Docstoc

and patient selection appropriateness of prostheses and its

Document Sample
and patient selection appropriateness of prostheses and its Powered By Docstoc
					                                 A review of
                   joint replacement surgery
                          and its outcomes:
              appropriateness of prostheses
                        and patient selection


                                   Prepared by

                          Professor Stephen Graves

                                         and

                                  Vanessa Wells

         for The Australian Centre For Health Research Ltd

                                   October 2006




                   ACHR
                      AUSTRALIAN CENTRE FOR HEALTH RESEARCH




     AUSTRALIAN CENTRE FOR HEALTH RESEARCH LTD ABN 87 116 781 965
114 Albert Road, South Melbourne Victoria 3205 Phone 03 9697 0283 Website www.achr.com.au
                                          Prepared by:

                                   PROFESSOR SE GRAVES

                                            DIRECTOR
                     AUSTRALIAN ORTHOPAEDIC ASSOCIATION
                     NATIONAL JOINT REPLACEMENT REGISTRY

                                                AND

                                       VANESSA WELLS

                         CLINICAL RESEARCH COORDINATOR
                         REPATRIATION GENERAL HOSPITAL
                           DAW PARK SOUTH AUSTRALIA



Special Note: The objective of the Australian Centre For Health Research Limited (ACHR) is to
undertake activities aimed at improving the provision of health and ageing services in Australia. To
this end, the ACHR commissions independent research on topics of importance in health and
ageing, in order to stimulate further exploration of areas of shared concern. The presentation of
factual matters, and the ideas and conclusions expressed in the independent research reflect the
views of the author of that research, and do not necessarily represent all of the views of ACHR or its
individual officers.

(c) 2006 Australian Centre For Health Research Limited
PREFACE

This discussion paper has been prepared at the request of the Australian Centre for Health Research
Ltd. Although the focus of this paper is on the outcomes of joint replacement surgery it has been
necessary to provide detail on changing incidence of joint replacement surgery, expenditure as well
as detailed information on the different categories of joint replacement and the methods of fixation
used. The reason for this is to define the practice of joint replacement surgery and provide
background information and understanding of the factors that directly effect outcomes.

It is also important to emphasise that the outcomes of joint replacement surgery in Australia are
generally good and comparable or better than what is being achieved in many other countries.
Outcomes, however, can always be improved and this document details how this can be achieved.




                                                 2
EXECUTIVE SUMMARY .............................................................................................................7
RECOMMENDATIONS ...............................................................................................................9
INTRODUCTION.......................................................................................................................10
AOA NJRR DATA .......................................................................................................................10
GOVERNMENT DATA ..............................................................................................................13
INTERNATIONAL DATA ..........................................................................................................13
   Swedish Hip Arthroplasty Registry.............................................................................................................13
   Swedish Knee arthroplasty Registry............................................................................................................13
   Norwegian Arthroplasty Registry................................................................................................................13
   Finnish Arthroplasty Register ....................................................................................................................14
   Canadian Joint Replacement Registry ........................................................................................................14
   England and Wales Registry ......................................................................................................................14
   New Zealand Joint Replacement Registry ...................................................................................................14
INCIDENCE OF JOINT REPLACEMENT .................................................................................16
FUTURE RATE OF JOINT REPLACEMENT.............................................................................18
PUBLIC VERSES PRIVATE JOINT REPLACEMENT................................................................19
EXPENDITURE ...........................................................................................................................20
DIFFERENT CATEGORIES OF JOINT REPLACEMENT .........................................................22
   Hip Replacement .......................................................................................................................................22
   Knee replacement.......................................................................................................................................23
INDICATIONS AND USE OF DIFFERENCE CATEGORIES OF JOINT REPLACEMENT ...23
   Hip replacement procedures ......................................................................................................................23
   Knee Replacement Procedures ...................................................................................................................25
CEMENTED AND CEMENTLESS FIXATION..........................................................................25
   Fixation of hip replacements ......................................................................................................................26
   Fixation of knee replacements....................................................................................................................27
REGIONAL DIFFERENCES AND TRENDS IN HIP REPLACEMENT.....................................28
   Partial hip replacement...............................................................................................................................29
   Total hip replacement ................................................................................................................................30
   Revision hip replacement...........................................................................................................................30
   Variation and trends in fixation of hip replacement...................................................................................31
REGIONAL DIFFERENCES AND TRENDS IN KNEE REPLACEMENT .................................34
MEASURES USED TO DETERMINE OUTCOME...................................................................36
OUTCOMES OF JOINT REPLACEMENT SURGERY IN AUSTRALIA ..................................38
   Global comparison ....................................................................................................................................38
   Hip replacement ........................................................................................................................................39
   Partial hip replacement...............................................................................................................................39
   Primary total hip replacement ....................................................................................................................40
   Knee replacement.......................................................................................................................................46
   Recently approved prostheses.....................................................................................................................50

                                                                       3
IMPLICATIONS OF IDENTIFIED VARIATION IN OUTCOMES..........................................51
   Comparison to patient and prostheses selection in Sweden.......................................................................51
   Why are there differences? .........................................................................................................................52
IMPACT OF ENHANCING OUTCOMES ................................................................................53
SUGGESTED APPROACHES TO ENHANCING OUTCOMES ..............................................53
   Need to maintain the Registry ....................................................................................................................53
   Profession developed guidelines for patient selection and prosthesis use ...................................................54
   Need to consider change to regulation of hip and knee prostheses ............................................................55
   Current regulation of medical devices in Australia .....................................................................................55
   Therapeutic Goods Administration ...........................................................................................................56
   Prostheses and Devices Committee............................................................................................................57
   Clinical Advisory Groups ..........................................................................................................................57
   Medical Services Advisory Committee (MSAC).........................................................................................57
   Recent international attempts to regulate and improve prostheses specific outcomes .................................58
   Suggested approach to regulatory change ...................................................................................................58
   Pre-market approval ...................................................................................................................................58
   Post-Market Surveillance ............................................................................................................................59
CONCLUSION ...........................................................................................................................60




                                                                       4
LIST OF TABLES

Table 1: Hip and knee procedures from 1994-1995 to 2004-2005. .................................................................. 16
Table 2: Incidence of hip and knee procedures per 100,000 population ........................................................... 17
Table 3: Public and private joint replacement ............................................................................................. 19
Table 4: Public and private hip replacement ............................................................................................... 20
Table 5: Public and private knee replacement ............................................................................................. 20
Table 6: Estimated total acute care expenditure 1999-2000 to 2002-2003 .......................................................... 21
Table 7: Estimated prostheses expenditure 1999-2000 to 2002-2003 ................................................................ 21
Table 8: Estimated prostheses expenditure (public and private) ....................................................................... 22
Table 9: Hip replacement and gender ....................................................................................................... 24
Table 10: Hip replacement and age .......................................................................................................... 24
Table 11: Knee replacement and gender .................................................................................................... 25
Table 12: Knee replacement and age......................................................................................................... 25
Table 13: Fixation — Primary partial hip replacement.................................................................................... 26
Table 14: Fixation — Primary total hip replacement ...................................................................................... 27
Table 15: Fixation — Primary unicompartment knee replacement .................................................................... 27
Table 16: Fixation — Primary total knee replacement .................................................................................... 28
Table 17: International Comparison of Proportion of Revision Procedures ....................................................... 38
Table 18: Revision rates of primary partial hip replacement ........................................................................... 39
Table 19: Revision rates by type of primary total hip replacement .................................................................... 40
Table 20: Revision rates of conventional total hip and resurfacing hip for osteoarthritis requiring revision excluding
     revisions for infection .................................................................................................................... 41
Table 21: Resurfacing hip systems requiring revision by age and sex ................................................................ 42
Table 22: Primary conventional total procedures for all diagnoses requiring revision by cement status ..................... 43
Table 23: Primary conventional total procedures for osteoarthritis requiring revision by cement status excluding infection
       ................................................................................................................................................. 43
Table 24: Resurfacing hip systems requiring revision.................................................................................... 44
Table 25: Variable outcome for the same femoral stem used with different acetabular components ......................... 45
Table 26: Unispacer procedures requiring revision ...................................................................................... 46
Table 27: Unicompartmental primary knee procedures requiring revision ........................................................ 47
Table 28: Comparison of unicompartmental primary knee procedures with three comparators M/G, Unix and Repicci
       ................................................................................................................................................. 49
Table 29: Comparison of new and established prostheses ............................................................................. 51




                                                                       5
LIST OF FIGURES

Figure 1: Schematic of the AOA NJRR...................................................................................................... 12
Figure 2: Hip and knee procedures from 1994-1995 to 2004-2005. ................................................................ 16
Figure 3: International comparison of number of procedures per 100,000 population ......................................... 18
Figure 4: Trends in hip replacement by state and territory ............................................................................. 28
Figure 5: Trends in primary partial hip replacement by state and territory ......................................................... 29
Figure 6: Trends in primary total hip replacement by state and territory ............................................................ 30
Figure 7: Trends in types of revision hip replacement by state and territory ....................................................... 31
Figure 8: Trends in unipolar monoblock fixation by state and territory ............................................................. 32
Figure 9: Trends in prosthesis fixation - femoral components used with unipolar modular prostheses by state and
     territory ...................................................................................................................................... 32
Figure 10: Trends in prosthesis fixation - femoral components used with bipolar prostheses by state and territory ...... 33
Figure 11: Trends in fixation of conventional hip replacement by state and territory ............................................ 33
Figure 12: Trends in knee replacement by state and territory .......................................................................... 34
Figure 13: Trends in fixation of unicompartment knee replacement by state and territory ..................................... 35
Figure 14: Trends in primary total knee fixation by state and territory .............................................................. 35
Figure 15: Trends in patella use for total knee replacement by state and territory ................................................ 36
Figure 16: Cumulative percentage of revision of Austin Moore and Thompson hip prostheses .............................. 40
Figure 17: Cumulative percentage of revision of conventional total hip and resurfacing hip for osteoarthritis excluding
     revisions for infection .................................................................................................................... 41
Figure 18: Cumulative percentage of revision of resurfacing hip by sex ............................................................. 42
Figure 19: Cumulative percentage of revision for total hip replacement by cement status excluding infection ............. 43
Figure 20: Cumulative percentage of revision for total hip replacement by cement status excluding infection for
     individuals 75 yrs and older ............................................................................................................ 44
Figure 21: Cumulative percentage of revision of cementless Margron hip prosthesis v other cementless femoral
     components ................................................................................................................................ 45
Figure 22: Comparative cumulative percentage of revision of unicompartmental procedures requiring revision by age
     (primary diagnosis OA excluding revisions for infection) ....................................................................... 46
Figure 23: Comparative cumulative percentage of revision of cemented, hybrid and cementless fixation of primary total
     knee replacement. ......................................................................................................................... 47
Figure 24: Cumulative percentage of revision of fixed and mobile ................................................................... 48
Figure 25: Revision rates for primary total knee replacements requiring revision by patella use............................... 48
Figure 26: Cumulative percentage of revision of Oxinium Cementless Genesis II total knee prosthesis.................... 50




                                                                      6
EXECUTIVE SUMMARY


The National Joint Replacement Registry monitors the outcome of all joint replacement procedures
undertaken in Australia. It is an initiative of the Australian Orthopaedic Association which was
established in 1999 and fully implemented nationally in 2002. It is a collaborative between
orthopaedics surgeons, government, all hospitals undertaking joint replacement surgery both public
and private and the orthopaedic industry. It is funded entirely by the Commonwealth. Its purpose
is to improve the results of joint replacement surgery. It does this by providing quality information
on the practice and outcomes, identifying reasons for revision and highlighting potential areas of
improvement. The Registry has already been responsible for a significant reduction in the
proportion of revision operations, equivalent to 1,200 fewer revisions a year. As well as the
considerable benefit to patients, this has reduced expenditure by $16-32 million per year. This
document highlights areas where further improvements can be made.


Approximately 65,000 joint replacements will be undertaken in 2006 at an estimated cost of well
over $1 billion. If the current rate of increase in joint replacement continues the number of
replacements is likely to double by 2016. The rate of hip replacement is less than some other
comparable countries but Australia has a high rate of knee replacement. Over 60% of joint
replacements are undertaken in the private sector and although the incidence of joint replacement is
increasing in both public and private the increase in the private sector is greater. The cost of
prostheses accounts for 35% of the total expenditure for each procedure. The non-prostheses
expenditure has declined but prostheses costs have increased particularly in the private sector. The
impact of the introduction of the Prostheses and Devices Committee has yet to be determined.
There is little regional variation in the age, gender and diagnosis of patients receiving joint
replacement. There is, however, significant state and territory variation in the choice of prosthesis
and surgical techniques.


The revision rate in Australia is comparable or better than most countries with the exception of
Sweden, which has half the revision rate. Sweden has had registries monitoring the outcome of hip
and knee replacements for over 30 years. The Australian Joint Registry has been modelled on these
registries. Analysis of Registry data and comparison to Swedish data has identified a number of
factors contributing to the current difference in the rate of revision. These include the use of large
numbers of different types of prostheses many of which have higher revision rates than the few well


                                                7
established prostheses used in Sweden. Australia also has a low use of cement fixation and this is
known to be associated with reduced revisions compared to cementless fixation. Some classes of
prostheses are more commonly used in Australia and the Registry has identified that these can have
higher revision rates, particularly in certain patient populations. There are also specific problems
related to the more frequent uptake of new prostheses when compared to Sweden. An analysis of
the outcome of all prostheses approved for use within Australia during the last four years indicates
that none of these prostheses has improved the outcome of joint replacement.


This discussion paper presents possible strategies to further enhance the outcomes of joint
replacement in this country. These include the maintenance of the Registry, development of
clinician guidelines and suggested alterations to the pre-market and post-market regulation. The aim
is to ensure that Australia has the best possible outcomes from this major surgery.




                                                8
RECOMMENDATIONS


  1. The Australian Orthopaedic Association National Joint Replacement Registry has clearly
     proven to be an effective method for determining the outcomes of joint replacement surgery
     and bringing about beneficial change in practice. Its effectiveness will be enhanced as time
     progresses. Its activities need to be ongoing and it should be provided with stable long term
     future funding to ensure its continued effectiveness.

  2. The pace of beneficial change in practice could be enhanced by the development of clinical
     guidelines by the profession for joint replacement surgery. The focus of these guidelines
     would be on appropriate patient and prostheses selection.

  3. There needs to be a re-evaluation of the regulatory activities governing the introduction of
     new hip and knee replacement prostheses into the Australian market. The aim of which
     would be to consider the introduction of the requirement of clinical trialling for new
     prostheses and that approval for use would be based on a proven clinical advantage over
     available prostheses.




                                             9
INTRODUCTION

Joint replacement surgery is a major health issue. Approximately 65,000 procedures will be
undertaken nationally in 2006. Most will be performed in the private sector. The estimated acute
care expenditure for this year will be well over $1 billion. During the last ten years the number of
hip and knee replacements undertaken each year has almost doubled. Available evidence suggests
that this rapid increase in the number of procedures undertaken each year will continue in the
foreseeable future.

Although hip and knee replacements are very successful operations, the Australian Orthopaedic
Association National Joint Replacement Registry (AOA NJRR) is now able to provide quality
evidence to demonstrate that the outcome of these procedures can be further enhanced. This will
improve outcomes for patients by decreasing the number of revision procedures and have a
significant cost benefit to the Australian community.

This discussion document reviews AOA NJRR and government data on joint replacement surgery
and compares this to available data from other countries. It identifies particular areas where
outcomes of joint replacement surgery can potentially be improved and discusses possible
approaches to achieve this.

AOA NJRR DATA

The Australian Orthopaedic Association established the National Joint Replacement Registry to
obtain quality data and use this to enhance outcomes. The justification for establishing the Registry
was in part based on the success of the Swedish National Hip and Knee Registries. These Registries
were established almost twenty years prior to the Australian Registry. The Commonwealth provided
funding to the AOA and the Registry commenced in September 1999. Data collection initially
started in South Australia and then other states were progressively added in a staged manner. New
South Wales, the most populous state was added last. Most hospitals undertaking hip and knee
replacements in Australia were contributing to the Registry by mid 2002. It was not until the end of
2003 however, that the first full calendar year of national data was collected. As of 30 June 2006 the
Registry has information on almost 300,000 hip and knee replacement procedures.

The Registry was established because data necessary to assess outcomes was not available from any
other source. The unique aspect of Registry data is that it links an individual patient, their diagnosis,
the operative joint (including side) and the specific prostheses used. All prosthetic components used
in each procedure are identified by catalogue and lot number. The use of catalogue and lot number
uniquely identifies each prosthetic component used. The Registry is then able to determine success
or otherwise of a procedure by linking that data to any subsequent procedures notified to the
Registry for that individual.

Notification of a subsequent procedure is an absolute indication that there has been a failure of the
previous procedure. When a subsequent procedure is notified, the reason for the revision is
provided and any new components used are identified. The identification of new components
establishes which of the components from the previous procedure have been replaced. This
information is then used to specifically link the replaced components to the reason for that revision.

The basic design of the AOA NJRR is very simple. It can be thought of as having two components.
The first is database maintenance and analysis. The second is the reporting mechanisms of the
Registry. Data on all hip and knee replacement procedures undertaken are collected into a stand-
                                               10
alone database. The data are collected from hospitals on paper forms. The Registry has the capacity
to accept electronic data but currently there is no hospital within Australia that is able to provide the
required data in this format. The paper forms are submitted to a central facility where they are
manually entered by Registry staff.

The AOA NJRR data also includes a comprehensive database of all prostheses available and used
since the establishment of the Registry. Each of the components is identified by catalogue and lot
number. There are additional databases within the Registry, many to do with data management and
verification.

The Registry has also been designed to handle additional information on procedures that may be
supplied to the Registry. These include the potential to handle additional clinical and radiological
outcomes data.

Considerable effort is made to ensure the quality of the data. All data being submitted is checked for
completeness and “integrity”. Completeness checks ensure that all the essential data have been
reported. This includes; date of birth, date of procedure, side on which the procedure was
performed and, so on. This process identifies missing data. Integrity checking is a determination of
the consistency of the data. It involves “intra procedure” component checking, to ensure that if a
primary procedure is being reported all the requisite components are included. Also an “inter
procedure” component check is undertaken. This is to establish that an individual has only one
primary procedure, hip or knee, on each side. Additionally each reported procedure is subjected to
many further integrity checks. These checks are repeated if any subsequent procedures are
submitted for an individual. The irregularities identified usually involve the identification of
incomplete or anomalous data. Vigorous attempts are then made to resolve all issues raised by the
integrated checking processes. This is done by returning queries to the reporting hospital for
resolution.

The Registry also works diligently to ensure that all hip and knee replacement procedures performed
are reported. It does this by checking the data received against state separation data. This is data
that is required to be provided to each state government following discharge of a patient from a
hospital within that state. The data is then subsequently provided to the Registry by individual state
governments and is cross-matched with Registry data to identify procedures that have not been
notified to the Registry. The Registry is then able to retrieve data for these “missing” procedures by
notifying the relevant hospital of the need to complete a registry form for the identified procedure.

The final piece of information incorporated into the Registry is vital status of individuals on the
Registry. This is needed for two reasons: first to assist in the “inter procedure” data integrity checks,
and second to allow valid statistical analysis of censored time to revision outcomes. The AOA NJRR
submits a file containing name, date of birth, gender and date of last contact to the Australian
Institute of Health and Welfare (AIHW) for people whose joint replacement data have been
submitted to the Registry. The AIHW uses this data to determine if there are any matches in the
National Death Index. The resulting matches are returned to the Registry and imported into the
database. This probabilistic matching process is not straightforward. Matches have a wide range of
certainty that vary from almost guaranteed to questionable matches. Sorting out the questionable
matches entails considerable manual checking. This is a lengthy process that is undertaken by the
Registry staff.




                                                11
The second component of the registry is the analysis of data. These analyses result in annual
reports, publications and data dissemination to hospitals, surgeons and manufacturing companies
through the ad hoc requesting process.

Figure 1: Schematic of the AOA NJRR




It must be emphasised that a critical factor in the successful establishment of the AOA NJRR and its
ongoing operational success is the central role of orthopaedic surgeons both individually and
through their professional organisation the Australian Orthopaedic Association (AOA). The
support of the AOA for the Joint Registry was necessary in obtaining cooperation of other
stakeholders and in gaining both the initial and ongoing funding for the Registry. The AOA and its
members also obtained the support of hospitals, orthopaedic companies and state governments.
Critically important is the ongoing role the AOA and its members have through the establishment
of the various governing committees and appointments to the AOA NJRR to ensure appropriate
data analysis and interpretation as well as the responsible accountable management of the Registry.

                                             12
GOVERNMENT DATA

Ten year Government data on joint replacement is presented as part of this discussion document. It
was obtained from each state and territory health department for the period 1 July 1994 to 30 June
2005. The requested information was for specific ICD-10-AM codes relating to hip and knee joint
replacement for all procedures undertaken in both public and private hospitals. These data have
been used solely for the purpose of providing information on the frequency of joint replacement for
a longer period than the AOA NJRR has been collecting data and, as such, are useful for examining
longer-term changes in the incidence of joint replacement surgery. These data do not provide any
prosthesis or outcome information.

INTERNATIONAL DATA

International data presented in this report has been largely obtained from National registries from
Sweden, Norway, Canada, England and Wales and New Zealand.

Swedish Hip Arthroplasty Registry
The Swedish National Hip Arthroplasty Registry commenced in 1979 with the purpose of
improving the outcome of primary hip replacements. All surgical units participate and
approximately 100% of hip replacements are registered. It monitors prostheses, surgical techniques
and prophylactic measures to minimize complications. It also provides an early warning system for
rapid implant failures. This information is provided back to all centers that perform total hip
replacement. Since 2001, all patients who have a total hip arthroplasty answer a questionnaire pre-
operatively and again after one, six, and ten years post-operatively. The questionnaire includes the
Charnley classification, EQ-5D and visual analog scales concerning pain and overall satisfaction and
is used by 31 of 81 units. Radiographic analysis is completed at six and ten years post-operatively.

All publications, annual reports, and scientific exhibitions, are communicated on a web site at
http://www.jru.orthop.gu.se/

Swedish Knee arthroplasty Registry
The Swedish Knee Registry commenced in 1975. All surgical units participate and over 90% of
knee replacements are registered. The initial aim of this Registry was to give early warning for
inferior prostheses. As with the Swedish Hip Registry, its principal focus is now on improving
outcomes of this surgical procedure. Link to the Swedish Knee Registry
http://www.ort.lu.se/knee/indexeng.html

Norwegian Arthroplasty Registry
The Norwegian Arthroplasty Registry started registration of total hip replacements in1987. In 1994
this was extended to include all types of prostheses. The main reason for establishing a nation-wide
registry was that throughout the 1970s, new hip prostheses were introduced without outcomes
evidence. After ten years, several of the prostheses were identified with high failure rates, but had, at
the time, already been used on large numbers of patients. The main purpose of the Registry is to
function as a surveillance tool to identify inferior implants as early as possible.
It is not compulsory to report joint replacements to the Norwegian Arthroplasty Registry.
Investigations have shown however that at least 95% of all joint replacements are reported. The
Annual Reports are published in Norwegian and English on the web at:
http://info.haukeland.no/nrl/#Publications


                                               13
Finnish Arthroplasty Register
This Registry was established in 1980. Since then it has been collecting information on total hip,
knee, ankle, shoulder, elbow, wrist and hand joint replacements. Health care authorities,
institutions and orthopedic units are obliged to provide the National Agency for Medicines with
information that is essential for maintenance of Registry data. Publication of the recent Implant
Yearbook on Orthopaedic Endoprosthesis and A study of the outcome of Arthroplasty Surgery in
Finland is available on the web at: http://www.nam.fi/english/publications/


Canadian Joint Replacement Registry
The CJRR is a national registry that collects information on total hip and total knee replacement
performed in Canada. It also follows joint replacement recipients over time in order to monitor
outcomes (including revision rates). The CJRR is collaboration between the Canadian Institute of
Health Information (CIHI) and the orthopaedic surgeons of Canada. Other key partners that
contributed to the successful development and implementation of the CJRR include the federal,
provincial, and territorial ministries of health, the Arthritis Society, the Canadian Arthritis Network
and the Ontario Joint Replacement Registry (OJRR). Currently, over 500 orthopaedic surgeons
from every province and one territory voluntarily participate in the CJRR.

The first data submission was in May 2001. The Registry captures information such as patient
demographics, the type of replacement, reason for replacement, surgical approach, fixation mode,
implant types, antibiotic use, deep vein thrombosis prophylaxis and operating room environment.
The data collection forms are sent directly to CIHI, where data verification and data entry are
completed.

The annual report provides information on hip and knee joint replacements performed in Canada.
It includes demographic, and surgery specific information but is yet to provide any outcome analysis.
The reports are available on the CIHI website:
http://secure.cihi.ca/cihiweb/dispPage.jsp?cw_page=services_cjrr_e


England and Wales Registry
The England and Wales National Registry was established in April 2003. The reason it was
established was to improve the outcomes of hip and knee joint replacement by monitoring
prosthesis outcomes and providing information to surgeons and other stakeholders. The
Department of Health and the Welsh Assembly Government established the Registry. Financing of
the Registry is through a levy placed on the sale of all total hip and knee prostheses. A Steering
Committee of 19 members, representing all stakeholders with an independent chair oversees the
registry’s activities. Contribution of information to the Registry is voluntary and approximately 60%
of hip and knee joint replacements are registered. The first annual report was published in
September 2004. This and subsequent reports can be found on the web at:
http://www.njrcentre.org.uk/Public/1st_annual_report.htm

New Zealand Joint Replacement Registry
The New Zealand Orthopaedic Association established this Registry so that technical information
about hip and knee surgery could be accurately recorded. In April 1998, the Register began as a
Christchurch based pilot study. A year later registrations expanded to include all surgical hospitals
throughout New Zealand. In January 2000 there was further expansion to include other joints. The
NZNJR now registers joint replacements for hip, knee, shoulder, elbow and ankle surgery.


                                               14
Approximately six months, two years, five years and then at three yearly intervals following surgery,
all registered patients are sent a questionnaire to measure outcomes and determine any
complications. This can now be answered on-line. The combination of technical data about the
joint implanted, and the individual patient assessment is analysed.




                                              15
INCIDENCE OF JOINT REPLACEMENT

Joint replacement is one of the most common major surgical procedures undertaken in Australia.
Over the last ten years the number of procedures has increased by 93.8%. Major increases have
occurred in both hip and knee replacement but the increase in knees has been considerably greater
(hips 61.9%, knees 138.4%). The yearly rate of change has been relatively constant with the number
of joint replacement procedures increasing by between 5-10% each year with the exception of 2001-
2002 where the increase was 13.4% (Table 1 and Figure 2).

Table 1: Hip and knee procedures from 1994-1995 to 2004-2005.
                                                                  Knee
                             Hip replacement                                               Total
 Year                                          % change           replacement   % change            % change
                             No.                                                           No.
                                                                  No.
  1994-1995                  18,635            N/A                13,371        N/A        32,006   N/A
  1995-1996                  19,132            2.7                14,542        8.8        33,674   5.2
  1996-1997                  20,127            5.2                15,456        6.3        35,583   5.7
  1997-1998                  21,379            6.2                17,317        12.0       38,696   8.7
  1998-1999                  21,800            2.0                18,832        8.7        40,632   5.0
  1999-2000                  22,717            4.2                19,936        5.9        42,653   5.0
  2000-2001                  24,285            6.9                22,252        11.6       46,537   9.1
  2001-2002                  26,689            9.9                26,089        17.2       52,778   13.4
  2002-2003                  27,833            4.3                28,003        7.3        55,836   5.8
  2003-2004                  29,165            4.8                29,899        6.8        59,064   5.8
  2004-2005                  30,166            3.4                31,870        6.6        62,036   5.0
  1994/95-2003/05                              61.9                             138.4               93.8

Figure 2: Hip and knee procedures from 1994-1995 to 2004-2005.


             35,000

             30,000

             25,000
   Numbers




             20,000

             15,000

             10,000

              5,000

                 0
                      1994- 1995- 1996- 1997- 1998- 1999- 2000- 2001- 2002- 2003- 2004-
                      1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
                                                          years


                                                      hips          knees



                                                             16
  Although change in absolute number of procedures provides important information, it does not
  reflect the true rate of change. In order to determine this the absolute numbers need to be corrected
  for change in population. The simplest approach to this is to express the number of procedures as a
  number per 100,000 population (Table 2). Expressing the data in this manner also allows
  comparison with international data. Australian data is available for the eight-year period between
  1997-1998 and 2004-2005. When hip and knee replacement procedures are combined there was an
  increase from 206.8 / 100,000 to 305.2 / 100,000 (47.6%).

  Hip procedures alone increased by 29.8%. Partial hips remained relatively constant, primary total
  hips increased by 40.5% and revision hips by 20%. All knee procedures increased by 69.5%.
  Revision knee procedures increased by 46.7%. It is not possible to calculate the change in the
  different types of primary knee replacement during this time as it was not required to identify the
  type of primary knee when notifying separation data to the government until 1999-2000. Prior to
  this time all primary knee replacements were reported as primary total knees. Primary knee
  procedures increased by 71.8% (Table 2).

  Table 2: Incidence of hip and knee procedures per 100,000 population

Type of joint replacement    1997 - 1998 1998 - 1999 1999 - 2000 2000 - 2001 2001 - 2002 2002 - 2003 2003 - 2004 2004 - 2005

population as at June 30th   18711300    18925900    19153400    19413200    19641000     19881500    20111300 20328600
Hip replacement
Partial                      26.4        26.8        27.6        28.2        28.5         28.5        29.2       28.0
Primary total                72.4        73.2        74.1        79.2        88.5         93.2        96.4       101.7
Revision                     15.5        15.2        16.9        17.7        18.9         18.3        19.4       18.6
Total hips                   114.3       115.2       118.6       125.1       135.9        140.0       145.0      148.4
Knee replacement
Patellar/trochlear           N/A         N/A         0.9         1.1         1.3          1.5         1.5        2.2
Unicompartment               N/A         N/A         11.3        14.4        16.5         17.9        17.5       16.0
Primary total                83.4        90.3        81.4        88.2        103.3        108.3       116.7      125.1
Revision                     9.2         9.2         10.4        10.9        11.7         13.1        13.0       13.5
Total knees                  92.5        99.5        104.1       114.6       132.8        140.8       148.7      156.8
Total                        206.8       214.7       222.7       239.7       268.7        280.8       293.7      305.2


  The most recent international figures available for other countries with national registries are for
  2004. Using 2004 Australian data to compare, it is evident that Australia has a lower rate of hip
  replacement compared to countries shown and this is 30% less than what was being undertaken in
  Sweden and New Zealand. Australia has the highest rate of knee replacement (Figure 3).




                                                         17
Figure 3: International comparison of number of procedures per 100,000 population


                        160
                        140
   Number per 100,000




                        120
                        100
                        80
                        60
                        40
                        20
                         0
                              Australia   Finland    New      Norw ay   Sw eden   Canada   Denmark
                                                    Zealand
                                                              Country              Hip      Knee



FUTURE RATE OF JOINT REPLACEMENT

The accuracy of predictions about the future rate of joint replacement is very much dependent on the
quality of information available to base the predictions on. Unfortunately it is difficult to be precise
about future rates as the disease burden of severe osteoarthritis within the Australian community is
unknown. This information is critical to determining a needs assessment for joint replacement
surgery. It involves establishing the disease burden of severe degenerative joint disease. Without
this knowledge it is difficult to be precise about the future rates of joint replacement surgery.

There are, however, a number of indicators as to what the rate is likely to be. It is clear that the
current yearly rate of increase in joint replacement surgery has been relatively constant during the
last ten years. At this point in time there is no obvious reason to suggest that this is likely to change.
If the current rate of increase were to continue then the number of joint replacement procedures
performed each year would double by 2016. The vast majority of joint replacement surgery is
undertaken for degenerative joint disease. Various estimates of the increasing disease burden due to
degenerative musculoskeletal conditions in the Australian community are that these conditions will
double before 2020. This tends to support the intuitive assumption that there is unlikely to be a
change in the current rate of increase. Most of these estimates are based largely on the anticipated
ageing of the Australian population. One underlying assumption that may be incorrect is that age
related incidence of degenerative joint disease will remain constant. There is evidence to suggest that
this may not be the case.

As previously mentioned, knee replacement has increased at double the rate of hip replacement.
This difference is not just a result of ageing - increasing numbers of younger patients are receiving
knee replacement surgery. The reason for the greater increase in the rate of knee replacement
remains uncertain and it should be an area of intensive investigation. There is clearly a problem and
the reasons behind it are not understood.

An important aspect that may also impact the rate of joint replacement surgery is the likely situation
that there remains significant unmet need within the Australian community. This is certainly true
for hip replacement as the rate of joint replacement in this country is 30% less than some other
developed countries. The rate of knee replacement in Australia is one of the highest national rates.
This may suggest that knee replacement surgery is undertaken at a lower clinical threshold than
                                                                18
elsewhere. Available evidence, however, suggests that this is not the case as Australian centres where
the clinical severity of patients undergoing knee replacement surgery has been assessed demonstrates
that severity is worse or comparable to patients receiving knee replacement elsewhere. This tends to
indicate that on average individuals receiving knee replacement in Australia have more severe disease
compared to overseas. The existence of unmet need is also supported by the presence of lengthy and
growing waiting lists within the public system. This is despite an increasing rate of joint replacement
within both public and private systems.

PUBLIC VERSES PRIVATE JOINT REPLACEMENT

Government data has been used to determine the change in public verses private joint replacement
surgery. The data is based on the facility where the surgery was undertaken. This data is only
available for the financial years 1997-1998 to 2004-2005.

There has been an increase in both public and private hip and knee replacements during this eight
year period. The increase observed in private joint replacement has been far greater (private 90.8%,
public 27.9%, Table 3). In 1997-1998 just over half of all joint replacement procedures where
undertaken in a private hospital (51.5%). In the 2004-2005 financial year this had increased to
61.3%.

Hip replacement increased by 41.1% during this period and knee replacement by 84%. Public hip
replacement increased by 17.8% and private by 67.8% (Table 4). Public knee replacement
procedures increased by 43.6%. Private knee replacements increased by 113.9% (Table 5).

When considering the yearly changes, the largest annual increases occurred in 2000-2001 and
2001-2002.

Table 3: Public and private joint replacement

    Year                  Public                 Private                   Total Joints (hip & knee)
    1997-1998                 18,777 (N/A)           19,919 (N/A)              38,696 (N/A)
    1998-1999                 19,195 (2.2%)          21,437 (7.6%)             40,632 (5.0%)
    1999-2000                 19,193 (0.0%)          23,460 (9.4%)             42,653 (5.0%)
    2000-2001                 19,290 (0.5%)          27,247 (16.1%)            46,537 (9.1%)
    2001-2002                 20,851 (8.1%)          31,937 (17.2%)            52,788 (13.4%)
    2002-2003                 21,797 (4.5%)          34,039 (6.6%)             55,836 (5.8%)
    2003-2004                 23,070 (5.8%)          35,994 (5.7%)             59,064 (5.8%)
    2004-2005                 24,022 (4.1%)          38,014 (5.6%)             62,036 (5.0%)
    1997/98-2004/05       27.9%                  90.8%                     60.3%




                                                19
Table 4: Public and private hip replacement

    Year                 Public                  Private                   Total (hip)
    1997-1998              11,417 (N/A)             9,962 (N/A)              21,379 (N/A)
    1998-1999              11,455 (0.3%)            10,345 (3.8%)            21,800 (2.9%)
    1999-2000              11,493 (0.3%)            11,224 (8.5%)            22,717 (4.2%)
    2000-2001              11,547 (0.5%)            12,738 (13.5%)           24,285 (6.9%)
    2001-2002              12,179 (5.5%)            14,510 (13.9%)           26,689 (9.9%)
    2002-2003              12,577 (3.3%)            15,256 (5.1%)            27,833 (4.3%)
    2003-2004              13,193 (4.9%)            15,972 (4.7%)            29,165 (4.8%)
    2004-2005              13,451 (2.0%)            16,715 (4.7%)            30,166 (3.4%)
    1997/98-2004/05      17.8%                   67.8%                     41.1%




Table 5: Public and private knee replacement

    Year                 Public                  Private                   Total (hip)
    1997-1998              7,360 (N/A)              9,957 (N/A)              17,317 (N/A)
    1998-1999              7,740 (5.2%)             11,092 (11.4%)           18,832 (8.7%)
    1999-2000              7,700 (-0.5%)            12,236 (10.3%)           19,936 (5.9%)
    2000-2001              7,743 (0.6%)             14,509 (18.6%)           22,252 (11.6%)
    2001-2002              8,672 (12.0%)            17,427 (20.1%)           26,099 (17.3%)
    2002-2003              9,220 (6.3%)             18,783 (7.8%)            28,003 (7.3%)
    2003-2004              9,877 (7.1%)             20,022 (6.6%)            29,899 (6.8%)
    2004-2005              10,571 (7.0%)            21,299 (6.4%)            31,870 (6.6%)
    1997/98-2004/05      43.6%                   113.9%                    84.0%



EXPENDITURE

The AOA NJRR has previously estimated total acute care and prostheses specific hip and knee
replacement expenditure in Australia by third party payers, (i.e. government and health insurers). It
has done this by utilising data from the National Hospital Cost Data Collection (NHCDC) as well
as the state and territory health department separation data. Adjustments for inflation have been
made by using the Consumer Price Index to determine constant Australian dollars to express all
expenditure.

The NHCDC calculations do not include fees for surgeons, assistants, anaesthetists, nor imaging
and pathology for the private sector. An estimate of these costs has been made using the Medical
Benefits Schedule and information obtained from radiological and clinical laboratories. The
expenditure estimates are calculations for inpatient management and do not include pre-operative
investigations or the cost of post-operative rehabilitation. Patient co-payments are also not included.

NHCDC is a sampling exercise that estimates detailed population and cost activity data. These data
are usually made available to the Registry 12 months after the end of the relevant financial year. As a
consequence expenditure estimates are a full financial year prior to the last available separation data.
The Registry has not been able to determine the estimated expenditure for 2003-2004, as the

                                               20
NHCDC has not been in a position to provide the required data at this point in time. The Registry
however does have available, estimates for the four financial years from 1999-2000 to 2002-2003.

Total acute care expenditure for 2002-2003 for all hip and knee replacements was estimated to be
$867.8 million. Compared to 1999-2000 the yearly estimated expenditure had increased by $213.8
million or 32.7% (Table 6). This is marginally higher than the increase in the number of
procedures during this period, 30.9% (Table 6 and Table 1)

Table 6: Estimated total acute care expenditure 1999-2000 to 2002-2003

                                        Hip                         Knee                   Total (hip + knee)
       Year
                            $Million       %           $Million       %              $Million          %

       1999-2000            349.1            N/A       305.1             N/A         654.2           N/A
       2000-2001            353.1            1.1%      304.5             -0.2%       657.6           0.6%
       2001-2002            417.5            18.4%     398.1             30.7%       815.6           24%
       2002-2003            466.5            11.7%     401.3             0.8%        867.8           6.4%
       1999-2003                             33.6%                       31.5%                       32.7%

Although it is apparent that the average expenditure per joint replacement procedure has remained
almost constant during this period, there were differences between hip and knee procedures.

Hip replacement expenditure alone was estimated to be $466.5 million for 2002-2003. The increase
in yearly estimated expenditure compared to 1999-2000 for hips was $117.4 million or 33.6%. The
number of hip replacement procedures undertaken each year increased by 22.5% during that same
period (Table 7 and Table 1).

Knee replacement expenditure was estimated at $401.3 million for 2002-2003, an increase in yearly
expenditure of $96.2 million or 31.5% compared to 1999-2000. The number of knee replacement
procedures increased by 40.5% (Table 7 and Table 1). These figures indicate that for hip
replacement there has been an increase in the average cost per procedure but that for knees there has
been a decrease.

Table 7: Estimated prostheses expenditure 1999-2000 to 2002-2003

                                    Hips                          Knee                  Total (Hip + Knee)
Year
                     $Million         %              $Million        %           $Million         %

1999-2000            87                N/A           88.9           N/A          175.9            N/A
2000-2001            95.5              9.8%          98.2           10.5%        193.7            10.1%
2001-2002            126.5             32.5%         146.7          49.4%        273.2            41.0%
2002-2003            153.1             21.0%         149.5          1.9%         302.6            10.8%
1999-2003                              76.0%                        68.2%                         72%
                                                                    %




                                                     21
Expenditure can be considered as being made up of two major components, the cost of the
prostheses and then other costs. Prostheses are a major component of total expenditure accounting
for 34.9% of the total. For both hip and knee replacement there was a significant increase in
prostheses costs during this four-year period. Hip replacement prostheses expenditure increased by
76% and knee prostheses expenditure by 68.2% (Table 7). The change in prostheses costs were
higher in the private sector (private 87.2%, public 39.9%) (Table 8)

Table 8: Estimated prostheses expenditure (public and private)

                                     Private                     Public           Total (Priv + Pub)
    Year
                          $Million      %             $Million      %       $Million        %

    1999-2000             119.5         N/A           56.4          N/A     175.9           N/A
    2000-2001             127.3         6.5%          66.4          17.7%   193.7           10.1%
    2001-2002             197.8         55.4%         75.4          13.6%   273.2           41.0%
    2002-2003             223.7         13.1%         78.9          4.6%    302.6           10.8%
    1999-2003                           87.2%                       39.9%                   72%



DIFFERENT CATEGORIES OF JOINT REPLACEMENT

There are different categories of hip and knee replacement and prior to discussing Australian
Orthopaedic National Joint Registry Data it is necessary to explain these categories.

Hip Replacement
There are a variety of different general categories of hip replacement procedures. They are primary
partial, primary total and revision procedures. Primary procedures are procedures that have been
undertaken for the first time. Revision procedures include all procedures (both primary and
revision) that have undergone subsequent surgery where at least one prosthetic component has been
removed or exchanged. Both primary partial hips and primary total hips are further classified into
different sub-categories. These sub-categories reflect a broad classification of the type of prostheses
used. There are many individual varieties of prostheses used within each of these sub-categories.
The sub-categories for primary partial hips are unipolar monoblock, unipolar modular and bipolar
procedures. Primary total hips are sub-categorised as conventional total hip, resurfacing and thrust
plate procedures.

Primary total hip replacement is the most common category of hip replacement (70.4%). Primary
partial hips account for 16.8% and revisions 12.7%. Within the primary total category 92.0% are
conventional total hip, 7.8% resurfacing and 0.1% thrust plate procedures. In the primary partial
category 59.4% are unipolar monoblock prostheses, 13.2% unipolar modular and 27.4% bipolar
prostheses (Table 9).




                                                 22
Knee replacement
As with hip replacement there are different categories of knee replacement procedures. There are
four different types of primary procedure; unispacer, patellar/trochlear, unicompartment and total
knee replacement. In addition there are also revision procedures. Unlike hip procedures there are
no sub-categories.

The knee is considered to have three anatomical joint compartments; the medial tibio-femoral joint,
the lateral tibio-femoral joint and the patello-femoral joint. Unicompartment knee replacements are
used to replace most commonly the medial or occasionally the lateral tibio-femoral joint,
patella/trochlear replacement the patello-femoral joint and the total knee replaces all three joint
compartments. A total knee replacement may or may not be combined with a patella component
which is used to resurface the articulating surface of the patella. The unispacer is a single
component prosthesis that is used as a spacing device between the joint surfaces of the medial tibio-
femoral joint. A revision knee procedure is a repeat procedure on either a previous primary or
revision procedure where one or more of the components are removed or exchanged.

Primary total knee replacement is the most common category of knee replacement (77.7%). The
proportional use of the other categories is; patella/trochlear (0.5%), unicompartmental (13.2%), and
revision procedures (8.6%). There have only been 39 unispacer procedures performed in this
country.

INDICATIONS AND USE OF DIFFERENCE CATEGORIES OF JOINT REPLACEMENT

Data from the AOA NJRR has been used to determine the indications and use of the different
categories of hip and knee replacement. It is based on an analysis of 269.712 procedures (130,909
hips and 138,803 knees) recorded by the Registry to the end of 2005.

Hip replacement procedures
There are differences in the indications for use for each of the general categories of hip procedures.
Primary partial hips are most commonly used for the treatment of fractures of the neck of the femur;
in particular displaced sub-capital fractures in the elderly. This diagnosis is the principle reason in
94.8% of all primary partial hips. Primary total hips are largely used in the treatment of chronic
joint disease, in particular arthritis, with osteoarthritis being the most common diagnosis (88.4%).
Revision procedures are undertaken when there has been a failure of a previous procedure. This
may have been a primary partial hip, primary total hip or a revision. The most common reason for
revision is loosening of one or more of the components (48.4%).

There are major differences in age and gender of individuals receiving the different categories of hip
replacement. This is a reflection of the different indications. Primary partial hips are most
commonly used in older females (74.5% female, 82.8% 75 years or older). Primary total hips are
also used more commonly in females (53.3%) but the age is generally younger than for primary
partial hip replacement with only 29.1% being 75 years or older. Revision procedures have a similar
gender distribution to primary total hips but the age differs in that 43.4% are 75 years or older
(Table 9 and Table 10).

There are differences in the age and gender pattern of use in the different sub-categories for primary
partial and primary total hip replacements. In the primary partial hip category, unipolar modular
and bipolar prostheses are used more frequently than unipolar monoblock prostheses in individuals
younger than 75 years. In the primary total hip category, resurfacing and the small number of thrust
plate procedures tend to be used in younger males (Table 9 and Table 10).
                                               23
        Table 9: Hip replacement and gender
                                                            Female                        Male                          Total
          Type of hip replacement
                                              No.              %              No.          %                 No.          %
          Unipolar Monoblock                  9385             74.9           3147          25.1             12532           56.9

          Unipolar Modular                    2586             74.8           873           25.2             3459            15.7

          Bipolar                             4435             73.5           1599          26.5             6034            27.4

          Primary Partial (16.8%)             16406            74.5           5619          25.5             22025           100.0

          Conventional Total                  46997            55.4           37875         44.6             84872           92.0

          Resurfacing                         2084             28.9           5121          71.1             7205            7.8

          Thrust Plate                        37               27.8           96            72.2             133             0.1

          Primary Total (70.4%)               49118            53.3           43092         46.7             92210           100.0

          Revision (12.7%)                    9023             54.1           7651          45.9             16674           100.0

          Total                               74547            56.9           56362         43.1             130909          100.0



        Table 10: Hip replacement and age
                                      <=54                 55-64              65-74               75-84               >=85                    Total
Type of hip replacement
                               No.        %         No.         %     No.         %        No.         %      No.         %          No.              %
Unipolar Monoblock             33        0.3        151        1.2    1028         8.2     5227       41.7    6093       48.6        12532            56.9
Unipolar Modular               65        1.9        175        5.1    629          18.2    1539       44.5    1051       30.4        3459             15.7
Bipolar                        161       2.7        398        6.6    1153         19.1    2725       45.2    1597       26.5        6034             27.4
Primary Partial                259       1.2        724        3.3    2810         12.8    9491       43.1    8741       39.7        22025            100.0
Conventional Total             10046     11.8       18831      22.2   29208        34.4    22765      26.8    4022       4.7         84872            92.0
Resurfacing                    3684      51.1       2785       38.7   688          9.5     48         0.7     .          .           7205             7.8
Thrust Plate                   55        41.4       54         40.6   22           16.5    2          1.5     .          .           133              0.1
Primary Total                  13785     14.9       21670      23.5   29918        32.4    22815      24.7    4022       4.4         92210            100.0
Revision                       1707      10.2       2672       16.0   5049         30.3    5608       33.6    1638       9.8         16674            100.0
Total                          15751     12.0       25066      19.1   37777        28.9    37914      29.0    14401      11.0        130909           100.0




                                                                       24
        Knee Replacement Procedures
        The indication for almost all primary knee replacement procedures is osteoarthritis (unispacer
        100%, patella/trochlear 98.8%, unicompartment 98.6% and primary total knee replacement
        96.4%). The principal cause for revision knee surgery is aseptic loosening (37.7%).

        There are differences in age and gender of individuals receiving the different categories of knee
        replacement. These differences relate to the relative indications for use of each of the categories.
        Unispacer, unicompartment and patella/trochlear replacements are often considered “lesser”
        procedures than a total, as the entire knee joint is not replaced. They are most often used in younger
        individuals when the evident osteoarthritis is more localised to a specific joint compartment.

        Table 11: Knee replacement and gender
                                                            Female                          Male                               Total
         Type of knee replacement
                                              No.              %               No.           %                  No.              %
         Unispacer                            19               48.7            20             51.3              39               0.0
         Patella/trochlear                    518              76.7            157            23.3              675              0.5
         Unicompartmental Knee                8937             48.7            9418           51.3              18355            13.2
         Primary Total Knee                   61903            57.4            45941          42.6              107844           77.7
         Revision Knee                        6166             51.9            5724           48.1              11890            8.6
         Total                                77543            55.9            61260          44.1              138803           100.0



        Table 12: Knee replacement and age
Type of knee                        <=54                  55-64              65-74             75-84                  >=85                  Total
replacement                  No.        %          No.         %      No.         %     No.         %       No.           %        No.          %
Unispacer                    18        46.2        17         43.6    3          7.7    1            2.6    .            .         39          0.0
Patella/trochlear            259       38.4        192        28.4    126        18.7   86           12.7   12           1.8       675         0.5
Unicompartmental             2502      13.6        5876       32.0    5969       32.5   3679         20.0   329          1.8       18355       13.2
Primary Total Knee           7234      6.7         24158      22.4    40929      38.0   32129        29.8   3394         3.1       107844      77.7
Revision Knee                1036      8.7         2416       20.3    3978       33.5   3920         33.0   540          4.5       11890       8.6
Total                        11049     8.0         32659      23.5    51005      36.7   30354        28.7   4275         3.1       138803      100.0


        CEMENTED AND CEMENTLESS FIXATION

        As well as the different categories and sub-categories, joint replacement procedures are divided into
        three broad types with respect to the way the different prosthetic components are fixed to bone.
        Fixation may be cemented, cementless or a combination of the two. The combination is referred to
        as hybrid fixation. Cemented prostheses are designed so that bone fixation is achieved using bone
        cement (methyl methacrylate). This is a polymer that is mixed at the time of surgery to a putty
        consistency. It is then applied, often with pressure to the bone surface that interfaces with the
        prosthesis. The prosthesis is then positioned under pressure and held. The cement interdigitates
        with the bone surface and depending on the design of the prosthesis also with the bone interface
        surface of the prosthesis. It sets hard in a matter of minutes and holds the prosthesis in place.

        A cementless component is specially designed so that bone interface surface of the prosthesis will
        allow bone to grow into the surface. This bone ingrowth provides long-term fixation of the
        prostheses. The bone ingrowth takes a variable period of time to occur and prior to this the
        cementless prosthesis relies on press fit fixation to maintain the position of the component.

                                                                          25
For both total hip and total knee replacement there are two major bone interface components. The
hip has a femoral and an acetabular component. The knee has a femoral and a tibial component.
Hybrid fixation is when one of these components is fixed using cement and the other using
cementless fixation. Hip replacement hybrid fixation usually involves cement fixation of the femoral
component and cementless fixation of the acetabular component. Hybrid fixation of the knee is
most commonly cement fixation of the tibial component and cementless fixation of the femoral
component. Partial hip replacements only have one component that fixes to bone which is on the
femoral side. Therefore there are only two alternatives for fixation of partial hip replacements; they
are either cemented or cementless.

Prostheses are specifically designed to be either cemented or cementless. As the bone interface
surface of cementless prostheses are considerably more complex with respect to manufacture so they
are accordingly also more expensive. Usually they are two or more times the cost of a cemented
component. Originally joint replacement prostheses were all cemented. With time, it became
apparent that a percentage of these prostheses became loose and the percentage increased the longer
it was from the time of the initial surgery. At that time the reason this occurred was not clearly
understood. One of the possibilities considered was a detrimental reaction to the cement at the
bone/cement surface interface. This was referred to as cement toxicity. In order to address this
issue prostheses were developed that did not require cement to maintain fixation.

Since that time understanding of the mechanisms of joint replacement failure has improved.
Although not yet fully understood it is clear that one major reason is the production of small wear
particles at the articulating and other surfaces of the prosthesis. These particles initiate a chronic
inflammatory reaction. The extent of the reaction is dependent on the nature and the number of the
particles produced. The consequence of this low-grade inflammation is that a variety of
inflammatory mediators are produced and they stimulate localised bone resorption to occur around
the prostheses. The prosthesis eventually becomes loose within the bone and causes pain. Most
joint replacements are revised because of loosening. This reaction occurs independently of the use
of cement.

In the Australian community the method of fixation varies depending on category and type of joint
replacement procedure. There is also considerable regional variation.

Fixation of hip replacements
There are differences in fixation depending on the category of prostheses. Partial hips are mostly
cementless (53.4%) but this varies depending on the different categories of partial hip replacement.
The most common category of partial hip, the unipolar monoblock prostheses, is most often used
cementless (75.9%). The other two types of partial hip, the unipolar modular prostheses and
bipolar prostheses, are most often used cemented (77.8% and 83.6%) (Table 13).

Table 13: Fixation — Primary partial hip replacement
                Unipolar Monoblock   Unipolar Modular          Bipolar           All Patients
Fixation
               No.       %         No.       %          No.       %      No.           %
Cemented       3038      13.7      2692      12.2       5044      22.8   10774         48.8
Cementless     9568      43.3      767       3.5        990       4.5    11325         51.2
Total          12606     57.0      3459      15.7       6034      27.3   22099         100.0




                                                26
The reasons for this difference in fixation for the different categories of partial hip replacement are
largely historical. The two most common unipolar monoblock prostheses are the Austin Moore and
the Thompson prostheses. These prostheses were originally designed over 50 years ago and have
been the main stay for the management of displaced subcapital fractures of the neck of femur in
elderly patients since that time. The essential difference between the two is that the Austin Moore is
a cementless prosthesis and the Thompson is cemented. The Austin Moore has been used more
frequently in Australia. Unipolar modular and Bipolar prostheses are more recent designs. They
utilise a conventional hip replacement stem, most often a cemented stem, combined with either a
single large femoral head (unipolar modular prosthesis) or a conventional femoral head and a
second head that fits over the conventional head and is sized appropriately to fit into the retained
acetabulum.

The most common method of fixation used in primary total hip replacement is also dependent on
the sub-category of replacement used. Conventional hips can be either cementless, cemented or
hybrid. Hybrid hips are where one of the major components are cemented and this is usually the
femoral component. In Australia cementless hip replacement is used most commonly (46.5% of all
primary total hips) followed by hybrids (31.9%) and then cemented hips (13.6%). Resurfacing hips
are nearly all hybrid fixation and the thrust plate procedures are cementless (Table 14).

Table 14: Fixation — Primary total hip replacement

                    Conventional Hips
Prosthesis                                                                       Resurfacing        Thrust Plate          Total
                         Cemented           Cementless            Hybrid
Fixation
                    No.      %          No.      %         No.       %      No.        %       No.        %        No.       %
ACT/NT              63       3.0        1234     59.7      554       26.8   216        10.4    .          .        2067      100.0
NSW                 1207     4.7        14899    57.6      7632      29.5   2132       8.2     11         0.0      25881     100.0
QLD                 4404     30.4       3839     26.5      5198      35.9   1036       7.2     .          .        14477     100.0
SA                  1946     18.8       3614     34.9      4095      39.6   696        6.7     .          .        10351     100.0
TAS                 310      9.3        2638     79.4      325       9.8    41         1.2     10         0.3      3324      100.0
VIC                 3846     14.8       11107    42.7      8215      31.6   2846       10.9    1          0.0      26015     100.0
WA                  784      7.8        5540     54.9      3422      33.9   238        2.4     111        1.1      10095     100.0
Australia           12560    13.6       42871    46.5      29441     31.9   7205       7.8     133        0.1      92210     100.0


Fixation of knee replacements
Fixation of the different types of knee replacement depends on the category of the knee replacement
procedure. Unispacer procedures have no fixation; they are placed between articulating surfaces of
the femur and the tibia on the medial side of the knee joint. Patella/trochlear replacements are
almost all cemented, as is the case with unicompartment knee replacements (90.9%) (Table 15).
The most common method of fixation for primary total knee is cement fixation (48.7%). Hybrid
(tibial component only 26.6%, femoral component only 0.8%) and cementless (23.8%).

Table 15: Fixation — Primary unicompartment knee replacement

        Fixation                                  Number                               %
            Tibial and femoral cementless         1531                                 8.3
            Tibial and femoral cemented           16689                                90.9
            Tibial only cemented                  71                                   0.4
            Femoral only cemented                 64                                   0.3
            Total                                 18355                                100.0

                                                             27
Table 16: Fixation — Primary total knee replacement
                                                                                Patella used
                                                 Total
      Fixation                                                   Patella cemented          Patella cementless
                                        Number     %          Number      %            Number        %
      Tibial and femoral cementless     25707      23.8       17348       67.5         2631          10.2
      Tibial and femoral cemented       52542      48.7       26376       50.2         52            0.1
      Tibial only cemented              28732      26.6       17817       62.0         358           1.2
      Femoral only cemented             864        0.8        429         49.7         9             1.0
      Total                             107845     100.0      61970       57.5         3050          2.8



REGIONAL DIFFERENCES AND TRENDS IN HIP REPLACEMENT

AOANJRR has been used to determine state and territory variation in use of the different categories
of hip replacement as well as any trends in change of that use. The data presented from the Registry
is for each calendar year from the beginning of 2001 to the end of 2005. The proportional use of
the three broad categories of hip replacement (primary partial, primary total and revision procedures)
has remained relatively constant during this time both nationally and within the states and
territories. There is only small variation in the proportional use of these categories of hip
replacement between the states and territories (Figure 4).

Figure 4: Trends in hip replacement by state and territory
    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005

                        2001
                        2002
                        2003
                        2004
                        2005
                                      2001
                                      2002
                                      2003
                                      2004
                                      2005

                                                 2001
                                                 2002
                                                 2003
                                                 2004
                                                 2005

                                                           2001
                                                           2002
                                                           2003
                                                           2004
                                                           2005

                                                                        2001
                                                                        2002
                                                                        2003
                                                                        2004
                                                                        2005

                                                                                     2001
                                                                                     2002
                                                                                     2003
                                                                                     2004
                                                                                     2005

                                                                                                  2001
                                                                                                  2002
                                                                                                  2003
                                                                                                  2004
                                                                                                  2005




                 AUST      NSW           VIC        QLD         WA           SA          TAS         ACT/NT


                                 Partial Hip               Total Hip                  Revision Hip




                                                     28
Partial hip replacement
The proportional use of partial hip replacement has remained constant during the last five years.
The main indication for this procedure is treatment of displaced subcapital fractures of the hip
particularly in the elderly. It is usually undertaken as an emergency procedure. The incidence of
this injury is anticipated to increase as the population ages. There has been an increase in the rate of
partial hip replacement per 100,000 population since 1997-1998 (26.4/100,000) peaking in 2003-
2004 (29.3/100,000) but declining in 2004-2005 to 28.0/100,000.

There are some minor regional variations in the proportion of partial hip replacements undertaken
but it is thought that these most likely reflect differences in the average age of the population in those
states and territories rather than any differences in patient selection (Figure 4).

There are however very significant changes and variation in regional patterns of use for the different
sub-categories of partial hip replacement. It has been mentioned that nationally monoblock partial
hips are the most common group of prostheses used (47.9% in 2005). Their use has however
declined from 67.8% in 2001. There is also a marginal reduction in the use of bipolar prostheses,
from a high of 30% in 2002 to 27% in 2005. Unipolar modular prostheses have increased from
10.3% in 2001 to 25.1% in 2005 (Figure 5). The proportional use of these sub-categories also
varies significantly between the different states and territories. All demonstrate reducing use of
monoblock prostheses and for most this is associated with an increase in use of unipolar modular
prostheses. Although the use of unipolar modular prostheses is increasing in most states bipolar
prostheses remains the most common alternative to unipolar monoblock prostheses with the
exception of South Australia and Tasmania where there is very little use of the bipolar (Figure 5).

Figure 5: Trends in primary partial hip replacement by state and territory
              100%


               80%


               60%


               40%


               20%


                0%
                     2001
                     2002
                     2003
                     2004
                     2005
                                2001
                                2002
                                2003
                                2004
                                2005
                                       2001
                                       2002
                                       2003
                                       2004
                                       2005
                                                2001
                                                2002
                                                2003
                                                2004
                                                2005
                                                           2001
                                                           2002
                                                           2003
                                                           2004
                                                           2005
                                                                     2001
                                                                     2002
                                                                     2003
                                                                     2004
                                                                     2005
                                                                             2001
                                                                             2002
                                                                             2003
                                                                             2004
                                                                             2005
                                                                                        2001
                                                                                        2002
                                                                                        2003
                                                                                        2004
                                                                                        2005




                       AUST      NSW      VIC        QLD        WA     SA        TAS     ACT/NT


                      Bipolar              Unipolar Monoblock               Unipolar Modular




                                                29
Total hip replacement
Conventional total hip is by far the most common type of primary total hip replacement. In 2005,
91% of all primary total hips were conventional hips. The proportion of conventional hips has
declined in recent years due to the increasing use of resurfacing procedures. A resurfacing procedure
differs from a conventional total hip replacement on the femoral side. A conventional total hip
involves resection of the femoral head that is replaced with a stem component inserted down the
femoral shaft and fitted with a smaller femoral head. The resurfacing procedure involves reshaping
the femoral head and covering it with a component that replaces the surface of the head. The use of
resurfacing procedures has increased from 5.6% of all primary total hips in 2001 to 8.9% in 2005.
There is considerable regional variability in the use of resurfacing procedures. It is as high as 17.1%
of all primary total hip procedures in the Australian Capital Territory/Northern Territory but as low
as 1.4 % in Western Australia (Figure 6). The third type of primary total hip replacement is the
Thrust plate procedure. This is a specific design of primary total hip prostheses that has limited use
in Australia.

Figure 6: Trends in primary total hip replacement by state and territory
              100%


               80%


               60%


               40%


               20%


                0%
                     2001
                     2002
                     2003
                     2004
                     2005
                              2001
                              2002
                              2003
                              2004
                              2005
                                        2001
                                        2002
                                        2003
                                        2004
                                        2005
                                                  2001
                                                  2002
                                                  2003
                                                  2004
                                                  2005
                                                             2001
                                                             2002
                                                             2003
                                                             2004
                                                             2005
                                                                       2001
                                                                       2002
                                                                       2003
                                                                       2004
                                                                       2005
                                                                                2001
                                                                                2002
                                                                                2003
                                                                                2004
                                                                                2005
                                                                                       2001
                                                                                       2002
                                                                                       2003
                                                                                       2004
                                                                                       2005


                      AUST      NSW         VIC        QLD      WA         SA    TAS    ACT/NT


                       Conventional Total Hip            Resurfacing Hip         Thrust Plate



Revision hip replacement
The proportion of revision hips has reduced from 13.6% in 2001 to 12.1% in 2005. Revisions are
categorised into minor or major revisions. A major revision involves the removal or exchange of a
major component. A major component is a component that interfaces with bone. This is the
femoral component or and acetabular cup or shell. There are two types of major revision. A partial
major revision involves the removal or exchange of one of these components. The other type of
major revision is a total revision, where both of these components are removed or exchanged. A
minor revision is a revision procedure where a major component is not removed or exchanged. In
general this involves replacement of the femoral head and or the acetabular insert with the exception
of when the femoral head is a resurfacing prostheses.




                                                  30
Most revisions are partial major revisions. In 2005, 52.3% of all revisions were major partial
revisions, 31.6% were major total revisions and 16.1% were minor revisions. The proportion of
major total revisions has decreased from 38.2% in 2001. Major total revisions are undertaken more
frequently in Queensland and Victoria. Minor revisions are more common in Western Australia
(Figure 7).

 Figure 7: Trends in types of revision hip replacement by state and territory

    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                     2001
                     2002
                     2003
                     2004
                     2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                          2001
                                          2002
                                          2003
                                          2004
                                          2005
                                                     2001
                                                     2002
                                                     2003
                                                     2004
                                                     2005
                                                                 2001
                                                                 2002
                                                                 2003
                                                                 2004
                                                                 2005
                                                                            2001
                                                                            2002
                                                                            2003
                                                                            2004
                                                                            2005
                                                                                        2001
                                                                                        2002
                                                                                        2003
                                                                                        2004
                                                                                        2005
             AUST      NSW        VIC        QLD         WA          SA         TAS     ACT/NT

                       Minor Revision   Major Total Revision   Major Partial Revision



Variation and trends in fixation of hip replacement
There are significant differences in the fixation used depending on the category of hip replacement
as well as the different subcategories. There is also significant state and territory variation.

Nationally cement fixation of monoblock partial hip replacement has increased from a low of 18.1%
in 2002 up to 34.6% in 2005. Two states Western Australia and Queensland have a considerably
higher rate of cement fixation (over 80%) compared to the other states and territories.

Cement fixation of unipolar modular prostheses and bipolar prostheses has been decreasing in
recent years although both have shown a slight increase in cement fixation during the last 12
months (Figures 7 and 8).




                                                   31
Figure 8: Trends in unipolar monoblock fixation by state and territory
    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                     2001
                     2002
                     2003
                     2004
                     2005
                                2001
                                2002
                                2003
                                2004
                                2005
                                             2001
                                             2002
                                             2003
                                             2004
                                             2005
                                                       2001
                                                       2002
                                                       2003
                                                       2004
                                                       2005
                                                                 2001
                                                                 2002
                                                                 2003
                                                                 2004
                                                                 2005
                                                                             2001
                                                                             2002
                                                                             2003
                                                                             2004
                                                                             2005
                                                                                            2001
                                                                                            2002
                                                                                            2003
                                                                                            2004
                                                                                            2005
             AUST       NSW         VIC         QLD       WA           SA          TAS       ACT/NT

                                          Cemented                 Cementless




 Figure 9: Trends in prosthesis fixation - femoral components used with unipolar modular prostheses
              by state and territory

    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                    2001
                    2002
                    2003
                    2004
                    2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                           2001
                                           2002
                                           2003
                                           2004
                                           2005
                                                      2001
                                                      2002
                                                      2003
                                                      2004
                                                      2005
                                                               2001
                                                               2002
                                                               2003
                                                               2004
                                                               2005
                                                                            2001
                                                                            2002
                                                                            2003
                                                                            2004
                                                                            2005
                                                                                     2001
                                                                                     2002
                                                                                     2003
                                                                                     2004
                                                                                     2005




            AUST       NSW        VIC         QLD      WA         SA         TAS         ACT/NT


                                    Cemented                    Cementless




                                                 32
 Figure 10: Trends in prosthesis fixation - femoral components used with bipolar prostheses by
             state and territory

    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                     2001
                     2002
                     2003
                     2004
                     2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                          2001
                                          2002
                                          2003
                                          2004
                                          2005
                                                      2001
                                                      2002
                                                      2003
                                                      2004
                                                      2005
                                                               2001
                                                               2002
                                                               2003
                                                               2004
                                                               2005
                                                                         2001
                                                                         2002
                                                                         2003
                                                                         2004
                                                                         2005
                                                                                      2001
                                                                                      2002
                                                                                      2003
                                                                                      2004
                                                                                      2005
             AUST      NSW         VIC       QLD          WA       SA        TAS      ACT/NT


                                      Cemented                  Cementless


There has been increasing use of cementless fixation for primary conventional total hip replacements
nationally and across all states and territories. This has been associated with a significant decrease in
the use of cemented fixation. The proportion of primary conventional hips with hybrid fixation has
remained relatively constant. There is however considerable variation in the fixation used between
the states and territories. Queensland has the highest use of cement fixation (24.5%). While
Western Australia, Tasmania, New South Wales and Australian Capital Territory/Northern
Territory have a very low use. The use of hybrid fixation is similar across the states and territories
with the exception of Tasmania which has a significantly lower use compared to other regions of the
country.

Figure 11: Trends in fixation of conventional hip replacement by state and territory
    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                    2001
                    2002
                    2003
                    2004
                    2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                          2001
                                          2002
                                          2003
                                          2004
                                          2005
                                                      2001
                                                      2002
                                                      2003
                                                      2004
                                                      2005
                                                               2001
                                                               2002
                                                               2003
                                                               2004
                                                               2005
                                                                        2001
                                                                        2002
                                                                        2003
                                                                        2004
                                                                        2005
                                                                                   2001
                                                                                   2002
                                                                                   2003
                                                                                   2004
                                                                                   2005




             AUST      NSW        VIC        QLD         WA       SA         TAS      ACT/NT


                           Cemented                   Hybrid             Cementless


                                                 33
REGIONAL DIFFERENCES AND TRENDS IN KNEE REPLACEMENT

The three major categories of knee replacement are unicompartmental, total and revision knee
replacement. The other types of knee replacement are used in small numbers. There has been a
decrease in the use of unicompartmental knee replacement from a high of 15.1% in 2002 to 11.8%
in 2005. There has also been a small decrease in the proportion of knee revision procedures from
9.3% in 2001 to 8.2 % in 2005. The principal regional variation in knee replacement is in the
unicompartmental category. The highest proportion of unicompartmental knees are undertaken in
South Australia (18.0%). This is considerably more than is undertaken in Queensland and
Western Australia (5.8% and 6.0% respectively). The proportion of revision procedures is similar
between the states and territories with the exception of Tasmania which has a low proportion (5.8%)
and South Australia a higher proportion (10.6%) (Figure 12).

Figure 12: Trends in knee replacement by state and territory
    100%


     80%


     60%


     40%


     20%


      0%
           2001
           2002
           2003
           2004
           2005
                    2001
                    2002
                    2003
                    2004
                    2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                          2001
                                          2002
                                          2003
                                          2004
                                          2005
                                                     2001
                                                     2002
                                                     2003
                                                     2004
                                                     2005
                                                                  2001
                                                                  2002
                                                                  2003
                                                                  2004
                                                                  2005
                                                                             2001
                                                                             2002
                                                                             2003
                                                                             2004
                                                                             2005
                                                                                        2001
                                                                                        2002
                                                                                        2003
                                                                                        2004
                                                                                        2005

             AUST      NSW         VIC       QLD         WA           SA          TAS   ACT/NT

                             Unicompartmental Knee   Total Knee   Revision Knee



Almost all unicompartmental knee prostheses are inserted utilising cement fixation. This is true of
almost all states apart from Victoria where over 30% of all unicompartmental knee replacements are
undertaken with cementless fixation. Hybrid fixation is rarely used (Figure 13).




                                                34
Figure 13: Trends in fixation of unicompartment knee replacement by state and territory
    100%


     80%


     60%


     40%


     20%


      0%
           2001
           2002
           2003
           2004
           2005
                    2001
                    2002
                    2003
                    2004
                    2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                         2001
                                         2002
                                         2003
                                         2004
                                         2005
                                                    2001
                                                    2002
                                                    2003
                                                    2004
                                                    2005
                                                                2001
                                                                2002
                                                                2003
                                                                2004
                                                                2005
                                                                        2001
                                                                        2002
                                                                        2003
                                                                        2004
                                                                        2005
                                                                                     2001
                                                                                     2002
                                                                                     2003
                                                                                     2004
                                                                                     2005
             AUST      NSW        VIC      QLD             WA     SA          TAS    ACT/NT

                             Cemented          Hybrid            Cementless



The approach to fixation in primary total knees is more variable. Cemented fixation is still the most
common approach (49.9% in 2005). Of the remaining procedures hybrid fixation is used in 26.3%
and cementless fixation in 23.9% of procedures (Figure 14).

Figure 14: Trends in primary total knee fixation by state and territory
    100%



     80%



     60%



     40%



     20%



      0%
           2001
           2002
           2003
           2004
           2005
                    2001
                    2002
                    2003
                    2004
                    2005
                               2001
                               2002
                               2003
                               2004
                               2005
                                         2001
                                         2002
                                         2003
                                         2004
                                         2005
                                                     2001
                                                     2002
                                                     2003
                                                     2004
                                                     2005
                                                                2001
                                                                2002
                                                                2003
                                                                2004
                                                                2005
                                                                          2001
                                                                          2002
                                                                          2003
                                                                          2004
                                                                          2005
                                                                                     2001
                                                                                     2002
                                                                                     2003
                                                                                     2004
                                                                                     2005




             AUST      NSW         VIC      QLD            WA      SA          TAS    ACT/NT

                             Cemented             Hybrid         Cementless




                                               35
The use of patella resurfacing is another area of variability in primary total knee replacement. Most
often a patella resurfacing is not undertaken as part of the procedure (58.5%). When the patella is
used it is usually cemented (38.7%). Cementless patella resurfacing is used in only 2.8% of primary
total knee procedures. The use of patella resurfacing in primary total knee replacement varies
between states and territories. Victoria has the lowest rate at 51.4% of all primary total knees and
South Australia the highest at 84.8%. There has been a marginal decrease in the use of patella
resurfacing in the last 12 months (Figure 15).

Figure 15: Trends in patella use for total knee replacement by state and territory
    100%
     90%
     80%
     70%
     60%
     50%
     40%
     30%
     20%
     10%
      0%
           2001
           2002
           2003
           2004
           2005
                     2001
                     2002
                     2003
                     2004
                     2005
                                 2001
                                 2002
                                 2003
                                 2004
                                 2005
                                                2001
                                                2002
                                                2003
                                                2004
                                                2005
                                                         2001
                                                         2002
                                                         2003
                                                         2004
                                                         2005
                                                                     2001
                                                                     2002
                                                                     2003
                                                                     2004
                                                                     2005
                                                                                  2001
                                                                                  2002
                                                                                  2003
                                                                                  2004
                                                                                  2005
                                                                                              2001
                                                                                              2002
                                                                                              2003
                                                                                              2004
                                                                                              2005
             AUST       NSW          VIC         QLD          WA          SA            TAS   ACT/NT

                              No Patella Used   Patella Cemented   Patella Cementless



MEASURES USED TO DETERMINE OUTCOME

The most unambiguous measure of a failed joint replacement procedure is the need to undertake
revision surgery. There is no doubt that when a revision procedure is undertaken that the previous
procedure has failed.

When considering the outcome of joint replacement surgery the least accurate measure is the
proportion of revision procedures undertaken. Although it is indicative of the revision burden it is
not a true outcome measure. It is a simple calculation of the number of revisions being undertaken
relative to the other procedures performed. It does not identify the risk of revision, which is the
chance that a procedure undertaken on an individual will be revised. The proportion of revision
procedures is dependent and significantly altered by the incidence of the other procedures. If all
primary procedures were ceased for a period, the proportion of revision procedures would increase
to 100%. This does not mean that the outcome of the previously undertaken procedures has
worsened. Conversely if there is an increase in the number of primary procedures, the proportion of
revision procedures will decrease and this does not mean that the outcome has improved. This said
however the proportion of revision procedures is a rough guide to success particularly when the rate
of joint replacement surgery is changing at a relative constant rate. The proportion of revision
procedures is often the only available information for comparison of results with other countries.



                                                  36
In order to determine the risk of revision a procedure needs to be identified and then followed until
the individual has died or has been revised. When an individual has died before revision has been
undertaken the data for that individual is then censored. This means that in any analysis to
determine risk of revision the outcome for that individual is only included until the time that they
have died. After that time the outcome is determined only on those individuals still at risk. The risk
of revision is a determination of the collective chance of being revised for all individuals that have
undergone the procedure that is being assessed and who are still alive at the time it is being assessed.

The AOA NJRR assesses risk of revision in a number of ways. Each has their own advantages and
disadvantages. The simplest is the percentage of individuals who have had a procedure that has
subsequently been revised. This is referred to as the percentage revision rate. Although this is a
good general indicator of risk it is limited because it does not include the effect of time. When
comparing procedures this approach is valid as long as the procedures being compared have been
undertaken for a similar time period and in similar proportions at different time intervals during
that period. It is not valid to use this approach to compare a procedure that has only been
undertaken for a short period compared to one that has been performed for a much greater length of
time.

In order to take time into account the AOA NJRR uses two further approaches. The first is a
determination of the revision rate for every observed 100-component years. Component years are
the number of procedures that have been undertaken multiplied by the period of time since the
operation. Two procedures undertaken five years ago have a value of ten component years. This
approach allows the incorporation of the effect of time. It is an excellent screening test to establish if
there is likely to be a relevant difference. A major limitation with this approach is if the risk of
revision is not uniform over time. It does not identify if the revision rate is variable at different time
points. To determine this, the AOA NJRR uses the second additional method for determining
outcome and that is the Kaplan-Meier estimates of survivorship. This is a graphical presentation of
the number of revisions over time and includes the curve as well as the 95% confidence intervals.

Finally, in order to determine if any observed differences are statistically significant, the Registry
undertakes analytical comparisons of prostheses’ survival using log-rank tests and proportional
hazards models.




                                                37
OUTCOMES OF JOINT REPLACEMENT SURGERY IN AUSTRALIA

Global comparison
As most countries do not have accurate figures on outcome and there is variation in the way it is
determined, the simplest approach is to compare the proportion of revision procedures. As already
mentioned there are significant limitations with this approach but it does provide some useful
insight. The country with the lowest proportion of revision procedures is Sweden. This is 7% for
hip and 6.4% for knee replacement. This low rate of revision is attributed to the longstanding
impact of the hip and knee registries in that country. At the time the Australian Orthopaedic
Association National Joint Replacement Registry was established the proportion of revision
procedures in Australia was 14.2% for hips and 10.0% for knees. In 2005 this had reduced to
12.1% for hips and 8.2% for knees. Australia has a higher proportion of knee revision procedures
compared to many other countries but appears to fair well for hip replacement apart from the
comparison to Sweden (Table 17).

Table 17: International Comparison of Proportion of Revision Procedures

                                      % Revision                % Revision
                 Country
                                      Hips                      Knees

                 Australia            12.1                      8.2
                 Denmark              12.8                      7.4
                 Finland              15.5                      7.7
                 New Zealand          12.9                      7.3
                 Norway               13.0                      10.0
                 Sweden               7.0                       6.4


Risk of revision surgery is a better measure of outcome. This is a difficult figure to obtain from most
national registries, as they have not been going long enough to determine this. These figures are
available for Sweden. It is well recognised that Sweden has the lowest rate of revision. As the
principal aim of the AOA NJRR is to improve outcomes then it is reasonable to benchmark
Australia against the Swedish results.

At present the risk of revision surgery for an individual receiving joint replacement surgery in
Australia can only be estimated. It is likely that this is between 20-25% for both hip and knee. In
Sweden this figure is 10% for both hip and knee. This indicates that there is significant potential to
reduce the rate of revision surgery within Australia. In order to achieve this it is necessary to
understand why this difference exists and identifying areas that can be improved.

There are many factors known to influence the outcome of joint replacement surgery. Age, gender,
diagnosis of patients, the type prosthesis and the surgical techniques used are just some of these.
Superimposed on this, is the rapid rate of change in medical technology. There is continual
development and use of new types of prostheses and surgical techniques and outcomes for many of
these remains to be established.




                                                   38
Hip replacement
The overall revision rate for hip replacement in Australia is unknown and it will not be established
until the Registry has recorded all primary procedures that are being revised. Currently most
revisions that are being undertaken have had the primary performed prior to the establishment of
the AOA NJRR. The Registry however has excellent early to mid-term data on primary procedures
undertaken since the Registry commenced data collection. Outcomes vary depending on the type of
procedure, patient diagnosis category of prostheses, method of fixation as well as individual variation
of different prostheses types. It is not relevant to compare the outcomes of primary partial hip
replacement and primary total hip replacement because, as mentioned previously, they are
undertaken for different reasons. Partial hips are mainly used in the management of fractured neck
of femur and primary totals for osteoarthritis. Only a few examples of important differences in
outcomes will be provided. A full analysis of outcomes is provided in the AOA NJRR annual
reports, available on their website.

Partial hip replacement
The three broad categories of partial hip replacement demonstrate significant differences in outcome.
Unipolar monoblock prostheses, the most common category of prostheses used for the treatment of
displaced sub-capital fracture, has the highest rate of revision (Table 18). Analysis of outcomes of
different types of unipolar monoblock prostheses indicate that the higher than anticipated rate of
revision for this category is due to only one of the prostheses types. That is the Austin Moore
prosthesis (Figure 16). The revision rate for the Austin Moore at three years is 6.1%. This
compares to the Thompson revision rate of 2.3% at the same time. Of further concern is that the
use of the Austin Moore is associated with a significant increase in mortality when compared to the
Thompson prosthesis. Further analysis has indicated that this difference is not related to the design
of the prostheses but to the method of fixation. An Austin Moore is inserted cementless and a
Thompson is cemented.

Table 18: Revision rates of primary partial hip replacement
                                                                               Revisions per
                                                                 Observed
                            Number     Total                                   100 observed
  Partial hip replacement                            % Revised   ‘component’                   Exact 95%CI
                            Revised    Number                                  ‘component’
                                                                 years
                                                                               years
  Unipolar Monoblock        333        10302         3.2         21100         1.6             (1.41, 1.76)
  Unipolar Modular          46         2286          2.0         4055          1.1             (0.83, 1.51)
  Bipolar                   104        4742          2.2         8731          1.2             (0.97, 1.44)
  Total                     483        17330         2.8         33886         1.4             (1.30, 1.56)




                                                39
Figure 16: Cumulative percentage of revision of Austin Moore and Thompson hip prostheses

                              0.09
                                       Log-rank test for equality over strata p-value .000000;
                                       Hazard Ratio (adjusted for age and sex; Austin-Moore Type Cementless
                                       v Thompson Type Cemented) = 2.204; 95% CI (1.621,2.997) p-value 0.0000
                              0.08


                              0.07


                              0.06
 Cumulative Percent Revised




                              0.05


                              0.04


                              0.03


                              0.02



                              0.01

                                                                                                              Austin-Moore Type Cementless
                              0.00                                                                            Thompson Type Cemented
                                 0.0        0.5         1.0    1.5        2.0        2.5         3.0    3.5         4.0      4.5       5.0
                                                                        years since primary procedure


Primary total hip replacement
As with partial hips, there are different outcomes for the three types of primary total hip
replacement. Conventional hips are revised less often than the other types of primary total hip
replacement (Table 19).

Table 19: Revision rates by type of primary total hip replacement
                                                                                                                           Revisions per
                                                                                                        Observed
                                                              Number         Total                                         100 observed
                              Type of hip replacement                                      % Revised    ‘component’                          Exact 95%CI
                                                              Revised        Number                                        ‘component’
                                                                                                        years
                                                                                                                           years
                          Conventional Total Hip              1275           65992         1.9          128708             1.0               (0.94, 1.05)
                          Resurfacing Hip                     118            5379          2.2          9118               1.3               (1.07, 1.55)
                          Thrust Plates                       2              112           1.8          263                0.8               (0.09, 2.74)
                          Total                               1878           88813         2.1          171975             1.1               (1.04, 1.14)

Osteoarthritis is the usual reason for a total hip replacement but there are others. It has been well
established that the outcome for diagnoses other than osteoarthritis are less satisfactory. In order to
compare the performance of different types of primary total hip it is usual that there is
standardisation of the diagnosis by undertaking the analysis only for osteoarthritis. This is
particularly important with this comparison as resurfacing hips are used more frequently for
osteoarthritis than conventional hips. It is also usual to exclude revisions that are the result of
infection as the occurrence of infection is usually independent of the type of prosthesis used. The
difference between resurfacing hip replacement and conventional hip replacement is more evident
when this analysis is done (Table 20 and Figure 17).

                                                                                      40
Table 20: Revision rates of conventional total hip and resurfacing hip for osteoarthritis requiring
       revision excluding revisions for infection
                                                                                                                                  Revisions per
                               Type of procedure for                                                        Observed
                                                              Number          Total         %                                     100 observed
                               Osteoarthritis excluding                                                     ‘component’                           Exact 95%CI
                                                              Revised         Number        Revised                               ‘component’
                               infection                                                                    years
                                                                                                                                  years
                              Resurfacing                     98              4974          2.0             8427                  1.2             (0.94, 1.42)
                              Conventional Total              900             57939         1.6             112848                0.8             (0.75, 0.85)
                              Total                           998             62913         1.6             121275                0.8             (0.77, 0.88)



Figure 17: Cumulative percentage of revision of conventional total hip and resurfacing hip for
            osteoarthritis excluding revisions for infection
                             0.05
                                      Log-rank test for equality over strata p-value .000559;
                                      Hazard Ratio (adjusted for age and sex; Resurfacing for OA
                                      v Conventional Total for OA) = 1.427; 95% CI (1.184,1.720) p-value 0.0002



                             0.04
Cumulative Percent Revised




                             0.03




                             0.02




                             0.01




                                                                                                            Resurfacing for OA
                             0.00                                                                           Conventional Total for OA
                                0.0        0.5       1.0       1.5      2.0       2.5        3.0      3.5         4.0       4.5         5.0
                                                                      years since primary procedure



Resurfacing hips differ from conventional total hips with respect to the femoral component. In a
conventional total hip the head and part of the neck of the femur is resected and replaced with a
prosthesis that has a stem inserted down the shaft of the femoral canal. A resurfacing hip retains the
femoral head and neck but involves reshaping the head and then covering it with the resurfacing
femoral component. The major reason for the higher early revision is subsequent fracture across the
retained femoral neck. This is a mode of failure that can not occur in a conventional total hip
replacement because a major portion of the neck is replaced by the femoral component. This risk is
reduced if there is no evidence of osteoporosis. Osteoporosis is associated with bone weakening and
is associated with a higher risk of fracture post resurfacing hip replacement. For this reason many
manufacturers of resurfacing hips recommend that this procedure should be undertaken in males
younger than 65. Further analysis of Registry data has demonstrated that resurfacing procedures on
individuals who are outside these recommendations have a significantly higher risk of revision
(Table 21 and Figure 18)




                                                                                           41
Table 21: Resurfacing hip systems requiring revision by age and sex
                                                                                                                        Revisions per
                                                                                                       Observed
                                                           Number        Total                                          100 observed
                             Sex           Age                                            % Revised    ‘component’                        Exact 95%CI
                                                           Revised       Number                                         ‘component’
                                                                                                       years
                                                                                                                        years
                         Male             <50              21            1165             1.8          2013             1.04              (0.65, 1.59)
                         Male             50-54            7             704              1.0          1169             0.60              (0.24, 1.23)
                         Male             55-59            7             779              0.9          1305             0.54              (0.22, 1.11)
                         Male             60-64            7             648              1.1          1053             0.66              (0.27, 1.37)
                         Male             >=65             21            496              4.2          829              2.53              (1.57, 3.87)
                         Female           <50              16            551              2.9          945              1.69              (0.97, 2.75)
                         Female           50-54            13            369              3.5          650              2.00              (1.07, 3.42)
                         Female           55-59            16            402              4.0          693              2.31              (1.32, 3.75)
                         Female           60-64            8             199              4.0          343              2.33              (1.01, 4.59)
                         Female           >=65             2             66               3.0          118              1.70              (0.21, 6.13)
                         Total            .                118           5379             2.2          9118             1.29              (1.07, 1.55)



Figure 18: Cumulative percentage of revision of resurfacing hip by sex
                             0.09
                                      Log-rank test for equality over strata p-value .000422;
                                      Hazard Ratio (adjusted for age and sex; Female
                                      v Male) = 1.788; 95% CI (1.322,2.420) p-value 0.0002
                             0.08


                             0.07


                             0.06
Cumulative Percent Revised




                             0.05


                             0.04


                             0.03


                             0.02


                             0.01

                                                                                                             Female
                             0.00                                                                            Male
                                0.0        0.5       1.0         1.5      2.0       2.5         3.0    3.5        4.0     4.5       5.0
                                                                       years since primary procedure


There is significant variation in outcome for conventional total hip depending on the method of
fixation used. Cemented conventional total hip is the least revised and cementless total hip the most
revised (Table 22, Table 23 and Figure 19). These differences are greater in older patients (Figure
20).




                                                                                      42
Table 22: Primary conventional total procedures for all diagnoses requiring revision by cement status
                                                                                    Revisions per
                                                                    Observed
                            Number      Total                                       100 observed
  Cement used                                         % Revised     ‘component’                        Exact 95%CI
                            Revised     Number                                      ‘component’
                                                                    years
                                                                                    years
  Cemented Total            172         10566         1.6           24002           0.7                (0.61, 0.83)
  Cementless Total          660         32084         2.1           57753           1.1                (1.06, 1.23)
  Hybrid                    443         23342         1.9           46953           0.9                (0.86, 1.04)
  Total                     1275        65992         1.9           128708          1.0                (0.94, 1.05)



Table 23: Primary conventional total procedures for osteoarthritis requiring revision by cement
       status excluding infection
                                                                                  Revisions per
                                                                  Observed
                         Number       Total      %                                100 observed
     Cement Used                                                  ‘component’                       Exact 95%CI
                         Revised      Number     Revised                          ‘component’
                                                                  years
                                                                                  years
    Cemented             134          10528      1.3              23968           0.6               (0.47, 0.66)
    Cementless           587          32011      1.8              57700           1.0               (0.94, 1.10)
    Hybrid               359          23258      1.5              46884           0.8               (0.69, 0.85)
    Total                1080         65797      1.6              128551          0.8               (0.79, 0.89)



Figure 19: Cumulative percentage of revision for total hip replacement by cement status excluding
            infection




                                                 43
Figure 20: Cumulative percentage of revision for total hip replacement by cement status excluding
            infection for individuals 75 yrs and older
                                       0.04
                                                Log-rank test for equality over strata p-value < 0.0001;
                                                Hazard Ratio (adjusted for age and sex;
                                                Hybrid v Cemented = 1.125; 95% CI (0.852,1.486) p-value = 0.4068
                                                Cementless v Cemented = 1.879; 95% CI (1.425,2.478) p-value < 0.0001
                                                Cementless v Hybrid = 1.670; 95% CI (1.347,2.071) p-value < 0.0001



                                       0.03
          Cumulative Percent Revised




                                       0.02




                                       0.01




                                                                                                                                   Cementless
                                                                                                                                   Hybrid
                                       0.00                                                                                        Cemented
                                          0.0        0.5       1.0       1.5       2.0        2.5       3.0       3.5       4.0          4.5      5.0
                                                                                 years since primary procedure



Not only is there a variation in the outcome between the different categories of hip replacement and
the method of fixation used, there is also considerable variation in the outcome of the different
prostheses types within the same category. Not all prostheses perform equally as well as others.
Currently in Australia there are at least 130 different types of hip replacement available on the
market with new prostheses being added continually to that list. A simple example of this is the
comparison of revision rates for the different types of hip resurfacing prostheses. It is apparent that
not all prostheses within this class perform equally well (Table 24). This prostheses specific
variation is evident in each of the different categories of hip replacement.

Table 24: Resurfacing hip systems requiring revision
                                                                                                                                  Revisions per
                                                                                                              Observed
Resurfacing                                              Number                Total                                              100 observed
                                         Resurfacing Cup                                    % Revised         ‘component’                               Exact 95%CI
Head                                                     Revised               Number                                             ‘component’
                                                                                                              years
                                                                                                                                  years
ASR                                     ASR                   18               503          3.6               446                 4.0                   (2.39, 6.38)
Adept                                   Adept                 0                19           0.0               3                   0.0                   (0.00, 129.6)
BHR                                     BHR                   126              5799         2.2               13487               0.9                   (0.78, 1.11)
Bionik                                  Bionik                0                12           0.0               3                   0.0                   (0.00, 143.3)
Conserve                                -                     2                2            100.0             3                   74.1                  (8.97, 267.6)
Conserve                                Conserve Plus         0                3            0.0               2                   0.0                   (0.00, 164.9)
Conserve Plus                           Conserve Plus         2                48           4.2               82                  2.4                   (0.29, 8.79)
Cormet 2000                             Cormet 2000           14               333          4.2               649                 2.2                   (1.18, 3.62)
Durom                                   Durom                 12               423          2.8               451                 2.7                   (1.38, 4.65)
Icon                                    Icon                  1                22           4.5               13                  7.5                   (0.19, 42.05)
Recap                                   Recap                 2                41           4.9               41                  4.9                   (0.59, 17.56)
Total                                                         177              7205         2.5               15179               1.2                   (1.00, 1.35)

                                                                                         44
  Further complicating the issue of varying performance of individual prostheses within a category is
  the practice for surgeons to mix and match acetabular and femoral components from different
  prostheses. Currently the Registry has recorded over 1,000 combinations of femoral and acetabular
  prostheses. Many of the combinations demonstrate different outcomes, some of which vary
  significantly. The Exeter stem is the most common femoral stem used in Australia and this has been
  combined with over 50 different acetabular components. The outcomes of these different
  combinations of acetabular component and Exeter stem vary up to a three fold when comparing
  revisions per 100-observed component years (Table 25).

  Table 25: Variable outcome for the same femoral stem used with different acetabular components
                                                                          Revisions per
                                                            Observed
Femoral      Acetabular      Number    Total                              100 observed
                                                % Revised   ‘component’                   Exact 95%CI
Component    Component       Revised   Number                             ‘component’
                                                            years
                                                                          years
Exeter V40   ABGII           13        933      1.4         2507          0.5             (0.28, 0.89)
Exeter V40   Duraloc         6         218      2.8         574           1.0             (0.38, 2.27)
Exeter V40   Hemispherical   1         102      1.0         98            1.0             (0.03, 5.69)
Exeter V40   Mallory-Head    7         557      1.3         1274          0.5             (0.22, 1.13)
Exeter V40   Reflection      5         177      2.8         467           1.1             (0.35, 2.50)
Exeter V40   Secur-Fit       7         150      4.7         387           1.8             (0.73, 3.73)
Exeter V40   Trident         114       7563     1.5         12022         0.9             (0.78, 1.14)
Exeter V40   Trilogy         6         237      2.5         460           1.3             (0.48, 2.84)
Exeter V40   Vitalock        26        1752     1.5         4353          0.6             (0.39, 0.88)

  Some types of prostheses perform well outside the range of other prostheses. The Registry identifies
  these using a statistically approach which highlights those prostheses that have double the
  anticipated rate of revision compared to all the other prostheses in that category of replacement.
  Each year the annual Registry report identifies those prostheses. An example of a hip prosthesis
  highlighted in the 2005 report was the Margron hip (Figure 21). This particular prosthesis has a
  risk of revision three times greater than other cementless hip replacements. At four years, the
  revision rate is 9.23% compared to 3.1% for all other cementless hips combined.

  Figure 21: Cumulative percentage of revision of cementless Margron hip prosthesis v other
              cementless femoral components




                                                45
Knee replacement
As with hip replacement there are differences in outcome related to the category and fixation
methods associated with knee replacement. Unispacer knee replacement is a new class of knee
prostheses that was available for only a short period in the Australian market. There were a number
of different types of unispacer prostheses. In the 2005 report the Registry identified that there was
almost a 50% revision rate within 12 months (Table 26). This category of prostheses is no longer
used or available in Australia.

Table 26: Unispacer procedures requiring revision
                                                                                                                   Revisions per
                                                                                          Observed
                                             Number      Total                                                     100 observed
                               Unispacer                               % Revised          ‘component’                              Exact 95%CI
                                             Revised     Number                                                    ‘component’
                                                                                          years
                                                                                                                   years
                              InterCushion   7           9             77.8               9                        77.7            (31.24, 160.1)
                              Unispacer      17          30            56.7               38                       44.5            (25.93, 71.26)
                              Total          24          39            61.5               47                       50.8            (32.57, 75.64)

There is considerable variation in the outcome of unicompartmental knee replacement depending
on the age of the patient receiving this type of prostheses The revision rate is high for younger
patients and progressively decreases with the age of the patient (Figure 22).

Figure 22: Comparative cumulative percentage of revision of unicompartmental procedures
            requiring revision by age (primary diagnosis OA excluding revisions for infection)

                             0.20




                             0.15
Cumulative Percent Revised




                             0.10




                             0.05




                                                                                                           < 55
                                                                                                           55-64
                                                                                                           65-74
                                                                                                           >=75
                             0.00
                                0.0    0.5   1.0   1.5        2.0     2.5        3.0           3.5   4.0           4.5     5.0
                                                          years since primary procedure




                                                                                 46
There is also considerable variation in the revision rates for the different prostheses used in this
category of primary knee replacement (Table 27).

Table 27: Unicompartmental primary knee procedures requiring revision
                                                                       Revisions per
                                                        Observed
                          Number    Total     %                        100 observed
     Model                                              ‘component’                    Exact 95%CI
                          Revised   Number    Revised                  ‘component’
                                                        years
                                                                       years
    AMC                   19        267       7.1       337            5.6             (3.40, 8.81)
    Active Knee           2         239       0.8       99             2.0             (0.24, 7.27)
    Advance               5         30        16.7      46             10.8            (3.51, 25.23)
    Allegretto Uni        85        1411      6.0       3907           2.2             (1.74, 2.69)
    BalanSys              5         152       3.3       119            4.2             (1.36, 9.81)
    Eius                  3         94        3.2       134            2.2             (0.46, 6.52)
    Endo-Model Sled       12        524       2.3       702            1.7             (0.88, 2.99)
    GRU                   26        938       2.8       1462           1.8             (1.16, 2.60)
    Genesis               50        1071      4.7       1984           2.5             (1.87, 3.32)
    HLS Uni Evolution     0         22        0.0       22             0.0             (0.00, 16.74)
    LCS                   4         26        15.4      118            3.4             (0.93, 8.69)
    M/G                   60        1626      3.7       3893           1.5             (1.18, 1.98)
    Natural Knee          16        143       11.2      414            3.9             (2.21, 6.27)
    Oxford 3              359       6569      5.5       16811          2.1             (1.92, 2.37)
    PFC Sigma             11        137       8.0       600            1.8             (0.92, 3.28)
    Preservation Fixed    75        1436      5.2       2852           2.6             (2.07, 3.30)
    Preservation Mobile   49        369       13.3      894            5.5             (4.06, 7.25)
    Repicci               67        1979      3.4       5161           1.3             (1.01, 1.65)
    Unix                  44        1211      3.6       2822           1.6             (1.13, 2.09)
    ZUK                   0         70        0.0       27             0.0             (0.00, 13.75)
    Total                 892       18314     4.9       42405          2.1             (1.97, 2.25)


Primary total knee replacement also shows variability depending on fixation, type of knee
replacement used and individual prostheses (Figure 23).

Figure 23: Comparative cumulative percentage of revision of cemented, hybrid and cementless
            fixation of primary total knee replacement.




                                               47
There are a variety of different technologies available in this category of knee replacement. These
relate to the movement of the components and the degree of knee stability that the components are
designed to provide. The outcomes of these different technologies differ with the more complicated
technologies tending to have a higher revision rate (Figure 24).

Figure 24: Cumulative percentage of revision of fixed and mobile
                             0.05
                                      Log-rank test for equality over strata p-value < 0.0001;
                                      Hazard Ratio (adjusted for age and sex; Mobile
                                      v Fixed) = 1.201; 95% CI (1.097,1.315) p-value = 0.0001



                             0.04
Cumulative Percent Revised




                             0.03




                             0.02




                             0.01




                                                                                                             Mobile
                             0.00                                                                            Fixed
                                0.0        0.5       1.0       1.5       2.0       2.5        3.0      3.5         4.0    4.5   5.0
                                                                       years since primary procedure



The Registry has also identified different outcomes related to technique with primary total knee
replacement. In this country it is common for knee replacement surgery to be undertaken without
replacing the patella. This is associated with an increased rate of revision compared to when a patellar
resurfacing prostheses is used (Figure 25).

Figure 25: Revision rates for primary total knee replacements requiring revision by patella use
                             0.05
                                      Log-rank test for equality over strata p-value < 0.0001;
                                      Hazard Ratio (adjusted for age and sex; No Patella
                                      v Patella) = 1.432; 95% CI (1.313,1.563) p-value < 0.0001



                             0.04
Cumulative Percent Revised




                             0.03




                             0.02




                             0.01




                                                                                                             No Patella
                             0.00                                                                            Patella
                                0.0        0.5       1.0       1.5       2.0       2.5        3.0      3.5         4.0    4.5   5.0
                                                                      years since primary procedure




                                                                                                 48
As with hip replacement there is also considerable variation in the outcome of different individual
types of knee replacement. Both types of unispacer prostheses had very high revision rates and are
no longer available on the Australian market. Individual patellar/trochlear and unicompartmental
prostheses show considerable variation and in the unicompartmental prostheses the majority of
prostheses have a significantly higher failure rate than the best three performing prostheses. Of the
16 other prostheses within this category, nine have a significantly higher revision rate (Table 28).

Table 28: Comparison of unicompartmental primary knee procedures with three comparators M/G,
       Unix and Repicci
                                                                               95% Upper
    Model                    Pr > Chi-   Hazard       95% Lower Confidence     Confidence Limit for
                             Square      Ratio        Limit for Hazard Ratio   Hazard Ratio
    AMC                      <.0001      3.618        2.237                    5.852
    Active Knee              0.4101      1.813        0.440                    7.470
    Advance                  <.0001      7.914        3.230                    19.387
    Allegretto               0.0017      1.513        1.168                    1.961
    Balansys                 0.0202      2.915        1.181                    7.192
    Eius                     0.3960      1.543        0.567                    4.200
    Endo-Model Sled          0.4829      1.236        0.684                    2.231
    GRU                      0.3557      1.218        0.802                    1.850
    Genesis                  0.0005      1.749        1.274                    2.400
    HLS Uni                  0.9716      0.000        0.000
    LCS                      0.2129      2.063        0.660                    6.445
    Natural knee             0.0003      2.612        1.560                    4.371
    Oxford 3                 <.0001      1.491        1.244                    1.787
    PFC Sigma                0.7618      1.101        0.591                    2.053
    Preservation Fixed       <.0001      1.795        1.372                    2.348
    Preservation Mobile      <.0001      3.337        2.416                    4.608

Individual primary total knee prostheses also show variable outcomes and as with all categories there
are a number of prostheses that have revision rates well above an acceptable level. The cementless
macrotextured oxinium femoral component is one example. This particular component is no longer
available on the market. The Registry has information on almost 500 of these prostheses with
almost half of these being revised within three years. The Oxinium Cementless Genesis II knee is
an example.




                                                 49
Figure 26: Cumulative percentage of revision of Oxinium Cementless Genesis II total knee
            prosthesis
                             0.80
                                      Hazard Ratio (adjusted for age and sex); Genesis II/MBK v
                             0.75     Genesis II/Genesis II) = 1.428; 95% CI (0.998,2.043) p-value = 0.0512
                             0.70     Hazard Ratio (adjusted for age and sex); Cementless Genesis II Oxinium/MBK v
                                      Cementless Genesis II Oxinium/Genesis II) = 1.563; 95% CI (1.013,2.414) p-value = 0.0438
                             0.65

                             0.60

                             0.55
Cumulative Percent Revised




                             0.50

                             0.45

                             0.40

                             0.35

                             0.30

                             0.25

                             0.20                                                                    GenesisII/GenesisII
                                                                                                     GenesisII/MBK
                             0.15                                                                    Cementless GenesisII Oxinium/GenesisII
                                                                                                     Cementless GenesisII Oxinium/MBK
                                                                                                     Cemented GenesisII Oxinium/GenesisII
                             0.10

                             0.05

                             0.00
                                0.0        0.5        1.0       1.5       2.0        2.5       3.0         3.5       4.0       4.5      5.0
                                                                        years since primary procedure



Recently approved prostheses
The Registry has determined the outcomes of prostheses approved in the last four years relative to
the best performing established prostheses in a number of prostheses categories. These included
unicompartmental knee replacement, primary cemented and cementless total hip and primary
cemented and cementless total knee. It has done this by comparing the performance of prostheses
that have a cumulative revision rate of three years or less and compared those to the three least
revised prostheses with a cumulative revision rate of four years or more and where there has been
more than 1000 procedures using each of these prostheses. The Registry undertook this analysis to
determine if any of the more recently approved prostheses had a better outcome than the well
performing established prostheses.

It became apparent during this analysis that new prostheses were almost universally at the top end of
the market with respect to cost in particular almost all were cementless components. It appears that
sponsors and manufacturers are not developing prostheses for cement fixation despite evidence from
the Swedish registries and now more recently the Australian registry that they have better outcomes.
Just over 100 separate components where examined. None of these performed better than the
comparators. About two thirds where not statistically different but many of these were trending to a
higher rate of revision. The remaining third had a statistically higher rate of revision (Table 29).




                                                                                                50
Table 29: Comparison of new and established prostheses

                                                              Compared to top 3 with CRR of 4
                                 Number of Prostheses with
              Prosthesis Type                                 or more years and over 1000
                                 CRR 3 years or less
                                                              procedures

                                                              Better     Same        Worse
             Uni Knee            14                           0          2           12
             Cemented TKR        4                            0          0           4
             Cementless TKR      12                           0          6           6
             Cemented THR        2                            0          0           2
             Cementless THR      71                           0          63          8
             Total               103                          0          71          32

IMPLICATIONS OF IDENTIFIED VARIATION IN OUTCOMES

It is now apparent that the Australian Orthopaedic Association National Joint Replacement Registry
is able to provide quality outcomes information on joint replacement surgery. This information is
accurate and enables the identification of areas where outcomes can be enhanced. This analysis is
assisted by comparison to Swedish data which can be used to determine differences in the practice of
joint replacement surgery between the two countries. The aim of the Australian Orthopaedic
Association is to utilise this data to identify approaches and strategies that ensures Australia achieves
the best outcomes worldwide. The Registry data and comparison to Swedish data identifies areas
that can be addressed to ensure enhanced outcomes in Australia.

Comparison to patient and prostheses selection in Sweden
The difference in hip replacement outcomes between Australia and Sweden is prostheses based.
There is little variation in the patient population undergoing hip replacement apart from the higher
incidence of use in Sweden. Gender and age characteristics as well as the diagnosis profile are
identical. It is also evident that there is no difference in comparative outcomes when the same
prosthesis is used. The implication of this is that the variation in outcome is not related to surgeon
experience, skill or quality of healthcare delivery. It is clear that the major difference relates to the
use and fixation of prostheses. In Sweden the vast majority of hip replacements are undertaken
using a limited range of prostheses with four femoral stems used in 85.7% of all primary procedures.
Each of these has well defined long-term outcomes. The method of fixation used is almost entirely
90% cement fixation, which has better outcomes than cementless fixation. Cementless fixation is
the least common method of fixation (3.4%) and when used it is confined largely to patients under
65 years of age with only 0.8% of patients over 60 years receiving cementless hip prostheses. The
difference in outcome between cementless and cemented fixation is most evident in the older
population. There is also very little use of resurfacing procedures and new prosthetic designs have
very little penetration into the Swedish market both of which are associated with increased revision
rates in the Australian Joint Registry data.

A similar situation is evident with knee replacement. There is less use of categories of knee
prostheses that are known to have a higher revision rate. In particular unicompartmental knee
replacement is used less often (9.0% Sweden 2004, 16% Australia 2004-2005) and there is almost
no use patellar/trochlear replacements. Total knee replacement is largely confined to the use of five
prostheses (94.4%) with established long-term outcomes. Almost all primary total knees in Sweden
are totally cemented (98.5%). As with the hip market there is very little penetration of new
prostheses some of which are known to have higher revision rates.

                                                 51
Why are there differences?
The reason for these differences relates to the influence of the long established hip and knee
registries in Sweden. The identification of the best performing prostheses and techniques has
known to significantly influence prostheses selection and the practice of joint surgery in that country.
It has been associated with a considerable reduction in patients requiring revision surgery and
significant cost savings. This effect was used in part to justify the establishment of the Australian
Joint Registry. The Swedish Government has recognised the importance of these registries with the
provision of permanent funding and the incorporation of these and other registries into a National
Competency Centre.

An important observation is that the Swedish experience has not influenced the practice of joint
replacement worldwide. This is certainly true in Australia with significant differences in prostheses
selection and fixation. It is not clear why this is the case. Possibilities include lack of awareness or
lack of ownership of the data. It can also be argued that the practice of joint surgery is now so
different that the relevance of the Swedish data is reduced. It is widely acknowledged however that
the Swedish registries have had a very beneficial impact within Sweden and there is a need to
replicate that experience in other countries. Currently Spain, Italy, Greece, Israel and the USA are
implementing national or large regional registries.




                                                52
IMPACT OF ENHANCING OUTCOMES

Revision procedures are associated with less satisfactory clinical outcomes, increased risk of revision
and an increased complication rate compared to primary procedures. In 2005, there were 6,119
revision procedures (3,461 hip revisions and 2,658 knee revisions). As with primary hip
replacement the absolute number of revisions is increasing by 5-10% each year. Enhancing
outcomes to comparable rates of revision in Sweden will eventually halve the number of revisions
undertaken in any one year. This not only has a significant impact on the welfare of individuals
receiving joint replacement but it is also associated with significant cost savings to the Australian
community.

At the current rate of revision, a proportional reduction of 1% in revision procedures (currently
12.1% for hips and 10% for knees) decreases the number of revisions by approximately 650 per
annum. Revision procedures are expensive and various estimates place the cost for most revisions at
between $25,000-50,000 per procedure. Reducing the number of revisions by 650 is therefore
associated with an estimated reduction in expenditure of between $16-32 million per annum.
Reducing the rate of revision to that currently experienced in Sweden has the potential to reduce
expenditure by $81-162 million per annum based on the current number of revisions and the
current estimated cost of each procedure.

SUGGESTED APPROACHES TO ENHANCING OUTCOMES

The Australian Joint Registry is now providing quality data that is identifying potential areas where
outcomes can be improved. Obtaining that improvement in the quickest time possible is dependent
on developing and implementing appropriate strategies to achieve this.

Need to maintain the Registry
Data from the Registry is able to provide comparative outcomes for different prostheses and the
surgical techniques used to implant those prostheses. It can identify the most effective prostheses to
use in particular clinical situations and it can do this for both primary and revision procedures. It is
also able to identify those prostheses with a less than satisfactory outcome. The ongoing provision of
this information to clinicians, governments, government bodies, hospitals, sponsors, manufacturers,
health insurance companies and the community is critical to achieving ongoing improvement. It is a
cost effective approach to maintaining and enhancing outcomes of joint replacement surgery. The
current budget for the Registry is approximately $1 million per annum. This is equivalent to just
over $15 per procedure or 0.1% of current annual expenditure on joint replacement. Since
implementation of the Registry the proportional revision rate has declined by 2.0%. Based on
figures already used this is equivalent to reducing the number of revisions by approximately 1,200
per year and an associated cost reduction of $30-60 million per annum.

Much of the decrease in the proportion of revision procedures can be attributed to the Registry.
Whenever the Registry has identified individual prosthesis or classes of prostheses associated with
higher than anticipated rates of revision, there is an immediate reduction in their use. Despite this
there is still considerable potential for further improvement as a number of areas of concern become
increasingly evident. In order to hasten the rate of improvement further strategies that utilise
Registry data should be considered for implementation.




                                               53
Profession developed guidelines for patient selection and prosthesis use
Registries provide data that can be utilised in various ways. One potential use is the development of
clinician guidelines that highlight recommendations as to the suitability and approach to joint
replacement in specific clinical situations. These guidelines could define the most appropriate class
of prosthesis, the best prostheses within that class, the method of fixation and the surgical technique
for prosthesis implantation. For guidelines to have the most impact they need to be developed and
endorsed by the professional body which in the Australian setting is the Australian Orthopaedic
Association.

There have been few attempts to develop guidelines although the best recognised guidelines are
those that were developed for primary hip replacement in 2000 by the National Institute of Clinical
Excellence (NICE) in the UK. These guidelines were restricted to the selection of prostheses.

NICE is an independent UK organisation responsible for providing guidance on the
implementation of health technologies including implantable medical devices. The NICE guidance
on the selection of prosthesis for primary total hip replacement states:

                Using the most recent available evidence of clinical effectiveness, the best
                prostheses (using long term viability as the determinant) demonstrate a
                revision rate (the rate at which they need to be replaced) of 10% or less at
                ten years. This should be regarded as the current ‘benchmark’ in the
                selection of prostheses for primary total hip replacement.

                The evidence used in support of any prosthesis, to establish whether or
                not it achieves this benchmark, should relate to data on ten or more
                years follow up from a number of centres, obtained via adequately sized,
                well conducted observational studies (preferably with consecutive patients
                from non-selected populations) or randomised controlled trials. Such
                evidence should have been published or be available for peer review.

The Institute also considers it reasonable to recommend consideration of prostheses with a
minimum of three years revision rate experience (collected as described in two) if their performance
is consistent with the benchmark of a 10% revision rate at ten years. Prostheses that achieve this
‘entry benchmark’ would then need to be subject to annual review (up to ten years) to ensure that
the revision rate remains consistent with the ten year benchmark.

Prostheses (cemented, cementless and hybrid) that have not been shown to achieve either of these
benchmarks, should be the subject of comparative clinical evaluation before they can be
recommended for routine use in the National Health Service (NHS).

There is currently more evidence of the long term viability of cemented prostheses which, in many
cases, occupy the lower end of the range of prostheses cost, than there is for cementless and hybrid
prosthesis.

It can be seen that these recommendations are limited and in part this reflects the problems
encounted by NICE in developing the guidelines. At the time it was suggested that specific
recommendations on the selection of hip prostheses for primary total hip replacement were difficult
to construct as the evidence base was generally poor and difficult to interpret. This was because few
studies undertaken were of best design to assess clinical effectiveness and few included long term (i.e.
more than ten years) follow-up data. There were often incremental changes in specific prostheses
                                               54
design over time and the interaction between surgeon, hospital team and device performance had
not been systematically assessed. They recommended that more evidence on the performance of hip
prostheses over longer periods of follow-up was required.

The NICE guidelines where originally planned to be reviewed in 2003. It was not until 2005
however that NICE released a communication indicating that the review was being deferred until
2008. The reason given was that it was too early for the recently established National Joint Registry
of England and Wales to produce useful evidence that could be used to contribute to the review
process. This reflects a change in thinking that acknowledges the importance and relevance of their
national registry in developing guidelines.

Many of the problems originally identified by NICE are overcome by utilising registry information.
Registries provide comparative data not just for procedures undertaken at one or a number of
specialised centres but for all procedures undertaken within the community. The analysis involves
large numbers of procedures and is able to distinguish effects on outcomes that are systemically or
prosthesis based.

The Australian National Registry has now been implemented for a sufficient time to have quality
outcomes of three or more years. As is implicit in the NICE guidelines, evaluation of outcomes at
three years is predictive of long term outcome. Current and future data can now be used to develop
Australian specific guidelines. It is the intention of the Australian Orthopaedic Association to
develop those guidelines and to base them on the Australian Joint Registry data.

Need to consider change to regulation of hip and knee prostheses
Recent evidence from the National Joint Registry indicates that at this time there has been no benefit
gained by the use of prostheses introduced in recent years to the Australian market. None of the
prostheses approved in the last four years has a better outcome than the best performed previously
available prostheses. Additionally it is also evident that some new prostheses have a significantly
higher risk of revision.

It can be reasonably argued therefore that the overall impact of introducing new prostheses in recent
years has been to increase the revision rate. The use of new prostheses has not only been associated
with increased expenditure due to additional revisions but also because these prostheses are in
general more expensive than established prostheses. It is a common argument by sponsors that a
premium should be paid for innovation and development associated with new prostheses.

The failure of recently approved prostheses to provide better clinical outcomes combined with the
less than satisfactory performance of some of these prostheses raises the question as too whether
there is a need to consider a change in current regulation of these devices.

Current regulation of medical devices in Australia
The Therapeutic Goods Administration (TGA) is the national regulator of therapeutic products and
the gateway through which therapeutic products enter the Australian health system. Regulation of
medical devices by the TGA is undertaken in a similar manner to other major regulatory bodies
internationally.




                                              55
Therapeutic Goods Administration
The medical devices program of the TGA is responsible for the regulation of therapeutic and
medical devices. The program has two key elements — pre-market assessment, ensuring the safety,
quality and performance of devices placed on the market; and post-market, monitoring of the
ongoing performance of devices in the market.

Assisting the TGA in its regulatory activities is the Medical Devices Evaluation Committee (MDEC).
This expert committee provides medical and scientific advice to the Federal Health Minister and the
TGA on issues relating to the safety, quality, and performance of medical devices supplied in
Australia including issues relating to pre-market conformity assessment and post-market monitoring.
The Committee also provides advice on the policies, procedures and priorities that should be
applied to the administration of the medical devices legislation.

Pre-market assessment by the TGA does not necessarily involve a full evaluation of the comparative
safety or effectiveness of an individual prosthesis. It uses a risk-based approach to the pre-market
assessment prior to entry of devices on to the Australian Register of Therapeutic Goods (ARTG),
and targeted and random sampling of low to medium risk devices that have been added to the
ARTG for post-market assessment of evidence to support that entry. It is not possible or necessary
to undertake a full evaluation of every therapeutic device. The level of evaluation depends on the
class of the product and the risk associated with a device determines that class. Hip and knee
replacements are ranked in a similar manner to most other countries as class IIb. This means that
these devices are classified as high risk. In September 2005 the European Commission released a
directive to change the classification of these devices to class III (extremely high risk) and that
member states implement this directive with defined transitional arrangements by September 2007.
The difference between these two classes is the increased level of prosthesis specific outcomes
evidence required for class III products before approval is given for release of the device by the
regulatory body.

The TGA also imposes other obligations on sponsors of products such as the responsibility for
reporting malfunctioning products and instigating recalls of those products when the TGA considers
that to be warranted. Critical to these activities is the need for post-market surveillance. In general,
performance of medical devices in the market is monitored in two ways. For selected devices,
samples of product are taken for testing against performance and quality standards. For all others,
the program operates the Incident Report and Investigation Scheme (IRIS) where reports of adverse
events associated with device use are received, analysed and investigated, in conjunction with the
device sponsor and manufacturer. When a problem is identified the sponsor or manufacturer
undertakes any necessary remedial action, with the TGA overseeing the process. In other words,
industry has the responsibility of post market surveillance. Under current legislation it is a
requirement of the device sponsor and manufacturer to notify the TGA if they become aware of any
adverse events.

For hip and knee replacements the advent of the National Joint Registry has established an
additional mechanism of post-market surveillance specific to this group of prostheses. All devices
implanted are now independently monitored for comparative outcomes as assessed by the rate of
revision of those devices. The National Joint Registry has for a number of years identified specific
prostheses that have a higher than anticipated rate of revision. This is defined as a revision rate that
is at least twice the rate of revision of comparable products. As the cut-off point for identification is
twice the revision rate, most of the identified products have a rate of revision that is significantly
higher than this. This information is publicly available through the annual reports and the posting
of those reports on the Registry website.
                                                56
In addition to making this information publicly available the Registry has also notified relevant
companies of their prostheses identified as having a higher than anticipated rate of revision. As far
as the Registry is aware no sponsor or manufacturer has subsequently notified the TGA of these
identified higher rates of revision. They are either unaware of their obligations under current
legislation or believe that the difference in the revision rate is not an event that requires notification.
This may in part be related to a lack of clarity as to what the triggers for notification should be. It
also could be related to sponsors and manufacturers believing that the underlying basis for the
difference in the revision rate is due to factors other than the prosthesis. It is known that many
factors other than the prosthesis have the potential to impact on the rate of revision. These include
amongst others, patient specific factors, factors related to the surgical technique or perhaps the level
of expertise and skill of the surgeon. Assessing the reasons for revision is a complex process and
there is certainly a need for fairness to sponsors and manufacturers in undertaking that assessment.
Balanced against this however is the more important imperative of ensuring safety of patients
receiving joint replacements.

As well as the TGA there are a number of other statutory committees that also provide advice on the
regulation and funding of implantable devices.

Prostheses and Devices Committee
Prostheses and Devices Committee (PDC) provides advice to the Minister on the listing of no-gap
and gap permitted prostheses and the benefit levels of prostheses and medical devices within the
private sector. As part of this process the PDC assesses devices primarily for relative clinical efficacy
based on the available evidence and data. It also categorises devices by MBS procedure. The PDC
is responsible for managing the various lists that comprise the prostheses schedule (the Prostheses
List).

Clinical Advisory Groups
Clinical Advisory Groups (CAGs) have been established to provide advice to the PDC in relation to
the relative clinical efficacy, and where practical, on cost effectiveness of prostheses and devices.
Where relevant, the CAGs also group prostheses into categories of similar clinical effectiveness to
assist consideration by the PDC.

The members of CAGs are comprised of expert clinicians in the clinical area to be examined.
Clinicians have been chosen from across Australia to ensure the best possible representation of
experience in surgery and the use of prostheses. In addition, each CAG has health fund nominees,
a non-aligned supplier nominee, a consumer representative and departmental advisors. Specific
CAGs have been established to assess hip and knee prostheses. Each application to list a new
product on the Prostheses List is assigned to two appropriate panel clinicians, in order to obtain
balanced evaluations. The difficulty for clinicians is that it is the usual situation that there is no
clinical outcome information specific to the individual prosthesis being considered. For hip and
knee prostheses the clinician provides advice on the appropriate grouping of the prosthesis. This
grouping is based on the technical features of the prosthesis.

Medical Services Advisory Committee (MSAC)
This committee was established in 1998 replacing the National Technology Advisory Panel and the
then Australian Health Technology Advisory Committee. This committee provides advice to the
Federal Health Minister on the strength of evidence pertaining to new or emerging medical
technologies and procedures in relation to its safety, clinical and cost effectiveness and under what
circumstance they should be supported by funding through the Medical Benefits Scheme. MSAC
                                                 57
considers applications for new medical services and makes decisions on the funding arrangements
for those services. The regulation of new hip and knee prostheses does not require MSAC
involvement as the implantation of these prostheses is covered by existing medical services funded
through the Medical Benefits Scheme.

Recent international attempts to regulate and improve prostheses specific outcomes
The recent up-grade to class III by the European Commission is an example of international
concern related to the use of these prostheses. The increasing number of countries implementing
registries to monitor outcomes is another. The development of guidelines by NICE is a further
example of attempts to enhance outcomes of joint replacement surgery by focusing on appropriate
prosthesis selection. NICE has identified the limited availability of adequate clinical information
and it also recognised that reliance on long-term clinical studies was a significant impediment to
innovation and that these studies were costly. At the time the original guidelines were developed
Radiostereometric analysis (RSA), which involves the insertion of radiographic markers in order to
assess prostheses movement and wear after implantation, was felt to be a promising development
that may assist in the more rapid assessment of potential long-term outcomes of these prostheses.
However, the Institute was undecided whether to recommend this as either a substitute for long-term
evidence or as an early warning indicator of expected poor long-term revision rates.

Since that time however there has been significant improvement in the understanding of the
potential role of RSA in the rapid assessment of the likely long-term outcome of new hip and knee
prostheses. There are many new implants developed using new material and designs, which have
been shown to function well invitro, but have not been validated in the clinical setting. Due to the
high level of accuracy, RSA plays an important role in the evaluation of new prostheses as studies
measuring implant fixation and wear can be performed using small numbers of patients with the
risk of clinical loosening and revision shown from six months to two years post-joint replacement.
Increased use of this approach has the potential to eliminate exposure of large numbers of patients
to new prostheses and long-term (ten years) follow-up before the validity of a prosthesis is
determined.

Suggested approach to regulatory change
It is clearly apparent that the current approach to regulation is unable to ensure new prostheses will perform
in a satisfactory manner. It is possible however to envisage a system of regulation that may significantly
reduce this risk. This would involve changes to both the pre-market approval process as well as post-market
surveillance.

Pre-market approval
A simple approach could be to follow the European lead and alter the classification of these devices
to class III. This means more detailed information on clinical outcomes would be required prior to
a prosthesis receiving approval to be added to the Australian Register of Therapeutic Goods. To
ensure the success of this approach it is necessary however to define the required clinical outcomes
necessary for approval. This is where the difficulty arises.

There are two major problems that need to be considered. Firstly, is it necessary for a new
prosthesis to perform better than established prostheses or is it acceptable if the prosthesis performs
the same? The current regulatory approach would allow approval if the prosthesis performed the
same. The effect of this is to allow an increasing number of new prostheses to be added to the
market without achieving an incremental increase in performance. Additionally it sets the
benchmark lower and thereby increases the risk of approving prostheses that have a less satisfactory


                                                  58
performance that becomes evident with increased numbers of procedures and the progression of
time.

The second and more important issue is what clinical outcomes should be required before approval
is given. Currently there are no set guidelines defining the benchmark for comparison, the clinical
outcomes indicative of satisfactory long-term outcome and the required approach to determining
these. Care must be taken in deciding these as a restrictive approach will limit innovation and
development and prevent patients from receiving the benefits of those developments.

Currently there are two possible approaches. The first is to rely on revision rate as an arbitrary
method of assessing clinical benefit. This however would require a large numbers of procedures and
monitoring over a long period of time for the majority of prostheses. It is only those prostheses that
have a very high early revision rate that would be identified with this approach. Although it is better
than the current approach it is impractical and in part defeats the purpose of limiting risk, as many
patients will still be exposed to unproven prostheses.

The second approach is to develop requirements based around RSA technology. This limits
exposure to a small number of patients and provides defined outcomes that can be benchmarked
against proven prostheses. These defined outcomes would include movement of the prostheses at
the bone-prosthesis interface and the wear of the articulating surface. These studies can be
undertaken within a short period of time (two years) and are strongly predictive of long-term
outcome. Additional information related to defined pre-clinical testing would also be required. The
pre-clinical information could be used to determine those prostheses permitted to progress to clinical
testing. It may be necessary to consider the introduction of a provisional registration step that limits
the number and use of a prosthesis to specific centres while RSA derived clinical outcomes are
obtained. Using this approach, defining the clinical outcomes criteria is a straightforward process.
It will be necessary to consult with clinicians, industry and other stakeholders to define these criteria.

There will however also be other issues to consider. The major one relates to minor design
modification to currently used prostheses. It is known that minor design changes can have a
significantly detrimental effect on outcomes and in general the safest option would be to require that
any minor design change requires repeat clinical testing. The increased scrutiny proposed by these
changes will reduce the risk of unsatisfactory prostheses being released on the Australian market.
They are however not so restrictive as to prevent innovation and development. Not only does this
approach protect individuals undergoing joint replacement but it also protects sponsors and
manufacturers by ensuring that the approved prostheses are more extensively tested and better
validated prior to general release.

Post-Market Surveillance
Although this approach reduces risk it does not eliminate it. Therefore the current requirements of
post-market surveillance need to be maintained. They also need to be maintained because the
performance of an individual prosthesis can change. In recent years there has been a number of
excellent examples where alterations to the manufacturing process has lead to increased rates of
revision of previously well performing prostheses. The relatively recent recalls of Zirconia femoral
heads and the Inter-op acetabular component are two of these.

The implementation of the National Joint Registry has placed Australia in a fortunate position. It is
not only an independent mechanism to identify both poor and changing performance but can also
identify optimal performing prostheses in different clinical settings. It is also able to establish if
higher revision rates are prosthesis specific or if they are due to patient or systemic factors. It
                                                59
removes bias from the interpretation of the analyses. The issue that remains to be resolved is how
this information is to be integrated and used by the regulatory and funding frameworks of the health
system.

As part of its current contract with the Commonwealth the Australian Orthopaedic Association has
agreed to develop mechanisms during the 2006-2007 financial year by which it reports directly to the
TGA. As part of this process it will be necessary to agree on what is the trigger for reporting. In
other words, what is the revision rate for a defined number of procedures over a defined period of
time before notification is required? The TGA will also need to determine what action may be
necessary once notification occurs and if this action should vary depending on the rate of revision.
Options could include ongoing monitoring, limitation of use to specific clinical settings or removal
from the Australian Register of Therapeutic Goods. It should be possible in consultation with
relevant stakeholders to establish the criteria for notification as well as the appropriate action
depending on the identified rate of revision.

In addition to developing the mechanisms for direct notification to the TGA the Australian
Orthopaedic Association has also agreed to integrate the PDC Prostheses List with the Registry
database. This will enable the PDC to have information available on use and comparative outcomes
that is presented in a manner relevant to the Prosthesis List. The purpose of this is to assist the
PDC in determining appropriate funding for classes of prostheses and the individual components
within those classes.


CONCLUSION

The outcomes of joint replacement surgery in Australia are satisfactory compared to most
international comparisons. The revision rate however remains significantly higher that that
experienced in Sweden. The difference is related to prostheses rather than patient selection. The
Australian Joint Registry is now capable of providing quality data that is able to define the areas
where outcomes can be improved. Developing strategies to address issues identified by the Registry
will enhance the rate of improvement. These strategies include the development of clinical
guidelines that outline best practice in prosthesis selection and surgical techniques. In addition
there is also a need to increase scrutiny of new prostheses prior to their addition to the Australian
Register of Therapeutic Goods and use Registry information to undertake more effective post-market
surveillance.




                                              60
1. Australian Orthopaedic Association National Joint Replacement Registry. Annual Report.
    Adelaide: AOA; 2005.
    http://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2005.pdf (accessed 7
    July 2006)
2. Freeman MAR, Plante-Bordenneuve P: Early migration and late aseptic failure of proximal
    femoral prostheses. J Bone Joint Surg, 1994, 76—B: 432—438.
3. Karrholm J, Malchau H, Snorrason F, Herberts P. Micromotion of femoral stems in total
    hip arthroplasty. J Bone Joint Surg (Am) 1994; 76(11): 1692—705
4. Karrholm J, Borssen B, Lowenhielm G, Snorrason F. Does early micromotion of femoral
    stem prostheses matter? 4—7 year stereoradiographic followup of 84 cemented prostheses. J
    Bone Joint Surg (Br), 1994, 76: 912.
5. Krismer M, Stockl B, Fisher M, Bauer R, Mayrhofer P, Ogon M. Early migration predicts
    late aseptic failure of hip sockets. JBJA 7—-B, 1996, 422—426.
6. NICE Technology Appraisal Guidance No. 2 Guidance on the Selection of Prostheses for
    Primary Total Hip Replacement. London: NICE; 2000.
    http://www.nice.org.uk/page.aspx?o=TA002guidance (accessed 14 June 2006)
7. NICE Technology Appraisal No TA44 Hip disease - metal on metal hip resurfacing:
    Guidance. London: NICE; 2002. http://www.nice.org.uk/page.aspx?o=TA044guidance
    (accessed 14 June 2006)
8. Therapeutic Goods Administration Business Plan, July 2006 — June 2007, Australian
    Government, Department of Health and Ageing.
    http://www.tga.gov.au/about/tgabp0607.pdf (accessed 7 July 2006)
9. Therapeutic Goods Administration Medical Device Regulatory Process, Commonwealth of
    Australia 2006. http://www.tga.gov.au/devices/process.pdf (accessed 7 July 2006)
10. Walker P, Mai SF, Cobb AG, Bentley G, Hua J. Prediction of clinical outcome of THR from
    migration measurements on standard radiographs. J Bone Joint Surg, 1995, 77B: 705—71.




                                        61