ISPOR Patient Registry SIG Forum ISPOR 14th Annual International

Document Sample
ISPOR Patient Registry SIG Forum ISPOR 14th Annual International Powered By Docstoc
					ISPOR Patient Registry SIG Forum

                                                                      Classification, Strategy & Design
                                                                               Working Group

            Patient Registry SIG                                 Goals:
                                                                    determine and define the most appropriate
                                                                    language for patient registry standardization, a
            Classification, Strategy & Design                       patient registry terminology (common language),
                                                                    universal patient registry characteristics and a
            Working Group                                           globally harmonized patient registry classification
            Chair: Chris L. Pashos PhD
            Vice President and Executive Director, HERQuLES         to establish good research practices related to
            Abt Bio-Pharma Solutions, Inc.                          choices of registry strategy and consequent design.

            ISPOR Taxonomy of Patient Registries                      Classification, Strategy & Design
                                                                               Working Group

                                                                 Establishment of 4 Project Teams:
              Each term will include a brief definition, a
              broader explanation, the associated values &
              uses, and conclude with a discussion of issues        Characteristics & Classifications
              or conflicts related to the term.                     of Patient Registries
                                                                    Design, Development & Implementation
              The issues/conflicts will be the basis
              of the Working Group’s Good                           Analysis
              Research Practices papers.                            Reporting & Publishing

             The Taxonomy Teams’ Methodology

            Identification of terms: hand-searched existing      Patient Registry SIG
              sources for terms:
                                                                 Classification, Strategy & Design
               – Berger et al, ISPOR Book of Terms (2003)
                                                                 Working Group
               – AHRQ, Registries for Evaluating Patient
                 Outcomes: A User’s Guide (2007)
               – CONSORT, ICJME, selected journal requirements   Team 1: Characteristics & Classifications
                 for authors
                                                                 Co-Chair: Dimitris Polygenis PharmD
                                                                 Co-Chair: Sally Thompson PhD, MSc

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

                   Characteristics & Classification

                Co-chair: Sally Thompson PhD, MSc                                           Developed working definition
                Director, Outcomes Research, Pfizer, Inc
                                                                                            Identified common elements and considerations
                Co-chair: Dimitris Polygenis PharmD
                Vice President, McKesson Specialty                                          Agreement as to what is NOT a registry
                McKesson Corporation
                                                                                            Identified commonly used registry designs and
                Grace Leung MPH                                                             applications
                Health Economist, Genentech
                                                                                            Set the stage for further discussion/clarification
                Neal Mantick
                Director, Registries, Abt Bio-Pharma Solutions, Inc.
                                                                                            (to follow in subsequent sections)

                             Registry Definition                                       Essential Characteristics of a Registry

                                                                                       Characteristics    Considerations

                Prospective observational study of                                     Observational      • Real world assessment

                subjects, with certain shared                                          Non-
                                                                                                          • No protocol-defined treatment/management, allocation
                                                                                                          of patients and patient visits
                characteristics, that collects ongoing                                 interventional     • Limited risk; ethics review/consent required however
                                                                                                          focus is on protection of personal health information
                and supporting data over time on well-                                                    • Dictated by patient and patient experience
                defined outcomes of interest for                                       Data Collection
                                                                                                          (i.e., heterogeneous and missing data)
                                                                                                          • Need to define key assessments and outcomes
                analysis and reporting                                                                    of interest

                                                                                                          • Baseline assessment critical
                                                                                                          • Longer-term observation period
                                                                                                          • Hypothesis generating versus hypothesis testing

                        Key Differences                                                             Key Differences
                  versus other Study Designs I                                               versus Other Study Designs II

                 Characteristic    Registry versus Traditional RCT                            Characteristic Registry versus Traditional RCT

            1    Treatment         Evaluate care in real-world setting                        Statistical       Hypothesis generating; no sample size
                                                                                        5     Analysis and      calculation; focus on ‘generalizability’
                 Time                                                                         Data Collection   Heterogeneous patients
            2                      Long-term outcomes collected

                                   Can involve large numbers of patients; ‘typical’           Patient Consent Focus on handling of personal health information
            3    Patients          patients seen in real-world setting                        & Ethics Review and not risk
                                   Limited inclusion/exclusion criteria
                                                                                                                Voluntary reporting of adverse events
                                   Do not require comparator/placebo; ‘typical’ care    7     Safety
                                                                                                                Unsolicited (vs. solicited) adverse event collection
                                   Open-label; no defined/mandated interventions or
            4    Methods
                                   data collection
                                   No random allocation of patients

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

                               Registry Classification I                                                                                            Registry Classification II

               Registry Type                                                                                                        Registry Type
                                             Design                  Measurement                     Application/Use                                    Design                    Measurement                           Application/Use
                 Sponsors                                                                                                             Sponsors
                                                                                                                                                                                                             •   Mandated by regulators to meet
                                  •    Prospective                                                                                                  •   Prospective
                                                                                                                                                                         •   Clinical outcomes as compared       specific goals and objectives in
             Simple Cohort        •    Non-interventional                                   •     Pregnancy registry                                •   Interventional
                                                                                                                                                                             to clinical studies                 minimizing known risks while
                                                                 •     Clinical                                                                     •   Population-
                                  •    Sample-based                                         •     Determine                                                              •   Safety information and              preserving benefits
                                                                       outcomes i.e.,                                              Risk                 based
             Epidemiologists      •    Collection of information                                  association/correlation                                                    adverse events compared to      •   Assessing product’s risk-benefit
                                                                       morbidity,                                                  Management       •   Use one or
                                                                                                                                                                             clinical studies                    balance
             Public Health             in population that share                                   between exposure and                                  more tools to
                                                                       mortality                                                                                         •   Compliance with prescribed      •   Developing and evaluating tools to
             Clinicians                common exposure (i.e.,                                     outcome                          Regulators           meet goal(s)
                                                                                                                                                                             management and prescribing          minimize risks while preserving
                                       pregnancy registry)                                                                         Manufacturers    •   May collect
                                                                                            •     Understand natural history                            info beyond
                                                                                                                                                                             protocols                           benefits
             Outcomes                                                                             of patient cohort that share                                           •   Impact of tools on ensuring     •   Making adjustments to risk
                                                                                                                                                                             compliance an outcomes              management tools to further improve
                                  •    Prospective                                                common characteristic i.e.,                           labelling
                                                                 •     Clinical                                                                                                                                  risk-benefit balance
             Epidemiologists      •    Non-interventional                                         social science research,
                                                                       outcomes i.e.,
             Policy makers        •    Population-based                                           population-based research,                        •   Prospective                                          •   Understand natural history of disease
                                                                       morbidity,                                                                   •   Non-                                                 •   Identify, compare and evaluate
             Governments          •    Collection of information                                  epidemiological research
                                                                       mortality                                                                        interventional                                           management patterns
             Public Health             in population                                        •     Examples: mortality,
                                                                                                                                                    •   Population-      •   Drug utilization and safety     •   Identify ‘signals’ relating to safety,
             Academia                                                                             literacy, access to medical      Disease
                                                                                                                                                        based            •   Outcomes – morbidity and            effectiveness and outcomes
                                                                                                  care, etc.                                        •   Collects             mortality                       •   Quantify burden of illness, QoL
             Safety               •    Prospective                                                                                 Regulators
                                                                 •     Adverse events •           Support product registration                          information in   •   Resource utilization            •   May be iterative in establishing and
             Surveillance         •    Non-interventional                                                                          Manufacturers
                                                                 •     Unexpected     •           Conduct post-marketing                                cohort of        •   Clinical management                 benchmarking best practices
                                  •    Sample-based                                                                                                     patients with                                        •   Assess screening, identification and
                                                                       adverse events             surveillance (‘real world
             Manufacturers        •    Collection of information                                                                                        common                                                   monitoring practices
                                                                 •     Serious                    setting’)
             Regulators                in patients receiving                                                                                            disease                                              •   Cost-effectiveness
                                                                       adverse events •           Identify ‘signals’
             Clinicians                common intervention

                             Registry Classification III

               Registry Type
                                          Design                 Measurement                         Application/Use               Patient Registry SIG
                                                            •    Safety and
                                                                                        •       Post-marketing surveillance
             Drug and Drug         •   Prospective (some         effectiveness
                                                                                        •       Compare effectiveness to
             Class                     retrospective)       •    Outcomes –
                                                                                                                                   Classification, Strategy & Design
                                   •   Sample-based              morbidity and
                                                                                        •       Study non-approved uses
             Clinicians            •   Collects information      mortality
                                                                                        •       Identify drug-related ‘signals’
                                       on patient cohort
                                       receiving common
                                                                 Resource utilization
                                                                                                Cost effectiveness
                                                                                                Willingness to pay
                                                                                                                                   Working Group
             Payers                    treatment                 management and
                                                                                        •       Reimbursement evaluation
                                                                 add-on therapy
                                                                                        •       Care mapping
             Management            •   Prospective/retrosp
                                                                                        •       Continuous quality improvement
                                       ective                •   Treatment and
                                                                                        •       Resource utilization and costing
                                   •   Collect information
                                       on common
                                                                                        •       Burden of illness                  Team 2: Design, Development &
                                                                                        •       Quality of care
             Health policy makers
             Health administrators •
                                                                 Resource utilization
                                                                                        •       Provider performance               Implementation
                                                                                        •       Health economic evaluation
             Academia                  based
                                                                                        •       Reimbursement evaluation
                                                             •   Direct costs i.e.,
                                                                 medical care, drug     •       Burden of illness
                                                                                                                                   Chair: Eric Gemmen, MA
                                   •   Prospective or
             Payers                    retrospective
                                                                 use, hospitalization   •       Cost of care                                                                                 Research,
                                                                                                                                   Senior Director, Medical Affairs, Epidemiology & Outcomes Research,
                                                             •   Productivity costs     •       Reimbursement evaluation
             Policy makers
                                   •   Sample-based
                                                                 i.e., absenteeism,     •       Health economic evaluations        Quintiles Late Phase & Safety Services
             Health administrators

        Design, Development & Implementation Members                                                                                                Achievements
                                                                                                                                          Identified 25 categories of terms
          Murtuza Bharmal PhD
          Associate Director, Quintiles Late Phase & Safety Services
          Maznah Dahlui MD, MPH
          Department of Social and Preventive Medicine, University of Malaya
                                                                      Malaya                                                                                          including:
                                                                                                                                          9 in Development section including:
          Nancy Dreyer PhD, MPH                                                                                                                registry purpose, funding and oversight,
          Donatus Ekwueme PhD                                                                                                                  stakeholders, scope, ethics and privacy, regulatory
          Senior Health Economist, U.S. Centers for Disease Control & Prevention (CDC)
                                                                      Prevention                                                               considerations, etc.
          Claudio Faria, PharmD MPH
          Associate Director of Clinical Research, UMass Medical School                                                                                  including:
                                                                                                                                          11 in Design including:
          Huan Huang PhD, MS
          Senior Analyst, Boston Health Economics
                                                                                                                                               research question(s), design characteristics, study
          Joanna Lis PhD, MBA                                                                                                                  population, data elements, data sources, data
          Manager of Health Economics Department, sanofi-aventis, Warsaw, Poland
                                                        sanofi-                                                                                collection materials & methods, guidelines &
          Yvonne Lis PhD                                                                                                                       standards, registry size and duration, etc.
          Director, Carter-Lis Associates Limited
          Anuprita D Patkar, PhD                                                                                                                                 including:
                                                                                                                                          5 in Implementation including:
          Associate Director, Health Economics & Reimbursement, ETHICON
          Gabriel Sandblom MD, PhD
                                                                                                                                               pre-launch issues, site support, data capture &
          Department of Surgery, University Hospital, Lund, Sweden                                                                                                        close-
                                                                                                                                               management, data lock, close-out
          Kathryn Starzyck MSc
          Associate Director of Scientific Affairs, Outcome

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

             160+ terms /definitions completed                                                        Challenges

               76 terms in Development section including:
                                                     including:              The terms ‘registry’ and ‘observational study’
                   exposure, feasibility, informed consent, IRB/ethics       are often used interchangeably, although
                   approval, target population, etc.                         registries are a subset of observational studies
               50 terms in Design including:                                  Moreover, the term ‘observational’ may differ in
                   observational, non-interventional, naturalistic,
                                    non-                                       meaning between Europe and the US
                   active/passive surveillance, historical control, etc.              European definition more strict
               37 terms in Implementation including:
                   site identification, regulatory documents, ICF-GCP,
                   database build, clinical research associate, query        Scope - keeping the terms specific to registries
                   resolution, loss to follow-up, source document
                                         follow-                             and not simply clinical studies, overall
                   verification (SDV), site close-out, etc.

                  Example: Patient Recruitment                                         Achievements & Next Steps

            Patient Recruitment                                            2009 ISPOR 14th Annual International Meeting in Orlando
                                                                             Workshop Presentation
              Brief Definition: The process of enrolling patients into
              the registry.
              Explanation: Registry participants are recruited on a        W12: A TAXONOMY FOR THE DESIGN, DEVELOPMENT AND
              disease basis or exposure/treatment basis                      IMPLEMENTATION OF PATIENT REGISTRIES
               – Only after treatment/prescribing decision has been
                 made by treating physician                                Available on the ISPOR website via the Research Digest or on the
              Issues:                                                        Patient Registry Classification, Strategy & Design Working Group
               – Can the existence of a product registry impact the          webpage
                 physician’s treatment/prescribing decision?
               – Is the enrolled sample representative of the target       Draft an article based on our findings and submit to ISPOR
                                                                             CONNECTIONS for September/October 2009 issue.
               – Are enrollment caps/limits at given sites appropriate?

                                                                                      Team 3: Analysis Members
                                                                           Co-Chair: Shital Kamble MS
                                                                           PhD Candidate, Health Services Research, University of North Carolina at Charlotte
                                                                           Co-Chair: Christopher Blanchette PhD
                                                                           Director, Center for Pharmacoeconomic and Outcomes Research
           Patient Registry SIG                                            Lovelace Respiratory Research Institute
                                                                           Maznah Dahlui MD, MPH
                                                                           Lecturer, Department of Social and Preventive Medicine
                                                                           University of Malaya
           Classification, Strategy & Design                               Donatus Ekwueme PhD
                                                                            Senior Health Economist, US Centers for Disease Control and Prevention
           Working Group                                                   Alex Exuzides PhD
                                                                            Director, ICON Clinical Research
                                                                           Eric Gemmen MA
                                                                           Senior Director, Medical Affairs, Epidemiology & Outcomes Research,
           Team 3: Analysis                                                Quintiles Late Phase & Safety Services
                                                                           Carl Gibbons BSc
                                                                            Research Analyst, Schering-Plough Ltd
                                                                           Joanna Lis PhD MSc
                                                                            Health Economics Manager, Sanofi-Aventis
           Co-Chair: Shital Kamble MS                                      Anuprita Patkar PhD
           Co-Chair: Chris Blanchette PhD                                   Associate Director, Health Economics & Reimbursement, ETHICON, J&J
                                                                           Matt Reaney MSc
                                                                           Health Outcomes Scientist, Endocrine, EuOR, Lilly UK

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

                            Achievements                                                    Achievements

            Identified 13 broad categories of terms                    70 analysis terms definitions completed

               Power/Sample Size Calculations                             Charlson Comorbidity Index, Elixhauser Comorbid
                                                                          Disease Adjustment Method, etc.
               Statistical Inference and Hypothesis Testing
                                                                          Regression- Ordinal Logit/Probit Models, Cox
               Main Analysis Techniques
                                                                          Proportional Hazards Models, Two-part Models,
               Treatment of Selection Bias                                Multilevel Models, etc.
               Treatment of Missing Observations                          Propensity Score Methods, Instrumental Variables
               Multiplicity Adjustments                                   Clinical Significance, Statistical Significance, etc.
               Systematic Reviews/ Meta-Analysis                          Missing Completely at Random (MCAR), Missing at
                                                                          Random (MAR), etc.

                             Achievements                                  Example: Multiplicity Adjustments

            2008 ISPOR 11th European Congress in                       Brief Definition: The process of adjusting for multiple
                                                                          statistical tests to correct for occurrence of false
              Athens Workshop                                             positives (i.e., Type I Error) that could emerge from
                                                                          investigators looking at many additional endpoints and
            W20: Use of Real World Data: Challenges in                    treatment group comparisons.
             the Use of Patient Registry Data
                                                                       Explanation: different methods of correcting for multiple
                                                                          testing procedures in clinical trials or observational
            Available on the ISPOR website via the Research                –   Bonferroni method,
              Digest or on the Patient Registry Classification,            –   Hochberg false discovery rate (FDR) method,
              Strategy & Design Working Group webpage                      –   Holm correction method,
                                                                           –   Westfall and Young Permutation (Hierarchical Testing),
                                                                           –   Bootstrap method

              Example: Multiplicity Adjustments

            Issues:                                                    Patient Registry SIG
               – Is multiplicity adjustment necessary?
                  - currently a trend to discount the multiplicity     Classification, Strategy & Design
                 problem and its effects                               Working Group
               – very few studies have described specific conditions
                 that demand the use of multiplicity adjustment as a
                 control measure
                                                                       Team 4: Reporting & Publishing

                                                                       Co-Chair: Kirsten Hall Long PhD
                                                                       Co-Chair: Diana Frame

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

                   Reporting & Publishing Members                                                      Achievements

           Co-Chair: Kirsten Hall Long PhD
           Senior Health Economist, Division of Health Care Policy & Research,
           College of Medicine, Mayo Clinic                                         Identified 3 broad categories of terms
           Co-Chair: Diana Frame                                                       Validity & Quality
           Independent Consultant, Frame Research LLC
                                                                                       Ethical Considerations
           John Ellison                                                                Public Reporting of Registry Data
           Senior Manager, Scientific Publications, Clinical Research Department,
           LifeScan, Inc.
           Huan Huang PhD, MS
           Senior Analyst, Boston Health Economics
           Siva Narayanan MS, MHS
           Vice President and Practice Leader, Treatment Performance Optimization
           – Global Portfolio, TNS Healthcare

                                  Achievements                                                         Achievements

            26 Reporting & Publishing terms defined                                 26 Reporting & Publishing terms
              Internal and external validity                                          Full draft of all terms completed
              Bias (selection, response, recall, attrition)
              Quality domains                                                         Brief Definition, Explanation, Value and Use, Issues,
                                                                                      and Bibliography sections for each term
              IRB / Ethics Committee (reporting considerations)
              Registry funding                                                        Beginning to identify cross-indexing with other
              Authorship                                                              sections
              Publication bias                                                        Looking ahead to Best Practices work
              Transparency                                                            ("preview" of this especially in Issues sections)

                 Reporting and Publishing Example                                       Reporting and Publishing Example

            Transparency                                                            Transparency
                                                                                      Issues: Observational studies have received less attention
               Brief Definition: A characteristic of the report                       than clinical trials in the quest for complete and timely
               defined by the presence and clarity of key                             reporting of study results. Postmarketing and registry studies
                                                                                      play an important role in evidence development, however,
               information on the rationale, methodology, and                         and efforts to increase transparency should encompass these
               support (including funding) used to collect, analyze,                  observational studies as well. Given limitations on manuscript
               and publish registry data.                                             length in print journals and the frequent complexity of study
               Explanation: Transparency in reporting facilitates                     design and analytic models in observational research, some
                                                                                      authors have called for availability of the detailed protocol,
               interpretation of the study by other researchers as                    coding definitions, and other methodologic detail on the
               well as health care decision-makers, and may lead                      journal’s or researcher's web site.
               to faster adoption of results.

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

                                                                                Analysis & Data Management
                                                                                       Working Group

                                                                         Establishment of 3 Project Teams:
            Patient Registries SIG
                                                                           Cost-effectiveness Analysis of Patient Registry Data Tea
            Analysis & Data Management                                     Analysis of Effectiveness in Registry Data Team
            Working Group                                                  Reporting Results from Registry Data for Publications

                                                                         Co-Chair: Mia Malmenäs MSc, Shire HGT
             Chair: Leanne Larson MHA                                    Co-Chair: Mike Novotny MBA, MA, Medrio
             Vice President, Sg2 Healthcare Intelligence

            Analysis & Data Managment Members
                                                                           Analysis Issues for Patient Registries

             Lusine Abrahamyan MD, MPH, University of Toronto
             Marg Hux, MSc, i3 Innovus                                    Our Focus – guidance & recommendations for analysis
             Michelle Pritchard Turner MS, ICON Lifecycle Sciences          within registries:
             Rebecca Gruhlkey MBA, Thomson Healthcare
             Isabelle Morin MSc, Shire HGT                                  Cost-effectiveness
             Nancy Dreyer PhD, MPH, Outcome                                 Effectiveness
             Claudio Faria PharmD, MPH, University of Massachusetts
                                                                            Reporting of registry data for publication
             Nandan Kenkeremath JD, BS, Leading Edge Policy & Strategy
             Peggy Schrammel, University of British Columbia
             Steven Takemoto PhD, University of California
             Fang Wang MD, PhD, GlaxoSmithKline
             Jaro Wex PhD, MD, BA, Pharmarchitecture Limited

                  Analysis Issues for Registries                                     Analysis of Registries

            Existing guidelines lack practical analysis guidance            Registries well suited to collect ‘real-world’ data
             – AHRQ Registry guidelines
                                                                            Incremental Effectiveness – issues identified
             – GRACE initiative
                                                                            Disease-related cost over fixed time frame
            Analysis issues differ by registry GOALS:                        – Identify relevant resources to include in cost
             – Natural history of disease                                    – Treatments, administration, safety and effectiveness
             – Treatment practice and effectiveness                               e.g. management of adverse events, long term
             – Burden of illness and cost-effectiveness of treatments             complications of progressing disease

             – Monitor treatment safety / harm
             – Quality of care

ISPOR 14th Annual International Meeting
ISPOR Patient Registry SIG Forum

          Good Research Practices for Cost Effectiveness                  Good Research Practices for Analysis of
                Analysis of Patient Registry Data                             Effectiveness in Registry Data

            Provide recommendations on:                                  Provide recommendations on suitable statistical
                                                                           approaches to registry data with a particular focus
              Dealing with missing data                                    on estimating the effectiveness of treatment
               – Variable Follow-up - balance long timeframe with          methods:
                 need to impute                                             –   covariate adjustments
               – Missing assessments, incomplete information                –   matching
                                                                            –   propensity scoring
              Dealing with potential bias – effect on cost
                                                                            –   missing data handling
               – Selection bias, ascertainment and measurement bias
                                                                            –   etc
              Cultural and country differences

              Good Research Practices for Reporting
            Results from Registry Data for Publications                                   Achievements

            Develop recommendations for reporting results                2009 ISPOR 14th Annual International Meeting in
              from registry data for publications and evaluate             Orlando Workshop Presentation
              the STROBE guideline for its appropriateness
              for patient registries                                     W9: ANALYSIS OF EFFECTIVENESS AND COST-
               – Develop a checklist for best practices                   EFFECTIVENESS IN PATIENT REGISTRIES
               – A point by point list analyzing the STROBE
                 guideline for relevance to registries
                                                                         Available on the ISPOR website via the Research
                                                                           Digest or on the Patient Registry Analysis
                                                                           Working Group webpage


            2009 ISPOR CONNECTIONS Article
            May – June issue

            Analysis of Effectiveness in Patient Registry Data
            By Mia Malmenäs MSc, Keith Lowton MSc, Isabelle Morin MSc,
               Shire Human Genetic Therapies, Stockholm, Sweden;
               Margaret Hux, MSc, i3 Innovus Burlington, ON, Canada,
               Lusine Abrahamyan MD, MPH, University of Toronto
               Mike Novotny MBA, MA, Medrio

ISPOR 14th Annual International Meeting

Shared By: