-indicated for any reason. If their use is appropriate,
 6   existing guidelines from regulatory authorities such as the MHRA advise the
 7   lowest dose be used over the shortest period of time. This should also
 8   discourage an over-reliance on drugs where there is no evidence to suggest
 9   that they treat the underlying disease process. In the early stages of RA a
10   person may need high doses of NSAIDs for symptomatic benefit, but if this
11   need continues, this may be a sign that disease control is not adequate.
12   Evidence of efficacy of DMARDs is the ability to decrease NSAID use, and
13   this needs to be reviewed throughout the course of the disease.
14
15   The GDG has been asked to update the current technology appraisal
16   guideline on COX2 drugs, and to consider the conclusions drawn by our
17   colleagues on the NICE osteoarthritis guidelines to determine whether factors
18   related to differences in efficacy or toxicity would be likely to result in
19   conclusions that are different from the OA guidelines. The key questions are:
20
21      1. How clinically and cost-effective are NSAIDs and COX2s in decreasing
22         the symptoms of RA?
23      2. What risks are associated with the use of NSAIDs and COX2s in RA?
24      3. Are there any differences between NSAIDs and COX2s for clinical- and
25         cost-efficacy and toxicity in RA?
26      4. Are there differences between individual drugs within the classes for
27         toxicity and clinical- and cost-efficacy in RA?
28      5. Is the combination of NSAID and proton-pump inhibitor more cost-
29         effective than COX2 selective drugs?
30      6. Are there any circumstances under which a combination of proton-
31         pump inhibitor and COX2 selective drug might be cost-effective in RA?

32   7.1.9 Clinical Methodological introduction
33   We looked for studies that investigated the efficacy and safety of NSAIDs and
34   Cox-2 selective drugs with respect to symptoms, function and quality of life in
35   patients with a recent onset of rheumatoid arthritis, and in established
36   disease. Trials on Cox-2 selective drugs were selected in accordance with the
37   criteria used for the NICE HTA on Cox-2s233 as it was within the remit of this
38   guideline to update the HTA. All trials published after the cut-off date used in
39   the HTA were considered as evidence. The criteria used in the TA were as
40   follows:
41           1. Cox-2s: celecoxib, rofecoxib, meloxicam and etodolac
42           2. Comparators: Placebo, NSAIDs, other Cox-2s
43           3. Doses: Licensed doses per day - celecoxib (200-400 mg), rofecoxib
44           (not specified for RA), meloxicam (15 mg) and etodolac (600 mg)
45           4. Treatment length: ≥ 4 weeks
46           5. Trial type: RCTs / SR / MA
47           6. Trial size: N >50 in each arm
48           7. Population: RA (any duration), any country
49

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              112 of 307
     DRAFT FOR CONSULTATION

 1   Papers looking at rofecoxib were excluded as evidence because rofecoxib is
 2   no longer licensed in the UK.
 3
 4   Due to the large volume of evidence on NSAIDs and other Cox-2s not listed in
 5   the TA, only RCTs/MA were selected which had been published in 1997
 6   onwards, were of a UK-relevant population, if the population was mixed
 7   arthritis there had to be >75% RA or RA subgroup analysis and looked at the
 8   following comparisons: NSAIDs or Cox-2s vs placebo, analgesics or NSAIDs
 9   (different or same NSAIDs including diff routes of administration and diff
10   doses). Some papers on COX-2s also report NSAID vs placebo arms and
11   thus these have been included in the NSAIDs evidence.
12
13   Twelve studies (11 RCTs and 1 extension of an RCT)234-246 were found that
14   fulfilled the criteria. One of the RCTs was published as two separate
15   papers243,245 reporting different outcomes and so the trial has only been
16   counted once. However results from both papers are reported and referenced
17   here. One of the studies246 was an extension study of an included trial238 and
18   was included as additional evidence. Two of the studies235,236 were excluded
19   due to methodological limitations.
20
21   NOTE: all trials found looked at patients with established RA patients, no
22   studies were found for patients with a recent onset of RA.
23
24
25   Cox-2 selective drugs (TA update)
26
27   1 MA235 and one RCT236 were found which fulfilled the criteria to update the
28   NICE TA on Cox-2s in RA patients. Both of these studies235,236 were excluded
29   due to methodological limitations and thus there were no papers included to
30   update the TA.
31
32   Other Cox-2s
33
34   Two RCTs237,238 and one extension study of an RCT246 were found which
35   looked at other Cox-2s in RA patients. Both RCTs were methodologically
36   sound randomised, parallel group studies with a 12-week treatment phase.
37   The first RCT237 looked at Cox-2s in N=891 patients with established RA. The
38   trial compared 3 different treatment arms (etoricoxib 90 mg, naproxen 1000
39   mg and placebo). The naproxen vs placebo arm will be reported in the
40   NSAIDs section. The second RCT238 looked at Cox-2s in N=816 patients with
41   established RA. The trial compared 3 different treatment arms (etoricoxib 90
42   mg, naproxen 1000 mg and placebo). The naproxen vs placebo arm will be
43   reported in the NSAIDs section. The extension study reported results of
44   patients who remained on etoricoxib 90 mg vs naproxen 1000 mg for the full
45   trial and extension period (121 weeks).
46
47   NSAIDs
48   9 RCTs234,237-245 were found which looked at NSAIDs in RA patients. Three of
49   these RCTs234,237,238 have been included in the Cox-2 sections, but they
50   additionally looked at NSAIDs vs placebo, and therefore these comparisons
51   are included as evidence in this section. One of the RCTs was published as
52   two separate papers243,245 reporting different outcomes and so the trial has
     Rheumatoid arthritis: full guideline DRAFT (August 2008)            113 of 307
     DRAFT FOR CONSULTATION

 1   only been counted once. However results from both papers are reported and
 2   referenced here.
 3
 4   All 9 included RCTs were methodologically sound randomised, double blind,
 5   fairly large (range: N=346 to N=1149) parallel group studies comparing
 6   naproxen 100 mg (500 mg twice/day) vs placebo in patients with established
 7   RA in a 12 week treatment phase (except for 1 RCT – Geussens et al.234
 8   which had a 26 week treatment phase and 1 RCT - Krug et al.242 which
 9   compared naproxen vs nabumetone). Some of the studies also had Cox-2
10   arms which were either reported as evidence in the previous Cox-2 sections
11   or consisted of Cox-2s that are not licensed in the UK.

12   7.1.10            Health economic methodological introduction
13           •    In the Brown et al HTA report247 a systematic review of economic
14                evaluations of Cox-2 drugs was performed. The nine papers found in
15                this report were appraised..248-256 A search was performed looking for
16                economic evaluations published subsequent to the HTA report but no
17                papers were found.

18   7.1.11            Clinical evidence statements
19   Cox-2 selective drugs (TA update)
20
21   No papers were included to update the NICE TA.
22
23   Other Cox-2s
24
25       •       Symptoms
26
     Study            Treatment      Follo    Outcomes                          Result -
                                     w-up                                       best
                                                                                treatment
     2                Etoricoxib 90 12    tender and swollen joint count,       Etoricoxib
     RCTs237,238      mg vs         weeks Pain (VAS) and ACR20                  (1 RCT all
     Level 1+         placebo             completers                            p<0.001, 1
                                                                                RCT all
                                                                                p<0.01).
     1 RCT            Etoricoxib 90 12    Tender and swollen joint count,       Etoricoxib
     and its          mg vs         weeks Pain (VAS) and ACR20                  (all p<0.01
     extension        naproxen            completers at 12 weeks                except
     study238,246                                                               swollen
     Level 1+                                                                   joints
                                                                                p=0.05)
                                     121   Tender and swollen joint count       NS
                                     weeks
27
28       •       Function
29
     Study            Treatment      Follo    Outcomes                          Result -
                                     w-up                                       best
                                                                                treatment

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                114 of 307
    DRAFT FOR CONSULTATION

    2              Etoricoxib 90 12    HAQ score / modified HAQ         Etoricoxib
    RCTs237,238    mg vs         weeks score                            (1 RCT
    Level 1+       placebo                                              p<0.001, 1
                                                                        RCT
                                                                        p<0.01)
    1 RCT237       Etoricoxib 90 12    HAQ score                        NS
    Level 1+       mg vs         weeks
    1 RCT 238      naproxen      12    modified HAQ score               Etoricoxib
    Level 1+                     weeks                                  (p<0.01)
1
2       •   Global assessment
3
    Study          Treatment    Follo   Outcomes                        Result -
                                w-up                                    best
                                                                        treatment
    2              Etoricoxib 90 12    Patient’s and investigator’s     Etoricoxib
    RCTs237,238    mg vs         weeks global assessment of disease     (p<0.001)
    Level 1+       placebo             activity

    1 RCT237       Etoricoxib 90 12    Patient’s and investigator’s     NS
    Level 1+       mg vs         weeks global assessment of disease
                   naproxen            activity

    1 RCT                       12    Patient’s and investigator’s      Etoricoxib
    and its                     weeks global assessment of disease      (p<0.01)
    extension                         activity
    study238,246                121   Patient’s and investigator’s      NS
    Level 1+                    weeks global assessment of disease
                                      activity
4
5       •   Biochemical markers
6
    Study          Treatment    Follo   Outcomes                        Result -
                                w-up                                    best
                                                                        treatment
    2              Etoricoxib 90 12    CRP                              Etoricoxib
    RCTs237,238    mg vs         weeks                                  (1 RCT
    Level 1+       placebo                                              p<0.05, 1
                                                                        RCT
                                                                        p<0.01)
    2              Etoricoxib 90 12    CRP                              NS
    RCTs237,238    mg vs         weeks
    Level 1+       naproxen
7
8       •   Withdrawals
9
    Study          Treatment    Follo   Outcomes                  Result - best
                                w-up                              treatment




    Rheumatoid arthritis: full guideline DRAFT (August 2008)          115 of 307
    DRAFT FOR CONSULTATION

    1 RCT237       Etoricoxib 90 12    Withdrawals due to lack of    Etoricoxib (1
    Level 1+       mg vs         weeks efficacy; total number of     RCT p<0.001, 1
                   placebo             withdrawals                   RCT p<0.01);
                                                                     similar

    1 RCT238                             total number of             Etoricoxib
    Level 1+                             withdrawals                 (better)

    2                                    Withdrawals due to AEs      NS
    RCTs237,238
    Level 1+
    1 RCT237       Etoricoxib 90 12    Withdrawals due to lack of    NS; Similar;
    Level 1+       mg vs         weeks efficacy; total number of     Similar
                   naproxen            withdrawals; withdrawals
                                       due to AEs
    1 RCT 238                    12    Withdrawals due to lack of    Etoricoxib
    Level 1+                     weeks efficacy; total number of     (p<0.01; better)
                                       withdrawals
1
2       •   AEs
3
    Study          Treatment     Follo   Outcomes                          Result -
                                 w-up                                      best
                                                                           treatment
    1 RCT237       Etoricoxib 90 12    Number of patients with SAEs        NS
    Level 1+       mg vs         weeks                                     Similar
                   placebo             GI nuisance symptoms                Placebo
                                                                           better
                                         Number of patients with drug-
                                         related AEs and hypertension
                                         AEs

    1 RCT 238                            Number of patients with drug-     NS
    Level 1+                             related AEs and for SAEs
                                                                           Similar
                                         Dyspepsia AEs                     Placebo
                                                                           better
                                         hypertension AEs

    1 RCT237       Etoricoxib 90 12    number of patients with: drug-      Similar
    Level 1+       mg vs         weeks related AEs, SAEs, GI
                   naproxen            nuisance symptoms and
                                       hypertension AEs
    1 RCT                        12    Dyspepsia AEs and                   Similar
    and its                      weeks Hypertension AEs
    extension                    121   total number of AEs                 Similar
    study238,246                 weeks
    Level 1+
4

5   NSAIDs
6
    Rheumatoid arthritis: full guideline DRAFT (August 2008)             116 of 307
    DRAFT FOR CONSULTATION

1      •    Symptoms
2
    Study          Treatment     Follo Outcomes                        Result - best
                                 w-up                                  treatment
    2              Naproxen vs   12    tender and swollen joint        Naproxen (1
    RCTs237,238    placebo       weeks count, Pain (VAS) and           RCT all
    Level 1+                           ACR20 completers                p<0.001
                                                                       except swollen
                                                                       joints p<0.05;
                                                                       1 RCT p<0.01)
    1 RCT234                             Tender joint count, ACR20     Naproxen (all
    Level 1++                            responder index, Pain         p<0.05)
                                         (VAS), Morning stiffness

                                         Swollen joint count           NS
    1 RCT240                     26      Swollen joint count           Naproxen
    Level 1+                     weeks                                 (p<0.05)
                                 13      Tender joint count; Pain      Naproxen
                                 and     (VAS)                         (p<0.05; p<0
                                 26                                    01)
                                 weeks
    1 RCT243,245                 12      Swollen joint count, ACR20    Naproxen (all:
    Level 1+                     weeks   responders, Pain (VAS) and    p<0.05)
                                         Morning stiffness
    1 RCT239                     12      ACR20, reduction in the       Naproxen (p≤
    Level 1+                     weeks   number of tender/painful      0.01; p<0.001)
                                         joints, Tender /Painful
                                         Joints; Pain (VAS), Morning
                                         stiffness
    1 RCT241                     12      Tender and painful joint      Naproxen (all
    Level 1+                     weeks   count, tender and painful     p≤ 0.001)
                                         joint score, swollen joint
                                         count and swollen joint
                                         score, ACR20 responders,
                                         ACR-N; Pain (VAS);
                                         Morning stiffness
    1 RCT244                     12      ACR20 responders, Tender      Naproxen (p≤
    Level 1+                     weeks   /Painful Joint Score;         0.001; p=0.03)
                                         Reduction in the number of
                                         tender/painful joints
    1 RCT242       Naproxen vs   12      Change in number of           NS
    Level 1+       nabumetone    weeks   tender, swollen and painful
                                         joints; Pain (VAS) and
                                         AIMS2 dimensions, Clinical
                                         change in number of joints
                                         involved (≥50% reduction)
                                         and Clinical change in
                                         number of tender, swollen
                                         and painful joints (≥50%
                                         reduction) and RADAR
                                         dimensions

    Rheumatoid arthritis: full guideline DRAFT (August 2008)           117 of 307
     DRAFT FOR CONSULTATION

1
2         •       Function
3
     Study              Treatment     Follo Outcomes                      Result - best
                                      w-up                                treatment
     4              Naproxen vs       12    HAQ or modified HAQ           Naproxen (2
     RCTs 237,238,2
                    placebo           weeks                               RCTs
     41,243,245
                                                                          p≤0.001; 1
     Level 1+                                                             RCT p<0.01; 1
                                                                          RCT p<0.05)
     1 RCT240                         13                                  Naproxen
     Level 1+                         and                                 (p<0.05)
                                      26
                                      weeks
4
5         •       Global assessment
6
     Study              Treatment     Follow-    Outcomes                 Result - best
                                      up                                  treatment
     7              Naproxen vs       All 12     Patient’s and            Naproxen (3
     RCTs 237,238,2
                    placebo           weeks      investigator’s global    RCTs
     40 239 241,244
                                                 assessment of disease    p≤0.001; 1
     Level 1+                                    activity                 RCT p<0.01; 1
     and 234                                                              RCT p<0.05; 1
     Level 1++                        (except                             RCT p<0.01
                                      Geusens                             13 weeks and
                                      240
                                          – 13                            p<0.05 26
                                      and 26                              weeks)
                                      weeks)

     1 RCT243,245                     12 weeks   Patient’s and            NS
     Level 1+                                    investigator’s global
                                                 assessment of disease
                                                 activity
     1 RCT242           Naproxen vs   12 weeks   Patient’s and            ns
     Level 1+           nabumetone               investigator’s global
                                                 assessment of disease
                                                 activity
7
8         •       QoL
9
     Study              Treatment     Follow-    Outcomes                 Result - best
                                      up                                  treatment
     1 RCT241           Naproxen vs   12 weeks   PTSS, SF-36 Physical     Naproxen
     Level 1+           placebo                  (all domains except      (p≤0.001;
                                                 general health); SF-36   p≤0.001;
                                                 Mental (all domains      p<0.01)
                                                 except role-emotional)

     1 RCT243,245                     12 weeks   SF-36 Physical (all      Naproxen
     Level 1+                                    domains); SF-36 Mental   (p<0.01;
                                                 (all domains)            p<0.05)
10
     Rheumatoid arthritis: full guideline DRAFT (August 2008)             118 of 307
    DRAFT FOR CONSULTATION

1        •       Biochemical markers
2
    Study             Treatment   Follow-     Outcomes                 Result - best
                                  up                                   treatment
    1 RCT238                      12 weeks    CRP                      Naproxen
    Level 1+                                                           (p<0.01)
    7             Naproxen vs     All 12      CRP                      NS
    RCTs237,239,2 placebo         weeks
    40 241,243-245

    Level 1+
    and 234
    Level 1++                     (except
                                  Geusens
                                  240
                                      – 13
                                  and 26
                                  weeks)

3
4        •       Withdrawals
5
    Study             Treatment   Follow- Outcomes                     Result - best
                                  up                                   treatment
    5             Naproxen vs     12      Withdrawals due to lack of   Naproxen (all
    RCTs237,243,2 placebo         weeks   efficacy                     SS except
    45 239,241
                                                                       Gibofsky =
    Level 1+                                                           better)
    and 234
    Level 1++
    1 RCT240                      26                                   Naproxen
    Level 1+                      weeks                                (better)
    2                             12         Total number of           Naproxen
    RCTs238,243,2                 weeks      withdrawals               (better)
    45

    Level 1+
    1 RCT237                                                           Similar
    Level 1+
    1 RCT240                      26                                   Naproxen
    Level 1+                      weeks                                (better)
    4                             12         Withdrawals due to AEs    NS
    RCTs237,238,2                 weeks
    43,245
           Level
    1+ and 234
    Level 1++
    1 RCT241                                                           Placebo
    Level 1+                                                           (better)
    1 RCT240                      26                                   Naproxen
    Level 1+                      weeks                                (better)
    1 RCT243,245                  12         Withdrawals due to GI     Placebo
    Level 1+                      weeks      AEs                       (better)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           119 of 307
    DRAFT FOR CONSULTATION

    1 RCT242        Naproxen vs   12         Withdrawals due to lack of   Similar
    Level 1+        nabumetone    weeks      efficacy                     NS
                                             Withdrawals due to
                                             treatment-related AEs and
                                             Total withdrawals
1
2
3         •   AEs
4
    Study           Treatment     Follow-     Outcomes                    Result - best
                                  up                                      treatment
    3               Naproxen vs   12 weeks    Drug-related AEs and        NS
    RCTs237,238     placebo                   SAEs
    Level 1+
    and 234
    Level 1++
    3                                         Total AEs                   Placebo
    RCTs239,243-                                                          (better)
    245

    Level 1+
    2                                         Dyspepsia AEs               Placebo
    RCTs238,241                                                           (better)
    Level 1+
    1 RCT234                                  1 or more AEs               NS
    Level 1++
    1 RCT241                      12 weeks    Percentage of patients      Similar
    Level 1+                                  with AEs
    1 RCT240                      26 weeks                                Placebo
    Level 1+                                                              (better)
    3 RCTs237-                    12 weeks    Hypertension AEs            Placebo
    239
                                                                          (better)
    Level 1+
    3                                                                     NS or similar
    RCTs241,243,2
    45

    Level 1+
    and 234
    Level 1++
    1 RCT240                      26 weeks                                Placebo
    Level 1+                                                              (better)
    1 RCT237                      12 weeks    GI AEs or nuisance          Similar
    Level 1+                                  symptoms
    2 RCTs243-                                                            Placebo
    245
                                                                          (better)
    Level 1+
    1 RCT240                      26 weeks                                Placebo
    Level 1+                                                              (better)
    1 RCT242        Naproxen vs   12 weeks    Number of patients with     Similar; NS
    Level 1+        nabumetone                ≥1 AE; Serious GI AEs
5
6         •   Use of rescue medication

    Rheumatoid arthritis: full guideline DRAFT (August 2008)              120 of 307
    DRAFT FOR CONSULTATION

    Study       Treatment      Follow-     Outcomes            Result - best
                               up                              treatment
    1 RCT234    Naproxen vs    12 weeks    Use of rescue       NS
    Level 1++   placebo                    medication

    1 RCT240                   26 weeks    Use of rescue       NS
    Level 1+                               medication
    1 RCT242    Naproxen vs    12 weeks    Use of rescue       NS
    Level 1+    nabumetone                 paracetamol
1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   121 of 307
    DRAFT FOR CONSULTATION




1   7.1.12       Health economic evidence statements
    COX-2 Inhibitors - taken from the Brown HTA Report and the NICE TA Cox-2 Assessment Report
    Study       Country       Comparators Patients Time       Model      ICER         Conclusions
                                                    horizon used
    Svarvar     Norway        Celecoxib vs OA and 1 year      ACCES      Per GI       Celecoxib dominates
        249
    2000                      NSAID         RA                Decision   event
                                                              Analytic   averted, per
                                                              model      LYG
    Chancellor Switzerland Celecoxib vs Arthritis 6           Decision   Per adverse Celecoxib dominates
        255
    2001                      NSAID                 months Analytic      event
                                                              with Monte
                                                              Carlo
    You         Hong Kong Celecoxib vs OA and 6               Decision   Expected     Celecoxib has lowest
        252
    2002                      NSAID         RA      months Analytic      Cost         expected cost
    El-Serag    USA           Celecoxib vs Cox-I    1 year    Decision   Per UGI      Cox-2 are dominant in high-
        256
    2002                      NSAID         users             Analytic   event        risk patients
    Spiegel     USA           Celecoxib or  OA and Lifetime Decision     Per QALY     Rofecoxib and celecoxib
        248
    2003                      rofecoxib vs  RA                Analytic                are cost-effective in high-
                              naproxen                                                risk patients (US$55k)
                                                                                      Average risk US$275k
    Zabinski    Canada        Celecoxib vs OA and 6           Decision   Expected     Celecoxib has lower
        253
    2001                      NSAID         RA      months Analytic      cost         expected cost than
                                                                                      NSAID+H2RA,
                                                                                      NSAID+misoprostol,
                                                                                      NSAID+PPI but more costly
                                                                                      than NSAID
    Maetzel     Canada        Naproxen vs   OA and 5 years Markov        Per QALY     Average risk
    2003251                   Rofecoxib,    RA                                        Celecoxib and Rofecoxib


    Rheumatoid arthritis: full guideline DRAFT (August 2008)    122 of 307
DRAFT FOR CONSULTATION



                          Diclofenac vs                                                   unlikely to be cost-effective
                          Celecoxib,                                                      (Can$271k and Can$125k
                          Ibuprofen vs                                                    respectively)
                          celecoxib.                                                      High risk
                                                                                           Rofecoxib dominates
                          High risk are                                                   naproxen +PPI
                          same as                                                         Celecoxib dominates
                          above + PPI                                                     ibuprofen +PPI
Fendrick     US           Cox-2 vs         Long      1 year   Markov          Per         US$32k per symptomatic
2002254                   generic          term                               symptomatic ulcer avoided
                          NSAID            NSAID                              ulcer
                          switched to      users                              avoided     US$57k per complicated
                          safer NSAID                                                     ulcer avoided
                          after AE

                          Cox-2 vs
                          safer NSAIDs
                          as first line
Brown        UK           Cox-2 vs         OA and    6        Probabilistic   Per ulcer      Mean ICER: £301 per
2006247                   Cox-I, along     RA        months   Decision        avoided, per   endoscopic ulcer avoided
                          with principal                      Analytic        serious GI     £22,843 per serious GI
                          GPA                                 model           event          event averted
                          strategies                                          avoided and    £12742 per LYG
                          (PPI, H2RA                                          per LYG
                          and
                          misoprostol)
Non selective NSAIDs
McCabe      UK            Nabumetone,      OA and    3        Decision        Per LYG        Co prescription after minor
1998250                   Ibuprofen        RA        months   tree                           AE: £2517per LYG



Rheumatoid arthritis: full guideline DRAFT (August 2008)         123 of 307
    DRAFT FOR CONSULTATION




                                                                            Switching after minor AE:
                                                                            £1880 per LYG
1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   124 of 307
     DRAFT FOR CONSULTATION


 1   Of the ten papers appraised the only UK Cox-2 economic evaluation was
 2   Brown et al247, and the only non-selective NSAID economic evaluation was
 3   McCabe et al.250 Direct comparison of the ICERs is impossible due to
 4   differences in comparator arms, model assumptions, cost data and health
 5   system setting. Three studies248,251,256 concluded that Cox-2s are cost
 6   effective for patients at a high risk of a gastrointestinal event. Two
 7   papers249,255 concluded that Cox-2s are dominant therapies over non selective
 8   NSAIDs. Two papers248,251determine that Cox-2s are not cost effective for
 9   patients with a normal risk of a gastrointestinal event. None of the papers
10   provide cost effectiveness analysis for an RA only population.

11   7.1.13         Summary of evidence statements and tables
12        •   NSAIDs and COX2 inhibitor drugs are useful in reducing symptoms
13            of RA (tender and swollen joint counts, pain (VAS), morning stiffness,
14            withdrawals due to lack of efficacy, achievement of ACR20,
15            improvement in function, patients and investigators global
16            assessment of disease activity) 234,237-239,241,243-245
17        •   Trials on large numbers of RA patients on NSAIDs and COX2
18            inhibitor drugs are available, but often over relatively short periods of
19            time, and in patients satisfying a number of exclusion criteria that
20            makes them only partially representative of RA patients in clinic234,237-
              239,241,243-245
21
22        •   In selected populations eligible to enter NSAID and COX2 trials, the
23            drugs are generally well tolerated234, though increased hypertension
24            and dyspepsia were reported in the active arms of some trials237-
              239,241,243-245
25
26        •   There is no consistent evidence of differences in efficacy when
27            NSAIDs are compared with other individual NSAIDs or COX2 inhibitor
28            drugs237,242

29   7.1.14          From evidence to recommendations
30   There is clear evidence to show that both NSAIDs and COX2 inhibitors are
31   effective in treating the symptoms of RA. The benefits obtained from these
32   drugs need to be balanced against their adverse effects, mainly on the gastro-
33   intestinal and cardiovascular systems. It was felt that whilst the data did not
34   suggest any consistent differences between NSAID and COX2 inhibitor for
35   efficacy, there were differences in toxicity profiles. There will be some patients
36   in whom these drugs are contraindicated, and others in whom they should
37   only be used with caution if they are clearly efficacious. In all cases, the GDG
38   felt that these drugs should be used in the lowest effective dose over the
39   shortest period of time, and on a “when necessary” basis rather than regularly.
40   In clinical practice, although there was no specific clinical trial evidence to
41   support this, strategies to minimise the use of these drugs were considered to
42   be important.
43
44   In symptomatic patients, the emphasis needs to be on monitoring disease
45   activity (see section 8.1), and if patients are requiring high doses of NSAIDS

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               125 of 307
     DRAFT FOR CONSULTATION


 1   or COX2 inhibitors on a regular basis, this should be regarded as an indication
 2   that disease modifying therapy may not be working satisfactorily, and that
 3   changes in treatment might be needed.
 4
 5   The GDG was cognisant of the changes that had been made to the NICE
 6   Technology Appraisal Guidance on COX2 inhibitors for patients with
 7   osteoarthritis8, and bore these in mind when updating this guidance for RA
 8   patients. It was felt that, in comparing the RA and OA populations, it was likely
 9   that gastrointestinal risks would be similar (unless the risk had been increased
10   by patients being co-prescribed steroids) but that cardiovascular risks would
11   be greater in the RA population (see section 8.2). The extensive health
12   modelling which had taken place for the OA Guideline was regarded as
13   unlikely to be very different for a RA population, and there should therefore be
14   a similar recommendation about the potential benefit of co-prescribing protein
15   pump inhibitors, which had been an important element of the cost-
16   effectiveness analysis.
17
18   The GDG also felt that there should be a recommendation for RA patients
19   similar to that for OA patients already receiving concomitant low-dose aspirin,
20   in that other analgesics should be considered before giving NSAIDs or COX2
21   inhibitors.

22   7.1.15        Recommendations
23   R 18 NSAIDs and COX2 inhibitors, unless they are specifically
24   contraindicated, should be used for the symptomatic treatment of people with
25   RA in the lowest effective dose and for the shortest duration of time on a
26   ’when necessary‘ basis.
27
28   R 19 If NSAIDs or COX2 inhibitors are not providing satisfactory symptom
29   control, the disease-modifying or biological drug regimen should be reviewed
30   in order to determine whether any modifications are needed.
31
32   R 20 A proton pump inhibitor** should be co-prescribed with NSAIDs or COX2
33   inhibitors for people with RA.
34
35   R 21 In a person with RA who needs to take concomitant low-dose aspirin,
36   and in whom pain relief is ineffective or insufficient, other analgesics should
37   be considered before substituting or adding an NSAID or COX2 inhibitor.

38   7.2    Glucocorticoids

39   7.2.1 Clinical introduction
40   Glucocorticoids in recent onset and established RA
41

     **
      This is based on health economic modelling with generic omeprazole at November 2007
     NHS drug tariff costs; therefore the cheapest available PPI should be used


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                126 of 307
     DRAFT FOR CONSULTATION


1    Glucocorticoids have been used in the management of RA for over 50 years.
2    When first introduced, enthusiasm for the efficacy of these drugs was
3    tempered by the severe side effects of the high dose regimes that were used.
4    Clinicians and patients continue to approach these drugs with caution.
5    However, some aspects of steroid use court little controversy:
6          • Intra-articular injections have a limited evidence-base, but their users
7             and receivers can testify that they are extremely useful for a flare in
8             one or more joints.
 9        •   For polyarticular flares, or at first presentation of the disease,
10            intramuscular/intra-articular, or short oral courses of steroids, can
11            decrease symptoms whilst waiting for other slower acting drugs to
12            take effect.
13        •   For severe extra-articular manifestations, intravenous steroids can
14            save critical organs (e.g. eyes in scleritis) or even life-threatening
15            complications on occasions (e.g. severe serositis or vasculitis),
16            though they should be used with immunosuppressives such as
17            cyclophosphamide.
18   Other questions are answered less easily:
19        • Should recent onset RA patients be treated with some form of
20           steroids (oral or intramuscular)?
21        •   Do the benefits of steroids out-weigh the disadvantages?
22        •   Do steroids have a lasting impact on symptoms, function of joints and
23            quality of life?
24        •   Should steroids be classified as disease modifying drugs?

25   7.2.2 Clinical methodological introduction
26   We looked for studies that investigated the efficacy and safety of
27   corticosteroids with respect to symptoms, joint damage, function and quality of
28   life in patients with a recent onset of RA or in patients with established RA.
29   Due to the large volume of evidence, only RCTs were selected which had a
30   sample size of N>50 and compared corticosteroids alone vs placebo or
31   corticosteroids in combination with a DMARD vs DMARD. Only studies that
32   compared steroids with placebo, or steroids plus DMARDs with DMARDs
33   were included in order to attempt to tease out the influence of steroids alone.
34   However, it should be noted that one study compared steroids with placebo in
35   the absence of other DMARDs, and that the other studies compared steroids
36   plus a DMARD with a DMARD alone. In some of these studies the DMARDs
37   were fixed by protocol and in others the concomitant DMARDs were left to the
38   treating physician. Disease activity inclusion criteria varied between studies.
39   Some studies included patients with up to 1 year of RA and others up to two
40   years. Consequently, because of all of these differences, pooling results
41   needs to be performed with caution.
42
43   Early disease was defined as two years or less and established disease as
44   more than two years. In order to be included in the established disease trials,
45   patients had to have active disease, but this had different definitions in


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               127 of 307
     DRAFT FOR CONSULTATION


 1   different trials. It was noted that one trial 257 used penicillamine in both arms,
 2   and another 258 used gold injections. Neither of these DMARDs are
 3   particularly popular currently because of toxicity concerns. There is no
 4   available evidence for much more popular DMARDs such as methotrexate or
 5   sulfasalazine. Papers assessing corticosteroids which were published in the
 6   1950’s were not included in the evidence because the databases searched
 7   only commence with publications from 1966 onwards.
 8
 9   Recent onset of RA:
10   Six RCTs259-266were found that fulfilled the criteria. Two of these RCTs were
11   each published as two separate papers reporting different outcomes and so
12   each trial has only been counted once, however results from both papers are
13   reported and referenced here.
14
15   All six trials were methodologically sound randomised, parallel group studies
16   using the oral corticosteroid prednisolone, but they differed with respect to the
17   following:
18          • Sample size (range: N=81 to N=259)
19        •   Blinding (5 RCTs double blind, 1 RCT unblinded)
20        •   Trial length (5 RCTs 2 years, 1 RCT 12 weeks; follow-up ranged from
21            immediate to 1 year post-treatment)
22        •   Treatment (1 RCT corticosteroid vs placebo, 5 RCTs corticosteroid +
23            DMARD vs DMARD)
24        •   Treatment regimen - dose
25
26   Established RA:
27   Five RCTs267 258 257 268,269 were found that fulfilled the criteria.
28
29   All five trials were methodologically sound randomised, parallel group studies
30   using corticosteroids, but they were very variable and had variable inclusion
31   criteria. The trials differed with respect to the following:
32          • Sample size (range: N=59 to N=137)
33        •   Blinding (3 RCTs double blind, 1 RCT single blind, 1 RCT
34            unblinded)\par
35        •   Trial length (range: single injection to 1 year; follow-up ranged from
36            immediate to 7 months post-treatment)
37        •   Treatment (1 RCT corticosteroid vs placebo, 3 RCTs corticosteroid +
38            DMARD vs DMARD)
39        •   Corticosteroid used (2 RCTs methylprednisolone, 1 RCT
40            prednisolone, 1 RCT depomedrone, 1 RCT rimexolone)
41        •   Route of corticosteroid administration (2 RCTs IM, 1 RCT IA, 1 RCT
42            IV, 1 RCT oral)
43        •   Treatment regimen - dose
44

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                128 of 307
     DRAFT FOR CONSULTATION


1    7.2.3 Health economic methodological introduction
2    Two studies were identified and appraised. Bae et al270 is a US based cost-
3    utility analysis of low-dose corticosteroids versus Cox-2 inhibitors and non-
4    selective NSAIDs. Verhoeven et al271is Dutch based cost-utility analysis of
5    step-down prednisolone, sulphasalazine and methotrexate versus
6    sulphasalazine.

7    7.2.4 Clinical evidence statements
 8   Recent onset of RA
 9
10   Symptoms and quality of life
11
12        •   Three RCTs (1 RCT261,262 1+; 263 1 RCT259 Level 1++) showed no
13            improvement in pain at 1 year or 2 years for oral prednisolone (and at
14            year 3 for the Hickling and Kirwan trial Level 1+).
15        •   One RCT263 showed sustained decreases of disease activity
16            (DAS28) at 6, 12 and 24 months in the prednisolone arm, and a
17            greater proportion of patients in DAS28 remission at the end of two
18            years, but not at the end of 1 year. Level 1+
19        •   Three RCTs (1 RCT261,262 Level 1+; 1 RCT264,265 Level 1++; 1 RCT
              266
20                Level 1+) demonstrated clinical and laboratory measures of joint
21            inflammation, with the exception of joint tenderness in the Van
22            Everidgen study, had not improved at 1 or 2 years. However, the
23            latter study showed no overall patient improvement or significant
24            changes in quality of life measures.
25        •   Three RCTs addressed adverse events and withdrawals over a two
26            year period (1 RCT 266 Level 1+; 1 RCT259 Level 1++; 263 Level 1++)
27            and found similar results for prednisolone versus placebo, although
28            Cappell found more withdrawals due to adverse events in the steroid
29            arm, and Wassenberg more withdrawals due to drug failure in the
30            steroid arm.
31        •   Two RCTs (1 RCT264,265 Level 1++; 263 Level 1+) showed that in the
32            prednisolone arms concomitant medication could be reduced.
33            Patients in the Van Everidgen and Svensson prednisolone arms were
34            able trial were able to decrease intra-articular steroids, and in the 263
35            trial were able to decrease their NSAIDs.
36
37   Joint Damage
38
39   Four RCTs (1 RCT261,262 Level 1+; 1 RCT 263 Level 1+; 1 RCT264,265 Level
40   1++; 1 RCT 266 Level 1+) showed decreased radiological damage (proportion
41   of erosions, progression of radiological damage) in the oral prednisolone arm
42   versus the placebo over 2 years. However, in only the Wassenberg and
43   Svensson trials was this divergence evident at 1 year. Joint space narrowing
44   failed to improve in the prednisolone arms of Wassenberg and Svensson, but
45   other components of composite damage indices did improve. In Kirwa Level


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               129 of 307
     DRAFT FOR CONSULTATION


 1   1+ the difference in erosions on hand x-rays remained significant at 3 years (1
 2   year after the withdrawal of steroids), but the progression of radiological
 3   damage was no different.
 4
 5   One RCT259 failed to show an improvement in radiological damage in the
 6   prednisolone arm at 1 or 2 years. Level 1++
 7
 8   Function
 9   Three RCTs (1 RCT264,265 Level 1++; 1 RCT259 Level 1++; 1 RCT 266 Level
10   1+) showed that functional scores were no different at 1 and 2 years in the
11   prednisolone arm. Van Everidgen showed an improvement in grip strength at
12   1 year on prednisolone, but this difference was lost at two years.
13
14   1 RCT261,262 showed no functional improvements in HAQ for budenoside 3mg
15   or 9mg at 3 months compared with placebo. Prednisolone 7.5.mg showed an
16   improvement in HAQ at 3 months, but this was not sustained at 1 or 2 years.
17   Level 1+
18
19   One RCT263 did show significant improvements in the prednisolone arm for
20   HAQ and Signals of Functional Impairment (SOFI index) at 6, 12 and 24
21   months. Level 1+
22
23   Established RA
24
25   Symptoms and quality of life
26
27        •   One RCT267 compared monthly depomedrone to placebo in addition
28            to the usual DMARD and found decreased DAS and swollen joints at
29            6 months, but both patient and physician’s global assessments were
30            no different. More patients withdrew from the placebo arm because of
31            lack of efficacy. Symptomatic benefits were not sustained at 1 or 2
32            years. No decrease in ESR occurred. Although the steroid arm had
33            more adverse events, there were no greater withdrawals from the trial
34            for these. Level 1++
35        •   One RCT258 looked at IM methylprednisolone at 0, 4 and 8 weeks of
36            initiation of IM gold therapy compared with placebo. Pain and joint
37            counts improved at 12 weeks (4 weeks post last injection), but not at
38            24 weeks. Joint counts improved at 12 weeks but not 24 weeks. ESR
39            decreased at 4 weeks in the steroid group but was not sustained
40            beyond this. Index of Disease Activity improved up to 12 weeks but
41            not at 24 weeks. The injections were well tolerated with no greater
42            number of withdrawals in the steroid arm. Level 1+
43        •   One RCT257 looked at varied doses of oral prednisolone given in a
44            tapered regime, starting at 30mg, and tailing this down to a dose
45            between 2.5mg and 15mg (determined by the patient) to control
46            disease. Although there were greater numbers of patients with more
47            than 20% and 50% clinical improvement at 3 months, this was not
48            sustained at 6 or 12 months. No improvements in ESR or CRP
49            occurred. Level 1+


     Rheumatoid arthritis: full guideline DRAFT (August 2008)            130 of 307
     DRAFT FOR CONSULTATION


1         •   One RCT260 compared 6 IV methylprednisolone injections (15mg/kg)
2             over 20 weeks with placebo and found no difference for a variety of
3             symptom outcomes after 20 weeks or 1 year (7 months post-
4             treatment). Level 1+
 5        •   One RCT268 found that tender and swollen joint counts significantly
 6            improved at 4 weeks for oral budesonide 3mg and 9 mg and oral
 7            prednisolone 7.5mg when added to usual treatment compared with
 8            placebo and usual treatment (all p<0.05). At 12 weeks, along with
 9            other measures of disease activity and quality of life (physicians and
10            patients global assessments, ACR20, SF-36 physical and mental
11            condition), this was only sustained for budesonide 9mg and
12            prednisolone 7.5mg (p<0.05). Pain and morning stiffness did not
13            improve at 12 weeks for oral budesonide 3mg and 9 mg. For
14            prednisolone 7.5mg, pain improved at 12 weeks (p<0.001) but
15            morning stiffness did not. CRP and ESR only improved at 12 weeks
16            on prednisolone (both p<0.001). Whilst the numbers of patients with
17            adverse events were similar to placebo, all three steroid arms had
18            greater numbers of adverse events than placebo, but with no greater
19            withdrawals. Level 1++
20        •   One RCT269 looked at different doses of intra-articular rimexolone
21            into RA knees and for higher doses found symptomatic benefit
22            compared to placebo for up to 84 days. The injection was well
23            tolerated. Level 1+
24
25   Joint Damage
26
27        •   One RCT267 compared monthly depomedrone to placebo in addition
28            to the usual DMARD, and found decreased radiological damage at 2
29            years in the steroid arm (analysis used % change from baseline due
30            to baseline differences – however when analysed using change in
31            actual scores there was NS between the groups). Level 1+
32        •   One RCT258 looked at IM methylprednisolone at 0, 4 and 8 weeks of
33            initiation of IM gold therapy compared with placebo, and found no
34            difference between the groups at 24 weeks for radiological damage.
35            Level 1+
36        •   One RCT257 looked at varied doses of oral prednisolone given in a
37            tapered regime, starting at 30mg, and tailing this down to a dose
38            between 2.5mg and 15mg (determined by the patient) to control
39            disease. The steroid arm showed less joint damage (delta Larsen
40            score) at 1 year, but a similar number of patients with joint damage
41            and progression of erosions. Level 1+
42
43   Function
44        • One RCT269 looked at different doses of intra-articular rimexolone
45           into RA knees and for higher doses found sustained functional
46           benefits (walking ability, range of movement) compared to placebo for
47           up to 84 days. Level 1+

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             131 of 307
     DRAFT FOR CONSULTATION


1         •   One RCT267 compared monthly depomedrone to placebo in addition
2             to the usual DMARD and found improved HAQs at 6 months, but not
3             sustained at 1 or 2 years. Level 1++
4         •   One RCT258 looked at IM methylprednisolone at 0, 4 and 8 weeks of
5             initiation of IM gold therapy compared with placebo. They found
6             improvements in HAQ in the steroid arm at 4 and 12 weeks, but not
7             at 8 and 24 weeks. There was no improvement in grip strength at any
8             stage for the steroid arm. Level 1+
 9        •    One RCT257 looked at varied doses of oral prednisolone given in a
10            tapered regime, starting at 30mg, and tailing this down to a dose
11            between 2.5mg and 15mg (determined by the patient) to control
12            disease. There was a significant improvement in HAQ and grip
13            strength in the steroid group at 1 year. Level 1+

14   7.2.5 Health economic evidence statements
15   Bae et al270 concluded that corticosteroids are less costly and more effective
16   than NSAIDs in the long-term treatment of RA patients, but differences
17   between the two treatments in both cost and health outcome are relatively
18   small. The study is US based hence resources used and their associated
19   costs could be very different in the UK. In the Verhoeven et al study271, the
20   analysis revealed that combined treatment with step-down prednisolone,
21   methotrexate and sulphasalazine to be more effective than sulphasalazine
22   alone at equal or lower cost. The combined treatment group had lower
23   expenses for non-protocol medication and in-patient care and lower costs
24   outside the health care system that offset the higher costs for protocol
25   medication and monitoring.

26   7.2.6 Summary of evidence statements
27    Recent onset of RA
28   In recent onset RA, low dose oral steroid regimes give symptomatic and
29   quality of life benefit for up to three months268, but this is usually not sustained
30   at one year or beyond 259,262-266. There is no good evidence for functional
31   improvements with steroids259,262,264-266. Low dose steroids are generally well
32   tolerated259,263,266, and small decreases in concomitant medications may be
33   possible}263 264,265. The majority of the trials suggest that steroids are disease
34   modifying in slowing radiological damage over two years262-266.
35
36   Established RA
37        • Knee joint injections of steroid give sustained benefit from a
38            symptomatic and functional viewpoint (although the steroid
39            concerned is not used in the UK)269.
40        •   IM, IV, or oral steroid routes are largely of value in the short term, do
41            not lead to sustained symptomatic benefits257,258,260,267.
42        •   Using oral steroids in variable doses to control symptoms may
43            improve function over 1 year257.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                132 of 307
     DRAFT FOR CONSULTATION


1         •   The evidence that the use of steroids in established RA may be
2             disease modifying is conflicting, some trials probably underpowered
3             showed no significant effect. This evidence base is not as strong as
4             the evidence base for early RA.258,267 257
5
6    Themes from both early RA and established disease
7        • Steroids usually only give short term symptomatic, functional and
8           quality of life benefit.
9    Most trials show evidence that steroids are disease modifying

10   7.2.7 From evidence to recommendations
11   Recent onset of RA
12   The GDG noted that there was a considerable mismatch between the
13   available data and what actually happens in clinical practice. For example,
14   both in the initial presentation of disease and during flare-ups, steroids (oral,
15   intra-muscular and intra-articular) are often used to obtain symptomatic
16   benefit and to achieve disease control whilst waiting for the more slowly-acting
17   DMARDS to take effect, despite the lack of evidence to support this. The
18   clinical efficacy of this approach is so well established that it is doubtful that
19   any future randomised controlled clinical trials would ever be conducted, and
20   the GDG felt that there should therefore be a recommendation endorsing this
21   use of steroids, both for those patients with newly diagnosed rheumatoid
22   arthritis who are not already receiving steroids as part of DMARD combination
23   therapy (see recommendation R22), and for the management of flare-ups in
24   those with recent onset or established disease.
25
26   Established RA
27   The GDG noted that the evidence for the use of steroids in established
28   disease was sparse, of limited quality, and that in two trials the much older
29   drugs penicillamine and gold had been used as comparators. There was a
30   need to establish the merits of combining steroids with drugs such as
31   methotrexate in established disease, where there was much less evidence for
32   disease modification by steroids than in early disease.
33
34   Although a consistent theme for both early and established disease is that the
35   symptomatic benefit produced by steroids is usually only short lasting, the
36   GDG noted that, in routine clinical practice, there are nevertheless some
37   patients who appear to be reliant on long-term low dose steroids since
38   withdrawing them results in flare-up of disease activity. This use of steroids in
39   this particular group of patients may have to be accepted, even though the
40   situation is not ideal, although attempts should always be made to replace the
41   steroids with other disease modifying drugs and to keep the steroids to the
42   lowest dose that controls symptoms. It was also felt important to emphasise
43   the specific potential serious complications associated with long-term steroid
44   therapy.

45   7.2.8 Recommendations
46

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              133 of 307
     DRAFT FOR CONSULTATION


 1   R 22 In people with newly diagnosed RA who are not already receiving
 2   glucocorticoids as part of DMARD combination therapy, short-term treatment
 3   with glucocorticoids (oral, intramuscular, intra-articular) should be considered
 4   for achieving rapid improvements in symptoms.
 5
 6   R 23 In people with recent onset or established rheumatoid arthritis, short-
 7   term treatment with glucocorticoids should be used for the management of
 8   flare-ups in order to rapidly decrease the symptoms of inflammation and re-
 9   establish disease control.
10
11   R 24 In people with established RA, long-term treatment with glucocorticoids
12   should only be continued when:
13        • The specific long-term complications of glucocorticoid therapy have
14           been fully discussed.
15         •    All other treatment options (including biologics) have been explored
16              and offered.
17         •    There is no other way of keeping their disease under control.

18   7.3       Disease modifying and biological drugs: when to
19   introduce, and optimal sequencing
20   Introducing disease-modifying drugs (DMARDS)

21   7.3.1 Clinical introduction
22   In previous decades, DMARDs were introduced when patients fulfilled the
23   ARA classification criteria for RA2, or when there were radiological erosions. It
24   is now recognised that the ARA classification criteria are not designed to
25   identify early RA (see section 4), and that persistent synovitis needs an
26   appropriate disease modifying drug intervention irrespective of the distribution
27   of joints affected, and the results of investigations. Consequently, the early
28   introduction of DMARDs has been advocated for any persistent idiopathic
29   synovitis. This decision is not always straightforward, because of diagnostic
30   and prognostic difficulties, and also concerns over toxicity associated with
31   DMARDs. There are also practical difficulties, with delays from symptom
32   onset, to presentation to General Practitioner, to referral to secondary care,
33   averaging 9 months in total in the UK. Irrespective of these diagnostic,
34   prognostic and practical difficulties, a concept of a “window of opportunity” has
35   emerged in the RA literature, expressing the belief that the earlier the
36   introduction of DMARDs, the greater the impact on long-term outcomes. Is
37   there any evidence to support this?
38
39   Having decided to start a disease modifying drug (DMARD), the
40   rheumatologist is then faced with the problem of determining which DMARD
41   should be used, and whether a monotherapy or combination of DMARDs
42   should be employed. Given appropriate information, people with RA may
43   prefer one drug over another, because of mode or frequency of
44   administration, lifestyle and perceived side effect profiles, and other
45   considerations (e.g. comorbidities, pregnancy). The clinician will also bring

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              134 of 307
     DRAFT FOR CONSULTATION


 1   their own clinical experience to these decisions, based on their interpretation
 2   of the evidence, and personal experience of efficacy and toxicity with different
 3   drug regimens. On occasions, the choice will be determined by DMARD
 4   toxicity profiles and patient co-morbidities. A large number of trials of a variety
 5   of monotherapies and combination therapies have now been published. It
 6   should be possible to determine whether there are drug approaches that have
 7   greater efficacy and tolerability than others, and all other considerations being
 8   equal, should be used in preference to other regimens for patients with a
 9   recent onset of RA. It should also be possible to address the cost-
10   effectiveness of different treatment strategies to determine whether some
11   should take priority over others.
12

13   7.3.2 Clinical methodological introduction (introducing DMARDS)
14   We looked for studies that investigated the efficacy and safety of early
15   introduction of DMARDs with respect to symptoms, joint damage, function and
16   quality of life in patients with a recent onset of RA. Due to the large volume of
17   evidence, only Level 1 and 2 studies (MA, RCT, Cohort and Case-control
18   studies) were selected which were of a UK-relevant population.
19
20   Ten studies were found that fulfilled the criteria. This consisted of 1 MA,272 6
21   RCTs,273-281 2 cohort studies (1 cohort50and 49; 1 cohort282) and 1 case-control
22   study283. Three of the RCTs (1 RCT275,276; 1 RCT273,274; 1 RCT278,279) and one
23   of the cohort studies49,50 were each published as 2 separate papers reporting
24   different outcomes or follow-up times, so these trials have only been counted
25   once, however results from both papers are reported and referenced here.
26   Two of the RCTs280,281 were excluded as evidence due to methodological
27   limitations. All other trials were methodologically sounds.
28
29   The SR/MA272 focused on 12 studies (6 follow-up studies of RCTs and 6
30   cohort studies) which compared early vs delayed treatment with DMARDs in
31   patients with a recent onset of RA (<2 years). The MA itself was well
32   conducted however, the 12 studies it included were of varying quality. Studies
33   included in the analysis differed with respect to:
34         • Study size (range N=23 to N=189)
35        •   Study quality (max score of 6) – [N=6 studies poor or reasonable
36            quality - score of 2 or 3 (1- or 1+); N=6 good quality - score of 4 or 5
37            (1++)]
38        •   Delay in DMARD initiation (difference in months in mean disease
39            duration at DMARD initiation between the two treatment arms) – 6 to
40            14 months
41        •   Study duration – length of follow-up (1 to 5.6 yrs)
42
43   All 4 included RCTs [1 RCT278,279; 1 RCT275,276; 1 RCT273,274; RCT277) had
44   variable inclusion criteria and differed with respect to the following:
45          • Sample size (range: N=120 to N=238)


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                135 of 307
     DRAFT FOR CONSULTATION


1              •    Blinding (2 RCTs double blind, 2 RCTs single blind/unblinded for
2                   some outcomes)
3              •    Trial length and follow-up (range: 36 weeks to 5 years)
4              •    Treatment – type of DMARDs used (1 RCT HCQ, 1 RCT auranofin, 1
5                   RCT SAARD vs delayed pyramid, 1 RCT 3 DMARDs –
6                   SSZ+MTZ+HCQ + CS vs 1 DMARD SSZ +/- CS)
7              •    Treatment regimen – single drugs and combinations compared
8              •    Treatment regimen – dose
 9   Cohort studies: The first cohort study49,50 included N=206 patients with a
10   recent onset of RA who were either a) promptly treated with DMARDs and
11   NSAIDs or b) had delayed treatment with DMARDs + NSAIDs. The mean
12   delay to treatment was approximately 4 months in the delayed treatment
13   group and patients were followed prospectively for 4 years. The second cohort
14   study282 included N=149 patients with a recent onset of RA who were treated
15   with DMARDs when they had either a) very early RA (mean duration 3.1
16   months) or b) early RA (mean duration 9.2 months). The patients were
17   followed prospectively for 3 years.
18
19   The case-control study283 included N=40 patients with a recent onset of RA
20   who were either a) promptly treated with DMARDs or b) had delayed
21   treatment with DMARDs. The mean delay to treatment in the delayed group
22   was approximately 9 months after the early treatment group and patients were
23   followed prospectively for 3 years.

24   7.3.3 Health economic methodological introduction (introducing
25   DMARDS)
26             •    Seven papers were identified and four were excluded for not being a
27                  cost-effectiveness analysis284-286 or not being an early RA
28                  population.287 The remaining three studies271,288,289met the inclusion
29                  criteria and were appraised.

30   7.3.4 Clinical evidence statements (introducing DMARDS)

31   Recent onset of RA
32
33        •        Symptoms / QoL (Early vs Delayed treatment)
34
     Study                   Outcomes                            Follow-up         Result -
                                                                                   best
                                                                                   treatment
     1 RCT (Borg et          Number of swollen joints and        2 years and 5     Early
     al., and                Beck depression score; RAI          years             (p<0.05)
     Egsmose et al)          (AUC over 5 years)
     275,276
                             Pain (VAS); Morning stiffness       2 years and 5     NS
     Level 1+                and general health (5 years)        years


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                    136 of 307
    DRAFT FOR CONSULTATION


    1 RCT (Van der      Pain, joint score (tender and        6 months and      Early
    Hejde et al and     swollen joints), percentage of       1 year            (p<0.05)
    Verstappen et       patients showing clinical
    al)278,279          improvement (joint score), well-
    Level 1+            being and morning stiffness
                        Joint score, Pain (VAS), General     Over 5 years      NS
                        well-being (VAS) and Morning         (AUC)
                        stiffness (mins)
                        number of patients showing           3, 6, 9, 12 and   NS
                        clinically relevant improvement      21 months
                        (≥20%)
    1 RCT (HERA         Composite scores of both joint       36 weeks (and     Early (all
    study – 2           index and pain index and for         mean over all     p<0.05)
    papers) 273,274     Clinically significant improvement   assessment
    Level 1++           (≥20%)                               times)
                        Pain index clinically significant    3 years follow-   Early
                        improvement (≥20%)                   up                (p<0.05)
                        AIMS psychological scale             36 weeks (and     NS
                                                             mean over all
                                                             assessment
                                                             times)
    1 cohort study      DAS score                            1 year            Early
    (Lard et al, Van                                                           (p<0.05)
    Aken et al)49,50
    Level 2+
                        DAS score and DAS (AUC)              2 years           NS
    1 cohort study      RAI and number of swollen joints     Over the 3        Early
    (Peltomaa et                                             years             (p<0.05)
    al) 282
    Level 2+
    1 case-control      Pain (VAS); ACR20, ACR50 and         3 months and      Early
    study (Nell et      ACR70                                3 years           (p<0.05)
    al)283              Tender joints; EULAR good            3 years           Early
     Level 2+           responders; DAS28 ≤ 3.2                                (p<0.05)
                        Swollen joints; EULAR response       3 months and      NS
                        rates                                3 years

1        •    Function (Early vs Delayed treatment)
2
    Study               Outcomes                             Follow-up         Result -
                                                                               best
                                                                               treatment
    1 RCT (Borg et      Kietel Functional score              2 years and 5     Early
    al., and                                                 years             (p<0.05)
    Egsmose et al)      HAQ                                  2 years           Early
    275,276
                                                                               (p<0.05)
    Level 1+        HAQ and grip strength                    5 years           NS
    1 RCT (Van der HAQ; HAQ clinical improvement;            6 months and      Early
    Hejde et al and grip strength                            1 year            (p<0.05)


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 137 of 307
    DRAFT FOR CONSULTATION


    Verstappen et       HAQ and grip strength                 Over 5 years      NS
    al)278,279                                                (AUC)
    Level 1+
    1 RCT (HERA         Physical function index               36 weeks (and     Early
    study – 2                                                 mean over all     (p=0.004)
    papers) 273,274                                           assessment
    Level 1++                                                 times)
                        Physical function index (clinically   3 years follow-   Early
                        significant improvement)              up
    1 cohort study      HAQ score (AUC)                       Over 2 years      Early
    (Lard et al, Van                                                            (p<0.05)
    Aken et al)49,50
    Level 2+
                        HAQ score                             2 years           NS
    1 case-control      HAQ                                   3 months and      Early
    study (Nell et                                            3 years           (p<0.05)
    al)283
    Level 2+
1
2         •   Joint damage (Early vs Delayed treatment)
3
    Study               Outcomes                              Follow-up         Result -
                                                                                best
                                                                                treatment
    1 MA (Finckh et Radiographic progression rate             Median follow-    Early
    al) 272                                                   up 3 years        (SMD -
    Level 1++                                                                   0.19, 95%
                                                                                CI -0.34 to
                                                                                -0.04)
    1 RCT (Borg et      Larsen Score; erosion score,          5 years           Early
    al., and            number of eroded joints and                             (p<0.05)
    Egsmose et al)      number of engaged joints
    275,276

    Level 1+
    1 RCT (Van der      Increase in radiographic damage       1 year and 5      NS
    Hejde et al and     (JSN, erosion and total               years
    Verstappen et       radiographic damage score)
    al)278,279
    Level 1+
    1 RCT               Increase in Larsen score              2 years           Early
    (Mottonen et al)                                                            (p<0.001)
    277

    Level 1+
    1 cohort study      Progressive joint destruction         1, 2 and 5        Early
    (Lard et al, Van    (Sharp score >5); radiographic        years             (p=0.005,
    Aken et al)49,50    joint damage (modified Sharp                            p=0.001
    Level 2+            score) over 2 years                                     and
                                                                                p=0.032)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 138 of 307
    DRAFT FOR CONSULTATION


                      Rate of progression                1, 2 and 3        Early
                                                         years             (better)
                                                         4 years           NS
                                                         Years 1-4 and     NS
                                                         2-4
                      Radiographic joint damage          6 months          Similar
    1 cohort study    Erosive disease and Larsen         over the 3        NS
    (Peltomaa et      scores                             years
    al) 282
    Level 2+
    1 case-control    Larsen score                       1, 2 and 3        Early
    study (Nell et                                       years             (p<0.05)
    al)283            Erosions (Larsen score ≥ 2)        3 years           Early
    Level 2+                                                               (p<0.05)
1
2      •    Global assessment (Early vs Delayed treatment)
3
    Study             Outcomes                           Follow-up         Result -
                                                                           best
                                                                           treatment
    1 RCT (HERA       Patient’s and physician’s global   36 weeks          Early
    study – 2         assessment of therapeutic                            (p=0.01
    papers) 273,274   benefit                            3 years follow-   and 0.032)
    Level 1++                                            up                NS
    1 case-control    Patient’s and physician’s global   3 months and      Early
    study (Nell et    assessment                         3 years           (p<0.05)
    al)283
    Level 2+
4
5      •    Biochemical markers (Early vs Delayed treatment)
6
    Study                          Outcomes              Follow-up         Result -
                                                                           best
                                                                           treatment
    1 RCT (Van der Heide et al     ESR                   1 year            Early
    and Verstappen et al)278,279                                           (p<0.05)
    Level 1+                       ESR (AUC)             Over 5 years      NS
                                                         (AUC)

                                   CRP                   6 months and      Early
                                                         1 year            (p<0.05)
    1 RCT (HERA study – 2          ESR                   36 weeks          NS
    papers) 273,274
    Level 1++
    1 cohort study (Lard et al,    CRP                   3 months          Early
    Van Aken et al)49,50                                                   (p<0.05)
    Level 2+

                                   CRP                   1 year and 2      NS
                                                         years

    Rheumatoid arthritis: full guideline DRAFT (August 2008)            139 of 307
    DRAFT FOR CONSULTATION


                                    CRP (AUC)                 Over the 2       Early
                                                              years
    1 cohort study (Peltomaa et CRP                           Over the 3       Early
    al) 282                                                   years            (p<0.05)
    Level 2+                    ESR                           Over the 3       NS
                                                              years
    1 case-control study (Nell      CRP levels and ESR        3 months and     Early
    et al)283                                                 3 years          (p<0.05)
    Level 2+
1
2        •    AEs (Early vs Delayed treatment)
3
    Study                           Outcomes                  Follow-up        Result -
                                                                               best
                                                                               treatment
    1 RCT (Van der Hejde et al      GI AEs                    1 year           NS
    and Verstappen et al)278,279
    Level 1+
    1 RCT (HERA study – 2           Clinically significant    36 weeks         NS
    papers) 273,274                 AEs
    Level 1++
                                    Total number of AEs       36 weeks         Similar
4
5        •    Withdrawals (Early vs Delayed treatment)
6
    Study               Outcomes                              Follow-up        Result -
                                                                               best
                                                                               treatment
    1 RCT (Borg et      Withdrawals due to lack of            2 years          Early
    al., and            response, patients still continuing                    (p<0.001)
    Egsmose et al)      on the original treatment
    275,276

    Level 1+            Withdrawals due to AEs                2 years          Delayed
                                                                               (p<0.01)
    1 RCT (Van der      Total number of withdrawals           2 years          Early
    Hejde et al and     (mainly due to lack of efficacy)                       (p<0.05)
    Verstappen et       and withdrawals due to AEs
    al)278,279
    Level 1+
    1 RCT (HERA         withdrawals due to lack of            36 weeks         Early
    study – 2           efficacy                                               (better)
    papers) 273,274     withdrawals due to AEs                36 weeks         NS
    Level 1++
    1 cohort study      Number of withdrawals                 2 years          Early
    (Lard et al, Van                                                           (better)
    Aken et al)49,50
    Level 2+
                        change in initial DMARD therapy       Over 2 years     Delayed
                        due to: AEs, lack of efficacy                          (better)


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 140 of 307
     DRAFT FOR CONSULTATION


     1 case-control      DMARD switching due to AEs or        Over 3 years      Delayed
     study (Nell et      lack of efficacy                                       (p<0.05)
     al)283
     Level 2+

1          •   Remission (Early vs Delayed treatment)
2
     Study               Outcomes                             Follow-up         Result -
                                                                                best
                                                                                treatment
     1 RCT (Van der      median lag time until first          1 year and 5      Early
     Hejde et al and     complete response (remission)        years             (p<0.05)
     Verstappen et       Number of patients achieving         Over 5 years      NS
     al)278,279          complete response (remission)
     Level 1+            over 5 years
     1 RCT               frequency of patients with           2 years           Early
     (Mottonen et al)    remission (in single treatment                         (p=0.021).
     277
                         group)
     Level 1+            frequency of patients with           2 years           NS
                         remission (in combination
                         treatment group)
     1 cohort study      number of patients in remission      over the 3        NS
     (Peltomaa et                                             years
     al) 282
     Level 2+
3

4    7.3.5 Health economic evidence statements (introducing
5    DMARDS)
 6   Grigor et al288 was the only study of a UK population, Verhoeven et al271 and
 7   Korthals et al289 were Dutch and Belgian respectively, and so their relevance
 8   to a UK is limited. Grigor et al has a high inclusion criteria for early RA (<5
 9   years) although the mean disease duration of the trial was 19 months. Grigor
10   et al estimates that the intensive step-up TICORA strategy dominates the
11   routine step-up strategy (sulfasalazine monotherapy with a DMARD added if
12   initially failed). Verhoeven et al and Korthals et al are both based on the
13   COBRA trial, and they estimate that combined step-down therapy or
14   prednisolone, methotrexate and sulphasalazine dominate sulphasalazine
15   monotherapy.

16   7.3.6 Summary of evidence statements (introducing DMARDS)
17   For symptoms, joint damage, function and quality of life, delay in introducing
18   DMARDs is inferior to early commencement275,276,278,279. 273,274 49,50,272,282,283
19   Prompt introduction of DMARDs can lead to benefits up to 5 years after the
20   drugs are introduced when compared with a delayed start275,276. Early
21   introduction of drugs also results in fewer adverse reactions and
22   withdrawals49,50,273-276,278,279,283. There was some evidence that combination
23   therapies could extend the window of opportunity for DMARDs to be effective

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                141 of 307
     DRAFT FOR CONSULTATION


1    when compared with monotherapies277. But the key message to emerge was
2    to start effective DMARD therapy as soon as possible.
3
4    Optimal sequencing of disease-modifying drugs (DMARDS)

5    7.3.7 Clinical methodological introduction (optimal sequencing of
6    DMARDs)
 7   We looked for studies that investigated the efficacy and safety of different
 8   sequences of DMARDs with respect to symptoms, joint damage, function and
 9   quality of life in patients with a recent onset of RA. Due to the large volume of
10   evidence, only Level 1 and 2 studies (MA, RCT, Cohort and Case-control
11   studies) were selected which were of a UK-relevant population and had a
12   sample size of N≥100.
13
14   Twelve studies were found that fulfilled the criteria. This consisted of 2
15   MA,290,291 9 RCTs,288,292-308 and 1 cohort study (prospective)309. Two of the
16   RCTs (1 RCT – The CIMESTRA study292,293; 1 RCT – The COBRA trial299,300
17   were each published as 2 separate papers. 2 RCTs (1 RCT – The FINRACO
18   study294-297; 1 RCT – The BeSt study302-305 were each published as 4 separate
19   papers. These all reported different outcomes or follow-up times, so these
20   trials have only been counted once, however results from all papers are
21   reported and referenced here. The MA291 and one of the RCTs298 were
22   excluded as evidence due to methodological limitations. All other trials were
23   methodologically sound.
24
25   The included studies were separated into two different groups in order to
26   analyse and compare the data. 1) Aggressive vs non-aggressive sequences
27   of DMARD treatment and 2) general sequences of DMARDs.
28
29   NOTE: For many of the trials, the sequences of the DMARDs may have
30   differed for patients even within the same arm of a trial, because often doses
31   and DMARDs used were adjusted or changed for individual patients during
32   the study depending upon whether they developed AEs or showed lack of
33   clinical efficacy. This was done to reflect clinical practice.
34
35
36   1) Aggressive vs non-aggressive DMARD treatment
37
38   Three RCTs (1 RCT – COBRA trial 299,300;1 RCT - The CIMESTRA
39   study292,293; 1 RCT – The FINRACO study294-297) all addressed aggressive vs
40   non-aggressive DMARD therapy.
41
42   The first RCT - the COBRA trial299,300 compared 2 different treatment arms.
43   Group 1 (aggressive): Patients given SSZ then CS and MTX were added then
44   tapered and withdrawn; Group 2 (non-aggressive): patients given SSZ and
45   continued on SSZ. The second RCT – CIMESTRA trial292,293 included N=163
46   patients with a recent onset of RA. The trial had 2 arms in which patients were
47   given either 1) aggressive DMARD treatment (SSZ given, then MTX +
48   prednisolone added) or 2) Non-aggressive DMARD treatment (SSZ given and

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              142 of 307
     DRAFT FOR CONSULTATION


 1   continued on this drug) in a 2 year treatment phase. The third RCT –
 2   FINRACO trial294-297 included N=199 patients with a recent onset of RA. The
 3   trial had 2 arms in which patients were given either 1) aggressive DMARD
 4   treatment (3 DMARDs + prednisolone) or 2) Non-aggressive DMARD
 5   treatment (Single DMARD +/- prednisolone) in a 5 year treatment phase.
 6
 7   2) General sequences of DMARDs
 8   One MA290, 6 RCTs (1 RCT – The BeSt study302-305; 1 RCT – The MASCOT
 9   study301; 1 RCT306; 1 RCT – the TICORA study288; 2 other RCTs307,308) and
10   one cohort study (prospective)309 all addressed sequences of DMARDs.
11
12   The MA290 focused on RCTs and quasi-randomised RCTs which compared
13   DMARD monotherapy vs DMARD combination therapy, in patients with both a
14   recent onset (<3 years) and established RA (>3 years). The MA itself was well
15   conducted and was methodologically sound. However, the 36 RCTs it
16   included were of varying quality. Studies included in the analysis were similar
17   with respect to:
18         • Study design (All RCTs/quasi-randomised CTs)
19        •   Intervention (DMARD)
20        •   Comparison group (combination therapy with 2 or more DMARDs or
21            1 DMARD + 1 biological agent)
22        •   Blinding (double blind assessment was performed)
23        •   Allocation concealment
24        •   ITT analysis was performed
25        •   Study size (all fairly small, N <100)
26   However, they differed with respect to:
27       • Study size (range N=11 to N=89)
28        •   Study quality – max Jadad score of 5 (N=30 studies good quality;
29            N=6 poor quality)
30        •   Study duration – variable, (exact lengths not mentioned)
31
32   NOTE: Because the MA pooled together papers which had patients of both
33   recent onset and established RA, it was decided that duplicate papers (ie.
34   papers picked up in our search which were already included in the MA) would
35   still be included as evidence in this section for ‘recent onset’ RA in order to
36   explore any effects within these sub-populations.
37
38   The 6 included RCTs all assessed sequences of DMARDs but had variable
39   inclusion criteria. The trials differed with respect to the following:
40         • Sample size (range: N=111 to N=508)
41        •   Blinding (1 RCT triple blind, 2 RCTs double blind, 2 RCTs single
42            blind, 1 RCT unblinded)
43        •   Trial length (1 RCT 6 months, 3 RCTs 18 months, 2 RCTs 2 years)
44        •   Treatment – type and sequences of DMARDs used and compared

     Rheumatoid arthritis: full guideline DRAFT (August 2008)            143 of 307
     DRAFT FOR CONSULTATION


1         •   Treatment regimen – doses of DMARDs
 2
 3   Treatments used
 4   1. The first RCT – the BeSt study302-305, compared 4 different treatment arms.
 5   Group 1: sequential monotherapy; Group 2: step-up combination therapy;
 6   Group 3: initial combination therapy with CS; and Group 4: initial combination
 7   therapy with infliximab. Therapies were changed depending on DAS score.
 8
 9   2. The second RCT, the MASCOT study301 compared 3 different treatment
10   arms. Group 1: patients were given SSZ then SSZ + MTX; Group 2 patients
11   were given SSZ and continued on SSZ; Group 3: patients were given SSZ
12   then MTX.
13
14   3. The third RCT306 compared 3 different treatment arms. Group 1: patients
15   were on DMARD then MTX was added; Group 2: patients were on DMARD
16   then CsA was added; Group 3: patients were on DMARD then SSZ was
17   added.
18
19   4. The fourth RCT – the TICORA study288 compared 2 different treatment
20   arms. Group 1 (Intensive): initial monotherapy with SSZ (escalating dose)
21   then triple therapy; Group 2 (Routine): initial monotherapy with SSZ then
22   alternative monotherapy or step-up combination therapy. Therapies were
23   changed depending on DAS score.
24
25   5. The fifth RCT308 compared 2 different treatment arms in patients who had
26   not previously been treated with DMARDs. Group 1: SC (subcutaneous) MTX;
27   Group 2: Oral MTX.
28
29   6. The sixth RCT307 compared 4 different treatment arms. Group 1: MTX
30   (starting 7.5 mg/week increasing incrementally to 15 mg/week). Group 2:
31   Step-down prednisolone started with MTX (60 mg/day initially, reduced to 7.5
32   mg at 6 weeks, 7.5 mg/day from 6-8 weeks, stopped by 34 weeks). Group 3:
33   ciclosporin started 3 months after MTX (initial dose 100 mg/day, increased
34   gradually to target dose of 3 mg/kg daily). Group 4: all treatments
35
36   The prospective cohort study309 included N=439 patients who were taken from
37   NOAR (The Norfolk Arthritis Register) and had been treated with either MTX
38   or SSZ as first-line therapy and were followed prospectively for 2 years.

39   7.3.8 Health economic methodological introduction (optimal
40   sequencing of DMARDs)
41        •   The search terms for these this question and the DMARD question
42            were combined in order to look at the cost-effectiveness evidence of
43            biologics and traditional DMARDs in early RA. To identify evidence
44            related to biologics, the HTA review by Chen et al284 was updated. No
45            studies addressed early RA, although this was addressed by the
46            BRAM model in the HTA report itself. One other study286 was
47            identified by updating the review.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)            144 of 307
     DRAFT FOR CONSULTATION


1    7.3.9 Clinical evidence statements (optimal sequencing of
2    DMARDs)

3    Recent onset of RA
4
5    1) AGGRESSIVE VS NON-AGGRESSIVE TREATMENT
6
7    Symptoms / QoL
8
     Study              Outcomes                      Follow-    Result - best
                                                      up         treatment
     COBRA              Pooled index, Tender and      28 weeks   Aggressive (all
     study299,300       swollen joint counts, pain               p≤0.001 except
     Level 1++          (VAS), DAS28 and MACTAR                  pain p=0.002)
                        score
                                                                 NS
                                                      56 weeks

                        DAS28 score reduction over    change     Aggressive (better)
                        time                          from 1-5
                                                      years

     CIMESTRA           ACR-N                         1 year     Aggressive
     study292,293                                                (p<0.05)
     Level 1++          ACR20                         1 year     Aggressive
                                                      2 years    (p<0.05)
                                                                 NS
                        ACR50                         1 year     NS

                                                      2 years    Aggressive
                                                                 (p<0.05)

                        DAS28                         2 years    NS

                        tender and swollen joints,    1 and 2    NS
                        ACR70 and pain (VAS)          years

     FINRACO            ACR50 and swollen joint       2 years    Aggressive
     study294-297       count                                    (p<0.05)
     Level 1+           DAS score; DAS28 AUC;         5 years    Aggressive
                        work disability                          (p<0.05, p<0.001,
                                                                 p<0.01)

 9       •   Function
10
     Study              Outcomes                      Follow-    Result - best
                                                      up         treatment



     Rheumatoid arthritis: full guideline DRAFT (August 2008)         145 of 307
    DRAFT FOR CONSULTATION


    COBRA            HAQ score and grip strength     28 weeks     Aggressive (both
    study299,300                                                  p<0.0001)
    Level 1++
                                                                  NS


                                                     56 weeks

    CIMESTRA         HAQ score; patients with        1 year       NS
    study292,293     HAQ score ≤0.25                 (and 2
    Level 1++                                        years
                                                     HAQ
                                                     score)
    FINRACO          physical function               2 years      NS
    study294-297
    Level 1+

1       •   Joint damage
2
    Study            Outcomes                   Follow-up              Result - best
                                                                       treatment
    COBRA            Erosion score; Total       28, 56 and 80          Aggressive (all
    study299,300     radiographic damage        weeks                  p≤0.01)
    Level 1++        (SHS) score

                     JSN                        28 weeks and           Aggressive
                                                over time (mean        (p<0.05)
                                                change from 1-5
                                                years)

                                                                       NS

                                                56 and 80 weeks

                     Sharp damage score,        28 weeks and           Aggressive
                     erosion score              over time (mean        (p<0.05)
                                                change from 1-5
                                                years)

    CIMESTRA         Radiographic               1 year                 NS
    study292,293     progression                2 years                Aggressive
    Level 1++                                                          (p=0.03)
                     Larsen score, Rate of    1 year                   NS
                     radiographic progression
                     and development of
                     bone erosions
                     Total Sharp Score,       2 years                  NS
                     erosion score, JSN and
                     progression since
                     baseline


    Rheumatoid arthritis: full guideline DRAFT (August 2008)             146 of 307
    DRAFT FOR CONSULTATION


    FINRACO          eroded joints, joint       2 years and 5     Aggressive (all
    study294-297     damage - Larsen score      years             p<0.01)
    Level 1+         and progression
                     (increase in Larsen
                     score)

1       •   Global assessment
2
    Study            Outcomes                   Follow-up         Result - best
                                                                  treatment
    COBRA            Assessor’s global          28 weeks          Aggressive
    study299,300     assessment                                   (p=0.0001)
    Level 1++                                   56 weeks          NS

                     Patient’s global           28 weeks and 56   NS
                     assessment                 weeks

    CIMESTRA         Patient’s and Physician’s 2 years            NS
    study292,293     global assessment
    Level 1++
    FINRACO          Patient’s and Physician’s 2 years            NS
    study294-297     overall assessments
    Level 1+
3
4       •   Biochemical markers
5
    Study            Outcomes                   Follow-up         Result - best
                                                                  treatment
    COBRA            ESR                        28 weeks          Aggressive
    study299,300                                                  (p=0.002)
    Level 1++                                   56 weeks          NS


    CIMESTRA         ESR and CRP                2 years           NS
    study292,293
    Level 1++
    FINRACO          ESR                        2 years           Aggressive
    study294-297                                                  (p<0.05)
    Level 1+

6       •   AEs
7
    Study            Outcomes                   Follow-up         Result - best
                                                                  treatment
    CIMESTRA         AEs                        1 year            Non-aggressive
    study292,293                                2 years           (better)
    Level 1++                                                     NS




    Rheumatoid arthritis: full guideline DRAFT (August 2008)        147 of 307
     DRAFT FOR CONSULTATION


     FINRACO             Number of patients with    2 years             NS
     study294-297        AEs; SAEs or GI AEs
     Level 1+
1
2        •    Withdrawals
3
     Study               Outcomes                   Follow-up           Result - best
                                                                        treatment
     COBRA               Total number of            56 weeks            Aggressive
     study299,300        withdrawals, withdrawals                       (better)
     Level 1++           due to AEs and due to
                         lack of efficacy

     CIMESTRA            withdrawals due to SAEs 2 years                Similar
     study292,293
     Level 1++

4        •    Remission
5            • 1 RCT, the FINRACO study294-297 found that sustained remission
6              protects against radiographic joint damage. Patients in sustained
7              remission had less radiographic progression over 2 years compared
8              with patients who were in remission at 6 months and lost it later. 1+
9
     Study               Outcomes                   Follow-up           Result - best
                                                                        treatment
     COBRA               Remission (ACR and         48 weeks, 1 year    NS
     study299,300        DAS)                       and 2 years
     Level 1++
                         Remission (EULAR)          2 years             NS

     FINRACO             patients with sustained    Over the 2 years    Aggressive
     study294-297        remission (ACR)                                (p=0.013 - OR
     Level 1+                                                           4.6, 95% CI 1.2 to
                                                                        17.0)
                         patients with sustained    Over the 2 years    Aggressive
                         remission (DAS28)                              (p<0.001- OR 5.6,
                                                                        95% CI 2.6 to
                                                                        11.6)
                         patients with sustained    Over the 2 years    Aggressive
                         remission (EULAR)                              (p<0.001 - OR
                                                                        5.4, 95% CI 2.7 to
                                                                        10.6).
10
11   2) GENERAL SEQUENCES

12   Symptoms / QoL
13




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              148 of 307
    DRAFT FOR CONSULTATION


1   •   1 RCT – the BeSt study302-305 found that after failure on initial MTX,
2       treatment with subsequent conventional DMARDs is unlikely to result in a
3       DAS ≤2.4 and allows progression of joint damage. Level 1+
4
    Study          Sequence                      Outcomes             Follow-up       Result - best
                                                                                      treatment
    MASCOT         SSZ then SSZ + MTX vs         DAS score            Change          Arm 1
    study          SSZ continuous                                     from 6 - 18     (p=0.039)
    (Capell et                                                        months
    al)301
                                                 Pain (VAS),          Change          NS
    Level 1++                                    Swollen joint        from 6 - 18
                                                 count, RAI,          months
                                                 ACR20, ACR50
                                                 and ACR70

                   SSZ then SSZ + MTX vs         DAS score and        Change          Arm 1 (all
                   SSZ then MTX                  Ritchie Articular    from 6 - 18     p<0.05)
                                                 Index                months

                                                 HAQ, Pain (VAS),     Change          NS
                                                 Swollen joint        from 6 - 18
                                                 count, ACR20,        months
                                                 ACR50 and
                                                 ACR70
                   SSZ then MTX vs SSZ           DAS score, RAI,      Change          NS
                   continuous                    Swollen joint        from 6 - 18
                                                 count, Pain          months
                                                 (VAS), ACR20,
                                                 ACR50 and
                                                 ACR70
    1 RCT          DMARD then DMARD +            Swollen joints;      18 months       Arm 1 (all
    (Ferraccioli   MTX vs DMARD then             Pain (VAS);                          p<0.05).
    et al)306      DMARD + SSZ                                        Between
                                                 Swollen joints       18-36           NS
    Level 1+                                                          months
                                                 Tender joints        18 months       Arm 1
                                                                      and             (p<0.05)
                                                                      between 18-
                                                                      36 months
                   DMARD then DMARD +            Pain (VAS);          18 months       Arm 1 (both
                   CsA vs DMARD then             Tender joint count                   p=0.001)
                   DMARD + SSZ                                        Between
                                                                      18-36           NS
                                                                      months
                                                 Swollen joints       18 months       NS
                                                                      and
                                                                      Between
                                                                      18-36
                                                                      months
    BeSt           Step-up combination vs        DAS44 of ≤2.4        1 year          NS or Arm 1
    study302-305   Sequential monotherapy                                             vs 2
                   or Initial combination with                        2 years         (p=0.004)
    Level 1+       IFX                                                                NS


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                        149 of 307
    DRAFT FOR CONSULTATION


                 Sequential monotherapy     DAS44 of ≤2.4       1 year and 2 NS
                 vs Initial combination                         years
                 with CS
                 Initial combination with   DAS44 of ≤2.4       1 year and 2 NS
                 CS vs Step-up                                  years
                 combination or Initial
                 combination with IFX
                 Initial combination with   DAS44 of ≤2.4       1 year and 2 NS
                 IFX vs Step-up                                 years
                 combination
    1 RCT        SC MTX vs Oral MTX         ACR20, ACR70,       24 weeks      SC (all
    (Braun et                               swollen joints                    p<0.05)
    al)308                                  ACR50, tender       24 weeks      NS
    Level 1++                               joints, DAS28
    1 RCT307     MTX increasing dose vs     SF-36, DAS28        2 years       Arm 1 (better)
    Level 1++    MTX + prednisolone         score
                 (decrease dose)
                 MTX increasing dose vs     SF-36, DAS28        2 years       Arm 1 (better)
                 MTX then add ciclosporin   score
                 MTX + prednisolone         SF-36, DAS28        2 years       Arm 2 (better)
                 (decrease dose) vs MTX     score
                 then add ciclosporin
                 MTX + prednisolone +       SF-36, DAS28        2 years       Arm 1 (better)
                 ciclosporin vs all other   score
                 arms
    TICORA       intensive vs Routine       Pain (VAS),         18 months     Arm 1
    study                                   ACR20, ACR50,                     (p<0.0001;
    (Grigor et                              ACR70 Disease                     p=0.0028;
    al) 288                                 activity score;                   p=0.0003;
    Level 1++                               Joint swelling;                   p=0.021)
                                            Joint tenderness;
                                            SF-12 physical
                                            domain
                                            SF-12 mental        18 months     NS
                                            domain
    1 cohort     starting on SSZ vs         swollen and         2 years       NS
    study        starting on MTX            tender joints
    (Hider et                               swollen and         5 years       Arm 1
    al)309                                  tender joints                     (p=0.01,
     Level 2+                                                                 p=0.02)
                                            DAS28               5 years       NS

1      •    Function
2
    Study        Sequence                   Outcomes            Follow-up     Result -
                                                                              best
                                                                              treatment
    MASCOT       SSZ then SSZ + MTX         HAQ score           Change        NS
    study        vs SSZ continuous or                           from 6 - 18
    (Capell et   SSZ then MTX                                   months




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                150 of 307
DRAFT FOR CONSULTATION


al)301       SSZ then MTX vs SSZ        HAQ score      Change        NS
             continuous                                from 6 - 18
Level 1++                                              months

BeSt         Step-up combination        D-HAQ          1 and 2       NS
study302-    vs Sequential                             years and
305
             monotherapy                               over time
                                                       (2 years)
Level 1+     Sequential             D-HAQ              1 year and    Arm 2 (all
             monotherapy vs Initial                    over time     p<0.001)
             combination with CS or                    (2 years)
             IFX
                                                       2 years       NS
             Initial combination with   D-HAQ          1 and 2       NS
             CS vs Step-up                             years
             combination or Initial                                  Arm 1 (all
             combination with IFX                      over time     p<0.001)
                                                       (2 years)
             Initial combination with   D-HAQ          1 and 2       NS
             IFX vs Step-up                            years
             combination                                             Arm 1 (all
                                                       over time     p<0.001)
                                                       (2 years)
1 RCT        SC MTX vs Oral MTX         HAQ            24 weeks      NS
(Braun et
al)308
Level 1++
1 RCT307     MTX increasing dose        HAQ            2 years       Arm 1
Level 1++    vs MTX + prednisolone                                   (better)
             (decrease dose)
             MTX increasing dose        HAQ            2 years       Both similar
             vs MTX then add
             ciclosporin
             MTX + prednisolone         HAQ            2 years       Arm 2
             (decrease dose) vs                                      (better)
             MTX then add
             ciclosporin
             MTX + prednisolone +       HAQ            2 years       Arm 1
             ciclosporin vs all other                                (better)
             arms
TICORA       intensive vs routine       HAQ            18 months     Arm 1
study                                                                (p=0.0025)
(Grigor et
al) 288
Level 1++
1 cohort     starting on SSZ vs         HAQ            2 and 5       NS
study        starting on MTX                           years
(Hider et
al)309
Level 2+

Rheumatoid arthritis: full guideline DRAFT (August 2008)           151 of 307
    DRAFT FOR CONSULTATION

1
2         •   Joint damage
3
4   •     1 RCT – the BeSt study302-305 found that patients who do not achieve and
5         maintain DAS ≤2.4 with MTX, regardless of the success of consecutive
6         treatment steps, develop significantly more radiographic joint damage
7         compared to patients with DAS ≤2.4 on initial MTX. Level 1+
8
9
    Study          Sequence                   Outcomes           Follow-up      Result -
                                                                                best
                                                                                treatment
    MASCOT         SSZ then SSZ + MTX         Total Sharp        Change         NS
    study          vs SSZ continuous or       score, total       from 6 - 18
    (Capell et     SSZ then MTX               erosions (hands    months
    al)301                                    and feet), JSN

    Level 1++      SSZ then MTX vs SSZ        Total Sharp        Change         NS
                   continuous                 score, total       from 6 - 18
                                              erosions (hands    months
                                              and feet), JSN

    BeSt           Step-up combination        SHS score          Over the 2     Arm 1
    study302-      vs Sequential              (radiographic      years          (p=0.044)
    305
                   monotherapy                progression)
                                              over time (2
    Level 1+                                  years,
                                              Number of          1 year         NS
                                              patients with no
                                              progression of
                                              total SHS (>
                                              SDD)
                                              Total SHS          1 and 2        NS
                                                                 years
                                          Total SHS,             2 years        NS
                                          erosion score
                                          and JSN score
                   Sequential             total SHS (>           1 year         NS
                   monotherapy vs Initial SDD)
                   combination with CS or Total SHS,             1 and 2        Arm 2 (all
                   IFX                    erosion score          years          p<0.05, all
                                                                                p<0.001)
                                              JSN score          1 year         Arm 2 (all
                                                                                p<0.05)
                                                                 2 years        NS
                                              SHS score          Over the 2     Arm 2 (all
                                              (radiographic      years          p<0.001)
                                              progression)
                   Initial combination with   total SHS (>       1 year         NS
                   CS vs Step-up              SDD)


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                  152 of 307
DRAFT FOR CONSULTATION


             combination                Total SHS,       1 and 2        Arm 1 (all
                                        erosion score    years          p<0.05, all
                                                                        p<0.001)
                                        JSN score        1 and 2        NS
                                                         years
                                        SHS score        Over the 2     Arm 1 (all
                                        (radiographic    years          p<0.001)
                                        progression)
             Initial combination with   total SHS (>     1 year         NS
             CS vs Initial              SDD)
             combination with IFX       Total SHS,       1 and 2        NS
                                        erosion score,   years
                                        JSN score
                                        SHS score        Over the 2     Similar
                                        (radiographic    years
                                        progression)
             Initial combination with   JSN, total SHS   1 year         Arm 1
             IFX vs Step-up             (<SDD) and                      (p<0.05, rest
             combination                total SHS (>                    p≤0.001)
                                        SDD)
                                        Total SHS,       1 and 2        Arm 1 (all
                                        erosion score    years          p<0.05, all
                                                                        p<0.001)
                                   JSN                   2 years        NS
                                   SHS score             Over the 2     Arm 1 (all
                                   (radiographic         years          p<0.001)
                                   progression)
1 RCT307     MTX increasing dose   Cases of new          2 years        Arm 2
             vs MTX + prednisolone erosions,                            (better)
Level 1++    (decrease dose)       change in
                                   Larsen score
             MTX increasing dose   Cases with            2 years        Arm 1
             vs MTX then add       erosions,                            (better)
             ciclosporin           Change in
                                   Larsen score
             MTX + prednisolone    Change in             2 years        Arm 1
             (decrease dose) vs    Larsen score                         (better)
             MTX then add          Cases of new
             ciclosporin           erosions
                                                                        NS
             MTX + prednisolone +       Cases of new     2 years        Arm 1
             ciclosporin vs all other   erosions,                       (better)
             arms                       change in
                                        Larsen score
TICORA       intensive vs routine       Erosion score,   18 months      Arm 1
study                                   TSS                             (p=0.002,
(Grigor et                                                              p=0.02)
al) 288                                 JSN              18 months      NS
Level 1++


Rheumatoid arthritis: full guideline DRAFT (August 2008)              153 of 307
    DRAFT FOR CONSULTATION


    1 cohort     starting on SSZ vs      Larsen score,      5 years       NS
    study        starting on MTX         percentage of
    (Hider et                            patients with
    al)309                               erosions
    Level 2+                             percentage of      5 years       Arm 1
                                         patients with
                                         erosions
                                         (propensity
                                         adjusted)
1
2      •    Global assessment
3
    Study        Sequence                Outcomes           Follow-up     Result -
                                                                          best
                                                                          treatment
    MASCOT       SSZ then SSZ + MTX      Patient’s and      Change        NS
    study        vs SSZ continuous       physician’s        from 6 - 18
    (Capell et                           global             months
    al)301                               assessment

    Level 1++    SSZ then SSZ + MTX      Patient’s and      Change        NS
                 vs SSZ then MTX         physician’s        from 6 - 18
                                         global             months
                                         assessment

                 SSZ then MTX vs SSZ     Patient’s and      Change        NS
                 continuous              physician’s        from 6 - 18
                                         global             months
                                         assessment

    1 RCT       DMARD then DMARD         Patient’s and      18 months     Arm 1
    (Ferracciol + MTX vs DMARD           physician’s                      (p<0.05).
    i et al)306 then DMARD + SSZ         global             18-36
                                         assessment         months        NS
    Level 1+

                 DMARD then DMARD        Patient’s and      18 months     Arm 1
                 + CsA vs DMARD then     physician’s                      (p<0.05).
                 DMARD + SSZ             global             18-36
                                         assessment         months        NS
    TICORA       intensive vs routine    Patient’s and      18 months     Arm 1
    study                                assessor’s                       (p<0.0001)
    (Grigor et                           global
    al) 288                              assessment of
    Level 1++                            disease activity
4
5      •    Biochemical markers
6
    Study        Sequence                Outcomes           Follow-up     Result -
                                                                          best
                                                                          treatment

    Rheumatoid arthritis: full guideline DRAFT (August 2008)            154 of 307
    DRAFT FOR CONSULTATION


    MASCOT        SSZ then SSZ + MTX     ESR, CRP          Change        NS
    study         vs SSZ continuous                        from 6 - 18
    (Capell et                                             months
    al)301
                  SSZ then SSZ + MTX     ESR               Change        Arm 1
    Level 1++     vs SSZ then MTX                          from 6 - 18   (p=0.033)
                                         CRP               months        NS

                  SSZ then MTX vs SSZ    ESR, CRP          Change        NS
                  continuous                               from 6 - 18
                                                           months

    1 RCT       DMARD then DMARD         ESR; CRP          18 months     Arm 1
    (Ferracciol + MTX vs DMARD                             and           (p=0.01,
    i et al)306 then DMARD + SSZ                           18-36         p=0.001;
                                                           months        both
    Level 1+                                                             p=0.001)
                  DMARD then DMARD       ESR               18 months     NS
                  + CsA vs DMARD then                      and 18-36
                  DMARD + SSZ                              months
                                         CRP               18 months     NS

                                                           18-36         Arm 1
                                                           months        (p=0.03)
    TICORA        intensive vs routine   ESR               18 months     Arm 1
    study                                                                (p=0.0007)
    (Grigor et
    al) 288                              CRP                             NS
    Level 1++
    1 cohort      starting on SSZ vs     CRP               5 years       NS
    study         starting on MTX
    (Hider et
    al)309
    Level 2+
1
2      •    AEs
3
    Study         Sequence               Outcomes          Follow-up     Result -
                                                                         best
                                                                         treatment
    1 RCT       DMARD then DMARD         Number of         18 months     NS
    (Ferracciol + MTX vs DMARD           patients with
    i et al)306 then DMARD + SSZ         AEs               36 months     Arm 2
                                                                         (better)
    Level 1+      DMARD then DMARD       Number of         18 months     NS
                  + CsA vs DMARD then    patients with
                  DMARD + SSZ            AEs               36 months     Arm 2
                                                                         (better)




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           155 of 307
    DRAFT FOR CONSULTATION


    BeSt           Step-up combination        Number of          1 year, 2     NS
    study302-      vs Sequential              patients with      years and
    305
                   monotherapy                AEs and SAEs       over time
                                                                 (2 years)
    Level 1+       Sequential                 Number of          1 year and    NS
                   monotherapy vs Initial     patients with      2 years
                   combination with CS or     AEs and SAEs
                   IFX
                   Initial combination with   Number of          1 year and    NS
                   CS vs Step-up              patients with      2 years
                   combination                AEs and SAEs
                   Initial combination with   Number of          1 year and    NS
                   CS vs Initial              patients with      2 years
                   combination with IFX       AEs and SAEs

                   Initial combination with   Number of          1 year and    NS
                   IFX vs Step-up             patients with      2 years
                   combination                AEs and SAEs
    TICORA         intensive vs routine       Number of AEs      18 months     Arm 1
    study                                                                      (better:
    (Grigor et                                                                 N=46 vs
    al) 288                                                                    N=85)
    Level 1++
    1 RCT          SC MTX vs Oral MTX         percentage of      24 weeks      NS
    (Braun et                                 patients with at
    al)308                                    least 1
    Level 1++                                 moderate AE
                                              percentage of      24 weeks      Similar
                                              patients with
                                              SAEs
1
2         •   Withdrawals
3
    Study          Sequence                   Outcomes           Follow-up     Result -
                                                                               best
                                                                               treatment
    MASCOT         SSZ then SSZ + MTX         Total number of    18 months     Similar
    study          vs SSZ continuous or       withdrawals,
    (Capell et     SSZ then MTX               withdrawals due
    al)301                                    to AEs and due
                                              to lack of
    Level 1++                                 efficacy

                   SSZ then MTX vs SSZ        Total number of                  Similar
                   continuous                 withdrawals;
                                              withdrawals due
                                              to AEs and due
                                              to lack of
                                              efficacy


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 156 of 307
    DRAFT FOR CONSULTATION


    BeSt        Step-up combination        Number of             2 years        Similar
    study302-   vs Sequential              withdrawals
    305
                monotherapy
                Sequential                 Number of             2 years        Similar
    Level 1+    monotherapy vs Initial     withdrawals
                combination with CS or
                IFX
                Initial combination with   Number of             2 years        Similar
                CS vs Step-up              withdrawals
                combination
                Initial combination with   Number of             2 years        Similar
                CS vs Initial              withdrawals
                combination with IFX
    1 RCT       DMARD then DMARD           Withdrawals           18 months      Arm 1
    (Ferracciol + MTX vs DMARD             Withdrawals           36 months      (better)
    i et al)306 then DMARD + SSZ           due to toxicity                      Arm 1
                                                                                (p=0.0001)
    Level 1+     DMARD then DMARD          Withdrawals           18 months      Arm 1
                 + CsA vs DMARD then       Withdrawals           36 months      (better)
                 DMARD + SSZ               due to toxicity                      Arm 1
                                                                                (p=0.0001)
    1 cohort     starting on SSZ vs        Proportion of         Over the 2     Arm 2
    study        starting on MTX           patients with no      years          (better)
    (Hider et                              change in
    al)309                                 treatment
    Level 2+

1      •    Remission
2
    Study        Sequence             Outcomes                      Follow- Result -
                                                                    up      best
                                                                            treatment
    MASCOT       SSZ then SSZ +       Percentage of patients        18      Arm 1
    study        MTX vs SSZ           with EULAR good               months (better)
    (Capell et   continuous           response and
    al)301                            percentage in remission

    Level 1++    SSZ then SSZ +       Percentage of patients        18          Similar
                 MTX vs SSZ then      with EULAR good               months
                 MTX                  response and
                                      percentage in remission

                 SSZ then MTX vs      Percentage of patients        18          Similar
                 SSZ continuous       with EULAR good               months
                                      response and
                                      percentage in remission

    1 RCT       DMARD then            Magnusson criteria (full      3 years     Arm 1
    (Ferracciol DMARD + MTX           response)                                 (better)
    i et al)306 vs DMARD then

    Rheumatoid arthritis: full guideline DRAFT (August 2008)                  157 of 307
     DRAFT FOR CONSULTATION


                  DMARD + SSZ           Number of patients in full 3 years     Similar
     Level 1+                           remission (ACR)
                  DMARD then            Magnusson criteria (full   3 years     Arm 1
                  DMARD + CsA vs        response)                              (better)
                  DMARD then
                  DMARD + SSZ           Number of patients in full 3 years     Similar
                                        remission (ACR)
     TICORA       intensive vs          EULAR good response, Over              Arm 1 (both
     study        routine               EULAR remission            the 2       p<0.0001)
     (Grigor et                                                    years
     al) 288
     Level 1++
     1 cohort     starting on SSZ       Percentage of patients in 2 and 5      NS
     study        vs starting on        remission                 years
     (Hider et    MTX
     al)309
     Level 2+

1    7.3.10        Health economics evidence statements (optimal
2    sequencing of DMARDs)
 3   Chen et al284 was the only study found that was UK based. The model was a
 4   Patient level simulation and considered adalimumab, etanercept, infliximab +
 5   MTX, adalimumab + MTX, etanercept + MTX against a sequence of
 6   DMARDs. Spalding et al developed a Markov model considering adalimumab,
 7   etanercept, infliximab + MTX, adalimumab + MTX against MTX.
 8
 9   For first line use of biologics, with our without MTX, Chen et al284estimate
10   ICERs well in excess of the usual NICE threshold. First line use of infliximab is
11   associated with extremely high ICERs in both studies. Adalimumab and
12   etanercept without MTX generate lower ICERs in both studies but these are
13   still unlikely to be considered cost effective. Third line use of each of the three
14   anti-TNFs, as reflected in current NICE guidance, may be cost effective
15   strategies depending on the threshold.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)               158 of 307
DRAFT FOR CONSULTATION


Drug           Comparator     Study        Date Time     ICER
                                                horizon
Adalimumab DMARD              Chen284      2006 Lifetime Adalimumab (no MTX)
           sequence                                            £35k per QALY

                                                           Adalimumab (with
                                                           MTX) £30k per QALY
                                                           (Third line (early RA
                                                           data))
                                                           Adalimumab (no MTX)
                                                                   £53k per QALY

                                                           Adalimumab (with
                                                           MTX) £170k per
                                                           QALY
                                                           (First line (Early RA
                                                           data))
               Methotrexate Spalding286 2006 Lifetime      Adalimumab (no MTX)
                                                                    $64k per QALY

                                                           Adalimumab (with
                                                           MTX) $195k per
                                                           QALY
Etanercept     DMARD          Chen284      2006 Lifetime   Etanercept (no MTX)
               sequence                                          £30k per QALY

                                                           Etanercept (with MTX)
                                                                   £28k per QALY
                                                           (Third line (early RA
                                                           data))
                                                           Etanercept (no MTX)
                                                                   £49k per QALY

                                                           Etanercept (with MTX)
                                                                    £78k per QALY
                                                           (First line (Early RA
                                                           data))
               Methotrexate Spalding286 2006 Lifetime      Etanercept (no MTX)
                                                                    $90k per QALY

Infliximab     DMARD          Chen284      2006 Lifetime   Infliximab (with MTX)
               sequence                                             £30k per QALY
                                                           (Third line (early RA
                                                           data))
                                                           Infliximab (with MTX)
                                                                    £650k per
                                                           QALY
                                                           (First line (Early RA
                                                           data))
               Methotrexate Spalding286 2006 Lifetime      Infliximab (with MTX)
                                                                    $410k per
Rheumatoid arthritis: full guideline DRAFT (August 2008)   QALY159 of 307
     DRAFT FOR CONSULTATION


1    7.3.11         Summary of evidence statements
2         •   For many patients, monotherapies such as methotrexate work well
              302-304 305
3                    .
4         •   Sequential monotherapies comparing methotrexate and
5             sulphasalazine do not show any obvious differences between the
6             drugs 301,309, although other studies comparing sulphasalazine with
7             methotrexate and cyclosporin in combination therapies suggest that
8             the latter two are superior for some outcomes to the former 306
 9        •   If a patient fails on methotrexate monotherapy, the chances they will
10            have a good response to other conventional DMARDs are less 305.
11        •   For symptoms, quality of life, ability to achieve remission, and
12            slowing joint damage, a variety of combination therapies (including
13            steroids and infliximab) appear to be superior to monotherapy 299,300
              292,293 294-297,301,310 302-304 290,305
14                                                    .
15        •   For other outcomes such as function there is less evidence of a
16            difference292-297,299-304 305.
17        •   There is no difference in tolerability between monotherapies and
18            combination therapies292,293,299,300 294-297,302-304,310 290,305.
19        •   Some studies show a convergence of outcomes between the arms of
20            the trial over time 292,293,299,300
21        •   Many of the studies showed similar improvements with a variety of
22            monotherapies and combination therapies, and suggested that the
23            type and combination of drug used was less important than the speed
24            and intensity of the DMARD introduction299,300 292,293 294-297,301,310 302-
              304 290,305
25                        . Intensity refers to the rapid escalation of DMARD to
26            therapeutic doses.

27   7.3.12         From evidence to recommendations
28   The GDG noted that there was overwhelming evidence to support the strategy
29   of early introduction of DMARDs. However, the GDG were mindful that all
30   trials had been conducted in patients with active inflammatory joint disease,
31   and there is currently no evidence to support the approach that must be taken
32   in mild synovitis, or palindromic arthritis where the disease activity waxes and
33   wanes with periods of activity and inactivity. It was therefore felt that the early
34   introduction of DMARDs in RA needed to relate specifically to active RA, and
35   the treatment of mild disease needed to be considered as a research
36   recommendation.
37
38   The key area that had been agreed for health economic modelling (see
39   2.8.1.9) was an assessment of the cost effectiveness of various DMARD
40   strategies in patients with recent onset of RA, including the initiation therapy
41   with combinations of disease modifying drugs. The results of this analysis
42   (see Appendix C) demonstrated clear benefits of a combination strategy,
43   compared with monotherapy, and demonstrated that step-down combinations
44   of DMARDs are likely to be very cost-effective or even cost-saving, and other

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                160 of 307
     DRAFT FOR CONSULTATION


 1   DMARD combinations are very likely to be cost-effective. The GDG felt that
 2   the opportunity should be taken of now recommending this approach as first
 3   line treatment unless specifically contraindicated.
 4
 5   There is evidence that methotrexate is at least as efficacious as other DMARD
 6   monotherapies, if not more so in some studies. Rapid acceleration of the drug
 7   can achieve excellent response in most patients. NICE require methotrexate
 8   to be tried before anti-TNF therapies can be considered. The GDG was also
 9   mindful of emerging evidence to suggest that if a patient fails on methotrexate,
10   they are unlikely to respond to other conventional DMARDs, and
11   consideration should be given to anti-TNF therapies once a patient has failed
12   on a trial of at least two conventional DMARDs (including methotrexate).
13
14   Taken all these considerations into account, the GDG felt that methotrexate
15   should be included as the first DMARD therapy, either as monotherapy or as
16   part of a combination of other therapies. Furthermore, the most successful
17   and cost-effective step-down and combination therapy regimes had all used
18   steroids in one form or another (either orally in a tapered dose, intra-
19   articularly, intra-muscularly or a combination of these approaches) and it was
20   therefore felt that steroids should be specifically mentioned in combination
21   regimes. Although the basecase analysis determined that a strategy of
22   monotherapy plus glucocorticoids is more costly and less effective than
23   DMARD monotherapy (due to the cost-differential not being covered by a
24   significant increase in ACR response rate) (Please see Appendix C for more
25   information) the GDG felt it was important to note that step up and step down
26   all include glucocorticoids in the regimen, so that no combination regime has
27   been shown to work without steroids in some form (oral, intramuscular or
28   intra-articular). Until studies have demonstrated that combination regimes can
29   be used efficaciously in the absence of steroids, the GDG felt that they should
30   be mentioned in the recommendation.
31
32   It was agreed that no lower limit should be placed on the amount of time that
33   needed to pass before the introduction of a DMARD in early RA; indeed such
34   therapy should be initiated as soon as possible. Although there was no good
35   evidence to support the concept of an upper limit by which time DMARD
36   therapy should have been started in order to achieve long-term benefits on
37   disease outcomes (the “window of opportunity” referred to in the Clinical
38   Introduction), the GDG felt that the initiation of DMARD therapy within three
39   months of the onset of persistent symptoms was supported by some data and
40   should therefore be recommended as an ideal target.
41
42   In keeping with this initial aggressive treatment of disease it was noted that
43   there was a need to try and decrease the doses of drugs once satisfactory
44   disease control had been achieved, and that this would in turn involve
45   appropriate monitoring (see recommendations in section 8.1)
46   The need for early initiation of therapy would also support the
47   recommendations about the need for urgent referral to specialist care of
48   people with persistent idiopathic synovitis, even in the absence of abnormal
49   tests, and especially if symptoms have already been present for more than
50   three months (see recommendations in section 8.2).

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             161 of 307
     DRAFT FOR CONSULTATION


 1   Although methotrexate appears to be the drug of choice for use in
 2   combination therapies in patients with a recent onset of rheumatoid arthritis, it
 3   was acknowledged that would be circumstances under which other drugs
 4   would be chosen (such as co-morbidities and contraindications to
 5   methotrexate), and perhaps used as monotherapies. Other DMARD
 6   monotherapies and combination therapies that exclude methotrexate can
 7   work very satisfactorily. It was therefore decided that where combination
 8   therapies were not indicated, the type of drug used is less important than the
 9   promptness of initiating therapy and the rapid escalation to, and maintenance
10   at a therapeutic dose of DMARD. The GDG felt that there should be a
11   recommendation that embraced these concepts

12   7.3.13            Recommendations
13   R 25 In people with recent onset of active rheumatoid arthritis, treatment with
14   a combination of disease modifying drugs (including methotrexate and
15   glucocorticoids) should be offered as first-line treatment as soon as possible,
16   ideally within three months of the onset of persistent symptoms.
17
18   R 26 In people with recent onset of rheumatoid arthritis receiving combination
19   DMARD therapy and in whom sustained and satisfactory levels of disease
20   control have been achieved, cautious attempts should be made to reduce
21   drug doses to levels which still maintain disease control.
22
23   R 27 In people with a recent onset of RA where combination therapy is not
24   appropriate†† and DMARD monotherapy is started, greater emphasis should
25   be placed on rapid escalation to a clinically effective dose, rather than the
26   choice of DMARD.
27

28   7.4       Disease modifying and biological drugs: relative merits
29   Anakinra

30   7.4.1 Clinical Introduction
31   Interleukin-1 (IL-1) is a pro-inflammatory cytokine. There is much evidence to
32   implicate this molecule in the pathogenesis of RA where it promotes cartilage
33   destruction and bone resorption. Anakinra is a recombinant form of human IL-
34   1 receptor antagonist that inhibits the activity of IL-1, thus theoretically
35   protecting both cartilage and bone. It is licensed for use in combination with
36   methotrexate in patients who have had an inadequate response to
37   methotrexate alone. NICE published a technology appraisal of anakinra in
38   November 2003 (TA 72). The appraisal committee could not recommend
39   anakinra on the balance of its clinical benefits and cost effectiveness, with the
40   cost per QALY estimated to be in excess of £69000. The technology appraisal
41   recommended further studies to evaluate the long-term effectiveness and
42   safety of anakinra, and comparative trials with DMARDS and TNFα inhibitors


     ††
          e.g. presence of comorbidities or pregnancy where certain drugs would be contraindicated

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                         162 of 307
     DRAFT FOR CONSULTATION


1    to guide clinical practice. The GDG have been asked to update this Anakinra
2    technology appraisal within this guideline.

3    7.4.2 Clinical methodological introduction
 4   We looked for studies that investigated the efficacy and safety of Anakinra
 5   alone or in combination with another DMARD versus placebo or other drug
 6   treatment with respect to symptoms, function, quality of life and ability to
 7   beneficially modify structural changes of RA. Trials were selected in
 8   accordance with the criteria used for the NICE HTA on Anakinra311 as it was
 9   within the remit of this guideline to update the HTA. All trials published after
10   the cut-off date used in the HTA were considered as evidence. Two
11   RCTs312,313 were found that fulfilled the criteria. The Cohen paper was the
12   same trial as one already included in the HTA, but has reported additional
13   outcomes. The GDG wished to address the issue of longer-term effects of
14   anakinra on patients with RA and identified 3 additional papers which were
15   extension studies of RCTs. One of these314 was excluded since it solely
16   reported AEs, the second and third315,316 were results of one trial but reported
17   different outcomes. These were thus included as additional evidence.
18   The 2 RCTs312,313 were both double blind, parallel group studies and
19   assessed patients who concurrently continued with their usual MTX treatment.
20
21   The first RCT (Cohen et al)312 was a 6-arm study comparing anakinra vs
22   placebo in N=419 patients. Patients were treated once a day with either
23   anakinra (5 groups with doses ranging from 0.04 mg/kg to 2.0 mg/kg) or
24   placebo (1 group) in a 12 week or 24 week treatment phase. The second RCT
25   (Genovese et al.)313 was a 3-arm study comparing anakinra + biologic vs
26   biologic in N=244 patients. Patients were treated with anakinra (100 mg four
27   times/week) + biologic (2 groups: etanercept 25 mg either once or twice/week)
28   vs biologic (1 group: etanercept, 25 mg twice/week) in a 24 week treatment
29   phase with follow-up at 4 weeks post-treatment or at time of early
30   discontinuation.
31
32   The extension study315,316 included patients who were originally randomised to
33   either anakinra or placebo for 24 weeks of treatment. In the extension phase
34   (which lasted 52 weeks – ie. total study was 76 weeks) all patients were given
35   anakinra (30 mg, 75 mg, 150 mg). Patients in the placebo group were
36   randomised to anakinra and those already in the anakinra groups remained
37   on the dose they were originally randomised to (30 mg, 75 mg, 150 mg).
38   Results are reported for week 48 (ie. 24 weeks into the extension phase).

39   7.4.3 Health economic methodological introduction
40   Health economic evidence was not formally reviewed as there is an existing
41   technology appraisal model. Drug costs and clinical evidence have not
42   changed enough for anakinra to become a cost-effective therapy.9

43   7.4.4 Clinical evidence statements:
44
45   Anakinra – extension studies

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             163 of 307
     DRAFT FOR CONSULTATION


1    Symptoms
2       • The extension study315,316 found the following at 48 weeks (24 weeks
3          into the extension phase of anakinra treatment):
 4              • In the original placebo group there was a significantly higher
 5                proportion of patients who achieved an ACR20 response
 6                (p=0.007) and sustained ACR20 response (p<0.001) compared
 7                with the response at week 24. For the individual doses of
 8                anakinra there were no significant differences in number of
 9                patients achieving ACR20 response compared with week 24.
10                However there were significant differences in number of patients
11                achieving sustained ACR20 response for anakinra 75mg
12                (p=0.016) and anakinra 150mg (p=0.022). Level 1+
13              • In the original placebo group ACR50 increased compared to
14                week 24, however ACR70 remained unchanged. Level 1+
15              • In the original placebo group there was a significant
16                improvement compared to week 24 for the number of swollen
17                joints (p<0.001), number of tender joints (p<0.001), pain
18                assessment (p<0.005), CRP (p<0.005), and ESR (p<0.001). For
19                the individual doses of anakinra there were significant
20                improvements compared to week 24 for anakinra 30mg: CRP
21                (p<0.05), and ESR (p<0.005); anakinra 75mg: number of
22                swollen joints (p<0.05), number of tender joints (p<0.005), and
23                ESR (p<0.001); anakinra 150mg: number of swollen joints
24                (<0.005), number of tender joints (p<0.05). Level 1+
25              • In the original anakinra groups combined, compared to week 24
26                there was NS difference in the proportion of patients who
27                achieved an ACR20 response or sustained ACR20 response.
28                There was also NS difference at week 24 for each of the
29                individual doses of anakinra. ACR50 increased compared to
30                week 24, however ACR70 remained unchanged. Level 1+
31              • In the original anakinra groups combined, compared to week 24
32                there was no difference for number of swollen joints, number of
33                tender joints, pain assessment, CRP (p<0.005), and ESR. In the
34                anakinra 150 mg group there was a significant deterioration in
35                assessment of pain (p<0.05). Level 1+
36
37        •   The extension study315,316 found the following at 48 weeks (24 weeks
38            into the extension phase of anakinra treatment):
39              • There was significantly less joint damage in all groups (those
40                originally randomised to placebo or anakinra) compared to that
41                at 24 weeks (p<0.001). Level 1+
42              • In the original placebo group there was a significant reduction in
43                TMSS, modified Sharp erosion score and modified Sharp joint
44                narrowing score for all anakinra doses (p<0.001). Level 1+
45              • In the original anakinra groups TMSS and modified Sharp
46                erosion score were significantly lower for the higher anakinra
47                doses (75 and 150 mg/day), with no significant difference for the



     Rheumatoid arthritis: full guideline DRAFT (August 2008)            164 of 307
     DRAFT FOR CONSULTATION


1                 30 mg/day dose and for the modified Sharp joint narrowing score
2                 at any dose. Level 1+

3    Function
4        • The extension study315,316 found the following at 48 weeks (24 weeks
5           into the extension phase of anakinra treatment):
 6              • In the original placebo group there was a significant
 7                improvement compared to week 24 for HAQ score (p<0.001).
 8                For the individual doses of anakinra there were significant
 9                improvements compared to week 24 for anakinra 30mg: HAQ
10                (p<0.005); anakinra 150mg: HAQ (p<0.005). Level 1+
11              • In the original anakinra groups combined, compared to week 24
12                there was a significant deterioration in the HAQ score (p<0.05).
13                In the anakinra 150 mg group there was a significant
14                deterioration in HAQ (p<0.05). Level 1+
15
16   Global assessment
17        •   The extension study315,316 found the following at 48 weeks (24 weeks
18            into the extension phase of anakinra treatment):
19              • In the original placebo group there was a significant
20                improvement compared to week 24 for patient global
21                assessment (p<0.05) and investigator assessment (p<0.05). For
22                the individual doses of anakinra there were significant
23                improvements compared to week 24 for anakinra 30mg: patient
24                global assessment (p<0.05). Level 1+
25              • In the original anakinra groups combined, compared to week 24
26                there was a significant deterioration in patient global assessment
27                and investigator assessment. In the anakinra 150 mg group
28                there was a significant deterioration in patients global
29                assessment. Level 1+
30
31   Study Withdrawals
32        •   The extension study315,316 found the following at 48 weeks (24 weeks
33            into the extension phase of anakinra treatment):
34              • Rates of withdrawal was similar compared to 24 weeks
35              • Withdrawals due to AEs was similar compared to 24 weeks
36              • Arthritis flare was the most common reason for withdrawal due
37                to AEs. Level 1+
38
39   Anakinra vs placebo
40
41   Function
42       • One RCT312 (N=419) found that for the outcome of HAQ-DI (change
43          from baseline), Anakinra 1.0 mg/kg and 2.0 mg/kg were significantly
44          better than placebo at 12 weeks (1.0 mg: -0.35, p<0.05; 2.0 mg: -
45          0.39, p<0.01) and 24 weeks, end of study (1.0 mg: -0.37, p<0.05; 2.0


     Rheumatoid arthritis: full guideline DRAFT (August 2008)            165 of 307
     DRAFT FOR CONSULTATION


1              mg: -0.51, p<0.01). However, there were NS differences between
2              Anakinra 0.04 mg/kg, 0.1 mg/kg, 0.4 mg/kg and placebo. Level 1++
3          •   The same RCT312 (N=419) found that for the outcome of percentage
4              of patients reporting no impairment of function (HAQ-DI = 0),
5              Anakinra 1.0 mg/kg was significantly better than placebo at week 24,
6              end of study (18.6% and 5.4% respectively, p<0.05; OR 4.76, 95% CI
7              1.1 to 20.0). However, there were NS differences between Anakinra
8              0.04 mg/kg, 0.1 mg/kg, 0.4 mg/kg, 2.0 mg/kg and placebo. Level 1++
 9
10   Anakinra + biologic vs biologic

11   Symptoms (all Level 1+)
12      • One RCT313 (N=244) found that for the outcome of ACR 20,
13         Etanercept was significantly better than anakinra + etanercept
14         once/week (68% and 51% respectively; OR 1.98, 95% CI 1.05 to
15         3.78; p=0.037) at 24 weeks (end of treatment). However there was
16         NS difference between etanercept and anakinra + etanercept
17         (twice/week).
18         •   The same RCT313 (N=244) found that for the outcome of ACR 50,
19             there was NS difference between Etanercept and anakinra +
20             etanercept (once/week or twice/week) at 24 weeks (end of
21             treatment).
22   •    The same RCT313 (N=244) found that for the outcome of ACR 70, there
23       was NS difference between Etanercept and anakinra + etanercept
24       (once/week or twice/week) at 24 weeks (end of treatment).
25         • The same RCT313 (N=244) found that for the outcome of EULAR
26            response, Etanercept was better than anakinra + etanercept
27            once/week (79% and 66% patients respectively) and twice/week
28            (79% and 73% patients respectively) at 24 weeks (end of treatment).
29         •   The same RCT313 (N=244) found that for the outcome of DAS score
30             (% reduction), Etanercept was similar to anakinra + etanercept
31             once/week (39% and 40% patients respectively) and twice/week
32             (39% and 41% patients respectively) at 24 weeks (end of treatment).
33   AEs
34         •   One RCT313 (N=244) found that for the outcomes of Number of
35             SAEs, number of infections and number of serious infections,
36             Etanercept was better than anakinra + etanercept once/week and
37             twice/week over 24 weeks (end of treatment). Level 1+
38   Study Withdrawals
39         •   One RCT313 (N=244) found that for the outcome of number of
40             withdrawals due to AEs, Etanercept was significantly better than
41             anakinra + etanercept once/week (0% and 8.6% respectively, p-value
42             not given) and twice/week (0% and 7.4% respectively, p value not
43             given) at 24 weeks (end of treatment).



     Rheumatoid arthritis: full guideline DRAFT (August 2008)           166 of 307
     DRAFT FOR CONSULTATION


1         •   The same RCT313 (N=244) found that for the outcome of number of
2             withdrawals, Etanercept was better than anakinra + etanercept
3             once/week (7% and 12% respectively) and twice/week (7% and 20%
4             respectively) at 24 weeks (end of treatment).

5    7.4.5 Summary of evidence statements
6         •   Anakinra in different doses works well in an extension study in
7             patients previously randomised to placebo for a variety of disease
8             activity measures, radiological 315,316 and functional outcomes
              312,315,316
9
10        •   The drug appears to be well tolerated 315,316
11        •   Etanercept was better alone than in combination with anakinra for
12            some measures of efficacy, tolerability and a reduced rate of
13            infections 313

14   7.4.6 From evidence to recommendations
15   Since publication of the NICE Technology Appraisal on anakinra in 2003, new
16   high quality data has been sparse. Extension studies have not shown any
17   evidence to suggest a significant improvement in efficacy of anakinra after the
18   first 24 weeks of treatment, or of any effect on decline in HAQ (the driver of
19   economic modelling in the NICE TA) after 24 weeks of follow-up. The GDG
20   noted that the cost of anakinra had reduced, but did not feel that this was
21   likely to bring it within the NICE cost effectiveness range if new health
22   economic modelling was undertaken.
23
24   It was also noted that the addition of etanercept to anakinra resulted in no
25   improvement of efficacy, with many outcome measures suggesting that the
26   combination was inferior to using etanercept alone. The combination also
27   resulted in a significant increase in serious infections. The GDG felt that it was
28   reasonable to extrapolate this finding to all anti-TNF therapies, given that they
29   all have a similar propensity for increasing the risk of infection.
30
31   The GDG concluded that, on the basis of the evidence reviewed, the current
32   TA recommendations should stand, but with an additional recommendation
33   cautioning against the co-prescribing of anakinra and anti-TNF therapy.

34   7.4.7 Recommendations
35   R 28 Anakinra is not recommended for the treatment of rheumatoid arthritis,
36   except in the context of a controlled, long-term clinical study
37
38   R 29 Patients currently receiving anakinra for rheumatoid arthritis may suffer
39   loss of wellbeing if their treatment were discontinued at a time they did not
40   anticipate. Therefore, patients should continue with anakinra until they and
41   their consultant consider it is appropriate to stop.
42



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              167 of 307
     DRAFT FOR CONSULTATION


1    R 30 In patients currently receiving anakinra for rheumatoid arthritis where a
2    change in treatment is being considered, the combination of anakinra and
3    anti-TNF therapy is not recommended.
4
5    Biological drugs vs conventional disease-modifying drugs in patients
6    with established rheumatoid arthritis where there is ongoing disease
7    activity.

8    7.4.8 Clinical Introduction
 9   Although biological therapies have had a tremendous impact on the
10   management of RA, so too have conventional DMARDs. Furthermore,
11   biological drugs are substantially more expensive than conventional DMARDs.
12   It is appropriate therefore to question the comparative efficacy of these new
13   therapies, and to determine what additional value they add to the
14   management of RA compared with cheaper and more established drugs.
15   Some biological drugs have already completed a NICE technology appraisal
16   such as the rituximab and the anti TNF drugs adalimumab, etanercept and
17   infliximab. Abatacept has recently been appraised and deemed not to be cost-
18   effective for use in the NHS, other biological drugs such as certolizumab pegol
19   have yet to begin the technology appraisal process

20   7.4.9 Clinical methodological introduction
21   We looked for studies that investigated the efficacy and safety of biological
22   drugs vs DMARDs (singly or in combination) with respect to symptoms, joint
23   damage, function and quality of life in patients with established RA. Due to the
24   large volume of evidence, only RCTs were selected which had a sample size
25   of N≥100 and were of a UK-relevant population.
26
27   Two MAs317,318, Ten RCTs319-332 and one extension study333 of 2 RCTs were
28   found that fulfilled the criteria. One of the RCTs (TEMPO trial) was published
29   as five separate papers320,321,328,331,332 reporting different outcomes and follow-
30   up times, so the trial has only been counted once, however results from all the
31   papers are reported and referenced here. All trials were methodologically
32   sound. All trials (including the MAs were in patients with established RA)
33
34   The first SR/MA317 focused on 6 double-blind RCTs with N=2381 patients
35   which compared adalimumab monotherapy or in combination with DMARDs
36   vs placebo or other DMARDs. Both the MA itself and the studies it included
37   were well conducted. Studies included in the analysis differed with respect to:
38         • Intervention – dose given and regimen
39        •   Study size (range N=54 to N=636)
40        •   Study duration – length of intervention (12 weeks to 52 weeks)
41   The second SR/MA318 focused on 3 double-blind RCTs with N=1040 patients
42   which compared anti-TNF + MTX vs MTX with a treatment time between 50 to
43   55 weeks. Both the MA itself and the studies it included were well conducted.
44   Studies included in the analysis differed with respect to:
45

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               168 of 307
     DRAFT FOR CONSULTATION


1         •   Intervention (1 RCT Infliximab + MTX, 1 RCT Etanercept + MTX, 1
2             RCT Adalimumab + MTX)
3         •   Intervention – dose given and regimen
4         •   Study size (range N=174 to N=459)
 5   NOTE: The two MAs317,318 included RCTs already included in this section.
 6   However it was felt important to report the RCTs separately in order to see the
 7   effects of the individual drugs for all outcome measures, since the MAs either
 8   pooled drug classes together or only pooled data for some outcomes (since
 9   not all the trials used the same outcome measures).
10
11   All 10 RCTs were parallel group but were very variable in terms of treatment
12   and had variable inclusion criteria. The trials differed with respect to the
13   following:
14         • Sample size (range: N=161 to N=2987)
15        •   Blinding (9 RCTs double blind, 1 RCT unblinded)
16        •   Trial length (range: 16 weeks to 4 years)
17        •   Treatment – type of biologics and DMARDs used
18        •   Treatment regimen – single drugs and combinations compared
19        •   Treatment regimen – dose
20   The included extension study333 included patients who were originally
21   randomised to either etanercept (10, 25 or 50 mg) or placebo in 2 RCTs. In
22   the 5-year extension phase, all patients were given etanercept 25 mg
23   twice/week.
24
25   NOTE: In most of the trials the population consisted of patients who were not
26   doing well on / not responding to their DMARD therapy. These patients were
27   then randomised to either continue on this DMARD or to take a biologic drug.
28   Therefore there may be some bias in the study design in favour of the biologic
29   drug because those in the trial arm that continue on the their usual DMARD
30   which is not working well, are unlikely to do well compared to those who are
31   put onto a biologic drug instead.

32   7.4.10        Health economic methodological introduction
33   In the HTA report “A systematic review of the effectiveness of adalimumab,
34   etanercept and infliximab for the treatment of rheumatoid arthritis in adults and
35   an economic evaluation of their cost-effectiveness” by Chen et al 2006284, a
36   review of publications relating to economic evaluations of biologics was
37   performed. In a search of publications subsequent to the review by Chen et
38   al284, twelve papers were found271,334-343 and were appraised.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             169 of 307
    DRAFT FOR CONSULTATION




1   Table 2 Summary of published economic analyses

    Study         Disease          County       TNF inhibitor(s) considered      Form of         Model       Time
                  Duration (yrs)                                                 economic        used        horizon
                                                                                 analysis
    Choi et al                     USA          Etanercept                       Cost-           Decision    6 months
    (2002) 337                                                                   effectiveness   tree
    Brennan et                     UK           Etanercept                       Cost-utility    Patient     Lifetime
                  Not given
    al            Failed 3.3.                                                                    level
    (2004)338     DMARDs
    Wong et al    Not stated       USA          Infliximab                       Cost-utility    Markov      Lifetime
    (2002) 336
    Kobelt et al Response:         Sweden       Infliximab                       Cost-utility    Markov      10 years
    (2003) 339   11yrs             and UK
                 Baseline: 8.2
                 months
    Kobelt et al 14.2              Sweden       Etanercept, infliximab           Cost utility    Not         Not
    (2004) 340                                                                                   applicabl   applicable
                                                                                                 e
    Chiou et al                    USA          Etanercept, infliximab,          Cost-utility    Decision    1 year
    (2004) 343                                  adalimumab                                       tree
    Welsing et    3.8 minimum      Netherland   Etanercept                       Cost-utility    Markov      5 years
    al (2004)                      s
    335

    Bansback     over 8yrs         Sweden       Etanercept, infliximab,          Cost-utility    Patient     Lifetime
    et al                                       adalimumab                                       level
    (2005) 334
    Kobelt et al 6.8               Sweden       Etanercept                       Cost-utility    Markov      10 years



    Rheumatoid arthritis: full guideline DRAFT (August 2008)              170 of 307
    DRAFT FOR CONSULTATION



    (2005) 344
    Spalding et   3 months         USA          Adalimumab, Etanercept,      Cost utility   Markov    Lifetime
    al                                          infliximab + methotrexate,
    (2006)286                                   adalimumab + methotrexate
    Tanno et al   11               Japan        Etanercept                   Cost utility   Markov    Lifetime
    (2006)342
    Marra et al   9                Canada       infliximab + methotrexate    Cost utility   Markov    10 years
    (2007)341
    Chen et al    Various          UK           Adalimumab, Etanercept,      Cost utility   Patient   lifetime
    (2006) 284                                  infliximab + methotrexate,                  level
                                                adalimumab +
                                                methotrexate, Etanercept+
                                                methotrexate
    Brennan et    14.1             UK           Etanercept, infliximab,      Cost utility   Patient   lifetime
    al                                          adalimumab                                  level
    (2007)345
1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)          171 of 307
    DRAFT FOR CONSULTATION



1   7.4.11        Clinical evidence statements
2   Established RA
3
4   A. POOLED DATA – DRUG CLASS
5
    Study        Intervention    Outcomes                       Result - best
                                                                treatment
    1 MA318      Anti-TNF + ACR 70; withdrawals due             Arm 1: (3 RCTs,
                 MTX vs MTX to lack of efficacy                 N=1040, RR 3.43,
    Level 1++                                                   95% CI 1.74 to
                                                                6.75, p=0.0004)
                                 withdrawals due to AEs         NS

                 IFX vs ADA      ACR20, ACR50, ACR70,           NS
                                 withdrawals due to AEs or
                                 lack of efficacy

                 ADA vs ETN      ACR20, ACR50, ACR70            ADA (p<0.001)
                                 withdrawals due to AEs         ETN (p=0.02)
                                 withdrawals due to lack of     NS
                                 efficacy
                 IFX vs ETN      ACR20                          IFX (p=0.001),
                                 ACR50, ACR70,                  NS
                                 withdrawals due to AEs or
                                 lack of efficacy
6
7   B. POOLED DATA – ADALIMUMAB
8
    Study       Intervention      Outcomes                       Follow-    Result - best
                                                                 up         treatment
    1 MA317     ADA sc 40mg       ACR50 and ACR70, HAQ,          24         Arm 1 (all
                eow + MTX (or     tender joints, patient pain    weeks      p<0.00001)
                DMARDs) vs        assessment
                placebo sc +
                MTX (or           Withdrawals; withdrawals                  NS
                DMARDs)           due to AEs; AEs and
                                  SAEs (all doses of ADA)

                                  significant heterogeneity                 ACR20
                ADA sc 20 mg      ACR20 and ACR50;               2          ADA:
                ew vs placebo     withdrawals (ADA all           weeks      p<0.0001 and
                                  doses)                                    p=0.04;
                                                                            p<0.00001)
                ADA sc 40 mg      ACR20 and ACR50                           ADA
                ew vs placebo                                               (p=0.0004 and
                                                                            p=0.009)
                ADA at 40mg       ACR20                          24/46      ADA
                eow vs placebo                                              (p=0.009)


    Rheumatoid arthritis: full guideline DRAFT (August 2008)               172 of 307
    DRAFT FOR CONSULTATION


                   ADA (all doses)    withdrawals                  weeks     ADA
                   vs placebo                                                (p<0.00001)
                                      SAEs, withdrawals due to               NS
                                      AEs,
                                      AEs                                    Significant
                                                                             heterogeneity
1
2   C. DATA FOR INDIVIDUAL DRUGS
3
4   1. ETANERCEPT
5
    Study         Intervent   Outcomes                                Follow- Result -
                  ion                                                 up      best
                                                                              treatment
    1             ETN vs      ACR20, ACR50, ACR70; DAS score,         24      ETN (all
    RCT324        SSZ         painful joints, swollen joints, Pain    weeks   p<0.05; all
    Level                     (VAS), Morning stiffness and                    p<0.01)
    1++                       EuroQoL, HAQ, patient’s and
                              physician’s global assessments, ESR,
                              CRP

                              withdrawals due to AEs                             NS

                  ETN +       ACR20, ACR50, ACR70; DAS score,                    Arm 1 (all
                  SSZ vs      painful joints, swollen joints, Pain               p<0.05; all
                  SSZ         (VAS), Morning stiffness and                       p<0.01)
                              EuroQoL, HAQ, patient’s and
                              physician’s global assessments, ESR,
                              CRP
                              withdrawals due to AEs                             NS
                  ETN +       ACR50                                              Arm 1
                  SSZ vs                                                         (p<0.05)
                  ETN         ACR20, ACR70, DAS score, painful                   NS
                              joints, swollen joints, Pain (VAS),
                              Morning stiffness and EuroQoL, HAQ,
                              patient’s and physician’s global
                              assessments, ESR, CRP, withdrawals
                              due to AEs
    1             ETN vs      ACR (AUC); patients achieving a          52        ETN
    RCT320,       MTX         major improvement of HAQ >0.8            weeks     (p=0.0034)
    321,328,332
                              patients achieving remission, HAQ,                 NS
    Level                     withdrawals due to AEs
    1++                       patient satisfaction, patients achieving 2 years   ETN (all
                              remission , Total Sharp Score, erosion             p<0.05)
                              score, JSN, total sharp score, erosion
                              score, patients with no erosions
                              number of painful joints                           Placebo




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                173 of 307
DRAFT FOR CONSULTATION


                         EQ5D, Patient global assessment of                    NS
                         overall RA activity, Patient General
                         Health Assessment, number of
                         swollen joints, Pain (VAS), patient and
                         physicians global assessment, HAQ,
                         CRP, AEs
                         ACR20, ACR50, ACR70, DAS score              52        NS
                                                                     weeks
                                                                     and at
                                                                     2 years

                         Patients achieving remission,               3 years   ETN (all
                         Radiographic progression (change                      p<0.05)
                         from baseline in TSS and erosion
                         score), JSN, total number of
                         withdrawals
                         Patients reporting 1/> AEs or number                  NS
                         of SAEs
1             ETN +      ACR20, ACR50, ACR70, patients               1,2 and   Arm 1(
RCT320,       MTX vs.    achieving remission, modified TSS,          3 years   most:
321,328,332
              MTX        JSN, erosion score                                    p<0.01)
Level                    DAS score, HAQ, HAQ clinical                1 and 2   Arm 1 (all
1++                      improvement of ≥0.22, HAQ major             years     p<0.05):
                         improvement of >0.8
                         incidence of AEs, withdrawals due to        1 year    NS
                         AEs
                         patient satisfaction, EQ5D, Patient         2 years   Arm 1 (all
                         global assessment of overall RA                       p<0.01)
                         activity, swollen joints, painful joints,
                         pain (VAS), patients and physicians
                         global assessment, HAQ, CRP;
                         patients with no radiographic
                         progression and no progression of
                         erosions
                         patients reporting 1/> AEs or number                  NS
                         of SAEs
                         patients with low disease activity (DAS     3 years   Arm 1
                         <2.4 and DAS <3.2), HAQ                               (most:
                         improvement, patients with no                         p<0.01)
                         disability (HAQ score 0); total number
                         of withdrawals and withdrawals due to
                         lack of efficacy
                         incidence of AEs, patients reporting                  NS
                         1/> AEs or number of SAEs
1             ETN +      patients achieving remission                1,2 and   Arm 1
RCT320,       MTX vs                                                 3 years   (p<0.0001,
321,328,332
              etanerce                                                         p<0.01,
Level         pt                                                               p<0.05)



Rheumatoid arthritis: full guideline DRAFT (August 2008)                  174 of 307
    DRAFT FOR CONSULTATION


    1++                  ACR20, ACR50, ACR70, ACR-N AUC 1 and 2               Arm 1
                         and DAS score, HAQ score (lower         years        (most
                         score); patients achieving clinical HAQ              p<0.01)
                         improvement of ≥0.22, HAQ major
                         improvement of >0.8; modified TSS
                         and erosion score
                         JSN                                     1 year       Arm 1
                                                                              (p<0.05)
                         incidence of AEs, withdrawals due to                 NS
                         AEs
                         EQ5D, Patient global assessment of         2 years   Arm 1 (all
                         overall RA activity, swollen joints,                 p<0.05)
                         painful joints, pain (VAS), patients and
                         physicians global assessment, HAQ,
                         CRP, patients with no radiographic
                         progression and no progression of
                         erosions
                         patients reporting 1/> AEs or number       2 years   NS
                         of SAEs
                         ACR20, ACR50, ACR70, patients with         3 years   Arm 1
                         low disease activity (DAS <2.4 and                   (p<0.001;
                         DAS <3.2); HAQ improvement and                       p<0.01;
                         patients with no disability (HAQ score               p<0.05)
                         0); JSN, patients with no radiographic
                         progression and no progression of
                         erosions, withdrawals and withdrawals
                         due to lack of efficacy
                         Patient General Health Assessment                    Arm 2
                         patient satisfaction, patients reporting             NS
                         1/> AEs or number of SAEs
    1                    ESR                                        16        Arm 1
    RCT329                                                          weeks     (p=0.001)
    Level                patients achieving ACR20, ACR50 or                   NS
    1+                   ACR70; proportion of patients with an
                         improvement in DAS28 of >1.2 units,
                         AEs
                         Number of flares, proportion of                      Similar
                         patients who experienced a clinical
                         remission, proportion of patients who
                         experienced a ‘good’ or ‘moderate’
                         EULAR response
1
2
3   EXTENSION STUDIES: ETANERCEPT
4
5   •   1 RCT extension study333found that patients treated with had stable
6       ACR20, increased ACR50 and ACR70; decreased patient’s and
7       physician’s global assessment, DAS and HAQ scores, ESR and CRP
8       levels; improved patient pain scores and reduced number of painful and
9       swollen joints. Additionally, the two most common reasons for

    Rheumatoid arthritis: full guideline DRAFT (August 2008)             175 of 307
     DRAFT FOR CONSULTATION


1        discontinuation were AEs and unsatisfactory response. There were no
2        predominant AEs leading to discontinuation. Level 3
3    •   1 RCT extension331 found that for the outcomes of DAS remission (<1.6)
4        and DAS low disease activity (<2.4) at 1 year: MTX with ETN added and
5        ETN + MTX was better than ETN with MTX added. Additionally, ETN +
6        MTX was better than MTX with ETN. Level 1++
7
8    2. ABATACEPT
9
     Study       Interventio   Outcomes                            Follow-      Result - best
                 n                                                 up           treatment
     1 RCT330    ABA 2mg +     3 of the 8 components of SF-        1 year       Arm 1 (all
     Level 1+    MTX vs        36 (including physical                           p<0.05)
                 MTX           functioning and bodily pain),
                               SF-36 physical summary
                               scores

                               patients improving in all SF-36                  Arm 1 (better)
                               scales (SS for 2/11
                               comparisons)

                 ABA 10mg      SF-36 bodily pain, vitality and                  Arm 1 (p<0.0001;
                 + MTX vs      physical functioning                             p<0.05)
                 MTX           components; SF-6D, SF-36
                               physical and mental

                               All other SF-36 components;                      Arm 1 (better)
                               patients improving in all SF-36
                               scales (SS for 10/11
                               comparisons)
                               SF-36 physical functioning,                      Arm 1 (p<0.05
                               role physical, bodily pain,                      p<0.001)
                               vitality, social functioning, SF-
                               36 physical component; SF-6D
                               patients improving in all SF-36                  Arm 1 (better)
                               scales (SS for 7/11
                               comparisons).
     1 RCT326    ABA +         All SF-36 subscales and                          Arm 1
                 DMARD vs      composite scores, fatigue,
                 placebo +     patients ‘doing better’ (except
                 DMARD         role functioning and mental
                               component), rate of change for
                               all outcomes (except role
                               emotional), HAQ-DI, rate of
                               change for HAQ
10




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                176 of 307
    DRAFT FOR CONSULTATION



1
2   3. ADALIMUMAB
    Study            Intervention            Outcomes                                                         Follow-up   Result - best
                                                                                                                          treatment
    1 RCT323         ADA + standard anti-    ACR20, ACR50 and ACR70                                           24 weeks    Arm 1 (all
    Level 1++        rheumatic therapy vs                                                                                 p≤0.001)
                     placebo                    incidence of AEs, SAEs, severe or life threatening AEs (all               NS
                                                did not vary according to ~ of DMARDs used); withdrawals
                                                                                                due to AEs

                     ADA + 1 or 2            ACR20 responses                                                              Arm 1 (p≤0.001)
                     DMARDS vs placebo

                     ADA + 0, 1 or 2         ACR50 and ACR70 responses                                                    Arm 1 (p≤0.001)
                     DMARDS vs placebo

    1 RCT322         ADA 20 mg + MTX vs      ACR20, ACR50, ACR70, tender and swollen joints, Patient’s        1 year      Arm 1 (all
    Level 1++        placebo + MTX           assessment of pain (VAS), patient and physicians’ global                     p≤0.001)
                                             assessment of disease activity, SF-36 (all domains), HAQ,
                                             radiographic progression – TSS and joint erosion score,
                                             CRP

                                             Total number of withdrawals,; Withdrawals due to lack of                     Arm 1 (better)
                                             efficacy and due to AEs

                                             JSN; patients reporting at least 1 AE and for rate of AEs                    NS; similar




    Rheumatoid arthritis: full guideline DRAFT (August 2008)          177 of 307
DRAFT FOR CONSULTATION



                 ADA 40 mg + MTX vs      ACR20, ACR50, ACR70, tender and swollen joints, patient’s               Arm 1 (all
                 placebo + MTX           assessment of pain (VAS), SF-36 (all domains except                     p≤0.001)
                                         emotional role), HAQ, radiographic progression (TSS, joint
                                         erosion score and JSN score), patient’s and physician’s
                                         global assessment of disease activity, CRP

                                         total number of withdrawals, withdrawals due to AEs                     Arm 1 (better)

                                         patients reporting at least 1 AE, rate of AEs, withdrawals              Similar
                                         due to lack of efficacy

1 RCT319         ADA 20 mg + MTX vs      ACR20, ACR50; Tender joints; swollen joints, patient’s       24 weeks   Arm 1 (p=0.003;
Level 1++        placebo + MTX           assessment of pain (VAS), patient’s and physician’s global              p=0.002; p≤0.001;
                                         assessment of disease activity, CRP; HAQ; better for SF-36              p=0.004)
                                         scores;

                                         ACR70 and Fatigue; withdrawals due to AEs                               NS; similar

                 ADA 40 mg + MTX vs      ACR20, ACR50, ACR70, tender and joints, patient’s                       Arm 1 (all
                 placebo + MTX           assessment of pain (VAS), Fatigue (FACIT), SF-36 (better),              p≤0.001)
                                         HAQ, patient’s and physician’s global assessment of
                                         disease activity, CRP

                                         Withdrawals due to AEs                                                  Similar

                 ADA 80 mg + MTX vs      ACR20, ACR50, ACR70 (p=0.02), tender and swollen joints,                Arm 1 (all
                 placebo + MTX           patient’s assessment of pain (VAS), Fatigue (FACIT), HAQ,               p≤0.001)
                                         patient’s and physician’s global assessment of disease
                                         activity, CRP; SF-36 score (better)




Rheumatoid arthritis: full guideline DRAFT (August 2008)          178 of 307
    DRAFT FOR CONSULTATION



                                             Withdrawals due to AEs                Similar

1
2
3




    Rheumatoid arthritis: full guideline DRAFT (August 2008)          179 of 307
    DRAFT FOR CONSULTATION

1   4. RITUXIMAB
2
    Study      Intervention   Outcomes                              Follow-  Result -
                                                                    up       best
                                                                             treatment
    1 RCT327   RTX vs         ACR20,                                24 weeks RTX (p<0.01
    Level      placebo +                                                     or not given)
    1+         MTX
                              % of patients with HAQ-DI                       RTX (better)
                              reductions ≥ 0.25

                              SAEs or withdrawals due to AEs                  NS

               RTX + CTX      ACR20 and ACR50                       24 and   Arm 1
               vs placebo +                                         48 weeks (p<0.01 and
               MTX                                                           p<0.05)
                              EULAR response and % of patients               Arm 1
                              with HAQ-DI reductions ≥ 0.25                  (better)

                              SAEs or withdrawals due to AEs                  NS

               RTX + MTX      ACR20, ACR50 and ACR70                          Arm 1
               vs placebo +                                                   (p<0.01 and
               MTX                                                            p<0.05)
                              EULAR response and % of patients                Arm 1
                              with HAQ-DI reductions ≥ 0.25                   (better)

                              SAEs or withdrawals due to AEs                  NS

3
4   5. INFLIXIMAB
5
    Study      Intervention   Outcomes                              Follow-  Result -
                                                                    up       best
                                                                             treatment
    1 RCT325   IFX 3mg/kg + ACR20, ACR50, ACR70, DAS28              22 weeks Arm 1 (all
    Level      MTX vs MTX score, proportion of patients in                   p<0.0001)
    1+         + placebo    remission

                              AEs                                             NS

                              When IFX dose increased: numbers                Similar
                              and types of SAEs, AEs and rates of
                              AEs

               IFX 10mg/kg    ACR20, ACR50, ACR70, DAS28                      Arm 1 (all
               + MTX vs       score, proportion of patients in                p<0.0001)
               MTX +          remission
               placebo
                              AEs                                             NS



    Rheumatoid arthritis: full guideline DRAFT (August 2008)         180 of 307
    DRAFT FOR CONSULTATION


                              When IFX dose increased: numbers               Similar
                              and types of SAEs, AEs and rates of
                              AEs

1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)        181 of 307
    DRAFT FOR CONSULTATION




1   7.4.12       Health economic evidence statements
2
    Drug            Comparator         Study                   Date   Time       ICER
                                                                      horizon
                    DMARD              Bansback                2005   Lifetime   ACR50/DAS28 good:
                    sequence                                                     €34,167 per QALY (MTX)
                                                                                 €34,922 per QALY (MTX) (from pooled
                                                                                 analysis)
                                                                                 €41,561 per QALY (monotherapy)
                                                                                 ACR20/DAS28 moderate:
                                                                                 €40,875 per QALY (+ MTX)
                                                                                 €44,018 per QALY (+ MTX) (from pooled
    Adalimumab                                                                   analysis)
                                                                                 €65,499 per QALY (monotherapy)
                    Anakinra           Chiou                   2004   1 year     Adalimumab alone dominated
                                                                                 Adalimumab + MTX dominated
                    DMARDs             Chen284                 2006   Lifetime   Adalimumab (no MTX)            £140,000
                                                                                 per QALY
                                                                                 Adalimumab (with MTX)          £64,000
                                                                                        per QALY
                                                                                 (Third line (late RA data))
                    Anakinra           Chiou                   2004   1 year     US$13,387 per QALY (monotherapy)
    Etanercept
                                                                                 US$7,925 per QALY (+ MTX)
                    DMARD              Brennan                 2004   Lifetime   £16,330 per QALY
                    sequence




    Rheumatoid arthritis: full guideline DRAFT (August 2008)            182 of 307
DRAFT FOR CONSULTATION



Drug            Comparator           Study                 Date   Time       ICER
                                                                  horizon
                DMARD                Bansback              2005   Lifetime   ACR50/DAS28 good:
                sequence                                                     €35,760 per QALY(+ MTX)
                                                                             €36,927 per QALY (monotherapy)
                                                                             ACR20/DAS28 moderate:
                                                                             €51,976 per QALY (+ MTX)
                                                                             €42,480 per QALY (monotherapy)
                Baseline level       Kobelt                2004   Not        After 3 months treatment:€43,500 per QALY
                (failed at least 2                                applicable After 6 weeks treatment: €36,900 per QALY
                DMARDs,
                including
                methotrexate)
                Methotrexate         Kobelt                2005   10 years   Etanercept alone dominated. Treatment for 2
                                                                             years, extrapolation to 10 years:
                                                                             Etan-MTX €37,331 per QALY.
                                                                             Treatment for 2 years, extrapolation to 5
                                                                             years: Etan-MTX €54,548 per QALY.
                                                                             Treatment for 10 years: Etan-MTX €46494 per
                                                                             QALY.
                                                                             Treatment for 5 years, extrapolation to 10
                                                                             years. Etan-MTX €47,316 per QALY




Rheumatoid arthritis: full guideline DRAFT (August 2008)           183 of 307
DRAFT FOR CONSULTATION



Drug            Comparator          Study                  Date   Time       ICER
                                                                  horizon
                Usual treatment,    Welsing                2004   5 years    Etanercept monotherapy dominated by
                leflunomide,                                                 leflunomide/etanercept combinations
                                                                             Etanercept vs usual treatment:
                                                                             €163,556 per QALY for Lef-Etan
                                                                             €297,151 per QALY for Etan-Lef
                                                                             Etanercept vs leflunomide
                                                                             €317,627 per QALY for Lef-Etan
                                                                             €517,061 per QALY for Etan-Lef
                Monotherapy         Choi                   2002   6 months   Etanercept – sulfasalazine: $41,900 per
                leflunomide,                                                 ACR20
                methotrexate,                                                Etanercept - methotrexate: $40,800 per
                sulfasalazine, no                                            ACR70WR
                second line
                agent
                Standard            Tanno                  2006   Lifetime   Etanercept
                therapy for                                                  Y2.50 million per QALY
                Japanese RA
                patients
                (methotrexate or
                sulfasalazine or
                methotrexate +
                sulfasalazine)
                DMARDs              Chen284                2006   Lifetime   Etanercept (no MTX)              £47,000
                                                                                    per QALY
                                                                             Etanercept (with MTX)            £50,000
                                                                                    per QALY
                                                                             (Third line (late RA data))



Rheumatoid arthritis: full guideline DRAFT (August 2008)            184 of 307
DRAFT FOR CONSULTATION



Drug            Comparator           Study                 Date   Time       ICER
                                                                  horizon
                Placebo and          Wong                  2002   Lifetime   $30,500 per QALY
                methotrexate
                Methotrexate         Kobelt                2003   10 years   For 1 year of treatment:
                                                                             €3,440 per QALY in Sweden
                                                                             €34,800 per QALY in UK
                Baseline level       Kobelt                2004   Not        After 3 months treatment €43,500 per QALY
                (failed at least 2                                applicable After 6 weeks treatment: €36,900 per QALY
                DMARDs,
                including
                methotrexate)
                DMARD                Bansback              2005   Lifetime   ACR50/DAS28 good:
                sequence                                                     €48,333 per QALY (+ MTX)
Infliximab
                                                                             ACR20/DAS28 moderate:
                                                                             €64,935 per QALY (+ MTX)
                Anakinra             Chiou                 2004   1 year     Infliximab + MTX dominated
                Methotrexate         Marra                 2007   10 years   Incremental cost per QALY depending on
                                                                             utilities used:
                                                                             HUI2-QALY $53,429
                                                                             HUI3-QALY $32,018
                                                                             SF-6D-QALY $69,826
                                                                             EQ-5D-QALY $46,322
                DMARDs               Chen284               2006   Lifetime   Infliximab (with MTX)            £140,000
                                                                             per QALY
                                                                             (Third line (late RA data))
Adalimumab/ DMARDs                   Brennan               2007   Lifetime   £23,882 per QALY (prob cost effective = 0.84)
etanercept/                                                                  (£32,013 at current discount rates, prob cost
infliximab                                                                   effective = 0.36)



Rheumatoid arthritis: full guideline DRAFT (August 2008)            185 of 307
     DRAFT FOR CONSULTATION


 1   Of the reviewed studies, only four284,338,339,345 are UK based. Of these four,
 2   Adalimumab reported an ICER greater than the NICE threshold284 . For
 3   etanercept, Brennan et al 2004338 reported an ICER below the threshold, but
 4   Chen et al284reported an ICER above the threshold. For infliximab, the Kobelt
 5   et al(5) ICER was under the threshold but the Chen et al ICER was above. As
 6   a group of anti-TNF drugs, Brennan et al 2007345 reported an ICER between
 7   £20-£30k. This study directly models from the British Society for
 8   Rheumatology Biologics Registry, but using the updated NICE discount rates
 9   the ICER would exceed £30k. A direct comparison of ICERs between the
10   studies is not possible as there is variations in model structure, assumptions,
11   comparator drugs, time horizon, source of preference weights and cycle
12   length.

13   7.4.13         Summary of evidence statements
14        •   In patients with active disease despite conventional disease
15            modifying drugs, the addition of a biological drug generally adds
16            significant benefits for symptom control, function and quality of life 324
              323,330 322 319 320 328 321 327 326,333
17                                                    .
18        •   In those studies addressing radiological damage, the combination of
19            biological drug and methotrexate with methotrexate alone favour the
20            combination320,322 328 .
21        •   The combination of anti-TNF with methotrexate was superior to anti-
22            TNF drug alone for symptomatic benefit, and in studies that
23            measured them, functional outcomes, quality of life and joint damage
              319,322 320 328 321,326
24
25        •   The only studies to compare biological therapy directly with
26            conventional disease modifying drug suggest that etanercept is
27            superior to sulphasalazine324, and rituximab is superior to
28            methotrexate327, for symptom control and functional benefit.
29        •   Comparisons of etanercept with methotrexate showed few
30            differences in symptom control or functional benefit, though
31            etanercept was superior to methotrexate for decreased radiological
32            progression 320 328

33   7.4.14          From evidence to recommendations
34   The available data does not answer the clinical question of whether a patient
35   who is not responding to DMARD therapy should go onto other conventional
36   DMARDs or onto a biological drug. There are no head to head trials of these
37   comparators.
38
39   The only studies to compare biological directly with conventional disease
40   modifying drug suggest that etanercept is superior to sulphasalazine, and
41   rituximab is superior to methotrexate, for symptom control and functional
42   benefit. However, a comparison of etanercept with methotrexate showed few
43   differences in symptom control or functional benefit, though etanercept was
44   superior to methotrexate for decreased radiological progression. This


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                186 of 307
     DRAFT FOR CONSULTATION


 1   suggests a disconnection between the ability of drugs on the one hand to
 2   influence disease activity, and on the other to slow radiological progression.
 3
 4   All the other studies show that in patients with active disease despite
 5   conventional disease modifying drugs, the addition of a biological drug
 6   generally adds significant benefits for symptom control, function and quality of
 7   life. This is in comparison with continuing on methotrexate, the response to
 8   which had not been ideal, because the patient had sufficiently active disease
 9   on the methotrexate to be included in the trial. In those studies addressing
10   radiological damage, the combination of biological drug and methotrexate with
11   methotrexate alone favour the combination. The combination of etanercept
12   and methotrexate was superior to either drug alone for symptomatic and
13   functional benefit, quality of life and joint damage. This is now reflected in
14   licence changes for etanercept and adalimumab to suggest that in the
15   absence of contraindications, methotrexate should be co-prescribed.

16   7.4.15        Recommendations
17   R 31 Biologic drugs should be offered in the treatment of people with
18   rheumatoid arthritis where disease control is not satisfactory with conventional
19   DMARD therapy, and such drugs should be used in accordance with NICE
20   Technology Appraisal Guidance 130 (October 2007).
21

22   7.5    Disease modifying and biological drugs: when to
23   withdraw them

24   7.5.1 Clinical introduction
25   There are a variety of reasons for stopping or reducing disease modifying
26   drugs. Some drugs may need to be stopped prior to a patient or their partner
27   conceiving (e.g. methotrexate and leflunomide) and during pregnancy and
28   breast-feeding. The drugs may be stopped or reduced due to side-effects, or
29   during inter-current illnesses. They may be stopped due to primary or
30   secondary loss of efficacy.
31
32   A further category of patients pose questions about the most appropriate
33   approach. These are patients with established RA with minimal or no disease
34   activity. The observation that their disease is controlled may be due to the
35   natural history of the disease, or due to the disease modifying drugs that they
36   are taking. If the former is suspected, it would be good clinical practice to
37   attempt to withdraw the medication. If the latter is the case, then such
38   attempts might result in an increase in disease activity. Furthermore, there is
39   some data to suggest that even in patients in remission, MRI scans can still
40   show disease activity and progression of damage, which might be intensified if
41   the disease modifying drug was withdrawn. However, the significance of this
42   observation is currently unknown.
43



     Rheumatoid arthritis: full guideline DRAFT (August 2008)             187 of 307
     DRAFT FOR CONSULTATION


1    Once established, it has been argued that RA never completely disappears.
2    Even if the data suggests complete withdrawal of disease modifying drugs is
3    not appropriate for most patients, is there any evidence to support the
4    common clinical approach of keeping patients on the smallest dose of drugs
5    that appears to keep the disease under satisfactory control?

6    7.5.2 Clinical methodological introduction
 7   We looked for studies that investigated the effect of withdrawing or titrating the
 8   dose of DMARDs or biologics with respect to symptoms, joint damage,
 9   function and quality of life in patients with established RA. Due to the paucity
10   of trials in this area, all study types were included as evidence.
11
12   Ten studies346-355 were found that fulfilled the criteria. The 10 studies
13   consisted of 7 RCTs346-351,355 and 3 case-series’ 352-354. One RCT355 was
14   excluded due to methodological limitations. No studies were found that
15   evaluated biologics.
16
17   The 6 included RCTs assessed RA patients who had already been treated
18   with DMARDs and were randomised to either continue on the treatment, have
19   treatment withdrawn (be given placebo) or have the dose of DMARD reduced.
20   Patients were then followed up for a period of time and outcomes were then
21   assessed. The RCTs differed with respect to the following:
22        • Sample size (range: N=10 to N=285)
23        •   Blinding (5 RCTs double blind, 1 RCT single blind)
24        •   Trial length (range: 6 months to 2 years)
25        •   Treatment (1 RCT 2nd line DMARDs vs placebo, 1 RCT MTX,
26            penicillamine or gold DMARDs vs placebo, 1 RCT MTX vs placebo, 1
27            RCT IM gold vs placebo, 1 RCT azathioprine vs placebo, 1 RCT D-
28            penicillamine at same dose vs D-penicillamine dose titrated)
29        •   Treatment regimen – dose and withdrawal / titration regimen
30   Although the RCTs were fairly sound methodologically, it is worth noting that
31   most of them were small trials (sample size <40) and did not perform ITT
32   analysis. Different dosing and titration regimens and the differing populations
33   in the studies may also limit direct comparisons between studies.
34
35   The 3 case-series’ assessed the effects of dose titration on patients already
36   receiving DMARDs or corticosteroids. The first case-series (Fleischmann et
37   al.)354 assessed patients already receiving MTX, treated them with infliximab
38   and if clinical improvement was seen at 22 weeks, the dose of MTX was then
39   tapered and patients were followed up and outcomes assessed at 1 year. The
40   second case-series (Tishler et al.)353 assessed patients already receiving
41   MTX with stable disease, the regime of MTX was then reduced from
42   once/week to once/fortnight. Patients were then followed up and outcomes
43   assessed at 1 year. The third case-series (Bacon et al.)352 assessed patients
44   already receiving corticosteroids (prednisolone) with stable disease. The dose
45   of corticosteroids was tapered and patients were then followed up and
46   outcomes assessed at 1 year.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              188 of 307
     DRAFT FOR CONSULTATION


1    7.5.3 Health economic methodological introduction
2    No health economic papers were identified for this question

3    7.5.4 Clinical evidence statements
 4
 5   Established RA
 6
 7   Symptoms / QoL (withdrawal vs continue treatment)
 8
 9        •   1 case-series 353 found that when the frequency of MTX treatment
10            was reduced in patients with clinical remission, there was no
11            deterioration in morning stiffness and Ritchie Articular Index. Level 3
12        •   1 case-series354 found that in patients who had infliximab added to
13            their current MTX treatment and who showed ≥40% clinical
14            improvement in arthritis, when MTX was tapered, significant
15            improvements from baseline were seen for tender and swollen joints.
16            Level 3
     Study (All Level    Outcomes                                Follow-up          Result -
     1+)                                                                            best
                                                                                    treatment
     1 RCT (deSilva      Pain; morning stiffness                 16, 24 and 32      Continue
     et al) 347                                                  weeks              (all p<0.05)

     1 RCT               swollen joints                          6 months           Continue
     (Gotzsche et                                                                   (p=0.03)
     al.)348             tender joints                           6 months           NS

     1 RCT (Kremer       tender and swollen joints               1 month            Continue
     et al) 349                                                                     (all p<0.05)
                         morning stiffness and evening           1 month            NS
                         fatigue
     1 RCT (Ten          Pain at rest (p<0.05), morning          1 year             Continue
     Wolde et al)350     stiffness and Ritchie articular index                      (p=0.031,
                                                                                    p=0.005,
                                                                                    p=0.000)


     1 RCT (Van der      RAI and number of swollen joints        Over the 3         Early
     Leeden et al) 351                                           years              (p<0.05)

17
18   Function (withdrawal vs continue treatment)
19
20        •   1 case-series353 found that when the frequency of MTX treatment
21            was reduced in patients with clinical remission, there was no
22            deterioration in grip strength. Level 3
23        •   1 case-series354 found that in patients who and infliximab added to
24            their current MTX treatment and who showed ≥40% clinical

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               189 of 307
     DRAFT FOR CONSULTATION


1             improvement in arthritis, when MTX was tapered, significant
2             improvements from baseline were seen for HAQ score. Level 3
     Study (All Level Outcomes                                    Follow-up       Result -
     1+)                                                                          best
                                                                                  treatment
     1 RCT (Kremer       grip strength                            1 month         NS
     et al) 349

     1 RCT (Ten          grip strength (right and left hands)     1 year          Continue
     Wolde et al)350     and HAQ score                                            (p<0.05,
                                                                                  p=0.014)


3
4    Joint damage (withdrawal vs continue treatment)
5
6         •   1 RCT351 found that there was NS difference between the group
7             withdrawn from gold treatment compared with the group continuing
8             on gold treatment for radiological score. 1+
 9
10   Global assessment (withdrawal vs continue treatment)
11
12        •   1 case-series353 found that when the frequency of MTX treatment
13            was reduced in patients with clinical remission, there was no
14            deterioration in doctor’s and patient’s global assessment of pain and
15            disease activity. 3
     Study (All Level     Outcomes                                   Follow-up           Result -
     1+)                                                                                 best
                                                                                         treatment
     1 RCT (deSilva et    Patient’s and clinician’s general          32 weeks            Continue (all
     al) 347              evaluation of response to therapy                              p<0.05)

     1 RCT (Gotzsche      Patient’s perception of well-being         6 months            Continue (p=
     et al.)348                                                                          p=0.002 )

                          Patient’s evaluation of the number of      6 months            NS
                          painful joints
     1 RCT (Kremer et     Physician’s and patient’s global           1 month             Continue (all
     al) 349              evaluation of pain and disease                                 p<0.05)
     1+                   activity
16
17   Remission (withdrawal vs continue treatment)
18
19        •   1 RCT350 found that risk of flare was associated with high
20            maintenance of second-line drugs, presence of painless swollen
21            joints and ever positive rheumatoid factor. Level 1+
22        •   1 case-series352 found that in 26% of patients who were on long-term
23            treatment with corticosteroids and whose arthritis appeared to be in


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               190 of 307
     DRAFT FOR CONSULTATION


1              remission, withdrawal of corticosteroid treatment was successful (no
2              reactivation of arthritis), however 61% had to have their corticosteroid
3              treatment reintroduced due to the return of active arthritis. Level 3
4          •   1 case-series353 found that in patients who were on treatment with
5              MTX and whose arthritis appeared to be in remission, reducing the
6              frequency of treatment was successful (arthritis did not deteriorate) in
7              87% of patients, however 13% had a flare of arthritis. Level 3
     Study (All Level      Outcomes                               Follow-up          Result -
     1+)                                                                             best
                                                                                     treatment
     1 RCT (Ahern et       Maintenance of remission               12 months          Continue
     al) 346                                                                         (better)

     1 RCT                 Number of patients experiencing        6 months           Continue
     (Gotzsche et          treatment failure                                         (60% vs
     al.)348                                                                         15.8%,
                                                                                     p=0.000001
                                                                                     )
     1 RCT (Kremer         Number of patients experiencing        1 month            Continue
     et al) 349            significant flare                                         (better)

 8
 9   Biochemical markers (withdrawal vs continue treatment)
10
11         •   1 case-series353 found that when the frequency of MTX treatment
12             was reduced in patients with clinical remission, there was no
13             deterioration in ECR or CRP levels. Level 3
14         •   1 case-series354 found that in patients who had infliximab added to
15             their current treatment MTX treatment and who showed ≥40% clinical
16             improvement in arthritis, when MTX was tapered, significant
17             improvements from baseline were seen for ESR and CRP levels.
18             Level 3
19
     Study (All           Outcomes             Follow-up                  Result - best
     Level 1+)                                                            treatment
     3 RCTs               ESR                  32 weeks (deSilva)         NS
     (deSilva,
     Kremer and                                1 month (Kremer)
     Van der
     Leeden)347,349,351                        over 3 years (Van der
                                               Leeden)

     1 RCT (Ahern         mean CRP level       1 month and 3 months       Continue
     et al) 346                                after clinical relapse     (p<0.05)




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                191 of 307
     DRAFT FOR CONSULTATION


     1 RCT (Ten          CRP and ESR          1 year                   Continue
     Wolde et al)350                                                   (p=0.008,
                                                                       p=0.000)


1
2    AEs (withdrawal vs continue treatment)
3
4         •   1 case-series354 found that in patients who had infliximab added to
5             their current MTX treatment and who showed ≥40% clinical
6             improvement in arthritis, when MTX was tapered, there was an 80%
7             incidence of AEs. Specific AEs were not specified but included
8             infection and infusion reactions. Level 3
9
     Study (All Level Outcomes                               Follow-up          Result -
     1+)                                                                        best
                                                                                treatment
     1 RCT               severity of reported side-effects   6 months           NS
     (Gotzsche et
     al.)348

     1 RCT (Ten          AEs                                 1 year             Similar
     Wolde et al)350

10
11   Withdrawals (withdrawal vs continue treatment)
12
     Study (All Level Outcomes                               Follow-up          Result -
     1+)                                                                        best
                                                                                treatment
     1 RCT (deSilva      Total number of withdrawals;        32 weeks           Continue
     et al) 347          withdrawals due to clinical                            (better)
                         deterioration

     1 RCT (Kremer       Number of withdrawals (both N=0)    1 month            NS
     et al) 349

     1 RCT (Van der      Number of withdrawals               Over the 3         Similar
     Leeden et al) 351                                       years              (N=5 and
                                                                                N=4)
13
14   Effect of reintroduction of withdrawn DMARDs
15
16        •   1 RCT346 found that when D-penicillamine was reintroduced at the
17            former dose in patients in the tapered dose group who experienced
18            flare, 87% achieved clinical remission again within 4 months. The
19            remaining 13% achieved remission when given a higher dose. Level
20            1+
21        •   1 RCT349 found that when MTX was reintroduced in patients in the
22            withdrawal group who experienced flare, all achieved improvement

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             192 of 307
     DRAFT FOR CONSULTATION


1             again in pain, morning stiffness and patient’s assessment of global
2             disease activity. Level 1+
3         •   1 case-series353 found that when MTX treatment was reintroduced at
4             the usual frequency in patients in the reduced dose frequency group
5             who had experienced flare, all achieved control of disease activity.
6             Level 3

7    7.5.5 Summary of evidence statements
 8        •   The studies used a variety of methodologies with different DMARDs.
 9            In some studies the patients had excellent disease control prior to
10            withdrawal of active DMARD346 352 347 350,353, and in others had
11            ongoing active disease354 348 349,351.
12        •   In some of the trials the active DMARD was tapered down352 353,354,
13            but in most was suddenly replaced with placebo346-351.
14        •   Which ever methodology was used, the majority of studies showed
15            the patients on placebo or lower doses of DMARD did not do as well
16            symptomatically, functionally or in quality of life347 348 349 350.346,352
17            There was insufficient data to address any impact on joint damage.
18        •   One study in which methotrexate could be reduced successfully was
19            in patients responding well to infliximab. 354. One study also
20            suggested that restoring the patient back to the original dose
21            controlling drug could result in restoration of disease improvement 346.

22   7.5.6 From evidence to recommendations
23   Based on the available evidence, the GDG felt that when a decision was
24   made to try and reduce dosages of disease modifying or biologic drugs, in
25   view of the possibility of a flare-up of disease it would be prudent to advise
26   that this should always be done with caution, and that arrangements were in
27   place for an urgent reassessment so that there could be a prompt return to
28   disease-controlling dosages of medication at the first sign of any flare-up.
29   Although the need for sufficient patient education to know when and how to
30   seek rapid access and help in the event of a flare-up in the context of well
31   controlled established disease is already covered elsewhere (see
32   recommendation R37), the GDG nevertheless felt that, for those specific
33   patients in whom a decision had been taken to decrease (or discontinue) their
34   disease modifying or biologic drugs, an extra recommendation about the need
35   for a prompt review being available (e.g. by rapid access to the named
36   member of the MDT, see section 6.1) was indicated.
37
38   It was also felt appropriate to extrapolate from the available evidence a more
39   general recommendation that the opportunity to try and decrease or
40   discontinue the dosages of current medication should always be explored
41   when additional drugs are being added to a treatment regimen.

42   7.5.7 Recommendations
43

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                193 of 307
     DRAFT FOR CONSULTATION


 1   R 32 In people with established rheumatoid arthritis whose disease is judged
 2   to be stable, reduction in dosage of disease modifying or biologic drugs
 3   should be performed with caution, and there should be a prompt return to
 4   disease-controlling dosages at the first sign of any flare-up of disease.
 5
 6   R 33 In people with established rheumatoid arthritis where new drugs are
 7   being introduced into their treatment regimen to improve disease control, the
 8   opportunity to decrease or discontinue their current drugs should be explored
 9   when disease control is established.
10
11   R 34 In any person with established rheumatoid arthritis in whom disease
12   modifying or biological drug doses are being decreased or discontinued,
13   arrangements should be in place for prompt review.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)           194 of 307
     DRAFT FOR CONSULTATION



1    8         Monitoring the course of RA

2    8.1       Monitoring disease

3    8.1.1 Clinical introduction
 4   Monitoring of disease activity in RA has traditionally been performed
 5   subjectively, and based on information about the inflammation shared
 6   between the patient and health care professional, such as the symptoms
 7   (pain, swelling, duration of morning stiffness, fever, weight loss), and the signs
 8   (joint swelling, heat and tenderness). This can be made more objective by
 9   laboratory tests of the inflammatory activity such as the C-reactive protein
10   (CRP), erythrocyte sedimentation rate (ESR). It has been acknowledged that
11   any individual measure of disease activity has limitations. For example, pain is
12   exclusively perceived by the patient, and this perception is influenced by a
13   host of factors beyond disease activity. Acute phase markers may be elevated
14   due to other intercurrent disease such as infection. No single measure of
15   disease activity satisfactorily encapsulates the complexity of the concept, or is
16   free from confounding so as to just measure RA disease activity.
17   Consequently amalgamations of single measures have been validated for use
18   in RA. Many of these originated from trials, but have increasingly been used in
19   clinical practice to more objectively inform decision making.
20
21   Because of the fluctuating and chronic nature of RA, there is limited
22   information from measuring disease activity at a single moment in time, and
23   evidence to show that the “area under the curve‡‡” of ongoing inflammation is
24   closely related to accumulating damage and disability. Furthermore, there is
25   no point in measuring disease activity if the results are ignored. High disease
26   activity suggests disease control is inadequate and demands an appropriate
27   response. Sustained low disease activity may enable a cautious reduction in
28   medication (see elsewhere in the guideline). In this section the best methods
29   of assessing ongoing disease activity are considered, and the appropriate
30   response to the information produced by these methods.

31   8.1.2 Clinical methodological introduction
32   We looked for studies that assessed what are the most effective methods of
33   measuring the ongoing disease activity of patients with RA (established
34   disease and recent onset of disease). Due to the large volume of evidence,
35   trials were selected that were of a UK relevant population, sample size N>70.
36   For papers assessing treatment adjusted by disease activity scores, only
37   those looking at composite scores were included and for papers assessing
38   disease activity measures, measurements had to have been taken over time
39   (ie. >2 time point assessments). Three RCTs288,356,357, one pooled analysis358
40   and 4 case-series’359-362 were found that fulfilled the inclusion criteria. All trials
41   were methodologically sound.

     ‡‡
          i.e. cumulative summation of measurements over time

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  195 of 307
     DRAFT FOR CONSULTATION

 1   Recent-onset RA
 2   Two RCTs288,357 and one case-series363 were found. The RCTs were both
 3   single blind, parallel group studies looking at adjusting treatment based on
 4   monitoring disease activity. The first RCT288 was performed in N=111 patients
 5   who were randomised to 2 different treatment arms: intensive monitoring
 6   (every month) by disease activity versus routine monitoring (every 3 months)
 7   over 18 months. The second RCT357 was performed in N=299 patients who
 8   were randomised to 2 different treatment arms: intensive monitoring (every
 9   month) by 20% response versus conventional monitoring (every 3 months)
10   over 2 years. The case-series357 looked at monitoring disease activity
11   measures over at least 3 years in N=110 patients.
12
13   Established RA
14   One cluster RCT356, 1 pooled analysis of 3 RCTs358 and 4 case-series’359-362
15   were found. The cluster RCT356 was a single blind, parallel group trial looking
16   at adjusting treatment based on monitoring disease activity in N=205 patients
17   who were randomised to 2 different treatment arms: systematic monitoring by
18   DAS28 versus Usual care (no monitoring/adjustment) over 24 weeks. The
19   pooled analysis358 looked at the data from 3 RCTs of leflunomide and
20   monitored disease activity measures over 1 year in N=1839 patients.
21
22   The 4 case-series all looked at monitoring disease activity measures over time
23   and differed with respect to: Sample size (range N=71 to N=233) and Length
24   of follow-up (range: 24 weeks to 30 months).

25   8.1.3 Health economic methodological introduction
26   No health economic papers were identified

27   8.1.4 Clinical evidence statements
28   Recent-onset RA
29
30   a) Monitoring: Disease activity measures
31        • One case-series363 found that time integrated CRP was significantly
32           correlated with radiological progression over 6 months, 1 year, 2
33           years and 3 years. Level 3
34
35   b) Treatment monitoring: adjustment by disease activity
36        •
37   Intensive strategy (adjusted by disease activity measure response) vs
38   routine strategy (adjusted by rheumatologist’s criteria)
39
     Study       Outcomes                                         Follow-up Result -
                                                                            best
                                                                            treatment




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             196 of 307
    DRAFT FOR CONSULTATION


    One          EULAR good response and remission              18 months    Intensive
    RCT288       ACR20, ACR50 and ACR70; Disease activity                    (p<0.0001
    Level        score; Joint swelling and tenderness                        unless
    1++          (p<0.01); Patient’s and assessor’s global                   stated)
                 assessment of disease activity; Pain (VAS);
                 HAQ (p=0.0025); ESR (p=0.0007); SF-12
                 physical domain (p=0.021); erosion score
                 (p=0.002) and Total sharp score (p=0.02).

                 Number of AEs, Higher prescription of IM and   18 months    Intensive
                 IA Corticosteroids and combination DMARDs                   (better)
                 and higher doses of MTX

                 CRP; SF-12 mental domain; JSN and doses        18 months    NS
                 of SSZ

    One          Number of patients reaching remission for 3    1 and 2      Intensive
    RCT357       months                                         years        (p<0.001,
    Level 1+                                                                 p=0.029)
                 Mean time until first period of remission;      1 year      Intensive
                 Duration of all periods of remission (p=0.025);             (p≤0.001
                 Median AUC (for morning stiffness (p=0.009),                unless
                 ESR (p=0.007), tender and swollen joint                     stated)
                 counts, VAS general well-being and pain
                 (VAS); modified ACR50 (p=0.018)
                 Use of NSAIDs                                   6 months    Intensive
                                                                 and 2
                                                                 years

                 Number of patients with AEs; number of AEs     2 years      Intensive
                                                                             (better)

                 Median AUC for Functional disability;          2 years      NS
                 Modified ACR50; Radiographic progression
                 and Number of IA CS
1
2   Established RA
3
4   a) Monitoring: Disease activity measures
5
    Study        Outcomes                                        Follow-up    Result -
                                                                              best
                                                                              measure
    1 pooled     SDAI change and HAQ change                      All times    Correlation
    analysis35                                                   up to 1      (p<0.0001)
    8
                                                                 year
    Level 1+




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           197 of 307
     DRAFT FOR CONSULTATION


                 SDAI and DAS28                                     All time-   Associatio
                                                                    points up   n
                 SDAI change and HAQ change                         to 6        (p<0.0001)
                                                                    months

     1 case-     RAI, summated change score;                        24 weeks    Best
     series359                                                                  clinical
     Level 3                                                                    measures
                 ESR, plasma viscosity                              24 weeks    Best
                                                                                laboratory
                                                                                measures
                 Grip strength and joint size                       24 weeks    Not good

     1 case-     CRP and Articular index (compared to ESR           24 weeks    Best
     series360   and Articular Index)                                           correlation
     Level 3
     1 case-     Change in KFI with therapy and: change in                      Correlation
     series361   RAI (p=0.001), morning stiffness (p=0.004),
     Level 3     swollen joint count (p=0.0005), CRP (p=0.03)
                 and LSI (p=0.002),
                 Change in KFI with therapy and: ESR or                         Not
                 change in time to onset of fatigue.                            correlate

                 Change in HFI and: change in RAI, morning                      Correlation
                 stiffness, swollen joint count, CRP and LSI                    (but not as
                                                                                good as
                                                                                KFI)

     1 case-     Clinical status (physical disability measured by               Best
     series362   rheumatologists) and: Disease activity score                   correlation
     Level 3     followed by RAI and the Mallya index                           s

                 Discriminate between high and low disease          2 years     Best
                 activity (based on use of DMARDs): Disease                     discriminat
                 activity score, followed by Riel index and the                 ors
                 Mallya Index
                 Increase in joint damage (erosions, JSN and        2 years     Best
                 total score) and: CRP, swollen joints, ESR,                    Correlation
                 disease activity score, Mallya index, Riel Index
                 and Grip strength.
 1
 2   b) Treatment monitoring: adjustment by disease activity
 3        • One RCT356 found that systematic monitoring + treatment adjustment
 4           was significantly better (all p<0.05) than usual care (no systematic
 5           monitoring or treatment adjustment) at 24 weeks for: Mean difference
 6           in proportion of patients with low disease activity (DAS28 <3.2);
 7           DMARD changes (higher); Patient global assessment of disease
 8           activity. However, there was NS difference for: Mean dose of non-oral
 9           steroids, prednisone and MTX dose; AEs; Pain (VAS) and disability.
10           Level 1+


     Rheumatoid arthritis: full guideline DRAFT (August 2008)             198 of 307
     DRAFT FOR CONSULTATION


1    8.1.5 Summary of evidence statements
 2   In recent onset RA time-integrated CRP predicts radiological progression 363
 3   and mean CRP correlates with articular index 360
 4
 5   In two studies of recent onset RA, intensive treatment strategies with the aim
 6   of keeping the Disease Activity Score to low levels of activity resulted in
 7   substantially better outcomes when compared with usual care for most
 8   measures of disease activity, remission, function and radiological progression
     288 357
 9           . A similar approach in established disease also resulted in improved
10   disease control 356.
11
12   In established disease, studies show high correlations between indices of
13   disease activity 358
14
15   In established disease changes in disease activity correlate with changes in
16   function 361 and indices that amalgamate several measures of disease activity
17   show greater validity than out-perform single measures of disease activity 359
     361
18
19   In established disease that disease activity index performs better than the Riel
20   Index and Mallya index for correlations with clinical status and joint damage,
21   and the ability to differentiate between low and high disease activity 362

22   8.1.6 From evidence to recommendations
23   The GDG noted that whilst there was no single measure nor composite
24   measure which was better than any other it seemed logical to recommend
25   both a laboratory measure of disease activity such as CRP and a well
26   validated composite score of disease activity such as DAS 28. It was also
27   noted that measurements of DAS28 were mandatory for initiating and
28   monitoring anti –TNF therapy according to current NICE guidance (see
29   section 7.3.13). The GDG felt very strongly that such objective numerical
30   assessments were preferable to descriptive words such as ‘better’, ‘worse’.
31   Furthermore such measurements would need to be made serially in order to
32   inform clinical decision making; this could include increasing therapy to
33   suppress disease activity when present or conversely cautiously decreasing
34   medication when disease activity was judged to be acceptably low.
35
36   The GDG were very impressed by the evidence for frequent (monthly)
37   measurements in patients with early active disease, where aggressive
38   treatment strategies aimed at keeping DAS28 to low levels produced
39   substantially better outcomes. Frequent monitoring of such patients in the
40   early stages of their disease were judged important in terms of better outcome
41   but patients with stable chronic well controlled disease would only need such
42   measurements infrequently.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             199 of 307
     DRAFT FOR CONSULTATION


1    8.1.7 Recommendations
2    R 35 In all people with rheumatoid arthritis, CRP and a composite score of
3    disease activity (such as DAS28) should be measured over time in order to
4    inform decision making about:
5          • increasing treatment, when indicated, to suppress disease activity
6          •    cautiously decreasing treatment, when indicated, when disease
7               activity is acceptably low.
 8   R 36 In all patients with early active disease, key constituent parts of disease
 9   activity and CRP should be measured monthly (using a composite score such
10   as DAS28) until an agreed level of disease control has been achieved by
11   appropriate treatment.

12   8.2       Content and frequency of review

13   8.2.1 Clinical introduction
14   RA is a chronic and unpredictable disease with fluctuations in activity. When
15   active, it has the propensity to damage the musculoskeletal and other
16   systems. The traditional approach to long term management has been to see
17   the patient at a frequency determined:
18         • by the problems at any one visit (active problems leading to more
19             frequent visits),
20         •    the practicalities of when clinic space allows appointments to be
21              made.
22   Such reviews have tended to focus on the immediate problems, lacked
23   structure, and have not paid enough attention to other less obvious disease
24   processes that might influence morbidity and mortality. The ARMA standards
25   of care have suggested an annual review should take place (ARMA Standards
26   of Care for people with Inflammatory Arthritis.364 An annual review could be
27   used to ensure that
28         • the disease status is addressed objectively, with formal measures of
29            disease activity, remission, damage and function.
30         •    current medications, educational needs, physical and psychosocial
31              issues, depression and fatigue are all addressed.
32         •    assessment and screening for complications of RA and other co-
33              morbidities takes place, especially osteoporosis and atherosclerosis.
34   In RA, problems outside the musculoskeletal system may be directly related to
35   the disease. For some patients the term RA is a misnomer, in that the disease
36   affects systems other than the musculoskeletal system. These manifestations
37   are related to:
38         • the joint inflammatory process itself, such as irreversible damage to
39            the cervical spine causing bone instability and deformity and potential
40            myelopathy.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              200 of 307
     DRAFT FOR CONSULTATION


1         •   the inflammatory process occurring in other organs and systems,
2             such as the eye (scleritis), the lung (pulmonary fibrosis and pleural
3             effusion), and vasculitis.
4    This section addresses the evidence for the content and frequency of a
5    regular review in order to ensure that disease control is optimised, and
6    problems directly and indirectly related to the disease are formally assessed
7    and objectively addressed, to ensure that no overt or covert problems lead to
8    a preventable impact on the quality of life for the patient.

9    8.2.2 Clinical methodological introduction
10   We looked for studies that assessed what should be the content of a regular
11   review of patients with established RA. Due to the sparsity of evidence, all
12   study types were looked at. Three RCTs365-368 were found that fulfilled the
13   inclusion criteria and no studies were found which addressed regular review in
14   terms of comorbidities. One of the RCTs366,368, was published as two separate
15   papers, reporting different follow-up times and so these have only been
16   counted once, however results from both the papers are reported and
17   referenced here.
18
19   Established RA
20   The 3 included RCTs365,366 367 were methodologically sound, single blind,
21   parallel group trials. The first RCT365 looked at annual and 4-montly review of
22   N=466 patients who were randomised to 2 different treatment arms: symptom
23   control/shared care setting versus aggressive treatment/hospital setting.
24   Patients in the symptom control group were assessed for HAQ score every 4
25   months and annually for OMERACT, OSRA, DAS28 and extra-articular
26   features. Patients in the aggressive treatment group were assessed for HAQ
27   score, ESR, CRP and tender/swollen joints every 4 months and annually for
28   OMERACT, OSRA, DAS28 and extra-articular features. The trial length was 3
29   years. The second RCT366,368 looked at review of N=209 patients who were
30   randomised to 2 different treatment arms: shared care with GP (no routine
31   hospital review but rapid access on request) setting versus traditional hospital
32   care (regular planned review every 3-4 months). The trial length was 2 years
33   and had a follow-up at 4 years. The third RCT367 looked at review of N=132
34   patients who were randomised to 3 different treatment arms: GP follow-up
35   (review on request) vs routine hospital follow-up (3-monthly review) vs OT
36   follow-up (3-monthly review). The trial length was 2 years.

37   8.2.3 Health economic methodological introduction
38   No health economic papers were identified

39   8.2.4 Clinical evidence statements
40   Established RA
41
42   Rapid access (shared care with GP) vs 3-4 month regular review
43   (traditional hospital care)



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              201 of 307
     DRAFT FOR CONSULTATION


1             •   One RCT366,368 found that rapid access was significantly better than
2                 traditional hospital regular review (every 3-4 months) for: Pain (VAS),
3                 change in pain (VAS) and self-efficacy score at 2 years (all p<0.05)
4                 and for ROM (right elbow, p<0.05) and patient satisfaction and
5                 confidence (p<0.01) at 4 years. However, there was NS difference for
6                 HAQ, Radiographic progression (Larsen scores) and Anxiety and
7                 depression at 2 years and for HAQ, Pain (VAS), morning stiffness,
8                 ROM (left elbow and both knees) and patient’s opinion of disease
9                 activity at 4 years. Level 1+
10   Symptom control vs aggressive treatment (both assessed at least every
11   4 months)
12       • One RCT365 found that at 3 years, aggressive treatment was
13          significantly better than symptom control for OSRA disease activity
14          score (OR -0.40, 95% CI -0.71 to -0.10, p=0.01). However, there was
15          NS difference for: HAQ, Patient’s and Physician’s global assessment,
16          tender and swollen joint count, Pain (VAS), DAS-28, Radiographic
17          progression (Larsen score), OSRA damage score, eroded joint count
18          and ESR. Level 1+
19   GP follow-up (on request) vs routine hospital follow-up (3-monthly) vs
20   OT follow-up (3-monthly)
21        • One RCT367 found that OT follow-up (3-monthly) was significantly
22           better than GP follow-up (on request) and routine hospital follow-up
23           (3-monthly) for: Articular Index at 2 years (p<0.05). However, there
24           were NS differences between the groups for: Articular Index at 1 year
25           and Functional capacity and ESR at 1 year and 2 years. Level 1+

26   8.2.5 Evidence to recommendations
27   The GDG noted the lack of consistency in the evidence relating to frequency
28   of review, place of review and assessment of aggressive treatment. It was
29   noted that no one approach would be suitable for everybody; regular review
30   may be suitable for some, whilst patient initiated review may be more suitable
31   for others. It was specifically felt that, for those people where it was thought
32   appropriate to offer patient initiated follow-up as when needed rather than
33   routine regular follow-up, it was essential that they were well educated about
34   their disease and knew when and how to obtain further help, for example by
35   contacting the named designated member of the MDT (see recommendations
36   in section 6.1). The GDG also noted that it was equally essential that where
37   patient-initiated follow-up was deemed appropriate, routine drug monitoring
38   must still take place as these people might otherwise only receive a routine
39   annual review.
40
41   The GDG were disappointed that there was no published evidence found for
42   the elements of what should be covered in a follow up. The GDG discussed
43   the GP Quality Outcomes Framework (QOF) points and a table from the
44   unpublished RA BSR guideline§§ that considers the interface between primary
45   and secondary care. GDG members who had been involved in drawing up the

     §§
          Unpublished at present but will be published by February 2009.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 202 of 307
     DRAFT FOR CONSULTATION


 1   BSR RA guidelines pointed out the lack of evidence to support the content of
 2   reviews. In the absence of any evidence, it was felt by the GDG that an
 3   annual review of the disease (with objective measures of activity, damage and
 4   function), complications and co-morbidities, was reasonable and that the
 5   content of the review should include assessment of disease status, primary
 6   prevention of ischaemic heart disease, osteoporosis assessment, making sure
 7   patient are not depressed, ensuring their cervical spine is stable, and
 8   checking for other possible organ involvement (e.g eye, lung, vasculitis)
 9   The GDG discussed that patients and healthcare professionals looking after
10   them are often unaware that they are at increased cardiovascular risk with
11   RA. Secondary care practitioners RA clinics need to be much more aware of
12   this and the GDG felt that the input of the RA Nurse Specialist is important
13   here. GDG felt that there was a need for awareness-raising in relation to RA
14   and co-morbid risks. Primary care especially needs to be aware that the of
15   management of RA is not just about managing the RA disease but also being
16   much more aware of the co-morbidities of ischaemic heart disease,
17   depression and osteoporosis.

18   8.2.6 Recommendations
19   R 37 People with satisfactorily controlled established RA should be offered
20   follow up at a frequency and location that are suitable to their needs, providing
21   they:
22         • have access to additional visits for disease flare-ups
23         •    have sufficient education to know when and how to seek rapid
24              access and help
25         •    receive ongoing drug monitoring
26         •    are offered an annual review.
27
28   R 38 All people with RA should be offered an annual review to include:
29        • assessment of disease activity and damage, and a measure of
30            functional ability (documenting possible serial decline of the disease)
31         •    checking for the development of co-morbidities such as hypertension,
32              ischaemic heart disease, osteoporosis and depression.
33         •    assessment of symptoms that might suggest complications of RA
34              such as disease of the cervical spine, lung, eyes and vasculitis
35         •    appropriate cross referral within the multidisciplinary team,
36         •    assessment of possible need for surgical referral (see section 8.3.6.).

37   8.3       Timing and referral for surgery

38   8.3.1 Clinical introduction
39   Despite recent advances in medical management of RA leading to a reduction
40   in the requirement for surgery, significant numbers of people with the disease
41   still go on to develop irreversible damage to joints and tendons. For patients

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                   203 of 307
     DRAFT FOR CONSULTATION


 1   with irreversible, or localised non-responding damage to the musculoskeletal
 2   system, surgical interventions can be an effective solution for pain relief,
 3   restoration of function, prevention of progressive deformity, and improvement
 4   in the quality of life. The timing of a surgical referral will clearly depend on the
 5   clinical urgency of the underlying problem. Myelopathy due to pressure on the
 6   spinal cord from cervical spine instability can at worst result in death. Short of
 7   this, the result can be irreversible neurological damage with a resultant
 8   disability. Clearly therefore cervical spine instability causing myelopathy
 9   requires urgent intervention. By contrast pressure on peripheral nerves, (e.g.
10   carpal tunnel syndrome) can result in pain and weakness but the urgency for
11   intervention is less than that of pressure on the spinal cord.
12
13   Some operations become technically more difficult or have worse outcomes if
14   the damage or deformity has progressed too far. This is one of many reasons
15   why a surgical opinion may need to be obtained early in disease progression,
16   in consultation with other appropriate members of the MDT.
17
18   As with all treatment, medical or surgical, the decision to undertake surgery
19   must be discussed with the patient. The potential risks and benefits of the
20   operation should be balanced against the risks and benefits of continuing
21   conservative management and the wishes of the informed patient.
22
23   Surgical and anaesthetic techniques have improved considerably. The risks of
24   intervention have decreased substantially, and the outcomes of most
25   operations are excellent with prolonged benefits (particularly hip and knee
26   replacements). Because RA affects many joints, it is important that the
27   surgeon has the knowledge to assess the overall problems of the RA patient,
28   be aware of the multisystem nature of the disease, the drugs that patients are
29   on and has the time and resources to assess patients in detail. While the
30   surgeon should have an overall ability to assess the patients needs, physical
31   and psychological, it is unlikely that any one surgeon will have the skillset
32   necessary to perform all the operations that may be necessary.
33
34   In this section, the focus is not on the type of intervention, which goes well
35   beyond the scope of this guideline, but on the timing of referral for a surgical
36   opinion.

37   8.3.2 Clinical methodological introduction
38   We looked for studies that assessed factors that determine the timing of
39   referral for surgery. Due to the small amount of evidence, all study types were
40   selected that were of a UK relevant population. Four cross-sectional
41   studies369-373 and one retrospective case-series374 were found that fulfilled the
42   inclusion criteria. One study369,370 was published as 2 separate papers looking
43   at different outcomes and different populations (physicians or patients) and so
44   has only been counted once but results from both papers have been included
45   and referenced here.
46
47   The first cross-sectional study369 performed a survey of N=1000 physicians
48   (N=500 rheumatologists and N=500 surgeons) and N=126 patients. The


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 204 of 307
     DRAFT FOR CONSULTATION


 1   survey focused on the physicians’ indications and timing of different types of
 2   surgery for rheumatoid hand disease and patients’ priorities and willingness to
 3   have hand surgery. The second cross-sectional study372 performed a survey
 4   of N=1379 RA patients and focused on the variables associated with having
 5   orthopaedic surgery or TJR. The third and fourth cross-sectional studies371,373
 6   performed surveys on N=62 and N=56 patients respectively and focused on
 7   predictors of RA patients having surgery.
 8   The retrospective case-series374 assessed the symptomology of N=111
 9   patients with rheumatoid cervical myelopathy who had undergone MRI or
10   cervical spine surgery or both.

11   8.3.3 Health economic methodological introduction
12   No health economic papers were identified

13   8.3.4 Clinical evidence statements
14   All studies were given an evidence grading of Level 3.

15   Indications for RA joint surgery - Physicians
16
     Study          Type of            Experts           Indications
                    surgery            opinions
     1 cross-       MCP joint          Hand surgeons     Most important: impaired hand
     sectional      arthroplasty       and               function followed by MCP joint
     study369,370                      rheumatologists   pain
                    Small joint        Hand surgeons     Progressive joint synovitis
                    synovectomy        Rheumatologists   Never indicated
                    Resection of       Hand surgeons     Impending tendon rupture
                    the distal ulna    and               followed by wrist pain
                                       rheumatologists
17
18   Indications for RA joint surgery – Patients
19
     Study          Patients’         Patients       Indications
                    opinions
     1 cross-       Willingness       Men and        Main concern: Hand pain followed by
     sectional      to have           Women (NS      hand function
     study369,370   surgery           difference)    Some concern: hand appearance
                    Concerns of       Women and      Women more concerned than men
                    inconvenienc      men
                    es, pain, risk
                    of
                    anaesthesia
                    and surgical
                    complication
                    s




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             205 of 307
DRAFT FOR CONSULTATION


1 cross-    Choosing       All patients     Associations: Age (patients older than
sectional   MCP                             50 years) and gender (female patients
study371    arthroplasty                    more likely).

                                            Predictors: Function followed by pain
                                            Not predictor: aesthetic consideration
1 cross-    Probability of All patients     Highest (univariate): female patients,
sectional   undergoing                      younger patients, those with long-term
study372    surgery                         disease, a poor functional ability,
                                            persistent active disease despite
                                            treatment, RF+ and presence of
                                            extraarticular complications and
                                            significant comorbidity

                                            Highest (multivariate): female gender,
                                            long-term disease (≥10 years), ACR
                                            functional grade III/IV and the
                                            presence of extraarticular
                                            complications
            Probability of All patients     Highest (Univariate): female patients,
            undergoing                      those with long-term disease,
            TJR                             functional class III/IV, persistent active
                                            disease despite treatment, presence
                                            of extraarticular complications and/or
                                            significant comorbidity

                                            Highest (multivariate): long-term
                                            disease (≥10 years), ACR functional
                                            grade III/IV and the presence of
                                            extraarticular complications
1 cross-    Hopes of       All patients     improving appearance and function
sectional   what surgery                    (44%), reducing pain (27%) and
study373    will do                         improving strength (15%)
            Important      All patients     ability to perform everyday activities
            aspects                         (75%), improvement of hand
                                            weakness (73%), ability to do one’s
                                            normal work (71%), reduction in hand
                                            (50%) and improvement of hand
                                            appearance (35%)
            What           All patients     function, pain, appearance and
            bothered                        weakness
            patients       Patients who     inability to work or do things with their
            most (hand     chose to have    hands
            RA)            surgery
                           Patients who     hand weakness and appearance
                           chose not to
                           have surgery




Rheumatoid arthritis: full guideline DRAFT (August 2008)            206 of 307
    DRAFT FOR CONSULTATION


                    patients’       Patients who     Surgery patients less likely to expect
                    post-           chose to have    difficulty with post-operative
                    operative       surgery vs not   rehabilitation; NS for belief in chance
                    expectations have surgery        of post-op complications
                    patients’       Patients who     Surgery patients more likely to expect
                    expectations chose to have       the ability to do more with their hands
                    for status 1    surgery vs not   in 1 year, to do more of their work,
                    year into the have surgery       have les pain and improved hand
                    future                           appearance
                    Most            Patients who     NS difference. However non-surgical
                    important       chose to have    patients valued their own opinion as
                    person to       surgery vs not   moist important and surgical valued
                    influence       have surgery     expert opinion more
                    surgical
                    decision
    1 case-         Patients with rheumatoid         Symptoms present (more):
    series374       cervical myelopathy who          paraesthesia, weakness,
                    underwent surgery vs             unsteadiness, and to exhibit extensor
                    conservative treatment           plantar reflexes, gait disturbances and
                                                     reduced power; Ranawat grades II
                                                     (NS) or III (SS) and not have normal
                                                     examination findings (SS)
                                                     MRI findings (more likely present):
                                                     Cord compression and impingement
                                                     on cord.
                                                     Less likely present: cervical
                                                     spondylosis; abnormal neurological
                                                     findings (but no compression or
                                                     impingement)
1
2   Timing for joint surgery
3
    Study           Type of          Experts              Timing
                    surgery          opinions
    1 cross-        MCP joint        Hand surgeons        Most appropriate time: Stage 3
    sectional       arthroplasty     and                  MCP joint disease
    study369,370                     rheumatologists
                    Extensor         Hand surgeons        Most appropriate time: 3-6
                    tenosynovecto    and                  months if the synovitis is
                    my               Rheumatologists      resistant to medical therapy
                                     Rheumatologists      appropriate after 12 months or
                                     vs Hand surgeons     more (26% vs 2%) and never
                                                          appropriate (8% vs 2%)

4   8.3.5 Summary of evidence statements
5        •      The amount of evidence to address the timing of surgical referral is
6               scanty and of limited quality. A survey of hand surgeons and
7               rheumatologists agreed on stage 3 MCP joint disease being the most



    Rheumatoid arthritis: full guideline DRAFT (August 2008)                207 of 307
     DRAFT FOR CONSULTATION


1             appropriate time for surgery, and 3-6 months of resistant synovitis for
2             extensor tenosynovectomy 369,370.
3         •   A survey of rheumatologists and hand surgeons found agreement on
4             MCP joint arthroplasty being indicated for function and pain, and
5             resection of the distal ulna indicated for tendon rupture and wrist pain.
6             However there was disagreement on the value of small joint
7             synovectomy with hand surgeons feeling that progressive joint
8             synovitis was an indication, and most rheumatologists feeling it was
9             never indicated 369,370
10        •   Most patients rank hand function and pain as the main concerns
              369,370 371
11                     and it is their hope that surgery will relieve these 373.
12        •   Hand appearance was of little importance in considering surgery in
13            one study 369,370, but improving function and appearance were main
14            hopes in another 373.
15        •   Cervical spine study – check evidence statement summarise
16            Hamilton study.

17   8.3.6 From evidence to recommendations
18   The GDG noted the lack of evidence which really addressed this question and
19   felt that it was extremely unlikely that any appropriate clinical trials would
20   specifically address this issue. Indeed the GDG felt that in some cases it
21   would be unethical in doing such trials. It was felt that there was currently an
22   enormous variation in thresholds for seeking surgeon’s involvement in the
23   formulation of management plans for patients who might eventually need
24   surgery.
25
26   The GDG felt that it would be appropriate to involve surgeons early, even if
27   possible surgery was not urgently indicated. It was noted that assessment for
28   referral for surgery was already recommended as part of an annual review
29   (see chapter 6 and section 8.2.6). It was agreed that there were 4 primary
30   reasons for considering referral for surgery:
31         • Persistent pain due to joint damage or other identifiable soft tissue
32             cause
33        •   Worsening function
34        •   Progressive deformity
35        •   Persisting localised synovitis
36   Additional reasons included tendon rupture, nerve compression, joint
37   instability (e.g. cervical), infection and incidental radiological findings such as
38   stress fracture. Although the need and urgency for possible surgical
39   intervention would vary with the underlying condition, the GDG felt that an
40   early surgical opinion should always be obtained, particularly in cases where
41   the outcome of a surgical procedure could be jeopardised by a delay in
42   surgical referral.
43



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 208 of 307
     DRAFT FOR CONSULTATION


 1   For tendon rupture, localised non-responding synovitis, nerve compression,
 2   septic arthritis and stress fractures it was felt appropriate to refer for a surgical
 3   opinion before damage had progressed too far. Multiple tendon rupture,
 4   persisting neurological deficit and uncorrectable deformity causing disability
 5   could be seen as a failure for the team. For more long term problems such as
 6   pain, function and deformity the need for referral was not so urgent but again
 7   should not be delayed until further irreversible changes had occurred.
 8
 9   There were some instances for example cervical myelopathy where urgent
10   action was clearly needed and even though the evidence for this condition
11   was from a retrospective study, the GDG felt that an urgent MRI scan and
12   surgical opinion would be needed in presence of appropriate symptoms or
13   signs.
14
15   The GDG noted that while in general the indications for joint replacement for
16   people with RA should be regarded in the same way as for people with OA.8
17   RA patients might be considerably younger and the wear and tear in the
18   arthroplasty could well be less in view of the polyarthritic nature of their
19   disease and reduced demands on the joint as a result. The GDG therefore felt
20   that there should be a specific recommendation pointing out that, in
21   comparison with patients with OA the comparative youthfulness of an RA
22   patient should not preclude consideration for joint replacement.

23   8.3.7 Recommendations
24   R 39 People with RA should be referred for an early specialist surgical opinion
25   if any of the following are failing to respond to optimal non-operative
26   multidisciplinary interventions:
27         • persistent pain due to joint damage or other identifiable soft tissue
28             cause
29        •   worsening function
30        •   progressive deformity
31        •   persisting localised synovitis.
32
33   R 40 People with the following complications of RA should be referred for a
34   specialist surgical opinion before damage or deformity becomes irreversible:
35        • imminent or actual tendon rupture
36        •   nerve compression (e.g. carpal tunnel syndrome)
37        •   stress fracture.
38
39   R 41 For those people with rheumatoid arthritis to whom surgery is offered, it
40   should be emphasised that the principal anticipated benefits are:
41        • relief of pain
42        •   improvement , or prevention of further deterioration, in function
43        •   prevention of deformity


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 209 of 307
     DRAFT FOR CONSULTATION


1     with cosmetic improvements being of lesser importance.
2
3    R 42 People with rheumatoid arthritis who have suspected or proven septic
4    arthritis (especially in a prosthetic joint) should receive urgent combined
5    medical and surgical management.
6
7    R 43 The development of any symptoms or signs suggestive of cervical
8    myelopathy*** should prompt a request for:
9         • an urgent MRI scan
10            •   referral for a specialist surgical opinion if indicated.
11
12   R 44 Concerns about the long-term durability of joint replacements should not
13   preclude the offering of such procedures to younger people with rheumatoid
14   arthritis.




     ***
           E.g. paraesthesiae, weakness, unsteadiness, reduced power, extensor plantars

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                  210 of 307
     DRAFT FOR CONSULTATION



1    9      Other aspects and treatment

2    9.1    Diet

3    9.1.1 Clinical introduction
 4   Every healthcare professional will know of people with rheumatoid arthritis
 5   who have experimented with their diet, and can testify to improvements in
 6   their disease activity. Some patients seem to have specific intolerances,
 7   where certain foods or drinks seem to cause flare-ups in their joints, and
 8   therefore they avoid them. Other books and internet sites advocate special
 9   inclusion or exclusion diets that make great claims about their efficacy.
10   Experimenting with diet gives the person with RA the opportunity to exercise
11   control over an important aspect of their life, and determine whether certain
12   foods included or excluded benefit them and their disease. This section does
13   not focus on weight reduction but looks at whether there is any evidence to
14   support the beneficial effects of specific dietary manipulations in helping to
15   control rheumatoid arthritis? Weight reduction was specifically addressed
16   within this evidence review. Weight reduction is covered within the
17   Osteoarthritis guideline.210

18   9.1.2 Clinical methodological introduction
19   We looked for studies that investigated the efficacy of different types of diet or
20   dietary supplements with respect to symptoms, joint damage, function and
21   quality of life in patients with RA (recent onset and established disease). Due
22   to the large volume of evidence, only MA and RCTs were selected which were
23   of a UK-relevant population and if the population was mixed arthritis there had
24   to be >75% RA or RA subgroup analysis.
25
26   Three MAs375-377 and 14 RCTs378-393 were found that fulfilled the inclusion
27   criteria. One of these RCTs was published as three separate papers388-390
28   reporting different outcomes and so the trial has only been counted once,
29   however results from all three papers are reported and referenced here. One
30   RCT393 was excluded due to methodological limitations. No studies were
31   found looking at patients with a recent onset of RA, all trials were conducted
32   using patients with established RA, and the two MAs used a trials which were
33   of a mixed population (recent onset and established RA).
34
35   Mixed population (Recent onset and Established RA)
36   Three MAs375-377 were found that fulfilled the criteria and focused on RCTs
37   which compared either omega-3 supplements and fish-oils376, herbal
38   therapies375, or lacto-vegetarian, vegan or Mediterranean diets377 in patients
39   with RA. All trials were methodologically sound.
40
41   The first SR/MA375 included 11 RCTs which were all double blind, placebo-
42   controlled trials. However, they differed with respect to:



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              211 of 307
     DRAFT FOR CONSULTATION


1         •   Intervention [N=7 RCTs used GLA (sources: evening primrose oil,
2             blackcurrant seed oil, borage seed oil); N=1 RCT used feverfew, N=1
3             RCT used Trypterygium wilfordii hook F, N=1 RCT used topical
4             capsaicin and N=1 RCT used Reumalex (contains willow bark)]
5         •   Study size (range N=20 to N=70)
6         •   Study quality – max score of 5 (N=10 studies reasonable to good
7             quality; N=1 study poor quality)
8         •   Study duration – length of intervention (4 weeks to 15 months)
 9   The second SR/MA376 included 17 RCTs which were all double blind, placebo
10   (inert-substance) controlled trials. However, they differed with respect to:
11         • Intervention – total omega-PUFA (not reported and range 1.7g to
12            9.6g)
13        •   Study size (range N=12 to N=90)
14        •   Study quality – max score of 5 (N=12 studies reasonable to good
15            quality; N=5 studies poor quality)
16        •   Study duration – length of intervention (1 month to 15 months)
17   The third MA377 was not a systematic review but a pooled analysis of 3
18   unblinded, placebo-controlled RCTs which differed with respect to:
19        • Design – (2 RCTs parallel; 1 RCT cross-over)
20        •   Intervention – (1 RCT Mediterranean diet vs western diet, 1 RCT
21            lacto-vegetarian diet vs normal diet, 1 RCT vegan diet vs control
22            period)
23        •   Study size (range N=22 to N=56)
24        •   Study duration – length of intervention (9 weeks to 4 months)
25
26   Established RA
27   14 RCTs377-392 were found that fulfilled the inclusion criteria. One of these
28   RCTs was published as three separate papers388-390 reporting different
29   outcomes and so the trial has only been counted once, however results from
30   all three papers are reported and referenced here. All trials were
31   methodologically sound.
32
33   All 14 RCTs were parallel group studies, but they differed with respect to the
34   following:
35         • Sample size (range: N=30 to N=116)
36        •   Blinding (6 RCTs double blind, 6 RCTs single blind, 2 RCTs
37            unblended/blinding not mentioned)
38        •   Trial length (Range: 4 weeks to 9 months; follow-up ranged from 2
39            months to 1 year post-treatment)
40        •   Treatment – diets (4 RCTs food intolerance and allergy; 4 RCTs
41            vegetarian or vegan; 2 RCTs vitamin and mineral supplements; 1
42            RCT fish oils;1 RCT Mediterranean; 1 RCT experimental; 1 RCT
43            calorie-restricted)

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             212 of 307
     DRAFT FOR CONSULTATION


1            •   Treatment regimen – dose and comparison

2    9.1.3 Health economic methodological introduction
3    No health economic papers were identified

4    9.1.4 Clinical evidence statements
5    Mixed population (Recent onset and Established RA)
6
7    GLA vs placebo
 8           •   1 MA375 found that GLA was significantly better than placebo for:
 9               Pain (VAS) and pain scale (0-4); Patient’s Global evaluation, morning
10               stiffness and joint tenderness. However there was NS difference for:
11               Pain (absolute score), morning stiffness (absolute score), joint
12               swelling and reduction in NSAID consumption. There was significant
13               heterogeneity for Physician’s global evaluation. Level 1++
14
15   Omega-3 PUFAs vs placebo
16           •   1 MA376 found that Omega-3 PUFAs were significantly better than
17               placebo for: Patient’s assessment of Pain, Morning stiffness, Number
18               of painful/tender joints, NSAID consumption. However there was NS
19               difference for RAI and physician’s assessment of Pain. Level 1++
20   Vegetarian and Mediterranean diets
21           •   1 pooled analysis377 found that Lacto-vegetarian, strictly vegetarian
22               and modified Cretan-Mediterranean diets were significantly better
23               than control diets for: univariate analysis - weight loss and pain.
24               Level 1+
25
26   Established RA
27
28   Gluten-free vegan diet followed by lacto-vegetarian diet
29           •   1 RCT388-390 found that responders were significantly better than non-
30               responders and controls for: pain, morning stiffness, HAQ, Number of
31               tender and swollen joints, global assessment and RAI. However there
32               was NS difference for radiographic score, grip strength and ESR.
33               Level 1++
34
     Study           Treatment            Follow-   Outcomes                                    Result - best
                                          up                                                    treatment
     1 RCT392        Fish oil             36        Pain (VAS); Reduction in daily NSAID         Fish
     Level 1++       supplements vs       weeks     requirement >30%
                     placebo
                                                    HAQ, morning stiffness, DAS28-CRP,          NS
                                                    CRP, Grip strength, Number or type of
                                                    AEs, Number of withdrawals and Type of
                                                    AEs leading to withdrawal

     1 RCT378        Mediterranean        12        DAS28, HAQ score, SF-36 Swollen joint       Mediterranean
                     diet vs usual diet   weeks     count, Pain (VAS), CRP level, withdrawals
                                                    and weight loss



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                213 of 307
DRAFT FOR CONSULTATION

                                           Tender joint count, ESR, Patients’ global    NS
                                           assessment of disease activity, Morning
                                           stiffness, SOFI score and GAT score
1 RCT379    Experimental diet    10        Morning stiffness, Grip strength, Walk
Level 1+    (to maintain or      weeks     time, Tender and swollen joints, Patient’s
            reduce weight)                 and physician’s global assessment, ESR
            vs placebo diet                and RF
1 RCT380    Experimental         6 weeks   Swollen joint count, Morning stiffness and   Experimental
Level 1+    calorie-restricted             Pain
            diet vs usual diet             BMI, weight, Tender joint count,             NS
                                           Physician’s global assessment, HAQ
                                           score and Larsen Score
                                           Withdrawals                                  Usual
1 RCT385    Vitamin E            12        Pain (morning, evening and after chosen      Vit E
Level 1++   supplementation      weeks     activity), response rates (pain in morning
            vs placebo                     and after chosen activity), patient’s and
                                           investigator’s global assessment of
                                           efficacy

                                           response rates (pain in the evening), RAI,   NS
                                           morning stiffness, swollen joints and AEs
                                 20        RAI, Pain measures, morning stiffness,       NS
                                 weeks     swollen joints and global assessment of
                                           efficacy
1 RCT386    Selenium vs          90 days   CRP, QoL (arm movements and health           Selenium
Level 1+    placebo                        perception)
                                           Pain (VAS), RAI, tender and swollen          NS
                                           joints, morning stiffness
                                           QoL (daily and social activities, mood,
                                           physical activity, symptoms and work) and
                                           AEs
1 RCT381    Elemental diet       4 weeks   Swollen joint count, ESR, General            Elemental
Level 1+    (hypoallergenic)               Assessment of Health and BMI
            vs usual diet        4 and     CRP, RAI, morning stiffness, Pain,           NS
                                 12
                                 weeks
                                 12        Swollen joint count, BMI, ESR, General       NS
                                 weeks     Assessment of Health
1 RCT382    Elemental diet vs    4 weeks   Average grip strength, RAI, and weight       Elemental
Level 1+    usual diet                     loss
                                           CRP                                          NS
1 RCT383    hypoallergenic,      24        Tender joint count, RAI, ESR                 Hypo
Level 1+    non-allergenic       weeks     BMI, Weight, Swollen joint count, Morning    NS
            diet vs control                stiffness, Pain severity (VAS), HAQ, CRP,
            diet                           Responders (Paulus Index – 20% and
                                           50%), Patient’s global assessment of
                                           disease
1 RCT384    allergen-free diet   4 weeks   Weight reduction                             Allergen-free
Level 1+    vs allergen-                                                                (all p<0.05)
            restricted diet                Swollen and tender joints, Global            NS
                                           assessment, Morning stiffness, RAI,
                                           Fatigue, Grip strength, Walking time,
                                           ESR, CRP, RF
1 RCT387    Vegan vs non-        1 year    ACR20 responders                             Vegan
Level 1++   vegan diet                                                                  (p<0.05)
                                           Radiographic progression                     NS
1 RCT391    Vegan vs non-        3         Weight reduction, Rheumatic pains, joint     Vegan (all
Level1+     vegan diet           months    swelling and morning stiffness               p<0.01)
                                           CRP, ESR, changes in disease activity,       NS
                                           mean amount of deterioration and ability
                                           to move




Rheumatoid arthritis: full guideline DRAFT (August 2008)                              214 of 307
     DRAFT FOR CONSULTATION

     1 RCT388-390   Vegan vs non-        13       Swollen joints, pain (VAS), grip strength,     Gluten-free (all
     Level 1++      vegan diet           months   weight reduction; global assessment,           p<0.02;
                    Gluten-free                   morning stiffness, HAQ, CRP and ESR            p<0.001)
                    vegan diet                    and RF (IgM) RAI, tender
                    followed by lacto-
                    vegetarian diet               RF
                    vs omnivorous                                                                NS
                    diet
1

2    9.1.5 Summary of evidence statements
3           •   Most trials show some benefit with a variety of dietary modifications
4               or supplementary additions, but no modifications seemed to exert
5               global improvements in disease activity and function 375 376 377,378
                380,385 386 381 382 383,384 387 391 388-390
6                      .                                    . The studies were conducted in
7               patients who were already established on conventional therapies, and
8               cannot be considered in isolation.
 9          •   The most impressive diet appeared to be gluten free diet followed by
10              vegetarian diet, although these patients also had physiotherapy 388-
                390
11                  . Other diets with benefits in disease activity include the “allergen –
12              free” diet384, elemental diet 381 and allergen restricted diet in some
13              patients 383, Mediterranean diet 377,378, omega-3 and PUFA’s376.

14   9.1.6 From evidence to recommendations
15   The GDG noted that many of the dietary interventions did seem to have
16   benefit when taken with conventional therapies. However, no diet produced
17   positive results in all of the patients in the trials, and there was insufficient
18   evidence to support the recommendation of a single diet. Some of the diets
19   might be unpopular with some patients, such as vegetarian diets, and some
20   might be unpalatable with understandably poor compliance, such as
21   elemental diets. It was felt that it would be helpful in a recommendation to give
22   some direction to RA patients. There was discussion about the evidence to
23   show that the principles of a Mediterranean diet* might be beneficial in people
24   with RA especially because of the impact of such a diet on cardiovascular risk
25   factors. Because
26         • people with RA are at an even greater risk of cardiovascular disease
27            than the rest of the population (add in number of CV section),
28          •   such a diet might be beneficial to the musculoskeletal symptoms of
29              RA,
30          •   this type of diet is more likely to be followed than some of the other
31              more unpalatable alternatives .
32   it was felt reasonable to recommend the principles of a Mediterranean diet

33   9.1.7 Recommendations
34   R 45 People with rheumatoid arthritis who wish to experiment with their diet
35   should be informed that there is no strong evidence that their arthritis will




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                  215 of 307
    DRAFT FOR CONSULTATION


1   benefit, but could be encouraged to follow the principles of the components of
2   a Mediterranean diet†††




    †††
       A Mediterranean diet is characterised by: a high consumption of fruit, vegetables, cereals
    and legumes394; less red meat and more fish than Western diets; includes a moderate intake
    of wine and the primary source of fat is olive oil.378


    Rheumatoid arthritis: full guideline DRAFT (August 2008)                         216 of 307
     DRAFT FOR CONSULTATION



1    9.2       Complementary and alternative interventions

2    9.2.1 Clinical introduction
 3   Conventional approaches to the management of RA are not always as
 4   successful as either healthcare professionals or patients would wish. Patients
 5   are frequently concerned about side-effects with prescribed drugs, as
 6   witnessed by calls to the National Rheumatoid Arthritis Society helpline on this
 7   issue. With the combination of less than total efficacy, and concern over
 8   toxicity with conventional drugs, it is understandable therefore that people with
 9   RA will explore other approaches to try to help themselves with their disease.
10   They are not short of suggested solutions, with a plethora of articles and
11   advertisements in newspapers, magazines, the internet, television and radio,
12   giving information on complementary, alternative and other non-
13   pharmacological interventions which are claimed to relieve or (misleadingly)
14   cure their arthritis. Surveys have shown that a substantial proportion of people
15   with RA will try complementary and alternative interventions395, perhaps
16   reflecting the lack of complete satisfaction with conventional approaches, and
17   also a desire to help themselves. Patients with a recent onset of inflammatory
18   arthritis may try alternative remedies first before presenting to their GP. Out of
19   all of the various therapies available, members of the National Rheumatoid
20   Arthritis Society were asked to list their preferences, and the top six were
21   selected from these for literature searches and these were acupuncture,
22   copper bracelets, aromatherapy, massage, reflexology and homeopathy.

23   9.2.2 Clinical methodological introduction
24   We looked for studies that assessed which aspects of complementary,
25   alternative and other non-pharmacological interventions are effective in
26   patients with RA (recent onset of disease and established disease). Due to
27   the small volume of evidence, trials of all types were selected that were of a
28   UK relevant population. One MA396, 2 RCTs397-399 and 1 case-series400 were
29   found that fulfilled the inclusion criteria. One RCT398,399 was published as two
30   separate papers reporting different follow-up times but was excluded due to
31   methodological limitations. All trials were methodologically sound.
32
33   Mixed arthritis (Recent-onset and established RA)
34   The Cochrane SR/MA396 was well conducted. The 2 RCTs included in the
35   analysis compared acupuncture vs placebo. However, they differed with
36   respect to:
37        • Intervention [N=1 RCT used acupuncture (needles manipulated);
38            N=1 RCT used electroacupuncture]
39         •    Study size (range N=20 and N=64)
40         •    Study quality – max score of 5 (N=1 study good quality; n=1 study
41              reasonable quality)
42         •    Study duration – length of intervention (N=1 RCT 5 weeks; N=1 RCT
43              3 months)


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              217 of 307
     DRAFT FOR CONSULTATION


1    The additional RCT397 assessed massage therapy vs standard care in N=22
2    patients (disease duration of patients not mentioned) in a 4-week treatment
3    phase.
4
5    Established RA
6    The case-series400 looked at plant-based homeopathic preparations +
7    antioxidants (Vitamin C 1000 mg and Vitamin E 800 mg intramuscularly) in
8    N=30 patients in a 5-week treatment phase.

9    9.2.3 Health economic methodological introduction
10   No health economic papers were identified

11   9.2.4 Clinical evidence statements
12   Mixed arthritis (Recent-onset and established RA)
13
14        •   1 MA396 found that electroacupuncture was significantly better than
15            placebo for: Pain (0-4 scale) at end of treatment-24 hours and at 4
16            month follow-up. Level 1++
17        •   The same MA396 found that there was NS difference between
18            Acupuncture and placebo at end of treatment-5 weeks for: Pain
19            (VAS) at end of treatment-5 weeks; swollen and tender joints at end
20            of treatment-5 weeks; Disease activity (DAS); Global Health
21            Questionnaire; ESR; CRP; Analgesic uptake; Patient’s global
22            assessment. Level 1++
23        •   1 RCT397 found that Hand massage was significantly better than
24            control (standard treatment) at 4 weeks (end of treatment, change
25            from baseline) for: Pain (VAS), Anxiety (STAI), Depression (POMS)
26            and Grip strength. Level 1+
27   Established RA
28       • One case-series400 found that at 5 weeks (end of study, change from
29           baseline), patients treated with Plant-based homeopathic
30           preparations + antioxidants had decreased Pain (VAS) change from
31           baseline -1.5, increased level of well-being (VAS) and decreased
32           restriction of movement. Additionally, there was successful reduction
33           of drugs that patients’ had been previously taking (all drugs causing
34           AEs were immediately eliminated – NSAIDs, MTX and/or
35           paracetamol). Level 3

36   9.2.5 From evidence to recommendations
37   The GDG felt that the evidence for any complementary therapy or medicines
38   being effective was lacking: the GDG was disappointed by this given the
39   overall popularity of these treatments. The only positive evidence surrounded
40   a small short term study of electroacupuncture and a 4 week study of hand
41   massage. A mixture of plant based homeopathic preparations in a short term
42   trial allowed a reduction in analgesia. However the feeling of the GDG was
43   that none of this constituted strong evidence for sound recommendations. The
44   GDG found that the benefit of complementary therapies was akin to that of

     Rheumatoid arthritis: full guideline DRAFT (August 2008)            218 of 307
     DRAFT FOR CONSULTATION


1    analgesics in providing short term benefit and that this should be reflected in
2    the recommendations. It was felt that the information on the lack of evidence
3    should be shared with patients, but if they chose to try complementary or
4    alternative treatments then it should be ensured that they do not take these to
5    the exclusion of conventional therapies. Some members mentioned that in
6    their experience a bias could develop against patients trying complementary
7    therapies and that this might influence the care given by the MDT. The GDG
8    felt that this attitude should be discouraged

9    9.2.6 Recommendations
10   R 46 People with rheumatoid arthritis wishing to try complementary or
11   alternative treatments should be informed that although some may provide
12   short-term symptomatic benefit the evidence for their long-term efficacy is
13   limited or non-existent.
14
15   R 47 If people with RA do decide to try complementary or alternative
16   treatments:
17        • they should be advised that these approaches should not replace
18            conventional treatment
19        •   this should neither prejudice the attitudes of individual members of
20            the MDT to the manner in which they relate to patients nor bias the
21            care they are offered.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             219 of 307
     DRAFT FOR CONSULTATION


1    10       Areas for future research
2    The cost-effectiveness of MRI and ultrasound should be evaluated for their
3    role in establishing the diagnosis and prognosis of small joint synovitis
4
5    Further evaluations, including cost-effectiveness studies, should take place of
6    the following in people with rheumatoid arthritis:
7          • Disease-related written information in early disease
8         •    Written self-management materials in early and established disease
9         •    Structured 1:1 information giving in early and established disease
10        •    Refresher courses and updates
11        •    Long-term benefits of group behavioural programmes
12        •    The Arthritis Self Management Programme and the Chronic Disease
13             Self Management Programme in early and established disease with
14             study populations recruited from out-patient clinics
15   The following further evaluations, including cost-effectiveness studies, should
16   take place of in people with rheumatoid arthritis:
17        • For exercise, the best methods of delivery, the optimal mode and
18            level of activity, and methods of maximising long-term concordance
19        •    The individual components of comprehensive physiotherapy, with
20             particular reference to the effectiveness of electrophysical agents,
21             and their optimal timing throughout the course of the disease
22   The following further evaluations, including cost-effectiveness studies, should
23   take place in people with rheumatoid arthritis:
24        • Comprehensive occupational therapy interventions
25        •    The usefulness of work rehabilitation programmes, both for those
26             currently in work but at risk of job loss and for those already
27             unemployed but needing help in returning to work
28   The role of DMARDs in the treatment of mild rheumatoid arthritis should be
29   assessed.
30
31   The cost effectiveness of aggressive early management with biologics versus
32   conventional DMARDs should be assessed.
33
34   Studies should be performed to address the effect of symptom duration on
35   outcome
36
37   The most appropriate treatment strategy for people who fail on their first anti-
38   TNF drug should be determined.
39
40   The components of the nurse specialist's role that exert the greatest benefits
41   on outcomes in early and established RA should be investigated.
42



     Rheumatoid arthritis: full guideline DRAFT (August 2008)               220 of 307
    DRAFT FOR CONSULTATION



1   11    Abbreviations
    Abbreviation        Description
    ABA                 Abatacept
    ACR                 American College of Rheumatology (see ARA)
    ACR20, 50, 70       ACR-criteria 20-50-70
    ADA                 Adalimumab
    ADL                 Activities of daily living
    AEs                 Adverse events
    AIMS                Arthritis Impact Measurement Scale
    AL-TENS             Acupuncture-like transcutaneous electrical nerve
                        stimulation
    Anti-CCP            Anti-cyclic citrullinated peptide
    Anti-TNF            Anti-Tumor Necrosis Factor
    ARA                 American Rheumatism Association (now ACR)
    ARMA                Arthritis and Musculoskeletal Alliance
    AUC                 Area Under the Curve
    ARC                 Arthritis Research Campaign
    BMI                 Body mass index
    BSR                 British Society of Rheumatology
    CBT                 Cognitive Behavioral Therapy
    CI                  Confidence interval (95% unless stated otherwise)
    COX-2               Cyclooxegenase-2
    CRP                 C-reactive protein
    CS                  Corticosteroid
    CsA                 Cyclosporin A
    CTX                 Cyclophosphamide
    CV                  Cardio-vascular
    DAS (DAS28,         Disease Activity Score
    DAS32)
    DMARD               Disease Modifying Antirheumatic Drug
    EA                  Electroacupuncture
    EQ-5D               EuroQol 5-dimensional outcomes questionnaire
    EOW, eow            Every other week
    ESR                 Erythrocyte Sedimentation Rate
    ETN                 Etanercept
    EULAR               The European League Against Rheumatism
    EW, ew              Every week
    GDG                 Guideline Development Group
    GI                  Gastrointestinal
    HAQ                 Stanford health assessment questionnaire score
    HCQ                 Hydroxychloroquine
    IA                  Inflammatory Arthritis, or Intra-Articular
    ICER                Incremental Cost-Effectiveness Ratio
    IFX                 Infliximab
    IM                  Intramuscular
    IRGL                Impact of Rheumatic diseases on General health and

    Rheumatoid arthritis: full guideline DRAFT (August 2008)         221 of 307
DRAFT FOR CONSULTATION


                    Lifestyle Questionnaire
ITT                 Intention to treat analysis
IV / iv             Intravenous
JSN                 Joint Space Narrowing
LASER               Light amplification by stimulated emission of radiation
MA                  Meta-analysis
MACTAR              McMaster Toronto Arthritis Patient Preference Disability
                    Questionnaire
MCP                 Metacarpophalangeal joint
MDT                 Multidisciplinary Team
MTP                 Metatarsophalangeal joint
MTX                 Methotrexate
MHRA                Medicines and Healthcare products Regulatory Agency
MI                  Myocardial infarction
NCC-CC              National Collaborating Centre for Chronic Conditions
NHS                 National Health Service; this guideline is intended for
                    use in the NHS in England and Wales
NICE                National Institute for Health and Clinical Excellence
NS                  Not significant (at the 5% level unless stated otherwise)
NSAID               Non-steroidal anti-inflammatory drugs
OMERACT             Outcome Measures in Rheumatoid Arthritis Clinical
                    Trials
OR                  Odds ratio
OT                  Occupational Therapy or Therapist
PIP                 Proximal interphalangeal joint
PPI                 Proton pump inhibitor
PPV                 Positive Predictive Value
PUFA                Polyunsaturated fatty acid
QALY                Quality of Life-adjusted year
RAI                 Ritchie Articular Index
RCT                 Randomised controlled trial
RF                  Rheumatoid Factor
ROM                 Range of motion
RR                  Relative risk
RTX                 Rituximab
SAARD               Slow acting antirheumatic drugs
SAEs                Serious Adverse Events
SC / sc             Subcutaneous
SDAI                Simplified Disease Activity Index
SHS                 Sharp /van der Heijde Score
SF-12               Short Form (12 point) questionnaire
SF-36               Short Form (36 point) questionnaire
SMD                 Standardised mean difference
SR                  Systematic review
SS                  Statistically significant
SSZ                 Sulphasalazine
TENS                Transcutaneous electrical nerve stimulation
TF                  Transdermal Fentanyl


Rheumatoid arthritis: full guideline DRAFT (August 2008)           222 of 307
    DRAFT FOR CONSULTATION


    TJR                 Total Joint Replacement
    TSS                 Total Sharp Score
    UA                  Undifferentiated Arthritis
    UPA                 Undifferentiated polyarthritis
    VAS                 Visual analogue scale
    WMD                 Weighted mean differences
1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   223 of 307
     DRAFT FOR CONSULTATION



1    12     Glossary
 2   Biological / Biologic Type of DMARD which targets pro-inflammatory
 3   cytokines that are involved in joint destruction (particularly TNF-alpha and IL-1
 4
 5   Clinically significant improvement Some trials define a dichotomous
 6   outcome of clinically significant pain relief as having been achieved above a
 7   specific threshold on a pain score, eg pain VAS. However, there is no
 8   standard threshold and each such trial should be considered individually.
 9
10   Cohort study A retrospective or prospective follow-up study. Groups of
11   individuals to be followed up are defined on the basis of presence or absence
12   of exposure to a
13   suspected risk factor or intervention. A cohort study can be comparative, in
14   which
15   case two or more groups are selected on the basis of differences in their
16   exposure
17   to the agent of interest.
18
19   Confidence interval (CI) A range of values which contain the true value for
20   the population with a stated ‘confidence’ (conventionally 95%). The interval is
21   calculated from sample data, and generally straddles the sample estimate.
22   The 95% confidence value means that if the study, and the method used to
23   calculate the interval, is repeated many times, then 95% of the calculated
24   intervals will actually contain the true value for the whole population.
25
26   Cochrane review The Cochrane Library consists of a regularly updated
27   collection of evidence-based medicine databases including the Cochrane
28   Database of Systematic Reviews (reviews of randomised controlled trials
29   prepared by the Cochrane Collaboration).
30
31   Cost–consequence analysis A type of economic evaluation where, for each
32   intervention, various health outcomes are reported in addition to cost, but
33   there is no overall measure of health gain.
34
35   Cost–effectiveness analysis An economic study design in which
36   consequences of different interventions are measured using a single outcome,
37   usually in natural units (for example, life-years gained, deaths avoided, heart
38   attacks avoided, cases detected). Alternative interventions are then compared
39   in terms of cost per unit of effectiveness.
40
41   Cost–utility analysis A form of cost–effectiveness analysis in which the units
42   of effectiveness are quality adjusted life-years (QALYs).
43
44   DMARD (Disease Modifying Antirheumatic Drug) treatment that can reduce or
45   prevent joint damage
46
47   Incremental cost The cost of one alternative less the cost of another.
48

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              224 of 307
     DRAFT FOR CONSULTATION


 1   Incremental cost effectiveness ratio (ICER) The ratio of the difference in
 2   costs between two alternatives to the difference in effectiveness between the
 3   same two alternatives.
 4
 5   Larsen Score Method of assessing radiographic joint damage cause by RA
 6
 7   Manual therapy A range of physiotherapy techniques where the affected joint
 8   (typically the hip) is manipulated and stretched beyond the range of motion
 9   that the person with osteoarthritis is able to use.
10
11   Meta-analysis A statistical technique for combining (pooling) the results of a
12   number of studies that address the same question and report on the same
13   outcomes to produce a summary result.
14
15   Methodological limitations Features of the design or reporting of a clinical
16   study which are known to be associated with risk of bias or lack of validity.
17   Where a study is reported in this guideline as having significant
18   methodological limitations, a recommendation has not been directly derived
19   from it.
20
21   Multivariate analysis of more than one variable at a time. Takes into account
22   the effects of all variables on the response of interest.
23
24   Observational study Retrospective or prospective study in which the
25   investigator observes the natural course of events with or without control
26   groups, for example cohort studies and case-control studies.
27
28   Odds ratio A measure of treatment effectiveness: the odds of an event
29   happening in the intervention group, divided by the odds of it happening in the
30   control group. The
31   ‘odds’ is the ratio of non-events to events.
32
33   p-values The probability that an observed difference could have occurred by
34   chance. A p-value of less than 0.05 is conventionally considered to be
35   ‘statistically significant’.
36
37   Quality of life Refers to the level of comfort, enjoyment and ability to pursue
38   daily activities.
39
40   Quality-of-life adjusted year (QALY) A measure of health outcome which
41   assigns to each period of time a weight, ranging from 0 to 1, corresponding to
42   the health-related quality of life during that period, where a weight of 1
43   corresponds to optimal health, and a weight of 0 corresponds to a health state
44   judged equivalent to death; these are then aggregated across time periods.
45
46   Randomised controlled trial (RCT A trial in which people are randomly
47   assigned to two (or more) groups: one (the ) experimental group) receiving the
48   treatment that is being tested, and the other (the comparison or control group)
49   receiving an alternative treatment, a placebo (dummy treatment) or no
50   treatment. The two groups are followed up to compare differences in

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             225 of 307
     DRAFT FOR CONSULTATION


 1   outcomes to see how effective the experimental treatment was. Such trial
 2   designs help minimize experimental bias.
 3
 4   Rescue medication In this guideline, this is an outcome recorded by some
 5   studies. The rate of rescue medication use is the rate at which participants
 6   had to use a stronger medication (typically for analgesia).
 7
 8   Self-management A term used for aspects of RA care which a person can do
 9   for themselves with advice from the primary care team, eg GP, nurse,
10   physiotherapist, occupational therapist and information leaflets.
11
12   Sensitivity analysis A measure of the extent to which small changes in
13   parameters and variables affect a result calculated from them. In this
14   guideline, sensitivity analysis is used in health economic modelling.
15
16   Stakeholder Any national organisation, including patient and carers’ groups,
17   healthcare professionals and commercial companies with an interest in the
18   guideline under development.
19
20   Statistical significance A result is deemed statistically significant if the
21   probability of the result occurring by chance is less than 1 in 20 (p<0.05).
22
23   Systematic review Research that summarises the evidence on a clearly
24   formulated question according to a pre-defined protocol using systematic and
25   explicit methods to identify, select and appraise relevant studies, and to
26   extract, collate and report their findings. It may or may not use statistical meta-
27   analysis.
28
29   Technology appraisal Formal ascertainment and review of the evidence
30   surrounding a health technology, restricted in the current document to
31   appraisals undertaken by NICE.
32
33   Univariate analysis which separately explores each variable in a data set
34
35   Utility A number between 0 and 1 that can be assigned to a particular state of
36   health,assessing the holistic impact on quality of life and allowing states to be
37   ranked in order of (average) patient preference.
38




     Rheumatoid arthritis: full guideline DRAFT (August 2008)               226 of 307
    DRAFT FOR CONSULTATION




1                             APPENDICES




2




3

4




    Rheumatoid arthritis: full guideline DRAFT (August 2008)   227 of 307
    DRAFT FOR CONSULTATION




1   Appendix A
    Question Question wording                                            Study Type          Databases and
    ID                                                                   Filters used        Years
    REFER1     In adults with a recent onset of an undifferentiated All study types          Medline 1950 –
               inflammatory arthritis, what features of the clinical                         2008
               presentation should a non specialist recognise in order                       Embase 1980 –
               to refer to specialist services, and how quickly should                       2008
               the referral be made in order to minimise the impact of                       Cochrane 1800 –
               disease on symptoms, joint damage, function and                               2008
               quality of life?                                                              Cinahl 1982 –2008

    INVEST     In adults with a recent onset of an undifferentiated      Systematic          Medline 1950 –
               inflammatory arthritis, what are the investigative        Reviews             2008
               procedures that need to be performed to recognise the     Randomised          Embase 1980 –
               disease early and minimise the impact on symptoms,        Controlled trials   2008
               joint damage, function and quality of life?               Observational       Cochrane 1800 –
                                                                         studies             2008
                                                                         Diagnosis           Cinahl 1982 –2008

    PROG       In adults with a recent onset of rheumatoid arthritis,    Systematic          Medline 1950 –
               and in established disease, what features of the          Reviews             2008
               clinical presentation help to identify the prognosis of   Randomised          Embase 1980 –
               the disease, and should those patients with poor          Controlled trials   2008
               prognosis be treated differently from the rest of the     Observational       Cochrane 1800 –
               patient population in order to minimise the impact of     studies             2008
               disease on symptoms, joint damage, function and           Prognosis           Cinahl 1982 –2008
               quality of life?




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                228 of 307
DRAFT FOR CONSULTATION



Question Question wording                                             Study Type            Databases and
ID                                                                    Filters used          Years
PATIENT    In adults with established rheumatoid arthritis, what is   Systematic            Medline 1950 –
           known of patient experiences of rheumatoid arthritis       Reviews               2008
           and its treatments, and how do patient perceptions and     Randomised            Embase 1980 –
           beliefs influence their preferences and outcomes with      Controlled trials     2008
           respect to symptoms, joint damage, function and            Observational         Cochrane 1800 –
           quality of life?                                           studies               2008
                                                                      Qualitative           Cinahl 1982 –2008
                                                                      studies incl. focus   Psychinfo 1806-
                                                                      groups                2008

EDU        In adults with recent onset of rheumatoid arthritis, and All study types         Medline 1950 –
           in established disease, what are the relative benefits of                        2008
           different patient information provision, and/or                                  Embase 1980 –
           educational methods and/or different patient self                                2008
           management programs i) in relation to each other, ii)                            Cochrane 1800 –
           versus no specific information provision/education with                          2008
           respect to symptoms, joint damage, function and                                  Cinahl 1982 –2008
           quality of life?                                                                 Psychinfo 1806-
                                                                                            2008

ANALG      In adults with a recent onset of rheumatoid arthritis,     Systematic            Medline 1950 –
           and in established disease, what are the benefits and      Reviews               2008
           harms of using analgesics on symptoms, joint damage,       Randomised            Embase 1980 –
           function and quality of life?                              Controlled trials     2008
                                                                      Observational         Cochrane 1800 –
                                                                      studies               2008
                                                                                            Cinahl 1982 –2008




Rheumatoid arthritis: full guideline DRAFT (August 2008)                 229 of 307
DRAFT FOR CONSULTATION



Question Question wording                                           Study Type          Databases and
ID                                                                  Filters used        Years
NSAID      In adults with a recent onset of rheumatoid arthritis,   Systematic          Medline 1950 –
           and in established disease, what are the benefits and    Reviews             2008
           harms of using non-steroidals and Cox-2 selective        Randomised          Embase 1980 –
           drugs on symptoms, joint damage, function and quality    Controlled trials   2008
           of life?                                                                     Cochrane 1800 –
                                                                                        2008
                                                                                        Cinahl 1982 –2008
CORTICO    In adults with a recent onset of rheumatoid arthritis,   Systematic          Medline 1950 –
           and in established disease, what are the benefits and    Reviews             2008
           harms of using corticosteroids on symptoms, joint        Randomised          Embase 1980 –
           damage, function and quality of life?                    Controlled trials   2008
                                                                                        Cochrane 1800 –
                                                                                        2008
                                                                                        Cinahl 1982 –2008
DRUG1      In adults with a recent onset of rheumatoid arthritis,   Systematic          Medline 1950 –
           what sequence of single and combined disease-            Reviews             2008
           modifying and biological drugs will exert maximum        Randomised          Embase 1980 –
           impact on symptoms, joint damage, function and           Controlled trials   2008
           quality of life?                                         Observational       Cochrane 1800 –
                                                                    studies             2008
                                                                                        Cinahl 1982 –2008

DRUG2      In adults with established rheumatoid arthritis, are     Systematic          Medline 1950 –
           there any circumstances under which disease-             Reviews             2008
           modifying drugs and biologics can be decreased           Randomised          Embase 1980 –
           and/or withdrawn without jeopardising the impact of      Controlled trials   2008
           disease on symptoms, joint damage, function and                              Cochrane 1800 –
           quality of life?                                                             2008
                                                                                        Cinahl 1982 –2008




Rheumatoid arthritis: full guideline DRAFT (August 2008)               230 of 307
DRAFT FOR CONSULTATION



Question Question wording                                           Study Type          Databases and
ID                                                                  Filters used        Years
DRUG3      In adults with established rheumatoid arthritis, are     Systematic          Medline 1950 –
           biological drugs more effective than conventional        Reviews             2008
           disease-modifying drugs (singly, or in combination) in   Randomised          Embase 1980 –
           patients where there is ongoing disease activity with    Controlled trials   2008
           respect to symptoms, joint damage, function and                              Cochrane 1800 –
           quality of life?                                                             2008
                                                                                        Cinahl 1982 –2008

DMARD      In adults with a recent onset of rheumatoid arthritis,   Systematic          Medline 1950 –
           what are the benefits and harms of early introduction    Reviews             2008
           of disease-modifying drugs on symptoms, joint            Randomised          Embase 1980 –
           damage, function and quality of life?                    Controlled trials   2008
                                                                    Observational       Cochrane 1800 –
                                                                    studies             2008
                                                                                        Cinahl 1982 –2008
ANAKINR    In adults with established rheumatoid arthritis is Systematic                Medline 2002 –
A          Anakinra an effective drug with respect to symptoms, Reviews                 2008
           joint damage, function and quality of life?            Randomised            Embase 2002 –
                                                                  Controlled trials     2008
                                                                  Observational         Cochrane 2002 –
                                                                  studies               2008
                                                                  Risk                  Cinahl 2002 –2008
MULTI      In adults with a recent onset of rheumatoid arthritis, All study types       Medline 1950 –
           what are the benefits and harms of multidisciplinary                         2008
           teams in order to minimise symptoms, joint damage,                           Embase 1980 –
           function and quality of life?                                                2008
                                                                                        Cochrane 1800 –
                                                                                        2008
                                                                                        Cinahl 1982 –2008




Rheumatoid arthritis: full guideline DRAFT (August 2008)               231 of 307
DRAFT FOR CONSULTATION



Question Question wording                                            Study Type          Databases and
ID                                                                   Filters used        Years
POD        In adults with a recent onset of rheumatoid arthritis,    Systematic          Medline 1950 –
           and in established disease, what aspects of podiatry      Reviews             2008
           can minimise the impact of disease on symptoms, joint     Randomised          Embase 1980 –
           damage, function and quality of life in the following     Controlled trials   2008
           groups?                                                   Observational       Cochrane 1800 –
                                                                     studies             2008
                                                                                         Cinahl 1982 –2008

DIET       In adults with established rheumatoid arthritis, what     Systematic          Medline 1950 –
           aspects of diet have evidence that they influence         Reviews             2008
           symptoms, joint damage, function and quality of life      Randomised          Embase 1980 –
           positively?                                               Controlled trials   2008
                                                                     Observational       Cochrane 1800 –
                                                                     studies             2008
                                                                                         Cinahl 1982 –2008
                                                                                         AMED 1985-2008
PHYSIO     In adults with a recent onset of rheumatoid arthritis,    Systematic          Medline 1950 –
           and in established disease, what aspects of               Reviews             2008
           physiotherapy can minimise the impact of disease on       Randomised          Embase 1980 –
           symptoms, joint damage, function and quality of life in   Controlled trials   2008
           the following groups?                                     Observational       Cochrane 1800 –
                                                                     studies             2008
                                                                                         Cinahl 1982 –2008




Rheumatoid arthritis: full guideline DRAFT (August 2008)                232 of 307
DRAFT FOR CONSULTATION



Question Question wording                                            Study Type          Databases and
ID                                                                   Filters used        Years
OCCU       In adults with a recent onset of rheumatoid arthritis,    Systematic          Medline 1950 –
           and in established disease,, what aspects of              Reviews             2008
           occupational therapy can minimise the impact of           Randomised          Embase 1980 –
           disease on symptoms, joint damage, function and           Controlled trials   2008
           quality of life in the following groups?                  Observational       Cochrane 1800 –
                                                                     studies             2008
                                                                                         Cinahl 1982 –2008

CAM        In adults with established rheumatoid arthritis, what     All study types     Medline 1950 –
           aspects of complementary, alternative and other non-                          2008
           pharmacological interventions (acupuncture, copper                            Embase 1980 –
           bracelets, aromatherapy and massage) have evidence                            2008
           that they influence symptoms, joint damage, function                          Cochrane 1800 –
           and quality of life positively?                                               2008
                                                                                         Cinahl 1982 –2008
                                                                                         AMED 1985-2008
MONIT      In adults with a recent onset of rheumatoid arthritis,    Systematic          Medline 1950 –
           and in established disease, what are the most effective   Reviews             2008
           methods to monitor the ongoing activity of the disease    Randomised          Embase 1980 –
           in order to minimise the impact of the disease on         Controlled trials   2008
           symptoms, joint damage, function and quality of life?     Observational       Cochrane 1800 –
                                                                     studies             2008
                                                                                         Cinahl 1982 –2008




Rheumatoid arthritis: full guideline DRAFT (August 2008)                233 of 307
    DRAFT FOR CONSULTATION



    Question Question wording                                            Study Type          Databases and
    ID                                                                   Filters used        Years
    REVIEW     In adults with established rheumatoid arthritis, what     Systematic          Medline 1950 –
               should be the content of a regular review in terms of:    Reviews             2008
               a) disease manifestations related to rheumatoid           Randomised          Embase 1980 –
               arthritis(e.g. extra-articular manifestations, cervical   Controlled trials   2008
               spine disease, secondary osteoarthritis, lymphoma,        Observational       Cochrane 1800 –
               amyloidosis)                                              studies             2008
               b) co-morbidities associated with rheumatoid arthritis    Risk                Cinahl 1982 –2008
               (e.g. cardiovascular disease, osteoporosis,
               depression), and for each component, what should be
               the frequency of review in order to minimise the impact
               of disease on symptoms, joint damage, function and
               quality of life?
    REFER2     In adults with established rheumatoid arthritis, what     All study types     Medline 1950 –
               factors should determine the timing of referral for                           2008
               surgery in order to minimise the impact of disease on                         Embase 1980 –
               symptoms, joint damage, function and quality of life?                         2008
                                                                                             Cochrane 1800 –
                                                                                             2008
                                                                                             Cinahl 1982 –2008

1
2   The final cut-off date for all searches was 13 June 2008




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                234 of 307
     DRAFT FOR CONSULTATION


1    Appendix B: SCOPE
2    Guideline title
3        • Rheumatoid arthritis: the management and treatment of rheumatoid
4            arthritis in adults
5    Short title
6        • Rheumatoid arthritis
 7   Background
 8   The National Institute for Health and Clinical Excellence (‘NICE’ or ‘the
 9   Institute’) has commissioned the National Collaborating Centre for Chronic
10   Conditions to develop a clinical guideline on the management and treatment
11   of rheumatoid arthritis in adults for use in the NHS in England and Wales. This
12   follows referral of the topic by the Department of Health (see appendix). The
13   guideline will provide recommendations for good practice that are based on
14   the best available evidence of clinical and cost effectiveness.
15
16   The Institute’s clinical guidelines will support the implementation of National
17   Service Frameworks (NSFs) in those aspects of care where a Framework has
18   been published. The statements in each NSF reflect the evidence that was
19   used at the time the Framework was prepared. The clinical guidelines and
20   technology appraisals published by the Institute after an NSF has been issued
21   will have the effect of updating the Framework.
22
23   NICE clinical guidelines support the role of healthcare professionals in
24   providing care in partnership with patients, taking account of their individual
25   needs and preferences, and ensuring that patients (and their carers and
26   families, where appropriate) can make informed decisions about their care
27   and treatment.
28
29   Clinical need for the guideline
30   Rheumatoid arthritis is a chronic, disabling autoimmune disease characterised
31   by inflammation of the synovial tissue of the peripheral joints, which causes
32   swelling, stiffness, pain and progressive joint destruction. For a small
33   proportion of patients, inflammatory disease outside the joints (for example,
34   eye and lung disease, vasculitis) can pose a significant problem.
35
36   Rheumatoid arthritis affects 1% of the population, or approximately 400,000
37   people in England and Wales. Of these, approximately 15% have severe
38   disease. Rheumatoid arthritis is three times more prevalent in women than in
39   men. The condition impacts severely on quality of life and is a considerable
40   economic burden. It is estimated that 40% of people with rheumatoid arthritis
41   will be unable to work within 5 years of diagnosis, and 50% within 10 years.
42   The life expectancy of people with rheumatoid arthritis is reduced by 5–10
43   years compared with that of people without the condition, and 35–50% of this
44   excess risk is accounted for by cardiovascular mortality.
45
46   A range of lifestyle, pharmacological, non-pharmacological, surgical and
47   rehabilitation interventions can help to decrease inflammation, limit damage

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               235 of 307
     DRAFT FOR CONSULTATION


 1   and improve function for people with rheumatoid arthritis, thereby maintaining
 2   or improving their quality of life.
 3
 4   The guideline
 5   The guideline development process is described in detail in two publications
 6   that are available from the NICE website (see ‘Further information’). ‘The
 7   guideline development process: an overview for stakeholders, the public and
 8   the NHS’ describes how organisations can become involved in the
 9   development of a guideline. ‘The guidelines manual’ provides advice on the
10   technical aspects of guideline development.
11
12   This document is the scope. It defines exactly what this guideline will (and will
13   not) examine, and what the guideline developers will consider. The scope is
14   based on the referral from the Department of Health (see appendix).
15
16   The areas that will be addressed by the guideline are described in the
17   following sections.
18
19   Population
20   Groups that will be covered
21       • Adults with rheumatoid arthritis.
22   Groups that will not be covered
23       • Patients with other chronic inflammatory polyarthritis.
24   Healthcare setting
25        • Primary care, secondary care and intermediate care.
26   Clinical management
27        • Identification of the prognostic factors that indicate patients at
28            greatest risk of persistent, damaging erosive and progressive
29            disease.
30        •   Pharmacological treatments for managing the condition including:
31              • cyclo-oxygenase (Cox) II inhibitors
32              • non-steroidal anti-inflammatory drugs (NSAIDS)
33              • corticosteroids
34              • disease-modifying anti-rheumatic drugs (DMARDS)
35              • biological drugs.
36        •   The guideline will attempt to identify the optimal sequencing of the
37            effective pharmacological agents.
38        •   (Note that guideline recommendations will normally fall within
39            licensed indications; exceptionally, and only where clearly supported
40            by evidence, use outside a licensed indication may be recommended.
41            The guideline will assume that prescribers will use a drug’s summary
42            of product characteristics to inform their decisions for individual
43            patients.)



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              236 of 307
     DRAFT FOR CONSULTATION


1         •   Non-pharmacological treatments relevant to rheumatoid arthritis, for
2             example:
3               • orthoses
4               • podiatry
5               • diet
6               • physiotherapy
7               • occupational therapy.
8         •   Pain management.
9         •   Timing of referral for surgery.
10        •   Support for patients and carers in managing rheumatoid arthritis,
11            through education, self management and the provision of information
12            and advice.
13        •   Principal complementary and alternative interventions or approaches
14            to care relevant to the guideline topic will be considered.
15        •   Where ineffective interventions and approaches to care are identified
16            and where robust and credible recommendations for re-positioning
17            the intervention for optimal use, or changing the approach to care to
18            make more efficient use of resources, can be made, they will be
19            clearly stated. If the resources released are substantial, consideration
20            will be given to listing such recommendations in the ‘Key priorities for
21            implementation’ section of the guideline.
22   Status
23   Scope
24       • This is the consultation draft of the scope. The consultation period is
25         14 November to 13 December 2006.
26        •   The guideline will update the following NICE technology appraisal
27            guidance, and will supersede the appraisals when it is published, with
28            regard to rheumatoid arthritis.
29             • Anakinra for rheumatoid arthritis. NICE technology appraisal
30                guidance 72 (2003). Available from: www.nice.org.uk/TA072
31             • Osteoarthritis and rheumatoid arthritis - cox II inhibitors: NICE
32                technology appraisal guidance 27 (2001). Available from:
33                www.nice.org.uk/TA027
34        •   Related NICE Technology Appraisals.
35              • Guidance on the use of etanercept and infliximab for the
36                treatment of rheumatoid arthritis. NICE technology appraisal
37                guidance 36 (2002). Available from: www.nice.org.uk/TA036
38                (review in progress, which will include adalimumab; expected
39                date of publication December 2007).
40              • Rituximab for the treatment of rheumatoid arthritis (publication
41                date to be confirmed).



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              237 of 307
     DRAFT FOR CONSULTATION


1              • Abatacept for the treatment of rheumatoid arthritis (publication
2                 date to be confirmed).
3              • Certolizumab pegol for rheumatoid arthritis (topic to be
4                 confirmed).
5         •   Related NICE clinical guidelines:
6              • Osteoarthritis: the care and management of adults with
7                 osteoarthritis (expected date of publication December 2007)
8         •   Related NICE interventional procedures :
 9             • Artificial metacarpophalangeal and interphalangeal joint
10               replacement for end-stage arthritis
11   Guideline
12       • The development of the guideline recommendations will begin in
13           June 2007.
14   Further information
15   Information on the guideline development process is provided in:
16        • ‘The guideline development process: an overview for stakeholders,
17            the public and the NHS’
18        •   The guidelines manual
19   These booklets are available as PDF files from the NICE website
20   (www.nice.org.uk/guidelinesmanual). Information on the progress of the
21   guideline will also be available from the website.
22
23
24




     Rheumatoid arthritis: full guideline DRAFT (August 2008)            238 of 307
     DRAFT FOR CONSULTATION



1    Appendix C
2
3

4

5

6

7          Report for the NCC-CC Clinical
8         Guideline for Rheumatoid Arthritis
9
10      Disease-Modifying Anti-Rheumatic Drug
11        Combinations in Early Rheumatoid
12      Arthritis: A Cost-Effectiveness Analysis
13
14
15
16
17




     Rheumatoid arthritis: full guideline DRAFT (August 2008)   239 of 307
     DRAFT FOR CONSULTATION



1    A.1. Introduction
 2   This cost-utility analysis evaluates the cost-effectiveness of combinations of
 3   disease-modifying anti-rheumatic drugs in the treatment of patients with early
 4   rheumatoid arthritis. The model also evaluates the cost-effectiveness of using
 5   glucocorticoids alongside DMARD monotherapy in patients with early
 6   rheumatoid arthritis.
 7
 8   This report details the drug combinations investigated, the parameters
 9   included within the model, and the structure of the model. The results
10   provided by the model are presented and discussed.
11

12   A.2. Background

13   Rheumatoid Arthritis (RA) is a chronic autoimmune inflammatory disorder
14   affecting approximately 0.8% of the adult UK population5. Being a progressive
15   disease and having a potentially early onset in a patient’s lifetime leads to
16   considerable societal and economic impact. Traditionally, patients have been
17   treated using Disease-Modifying Anti-Rheumatic Drugs (DMARDs), anti-
18   inflammatories, glucocorticoids, and analgesics. These treatments are
19   relatively cheap, with drug costs representing about 13-15% of the total direct
20   costs of treatment for RA401. However newer, more effective, more expensive
21   and more tolerable anti-TNF (biologic) drugs have been produced which may
22   in time see the movement of more patients to more expensive treatments. As
23   per current NICE guidance,402 anti-TNF drugs are not available for patients
24   until they have failed two traditional DMARDs (one being Methotrexate), and
25   so the use of DMARDS in early RA has a significant impact on both the future
26   costs and future benefits accrued.
27
28   A large number of trials over the last 10-15 years have suggested an
29   increased benefit in patients taking DMARDs in combination, instead of
30   sequentially with the aim of rapidly bringing a patient’s disease under control.
31   Traditional approaches to treating RA have focused on drug escalation, with
32   DMARDs and steroids introduced sequentially and dosages titrated upwards.
33   Whilst this traditional approach can see patients achieve disease-control,
34   there is now much support to believe, both in clinical trials and clinical
35   practice, that a ‘slow’ sequential monotherapy results in unsatisfactory long-
36   term results as slow treatment and a failure to control a patients’ disease from
37   the outset can see poorer symptom control and irreversible damage to a
38   patients joints.278 Combinations of DMARDs and rapid changing of ineffective
39   DMARDs will hope to bring a patients disease under control much more
40   quickly.
41
42   It is therefore important to evaluate the cost-effectiveness of these
43   combinations of DMARDs as they will impact on the lifetime costs and
44   benefits for the patient, especially as patients may fail to respond to two
45   DMARDs more quickly than if they were taken sequentially and therefore
46   move to more expensive biologic therapies earlier.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             240 of 307
     DRAFT FOR CONSULTATION



1    A.3. Comparator Treatments
 2   This analysis compares a range of combination DMARD therapies clinically
 3   evaluated in published studies against DMARD monotherapy, which is
 4   assumed to be standard clinical practice. The analysis also compares the use
 5   of glucocorticoids in combination with a DMARD, against DMARD
 6   monotherapy. The analysis will group the combinations by their general trend,
 7   and so the focus of the cost-effectiveness analysis will be comparing types of
 8   combination strategies, instead of modelling the numerous trials, many of
 9   which have similar protocols. This is based on the clinical opinion that there
10   are no statistically significant differences in the efficacy of specific DMARDs.
11   This allows for a full cost-effectiveness analysis with comparison to DMARD
12   monotherapy as well as all alternative combination strategies.
13
14   Trials
15   A systematic review of the clinical and cost-effectiveness evidence was
16   performed as part of the Guideline Development The clinical searches were
17   broadened to allow for evidence from non-UK populations and small studies.
18   The defined aim of the review was to look for studies (see table 1 below) that
19   “…investigated the efficacy and safety of early introduction of DMARDs with
20   respect to symptoms, joint damage, function and quality of life in patients with
21   a recent onset of RA.” Excluded were trials that were not a meta-analysis,
22   randomised controlled trial, cohort study or case-control study and also trials
23   that did not report HAQ disease levels and ACR 20 or 50 response rates.
24
25   Table 1 - Combination DMARD Trials
     Name                  Arms                         Reference
         304
     BEST                  1. Monotherapy               Goekoop-Ruiterman YP, de Vries-Bouwstra JK, Allaart CF, et al.
                           2. Step Up Combination       Clinical and radiographic outcomes of four different treatment
                           (Combo)                      strategies in patients with early rheumatoid arthritis (the BeSt
                           3. Parallel Combo            study): a randomized, controlled trial. Arthritis Rheum
                                                        2005;52:3381-90.
               288         1.   Routine Step Up         Grigor C, Capell H, Stirling A et al. Effect of a treatment strategy of
     TICORA
                           2.   Intensive Step Up       tight control for rheumatoid arthritis (the TICORA study): a single-
                                                        blind randomised controlled trial. Lancet 2004; 364(9430):263-9.
                403
     TICORA2               1.   Intensive Step Up       Saunders SA, Capell HA, Stirling A et al. Triple Therapy in Early
                           2.   Triple Parallel Combo   Active Rheumatoid Arthritis. A randomized, single-blind, controlled
                                                        trial comparing step-up and parallel treatment strategies. Arthritis &
                                                        Rheumatism 2008; 58(5) May 1310-1317
                     404
     Dougados              1.   Sulphasalazine (SSZ)    Dougados M, Combe B, Cantagrel A et al. Combination therapy in
                                monotherapy             early rheumatoid arthritis: a randomised, controlled, double blind
                           2.   Methotrexate (MTX)      52 week clinical trial of sulphasalazine and methotrexate compared
                                monotherapy             with the single components. Annals of the Rheumatic Diseases
                           3.   Parallel Combo          1999; 58; 220-225
               405
     Gerards               1.   Monotherapy             Gerards AH, Landewe RBM, Prins APA et al. Cyclosporin A
                           2.   Parallel Combo          monotherapy versus cyclosporin A and methotrexate combination
                                                        therapy in patients with early rheumatoid arthritis: a double blind
                                                        randomised placebo controlled trial. Annals of the Rheumatic
                                                        Diseases 2003; 62; 291-296
                     292
     CIMESTRA              1.   Steroid + Monotherapy   Hetland ML, Stengaard-Pedersen K, Junker P et al. Combination
                           2.   Parallel Combo          treatment with methotrexate, cyclosporine, and intraarticular
                                                        betamethasone compared with methotrexate and intraarticular
                                                        betamethasone in early active rheumatoid arthritis: an investigator-
                                                        initiated, multicenter, randomized, double-blind, parallel-group,
                                                        placebo-controlled study. Arthritis Rheum 2006; 54(5):1401-9.
                     84
     Proudman              1.   Monotherapy             Proudman SM, Conaghan PG, Richardson C et al. Treatment of
                           2.   Parallel Combo          poor-prognosis early rheumatoid arthritis. Arthritis & Rheumatism.
                                                        2000, 43(8) 1809-1819
               406         1.   Monotherapy             Miranda JM, Alvarex-Nemegyei J, Saavedra MA et al. A
     Miranda
                           2.   Parallel Combo          Randomized, Double-Blind, Multicenter, Controlled Clinical Trial of


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                                 241 of 307
     DRAFT FOR CONSULTATION

                                                            Cyclosporine Plus Chloroquine vs Cyclosporine Plus Placebo in
                                                            Early- Onset Rheumatoid Arthritis. Archives of Medical Research
                                                            35 (2004) 36-42
                       407
     Sarzi-Puttini           1. Monotherapy                 Sarzi-Puttini P, D’Ingianna E, Fumagalli M et al. An open,
                             2. Parallel Combo              randomized comparison study of cyclosporine A, cyclosporine A +
                             (Hydroxychloroquine)           methotrexate and cyclosporine A + Hydroxychloroquine in the
                             3. Parallel Combo (MTX)        treatment of early severe rheumatoid arthritis. Rheumatology
                                                            International (2005) 25; 15-22
                   408       1.    Monotherapy              Marchesoni A, Battafarano N, Arreghini M et al. Radiographic
     Marchesoni
                             2.    Parallel Combo           progression in early rheumatoid arthritis: a 12-month randomized
                                                            controlled study comparing the combination of cyclosporin and
                                                            methotrexate with methotrexate alone. Rheumatology 2003; 42;
                                                            1545-1549
     Van den                 1. Monotherapy                 Van den Borne BEEM, Landewe RBM, Goeithe HS et al.
           409
     Borne                   2. Parallel Combo (low         Combination Therapy in Recent Onset Rheumatoid Arthritis: A
                             dose)                          Randomized Double Blind Trial of the Addition of Low Dose
                             3. Parallel Combo (high        Cyclosporine to Patients Treated with Low Dose Chloroquine. The
                             dose)                          Journal of Rheumatology, 1998, 25(8) 1493-1498
                 296
     Fin-RACo                1. Steroid + Monotherapy       Möttönen T, Hannonen P, Leirisalo-Repo M et al. Comparison of
                             2. Step Down Combo             combination therapy with single-drug therapy in early rheumatoid
                                                            arthritis: a randomised trial. FIN-RACo trial group. Lancet 1999;
                                                            353:1568–73.
             299
     COBRA                   1.    Monotherapy              Boers M, Verhoeven AC, Markusse HM et al. Randomised
                             2.    Step Down Combo          comparison of combined step-down prednisolone, methotrexate
                                                            and sulphasalazine with sulphasalazine alone in early rheumatoid
                                                            arthritis. Lancet 1997; 350:309–18.
1

2    A.3.1 Treatment Regimens
 3   The range of treatment regimens, with different dosages, specific drugs and
 4   overall ‘styles’ for adding on or removing drugs from the regimen means that
 5   modelling each trial separately would be unnecessary in the context of the
 6   policy decisions being made. The majority of the trials are relatively small
 7   (n<100) and so it is unlikely that one trial will provide the weight to make a
 8   confident conclusion about a specific DMARD regimen. To provide a useful
 9   model without over-simplifying the complex nature and treatment of RA has
10   been attempted by grouping together the trials into groups of similar treatment
11   regimens. Many trials provide very similar strategies and so combining
12   formally combining them in a meta-analysis will provide more weight to the
13   conclusions (see table 2). The methods used to combine trials into strategies
14   and how comparison is drawn between different groups is explained in detail
15   later in the report under “6-month Response”.
16
17   Table 2- Combination Strategy Groups
     Treatment Group              Explanation                                      Trial Arms Included
     Control Arm                  DMARD monotherapy (First line Methotrexate       Dougados – SSZ
                                  15mg/week, Second line Sulphasalazine 1g/day)    Dougados – MTX
                                                                                   Capell
                                                                                   Van den Borne
                                                                                   BEST
                                                                                   Gerards
                                                                                   Proudman
                                                                                   Miranda
                                                                                   Sarzi-Puttini
                                                                                   Marchesoni
                                                                                   COBRA
     Step Up                      One or more DMARDs are added (stepped-up)        BEST – Step Up Arm
     Combination                  on top of a previous monotherapy regimen         TICORA – Routine Arm
     Parallel                     Two or more DMARDs are given in combination      Dougados
     Combination                                                                   BEST – Parallel Arm
                                                                                   Gerards
                                                                                   CIMESTRA
                                                                                   Proudman
                                                                                   Miranda


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                                 242 of 307
DRAFT FOR CONSULTATION

                                                                Sarzi-Puttini CSA + HCQ
                                                                Sarzi-Puttini CSA + MTX
                                                                Marchesoni
                                                                Van den Borne (two arms averaged)
                                                                TICORA 2 – Triple Therapy
Step Down        Initial parallel combination before specific   COBRA
Combination      DMARDs are removed                             Fin-RACo
Intensive Step   A parallel strategy where dosages of DMARDs    TICORA – Intensive Arm
Up Combination   are rapidly increased to above their           TICORA2 – Intensive Arm
                 recommended BNF dose
Steroid +        Where glucocorticoids are used alongside a     Capell
Monotherapy      DMARD monotherapy regimen                      Fin-RACo Monotherapy Arm
                                                                CIMESTRA




Rheumatoid arthritis: full guideline DRAFT (August 2008)                                 243 of 307
     DRAFT FOR CONSULTATION



1    A.4. Model Structure
 2   The economic model is in the form of a Patient Level Simulation - that is it
 3   models the course of disease over time with changes to important variables at
 4   time-points where events (starting treatment, withdrawing from treatment,
 5   death) occur. An essential part of the model is that it tracks individual patients
 6   one at a time. The time at which events occur and the results of those events
 7   is dependent on the patients past history and a range of characteristics
 8   specific to the individual.
 9
10   Individual patients are followed from the time of starting a first-line DMARD
11   strategy until death, with any changes calculated every six months. Unlike
12   cohort model approaches, where a population is tracked through a model,
13   individuals are modelled one at a time. At every 6-month point the route taken
14   by that individual patient is determined by a random number generator and
15   the assigned probability of each event.
16
17   A sufficient number of hypothetical patients are sent through the model to give
18   overall precision to the estimates of mean cost and utility. The expected costs
19   and benefits of each comparison are calculated, and this allows the
20   assessment of the overall cost-effectiveness of each strategy in terms of cost
21   per QALY.
22

23   A.4.1 Duration
24   The model has a lifetime duration, to allow all possible costs and benefits that
25   are accrued to be captured by the model. With RA being a lifelong chronic
26   condition, and evidence showing early treatment having a lifelong impact on
27   the future progression of the disease278 it is important that the model
28   aggregates costs and benefits accrued in a patient’s lifetime. For
29   computational purposes, the lifetime duration is assumed to be fifty years from
30   the disease onset, as the mean patient age is 54.8 years (SD = 13.6).
31
32   In the model, HAQ and cost profile are calculated at 6-month time points
33   (cycle-length). Initial response to a treatment is assumed to occur in the first
34   time period (6 months). Depending on the age of the patient, the model ‘runs’
35   through the patients lifetime to calculate their expected costs and utility scores
36   for each remaining cycle (calculated by life tables).
37

38   A.4.2 Clinical Pathway
39   The model tracks patients over 6-monthly intervals from the initial decision to
40   start a patient on a specific DMARD combination strategy or monotherapy,
41   until the end of their lifetime. As the patients who enter the model are DMARD
42   and biologic naïve, the model is actually tracking patients with RA through the
43   whole of their medical treatment for RA. The model runs each hypothetical
44   patient through all arms (all combinations and monotherapy). Figure 1 is a
45   decision algorithm which illustrates the model.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               244 of 307
     DRAFT FOR CONSULTATION


 1
 2   Patients who begin a new treatment are then simulated as to whether they are
 3   an ACR20, 50 or a non-responder. This initial response to treatment is
 4   mapped to the patients HAQ profile, which is converted to a utility estimate.
 5   Patients who do not achieve an ACR20 or 50 response at 6 months progress
 6   to their next treatment (2nd DMARD monotherapy or biologics). Patients who
 7   do achieve an ACR20 or 50 response at 6 months are then simulated through
 8   a period on treatment, which is determined by the 6-month withdrawal rate of
 9   that strategy due to a loss of efficacy or an adverse event. Patients who are
10   on treatment for longer than 6 months see a gradual worsening of their HAQ
11   until they are withdrawn from treatment, at which time their HAQ rebounds by
12   the same amount as the initial gain at 6 months. These fundamental
13   assumptions are discussed in more detail later in the report, and investigated
14   in the sensitivity analysis.
15
16   Once a patient has been withdrawn from either their 2nd monotherapy DMARD
17   or their initial combination, they progress to biologic therapies, which are
18   modelled as assigned lifetime costs and QALYs from an existing model of
19   biologics. For further detail about how this is modelled, please see the section
20   on “Biologics” later in this report.
21
22   The level of a patient’s disease in the model is measured in their Health
23   Assessment Questionnaire (HAQ) score, for more details on how this, see the
24   “HAQ” subsection in the Model Inputs chapter.
25

                                                       New Treatment




           Utility (DMARDs)
                                                Simulate 6 month ACR
           Function of:
                                                      response
           •   ACR response
           •   HAQ progression
                                                                              ACR 20 or
                                                                             50 response
                                            No ACR
                                            response


                                                            Continue treatment
                                                           until lack of efficacy or
                                                                adverse event




                                                                        No
                                                         Eligible for
                                                         Biologics?
           Utility (Biologics)
                                                  Yes
           •    QALYs assigned
                from BSRBR                     Biologics (Brennan Model)
                Model

26

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                              245 of 307
     DRAFT FOR CONSULTATION

1    Figure 1 - Clinical Algorithm for DMARD Treatments in Model

2    A.4.3 Limitations
 3
 4   Structurally, the model has a number of notable limitations. The first is that
 5   although the model is a patient-level simulation, the clinical data is based on
 6   clinical trials, not a registry and so patient covariates are not included to
 7   determine differences in clinical response or treatment withdrawal. Secondly
 8   the NICE guidance on biologic therapies specifies that patients only progress
 9   to biologics after failing two DMARDs (one being Methotrexate and both tried
10   for at least 6-months) and a patient having a Disease Activity Score (DAS) of
11   more than 5.1. As the model is HAQ based, and there is no available
12   conversion between HAQ and DAS, this requirement was left out of the model
13   and it was assumed that patients who had failed two DMARDs would be
14   eligible for biologics.
15
16
17




     Rheumatoid arthritis: full guideline DRAFT (August 2008)            246 of 307
     DRAFT FOR CONSULTATION



1    A.5. Model Inputs

2    A.5.1 6-month Response
 3   As explained in the clinical pathways section, the model relies on clinical trial
 4   data for the initial probability of an ACR 20 or 50 response at 6-months, and
 5   the probability of a patient withdrawing at 6-months for a loss of efficacy or an
 6   adverse event. Because no one trial exists that compares all the different
 7   combinations of interest, ACR response data has to be taken from a number
 8   of trials, which each have different trial and patient baseline characteristics
 9   and compare combination DMARDs to different control arms.
10
11   To account for this range of direct and indirect evidence, it is synthesised in a
12   mixed treatment comparison410,411 (also known as a network meta-
13   analysis412). This method allows the different strategies to be compared to
14   each other, including where no primary trials have made such a comparison.
15   A total of thirteen trials comparing six treatment strategies were included in
16   the analysis. Not all the studies reported both ACR20 and 50, and for one
17   study304 the authors were contacted directly for ACR50 response data,
18   because the study publications reported that this had been assessed within
19   the trial but had not been reported. There are 15 pairwise comparisons
20   between the 6 treatment strategies. Only one comparison, between
21   monotherapy and parallel combination, had more than a single study
22   providing direct evidence. Eight studies provided ACR20 data for this
23   comparison and 6 reported ACR50 data. A standard random effects meta
24   analysis of these data for comparison with the mixed treatment model was
25   undertaken. 6 of the remaining pairwise comparisons are made directly by the
26   trials identified for ACR20. 7 of these comparisons were made for ACR50
27   (See Figure 2).
28
29   Table 3 - ACR Response Data
     Strategy                ACR20                    ACR50           Trial                    6-months
                     Log Odds        SD      Log Odds         SD                               or 1 year
                     Ratio                   Ratio                                             ACR
     Monotherapy     -               -       -                -       Dougados – SSZ           1 year
                                                                      Dougados – MTX           1 year
                                                                      Capell                   6 month
                                                                      Van den Borne            24 weeks
                                                                                                       ‡‡‡
                                                                      BEST                     6 month
                                                                      Gerards                  48 weeks
                                                                      Proudman                 48 weeks
                                                                      Miranda                  1 year
                                                                      Sarzi-Puttini            1 year
                                                                      Marchesoni               1 year
                                                                      COBRA                    28 weeks
     Step-up         0.010           0.232   -0.318           0.862   BEST – Step Up Arm       6 month
                                                                      TICORA – Routine Arm     1 year
     Parallel        0.568           0.568   0.890            0.358   Dougados                 1 year
                                                                      BEST – Parallel Arm      6 month
                                                                      Gerards                  48 weeks
                                                                      CIMESTRA                 6 month
                                                                      Proudman                 48 weeks
                                                                      Miranda                  1 year

     ‡‡‡
           Data provided through correspondence with author

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                247 of 307
     DRAFT FOR CONSULTATION


                                                         Sarzi-Puttini CSA + HCQ               1 year
                                                         Sarzi-Puttini CSA + MTX               1 year
                                                         Marchesoni                            1 year
                                                         Van den Borne (two arms averaged)     24 weeks
                                                         TICORA 2 – Triple Therapy             1 year
     Intensive     1.197        0.612   1.501    1.287   TICORA – Intensive Arm                1 year
                                                         TICORA2 – Intensive Arm               1 year
     Step-down     1.003        0.344   0.951    0.709   COBRA                                 28 weeks
                                                         Fin-RACo                              6 month
     Steroid       0.136        0.248   0.280    0.749   Capell                                6 month
                                                         Fin-RACo Monotherapy Arm              6 month
                                                         CIMESTRA                              6 month
1
2    The log odds ratios are relative to a probability of ACR20 response in the
3    monotherapy arm of 0.502, and an ACR50 response of 0.315. These values
4    are taken from a weighted average of the trials arms reporting a DMARD
5    monotherapy arm. In the model, the log odds ratios were parameterised with
6    normal distributions and the associated standard deviations. The log odds
7    ratios were converted back to probabilities and sampled for the patient level
8    simulation (see table 4).
9
10   Table 4 - Pairwise odds ratios between six DMARD treatment strategies
11   obtained by random effects mixed treatment comparisons
      Treatment comparison                                       Mixed treatment comparison
      Parallel versus monotherapy               ACR20         1.72    (1.11 to 2.51)
                                                ACR50         2.50    (1.14 to 4.80)
      Step up versus monotherapy                ACR20         1.55    (0.46 to 3.96)
                                                ACR50         1.39    (0.16 to 5.30)
      Step down versus monotherapy              ACR20         3.25    (0.84 to 8.87)
                                                ACR50         3.34    (0.59 to 10.20)
      Parallel intensive versus monotherapy     ACR20        13.02    (1.35 to 50.01)
                                                ACR50        25.57    (0.55 to 105.40)
      Steroid versus monotherapy                ACR20         1.23    (0.47 to 2.53)
                                                ACR50         1.72    (0.27 to 5.68)
      Step up versus parallel                   ACR20         0.93    (0.26 to 2.53)
                                                ACR50         0.65    (0.06 to 2.58)
      Step down versus parallel                 ACR20         1.98    (0.48 to 5.73)
                                                ACR50         1.49    (0.24 to 4.75)
      Parallel intensive versus parallel        ACR20         7.87    (0.80 to 31.23)
                                                ACR50        15.40    (0.22 to 48.71)
      Steroid versus parallel                   ACR20         0.73    (0.28 to 1.54)
                                                ACR50         0.72    (0.12 to 2.39)
      Step down versus step up                  ACR20         2.87    (0.40 to 9.92)
                                                ACR50         5.79    (0.27 to 26.65)
      Parallel intensive versus step up         ACR20         8.28    (1.51 to 27.42)
                                                ACR50        13.14    (1.24 to 52.97)
      Steroid versus step up                    ACR20         1.07    (0.20 to 3.09)
                                                ACR50         3.14    (0.14 to 14.60)
      Parallel intensive versus step down       ACR20         5.94    (0.36 to 26.85)
                                                ACR50        14.85    (0.15 to 63.79)
      Steroid versus step down                  ACR20         0.57    (0.09 to 1.76)
                                                ACR50         0.65    (0.12 to 2.13)
      Steroid versus parallel intensive         ACR20         0.23    (0.02 to 0.95)
                                                ACR50         0.75    (0.01 to 3.40)


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                248 of 307
     DRAFT FOR CONSULTATION

1    Note: figures in brackets are 95% credible intervals.
2

          ACR 20 Response
                                 Hetland                      BeSt
                                                   Parallel
                                                                           Step-up
                  Steroid

                                               8 Trial
                                               Arms             TICORA2
                              Capell

                                                                                TICORA


                               COBRA          Monotherapy


                                                                          Intensive
               Step-down                                                   Step-up




          ACR 50 Response
                                 Hetland                      BeSt
                                                   Parallel
                                                                           Step-up
                  Steroid

                                               6 Trial
                                               Arms             TICORA2
                              Capell

            FinRACo                                                             TICORA


                               COBRA          Monotherapy


                                                                      Intensive
               Step-down                                               Step-up



 3
 4   Figure 2 - Mixed Treatment Comparison
 5
 6   A limitation of the data available is that thirteen arms provide ACR20 and 50
 7   response data at 6 months, with the remaining eighteen (of 31) arms
 8   providing ACR20 and 50 response data at 1 year. To allow for the events to
 9   occur after 6 months, which is the minimum time a 2nd DMARD can be
10   attempted before moving to biologics, along with the clinical assumption that
11   the vast majority of clinical response to a DMARD will be seen in the first 6
12   months, 1 year data was used to populate holes where 6 months data was not
13   available.
14


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                    249 of 307
     DRAFT FOR CONSULTATION


1    A.5.2   Withdrawal
 2   The 6-month withdrawal rate was estimated (see table 5) for each treatment
 3   strategy by taking a weighted average of the patients withdrawn from each
 4   arm of the trials. 6-month withdrawals were specifically for a loss of efficacy or
 5   an adverse event. If 6-month rates were not given then the number of patients
 6   who were withdrawn annually was halved. In the model, the probability of
 7   withdrawing was held constant for each 6-month time period that a patient
 8   was on a treatment. A random-number generator was used to simulate when
 9   a patient would withdraw from a treatment. Beta distributions were assigned
10   to reflect the uncertainty from the sum of withdrawals and trial participants in
11   each treatment strategy.
12
13   Table 5 – 6 month Withdrawal Rates
                              Strategy      6-month withdrawal rate
                              Monotherapy   0.095
                              Step-up       0.020
                              Parallel      0.060
                              Intensive     0.049
                              Step-down     0.026
                              Steroid       0.074
14
15   For the monotherapy arm, both the probability of achieving an ACR 20 or 50
16   response, and the rate of withdrawal were assumed to be the same for both
17   1st DMARD and 2nd DMARD treatment before biologics.

18   A.5.3 HAQ
19   The most recorded measure of disease activity in the trials is the Health
20   Assessment Questionnaire (HAQ), a measure of functional ability. HAQ is
21   used as the measure of a patients level of disease severity in all recent cost-
22   effectiveness analyses of RA and also the relationship of HAQ to costs and
23   QALYs is known.
24
25   ACR response is the most commonly reported measure of response in the
26   clinical trials and so has been incorporated into the model as an indication on
27   the clinical impact at 6-months of a treatment strategy on their level of disease
28   activity413. The ACR response criteria defines a ACR 20 or ACR 50 response
29   as a 20% or 50% reduction in their disease activity, as defined as a 20% or
30   50% reduction of their swollen joint count, plus an improvement in three of five
31   other core measures. The 20% or 50% response cannot be seen as a 20% or
32   50% improvement in a patients HAQ level, as they are fundamentally different
33   measures of a patient’s disease activity. Regression of a large longitudinal
34   outcomes bank, the National Databank for Rheumatic Diseases (NDRD) was
35   performed in a previous cost-effectiveness analysis and estimated a
36   percentage HAQ improvement for an ACR 20 and ACR 50 response at 6-
37   months to be 37.8% and 85.3% respectively414. Therefore in the model when
38   a patient is simulated as having an ACR 20 response at 6-months, their HAQ
39   has improved by 37.8% from their baseline HAQ level,
40
41   Once a patient has had their ACR response simulated, their disease level
42   (HAQ) increases at an annual rate of 0.0418, as estimated in a meta-analysis


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               250 of 307
     DRAFT FOR CONSULTATION


 1   of long one natural disease data345. This data only looks at the annual
 2   progression of a patient’s level of HAQ when on monotherapy DMARDs, but
 3   for the baseline case is assumed across all treatment strategies. This
 4   assumption will be explored in the sensitivity analysis, with a differential rate
 5   applied to monotherapy and combinations.
 6
 7   When a patient withdraws from treatment, it is assumed that their HAQ
 8   rebounds back by the same amount that they may have gained from a 6-
 9   month ACR 20 or 50 response. So they will not rebound back to their original
10   baseline HAQ level, as the time spent on treatment will assume that there is a
11   gradual worsening of their disease level over time.

12   A.5.4 Outcomes
13   Quality Adjusted Life Years (QALYs) are the final outcome of the model, to
14   allow for a full cost-effectiveness analysis of the alternative treatments. A
15   QALY is calculated by a utility score for patients Health Related Quality of Life
16   (HRQoL) multiplied by a year. The majority of the trials used to gain efficacy
17   data for the DMARD combinations do not report any final HRQoL outcome
18   measures, and so a reported intermediate disease-specific outcome measure
19   has to be statistically ‘mapped’ to a utility estimate. Mapping is not an ideal
20   solution, as disease-specific health outcomes do not fully capture the impact
21   of treatments on a patient’s Health Related Quality of Life (HRQoL). However
22   mapping is a common approach adopted in the modelling of Rheumatoid
23   Arthritis and this approach has been observed in all past RA technology
24   appraisals.
25
26   The most recorded measure of disease activity in the trials is the Health
27   Assessment Questionnaire (HAQ), and so is used to track the patient’s level
28   of disease activity through the model. HAQ may not capture all dimensions of
29   HRQoL, especially pain. There is a risk that it may underestimate the impact
30   of RA on patients, and the improvements that a treatment may provide.
31   However in the Abatacept Single Technology Appraisal (STA), the Appraisal
32   Committee notes that whilst it has limitations, “it is the best means of
33   estimating utility for the purpose of the economic analysis given the available
34   data”.§§§ HAQ correlates highly to utility measures such as the EQ5D, and so it
35   is seen as appropriate to estimate a patients utility through mapping from
36   HAQ. Mapping tools are available to derive utility from HAQ scores, for this
37   model a regression analysis of US National Databank for Rheumatic Diseases
38   (NDRD) is used. Utilities were mapped from HAQ-DI values of nearly 89,000
39   RA patients who were simultaneously administered the EuroQol, HAQ, pain
40   indices and SF-36 (see table 6). The HAQ scores were grouped into quarter-
41   point categories from 0-3 and the associated model estimates for EQ-5D
42   weighted health index are provided.
43
44   Table 6 - Wolfe NDRD Utility Regression
                               HAQ category     N     Mean utility (SD)    SE
                               0 - <0.25      13249   0.857 (0.16)        0.001
                               0.25 - <0.50    8305   0.803 (0.13)        0.001


     §§§§§§
              Section 4.3.10 http://www.nice.org.uk/page.aspx?o=388554

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                     251 of 307
        DRAFT FOR CONSULTATION

                                       0.50 - <0.75         8243   0.762 (0.14)           0.001
                                       0.75 - <1.00         8717   0.713 (0.15)           0.002
                                       1.00 - <1.25         9390   0.657 (0.15)           0.002
                                       1.25 - <1.50         9596   0.590 (0.18)           0.002
                                       1.50 - <1.75         9144   0.511 (0.19)           0.002
                                       1.75 - <2.00         8113   0.427 (0.21)           0.002
                                       2.00 - <2.25         5787   0.333 (0.24)           0.003
                                       2.25 - <2.50         3110   0.229 (0.25)           0.002
                                       2.50 - <2.75         1486   0.120 (0.27)           0.004
                                       2.75 - <3.00          742   0.034 (0.33)           0.012
1
2       As well as the regression model, two linear transformations (Bansback et al
3       2007334 and Birmingham Rheumatoid Arthritis Model used for the Biologics
4       Technology Appraisal415) were taken from the literature and used in the
5       sensitivity analysis of the model. Table 7 gives the details of the linear
6       transformations. The basecase analysis uses the NDRD regression data as it
7       is from a large sample of patients, and a common critique of linear
8       transformations is they do not take a clear linear correlation.
9
10      Table 7 - Linear HAQ to Utility Transformations
                                               Reference     Conversion
                                               Bansback      Utility = 0.76 - 0.28HAQ
                                               BRAM/Hurst    Utility = 0.862 - 0.327HAQ


11      A.5.5 Adverse Events
12      The implications of adverse drug events on costs or patient utility were not
13      introduced directly into the model but were reflected indirectly in the treatment
14      withdrawal rates.

15      A.5.6 Resource Use
16      A HAQ to resource use estimation was used to estimate the long run extra
17      usage of NHS resources, taken from the Resource Utilisation Norfolk Arthritis
18      Register (NOAR)413. Table 8 shows the average resource use (hospital days,
19      outpatient visits and % of joint replacements) grouped by patients HAQ level.
20      These rates are combined with national costs to give an expected cost of
21      additional direct resource use as a function of a patients HAQ level. The
22      factoring of resource usage by HAQ level indirectly accounts for adverse
23      events and associated complications due to having a high level of disease
24      activity.
25
26      Table 8 - HAQ to Resource use
     HAQ        Hospital Days (year)       Number of outpatient visits (year)      % patients who require joint replacement
     0          0.2                        0.6                                     0.3
     0-1        0.5                        1                                       0.8
     1-2        1.2                        1.5                                     2.3
     2-3        5.1                        2.1                                     4
                                                i
     Hospital days                     £235.00
                                              i
     Outpatient visit                  £85.00
                                                 ****
     Joint replacement                 £7,410.32
27


        ****
           BMS abatacept submission, weighted average of all primary and revisional joint replacements, uprated from 2005
        to 2007 using 5.5% inflator

        Rheumatoid arthritis: full guideline DRAFT (August 2008)                                          252 of 307
       DRAFT FOR CONSULTATION


1      The estimated annual cost of resources used related to HAQ is £120.23,
2      £261.78, £579.94 and £1673.41 for a patient with a HAQ of 0, 0-1, 1-2 and 2-
3      3, respectively.

4      A.5.7 Unit Costs
 5     All unit costs are taken from literature and national unit costs (See Table 9).
 6     The cost of treatment for each trial was calculated, and the cheapest trial arm
 7     was used as the strategy cost in the model. This was because some trials
 8     used more expensive DMARDs, and the underlying assumption is that there
 9     is no significant difference in the efficacy of particular DMARDs and so the
10     cheapest is to be used.
11
12     Table 9 - Costs (BNF 2008)
Drug                             approx dose       price          Pack size                          Cost
                                    mg/day        £/pack   mg/tablet   tablet/packet   Drugs £/day     Drugs £/month
Sulfasalazine                            2000    £18.33         500              112   £ 0.65          £   19.93
Methotrexate                              1.15   £ 3.27          2.5              28   £ 0.05          £   1.65
Hydroxychloroquine                         300   £ 5.46         200               60   £ 0.14          £   4.15
Prednisolone tablets                        10   £ 0.60            5              28   £ 0.04          £   1.30
Prednisolone injections mg/mL                    £ 5.73           25               1
Leflunomide                                  20  £51.13           20              30   £   1.70        £    51.88
Cyclosporine                               175   £13.86           25              30   £   3.23        £    98.43
Betamethasone injections mg/mL                   £ 1.22            4               1
Folic Acid                                  0.7  £ 0.40            5              28   £   0.01        £    0.06
Outpatient Attendance (day case weighted average cost)                 £85.00                         2007 PSSRU
13
14     In general each trial requires review and adjustment of the treatment at 3-
15     monthly intervals. The intensive step-up strategy in the TICORA trial requires
16     a second review after 3 months and from then it required monthly reviews for
17     adjustment of the treatment and for administration of prednisolone injections,
18     which is why the unit costs associated with the intensive step-up strategy are
19     over three times higher than the second most costly treatment strategy. Costs
20     included are drug costs, doctor/hospital consultation costs and drug
21     administration costs. The 6-month total drug, administration and review costs
22     are given in Table 10.
23
24     Table 10 - 6 month Strategy Costs
                                      Strategy              6-month cost
                                      Monotherapy           £251.40
                                      Steroid               £269.98
                                      Intensive Step-up     £766.35
                                      Step Down             £269.29
                                      Step Up               £266.93
                                      Parallel              £263.56
25

26     A.5.8 Population
27     The baseline characteristics (see table 11) of patients with early RA are taken
28     from the Early Rheumatoid Arthritis Study (ERAS)401. The study provides
29     mean estimates and the standard deviation, from which the uncertainty is
30     used to sample random patients in the model.
31


       Rheumatoid arthritis: full guideline DRAFT (August 2008)                                   253 of 307
     DRAFT FOR CONSULTATION

1    Table 11 - ERAS Baseline Patient Characteristics
                        Parameter                  Value    SD

                        Age (years)                54.8     13.6
                        Sex (proportion female)    66.6%    -
                        Disease Duration (years)   0.68     0.508
                        HAQ at baseline            1.11     0.7
 2
 3   It is noted that there is likely to be a difference between the baseline
 4   characteristics of patients in the UK with RA, than those selected for the
 5   clinical trials used in this analysis. However it is assumed appropriate to base
 6   the analysis on UK patients as there is no clinical evidence suggesting a
 7   difference in the treatment effectiveness between patients with different
 8   baseline characteristics. The life expectancy of patients is determined from
 9   standard UK life tables416,416.
10

11   A.5.9 Biologics
12   Instead of recreating a lifetime model of the complete patient pathway of care
13   for RA, the progression of patients to biologic therapies who have failed at
14   least two DMARDs is modelled by ‘bolting on’ estimated costs and QALYs
15   from an existing model. These values are taken from the patient level
16   simulation model by Brennan et al345. The Brennan model provides registry-
17   based evidence that biologics are cost-effective as a 3rd line strategy, with a
18   baseline ICER of £23,882 per QALY.
19
20   The baseline lifetime cost for biologics is estimated as £57,919, and result in
21   5.1514 QALYs. Note that these values are discounted at the old NICE rate of
22   6% for costs and 1.5% for QALYs, and this is maintained for the biologic
23   components of this model as these discount rates were used when NICE
24   determined that biologics were cost-effective as 3rd line treatments for RA. It is
25   seen as appropriate to keep the biologic costs and QALYs discounted at the
26   old rates to maintain biologics as a cost-effective strategy for 3rd line
27   treatment. Sub-group sensitivity analysis of the Brennan model indicates that
28   updating the model to the new discount rates of 3.5% for costs and QALYs
29   would return an ICER of £32,013. Also the Brennan model provides sub-
30   group analysis across patient covariates such as age, HAQ level, disease
31   duration and gender and so the model could potentially assign costs and
32   QALYs that were proportionally adjusted by the sensitivity analysis reported.
33   However as the sub-group analysis was only performed one-way on each of
34   the covariates, it would be inappropriate to use them at all as the combined
35   effect of a change across all covariates is not known.
36

37   A.5.10 Discounting
38   The estimated costs and QALYs were discounted at 3.5% per year, as per the
39   NICE reference case.
40




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              254 of 307
    DRAFT FOR CONSULTATION


1   A.5.11 Sensitivity analysis
2   Probabilistic Sensitivity Analysis (PSA) was performed. PSA quantifies
3   uncertainty in the model by assigning distributions to parameters. Normal
4   distributions were assigned to the patient baseline characteristics, the ACR
5   response probabilities, log-odds ratios and for annual HAQ progression rates.
6   Beta distributions were assigned to withdrawal rates. All other parameters
7   were held constant. Sub-group sensitivity analysis was performed, along with
8   one-way sensitivity analysis of a number of parameters and structural
9   assumptions.




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           255 of 307
     DRAFT FOR CONSULTATION



1    A.6. Results

2    A.6.1 Basecase analysis
3
4    The model results for 100 patients run through 1000 simulations of the patient
5    level simulation are shown in Table 12.
6
7    Table 12 - Model Results: Basecase Analysis
      Strategy      Cost       QALYs     Cost         QALY         ICER          Net Benefit    Net Benefit
                                         difference   difference                                Rank
      Monotherapy    £55,996     13.73                                             £410,863               2
      Step Up        £50,791     11.91     - £5,205         -1.8        £2,852     £306,493               5
      Parallel       £55,573     13.42       - £423         -0.3        £1,356     £347,098               4
      Intensive      £61,046     15.77       £5,050          2.0        £2,482     £356,030               3
      Step Down      £48,849     15.32     - £7,147          1.6   Cost Saving     £296,191               6
      Steroid        £57,468     11.79       £1,472         -1.9   Dominated       £412,025               1
 8
 9   It is clear from the results that step-down combination is a dominating/cost-
10   saving strategy, with an expected 1.6 QALYs gained whilst saving £7,147,
11   when compared to monotherapy. The results also show that Steroids plus
12   DMARD monotherapy is a dominated strategy, which is more costly and less
13   effective than DMARD monotherapy. Whilst the steroid strategy had a slightly
14   better chance achieving an ACR response, the small difference was
15   countered by patients in monotherapy having to go through the strategy twice
16   before moving to biologics, and therefore having another chance of seeing an
17   ACR response and the associated gains in QALYs before moving to biologics.
18




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                  256 of 307
     DRAFT FOR CONSULTATION


                                                                                        Mean treatment effects and costs


                                                           £70,000



                                                           £60,000
            Mean cost (2007 UK £ per person, discounted)




                                                           £50,000



                                                           £40,000



                                                           £30,000



                                                           £20,000



                                                           £10,000



                                                              £-
                                                                     0.0   2.0    4.0        6.0       8.0       10.0      12.0       14.0       16.0   18.0
                                                                                         Mean effect (QALYs per person, discounted)

                                                                            CONTROL     STEP UP    PARALLEL   INTENSIVE    STEP DOWN         STEROID
1
2    Figure 3 - Mean Treatment Costs and Benefits: Basecase Analysis
 3
 4   The remaining strategies; step up, intensive step up and parallel combination
 5   all had estimated ICERs of under £3,000 per QALY. NICE recommends that
 6   advisory bodies should apply a cost-effectiveness threshold of between
 7   £20,000 and £30,000 per QALY†††† and so in this case step up, intensive step
 8   up and parallel combination are cost-effective strategies when compared to
 9   DMARD monotherapy.
10
11   The model was also run deterministically, using the point estimates for
12   parameters. The results of a deterministic run of 100 patients through the
13   basecase model are given in Table 13. The results indicate that through a run
14   of 100 patients, the results are generally similar to those when run
15   probabilistically, although intensive step-up strategy becomes cost saving,
16   due to the costs of this treatment being less than the monotherapy arm whilst
17   maintaining a similar gain in QALYs. In terms of ACR20 and 50 response at 6
18   months, intensive step-up is a highly effective strategy, and this is attributable
19   to gains in QALY and reductions in HAQ that are likely to see the significantly
20   lower resource costs erase the difference in treatment costs. 6-months on
21   monotherapy with HAQ>2 costs £1088.11, whilst 6-months on monotherapy
22   with HAQ<1 costs £897.27.
23



     ††††
                      Social value judgements: principles for the development of NICE guidance

     http://www.nice.org.uk/media/873/2F/SocialValueJudgementsDec05.pdf

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                                                                    257 of 307
     DRAFT FOR CONSULTATION

1    Table 13 - Basecase Analysis: Deterministic Results
                Strategy        Cost         QALYs       Cost         QALY          ICER
                                                         difference   difference
                Monotherapy       £57,106       14.41
                Step Up           £48,258       11.73      - £8,848         -2.67         £3,308
                Parallel          £56,822       13.11        - £284         -1.29          £219
                Intensive         £54,570       15.92      - £2,536          1.51   Cost Saving
                Step Down         £51,986       16.57      - £5,120          2.16   Cost Saving
                Steroid           £60,502       11.67        £3,396         -2.74   Dominated
2
3    Performing a mixed treatment comparison allows full incremental cost-
4    effectiveness analysis between each of the comparators. The results of the
5    full incremental analysis are given in Table 14.
6
7    Table 14 - Incremental Basecase Analysis
                Strategy      Cost        QALY      Net Benefit     ICER           Comparator
                Step Down      £48,849      15.3        £410,863 Cost-saving       Monotherapy
                Intensive      £61,046      15.8        £412,025        £27,392 Step-down
 8                  (Note: Comparisons are against the next best, non-dominated treatment strategy)
 9
10   Compared with the cheapest strategy, step down combination, all strategies
11   re dominated by the laws of simple domination apart from intensive step-up.
12   This means that the remaining strategies are more expensive and less
13   effective, including DMARD monotherapy. When compared to step down
14   combination, intensive step up has an incremental cost-effectiveness ratio of
15   £27,392 per QALY.
16

17   A.6.2 Uncertainty
18   Probabilistic sensitivity analysis (PSA) was performed to reflect the
19   uncertainty in the input parameters of the model and determine what this
20   means for the results of the model. The results of the PSA are showing in
21   Figure 4. The cost-effectiveness acceptability curves (CEACs) show the
22   estimated probability that a treatment option is cost-effective given the amount
23   that we are willing to pay for a QALY (the cost-effectiveness threshold shown
24   on the horizontal axis). This helps to reinforce the conclusion that at a NICE
25   cost-effectiveness threshold of around £20,000 to £30,000 per QALY, that
26   step-down combination is the most likely cost-effective strategy when
27   compared to monotherapy, with a probability of over 80%.
28
29   The CEAC also illustrates that there is considerable uncertainty around the
30   relatively ranking of the other combination strategies. Figure 5 is an
31   incremental CEAC, which estimates which treatment is likely to be the most
32   cost-effective when compared to all other comparators. It shows that at a
33   threshold of £20,000 per QALY, step-down combination is the most likely to
34   be the most cost-effective.
35




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                                      258 of 307
    DRAFT FOR CONSULTATION


                                                                Common Baseline CEAC

                                      1.200

                                      1.000
         Probability Cost Effective




                                      0.800

                                      0.600

                                      0.400

                                      0.200

                                      0.000
                                              0     20000        40000       60000        80000         100000      120000
                                                            Cost Effectiveness Threshold (£ per QALY)

                                                  STEP UP    PARALLEL      INTENSIVE     STEP DOWN        STEROID
1
2   Figure 4 - Baseline Analysis - CEAC Common Baseline
3
                                                                  Incremental CEAC
                                      0.900

                                      0.800
     Probability Cost Effective




                                      0.700

                                      0.600

                                      0.500

                                      0.400

                                      0.300

                                      0.200

                                      0.100

                                      0.000
                                              0     20000       40000       60000       80000       100000        120000

                                                       Cost Effectiveness Threshold (£ per QALY)

                                  CONTROL         STEP UP    PARALLEL      INTENSIVE      STEP DOWN          STEROID
4
5   Figure 5 - Baseline Analysis - Incremental CEAC
6




    Rheumatoid arthritis: full guideline DRAFT (August 2008)                                                     259 of 307
      DRAFT FOR CONSULTATION



1     A.7. Discussion
2
3     The basecase analysis in the previous section provides convincing evidence
4     that a step-down combination strategy is an effective and cost-effective first
5     line strategy for patients with RA. However this conclusion is based on a
6     number of clinical assumptions which are tested through sensitivity analysis.
7

8     A.7.1 Treatment Withdrawal
 9    A key assumption of the model is that patients who do not achieve an ACR 20
10    or 50 response after 6-months will either move to biologics or begin a 2nd line
11    DMARD monotherapy. This may be clinically inappropriate with a 6-month
12    ACR response not fully capturing patients who may have yet to respond to
13    their treatment, as well as capturing those who may have achieved a sub-
14    ACR20 response but still achieved some form of disease control. Therefore
15    the model was run with none responders continuing on their treatment until
16    they are withdrawn due to a loss of efficacy or an adverse event. The model
17    was run probabilistically, with 100 patients sent through 1000 samples and the
18    results are given below in Table 15.
19
20    Table 15 - Results - No Switch For Non-Responders
                                        Cost         QALY
                     Cost      QALYs    Difference   Difference   ICER           Net Benefit   Net Benefit Rank
     Monotherapy    £ 53,569    14.58                                              £470,889                       1
     Step Up        £ 45,578    15.37      -£7,991          0.8   Cost Saving      £415,487                       4
     Parallel       £ 54,975    16.19       £1,405          1.6           £872     £430,693                       3
     Intensive      £ 62,245    17.25       £8,676          2.7         £3,244     £383,742                       6
     Step Down      £ 46,718    17.25      -£6,851          2.7   Cost Saving      £394,261                       5
     Steroid        £ 57,457    15.06       £3,888          0.5         £8,096     £455,295                       2
21
22    Interestingly, both step-up and step-down combinations are now a dominant
23    strategy. This is because, referring back to Table 5, these strategies have
24    very low rates of withdrawals compared to the other strategies. Spending a
25    much longer time on DMARD therapy gives a patient a long time for QALYs to
26    accrue before withdrawing and then being assigned costs and QALYs for
27    biologics, assuming they have not died before the end of their treatment.
28
29    Table 16 compares the estimated length on treatment between the basecase
30    analysis and a scenario when non-responders do not switch after 6 months.
31
32    Table 16 - Time on DMARDs analysis
                               None-responders Switch                     None-responders do not Switch
                    Cost       QALY     ICER         Time on      Cost         QALY    ICER           Time on
                                                     DMARDs                                           DMARDs
      Monotherapy £57,106        14.41                   12.29      £ 53,569    14.58                    17.97
      Step Up      £48,258       11.73       £3,308      18.66      £ 45,578    15.37   Cost Saving      28.94
      Parallel     £56,822       13.11        £219       12.13      £ 54,975    16.19          £872      17.13
      Intensive    £54,570       15.92 Cost Saving       14.77      £ 62,245    17.25        £3,244      27.86
      Step Down    £51,986       16.57 Cost Saving       23.78      £ 46,718    17.25   Cost Saving      23.18
      Steroid      £60,502       11.67 Dominated          8.73      £ 57,457    15.06        £8,096       9.38
33    (Time on DMARDs is in half-years)
34

      Rheumatoid arthritis: full guideline DRAFT (August 2008)                                  260 of 307
     DRAFT FOR CONSULTATION


 1   A conclusion from this is that there is clinical benefit in finding a DMARD that
 2   is effective for a patient. The standard clinical belief is that patients that are on
 3   drawn out monotherapy treatment will be on DMARDs much longer than
 4   combinations before they switch to biologics. However the evidence from the
 5   trials is that monotherapy has the greatest rate of withdrawals for loss of
 6   efficacy or for an adverse event. This could be because of the lower rate of
 7   clinical response seeing withdrawals, or just random variation due to the
 8   generally small nature of the trials. Either way it supports the clinical evidence
 9   suggesting that combination DMARDs are no more toxic or less safe than
10   monotherapy DMARDs, and that there is no greater risk of adverse events
11   (see Clinical Evidence for DMARDs in The Guideline document).

12   A.7.2 HAQ Progression
13   The second key assumption of the model is that all patients on treatment will
14   see an increase in their HAQ by a given annual rate345 . However this data is
15   based on DMARD monotherapy evidence, and as yet there has not been a
16   long-run formal analysis of the progression RA in patients on combinations
17   DMARDs. The follow-up evidence in BeSt and COBRA suggest that there is
18   not a significant annual increase in their HAQ from 6-months until the end of
19   follow-up. For this reason, the model was re-run with the HAQ annual
20   progression assigned to the monotherapy group (see table 17), whilst the
21   combination DMARD group saw no annual progression when they had an
22   ACR20 or 50 response. As per the basecase analysis, non-ACR responders
23   were switched to their relevant second line therapy.
24
25   The model was run probabilistically with 100 patients through 1000
26   simulations.
27
28   Table 17 - Probabilistic Analysis - HAQ Progression
                       Strategy      Cost             QALY      ICER
                       Monotherapy          £56,545     13.01   £-
                       Step Up              £51,061     11.74   £ 4,318
                       Parallel             £55,612     13.35   Cost Saving
                       Intensive            £61,160     15.67   £1,732
                       Step Down            £48,935     15.08   Cost Saving
                       Steroid              £57,546     11.75   Dominated
29
30   As expected, the results see a general improvement in the cost-effectiveness
31   of the combination strategies when compared to monotherapy. Steroid plus
32   monotherapy is still a dominated strategy, as the relatively fast withdrawal
33   does not allow much benefit to accrue over time on treatment when compared
34   to monotherapy. Parallel combination is now a dominating strategy, along with
35   step down, whilst step up and intensive are highly cost-effective when
36   compared to monotherapy. Until longer follow-ups and dedicated analysis on
37   the HAQ progression of patients on combinations are available, the issue of
38   how to model the long term benefits of DMARDs will still exist.

39   A.7.3 Subgroup Analysis
40

     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 261 of 307
     DRAFT FOR CONSULTATION


1    Deterministic subgroup analysis was performed to investigate whether
2    baseline HAQ or patient age would have any strong influence on the
3    outcomes and cost-effectiveness of each strategy. Tables 18-20 give the
4    results for the analysis when patients are split between baseline HAQ 0-1, 1-
5    2, and 2-3. The results show that for steroid strategy, it is a dominated option
6    for patients with a HAQ less than 1 but is cost-effective for patients with active
7    disease. The remaining strategies are cost-effective (<£20,000 per QALY) or
8    cost-saving across all baseline HAQ groups.
9
10   Table 18 - Deterministic Subgroup Analysis - HAQ<1
                      Strategy      Cost                QALY      ICER
                      Monotherapy             £58,064     13.73   £-
                      Step Up                 £48,169     12.34   £7,140
                      Parallel                £54,683     12.87   £3,957
                      Intensive               £66,772     16.03   £3,781
                      Step Down               £48,008     15.36   Cost Saving
                      Steroid                 £59,200      9.66   Dominated

11
12   Table 19 - Deterministic Subgroup Analysis - HAQ 1-2
                      Strategy      Cost                QALY      ICER
                      Monotherapy   £ 57,880              12.49   £      -
                      Step Up       £ 44,544              13.76   Cost Saving
                      Parallel      £ 56,459              12.11   £     3,742
                      Intensive     £ 62,996              15.24   £     1,857
                      Step Down     £ 47,159              16.28   Cost Saving
                      Steroid       £ 59,638              13.18   £     2,556
13
14   Table 20 - Deterministic Subgroup Analysis - HAQ 2-3
                      Strategy      Cost                QALY      ICER
                      Monotherapy   £57,643               12.73   £-
                      Step Up       £49,072               13.16   Cost Saving
                      Parallel      £55,385               14.20   Cost Saving
                      Intensive     £63,387               16.69   £1,448
                      Step Down     £46,746               15.83   Cost Saving
                      Steroid       £57,552                9.03   £25
15
16   Tables 21 and 22 show how the cost-effectiveness of each strategy varies
17   depending on the age of the patient as they enter the model. Understandably
18   the patients aged over 79 generally accrue less QALYs. Interestingly the
19   parallel strategy is dominated in the younger group, whilst being a very cost-
20   effective strategy in the older group. Because the clinical effectiveness is
21   assumed not to vary by age, this wide difference is due to the wide
22   uncertainty in the model and provides clear evidence that drawing conclusions
23   from point estimates, and having no grasp of the uncertainty in the model, is
24   inappropriate.
25
26   Table 21 - Deterministic Subgroup Analysis - Age <50
                      Strategy      Cost                QALY      ICER


     Rheumatoid arthritis: full guideline DRAFT (August 2008)                   262 of 307
     DRAFT FOR CONSULTATION


                      Monotherapy   £ 53,621       12.81    £        -
                      Step Up       £ 53,556       12.15    £      98
                      Parallel      £ 55,659       12.40    Dominated
                      Intensive     £ 63,603       16.26    £    2,890
                      Step Down     £ 49,267       15.81    Cost saving
                      Steroid       £ 59,023       11.85    Dominated
1
2    Table 22 - Deterministic Subgroup Analysis - Age >79
                      Strategy      Cost       QALY         ICER
                      Monotherapy   £ 56,048       13.76    £        -
                      Step Up       £ 47,646       11.64    £    3,970
                      Parallel      £ 53,596       13.18    £    4,218
                      Intensive     £ 64,273       15.87    £    3,899
                      Step Down     £ 45,294       14.96    cost saving
                      Steroid       £ 58,821       10.78    dominated
3

4    A.7.4 HAQ to Utility
5    Whilst the basecase analysis uses the Wolfe NDRD analysis to convert HAQ
6    to utility, there are two published linear conversions that transform HAQ into a
7    utility score. The model was run deterministically through both of these
8    transformation tools.
9
10   Table 23 - Deterministic - Bansback Utility
                      Strategy      Cost       QALY         ICER
                      Monotherapy   £58,017        10.79    £-
                      Step Up       £52,075          9.81   £ 6,051
                      Parallel      £57,671        10.93    cost saving
                      Intensive     £65,428        12.39    £4,637
                      Step Down     £48,339        11.58    cost saving
                      Steroid       £59,512          8.79   dominated

11
12   Table 24 - Deterministic - Hurst/BRAM Utility
                      Strategy      Cost       QALY         ICER
                      Monotherapy   £55,077        10.44    £-
                      Step Up       £53,706        10.08    £3,898
                      Parallel      £57,886        10.79    £7,920
                      Intensive     £64,129        13.68    £2,790
                      Step Down     £49,578        12.01    cost saving
                      Steroid       £56,926        10.67    £7,727
13
14   Both tables show that the resulting QALYs estimated are lower than the
15   basecase analysis. However the difference in the costs would not have been
16   caused by the uncertainty in the estimates and again show that drawing firm
17   conclusions from a deterministic run of 100 patients is inappropriate.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)             263 of 307
     DRAFT FOR CONSULTATION


1    A.7.5 Cost
 2   The TICORA trial, that populated the intensive step-up strategy, included a
 3   cost-effectiveness analysis288 which found the intensive arm to be more
 4   effective at no additional cost. There were higher inpatient costs in the
 5   standard step up arm which dominated the higher prescribing costs in the
 6   intensive arm. Because the costs of the strategies were estimated using the
 7   prescribing, administration and review costs of each of the strategies and
 8   direct extra admissions or admissions due to adverse events not being
 9   directly estimated, the modelled cost of intensive step up was higher than the
10   standard step up strategy. The difference in cost was met by an appropriate
11   gain in QALYs which sees intensive step-up as a very cost effective strategy.
12   However it is important to elaborate on the reasons why the resulting
13   conclusions are different from the original trial analysis.
14
15   It is also important to make reference to the estimation of costs of the
16   strategies. The cost of each trial strategy was estimated, and then the
17   cheapest strategy was taken for each of the groups. The dosages were
18   generally similar, and so the fundamental assumption is that with no
19   significant difference between the safety, efficacy and tolerability of the
20   DMARDs, the cheapest drug(s) should be chosen to construct a DMARD
21   combination. Apart from the intensive step-up strategy which required monthly
22   monitoring and drug administration, the other strategies all had very similar
23   costs, due to the relatively low costs of the drugs, and the key component
24   being the number of admissions required.
25
26   Finally, in the context of actual clinical practise, it may not be appropriate that
27   at the beginning of therapies that patients require only 3-monthly monitoring
28   (see table 25). The model was re-run so that across all strategies monthly
29   outpatient attendances were required and this extra cost was factored in. If a
30   patient responded then that patient would then return to a rate of attendance
31   as determined in the trial (All 3-monthly except for TICORA intensive step-up
32   which required monthly attendances). The model was re-run probabilistically,
33   with these extra monitoring costs, through 100 patients and 1000 simulations.
34
35   Table 25 - Monthly monitoring for 1st 6 months
                      Strategy      Cost             QALY      ICER
                      Monotherapy          £56,865     13.05   £ -
                      Step Up              £51,941     10.89   £2,276
                      Parallel             £56,657     12.60    £458
                      Intensive            £62,199     14.55   £ 3,556
                      Step Down            £50,026     13.99   Cost Saving
                      Steroid              £58,142     11.26   Dominated
36
37   When compared to the basecase analysis, the results of the extra monitoring
38   for the first 6 months are no difference in terms of the cost-effectiveness of
39   any of the strategies, but there is a small increase in the cost of all strategies.
40   It is important that any model accurately reflects clinical practice and so these
41   results assume that extra monitoring is required, both for the risk of adverse



     Rheumatoid arthritis: full guideline DRAFT (August 2008)                264 of 307
     DRAFT FOR CONSULTATION


1    events, and also to formally assess the level of function and disease to
2    establish whether there has been a response to treatment.

3    A.8. Conclusions
 4
 5   This cost-effectiveness analysis has provided conclusions on the use of
 6   combination DMARD strategies compared to DMARD monotherapy. The
 7   analysis has also provided conclusions on the use of steroids alongside
 8   DMARD monotherapy when compared to monotherapy alone. The analysis
 9   was focused on patients with Rheumatoid Arthritis who were both DMARD
10   and biologic naïve.
11
12   The analysis took a lifetime perspective, which required the estimation of
13   costs and QALYs for patients who had failed two DMARDs and were eligible
14   for biologic therapies, and the subsequent costs and QALYs associated with
15   biologic therapy. The analysis was based on two clinical assumptions, firstly
16   that patients who did not achieve an ACR 20 or 50 response after 6 months
17   moved to 2nd DMARD or biologics. Secondly, if a patient achieved an ACR 20
18   or 50 response then their HAQ worsened at a given rate. The ACR 20 or 50
19   response evidence is taken from thirteen randomised controlled clinical trials
20   and mixed treatment comparison techniques are applied to allow the trials to
21   be compared to each other. The specific protocols used in the trials are used
22   to estimate the cost of each strategy, and the withdrawal rates for a loss of
23   efficacy or an adverse event are taken from a weighted average of six month
24   withdrawals from the trials.
25
26   A clear result from the basecase analysis is that it is cost-effective to provide
27   a combination strategy of DMARDs as first line treatment for RA. The analysis
28   shows that a step-down combination of DMARDs is likely to be a cost-saving
29   strategy, and more than likely to be the most cost-effective strategy given a
30   NICE threshold of between £20,000 and £30,000 per QALY. Incidentally, one
31   of the two trials that provided the evidence for a step-down combination
32   strategy, COBRA299 provided a cost-effectiveness analysis in a Dutch setting
33   and reported step-down combination to be a cost-saving strategy. In the
34   COBRA trial, step-down therapy was performed using initial prednisolone,
35   methotrexate and sulphasalazine with prednisolone and methotrexate tapered
36   and stopped from week 28 to week 40. In the Fin-RaCo296 trial, patients were
37   also given prednisolone, methotrexate and hydroxychloroquine which was
38   increased if there was a poor response, and if not the dosages were tapered
39   and stopped from between 9 to 18 months.
40
41   The basecase analysis determined that a strategy of monotherapy plus
42   glucocorticoids is dominated (more costly and less effective) by DMARD
43   monotherapy. This is due to the cost-differential not being covered by a
44   significant increase in ACR response rate. However it is important to note that
45   intensive step up, step up and step down all include glucocorticoids in the
46   regimen and so steroids contribute to the effectiveness of a DMARD
47   combination strategy.
48

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              265 of 307
     DRAFT FOR CONSULTATION


 1   However it is important to note the uncertainties over the rates of withdrawals
 2   and the probability of an ACR response. However PSA provides confirmation
 3   that at a threshold of £20,000 per QALY, step-down has an 86% likelihood of
 4   being cost-effective, whilst intensive step-up has an 80% likelihood, and
 5   parallel combination having an 40% likelihood. Incremental PSA analysis
 6   suggests that at a £20,000 per QALY threshold, step down is more likely to be
 7   the most cost-effective strategy than intensive step up.
 8
 9   Sub-group analysis was performed, modelling groups of patients according to
10   their age and their baseline HAQ. The analysis emphasised the dangers of
11   drawing conclusions of running a model deterministically, as there is great
12   uncertainty in the model that cannot be accounted for when run using point
13   estimates. The combinations therapies all maintained being cost-effective
14   strategies when compared to baseline. The glucocorticoid alongside
15   monotherapy group was found to become a cost-effective strategy when the
16   model was run in patients with a baseline HAQ level above 1. Sensitivity
17   analysis was performed on the conversion of HAQ to utility scores. The model
18   was run through two linear transformations, both of which did not cause any
19   DMARD combinations to become not cost-effective, and again highlight the
20   uncertainty in the parameters.
21
22   A key assumption in the model is that all patients that respond to treatment
23   will see a annual increase in their HAQ level. However the evidence for this
24   assumption is limited to DMARD monotherapy, and so sensitivity analysis was
25   performed to determine the cost-effectiveness of combinations DMARDs
26   when they see a constant annual level of HAQ, and whilst monotherapy
27   remains at the annual rate of increase. The re-run of the model concluded that
28   parallel combination becomes a cost-saving strategy, steroid plus
29   monotherapy remains dominated, and the other combinations all become
30   more cost-effective, when compared to monotherapy. The sensitivity analysis
31   has provided intuitive changes in the results, providing reassurance on the
32   overall validity of the model.
33
34   This analysis has determined that step-down combinations of DMARDs are
35   likely to be very cost-effective or even cost-saving, and other DMARD
36   combinations are very likely to be cost-effective. However it has also
37   highlighted the high level of uncertainty over the clinical evidence used to
38   populate the model, as well as the structural/clinical decisions used to
39   determine how patients withdraw from treatment, and how their disease level
40   behaves whilst on treatment.
41
42   Further research could be done to assess the underlying reasons for patients
43   withdrawing from treatment, which is a key driver for the results of this model.
44   Further research is required to assess the long-term disease activity of
45   patients on monotherapy DMARDs and combination DMARDs, as this has a
46   substantial impact on the QALYs that a patient accrues whilst on treatment.
47   Also the model could be expanded to fully incorporate biologic therapies into
48   the analysis, providing a complete model of drug therapy for patients with RA.
49   This could be done to either completely model the cost-effectiveness of all RA
50   treatment strategies, or to assess the cost-effectiveness of biologic therapies

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             266 of 307
    DRAFT FOR CONSULTATION


1   against more effective DMARD combinations, as opposed to the common
2   approach comparing the cost-effectiveness of biologics against monotherapy
3   DMARDs.




    Rheumatoid arthritis: full guideline DRAFT (August 2008)         267 of 307
    DRAFT FOR CONSULTATION


1   Appendix D
2
    Group        Personal             Personal   Non-personal         Personal non-
    member       pecuniary            family     pecuniary            pecuniary
                 interest             interest   interest             interest
    Raashid      Consultancy for      None       Abbott & Schering    None
    Luqmani      Euro Nippon                     Plough have
                 Kayaku on                       provided financial
                 evaluating role of              support to the MSc
                 gusperimus – now                course in Oxford,
                 terminated - in                 WHKH and RUN
                 Wegener’s
                 granulamatosis
    Louise       None                            none                 I am a member of
    Warburton                                                         the Primary Care
                                                                      Rheumatology
                                                                      Society and on the
                                                                      Editorial Board of
                                                                      Rheumatology in
                                                                      Practice. I have
                                                                      inflammatory
                                                                      arthritis and am on
                                                                      anti-TNF therapy
    Patrick      Consultancy,         None       None                 Medical Advisor for
    Kiely        honoraria for                                        National
                 lecturing for:                                       Rheumatoid
                 Schering Plough                                      Arthritis Society
                 Roche                                                (NRAS)
                 Bristol Myers
                 Squibb
    Chris        None                 None       My department        My wife is a sales
    Deighton                                     has received         representative for
                                                 sponsorship for      Novartis
                                                 meetings and
                                                 unrestricted
                                                 research grants
                                                 from Wyeth,
                                                 Abbott and
                                                 Schering-Plough
    Isabel       None                 None       None                 None
    Raiman
    Ailsa        None                 None       NRAS has             None
    Bosworth                                     received
                                                 educational grants
                                                 from Roche,
                                                 Schering Plough,
                                                 Abbott, Wyeth, and
                                                 very small grants
                                                 from UCB and
                                                 BMS
    Alison       None                 None       None                 None
    Hammond
    Jonathan     None                 None       None                 None
    Tosh
    Jane Hall    None                 None       None                 None
    Enid Quest   None                 None       None                 None
    David        None                 None       None                 None
    Morgan

    Rheumatoid arthritis: full guideline DRAFT (August 2008)                 268 of 307
    DRAFT FOR CONSULTATION


    Group         Personal               Personal   Non-personal   Personal non-
    member        pecuniary              family     pecuniary      pecuniary
                  interest               interest   interest       interest
    Sheena        None                   None       None           None
    Hennell
    Michael       None                   None       None           None
    Rudolf
    Colin Howie   None                   None       None           None
    Andrew        Education contract:    None       None           None
    Robinson      de Puy, Johnson &
                  Johnson (non
                  specific)
                  Consultancy
                  contract:
                  Orthomimetics Ltd
                  (non specific) Glaxo
                  SmithKlein
                  shareholding – non
                  specific
    Anthony       None                   None       None           None
    Redmond
1




    Rheumatoid arthritis: full guideline DRAFT (August 2008)              269 of 307
     DRAFT FOR CONSULTATION


1    REFERENCES
2       1 National Institute for Health and Clinical Excellence. Guidelines
3         Manual. UK: London: National Institute for Health and Clinical
4         Excellence, 2007.

5       2 Arnett FC, Edworthy SM, Bloch DA et al. The American Rheumatism
6         Association 1987 revised criteria for the classification of rheumatoid
7         arthritis. Arthritis & Rheumatism. 1988; 31(3):315-324.

 8      3 Combe B, Landewe R, Lukas C et al. EULAR recommendations for the
 9        management of early arthritis: report of a task force of the European
10        Standing Committee for International Clinical Studies Including
11        Therapeutics (ESCISIT). Annals of the Rheumatic Diseases. 2007;
12        66(1):34-45.

13      4 Lawrence JS. Prevalence of rheumatoid arthritis. Annals of the
14        Rheumatic Diseases. 1961; 20:11-17.

15      5 Symmons D, Turner G, Webb R et al. The prevalence of rheumatoid
16        arthritis in the United Kingdom: new estimates for a new century.
17        Rheumatology. 2002; 41(7):793-800.

18      6 Symmons DP, Barrett EM, Bankhead CR et al. The incidence of
19        rheumatoid arthritis in the United Kingdom: Results from the Norfolk
20        Arthritis Register. British Journal of Rheumatology. 1994; 33(8):735-
21        739.

22      7 British Society for Rheumatology. Clinical Guidelines:Rheumatoid
23        Arthritis. Available from: British Society for Rheumatology. Last
24        accessed on: 2008 July 30.

25      8 National Institute for Health and Clinical Effectiveness. Osteoarthritis:
26        the care and management of Osteoarthritis in adults. (59). London:
27        National Institute for Health and Clinical Effectiveness, 2008.

28      9 National Institute for Health and Clinical Effectiveness. Anakinra for
29        rheumatoid Arhritis. (TA72)., 2003.

30     10 National Institute for Health and Clinical Effectiveness. Rituximab for
31        the treatment of Rheumatoid Arthritis. (TA126). London: NICE, 2007.

32     11 National Institute for Health and Clinical Effectiveness. Abatacept for
33        the Treatment of Rheumatoid Arthritis. (TA141). London: National
34        Institute for Health and Clinical Effectiveness, 2008.

35     12 Pugner KM, Scott DI, Holmes JW et al. The costs of rheumatoid
36        arthritis: an international long-term view. Seminars in Arthritis &
37        Rheumatism. 2000; 29(5):305-320.

38     13 National Rheumatoid Arthritis Society. I want to work:a self-help guide
39        for people with Rheumatoid Arthritis. 2007. London, NRAS.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)             270 of 307
     DRAFT FOR CONSULTATION


1      14 Emery P, Breedveld FC, Dougados M et al. Early referral
2         recommendation for newly diagnosed rheumatoid arthritis: evidence
3         based development of a clinical guide. Annals of the Rheumatic
4         Diseases. 2002; 61(4):290-297.

5      15 Kumar K, Daley E, Carruthers DM et al. Delay in presentation to
6         primary care physicians is the main reason why patients with
7         rheumatoid arthritis are seen late by rheumatologists. Rheumatology.
8         2007; 46(9):1438-1440.

 9     16 Van der Horst- Bruinsma IE, Speyer I, Visser H et al. Diagnosis and
10        course of early-onset arthritis: Results of a special early arthritis clinic
11        compared to routine patient care. British Journal of Rheumatology.
12        1998; 37(10):1084-1088.

13     17 Kaarela K. Prognostic factors and diagnostic criteria in early
14        rheumatoid arthritis. Scandinavian Journal of Rheumatology -
15        Supplement. 1985; 57:1-54.

16     18 Machold KP, Stamm TA, Eberl GJM et al. Very recent onset arthritis -
17        Clinical, laboratory, and radiological findings during the first year of
18        disease. Journal of Rheumatology. 2002; 29(11):2278-2287.

19     19 Arndt U, Behrens F, Ziswiler HR et al. Observational study of a patient
20        and doctor directed pre-referral questionnaire for an early arthritis
21        clinic. Rheumatology International. 2007; 28(1):21-26.

22     20 Devlin J, Gough A, Huissoon A et al. The outcome of knee synovitis in
23        early arthritis provides guidelines for management. Clinical
24        Rheumatology. 2000; 19(2):82-85.

25     21 Harrison BJ, Symmons DPM, Barrett EM et al. The performance of the
26        1987 ARA classification criteria for rheumatoid arthritis in a population
27        based cohort of patients with early inflammatory polyarthritis. Journal of
28        Rheumatology. 1998; 25(12):2324-2330.

29     22 Kaarela K, Hameenkorpi R, Isomaki H. The value of the diagnostic
30        criteria in rheumatoid arthritis. Scandinavian Journal of Rheumatology.
31        1983; 12(1):43-45.

32     23 El Miedany Y, Palmer D, El Gaafary M. Diagnosis of early arthritis:
33        outcomes of a nurse-led clinic. British Journal of Nursing. 2006;
34        15(7):394-399.

35     24 Alarcon GS, Willkens RF, Ward JR et al. Early undifferentiated
36        connective tissue disease. IV.Musculoskeletal manifestations in a large
37        cohort of patients with undifferentiated connective tissue diseases
38        compared with cohorts of patients with well-established connective
39        tissue diseases: followup analyses in patients with unexplained
40        polyarthritis and patients with rheumatoid arthritis at baseline. Arthritis
41        & Rheumatism. 1996; 39(3):403-414.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               271 of 307
     DRAFT FOR CONSULTATION


1      25 Gormley GJ, Steele WK, Gilliland A et al. Can diagnostic triage by
2         general practitioners or rheumatology nurses improve the positive
3         predictive value of referrals to early arthritis clinics? Rheumatology.
4         2003; 42(6):763-768.

5      26 Duer A, Ostergaard M, Horslev PK et al. Magnetic resonance imaging
6         and bone scintigraphy in the differential diagnosis of unclassified
7         arthritis. Annals of the Rheumatic Diseases. 2008; 67(1):48-51.

 8     27 Van der Helm-van Mil AH, Le Cessie S, Van Dongen H et al. A
 9        prediction rule for disease outcome in patients with recent-onset
10        undifferentiated arthritis: how to guide individual treatment decisions.
11        Arthritis & Rheumatism. 2007; 56(2):433-440.

12     28 Houssien DA, Scott DL. Early referral and outcome in rheumatoid
13        arthritis. Scandinavian Journal of Rheumatology. 1998; 27(4):300-302.

14     29 Irvine S, Munro R, Porter D. Early referral, diagnosis, and treatment of
15        rheumatoid arthritis: evidence for changing medical practice. Annals of
16        the Rheumatic Diseases. 1999; 58(8):510-513.

17     30 Conaghan PG, O'Connor P, McGonagle D et al. Elucidation of the
18        relationship between synovitis and bone damage: a randomized
19        magnetic resonance imaging study of individual joints in patients with
20        early rheumatoid arthritis. Arthritis & Rheumatism. 2003; 48(1):64-71.

21     31 Wakefield RJ, Gibbon WW, Conaghan PG et al. The value of
22        sonography in the detection of bone erosions in patients with
23        rheumatoid arthritis: a comparison with conventional radiography.
24        Arthritis & Rheumatism. 2000; 43(12):2762-2770.

25     32 Avouac J, Gossec L, Dougados M. Diagnostic and predictive value of
26        anti-cyclic citrullinated protein antibodies in rheumatoid arthritis: a
27        systematic literature review. Annals of the Rheumatic Diseases. 2006;
28        65(7):845-851.

29     33 Nishimura K, Sugiyama D, Kogata Y et al. Meta-analysis: diagnostic
30        accuracy of anti-cyclic citrullinated peptide antibody and rheumatoid
31        factor for rheumatoid arthritis. ANN INTERN MED. 2007; 146(11):797-
32        808.

33     34 Aho K, Palosuo T, Heliovaara M et al. Antifilaggrin antibodies within
34        "normal" range predict rheumatoid arthritis in a linear fashion. Journal
35        of Rheumatology. 2000; 27(12):2743-2746.

36     35 Aho K, Palosuo T, Knekt P et al. Serum C-reactive protein does not
37        predict rheumatoid arthritis. Journal of Rheumatology. 2000;
38        27(5):1136-1138.

39     36 Koivula MK, Heliovaara M, Ramberg J et al. Autoantibodies binding to
40        citrullinated telopeptide of type II collagen and to cyclic citrullinated
41        peptides predict synergistically the development of seropositive

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             272 of 307
     DRAFT FOR CONSULTATION


1          rheumatoid arthritis. Annals of the Rheumatic Diseases. 2007;
2          66(11):1450-1455.

3      37 Van Aken J, Van Dongen H, Le Cessie S et al. Comparison of long
4         term outcome of patients with rheumatoid arthritis presenting with
5         undifferentiated arthritis or with rheumatoid arthritis: an observational
6         cohort study. Annals of the Rheumatic Diseases. 2006; 65(1):20-25.

7      38 Aho K, Palosuo T, Lukka M et al. Antifilaggrin antibodies in recent-
8         onset arthritis. Scandinavian Journal of Rheumatology. 1999;
9         28(2):113-116.

10     39 Bayliss CE, Dawkins RL, Cullity G et al. Laboratory diagnosis of
11        rheumatoid arthritis. Prospective study of 85 patients. Annals of the
12        Rheumatic Diseases. 1975; 34(5):395-402.

13     40 Devauchelle-Pensec V, Berthelot JM, Jousse S et al. Performance of
14        hand radiographs in predicting the diagnosis in patients with early
15        arthritis. Journal of Rheumatology. 2006; 33(8):1511-1515.

16     41 Hoffman IE, Peene I, Pottel H et al. Diagnostic performance and
17        predictive value of rheumatoid factor, anti-citrullinated peptide
18        antibodies, and the HLA shared epitope for diagnosis of rheumatoid
19        arthritis. Clinical Chemistry. 2005; 51(1):261-263.

20     42 Quinn MA, Green MJ, Marzo OH et al. Prognostic factors in a large
21        cohort of patients with early undifferentiated inflammatory arthritis after
22        application of a structured management protocol. Arthritis &
23        Rheumatism. 2003; 48(11):3039-3045.

24     43 Saraux A, Valls I, Voisin V et al. How useful are tests for rheumatoid
25        factors, antiperinuclear factors, antikeratin antibody, and the HLA DR4
26        antigen for the diagnosis of rheumatoid arthritis? Revue du
27        Rhumatisme (English Edition). 1995; 62(1):16-20.

28     44 Solau-Gervais E, Legrand JL, Cortet B et al. Magnetic resonance
29        imaging of the hand for the diagnosis of rheumatoid arthritis in the
30        absence of anti-cyclic citrullinated peptide antibodies: A prospective
31        study. Journal of Rheumatology. 2006; 33(9):1760-1765.

32     45 Verpoort KN, Jol-Van D, Papendrecht-Van D et al. Isotype distribution
33        of anti-cyclic citrullinated peptide antibodies in undifferentiated arthritis
34        and rheumatoid arthritis reflects an ongoing immune response. Arthritis
35        & Rheumatism. 2006; 54(12):3799-3808.

36     46 Wolfe F, Ross K, Hawley DJ et al. The prognosis of rheumatoid arthritis
37        and undifferentiated polyarthritis syndrome in the clinic: a study of 1141
38        patients. Journal of Rheumatology. 1993; 20(12):2005-2009.

39     47 Young A, Sumar N, Bodman K et al. Agalactosyl IgG: an aid to
40        differential diagnosis in early synovitis. Arthritis & Rheumatism. 1991;
41        34(11):1425-1429.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)               273 of 307
     DRAFT FOR CONSULTATION


1      48 Van der Cruyssen B, Hoffman IEA, Peene I et al. Prediction models for
2         rheumatoid arthritis during diagnostic investigation: Evaluation of
3         combinations of rheumatoid factor, anti-citrullinated protein/peptide
4         antibodies and the human leucocyte antigen-shared epitope. Annals of
5         the Rheumatic Diseases. 2007; 66(3):364-369.

6      49 Van Aken J, Lard LR, Le Cessie S et al. Radiological outcome after
7         four years of early versus delayed treatment strategy in patients with
8         recent onset rheumatoid arthritis. Annals of the Rheumatic Diseases.
9         2004; 63(3):274-279.

10     50 Lard LR, Visser H, Speyer I et al. Early versus delayed treatment in
11        patients with recent-onset rheumatoid arthritis: comparison of two
12        cohorts who received different treatment strategies. American Journal
13        of Medicine. 2001; 111(6):446-451.

14     51 Aletaha D, Ward MM. Duration of rheumatoid arthritis influences the
15        degree of functional improvement in clinical trials. Annals of the
16        Rheumatic Diseases. 2006; 65(2):227-233.

17     52 Bukhari M, Lunt M, Harrison BJ et al. Rheumatoid factor is the major
18        predictor of increasing severity of radiographic erosions in rheumatoid
19        arthritis: results from the Norfolk Arthritis Register Study, a large
20        inception cohort. Arthritis & Rheumatism. 2002; 46(4):906-912.

21     53 Dixey J, Solymossy C, Young A. Is It Possible to Predict Radiological
22        Damage in Early Rheumatoid Arthritis (RA)? A Report on the
23        Occurrence, Progression, and Prognostic Factors of Radiological
24        Erosions over the First 3 Years in 866 Patients from the Early RA
25        Study (ERAS). Journal of Rheumatology. 2004; 31(suppl 69):48-54.

26     54 Forslind K, Hafstrom I, Ahlmen M et al. Sex: a major predictor of
27        remission in early rheumatoid arthritis? Annals of the Rheumatic
28        Diseases. 2007; 66(1):46-52.

29     55 Forslind K, Ahlmen M, Eberhardt K et al. Prediction of radiological
30        outcome in early rheumatoid arthritis in clinical practice: Role of
31        antibodies to citrullinated peptides (anti-CCP). Annals of the Rheumatic
32        Diseases. 2004; 63(9):1090-1095.

33     56 Guillemin F, Gerard N, van LM et al. Prognostic factors for joint
34        destruction in rheumatoid arthritis: a prospective longitudinal study of
35        318 patients. Journal of Rheumatology. 2003; 30(12):2585-2589.

36     57 Van Jaarsveld CH, ter Borg EJ, Jacobs JW et al. The prognostic value
37        of the antiperinuclear factor, anti-citrullinated peptide antibodies and
38        rheumatoid factor in early rheumatoid arthritis. Clinical & Experimental
39        Rheumatology. 1999; 17(6):689-697.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)            274 of 307
     DRAFT FOR CONSULTATION


1      58 Kroot EJ, De Jong B, Van Leeuwen MA et al. The prognostic value of
2         anti-cyclic citrullinated peptide antibody in patients with recent-onset
3         rheumatoid arthritis. Arthritis & Rheumatism. 2000; 43(8):1831-1835.

4      59 Linn-Rasker SP, Van der Helm-van Mil AH, Breedveld FC et al.
5         Arthritis of the large joints - in particular, the knee - at first presentation
6         is predictive for a high level of radiological destruction of the small
7         joints in rheumatoid arthritis. Annals of the Rheumatic Diseases. 2007;
8         66(5):646-650.

 9     60 Odegard S, Landewe R, van d et al. Association of early radiographic
10        damage with impaired physical function in rheumatoid arthritis: A ten-
11        year, longitudinal observational study in 238 patients. Arthritis &
12        Rheumatism. 2006; 54(1):68-75.

13     61 Plant MJ, Williams AL, O'Sullivan MM et al. Relationship between time-
14        integrated C-reactive protein levels and radiologic progression in
15        patients with rheumatoid arthritis. Arthritis & Rheumatism. 2000;
16        43(7):1473-1477.

17     62 Priolo F, Bacarini L, Cammisa M et al. Radiographic changes in the
18        feet of patients with early rheumatoid arthritis. GRISAR (Gruppo
19        Reumatologi Italiani Studio Artrite Reumatoide). Journal of
20        Rheumatology. 1997; 24(11):2113-2118.

21     63 Ronnelid J, Wick MC, Lampa J et al. Longitudinal analysis of
22        citrullinated protein/peptide antibodies (anti-CP) during 5 year follow up
23        in early rheumatoid arthritis: anti-CP status predicts worse disease
24        activity and greater radiological progression. Annals of the Rheumatic
25        Diseases. 2005; 64(12):1744-1749.

26     64 Turesson C, O'Fallon WM, Crowson CS et al. Occurrence of
27        extraarticular disease manifestations is associated with excess
28        mortality in a community based cohort of patients with rheumatoid
29        arthritis. Journal of Rheumatology. 2002; 29(1):62-67.

30     65 Uhlig T, Smedstad LM, Vaglum P. The course of rheumatoid arthritis
31        and predictors of psychological, physical and radiographic outcome
32        after 5 years of follow-up. Rheumatology. 2000; 39(7):732-741.

33     66 Wallberg-Jonsson S, Johansson H, Ohman ML et al. Extent of
34        inflammation predicts cardiovascular disease and overall mortality in
35        seropositive rheumatoid arthritis. A retrospective cohort study from
36        disease onset. Journal of Rheumatology. 1999; 26(12):2562-2571.

37     67 Wolfe F, Sharp JT. Radiographic outcome of recent-onset rheumatoid
38        arthritis. A 19-year study of radiographic progression. Arthritis &
39        Rheumatism. 1998; 41(9):1571-1582.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)                 275 of 307
     DRAFT FOR CONSULTATION


1      68 Wolfe F, Michaud K, Gefeller O et al. Predicting mortality in patients
2         with rheumatoid arthritis. Arthritis & Rheumatism. 2003; 48(6):1530-
3         1542.

4      69 Young A, Bielawska C, Corbett M et al. A prospective study of early
5         onset rheumatoid arthritis over fifteen years: prognostic features and
6         outcome. Clinical Rheumatology. 1987; 6(suppl 2):12-19.

 7     70 Innala L, Kokkonen H, Eriksson C et al. Antibodies against mutated
 8        citrullinated vimentin are a better predictor of disease activity at 24
 9        months in early rheumatoid arthritis than antibodies against cyclic
10        citrullinated peptides. Journal of Rheumatology. 2008; 35(6):1002-
11        1008.

12     71 Mathsson L, Mullazehi M, Wick MC et al. Antibodies against
13        citrullinated vimentin in rheumatoid arthritis: Higher sensitivity and
14        extended prognostic value concerning future radiographic progression
15        as compared with antibodies against cyclic citrullinated peptides.
16        Arthritis & Rheumatism. 2008; 58(1):36-45.

17     72 Syversen SW, Gaarder PI, Goll GL et al. High anti-cyclic citrullinated
18        peptide levels and an algorithm of four variables predict radiographic
19        progression in patients with rheumatoid arthritis: Results from a 10-
20        year longitudinal study. Annals of the Rheumatic Diseases. 2008;
21        67(2):212-217.

22     73 De Vries-Bouwstra JK, Goekoop-Ruiterman YP, Verpoort KN et al.
23        Progression of joint damage in early rheumatoid arthritis: association
24        with HLA-DRB1, rheumatoid factor, and anti-citrullinated protein
25        antibodies in relation to different treatment strategies. Arthritis &
26        Rheumatism. 2008; 58(5):1293-1298.

27     74 Bas S, Perneger TV, Mikhnevitch E et al. Association of rheumatoid
28        factors and anti-filaggrin antibodies with severity of erosions in
29        rheumatoid arthritis. Rheumatology. 2000; 39(10):1082-1088.

30     75 Leigh JP, Fries JF. Mortality predictors among 263 patients with
31        rheumatoid arthritis. Journal of Rheumatology. 1991; 18(9):1307-1312.

32     76 Leigh JP, Fries JF. Predictors of disability in a longitudinal sample of
33        patients with rheumatoid arthritis. Annals of the Rheumatic Diseases.
34        1992; 51(5):581-587.

35     77 Plant MJ, O'Sullivan MM, Lewis PA et al. What factors influence
36        functional ability in patients with rheumatoid arthritis. Do they alter over
37        time? Rheumatology. 2005; 44(9):1181-1185.

38     78 Sherrer YS, Bloch DA, Mitchell DM et al. Disability in rheumatoid
39        arthritis: comparison of prognostic factors across three populations.
40        Journal of Rheumatology. 1987; 14(4):705-709.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              276 of 307
     DRAFT FOR CONSULTATION


1      79 Sihvonen S, Korpela M, Mustila A et al. The predictive value of
2         rheumatoid factor isotypes, anti-cyclic citrullinated peptide antibodies,
3         and antineutrophil cytoplasmic antibodies for mortality in patients with
4         rheumatoid arthritis. Journal of Rheumatology. 2005; 32(11):2089-
5         2094.

6      80 Strating MM, Van Schuur WH, Suurmeijer TP. Predictors of functional
7         disability in rheumatoid arthritis: results from a 13-year prospective
8         study. Disability & Rehabilitation. 2007; 29(10):805-815.

 9     81 Turesson C, McClelland RL, Christianson TJH et al. Severe extra-
10        articular disease manifestations are associated with an increased risk
11        of first ever cardiovascular events in patients with rheumatoid arthritis.
12        Annals of the Rheumatic Diseases. 2007; 66(1):70-75.

13     82 Wolfe F, Zwillich SH. The long-term outcomes of rheumatoid arthritis: a
14        23-year prospective, longitudinal study of total joint replacement and its
15        predictors in 1,600 patients with rheumatoid arthritis. Arthritis &
16        Rheumatism. 1998; 41(6):1072-1082.

17     83 Mancarella L, Bobbio PF, Ceccarelli F et al. Good clinical response,
18        remission, and predictors of remission in rheumatoid arthritis patients
19        treated with tumor necrosis factor-alpha blockers: the GISEA
20        study.[erratum appears in J Rheumatol. 2007 Sep;34(9):1947]. Journal
21        of Rheumatology. 2007; 34(8):1670-1673.

22     84 Proudman SM, Conaghan PG, Richardson C et al. Treatment of poor-
23        prognosis early rheumatoid arthritis. A randomized study of treatment
24        with methotrexate, cyclosporin A, and intraarticular corticosteroids
25        compared with sulfasalazine alone. Arthritis & Rheumatism. 2000;
26        43(8):1809-1819.

27     85 Quinn MA, Conaghan PG, O'Connor PJ et al. Very early treatment with
28        infliximab in addition to methotrexate in early, poor-prognosis
29        rheumatoid arthritis reduces magnetic resonance imaging evidence of
30        synovitis and damage, with sustained benefit after infliximab
31        withdrawal: results from a twelve-month randomized, double-blind,
32        placebo-controlled trial. Arthritis & Rheumatism. 2005; 52(1):27-35.

33     86 Van Dongen H, Van Aken J., Lard LR et al. Efficacy of methotrexate
34        treatment in patients with probable rheumatoid arthritis: A double-blind,
35        randomized, placebo-controlled trial. Arthritis & Rheumatism. 56; 1424-
36        1432. 2008.

37     87 Blalock SJ, Orlando M, Mutran EJ et al. Effect of satisfaction with one's
38        abilities on positive and negative affect among individuals with recently
39        diagnosed rheumatoid arthritis. Arthritis Care & Research. 1998;
40        11(3):158-165.

41     88 Nagyova I, Stewart RE, Macejova Z et al. The impact of pain on
42        psychological well-being in rheumatoid arthritis: the mediating effects of

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             277 of 307
     DRAFT FOR CONSULTATION


1          self-esteem and adjustment to disease. Patient Education &
2          Counseling. 2005; 58(1):55-62.

3      89 Smedstad LM, Kvien TK, Moum T et al. Correlates of patients' global
4         assessment of arthritis impact. A 2-year study of 216 patients with RA.
5         Scandinavian Journal of Rheumatology. 1997; 26(4):259-265.

6      90 Suurmeijer TP, Waltz M, Moum T et al. Quality of life profiles in the first
7         years of rheumatoid arthritis: results from the EURIDISS longitudinal
8         study. Arthritis & Rheumatism. 2001; 45(2):111-121.

 9     91 Thyberg I, Skogh T, Hass UAM et al. Recent-onset rheumatoid
10        arthritis: A 1-year observational study of correlations between health-
11        related quality of life and clinical/laboratory data. Journal of
12        Rehabilitation Medicine. 2005; 37(3):159-165.

13     92 Wong M, Mulherin D, Sousa KH. The influence of medication beliefs
14        and other psychosocial factors on early discontinuation of disease-
15        modifying anti-rheumatic drugs. Musculoskeletal Care. 2007; 5(3):148-
16        159.

17     93 Goekoop-Ruiterman YP. Patient preferences for treatment: report from
18        a randomised comparison of treatment strategies in early rheumatoid
19        arthritis (BeSt trial). Annals of the Rheumatic Diseases. 2007;
20        66(9):1227-1232.

21     94 Bath J, Hooper J, Giles M et al. Patient perceptions of rheumatoid
22        arthritis. Nursing Standard. 1999; 14(3):35-38.

23     95 Carr A, Hewlett S, Hughes R et al. Rheumatology outcomes: the
24        patient's perspective. Journal of Rheumatology. 2003; 30(4):880-883.

25     96 Covic T, Adamson B, Hough M. the impact of passive coping on
26        rheumatoid arthritis pain. Rheumatology. 2000; 39(9):1027.

27     97 Heiberg T, Kvien TK. Preferences for improved health examined in
28        1,024 patients with rheumatoid arthritis: pain has highest priority.
29        Arthritis & Rheumatism. 2002; 47(4):391-397.

30     98 Iaquinta ML, Larrabee JH. Phenomenological lived experience of
31        patients with rheumatoid arthritis. Journal of Nursing Care Quality.
32        2004; 19(3):280-289.

33     99 Jacobi CE, Boshuizen HC, Rupp I et al. Quality of rheumatoid arthritis
34        care: the patient's perspective. International Journal for Quality in
35        Health Care. 2004; 16(1):73-81.

36    100 Katz PP, Morris A, Yelin EH. Prevalence and predictors of disability in
37        valued life activities among individuals with rheumatoid arthritis. Annals
38        of the Rheumatic Diseases. 2006; 65(6):763-769.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)             278 of 307
     DRAFT FOR CONSULTATION


1     101 Kjeken I, Dagfinrud H, Mowinckel P et al. Rheumatology care:
2         Involvement in medical decisions, received information, satisfaction
3         with care, and unmet health care needs in patients with rheumatoid
4         arthritis and ankylosing spondylitis. Arthritis & Rheumatism. 2006;
5         55(3):394-401.

6     102 Neame R, Hammond A, Deighton C. Need for information and for
7         involvement in decision making among patients with rheumatoid
8         arthritis: a questionnaire survey. Arthritis & Rheumatism. 2005;
9         53(2):249-255.

10    103 Neugebauer A, Katz PP, Pasch LA. Effect of valued activity disability,
11        social comparisons, and satisfaction with ability on depressive
12        symptoms in rheumatoid arthritis. Health Psychology. 2003; 22(3):253-
13        262.

14    104 Rupp I, Boshuizen HC, Jacobi CE et al. Impact of fatigue on health-
15        related quality of life in rheumatoid arthritis. Arthritis & Rheumatism.
16        2004; 51(4):578-585.

17    105 Rupp I, Boshuizen HC, Roorda LD et al. Poor and good health
18        outcomes in rheumatoid arthritis: the role of comorbidity. Journal of
19        Rheumatology. 2006; 33(8):1488-1495.

20    106 Rupp I, Boshuizen HC, Dinant HJ et al. Disability and health-related
21        quality of life among patients with rheumatoid arthritis: association with
22        radiographic joint damage, disease activity, pain, and depressive
23        symptoms. Scandinavian Journal of Rheumatology. 2006; 35(3):175-
24        181.

25    107 Slatkowsky-Christensen B, Mowinckel P, Loge JH et al. Health-related
26        quality of life in women with symptomatic hand osteoarthritis: A
27        comparison with rheumatoid arthritis patients, healthy controls, and
28        normative data. Arthritis Care & Research. 2007; 57(8):1404-1409.

29    108 Katz PP, Morris A. Use of accommodations for valued life activities:
30        prevalence and effects on disability scores. Arthritis & Rheumatism.
31        2007; 57(5):730-737.

32    109 Pouchot J, Le Parc JM, Queffelec L et al. Perceptions in 7700 patients
33        with rheumatoid arthritis compared to their families and physicians.
34        Joint, Bone, Spine: Revue du Rhumatisme. 2007; 74(6):622-626.

35    110 Reinseth L, Espnes GA. Women with rheumatoid arthritis: non-
36        vocational activities and quality of life. SCAND J OCCUP THER. 2007;
37        14(2):108-115.

38    111 Sterba KR, DeVellis RF, Lewis MA et al. Effect of couple illness
39        perception congruence on psychological adjustment in women with
40        rheumatoid arthritis. Health Psychology. 2008; 27(2):221-229.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)              279 of 307
     DRAFT FOR CONSULTATION


1     112 Strating MMH, van Duijn MAJ, Van Schuur WH et al. The differential
2         effects of rheumatoid arthritis on distress among patients and partners.
3         Psychology & Health. 2007; 22(3):361-379.

4     113 Ward V, Hill J, Hale C et al. Patient priorities of care in rheumatology
5         outpatient clinics: a qualitative study. Musculoskeletal Care. 2007;
6         5(4):216-228.

 7    114 Uhlig T, Loge JH, Kristiansen IS et al. Quantification of reduced health-
 8        related quality of life in patients with rheumatoid arthritis compared to
 9        the general population. Journal of Rheumatology. 2007; 34(6):1241-
10        1247.

11    115 Williams AE, Nester CJ, Ravey MI. Rheumatoid arthritis patients'
12        experiences of wearing therapeutic footwear - A qualitative
13        investigation. BMC Musculoskeletal Disorders. 2007; 8

14    116 Verduin PJ, de Bock GH, Vliet Vlieland TP et al. Purpose in life in
15        patients with rheumatoid arthritis. Clinical Rheumatology. 2008;
16        27(7):899-908.

17    117 Wolfe F, Michaud K. Resistance of rheumatoid arthritis patients to
18        changing therapy: discordance between disease activity and patients'
19        treatment choices. Arthritis & Rheumatism. 2007; 56(7):2135-2142.

20    118 Barlow JH, Cullen LA, Rowe IF. Comparison of knowledge and
21        psychological well-being between patients with a short disease
22        duration (< or = 1 year) and patients with more established rheumatoid
23        arthritis (> or = 10 years duration). Patient Education & Counseling.
24        1999; 38(3):195-203.

25    119 Chilton F, Collett RA. Treatment choices, preferences and decision-
26        making by patients with rheumatoid arthritis. Musculoskeletal Care.
27        2008; 6(1):1-14.

28    120 Stamm T, Lovelock L, Stew G et al. I have mastered the challenge of
29        living with a chronic disease: Life stories of people with rheumatoid
30        arthritis. Qualitative Health Research. 2008; 18(5):658-669.

31    121 Treharne GJ, Lyons AC, Booth DA et al. Psychological well-being
32        across 1 year with rheumatoid arthritis: coping resources as buffers of
33        perceived stress. British Journal of Health Psychology. 2007;
34        12(Pt:3):3-45.

35    122 Veehof MM, Taal E, Willems MJ et al. Determinants of the use of wrist
36        working splints in rheumatoid arthritis. Arthritis Care & Research. 2008;
37        59(4):531-536.

38    123 Lempp H, Scott D, Kingsley G. The personal impact of rheumatoid
39        arthritis on patients' identity: a qualitative study. Chronic Illness. 2006;
40        2(2):109-120.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               280 of 307
     DRAFT FOR CONSULTATION


1     124 Lempp H, Scott DL, Kingsley GH. Patients' views on the quality of
2         health care for rheumatoid arthritis. Rheumatology. 2006; 45(12):1522-
3         1528.

4     125 Burckhardt CS, Lorig K, Moncur C et al. Arthritis and musculoskeletal
5         patient education standards. Arthritis Foundation. Arthritis Care &
6         Research. 1994; 7(1):1-4.

7     126 Barlow JH. How to use education as an intervention in osteoarthritis.
8         Best Practice & Research in Clinical Rheumatology. 2001; 15(4):545-
9         558.

10    127 National Institute for Health and Clinical Effectiveness. Computerised
11        cognitive behaviour therapy for Depression and Anxiety. (TA97).
12        London: National Institute for Health and Clinical Effectiveness, 2006.

13    128 Barlow JH, Turner AP, Wright CC. A randomized controlled study of
14        the Arthritis Self-Management Programme in the UK. Health Education
15        Research. 2000; 15(6):665-680.

16    129 Barlow JH, Turner AP, Wright CC. Long-term outcomes of an arthritis
17        self-management programme. British Journal of Rheumatology. 1998;
18        37(12):1315-1319.

19    130 Riemsma RP, Kirwan JR, Taal E et al. Patient education for adults with
20        rheumatoid arthritis. COCHRANE DATABASE SYST REV.
21        2003;(2):CD003688.

22    131 Brus HL, van de Laar MA, Taal E et al. Effects of patient education on
23        compliance with basic treatment regimens and health in recent onset
24        active rheumatoid arthritis. Annals of the Rheumatic Diseases. 1998;
25        57(3):146-151.

26    132 Freeman K, Hammond A, Lincoln NB. Use of cognitive-behavioural
27        arthritis education programmes in newly diagnosed rheumatoid
28        arthritis. Clinical Rehabilitation. 2002; 16(8):828-836.

29    133 Hammond A, Freeman K. One-year outcomes of a randomized
30        controlled trial of an educational-behavioural joint protection
31        programme for people with rheumatoid arthritis. Rheumatology. 2001;
32        40(9):1044-1051.

33    134 Hammond A, Freeman K. The long-term outcomes from a randomized
34        controlled trial of an educational-behavioural joint protection
35        programme for people with rheumatoid arthritis. Clinical Rehabilitation.
36        2004; 18(5):520-528.

37    135 Hammond A, Young A, Kidao R. A randomised controlled trial of
38        occupational therapy for people with early rheumatoid arthritis. Annals
39        of the Rheumatic Diseases. 2004; 63(1):23-30.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)           281 of 307
     DRAFT FOR CONSULTATION


1     136 Barlow JH, Pennington DC, Bishop PE. Patient education leaflets for
2         people with rheumatoid arthritis: A controlled study. Psychology Health
3         & Medicine. 1997; 2(3):221-235.

4     137 Helliwell PS, O'Hara M, Holdsworth J et al. A 12-month randomized
5         controlled trial of patient education on radiographic changes and quality
6         of life in early rheumatoid arthritis. Rheumatology. 1999; 38(4):303-
7         308.

 8    138 Hill J, Bird H, Johnson S. Effect of patient education on adherence to
 9        drug treatment for rheumatoid arthritis: a randomised controlled trial.
10        Annals of the Rheumatic Diseases. 2001; 60(9):869-875.

11    139 Lindroth Y, Brattstrom M, Bellman I et al. A problem-based education
12        program for patients with rheumatoid arthritis: evaluation after three
13        and twelve months. Arthritis Care & Research. 1997; 10(5):325-332.

14    140 Riemsma RP, Taal E, Rasker JJ. Group education for patients with
15        rheumatoid arthritis and their partners. Arthritis & Rheumatism. 2003;
16        49(4):556-566.

17    141 Riemsma RP, Taal E, Brus HL et al. Coordinated individual education
18        with an arthritis passport for patients with rheumatoid arthritis. Arthritis
19        Care & Research. 1997; 10(4):238-249.

20    142 Van Lankveld W, van HT, Naring G et al. Partner participation in
21        cognitive-behavioral self-management group treatment for patients with
22        rheumatoid arthritis. Journal of Rheumatology. 2004; 31(9):1738-1745.

23    143 Walker D, Adebajo A, Heslop P et al. Patient education in rheumatoid
24        arthritis: the effectiveness of the ARC booklet and the mind map.
25        Rheumatology. 2007; 46(10):1593-1596.

26    144 Giraudet-Le Quintrec JS. Effect of a collective educational program for
27        patients with rheumatoid arthritis: a prospective 12-month randomized
28        controlled trial. Journal of Rheumatology. 2007; 34(8):1684-1691.

29    145 Masiero S, Boniolo A, Wassermann L et al. Effects of an educational-
30        behavioral joint protection program on people with moderate to severe
31        rheumatoid arthritis: a randomized controlled trial. Clinical
32        Rheumatology. 2007; 26(12):2043-2050.

33    146 Ahlmen M, Sullivan M, Bjelle A. Team versus non-team outpatient care
34        in rheumatoid arthritis. A comprehensive outcome evaluation including
35        an overall health measure. Arthritis & Rheumatism. 1988; 31(4):471-
36        479.

37    147 Vliet Vlieland TP, Breedveld FC, Hazes JM. The two-year follow-up of
38        a randomized comparison of in-patient multidisciplinary team care and
39        routine out-patient care for active rheumatoid arthritis. British Journal of
40        Rheumatology. 1997; 36(1):82-85.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)               282 of 307
     DRAFT FOR CONSULTATION


1     148 Vliet Vlieland TP, Zwinderman AH, Vandenbroucke JP et al. A
2         randomized clinical trial of in-patient multidisciplinary treatment versus
3         routine out-patient care in active rheumatoid arthritis. Rheumatology.
4         1996; 35(5):475-482.

5     149 Scholten C, Brodowicz T, Graninger W et al. Persistent functional and
6         social benefit 5 years after a multidisciplinary arthritis training program.
7         Archives of Physical Medicine & Rehabilitation. 1999; 80(10):1282-
8         1287.

 9    150 Schned ES, Doyle MA, Glickstein SL et al. Team managed outpatient
10        care for early onset chronic inflammatory arthritis. Journal of
11        Rheumatology. 1995; 22(6):1141-1148.

12    151 Feinberg JR, Brandt KD. Allied health team management of
13        rheumatoid arthritis patients. American Journal of Occupational
14        Therapy. 1984; 38(9):613-620.

15    152 Jacobsson LTH, Frithiof M, Olofsson Y et al. Evaluation of a structured
16        multidisciplinary day care program in rheumatoid arthritis.
17        Scandinavian Journal of Rheumatology. 1998; 27(2):117-124.

18    153 Prier A, Berenbaum F, Karneff A et al. Multidisciplinary day hospital
19        treatment of rheumatoid arthritis patients. Evaluation after two years.
20        Revue du Rhumatisme (English Edition). 1997; 64(7-9):443-450.

21    154 Nordmark B, Blomqvist P, Andersson B et al. A two-year follow-up of
22        work capacity in early rheumatoid arthritis: a study of multidisciplinary
23        team care with emphasis on vocational support. Scandinavian Journal
24        of Rheumatology. 2006; 35(1):7-14.

25    155 Van den Hout WB, Tijhuis GJ, Hazes JM et al. Cost effectiveness and
26        cost utility analysis of multidisciplinary care in patients with rheumatoid
27        arthritis: a randomised comparison of clinical nurse specialist care,
28        inpatient team care, and day patient team care. Annals of the
29        Rheumatic Diseases. 2003; 62(4):308-315.

30    156 Gabriel SE, Wagner JL, Zinsmeister AR et al. Is rheumatoid arthritis
31        care more costly when provided by rheumatologists compared with
32        generalists? Arthritis & Rheumatism. 2001; 44(7):1504-1514.

33    157 Li LC, Maetzel A, Davis AM et al. Primary therapist model for patients
34        referred for rheumatoid arthritis rehabilitation: a cost-effectiveness
35        analysis. Arthritis & Rheumatism. 2006; 55(3):402-410.

36    158 Sokka T, Hakkinen A, Kautiainen H et al. Physical inactivity in patients
37        with rheumatoid arthritis: data from twenty-one countries in a cross-
38        sectional, international study. Arthritis & Rheumatism. 2008; 59(1):42-
39        50.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              283 of 307
     DRAFT FOR CONSULTATION


1     159 Turesson C, Matteson EL. Cardiovascular risk factors, fitness and
2         physical activity in rheumatic diseases. Current Opinion in
3         Rheumatology. 2007; 19(2):190-196.

4     160 Huusko TM, Korpela M, Karppi P et al. Threefold increased risk of hip
5         fractures with rheumatoid arthritis in Central Finland. Annals of the
6         Rheumatic Diseases. 2001; 60(5):521-522.

 7    161 Hakkinen A, Sokka T, Kotaniemi A et al. Muscle strength
 8        characteristics and central bone mineral density in women with recent
 9        onset rheumatoid arthritis compared with healthy controls.
10        Scandinavian Journal of Rheumatology. 1999; 28(3):145-151.

11    162 American Academy of Orthopaedic Manual Physical Therapists. Home
12        page. Available from: American Academy of Orthopaedic Manual
13        Physical Therapists. Last accessed on: 2008 July 28.

14    163 World Confederation of Physical Therapists. World Confederation of
15        Physical Therapists: a message from the WCPT president. Available
16        from: World Confederation of Physical Therapists. Last accessed on:
17        2008 July 28.

18    164 Robinson V, Brosseau L, Casimiro L et al. Thermotherapy for treating
19        rheumatoid arthritis. COCHRANE DATABASE SYST REV.
20        2002;(2):CD002826.

21    165 Brosseau L, Welch V, Wells G et al. Low level laser therapy (classes I,
22        II and III) in the treatment of rheumatoid arthritis. COCHRANE
23        DATABASE SYST REV. 2005;(4):CD002049.

24    166 Brosseau L, Judd MG, Marchand S et al. Transcutaneous electrical
25        nerve stimulation (TENS) for the treatment of rheumatoid arthritis in the
26        hand. COCHRANE DATABASE SYST REV. 2003;(2):CD004377.

27    167 Casimiro L, Brosseau L, Robinson V et al. Therapeutic ultrasound for
28        the treatment of rheumatoid arthritis. COCHRANE DATABASE SYST
29        REV. 2002;(3):CD003787.

30    168 Han A, Robinson V, Judd M et al. Tai chi for treating rheumatoid
31        arthritis. COCHRANE DATABASE SYST REV. 2004;(3):CD004849.

32    169 Bearne LM, Scott DL, Hurley MV. Exercise can reverse quadriceps
33        sensorimotor dysfunction that is associated with rheumatoid arthritis
34        without exacerbating disease activity. Rheumatology. 2002; 41(2):157-
35        166.

36    170 Hall J, Skevington SM, Maddison PJ et al. A randomized and controlled
37        trial of hydrotherapy in rheumatoid arthritis. Arthritis Care & Research.
38        1996; 9(3):206-215.

39    171 Harris R, Millard JB. Paraffin-wax baths in the treatment of rheumatoid
40        arthritis. Annals of the Rheumatic Diseases. 1955; 14(3):278-282.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)           284 of 307
     DRAFT FOR CONSULTATION


1     172 Bell MJ, Lineker SC, Wilkins AL et al. A randomized controlled trial to
2         evaluate the efficacy of community based physical therapy in the
3         treatment of people with rheumatoid arthritis. Journal of Rheumatology.
4         1998; 25(2):231-237.

5     173 Lineker SC, Bell MJ, Wilkins AL et al. Improvements following short
6         term home based physical therapy are maintained at one year in
7         people with moderate to severe rheumatoid arthritis. Journal of
8         Rheumatology. 2001; 28(1):165-168.

 9    174 Buljina AI, Taljanovic MS, Avdic DM et al. Physical and exercise
10        therapy for treatment of the rheumatoid hand. Arthritis & Rheumatism.
11        2001; 45(4):392-397.

12    175 O'Brien AV, Jones P, Mullis R et al. Conservative hand therapy
13        treatments in rheumatoid arthritis--a randomized controlled trial.
14        Rheumatology. 2006; 45(5):577-583.

15    176 Eversden L. A pragmatic randomised controlled trial of hydrotherapy
16        and land exercises on overall well being and quality of life in
17        rheumatoid arthritis. BMC Musculoskeletal Disorders. 2007; 8:23-29.

18    177 Hoenig H, Groff G, Pratt K et al. A randomized controlled trial of home
19        exercise on the rheumatoid hand. Journal of Rheumatology. 1993;
20        20(5):785-789.

21    178 De Jong Z, Munneke M, Zwinderman AH et al. Long term high intensity
22        exercise and damage of small joints in rheumatoid arthritis. Annals of
23        the Rheumatic Diseases. 2004; 63(11):1399-1405.

24    179 De Jong Z, Munneke M, Lems WF et al. Slowing of bone loss in
25        patients with rheumatoid arthritis by long-term high-intensity exercise:
26        results of a randomized, controlled trial. Arthritis & Rheumatism. 2004;
27        50(4):1066-1076.

28    180 De Jong Z, Munneke M, Zwinderman AH et al. Is a long-term high-
29        intensity exercise program effective and safe in patients with
30        rheumatoid arthritis? Results of a randomized controlled trial. Arthritis &
31        Rheumatism. 2003; 48(9):2415-2424.

32    181 Neuberger GB, Aaronson LS, Gajewski B et al. Predictors of exercise
33        and effects of exercise on symptoms, function, aerobic fitness, and
34        disease outcomes of rheumatoid arthritis. Arthritis & Rheumatism.
35        2007; 57(6):943-952.

36    182 Hansen TM, Hansen G. Longterm physical training in rheumatoid
37        arthritis. A randomized trial with different training programs and blinded
38        observers. Scandinavian Journal of Rheumatology. 1993; 22(3):107-
39        112.

40    183 Van den Berg MH, Ronday HK, Peeters AJ et al. Using internet
41        technology to deliver a home-based physical activity intervention for

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             285 of 307
     DRAFT FOR CONSULTATION


1          patients with rheumatoid arthritis: A randomized controlled trial. Arthritis
2          & Rheumatism. 2006; 55(6):935-945.

3     184 Hakkinen A, Sokka T, Hannonen P. A home-based two-year strength
4         training period in early rheumatoid arthritis led to good long-term
5         compliance: a five-year followup. Arthritis & Rheumatism. 2004;
6         51(1):56-62.

 7    185 Hakkinen A, Sokka T, Lietsalmi AM et al. Effects of dynamic strength
 8        training on physical function, Valpar 9 work sample test, and working
 9        capacity in patients with recent-onset rheumatoid arthritis. Arthritis &
10        Rheumatism. 2003; 49(1):71-77.

11    186 Hakkinen A, Sokka T, Kotaniemi A et al. A randomized two-year study
12        of the effects of dynamic strength training on muscle strength, disease
13        activity, functional capacity, and bone mineral density in early
14        rheumatoid arthritis. Arthritis & Rheumatism. 2001; 44(3):515-522.

15    187 Hakkinen A, Sokka T, Kautiainen H et al. Sustained maintenance of
16        exercise induced muscle strength gains and normal bone mineral
17        density in patients with early rheumatoid arthritis: a 5 year follow up.
18        Annals of the Rheumatic Diseases. 2004; 63(8):910-916.

19    188 Van den Ende CH, Hazes JM, Le CS et al. Comparison of high and low
20        intensity training in well controlled rheumatoid arthritis. Results of a
21        randomised clinical trial. Annals of the Rheumatic Diseases. 1996;
22        55(11):798-805.

23    189 Van den Ende CH, Breedveld FC, Le CS et al. Effect of intensive
24        exercise on patients with active rheumatoid arthritis: a randomised
25        clinical trial. Annals of the Rheumatic Diseases. 2000; 59(8):615-621.

26    190 Brodin N, Eurenius E, Jensen I et al. Coaching patients with early
27        rheumatoid arthritis to healthy physical activity: A multicenter,
28        randomized, controlled study. Arthritis Care & Research. 2008;
29        59(3):325-331.

30    191 Hirvonen HE, Mikkelsson MK, Kautiainen H et al. Effectiveness of
31        different cryotherapies on pain and disease activity in active
32        rheumatoid arthritis. A randomised single blinded controlled trial.
33        Clinical & Experimental Rheumatology. 2006; 24(3):295-301.

34    192 Munneke M, de JZ, Zwinderman AH et al. Effect of a high-intensity
35        weight-bearing exercise program on radiologic damage progression of
36        the large joints in subgroups of patients with rheumatoid arthritis.
37        Arthritis & Rheumatism. 2005; 53(3):410-417.

38    193 Van den Hout WB, de JZ, Munneke M et al. Cost-utility and cost-
39        effectiveness analyses of a long-term, high-intensity exercise program
40        compared with conventional physical therapy in patients with
41        rheumatoid arthritis. Arthritis & Rheumatism. 2005; 53(1):39-47.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              286 of 307
     DRAFT FOR CONSULTATION


1     194 Eberhardt KB, Fex E. Functional impairment and disability in early
2         rheumatoid arthritis--development over 5 years. Journal of
3         Rheumatology. 1995; 22(6):1037-1042.

4     195 Reisine ST, Goodenow C, Grady KE. The impact of rheumatoid
5         arthritis on the homemaker. Social Science & Medicine. 1987;
6         25(1):89-95.

 7    196 Jordan JM, Bernard SL, Callahan LF et al. Self-reported arthritis-
 8        related disruptions in sleep and daily life and the use of medical,
 9        complementary, and self-care strategies for arthritis: the National
10        Survey of Self-care and Aging. Archives of Family Medicine. 2000;
11        9(2):143-149.

12    197 Katz PP, Yelin EH. Life activities of persons with rheumatoid arthritis
13        with and without depressive symptoms. Arthritis Care & Research.
14        1994; 7(2):69-77.

15    198 Fex E, Larsson BM, Nived K et al. Effect of rheumatoid arthritis on work
16        status and social and leisure time activities in patients followed 8 years
17        from onset. Journal of Rheumatology. 1998; 25(1):44-50.

18    199 Astin JA, Beckner W, Soeken K et al. Psychological interventions for
19        rheumatoid arthritis: a meta-analysis of randomized controlled trials.
20        Arthritis & Rheumatism. 2002; 47(3):291-302.

21    200 Egan M, Brosseau L, Farmer M et al. Splints and orthoses in the
22        treatment of rheumatoid arthritis. COCHRANE DATABASE SYST REV.
23        2001;(4):CD004018.

24    201 Helewa A, Goldsmith CH, Lee P et al. Effects of occupational therapy
25        home service on patients with rheumatoid arthritis. Lancet. 1991;
26        337(8755):1453-1456.

27    202 Haskett S, Backman C, Porter B et al. A crossover trial of custom-
28        made and commercially available wrist splints in adults with
29        inflammatory arthritis. Arthritis & Rheumatism. 2004; 51(5):792-799.

30    203 Evers AW, Kraaimaat FW, van Riel PL et al. Tailored cognitive-
31        behavioral therapy in early rheumatoid arthritis for patients at risk: a
32        randomized controlled trial. Pain. 2002; 100(1-2):141-153.

33    204 Parker JC, Frank RG, Beck NC et al. Pain management in rheumatoid
34        arthritis patients. A cognitive-behavioral approach. Arthritis &
35        Rheumatism. 1988; 31(5):593-601.

36    205 Pradhan EK, Baumgarten M, Langenberg P et al. Effect of
37        Mindfulness-Based Stress Reduction in rheumatoid arthritis patients.
38        Arthritis & Rheumatism. 2007; 57(7):1134-1142.

39    206 Sharpe L, Sensky T, Timberlake N et al. Long-term efficacy of a
40        cognitive behavioural treatment from a randomized controlled trial for

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              287 of 307
     DRAFT FOR CONSULTATION


1          patients recently diagnosed with rheumatoid arthritis. Rheumatology.
2          2003; 42(3):435-441.

3     207 Sharpe L, Allard S, Sensky T. Five-year followup of a cognitive-
4         behavioral intervention for patients with recently-diagnosed rheumatoid
5         arthritis: Effects on health care utilization. Arthritis Care & Research.
6         2008; 59(3):311-316.

 7    208 Zautra AJ, Davis MC, Reich JW et al. Comparison of cognitive
 8        behavioral and mindfulness meditation interventions on adaptation to
 9        rheumatoid arthritis for patients with and without history of recurrent
10        depression. Journal of Consulting & Clinical Psychology. 2008;
11        76(3):408-421.

12    209 Kjeken I, Moller G, Kvien TK. Use of commercially produced elastic
13        wrist orthoses in chronic arthritis: a controlled study. Arthritis Care &
14        Research. 1995; 8(2):108-113.

15    210 National Institute for Health and Clinical Excellence. Depression(
16        amended):management of depression in primary and secondary care.
17        (CG23 (amended)). London: NICE, 2007.

18    211 Michelson J, Easley M, Wigley FM et al. Foot and ankle problems in
19        rheumatoid arthritis. Foot & Ankle International. 1994; 15(11):608-613.

20    212 Wickman AM, Pinzur MS, Kadanoff R et al. Health-related quality of life
21        for patients with rheumatoid arthritis foot involvement. Foot & Ankle
22        International. 2004; 25(1):19-26.

23    213 Kerry RM, Holt GM, Stockley I. The foot in chronic rheumatoid arthritis:
24        a continuing problem. Foot. 1994; 4(4):201-203.

25    214 Redmond AC, Waxman R, Helliwell PS. Provision of foot health
26        services in rheumatology in the UK. Rheumatology. 2006; 45(5):571-
27        576.

28    215 Davys HJ, Turner DE, Helliwell PS et al. Debridement of plantar
29        callosities in rheumatoid arthritis: a randomized controlled trial.
30        Rheumatology. 2005; 44(2):207-210.

31    216 Woodburn J, Barker S, Helliwell PS. A randomized controlled trial of
32        foot orthoses in rheumatoid arthritis. Journal of Rheumatology. 2002;
33        29(7):1377-1383.

34    217 Williams AE, Rome K, Nester CJ. A clinical trial of specialist footwear
35        for patients with rheumatoid arthritis. Rheumatology. 2007; 46(2):302-
36        307.

37    218 Woodburn J, Helliwell PS, Barker S. Changes in 3D joint kinematics
38        support the continuous use of orthoses in the management of painful
39        rearfoot deformity in rheumatoid arthritis. Journal of Rheumatology.
40        2003; 30(11):2356-2364.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              288 of 307
     DRAFT FOR CONSULTATION


1     219 Moncur C, Ward JR. Heat-moldable shoes for management of forefoot
2         problems in rheumatoid arthritis. Arthritis Care & Research. 1990;
3         3(4):222-226.

4     220 Verstappen SM, Jacobs JW, Kruize AA et al. Trends in economic
5         consequences of rheumatoid arthritis over two subsequent years.
6         Rheumatology. 2007; 46(6):968-974.

7     221 Nordenskiold U. Evaluation of assistive devices after a course in joint
8         protection. International Journal of Technology Assessment in Health
9         Care. 1994; 10(2):293-304.

10    222 Clark H, Rome K, Plant M et al. A critical review of foot orthoses in the
11        rheumatoid arthritic foot. Rheumatology. 2006; 45(2):139-145.

12    223 Chalmers AC, Busby C, Goyert J et al. Metatarsalgia and rheumatoid
13        arthritis--a randomized, single blind, sequential trial comparing 2 types
14        of foot orthoses and supportive shoes. Journal of Rheumatology. 2000;
15        27(7):1643-1647.

16    224 Glowinski J, Boccard E. Placebo-controlled study of the analgesic
17        efficacy of a paracetamol 500 mg/Codeine 30 mg combination together
18        with low-dose vs high-dose diclofenac in rheumatoid arthritis. Clinical
19        Drug Investigation. 1999; 18(3):189-197.

20    225 Herrero-Beaumont G, Bjorneboe O, Richarz U. Transdermal fentanyl
21        for the treatment of pain caused by rheumatoid arthritis. Rheumatology
22        International. 2004; 24(6):325-332.

23    226 Berliner MN, Giesecke T, Bornhovd KD. Impact of transdermal fentanyl
24        on quality of life in rheumatoid arthritis. Clinical Journal of Pain. 2007;
25        23(6):530-534.

26    227 Seideman P. Additive effect of combined naproxen and paracetamol in
27        rheumatoid arthritis. British Journal of Rheumatology. 1993;
28        32(12):1077-1082.

29    228 Seideman P, Melander A. Equianalgesic effects of paracetamol and
30        indomethacin in rheumatoid arthritis. British Journal of Rheumatology.
31        1988; 27(2):117-122.

32    229 Frank RG, Kashani JH, Parker JC et al. Antidepressant analgesia in
33        rheumatoid arthritis. Journal of Rheumatology. 1988; 15(11):1632-
34        1638.

35    230 Emery P, Gibson T. A double-blind study of the simple analgesic
36        nefopam in rheumatoid arthritis. British Journal of Rheumatology. 1986;
37        25(1):72-76.

38    231 Grace EM, Bellamy N, Kassam Y et al. Controlled, double-blind,
39        randomized trial of amitriptyline in relieving articular pain and


     Rheumatoid arthritis: full guideline DRAFT (August 2008)             289 of 307
     DRAFT FOR CONSULTATION


1          tenderness in patients with rheumatoid arthritis. Current Medical
2          Research & Opinion. 1985; 9(6):426-429.

3     232 Blake DR, Robson P, Ho M et al. Preliminary assessment of the
4         efficacy, tolerability and safety of a cannabis-based medicine (Sativex)
5         in the treatment of pain caused by rheumatoid arthritis. Rheumatology.
6         2006; 45(1):50-52.

 7    233 National Institute for Health and Clinical Effectiveness. Guidance on
 8        the use of cyclo-oxygenase (cox) II selective
 9        inhibitors,celecoxib,rofecoxib,meloxicam and etodolac for osteoarthritis
10        and rheumatoid arthritis. (TA27). London: NICE, 2001.

11    234 Geusens PP, Truitt K, Sfikakis P et al. A placebo and active
12        comparator-controlled trial of rofecoxib for the treatment of rheumatoid
13        arthritis. Scandinavian Journal of Rheumatology. 2002; 31(4):230-238.

14    235 Porzio F. Meta-analysis of two double-blind comparative studies with
15        the sustained-release form of etodolac in rheumatoid arthritis.
16        Rheumatology International. 1993; 13(2 suppl):1-30.

17    236 Shi W. Safety and efficacy of oral nonsteroidal anti-inflammatory drugs
18        in patients with rheumatoid arthritis: a six-month randomised study.
19        Clinical Drug Investigation. 2004; 24(2):89-101.

20    237 Collantes E, Curtis SP, Lee KW et al. A multinational randomized,
21        controlled, clinical trial of etoricoxib inthetreatment of rheumatoid
22        arthritis. BMC Family Practice. 2002; 3:1-10.

23    238 Matsumoto AK, Melian A, Mandel DR et al. A randomized, controlled,
24        clinical trial of etoricoxib in the treatment of rheumatoid arthritis. Journal
25        of Rheumatology. 2002; 29(8):1623-1630.

26    239 Bensen W, Weaver A, Espinoza L et al. Efficacy and safety of
27        valdecoxib in treating the signs and symptoms of rheumatoid arthritis: a
28        randomized, controlled comparison with placebo and naproxen.
29        Rheumatology. 2002; 41(9):1008-1016.

30    240 Geusens P, Alten R, Rovensky J et al. Efficacy, safety and tolerability
31        of lumiracoxib in patients with rheumatoid arthritis. International Journal
32        of Clinical Practice. 2004; 58(11):1033-1041.

33    241 Gibofsky A, Rodrigues J, Fiechtner J et al. Efficacy and tolerability of
34        valdecoxib in treating the signs and symptoms of severe rheumatoid
35        arthritis: a 12-week, multicenter, randomized, double-blind, placebo-
36        controlled study. Clinical Therapeutics. 2007; 29(6):1071-1085.

37    242 Krug H, Broadwell LK, Berry M et al. Tolerability and efficacy of
38        nabumetone and naproxen in the treatment of rheumatoid arthritis.
39        Clinical Therapeutics. 2000; 22(1):40-52.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)               290 of 307
     DRAFT FOR CONSULTATION


1     243 Simon LS, Weaver AL, Graham DY et al. Anti-inflammatory and upper
2         gastrointestinal effects of celecoxib in rheumatoid arthritis: a
3         randomized controlled trial. Journal of the American Medical
4         Association. 1999; 282(20):1921-1928.

5     244 Williams GW, Kivitz AJ, Brown MT et al. A comparison of valdecoxib
6         and naproxen in the treatment of rheumatoid arthritis symptoms.
7         Clinical Therapeutics. 2006; 28(2):204-221.

 8    245 Zhao SZ, Fiechtner JI, Tindall EA et al. Evaluation of health-related
 9        quality of life of rheumatoid arthritis patients treated with celecoxib.
10        Arthritis Care & Research. 2000; 13(2):112-121.

11    246 Matsumoto A, Melian A, Shah A et al. Etoricoxib versus naproxen in
12        patients with rheumatoid arthritis: a prospective, randomized,
13        comparator-controlled 121-week trial. Current Medical Research &
14        Opinion. 2007; 23(9):2259-2268.

15    247 Brown TJ, Hooper L, Elliott RA et al. A comparison of the cost-
16        effectiveness of five strategies for the prevention of non-steroidal anti-
17        inflammatory drug-induced gastrointestinal toxicity: a systematic review
18        with economic modelling. Health Technology Assessment. 2006;
19        10(38):iii-xiii, 1.

20    248 Spiegel BM, Targownik L, Dulai GS et al. The cost-effectiveness of
21        cyclooxygenase-2 selective inhibitors in the management of chronic
22        arthritis. ANN INTERN MED. 2003; 138(10):795-806.

23    249 Svarvar P, Aly A. Use of the ACCES model to predict the health
24        economic impact of celecoxib in patients with osteoarthritis or
25        rheumatoid arthritis in Norway. Rheumatology. 2000; 39(suppl 2):43-
26        50.

27    250 McCabe CJ, Akehurst RL, Kirsch J et al. Choice of NSAID and
28        management strategy in rheumatoid arthritis and osteoarthritis. The
29        impact on costs and outcomes in the UK. Pharmacoeconomics. 1998;
30        14(2):191-199.

31    251 Maetzel A, Krahn M, Naglie G. The cost effectiveness of rofecoxib and
32        celecoxib in patients with osteoarthritis or rheumatoid arthritis. Arthritis
33        & Rheumatism. 2003; 49(3):283-292.

34    252 You JH, Lee KK, Chan TY et al. Arthritis treatment in Hong Kong--cost
35        analysis of celecoxib versus conventional NSAIDS, with or without
36        gastroprotective agents. Alimentary Pharmacology & Therapeutics.
37        2002; 16(12):2089-2096.

38    253 Zabinski RA, Burke TA, Johnson J et al. An economic model for
39        determining the costs and consequences of using various treatment
40        alternatives for the management of arthritis in Canada.
41        Pharmacoeconomics. 2001; 19(suppl 1):49-58.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              291 of 307
     DRAFT FOR CONSULTATION


1     254 Fendrick AM, Bandekar RR, Chernew ME et al. Role of initial NSAID
2         choice and patient risk factors in the prevention of NSAID gastropathy:
3         a decision analysis. Arthritis & Rheumatism. 2002; 47(1):36-43.

4     255 Chancellor JV, Hunsche E, de CE et al. Economic evaluation of
5         celecoxib, a new cyclo-oxygenase 2 specific inhibitor, in Switzerland.
6         Pharmacoeconomics. 2001; 19(suppl 1):59-75.

 7    256 El-Serag HB, Graham DY, Richardson P et al. Prevention of
 8        complicated ulcer disease among chronic users of nonsteroidal anti-
 9        inflammatory drugs: the use of a nomogram in cost-effectiveness
10        analysis. Archives of Internal Medicine. 2002; 162(18):2105-2110.

11    257 Hansen M, Podenphant J, Florescu A et al. A randomised trial of
12        differentiated prednisolone treatment in active rheumatoid arthritis.
13        Clinical benefits and skeletal side effects. Annals of the Rheumatic
14        Diseases. 1999; 58(11):713-718.

15    258 Corkill MM, Kirkham BW, Chikanza IC et al. Intramuscular depot
16        methylprednisolone induction of chrysotherapy in rheumatoid arthritis:
17        a 24-week randomized controlled trial. British Journal of
18        Rheumatology. 1990; 29(4):274-279.

19    259 Capell HA, Madhok R, Hunter JA et al. Lack of radiological and clinical
20        benefit over two years of low dose prednisolone for rheumatoid
21        arthritis: results of a randomised controlled trial. Annals of the
22        Rheumatic Diseases. 2004; 63(7):797-803.

23    260 Hansen TM, Kryger P, Elling H et al. Double blind placebo controlled
24        trial of pulse treatment with methylprednisolone combined with disease
25        modifying drugs in rheumatoid arthritis. British Medical Journal. 1990;
26        301(6746):268-270.

27    261 Hickling P, Jacoby RK, Kirwan JR et al. Joint destruction after
28        glucocorticoids are withdrawn in early rheumatoid arthritis. British
29        Journal of Rheumatology. 1998; 37(9):930-936.

30    262 Kirwan JR, Byron M, Dieppe P et al. The effect of glucocorticoids on
31        joint destruction in rheumatoid arthritis. New England Journal of
32        Medicine. 1995; 333(3):142-146.

33    263 Svensson B, Boonen A, Albertsson K et al. Low-dose prednisolone in
34        addition to the initial disease-modifying antirheumatic drug in patients
35        with early active rheumatoid arthritis reduces joint destruction and
36        increases the remission rate: A two-year randomized trial. Arthritis &
37        Rheumatism. 2005; 52(11):3360-3370.

38    264 Van Everdingen AA, Jacobs JW, Siewertsz van Reesema DR et al.
39        Low-dose prednisone therapy for patients with early active rheumatoid
40        arthritis: clinical efficacy, disease-modifying properties, and side



     Rheumatoid arthritis: full guideline DRAFT (August 2008)            292 of 307
     DRAFT FOR CONSULTATION


1          effects: a randomized, double-blind, placebo-controlled clinical trial.
2          ANN INTERN MED. 2002; 136(1):1-12.

3     265 Van Everdingen AA, Siewertsz van Reesema DR, Jacobs JW et al.
4         The clinical effect of glucocorticoids in patients with rheumatoid arthritis
5         may be masked by decreased use of additional therapies. Arthritis &
6         Rheumatism. 2004; 51(2):233-238.

 7    266 Wassenberg S, Rau R, Steinfeld P et al. Very low-dose prednisolone in
 8        early rheumatoid arthritis retards radiographic progression over two
 9        years: A multicenter, double-blind, placebo-controlled trial. Arthritis &
10        Rheumatism. 2005; 52(11):3371-3380.

11    267 Choy EH, Kingsley GH, Khoshaba B et al. A two year randomised
12        controlled trial of intramuscular depot steroids in patients with
13        established rheumatoid arthritis who have shown an incomplete
14        response to disease modifying antirheumatic drugs. Annals of the
15        Rheumatic Diseases. 2005; 64(9):1288-1293.

16    268 Kirwan JR, Hallgren R, Mielants H et al. A randomised placebo
17        controlled 12 week trial of budesonide and prednisolone in rheumatoid
18        arthritis. Annals of the Rheumatic Diseases. 2004; 63(6):688-695.

19    269 Van Vliet-Daskalopoulu E, Jentjens T, Scheffer RTC. Intra-articular
20        rimexolone in the rheumatoid knee: A placebo-controlled, double-blind,
21        multicentre trial of three doses. British Journal of Rheumatology. 1987;
22        26(6):450-453.

23    270 Bae SC, Corzillius M, Kuntz KM et al. Cost-effectiveness of low dose
24        corticosteroids versus non-steroidal anti-inflammatory drugs and COX-
25        2 specific inhibitors in the long-term treatment of rheumatoid arthritis.
26        Rheumatology. 2003; 42(1):46-53.

27    271 Verhoeven AC, Bibo JC, Boers M et al. Cost-effectiveness and cost-
28        utility of combination therapy in early rheumatoid arthritis: randomized
29        comparison of combined step-down prednisolone, methotrexate and
30        sulphasalazine with sulphasalazine alone. COBRA Trial Group. British
31        Journal of Rheumatology. 1998; 37(10):1102-1109.

32    272 Finckh A, Liang MH, van Herckenrode CM et al. Long-term impact of
33        early treatment on radiographic progression in rheumatoid arthritis: A
34        meta-analysis. Arthritis & Rheumatism. 2006; 55(6):864-872.

35    273 Anon. A randomized trial of hydroxychloroquine in early rheumatoid
36        arthritis: the HERA Study. American Journal of Medicine. 1995;
37        98(2):156-168.

38    274 Tsakona E, Fitzgerald AA. Consequences of delayed therapy with
39        second-line agents in rheumatoid arthritis: a 3 year followup on the
40        hydroxychloroquine in early rheumatoid arthritis (HERA) study. Journal
41        of Rheumatology. 2000; 27(3):623-629.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              293 of 307
     DRAFT FOR CONSULTATION


1     275 Borg G, Allander E, Lund B et al. Auranofin improves outcome in early
2         rheumatoid arthritis. Results from a 2-year, double blind placebo
3         controlled study. Journal of Rheumatology. 1988; 15(12):1747-1754.

4     276 Egsmose C, Lund B, Borg G et al. Patients with rheumatoid arthritis
5         benefit from early 2nd line therapy: 5 year followup of a prospective
6         double blind placebo controlled study. Journal of Rheumatology. 1995;
7         22(12):2208-2213.

 8    277 Mottonen T, Hannonen P, Korpela M et al. Delay to institution of
 9        therapy and induction of remission using single-drug or combination-
10        disease-modifying antirheumatic drug therapy in early rheumatoid
11        arthritis. Arthritis & Rheumatism. 2002; 46(4):894-898.

12    278 Van der Heide A, Jacobs JW, Bijlsma JW et al. The effectiveness of
13        early treatment with "second-line" antirheumatic drugs. A randomized,
14        controlled trial. ANN INTERN MED. 1996; 124(8):699-707.

15    279 Verstappen SM, Jacobs JW, Bijlsma JW et al. Five-year followup of
16        rheumatoid arthritis patients after early treatment with disease-
17        modifying antirheumatic drugs versus treatment according to the
18        pyramid approach in the first year. Arthritis & Rheumatism. 2003;
19        48(7):1797-1807.

20    280 Buckland-Wright JC, Clarke GS, Chikanza IC et al. Quantitative
21        microfocal radiography detects changes in erosion area in patients with
22        early rheumatoid arthritis treated with myocrisine. Journal of
23        Rheumatology. 1993; 20(2):243-247.

24    281 Choy EH, Scott DL, Kingsley GH et al. Treating rheumatoid arthritis
25        early with disease modifying drugs reduces joint damage: a
26        randomised double blind trial of sulphasalazine vs diclofenac sodium.
27        Clinical & Experimental Rheumatology. 2002; 20(3):351-358.

28    282 Peltomaa R, Paimela L, Helve T et al. Effect of treatment on the
29        outcome of very early rheumatoid arthritis. Scandinavian Journal of
30        Rheumatology. 2001; 30(3):143-148.

31    283 Nell VP, Machold KP, Eberl G et al. Benefit of very early referral and
32        very early therapy with disease-modifying anti-rheumatic drugs in
33        patients with early rheumatoid arthritis. Rheumatology. 2004;
34        43(7):906-914.

35    284 Chen YF, Jobanputra P, Barton P et al. A systematic review of the
36        effectiveness of adalimumab, etanercept and infliximab for the
37        treatment of rheumatoid arthritis in adults and an economic evaluation
38        of their cost-effectiveness. Health Technology Assessment. 2006;
39        10(42):iii-iiv.

40    285 Ryan L, Brooks P. Disease-modifying antirheumatic drugs. Current
41        Opinion in Rheumatology. 1999; 11(3):161-166.


     Rheumatoid arthritis: full guideline DRAFT (August 2008)           294 of 307
     DRAFT FOR CONSULTATION


1     286 Spalding JR, Hay J. Cost effectiveness of tumour necrosis factor-alpha
2         inhibitors as first-line agents in rheumatoid arthritis.
3         Pharmacoeconomics. 2006; 24(12):1221-1232.

4     287 Yen JH. Treatment of early rheumatoid arthritis in developing countries.
5         Biologics or disease-modifying anti-rheumatic drugs? Biomedicine &
6         Pharmacotherapy. 2006; 60(10):688-692.

7     288 Grigor C, Capell H, Stirling A et al. Effect of a treatment strategy of tight
8         control for rheumatoid arthritis (the TICORA study): A single-blind
9         randomised controlled trial. Lancet. 2004; 364(9430):263-269.

10    289 Korthals-de Bos I, Van Tulder M, Boers M et al. Indirect and total costs
11        of early rheumatoid arthritis: a randomized comparison of combined
12        step-down prednisolone, methotrexate, and sulfasalazine with
13        sulfasalazine alone. Journal of Rheumatology. 2004; 31(9):1709-1716.

14    290 Choy EH, Smith C, Dore CJ et al. A meta-analysis of the efficacy and
15        toxicity of combining disease-modifying anti-rheumatic drugs in
16        rheumatoid arthritis based on patient withdrawal. Rheumatology. 2005;
17        44(11):1414-1421.

18    291 Felson DT, Anderson JJ, Meenan RF. The efficacy and toxicity of
19        combination therapy in rheumatoid arthritis. A meta-analysis. Arthritis &
20        Rheumatism. 1994; 37(10):1487-1491.

21    292 Hetland ML, Stengaard-Pedersen K. Combination treatment with
22        methotrexate, cyclosporine, and intraarticular betamethasone
23        compared with methotrexate and intraarticular betamethasone in early
24        active rheumatoid arthritis: an investigator-initiated, multicenter,
25        randomized, double-blind, parallel-group, placebo-controlled study.
26        Arthritis & Rheumatism. 2006; 54(5):1401-1409.

27    293 Hetland ML, Stengaard-Pedersen K, Junker P et al. Aggressive
28        combination therapy with intraarticular glucocorticoid injections and
29        conventional DMARDs in early rheumatoid arthritis Two Year Clinical
30        and Radiographic Results From The CIMESTRA Study. Annals of the
31        Rheumatic Diseases. 2007; 66

32    294 Korpela M, Laasonen L, Hannonen P et al. Retardation of joint damage
33        in patients with early rheumatoid arthritis by initial aggressive treatment
34        with disease-modifying antirheumatic drugs: five-year experience from
35        the FIN-RACo study. Arthritis & Rheumatism. 2004; 50(7):2072-2081.

36    295 Makinen H, Kautiainen H, Hannonen P et al. Sustained remission and
37        reduced radiographic progression with combination disease modifying
38        antirheumatic drugs in early rheumatoid arthritis. Journal of
39        Rheumatology. 2007; 34(2):316-321.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)              295 of 307
     DRAFT FOR CONSULTATION


1     296 Mottonen T, Hannonen P, LeirisaloRepo M et al. Comparison of
2         combination therapy with single-drug therapy in early rheumatoid
3         arthritis: A randomised trial. Lancet. 1999; 353(9164):1568-1573.

4     297 Puolakka K. Impact of initial aggressive drug treatment with a
5         combination of disease-modifying antirheumatic drugs on the
6         development of work disability in early rheumatoid arthritis: a five-year
7         randomized followup trial. Arthritis & Rheumatism. 2004; 50(1):55-62.

 8    298 Van Jaarsveld CH, Jacobs JW, Van der Veen MJ et al. Aggressive
 9        treatment in early rheumatoid arthritis: a randomised controlled trial. On
10        behalf of the Rheumatic Research Foundation Utrecht, The
11        Netherlands. Annals of the Rheumatic Diseases. 2000; 59(6):468-477.

12    299 Boers M. Randomised comparison of combined step-down
13        prednisolone, methotrexate and sulphasalazine with sulphasalazine
14        alone in early rheumatoid arthritis. Lancet. 1997; 350(9074):309-318.

15    300 Landewe RB, Boers M, Verhoeven AC et al. COBRA combination
16        therapy in patients with early rheumatoid arthritis: long-term structural
17        benefits of a brief intervention. Arthritis & Rheumatism. 2002;
18        46(2):347-356.

19    301 Capell HA, Madhok R, Porter DR et al. Combination therapy with
20        sulfasalazine and methotrexate is more effective than either drug alone
21        in patients with rheumatoid arthritis with a suboptimal response to
22        sulfasalazine: results from the double-blind placebo-controlled
23        MASCOT study. Annals of the Rheumatic Diseases. 2007; 66(2):235-
24        241.

25    302 Allaart CF, Goekoop-Ruiterman YP, De Vries-Bouwstra JK et al.
26        Aiming at low disease activity in rheumatoid arthritis with initial
27        combination therapy or initial monotherapy strategies: the BeSt study.
28        Clinical & Experimental Rheumatology. 2006; 24(6 suppl 43):1-77.

29    303 Goekoop-Ruiterman YP, de Vries BJK, Allaart CF et al. Comparison of
30        treatment strategies in early rheumatoid arthritis: a randomized trial.
31        ANN INTERN MED. 2007; 146(6):406-415.

32    304 Goekoop-Ruiterman YP, De Vries-Bouwstra JK, Allaart CF et al.
33        Clinical and radiographic outcomes of four different treatment
34        strategies in patients with early rheumatoid arthritis (the best study): A
35        randomized, controlled trial. Arthritis & Rheumatism. 2005;
36        52(11):3381-3390.

37    305 Van der Kooji SM, De Vries-Bouwstra JK, Goekoop-Ruiterman YP et
38        al. Limited efficacy of conventional DMARDs after initial methotrexate
39        failure in patients with recent onset rheumatoid arthritis treated
40        according to the disease activity score. Annals of the Rheumatic
41        Diseases. 2007; 66(10):1356-1362.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)             296 of 307
     DRAFT FOR CONSULTATION


1     306 Ferraccioli GF, Gremese E, Tomietto P et al. Analysis of
2         improvements, full responses, remission and toxicity in rheumatoid
3         patients treated with step-up combination therapy (methotrexate,
4         cyclosporin A, sulphasalazine) or monotherapy for three years.
5         Rheumatology. 2002; 41(8):892-898.

6     307 Choy EH, Smith CM, Farewell V et al. Factorial randomised controlled
7         trial of glucocorticoids and combination disease modifying drugs in
8         early rheumatoid arthritis. Annals of the Rheumatic Diseases. 2008;
9         67(5):656-663.

10    308 Braun J, Kastner P, Flaxenberg P et al. Comparison of the clinical
11        efficacy and safety of subcutaneous versus oral administration of
12        methotrexate in patients with active rheumatoid arthritis: results of a
13        six-month, multicenter, randomized, double-blind, controlled, phase IV
14        trial. Arthritis & Rheumatism. 2008; 58(1):73-81.

15    309 Hider SL, Silman A, Bunn D et al. Comparing the long-term clinical
16        outcome of treatment with methotrexate or sulfasalazine prescribed as
17        the first disease-modifying antirheumatic drug in patients with
18        inflammatory polyarthritis. Annals of the Rheumatic Diseases. 2006;
19        65(11):1449-1455.

20    310 Stenger AA, Van Leeuwen MA, Houtman PM et al. Early effective
21        suppression of inflammation in rheumatoid arthritis reduces
22        radiographic progression. British Journal of Rheumatology. 1998;
23        37(11):1157-1163.

24    311 Clark W, Jobanputra P, Barton P et al. The clinical and cost-
25        effectiveness of anakinra for the treatment of rheumatoid arthritis in
26        adults: a systematic review and economic analysis. Health Technology
27        Assessment. 2004; 8(18):iii-iiv.

28    312 Cohen SB, Woolley JM, Chan W et al. Interleukin 1 receptor antagonist
29        anakinra improves functional status in patients with rheumatoid
30        arthritis. Journal of Rheumatology. 2003; 30(2):225-231.

31    313 Genovese MC, Cohen S, Moreland L et al. Combination therapy with
32        etanercept and anakinra in the treatment of patients with rheumatoid
33        arthritis who have been treated unsuccessfully with methotrexate.
34        Arthritis & Rheumatism. 2004; 50(5):1412-1419.

35    314 Fleischmann RM, Tesser J, Schiff MH et al. Safety of extended
36        treatment with anakinra in patients with rheumatoid arthritis. Annals of
37        the Rheumatic Diseases. 2006; 65(8):1006-1012.

38    315 Bresnihan B, Newmark R, Robbins S et al. Effects of anakinra
39        monotherapy on joint damage in patients with rheumatoid arthritis.
40        Extension of a 24-week randomized, placebo-controlled trial. Journal of
41        Rheumatology. 2004; 31(6):1103-1111.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)           297 of 307
     DRAFT FOR CONSULTATION


1     316 Nuki G, Bresnihan B, Bear MB et al. Long-term safety and
2         maintenance of clinical improvement following treatment with anakinra
3         (recombinant human interleukin-1 receptor antagonist) in patients with
4         rheumatoid arthritis: extension phase of a randomized, double-blind,
5         placebo-controlled trial. Arthritis & Rheumatism. 2002; 46(11):2838-
6         2846.

7     317 Navarro-Sarabia F, Ariza AR, Hernandez CB et al. Adalimumab for
8         treating rheumatoid arthritis. COCHRANE DATABASE SYST REV.
9         2005;(3):CD005113.

10    318 Lee YH, Woo JH, Rho YH et al. Meta-analysis of the combination of
11        TNF inhibitors plus MTX compared to MTX monotherapy, and the
12        adjusted indirect comparison of TNF inhibitors in patients suffering from
13        active rheumatoid arthritis. Rheumatology International. 2008;
14        28(6):553-559.

15    319 Weinblatt ME, Keystone EC, Furst DE et al. Adalimumab, a fully
16        human anti-tumor necrosis factor alpha monoclonal antibody, for the
17        treatment of rheumatoid arthritis in patients taking concomitant
18        methotrexate: the ARMADA trial. Arthritis & Rheumatism. 2003;
19        48(1):35-45.

20    320 Klareskog L, Van Der HD, de Jager JP et al. Therapeutic effect of the
21        combination of etanercept and methotrexate compared with each
22        treatment alone in patients with rheumatoid arthritis: double-blind
23        randomised controlled trial. Lancet. 2004; 363(9410):675-681.

24    321 Van der Heijde D, Klareskog L, Singh A et al. Patient reported
25        outcomes in a trial of combination therapy with etanercept and
26        methotrexate for rheumatoid arthritis: the TEMPO trial. Annals of the
27        Rheumatic Diseases. 2006; 65(3):328-334.

28    322 Keystone EC, Kavanaugh AF, Sharp JT et al. Radiographic, clinical,
29        and functional outcomes of treatment with adalimumab (a human anti-
30        tumor necrosis factor monoclonal antibody) in patients with active
31        rheumatoid arthritis receiving concomitant methotrexate therapy: a
32        randomized, placebo-controlled, 52-week trial. Arthritis & Rheumatism.
33        2004; 50(5):1400-1411.

34    323 Furst DE, Schiff MH, Fleischmann RM et al. Adalimumab, a fully
35        human anti tumor necrosis factor-alpha monoclonal antibody, and
36        concomitant standard antirheumatic therapy for the treatment of
37        rheumatoid arthritis: results of STAR (Safety Trial of Adalimumab in
38        Rheumatoid Arthritis). Journal of Rheumatology. 2003; 30(12):2563-
39        2571.

40    324 Combe B, Codreanu C, Fiocco U et al. Etanercept and sulfasalazine,
41        alone and combined, in patients with active rheumatoid arthritis despite
42        receiving sulfasalazine: a double-blind comparison. Annals of the
43        Rheumatic Diseases. 2006; 65(10):1357-1362.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)           298 of 307
     DRAFT FOR CONSULTATION


1     325 Westhovens R, Yocum D, Han J et al. The safety of infliximab,
2         combined with background treatments, among patients with
3         rheumatoid arthritis and various comorbidities: a large, randomized,
4         placebo-controlled trial. Arthritis & Rheumatism. 2006; 54(4):1075-
5         1086.

 6    326 Westhovens R, Cole JC, Li T et al. Improved health-related quality of
 7        life for rheumatoid arthritis patients treated with abatacept who have
 8        inadequate response to anti-TNF therapy in a double-blind, placebo-
 9        controlled, multicentre randomized clinical trial. Rheumatology. 2006;
10        45(10):1238-1246.

11    327 Strand V, Balbir GA, Pavelka K et al. Sustained benefit in rheumatoid
12        arthritis following one course of rituximab: improvements in physical
13        function over 2 years. Rheumatology. 2006; 45(12):1505-1513.

14    328 Van der Heijde D, Klareskog L, Rodriguez VV et al. Comparison of
15        etanercept and methotrexate, alone and combined, in the treatment of
16        rheumatoid arthritis: two-year clinical and radiographic results from the
17        TEMPO study, a double-blind, randomized trial. Arthritis &
18        Rheumatism. 2006; 54(4):1063-1074.

19    329 Van Riel P, Taggart AJ, Sany J et al. Efficacy and safety of
20        combination etanercept and methotrexate versus etanercept alone in
21        patients with rheumatoid arthritis with an inadequate response to
22        methotrexate: the ADORE study. Annals of the Rheumatic Diseases.
23        2006; 65(11):1478-1483.

24    330 Emery P, Kosinski M, Li T et al. Treatment of rheumatoid arthritis
25        patients with abatacept and methotrexate significantly improved health-
26        related quality of life. Journal of Rheumatology. 2006; 33(4):681-689.

27    331 Van der Heijde D, Burmester G, Melo GJ et al. The safety and efficacy
28        of adding etanercept to methotrexate or methotrexate to etanercept in
29        moderately active rheumatoid arthritis patients previously treated with
30        monotherapy. Annals of the Rheumatic Diseases. 2008; 67(2):182-188.

31    332 Van der Heijde D, Klareskog L, Landewe R et al. Disease remission
32        and sustained halting of radiographic progression with combination
33        etanercept and methotrexate in patients with rheumatoid arthritis.
34        Arthritis & Rheumatism. 2007; 56(12):3928-3939.

35    333 Klareskog L, Gaubitz M, Rodriguez VV et al. A long-term, open-label
36        trial of the safety and efficacy of etanercept (Enbrel) in patients with
37        rheumatoid arthritis not treated with other disease-modifying
38        antirheumatic drugs. Annals of the Rheumatic Diseases. 2006;
39        65(12):1578-1584.

40    334 Bansback NJ, Brennan A, Ghatnekar O. Cost effectiveness of
41        adalimumab in the treatment of patients with moderate to severe



     Rheumatoid arthritis: full guideline DRAFT (August 2008)             299 of 307
     DRAFT FOR CONSULTATION


1          rheumatoid arthritis in Sweden. Annals of the Rheumatic Diseases.
2          2005; 64(7):995-1002.

3     335 Welsing PM, Severens JL, Hartman M et al. Modeling the 5-year cost
4         effectiveness of treatment strategies including tumor necrosis factor-
5         blocking agents and leflunomide for treating rheumatoid arthritis in the
6         Netherlands. Arthritis & Rheumatism. 2004; 51(6):964-973.

7     336 Wong JB, Singh G, Kavanaugh A. Estimating the cost-effectiveness of
8         54 weeks of infliximab for rheumatoid arthritis. American Journal of
9         Medicine. 2002; 113(5):400-408.

10    337 Choi HK, Seeger JD, Kuntz KM. A cost effectiveness analysis of
11        treatment options for methotrexate-naive rheumatoid arthritis. Journal
12        of Rheumatology. 2002; 29(6):1156-1165.

13    338 Brennan A, Bansback N, Reynolds A et al. Modelling the cost-
14        effectiveness of etanercept in adults with rheumatoid arthritis in the UK.
15        Rheumatology. 2004; 43(1):62-72.

16    339 Kobelt G, Jonsson L, Young A et al. The cost-effectiveness of
17        infliximab (Remicade) in the treatment of rheumatoid arthritis in
18        Sweden and the United Kingdom based on the ATTRACT study.
19        Rheumatology. 2003; 42(2):326-335.

20    340 Kobelt G, Eberhardt K, Geborek P. TNF inhibitors in the treatment of
21        rheumatoid arthritis in clinical practice: costs and outcomes in a follow
22        up study of patients with RA treated with etanercept or infliximab in
23        southern Sweden. Annals of the Rheumatic Diseases. 2004; 63(1):4-
24        10.

25    341 Marra CA, Marion SA, Guh DP et al. Not all "quality-adjusted life years"
26        are equal. Journal of Clinical Epidemiology. 2007; 60(6):616-624.

27    342 Tanno M, Nakamura I, Ito K et al. Modeling and cost-effectiveness
28        analysis of etanercept in adults with rheumatoid arthritis in Japan: a
29        preliminary analysis. Modern Rheumatology. 2006; 16(2):77-84.

30    343 Chiou CF, Choi J, Reyes CM. Cost-effectiveness analysis of biological
31        treatments for rheumatoid arthritis. Expert Review of
32        Pharmacoeconomics and Outcomes Research 4(3)()(pp 307-315),
33        2004 Date of Publication: Jun 2004. 2004;(3):307-315.

34    344 Kobelt G, Lindgren P, Singh A et al. Cost effectiveness of etanercept
35        (Enbrel) in combination with methotrexate in the treatment of active
36        rheumatoid arthritis based on the TEMPO trial. Annals of the
37        Rheumatic Diseases. 2005; 64(8):1174-1179.

38    345 Brennan A, Bansback N, Nixon R et al. Modelling the cost
39        effectiveness of TNF-alpha antagonists in the management of
40        rheumatoid arthritis: results from the British Society for Rheumatology
41        Biologics Registry. Rheumatology. 2007; 46(8):1345-1354.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)            300 of 307
     DRAFT FOR CONSULTATION


1     346 Ahern MJ, Hall ND, Case K et al. D-penicillamine withdrawal in
2         rheumatoid arthritis. Annals of the Rheumatic Diseases. 1984;
3         43(2):213-217.

4     347 De Silva M, Hazleman BL. Long-term azathioprine in rheumatoid
5         arthritis: A double-blind study. Annals of the Rheumatic Diseases.
6         1981; 40(6):560-563.

 7    348 Gotzsche PC, Hansen M, Stoltenberg M et al. Randomized, placebo
 8        controlled trial of withdrawal of slow-acting antirheumatic drugs and of
 9        observer bias in rheumatoid arthritis. Scandinavian Journal of
10        Rheumatology. 1996; 25(4):194-199.

11    349 Kremer JM, Rynes RI, Bartholomew LE. Severe flare of rheumatoid
12        arthritis after discontinuation of long-term methotrexate therapy.
13        Double-blind study. American Journal of Medicine. 1987; 82(4):781-
14        786.

15    350 Ten Wolde S, Breedveld FC, Hermans J et al. Randomised placebo-
16        controlled study of stopping second-line drugs in rheumatoid arthritis.
17        Lancet. 1996; 347(8998):347-352.

18    351 Van der Leeden H, Dijkmans BA, Hermans J et al. A double-blind study
19        on the effect of discontinuation of gold therapy in patients with
20        rheumatoid arthritis. Clinical Rheumatology. 1986; 5(1):56-61.

21    352 Bacon PA, Myles AB, Beardwell CG et al. Corticosteroid withdrawal in
22        rheumatoid arthritis. Lancet. 1966; 2(7470):935-937.

23    353 Tishler M, Caspi D, Yaron M. Methotrexate treatment of rheumatoid
24        arthritis: is a fortnightly maintenance schedule enough? Annals of the
25        Rheumatic Diseases. 1992; 51(12):1330-1331.

26    354 Fleischmann RM, Cohen SB, Moreland LW et al. Methotrexate dosage
27        reduction in patients with rheumatoid arthritis beginning therapy with
28        infliximab: the Infliximab Rheumatoid Arthritis Methotrexate Tapering
29        (iRAMT) trial. Current Medical Research & Opinion. 2005; 21(8):1181-
30        1190.

31    355 Tengstrand B, Larsson E, Klareskog L et al. Randomized withdrawal of
32        long-term prednisolone treatment in rheumatoid arthritis: effects on
33        inflammation and bone mineral density. Scandinavian Journal of
34        Rheumatology. 2007; 36(5):351-358.

35    356 Fransen J, Moens HB, Speyer I et al. Effectiveness of systematic
36        monitoring of rheumatoid arthritis disease activity in daily practice: a
37        multicentre, cluster randomised controlled trial. Annals of the
38        Rheumatic Diseases. 2005; 64(9):1294-1298.

39    357 Verstappen SM, Jacobs JW, Van der Veen MJ et al. Intensive
40        treatment with methotrexate in early rheumatoid arthritis: aiming for
41        remission. Computer Assisted Management in Early Rheumatoid

     Rheumatoid arthritis: full guideline DRAFT (August 2008)             301 of 307
     DRAFT FOR CONSULTATION


1          Arthritis (CAMERA, an open-label strategy trial). Annals of the
2          Rheumatic Diseases. 2007; 66(11):1443-1449.

3     358 Smolen JS, Breedveld FC, Schiff MH et al. A simplified disease activity
4         index for rheumatoid arthritis for use in clinical practice. Rheumatology.
5         2003; 42(2):244-257.

6     359 Dixon JS, Hayes S, Constable PDL et al. What are the 'best'
7         measurements for monitoring patients during short-term second-line
8         therapy? British Journal of Rheumatology. 1988; 27(1):37-43.

 9    360 Dixon JS, Bird HA, Sitton NG et al. C-reactive protein in the serial
10        assessment of disease activity in rheumatoid arthritis. Scandinavian
11        Journal of Rheumatology. 1984; 13(1):39-44.

12    361 Kalla AA, Smith PR, Brown GM et al. Responsiveness of Keitel
13        functional index compared with laboratory measures of disease activity
14        in rheumatoid arthritis. British Journal of Rheumatology. 1995;
15        34(2):141-149.

16    362 Van der Heijde D, van't Hof MA, van Riel PL et al. Validity of single
17        variables and composite indices for measuring disease activity in
18        rheumatoid arthritis. Annals of the Rheumatic Diseases. 1992;
19        51(2):177-181.

20    363 Van Leeuwen MA, Van Rijswijk M, Van der Heijde D et al. The acute-
21        phase response in relation to radiographic progression in early
22        rheumatoid arthritis: A prospective study during the first three years of
23        the disease. British Journal of Rheumatology. 1993; 32(6):9-13.

24    364 Arthritis and Musculoskeletal Alliance. Standards of care for people
25        with inflammatory arthritis. London: ARMA, 2004.

26    365 Symmons D, Tricker K, Harrison M et al. Patients with stable long-
27        standing rheumatoid arthritis continue to deteriorate despite intensified
28        treatment with traditional disease modifying anti-rheumatic drugs--
29        results of the British Rheumatoid Outcome Study Group randomized
30        controlled clinical trial. Rheumatology. 2006; 45(5):558-565.

31    366 Hewlett S, Mitchell K, Haynes J et al. Patient-initiated hospital follow-up
32        for rheumatoid arthritis. Rheumatology. 2000; 39(9):990-997.

33    367 Mowat AG, Nichols PJ, Hollings EM et al. A comparison of follow-up
34        regimes in rheumatoid arthritis. Annals of the Rheumatic Diseases.
35        1980; 39(1):12-17.

36    368 Kirwan JR, Mitchell K, Hewlett S et al. Clinical and psychological
37        outcome from a randomized controlled trial of patient-initiated direct-
38        access hospital follow-up for rheumatoid arthritis extended to 4 years.
39        Rheumatology. 2003; 42(3):422-426.



     Rheumatoid arthritis: full guideline DRAFT (August 2008)             302 of 307
     DRAFT FOR CONSULTATION


1     369 Alderman AK, Ubel PA, Kim HM et al. Surgical management of the
2         rheumatoid hand: consensus and controversy among rheumatologists
3         and hand surgeons. Journal of Rheumatology. 2003; 30(7):1464-1472.

4     370 Alderman AK, Arora AS, Kuhn L et al. An analysis of women's and
5         men's surgical priorities and willingness to have rheumatoid hand
6         surgery. Journal of Hand Surgery - American Volume. 2006;
7         31(9):1447-1453.

 8    371 Chung KC, Kotsis SV, Kim HM et al. Reasons why rheumatoid arthritis
 9        patients seek surgical treatment for hand deformities. Journal of Hand
10        Surgery - American Volume. 2006; 31(2):289-294.

11    372 Loza E, Abasolo L, Clemente D et al. Variability in the use of
12        orthopedic surgery in patients with rheumatoid arthritis in Spain.
13        Journal of Rheumatology. 2007; 34(7):1485-1490.

14    373 Mandl LA, Burke FD, Shaw Wilgis EF et al. Could preoperative
15        preferences and expectations influence surgical decision making?
16        Rheumatoid arthritis patients contemplating metacarpophalangeal joint
17        arthroplasty. Plastic & Reconstructive Surgery. 2008; 121(1):175-180.

18    374 Hamilton JD, Gordon MM, McInnes IB et al. Improved medical and
19        surgical management of cervical spine disease in patients with
20        rheumatoid arthritis over 10 years. Annals of the Rheumatic Diseases.
21        2000; 59(6):434-438.

22    375 Little C, Parsons T. Herbal therapy for treating rheumatoid arthritis.
23        COCHRANE DATABASE SYST REV. 2000;(4):CD002948.

24    376 Goldberg RJ, Katz J. A meta-analysis of the analgesic effects of
25        omega-3 polyunsaturated fatty acid supplementation for inflammatory
26        joint pain. Pain. 2007; 129(1-2):210-223.

27    377 Skoldstam L, Brudin L, Hagfors L et al. Weight reduction is not a major
28        reason for improvement in rheumatoid arthritis from lacto-vegetarian,
29        vegan or Mediterranean diets. Nutrition Journal. 2005; 4(15)

30    378 Skoldstam L, Hagfors L, Johansson G. An experimental study of a
31        Mediterranean diet intervention for patients with rheumatoid arthritis.
32        Annals of the Rheumatic Diseases. 2003; 62(3):208-214.

33    379 Panush RS, Carter RL, Katz P et al. Diet therapy for rheumatoid
34        arthritis. Arthritis & Rheumatism. 1983; 26(4):462-471.

35    380 Hansen GVO, Nielsen L, Kluger E et al. Nutritional status of Danish
36        rheumatoid arthritis patients and effects of a diet adjusted in energy
37        intake, fish-meal, and antioxidants. Scandinavian Journal of
38        Rheumatology. 1996; 25(5):325-330.




     Rheumatoid arthritis: full guideline DRAFT (August 2008)            303 of 307
     DRAFT FOR CONSULTATION


1     381 Holst-Jensen SE, Pfeiffer-Jensen M, Monsrud M et al. Treatment of
2         rheumatoid arthritis with a peptide diet: a randomized, controlled trial.
3         Scandinavian Journal of Rheumatology. 1998; 27(5):329-336.

4     382 Kavanagh R, Workman E, Nash P et al. The effects of elemental diet
5         and subsequent food reintroduction on rheumatoid arthritis. British
6         Journal of Rheumatology. 1995; 34(3):270-273.

 7    383 Sarzi-Puttini P, Comi D, Boccassini L et al. Diet therapy for rheumatoid
 8        arthritis: A controlled double-blind study of two different dietary
 9        regimens. Scandinavian Journal of Rheumatology. 2000; 29(5):302-
10        307.

11    384 Van de Laar M, van der Korst JK. Food intolerance in rheumatoid
12        arthritis. I. A double blind, controlled trial of the clinical effects of
13        elimination of milk allergens and azo dyes. Annals of the Rheumatic
14        Diseases. 1992; 51(3):298-302.

15    385 Edmonds SE, Winyard PG. Putative analgesic activity of repeated oral
16        doses of vitamin E in the treatment of rheumatoid arthritis. Results of a
17        prospective placebo controlled double blind trial. Annals of the
18        Rheumatic Diseases. 1997; 56(11):649-655.

19    386 Peretz A, Siderova V, Neve J. Selenium supplementation in
20        rheumatoid arthritis investigated in a double blind, placebo-controlled
21        trial. Scandinavian Journal of Rheumatology. 2001; 30(4):-212.

22    387 Hafstrom I, Ringertz B, Spangberg A et al. A vegan diet free of gluten
23        improves the signs and symptoms of rheumatoid arthritis: the effects
24        on arthritis correlate with a reduction in antibodies to food antigens.
25        Rheumatology. 2001; 40(10):1175-1179.

26    388 Kjeldsen-Kragh J., Haugen M, Borchgrevink CF et al. Vegetarian diet
27        for patients with rheumatoid arthritis--status: two years after
28        introduction of the diet. Clinical Rheumatology. 1994; 13(3):475-482.

29    389 Kjeldsen-Kragh J, Mellbye OJ, Haugen M et al. Changes in laboratory
30        variables in rheumatoid arthritis patients during a trial of fasting and
31        one-year vegetarian diet. Scandinavian Journal of Rheumatology.
32        1995; 24(2):85-93.

33    390 Kjeldsen-Kragh J, Haugen M, Borchgrevink CF et al. Controlled trial of
34        fasting and one-year vegetarian diet in rheumatoid arthritis. Lancet.
35        1991; 338(8772):899-902.

36    391 Nenonen MT, Helve TA, Rauma AL et al. Uncooked, lactobacilli-rich,
37        vegan food and rheumatoid arthritis. British Journal of Rheumatology.
38        1998; 37(3):274-281.

39    392 Galarraga B, Ho M, Youssef HM et al. Cod liver oil (n-3 fatty acids) as
40        an non-steroidal anti-inflammatory drug sparing agent in rheumatoid
41        arthritis. Rheumatology. 2008; 47(5):665-669.

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              304 of 307
     DRAFT FOR CONSULTATION


1     393 Elkan AC, Sjoberg B, Kolsrud B et al. Gluten-free vegan diet induces
2         decreased LDL and oxidized LDL levels and raised atheroprotective
3         natural antibodies against phosphorylcholine in patients with
4         rheumatoid arthritis: A randomized study. Arthritis Research &
5         Therapy. 2008; 10(2)

6     394 Kromhout D, Keys A, Aravanis C et al. Food consumption patterns in
7         the 1960s in seven countries. American Journal of Clinical Nutrition.
8         1989; 49(5):889-894.

 9    395 Ernst E. Musculoskeletal conditions and complementary/alternative
10        medicine. Best Practice & Research in Clinical Rheumatology. 2004;
11        18(4):539-556.

12    396 Casimiro L, Brosseau L, Milne S et al. Acupuncture and
13        electroacupuncture for the treatment of Rheumatoid Arthritis.
14        COCHRANE DATABASE SYST REV. 2005;(4):CD003788.

15    397 Field T, Diego M, Hernandez RM et al. Hand arthritis pain is reduced
16        by massage therapy. Journal of Bodywork & Movement Therapies.
17        2007; 11(1):21-24.

18    398 Gibson RG, Gibson SL, MacNeill AD et al. Homoeopathic therapy in
19        rheumatoid arthritis: evaluation by double-blind clinical therapeutic trial.
20        British Journal of Clinical Pharmacology. 1980; 9(5):453-459.

21    399 Gibson RG, Gibson SLM, MacNeill AD et al. The place for non-
22        pharmaceutical therapy in chronic rheumatoid arthritis: A critical study
23        of homoeopathy. British Homoeopathic Journal. 1980; 69(3):121-133.

24    400 Freye E, Latasch L. Analgesic therapy of rheumatoid arthritis - Part II:
25        A study of combined allopathic and homeopathic therapy. Biomedical
26        Therapy. 2000; 18(2):193-196.

27    401 Kobelt G, Jonsson L, Lindgren P et al. Modeling the progression of
28        rheumatoid arthritis: a two-country model to estimate costs and
29        consequences of rheumatoid arthritis. Arthritis & Rheumatism. 2002;
30        46(9):2310-2319.

31    402 National Institute for Health and Clinical Effectiveness.
32        Adalimumab,etanercept and infliximab for the treatment of rheumatoid
33        arthritis. (TA130). London: NICE, 2007.

34    403 Saunders SA, Capell HA, Stirling A et al. Triple therapy in early active
35        rheumatoid arthritis: a randomized, single-blind, controlled trial
36        comparing step-up and parallel treatment strategies. Arthritis &
37        Rheumatism. 2008; 58(5):1310-1317.

38    404 Dougados M, Combe B, Cantagrel A et al. Combination therapy in
39        early rheumatoid arthritis: a randomised, controlled, double blind 52
40        week clinical trial of sulphasalazine and methotrexate compared with


     Rheumatoid arthritis: full guideline DRAFT (August 2008)              305 of 307
     DRAFT FOR CONSULTATION


1          the single components. Annals of the Rheumatic Diseases. 1999;
2          58(4):220-225.

3     405 Gerards AH, Landewe RB, Prins AP et al. Cyclosporin A monotherapy
4         versus cyclosporin A and methotrexate combination therapy in patients
5         with early rheumatoid arthritis: a double blind randomised placebo
6         controlled trial. Annals of the Rheumatic Diseases. 2003; 62(4):291-
7         296.

 8    406 Miranda JM, varez-Nemegyei J, Saavedra MA et al. A randomized,
 9        double-blind, multicenter, controlled clinical trial of cyclosporine plus
10        chloroquine vs. cyclosporine plus placebo in early-onset rheumatoid
11        arthritis. Archives of Medical Research. 2004; 35(1):36-42.

12    407 Sarzi-Puttini P, D'Ingianna E, Fumagalli M et al. An open, randomized
13        comparison study of cyclosporine A, cyclosporine A + methotrexate
14        and cyclosporine A + hydroxychloroquine in the treatment of early
15        severe rheumatoid arthritis. Rheumatology International. 2005;
16        25(1):15-22.

17    408 Marchesoni A, Battafarano N, Arreghini M et al. Radiographic
18        progression in early rheumatoid arthritis: a 12-month randomized
19        controlled study comparing the combination of cyclosporin and
20        methotrexate with methotrexate alone. Rheumatology. 2003;
21        42(12):1545-1549.

22    409 Van den Borne B E, Landewe RB, Goei THS et al. Combination
23        therapy in recent onset rheumatoid arthritis: a randomized double blind
24        trial of the addition of low dose cyclosporine to patients treated with low
25        dose chloroquine. Journal of Rheumatology. 1998; 25(8):1493-1498.

26    410 Lu G, Ades AE. Combination of direct and indirect evidence in mixed
27        treatment comparisons. Statistics in Medicine. 2004; 23(20):3105-
28        3124.

29    411 Caldwell DM, Ades AE, Higgins JP. Simultaneous comparison of
30        multiple treatments: combining direct and indirect evidence. British
31        Medical Journal. 2005; 331(7521):897-900.

32    412 Lumley T. Network meta-analysis for indirect treatment comparisons.
33        Statistics in Medicine. 2002; 21(16):2313-2324.

34    413 Barbieri M, Wong JB, Drummond M. The cost effectiveness of
35        infliximab for severe treatment-resistant rheumatoid arthritis in the UK.
36        Pharmacoeconomics. 2005; 23(6):607-618.

37    414 Wailoo AJ, Bansback N, Brennan A et al. Biologic drugs for rheumatoid
38        arthritis in the Medicare program: a cost-effectiveness analysis.
39        Arthritis & Rheumatism. 2008; 58(4):939-946.

40    415 Hurst NP, Kind P, Ruta D et al. Measuring health-related quality of life
41        in rheumatoid arthritis: validity, responsiveness and reliability of

     Rheumatoid arthritis: full guideline DRAFT (August 2008)              306 of 307
    DRAFT FOR CONSULTATION


1         EuroQol (EQ-5D). British Journal of Rheumatology. 1997; 36(5):551-
2         559.

3    416 Office of National Statistics. Standard Interim UK Life Tables. 2006.
4        London, ONS.
5        http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=14459

6
7




    Rheumatoid arthritis: full guideline DRAFT (August 2008)           307 of 307