Workshops
Monday, June 2, 2014
5:00 PM - 6:00 PM
WORKSHOPS - SESSION I
Clinical Outcomes Research

5:00 PM - 6:00 PM
Room: Room 710b (7th Floor)

W1: CHANGING THE DRUG DEVELOPMENT PARADIGM

Discussion Leaders:

Adrian Towse, MA, MPhil, Director, Office of Health Economics, London, UK

Penny Mohr, MA, Senior Vice President, Program Development, Center for Medical Technology Policy, Baltimore, MD, USA

Donna Messner, PhD, Research Director, Center for Medical Technology Policy, Baltimore, MD, USA

PURPOSE:

Payers and health care systems are demanding comparative or relative effectiveness research (CER/RE) to aid coverage and pricing decisions.  Drug development costs have increased markedly in recent years, causing manufacturers to seek creative solutions to improve efficiency and get drugs used earlier. The environment for conducting CER/RE research is changing rapidly-with innovative study designs and growing access to electronic data. The time is ripe for a new drug development paradigm (NDDP). This workshop will (i) set out how the future environment for CER/RE research may look (based on a scenario-building exercise involving an intensive modified-Delphi process including key informant interviews with 90 payers, regulators, patient advocates, experts in health information technology, health system change, research methods and personalized medicine, senior executives in the life science industry, and thought leaders in CER and RE in the United States and Europe), (ii) summarise the different proposals put forward in the literature for radical change to the drug development paradigm, and (iii) explore how one paradigm (based on facilitated discussions between executives and senior leaders from five major life science companies) would tackle generating evidence pre- and post-launch for two drug archetypes. The purpose is to equip participants with an understanding of how evidence generation may change and the implications for the development of drugs.

DESCRIPTION:

This workshop will be in four parts: (i) presentation of US and EU scenarios for the evolution of CER/RE evidence requirements from payers and HTA bodies (ii) presentation of literature review of radical changes to the drug development paradigm (iii) presentation of one new paradigm and how it would be applied to two different drug archetypes  (iv) workshop participants will be expected to debate the question as to the type of studies in the new paradigm that are relevant to the two drug archetypes.

Economic Outcomes Research

5:00 PM - 6:00 PM
Room: Room 520ad (5th Floor)

W2: A NEW VALIDATION-ASSESSMENT TOOL FOR HEALTH-ECONOMIC DECISION MODELS

Discussion Leaders:

Pepijn Vemer, MSc, Researcher, Epidemiology, UMC Groningen, Groningen, The Netherlands

George Van Voorn, PhD, Senior Researcher, Biometrics, Wageningen University & Research, Wageningen, The Netherlands

Isaac Corro Ramos, PhD, Senior Researcher, Institute of Medical Technology Assessment (iMTA), Erasmus University, Rotterdam, The Netherlands

Josephine A. Mauskopf, PhD, MHA, MA, Vice President, Health Economics, RTI Health Solutions, Research Triangle Park, NC, USA

PURPOSE:

To share a first view of a new validation-assessment tool, to discuss how it will help to improve the model review process and support future decision making, and to discuss the added value compared with other available instruments. Audience members are actively invited to comment on the draft tool and make suggestions of improvement.

DESCRIPTION:

A validation-assessment tool is being developed for decision makers to transparently and consistently evaluate the validation status of different health-economic decision models. It is designed as a list of validation techniques covering all relevant aspects of model validation to be filled in by modelers, where they will indicate if and how each validation technique has been executed and what the outcomes were. It is then used by decision makers and their advisors to gain insight in the validation status of a model, without having to validate the model themselves. The tool is envisioned to be part of, for example, a submission dossier. The project purposely included experts from outside the health-economics field, in order to gain new insights. Starting from a gross list of validation techniques, the tool was developed using a Delphi panel of more than 50 international experts in several rounds.  The workshop will start with an overview of the importance of model validation and the problems associated with it, such as consistency, transparency, and acceptability. Next, a first version of this tool will be shared, with a discussion on how it may help improve the model review process and support decision making. Next, we will actively invite audience members to add their own expertise by commenting on the draft tool. The workshop will finish with a discussion of the added value compared to other available instruments, among others the 2006 Phillips guidelines and the ISPOR Assessing Modeling Studies for Health Care Decisions Task Force.

Health Care Policy Development Using Outcomes Research

5:00 PM - 6:00 PM
Room: Room 518abc (5th Floor)

W3: USE OF NETWORK META-ANALYSIS IN DIFFERENT SETTINGS: METHODOLOGICAL AND POLICY CONSIDERATIONS

Discussion Leaders:

Huseyin Naci, PhD, MHS, Fellow in Pharmaceutical Policy Research, Department of Population Medicine, Harvard Medical School, Boston, MA, USA

James D Chambers, PhD, MPharm, MSc, Assistant Professor, The Center for the Evaluation of Value and Risk in Health, Institute for Clinical Research and Health Policy Studies, Tufts Medical Center, Boston, MA, USA

Jeroen P Jansen, PhD, Director, Redwood Outcomes, San Francisco, CA, USA

PURPOSE:

Clinical evidence synthesized using network meta-analysis (NMA) often informs coverage and reimbursement decisions in the health technology assessment (HTA) setting. In this workshop, our objective is to review the relevance of NMA to decisions beyond HTA and outline methodological and policy relevant issues that arise when conducting NMA to support evidence-based decision making in different settings.

DESCRIPTION:

Comparative evidence on the benefits and harms of health care interventions generated from clinical trials rarely exist. NMA is increasingly used to generate and synthesize evidence on the comparative effectiveness of multiple interventions. In this workshop we will first provide an overview of different types of decisions that rely on comparative clinical evidence. We will then focus on study design considerations when conducting NMAs to inform decision making in different settings, particularly focusing on decisions regarding drug approval (e.g., European Medicines Agency); HTA (e.g., UK National Institute for Health and Care Excellence); pharmaceutical research and development (e.g., pharmaceutical industry), value-based insurance design (e.g., managed care); and pharmacological treatment in clinical practice. Using case studies of NMAs conducted across different settings, we will compare the evidence base and analytic approach adopted in NMAs, and will discuss the internal and external validity of findings. We will illustrate how differences in analytic approach may have policy implications, particularly if NMAs conducted in different settings produce potentially discrepant findings. We will caution participants about the dangers of a mechanistic synthesis of evidence, without a careful exploration of possible differences in perspective. Participants will gain a thorough understanding of how NMA can be used to support evidence-based decisions in different settings.


5:00 PM - 6:00 PM
Room: Room 519ab (5th Floor)

W4: MODELLING ALCHEMY: THE IMPACT OF UNORTHODOX TRIAL DESIGN ON HEALTH TECHNOLOGY APPRAISAL STRATEGY

Discussion Leaders:

Rachel Beckerman, PhD, Principal, Value Demonstration, CBPartners, New York, NY, USA

Meghan Gallagher, Director, Global Health Economics and Outcomes Research, Oncology, Sanofi, Cambridge, MA, USA

Anthony Hatswell, MSc, Principle Consulting Economist, BresMed, Sheffield, UK

Marc Bardou, MD, PhD, Gastroenterologist, Centre Hospitalier Universitaire Le Bocage, Dijon, France

PURPOSE:

The purpose of this workshop is to understand the manufacturer, payer, and academic perspectives on unconventional clinical trial design in orphan drug development. The issues are ethical, statistical, and policy-related in nature, and present complex challenges from an economic modeling and health technology appraisal standpoint. Through the discussion of recent case studies, this workshop will look at possible solutions to the challenges involved in the evaluation of evidence stemming from unusual trial designs.

DESCRIPTION:

Over the last decade, pharmaceutical manufacturers have increasingly focused development on therapies that address small patient populations with high unmet need, including the identification of disease subtypes, and the development of therapies for orphan designations.  In developing therapies for these indications, several challenges may arise from an evidence generation standpoint, including a limited patient population, ethical obligations to avoid placebo controls, lack of well-defined standard of care, or strong efficacy results leading to early termination of the trial. This often means that the ‘gold standard’ template of a double-blind RCT may be inappropriate, resulting in trial designs such as non-comparative single-arm studies, open-label designs, patient cross-over, and early un-blinding. At HTA submission, manufacturers then have a difficult task in demonstrating a product’s comparative clinical or economic benefit, while payers need to make decisions on a less than ideal evidence base.  This workshop will explore the reasons behind unorthodox trial designs, look at the statistical and modelling issues arising from such trials, examine how recent products with these evidence packages have fared across countries, and highlight the decision-making trade-offs that must be made when evaluating such evidence from the perspective of a payer.  Via group discussion, participants in this workshop will be encouraged to contribute their experience and perspective, in order to understand how best to deal with the consequences of more unusual trial designs.

Patient-Reported Outcomes & Patient Preference Research

5:00 PM - 6:00 PM
Room: Room 710a (7th Floor)

W5: REGULATORY QUALIFICATION OF CLINICAL OUTCOME ASSESSMENT TOOLS FOR DOCUMENTING TREATMENT BENEFIT IN CLINICAL TRIALS: CORE ISSUES

Discussion Leaders:

Stephen Joel Coons, PhD, Executive Director, Patient-Reported Outcome (PRO) Consortium, Critical Path Institute, Tucson, AZ, USA

Ari Gnanasakthy, MSc, Head, Patient Reported Outcomes, Novartis Pharmaceuticals, East Hanover, NJ, USA

Ashley Slagle, PhD, MS, Endpoints Reviewer, SEALD, ONDIO, CDER, U.S. Food & Drug Administration (FDA), Silver Spring, MD, USA

Elisabeth Piault, PharmD, MA, Senior Patient Reported Outcomes Scientist, Genentech, San Francisco, CA, USA

PURPOSE:

The US Food and Drug Administration (FDA) released two guidance documents with significant relevance to the use of clinical outcome assessments (COAs) to support medical product label claims.  The Patient-Reported Outcome (PRO) Consortium has developed core messages aimed at increasing understanding of the terminology, processes, and regulatory expectations described in these documents.  These core messages will be discussed with workshop participants.

DESCRIPTION:

COAs, which include patient-reported, clinician-reported, observer-reported, and performance outcome measures, are receiving increasing attention as efficacy endpoints in clinical trials.  Although not required, COA qualification is supported by the FDA to enhance the quality and number of publicly available COAs.  As stated by the FDA’s Center for Drug Evaluation and Research, COA qualification is based on a review of the evidence to support the conclusion that the COA is a well-defined and reliable assessment of a specified concept of interest for use in adequate and well-controlled studies in a specified context of use.  The FDA’s guidance documents titled “Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims” and “Qualification Process for Drug Development Tools” provide important information regarding regulatory expectations for COAs as efficacy endpoints in confirmatory trials.  In addition, the FDA COA qualification program webpage has further detail regarding the process and supporting evidence required.  However, to optimize the use and usefulness of these information sources, more uniform communication, understanding, and interpretation of the content must be achieved.  Discussion leaders from the FDA, pharmaceutical industry, and Critical Path Institute will discuss the types of COAs, distinguish between COA qualification and the regulatory acceptance of a COA in the investigational new drug/ new drug application/ biological license application path, and provide clarity around definitions of important terms (e.g., concept of interest, context of use, treatment benefit).  Participants will be encouraged to contribute during a facilitated discussion.

Use of Real World Data

5:00 PM - 6:00 PM
Room: Room 520be (5th Floor)

W6: A REALISTIC APPROACH TO WORKING WITH ONCOLOGY ELECTRONIC MEDICAL RECORD (EMR) DATA IN OUTCOMES RESEARCH

Discussion Leaders:

Kathy Foley, PhD, Senior Director, Strategic Consulting, Life Sciences, Truven Health Analytics, Cambridge, MA, USA

Katherine B Winfree, PhD, MPH, Senior Research Scientist, Eli Lilly and Company, Indianapolis, IN, USA

Claudio Faria, PharmD, MPH, Director, Health Economics & Outcomes Research, Eisai, Inc., Woodcliff Lake, NJ, USA

Leigh G. Hansen, MS, MBA, Vice President, Market Planning & Strategy - Life Sciences, Truven Health Analytics, Northwood, NH, USA

PURPOSE:

The purpose of this workshop is to describe key lessons learned from working with cancer- specific electronic medical records (EMR) and to identify the best uses of EMR data in oncology outcomes research.

DESCRIPTION:

The use of oncology-specific EMR data for outcomes research is rapidly growing.   As drug development increasingly focuses on personalized medicines, the need for data that include biomarkers and clinical details on disease state and progression are becoming increasingly important for oncology outcomes research. As a result, oncology EMR vendors and networks are making their data available for research. However, as with any data, there are pros and cons to working with EMR.   This workshop will provide an overview of key challenges in working with oncology EMR, recommendations for addressing those challenges, and suggestions as to the research questions best answered with oncology EMR data.  Dr. Faria will present on identification of biomarker status and adverse events.  Dr. Winfree’s presentation will focus on the identification of disease progression status and survival endpoints.    Dr. Foley will discuss the identification of relapse/refractory status among patients with hematological cancers. After reviewing the different topics, the presenters will discuss the best uses of EMR data for oncology research. Ms. Hansen will moderate the session.  The audience will be asked to provide additional examples of challenges and solutions to working with EMR data and will also be encouraged to ask questions about particular challenges they face with oncology EMR data.


5:00 PM - 6:00 PM
Room: Room 520cf (5th Floor)

W7: ESTIMATING HETEROGENEITY OF TREATMENT EFFECTS USING DECOMPOSITION TECHNIQUES WITH OBSERVATIONAL DATA: FROM IVORY TOWER TO THE REAL WORLD

Discussion Leaders:

Lee Brekke, PhD, Director, Health Economics and Outcomes Research, OptumInsight, Eden Prairie, MN, USA

Michael Grabner, PhD, Research Manager, Health Economics & Outcomes Research, HealthCore, Inc., Wilmington, DE, USA

Wenhui Wei, PhD, MS, MBA, Director, Evidence Based Medicine, Sanofi U.S., Bridgewater, NJ, USA

PURPOSE:

This workshop will introduce the use of decomposition techniques to estimate heterogeneity of treatment effects from a theoretical standpoint and with a practical example in which patients with type 2 diabetes from two large US health plans who initiated different injectable therapies were studied using their linked administrative claims and medical records data.

DESCRIPTION:

With the wide variety of therapies available to treat common chronic diseases and the limitations of clinical randomized controlled trials, comparative effectiveness research (CER) using observational data is increasingly being used to identify the most effective treatments. A key component of CER is to understand the heterogeneity of treatment response, i.e. to determine the characteristics that make a patient more likely to respond to a given therapy. Decomposition methods separate the treatment difference between two groups into a part that is “explained” by group differences in characteristics, such as demographics or comorbidities, and a residual part that can be used as a measure of the overall treatment effect. Detailed decomposition can be used to allocate the individual contributions of each predictor to each of the components of the decomposition, thus providing a measure of the heterogeneity of the treatment effect and serving as a basis for a personalized medicine approach. In this workshop the audience will not only be introduced to the theoretical framework of the decomposition technique, but also be able to see how this technique is applied in a practical example of the INITIATOR study, which is a large observational study of patients with type 2 diabetes using data from two large US health plans who initiated either insulin glargine via disposable pen or liraglutide.  Methodological and other issues associated with the use of these methods will be discussed. Audience participation will be encouraged through the use of this real-world example.

Tuesday, June 3, 2014
3:45 PM - 4:45 PM
WORKSHOPS - SESSION II
Clinical Outcomes Research

3:45 PM - 4:45 PM
Room: Room 518abc (5th Floor)

W8: MULTIVARIATE META-ANALYSIS: USE AND APPLICATIONS

Discussion Leaders:

Sonya J Snedecor, PhD, Director, Health Economics, Pharmerit International, Bethesda, MD, USA

Yin Wan, MS, Associate Scientist, Pharmerit International, Bethesda, MD, USA

Joseph C. Cappelleri, PhD, MPH, MS, Senior Director, Biostatistics, Pfizer Inc, Groton, CT, USA

John William Stevens, PhD, Statistician, School of Health and Related Research (ScHARR), University of Sheffield, Sheffield, UK

PURPOSE:

Introduce methodology and application of multivariate meta-analysis.

DESCRIPTION:

Pairwise multivariate meta-analysis (PMVMA) simultaneously synthesizes evidence of multiple outcomes reported in different clinical trials.  PMVMA is useful as it accounts for correlations among the outcomes which are ignored when performing meta-analyses on each outcome separately.  Typical applications of PMVMAs have been in the analysis of multiple correlated outcomes within a trial, such as overall survival and disease-free survival.  Advantages of PMVMA include generating the joint distribution of treatment effects within a single modeling framework, borrowing strength about treatment effects when there are missing outcomes in some studies, and smaller variances of treatment effects for each outcome.  MVMAs can be challenging as they are more complex and more difficult to execute.  Additionally, data required for MVMA, such as correlations between outcomes and between studies are often not reported.  There are several options for fitting multivariate models, including maximum likelihood, multivariate method of moments, generalized least squares, and Bayesian methods (requiring priors on all parameters, including covariances). This workshop will introduce the audience to the concepts of MVMA and methods for their execution/implementation, including strategies for overcoming the lack of reporting of within or between-study correlation data and using MVMA results in the context of health economic evaluations.  Bivariate pairwise meta-analysis models (of two outcomes) will be demonstrated in SAS and WinBUGS and compared to independent outcome analyses.  Tri-variate (three outcomes) pairwise meta-analysis methods will also be introduced.  By the end of this workshop, participants will be able to perform an example analysis in SAS and WinBUGS, and will be invited to share their experiences.

Economic Outcomes Research

3:45 PM - 4:45 PM
Room: Room 519ab (5th Floor)

W9: MODELLING ISSUES IN HIV

Discussion Leaders:

Nicolas Despiégel, MSc, Associate Director, Health Economics & Outcomes Research, OptumInsight, Nanterre, France

Milton Weinstein, PhD, Professor, Department of Health Policy and Management, Harvard School of Public Health, Boston, MA, USA

Miranda Murray, PhD, HO Lead, ViiV Healthcare, Brentford, UK

Rodrigo Refoios Camejo, PhD, Director, Value Evidence & Outcomes, GlaxoSmithKline, Brentford, UK

PURPOSE:

To discuss issues encountered when modelling HIV; provide solutions of how these can be addressed.

DESCRIPTION:

In recent years HIV infection has transformed from an acute life threatening illness to a chronic condition, managed with a combination of highly active drugs. Most regimens are extremely effective, but need to be taken long-term and have side-effects.  Consequently, effective treatment is less limited by efficacy than by poor adherence to prescribed medicines and incomplete retention in care. Moreover, patients who take their medications can expect to live long lives, meaning that comorbidities and competing risks associated with aging play an increasing role in the care – and costs of care - of HIV patients. Competing risks are especially important among HIV patients whose risk characteristics predispose them to an increased incidence of chronic conditions (cardiovascular disease, liver disease, cancer) and morbidity/mortality associated with risky sexual behaviour and substance abuse.  These changes mean that the modelling of HIV has needed to change to include these new drivers of clinical outcomes and costs. This workshop will identify the necessary modifications to accommodate these changes and discuss matters such as: the relevance of integrating long-term non-AIDS disease (most patients die of non-AIDS-related causes); how to define and include lifetime treatment algorithms and disease management strategies (treatment choice is very individualised; depending on many factors including treatment history, tolerance of side-effects); how to extrapolate beyond the clinical trial duration (given that patients who fail to adhere and develop resistance may be treated with subsequent treatment lines); how to integrate adherence to treatment, given that lack of compliance may lead to resistance and virologic failure, affecting different treatments in different ways; finally how to reflect the competing risks and costs of chronic diseases associated with aging. The audience will be asked to comment on the solutions suggested.

Health Care Policy Development Using Outcomes Research

3:45 PM - 4:45 PM
Room: Room 710b (7th Floor)

W10: IDENTIFYING HIGH-VALUE TREATMENT POPULATIONS IN CLINICAL TRIALS AND REAL-WORLD DATA

Discussion Leaders:

James Signorovitch, PhD, Vice President, Analysis Group, Inc., Boston, MA, USA

James Shaw, PhD, Associate Director, Cardiovascular/Metabolics, Global Health Economics and Outcomes Research, Bristol-Myers Squibb, Princeton, NJ, USA

Keith A. Betts, PhD, Manager, Analysis Group, Inc., Boston, MA, USA

Eric Q. Wu, PhD, Managing Principal, Analysis Group, Inc., Boston, MA, USA

PURPOSE:

As innovative treatments enter crowded markets with limited resources, their use is often limited to a portion of the population.  Ideally, the patients who receive the treatment should all have a higher expected benefit than the patients who do not – that is, the treatment should be allocated efficiently.  However, in practice, the evidence base for deciding who to treat is usually limited, which may result in inefficient treatment allocation.  This, in turn, results in missed opportunities for clinical benefits to patients and missed economic benefits for payers, manufacturers and society.

DESCRIPTION:

Through multiple real-world applications, this workshop will illustrate how to generate and interpret evidence for patient-specific treatment effects.  In addition, we will show how to estimate the efficiency frontier for treatment allocation within a population, i.e., the greatest level of health benefit that can be achieved for a given cost via optimal treatment allocation among patients.  We will provide intuitive descriptions of the recently-developed statistical methods that enable estimation of efficiency frontiers and population-specific treatment effects, and will explain how these methods avoid the low power and potential for cherry-picking inherent in traditional subpopulation analyses.  Real-data examples in clinical trials will include the identification of high value sub-populations that maximize the population size in which a treatment can achieve a given incremental cost-effectiveness ratio.  In addition, we will show how to measure the efficiency, or inefficiency, of subpopulations proposed by a health technology assessment authority (e.g., restricting reimbursement to patients exceeding a disease severity threshold or failing a prior treatment).  Using real-world data, we will give examples in which actual use of a treatment is inefficient, and there are potential opportunities to improve population-level health outcomes, reduce costs, or both. The audience will participate through discussion of real-world examples and from the methodological and decision-maker perspectives.


3:45 PM - 4:45 PM
Room: Room 710a (7th Floor)

W11: HOW CAN WE USE RANDOMIZED TRIAL DATA TO ASSESS HETEROGENEITY OF TREATMENT EFFECTS? LET ME COUNT THE WAYS

Discussion Leaders:

Robert W. Dubois, MD, PhD, Chief Scientist, National Pharmaceutical Council, Washington, DC, USA

C Daniel Mullins, PhD, Professor, Department of Pharmaceutical Health Services Research, University of Maryland School of Pharmacy, Baltimore, MD, USA

David M Kent, MD, MS, Professor of Medicine, Predictive Analytics and Comparative Effectiveness Center, Institute for Clinical Research and Health Policy Studies, Tufts Medical Center/Tufts University School of Medicine, Boston, MA, USA

Richard Willke, PhD, Group Lead, Outcomes & Evidence, Pfizer Inc., Peapack, NJ, USA

PURPOSE:

A variety of methods are available for exploring heterogeneity of treatment effect (HTE) from both prospectively designed studies aimed to test for HTE as well as retrospective analysis of individual or published trial data.  This workshop will discuss a range of research approaches for evaluating HTE, and will demonstrate a selected set of these approaches including:  1) Systematic review of clinical trial reports and other sources;  2) Repeated period cross-over studies and series of n-of-1 trials; and 3) Various types of individual patient data and meta-analysis.   By delineating the strengths and limitations of these methods, this workshop is meant to better enable outcomes researchers to understand aspects of HTE in the context of patient-centered outcomes research.

DESCRIPTION:

Estimation procedures for identifying or testing for HTE may not be feasible or practical within regulatory clinical trials. However, alternative approaches are not always sufficiently robust to justify treatment predictions for subgroups or individuals.  This session will discuss research methods for examining HTE based upon the intent of the study (e.g. exploratory vs. confirmatory).  Dr. Dubois will moderate the session and provide an overview of methodological approaches to the estimation of HTE across study types.  Dr. Mullins will demonstrate the potential usefulness and challenges associated with identifying evidence of HTE from systematic literature reviews, using stage-4 prostate cancer as an example to explore patient-level specific HTE factors.  Dr. Kent will discuss the role of repeated cross-over studies and n-of-1 trials on HTE, and demonstrate the use of these methods in the examination of HTE in selected case examples.  Dr.  Willke will discuss selected analytic methods such as quantile regression, finite mixture modeling, and meta-regression.  The audience will be asked to help identify key strengths and limitations of these methodologies, and to share their views on how these methods can best be applied.

Patient-Reported Outcomes & Patient Preference Research

3:45 PM - 4:45 PM
Room: Room 520ad (5th Floor)

W12: A FRAMEWORK FOR CONDUCTING INITIAL MEDICATION ADHERENCE RESEARCH

Discussion Leaders:

David Hutchins, MBA, Executive Advisor, Caremark, Scottsdale, AZ, USA

Craig S Roberts, PharmD, MBA, Senior Director, Outcomes Research, Pfizer Inc., New York, NY, USA

John E Zeber, PhD, MHA, Investigator & Associate Professor, Center for Applied Health Research, Scott & White Healthcare, Temple, TX, USA

Andrew M. Peterson, PharmD, PhD, Dean, Mayes College of Healthcare Business and Policy, University of the Sciences, Philadelphia, PA, USA

PURPOSE:

 To provide guidance on measuring initial medication adherence (IMA), including developing standard nomenclature and key components of quality IMA research. This workshop will be useful to a diverse audience of investigators, health system practitioners, patients, and policy makers

DESCRIPTION:

Medication adherence is a significant health care issue improving clinical symptoms, lowering mortality, and increasing health care cost and utilization.  It is composed of ongoing and initial medication adherence. Ongoing medication adherence measures what happens after an initial prescription is filled and has been the focus of most research associating adherence with beneficial factors.  Initial medication adherence measures filling rates of initial prescriptions within a therapeutic class encompassing both un-presented prescriptions, and unclaimed prescriptions. Researchers have employed numerous methods and used various terms for measuring IMA, some relying only  on un-presented or unclaimed prescriptions. Both provide information on IMA and both are necessary for a complete IMA measurement. Based upon a review of published initial adherence research, this ISPOR Working Group has identified facilitators and barriers to measuring IMA such as (patient, pharmacist, prescriber, and system), study design parameters (core, supplemental, and ancillary), and the definition of IMA.  A set of recommendations to guide future investigators into producing high quality initial adherence research will also be presented. The first 30 minutes will include a discussion of ISPOR Working Group’s observations of terminology, perspective and definitions of IMA, audience interaction (brainstorming/audience polling) to discuss terms used for IMA by audience members and discussion of definitions-consensus development, and a discussion on the state of current research.  The next 25 minutes include a review of recommended framework for conducting IMA studies, review of the core, supplemental and ancillary parameters associated with measuring IMA, and audience interaction (open discussion).  The workshop will conclude with a discussion on the recommended final steps.

Use of Real World Data

3:45 PM - 4:45 PM
Room: Room 520be (5th Floor)

W13: EVIDENCE SYNTHESIS AND DECISION MAKING WHEN RANDOMISED EVIDENCE DOES NOT SUFFICE

Discussion Leaders:

Doug Coyle, PhD, Professor, Epidemiology and Community Medicine, University of Ottawa, Ottawa, ON, Canada

Karen M Lee, MA, Director, Health Economics, Canadian Agency for Drugs and Technologies in Health, Ottawa, ON, Canada

Brian Hutton, PhD, Adjunct Professor, Faculty of Health Sciences, University of Ottawa, Ottawa, ON, Canada

Kristian Thorlund, PhD, MSc, Director, Redwood Outcomes, Vancouver, BC, Canada

PURPOSE:

To discuss issues when, in the reimbursement process, randomized clinical evidence is not available or highly sparse. To highlighted examples where decision making had to consider the relevance of evidence sources beyond randomized clinical trials. To illustrate possible solutions and discuss their pros and cons. To have participants engage in discussion around other methods that might be used to address these limitations

DESCRIPTION:

Lately, the clinical questions that need to be addressed in the context of health technology assessment (HTA) and decision making are often not addressed by randomized clinical trials. Further, the first randomized clinical trials (RCTs) for new therapeutic agents or the use of known therapeutic agents for new indications rarely provide adequate information for addressing the evidence needs of health care decision makers. Meta-analysis and network meta-analysis are frequently used to synthesize findings of multiple RCTs of the interventions of interest to inform decision making. However, RCTs often only represent a small subset of the available evidence.  Therefore, there is the potential to expand the evidence-base beyond that of RCTs. This workshop will focus on particular examples from the Canadian setting where the RCT evidence was insufficient to form a basis for recommendations. Examples will cover HTAs where the following sources of non-RCT evidence could be considered 1) observational efficacy and safety data; 2) non-comparative clinical trials; and 3) long-term observational data. Statistical and epidemiological solutions will be presented for incorporating each of these three types of data sources, and their particular merits and limitations for the given examples will be discussed. However, the potential limitations of such approaches will be discussed. Participants will comment on the solutions and be encouraged to bring forward other options for discussion.


3:45 PM - 4:45 PM
Room: Room 520cf (5th Floor)

W14: SENSITIVITY OF ESTIMATED TREATMENT EFFECTS FROM PROPENSITY SCORE MATCHING ANALYSES

Discussion Leaders:

Bijan J. Borah, PhD, Assistant Professor, Mayo College of Medicine, Department of Health Sciences, Division of Health Care Policy and Research, Mayo Clinic, Rochester, MN, USA

Henry Henk, PhD, Senior Fellow, OptumInsight, Eden Prairie, MN, USA

William H. Crown, PhD, Chief Scientific Officer, Optum Labs, Cambridge, MA, USA

PURPOSE:

Provide a tutorial on how sensitivity analyses of the estimated treatment effects using propensity score matching (PSM) can be implemented.  

DESCRIPTION:

Propensity score matching (PSM) has been widely used in observational comparative effectiveness research (O-CER) as a tool for estimating treatment effect of an intervention compared to alternative intervention(s). The validity of PSM-based treatment effect estimates relies on the untestable assumption of conditional independence, that is, treatment assignment is independent of outcomes conditional on observed covariates. However, when unobserved factors impact both treatment assignment and outcomes simultaneously, the estimated treatment effects from a PSM analysis become questionable.  Rosenbaum bounds approach is a way to assess the sensitivity of a potential unmeasured confounder on the probability of receiving the treatment (intervention) so as to invalidate the estimated treatment effect. However, this important aspect of the O-CER has largely been ignored in the literature, potentially due to lack of implementation tools. The purpose of this workshop is to demonstrate and provide a tutorial on how sensitivity analyses of the estimated treatment effects using PSM can be implemented using existing software (Stata and/or R).  Particular focus will be placed on interpreting sensitivity analysis results. The workshop will be interactive in nature that will encourage audience participation through two worked-out examples. Participants will be asked to share the specific research questions that they would likely implement in this sensitivity analysis. The two examples will use claims data from large commercial payers in the United States. The examples will consider both binary and continuous outcomes. By the end of the workshop, it expected that the participants will be fairly comfortable with the tools to conduct sensitivity analyses of the estimated treated effects from PSM, thereby providing greater confidence to the decision maker who would make decisions around the intervention(s) being compared.

5:00 PM - 6:00 PM
WORKSHOPS - SESSION III
Clinical Outcomes Research

5:00 PM - 6:00 PM
Room: Room 710b (7th Floor)

W15: ENHANCING META-ANALYSIS BY CONSIDERING THE CORRELATION BETWEEN TWO OUTCOMES

Discussion Leaders:

Julie Roïz, MSc, Director, UK Operations, Creativ-Ceutical, London, UK

Julie Dorey, MSc, Manager, US Operations, Creativ-Ceutical USA, Chicago, IL, USA

Anne-Lise Vataire, PhD, Lead analyst, Creativ-Ceutical, Paris, France

PURPOSE:

Several methodological articles on bivariate meta-analysis have been published, but this technique remains rarely used although many trials include several correlated outcomes. Taking into account the correlation between outcomes in a meta-analysis offers potential to improve estimates by borrowing strength across outcomes and reducing potential bias when one outcome is not reported in all studies, e.g. in presence of selective reporting. This workshop will present methods for bivariate meta-analysis and review criteria to take into consideration when choosing a meta-analysis method.

DESCRIPTION:

We will provide an overview of published models for meta-analysis with two endpoints, including fixed-effect and random-effect model. Different ways of specifying the correlation between outcomes have been proposed. The empirical correlation between outcomes will be presented for several published meta-analyses, in different therapeutic areas. The correlation between progression-free survival and overall survival in oncology will be studied in particular. The concepts of between-studies correlation and within-studies correlation will be explained. The relationship between variances for the two endpoints will be described using some examples and the role of this relationship in bivariate meta-analysis will be presented. We will then introduce a bivariate method using an imputation of the variance for studies for which one endpoint is not reported. The results of bivariate meta-analysis models and standard univariate models will be presented and compared in several cases, for different numbers of studies, different levels of heterogeneity between studies, different values of correlation between endpoints, and with missing values at random or not at random. Finally, we will discuss ways to incorporate results of bivariate meta-analysis in cost-effectiveness models. Participants will be encouraged to comment on applicability and usefulness of these methods in different therapeutic areas.

Economic Outcomes Research

5:00 PM - 6:00 PM
Room: Room 519ab (5th Floor)

W16: CHANGING CYCLE LENGTHS IN MARKOV MODELS: DOING IT THE RIGHT WAY

Discussion Leaders:

Jagpreet Chhatwal, PhD, Assistant Professor, Health Services Research, MD Anderson Cancer Center, Houston, TX, USA

Elamin H Elbasha, PhD, Distinguished Scientist, Merck Sharp & Dohme Corp., Whitehouse Station, NJ, USA

PURPOSE:

Markov models are widely used for cost-effectiveness analysis and medical decision making because of their simplicity. In these models, time advances in discrete time-steps (i.e. cycles), and only a single transition can occur within a cycle based on a predefined transition probability. These probabilities are often defined over longer time intervals and need to be converted into desired shorter length probabilities.  The purpose of this workshop is to: (1) show that the commonly used formula of changing probabilities to different cycle lengths is incorrect; and (2) provide the correct approach.

DESCRIPTION:

The ISPOR-SMDM Modeling Good Research Practices Task Force report recommends that the cycle length should be short enough to represent the frequency of clinical events and interventions. The commonly used approach of converting transition probabilities to a different cycle length in Markov models is to first convert transition probabilities into rates, divide rates according to the new cycle length, and convert rates back into new probabilities. However, this approach is not applicable to models with more than two states. This workshop will highlight the bias introduced by using the incorrect approach in the outcomes of Markov models. Further, we will provide the correct method based on taking the roots of matrices to change cycle lengths. Using several examples, we will show bias in the transition probabilities, costs, and quality-adjusted life years with incorrect approach. Contrary to the common belief, we show that the bias increases with shorter cycle lengths when commonly used approach is used. The workshop will be interactive and also provide participants a stand-alone tool to perform conversions correctly.

Health Care Policy Development Using Outcomes Research

5:00 PM - 6:00 PM
Room: Room 710a (7th Floor)

W17: UNDERSTANDING AND MODELING BUSINESS DECISIONS IN MARKET ACCESS AND REIMBURSEMENT USING MULTI-CRITERIA DECISION ANALYSIS TECHNIQUES

Discussion Leaders:

Maarten J. IJzerman, PhD, Professor & Chair, Department of Health Technology & Services Research, University of Twente, Enschede, The Netherlands

Kevin Marsh, PhD, Director, Modelling and Simulation and Senior Research Scientist, Health Economics, Evidera, London, UK

Ansgar Hebborn, PhD, Head, Global Market Access Policy, F. Hoffmann-La Roche AG, Basel, Switzerland

Tereza Lanitis, MSc, Senior Research Associate, Evidera, London, UK

PURPOSE:

To introduce MCDA as a method to support business decisions and to discuss the opportunities of applying it to model payer’s decisions

DESCRIPTION:

Multi-criteria decision analysis (MCDA) in supporting health care decision making has been increasingly recognized. Various techniques for MCDA exist and their predictive validity has been assessed. Yet, confusion exists for what purpose MCDA can be used. Previous ISPOR sessions have asked whether MCDA can replace cost-effectiveness analysis or discussed its role in benefit-risk assessment. In these applications it is assumed MCDA is an appropriate way to obtain an aggregate function of benefit considering multiple therapeutic outcomes. However, another widely used application of MCDA is to support the strategic and business decision-making processes. This workshop will therefore illustrate how MCDA may be used for business decisions with an emphasis on market access and reimbursement decisions. The workshop starts with an introduction to MCDA and review of its use in health care by Tereza Lanitis, including examples of studies investigating the use of MCDA being used to support business decisions. This will be followed by Maarten IJzerman speaking about the methodological issues of using MCDA for modeling business decisions in general and reimbursement decisions in particular. His presentation will discuss the set of criteria relevant for such decisions and methods to deal with intangible criteria.  The final presentation by Ansgar Hebborn deals with the companies’ internal decision-making processes. He will focus on business decisions with an emphasis on the modeling of external decision-maker preferences to support pricing, reimbursement, and funding decisions. He will discuss some of the challenges of adopting an MCDA framework, including the difficulty synthesizing MCDA outcomes across multiple countries.  The audience will be invited to share their experiences of using MCDA to support business decisions and their thoughts of the feasibility of doing so.

Patient-Reported Outcomes & Patient Preference Research

5:00 PM - 6:00 PM
Room: Room 518abc (5th Floor)

W18: PATIENT-REPORTED OUTCOMES (PROs) – USE IN ELECTRONIC MEDICAL RECORDS (EMR) AND IMPLICATIONS FOR COMPARATIVE EFFECTIVENESS RESEARCH (CER)

Discussion Leaders:

Tara Symonds, PhD, Head of PRO Center of Excellence, Pfizer Ltd., Tadworth, Surrey, UK

Bryce B. Reeve, PhD, Associate Professor, Lineberger Comprehensive Cancer Center & Department of Health Policy and Management, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Joseph C Cappelleri, PhD, MPH, MS, Senior Director, Biostatistics, Pfizer Inc., Groton, CT, USA

I Katzan, MD, Director, Neurological Institute Center for Outcomes Research & Evaluation, Cleveland Clinic, OH, USA

PURPOSE:

With the formation of PCORI (Patient-Centered Outcomes Research Institute) bringing emphasis to the relevance of patient-centered outcomes in health care research, specifically CER, there has been an upsurge in systems to capture PRO data within EMRs. Further incentive for integrating PROs within EMRs comes from the recent Meaningful Use criteria from the Office of the National Coordinator for Health Information Technology.  The purpose of this workshop is to identify some of the potential challenges in collecting patient-reported data within an EMR system and to discuss practical solutions for successful use.

DESCRIPTION:

A blind plunge into using PRO assessments in EMR systems to conduct real-world comparative studies, without consideration and action to prevent methodological or operational challenges, may lead to “garbage in, garbage out” and hence lead to questionable results and wasted resources.  A priori plans for how to deal with PRO data is becoming increasingly important. The discussion leaders will outline and provide practical applications of some of the methodological issues that need to be addressed such as validity and reliability, interpretation of the data, response shift, and missing data.  Speakers will debate use of computer-adaptive testing to address patient burden, a significant concern which may limit researchers’ ability to assess key patient-centered outcomes.  Finally, the operational aspects of developing an EMR system that captures PRO data will be discussed.  These practical issues will address how often data should be collected; how it should be compiled, stored, and accessed for analysis; how to set up the data capture with meta-data fields; and how to implement  site and patient training.  Panelists and the audience will hold a discussion on identifying critical gaps and needs for better integration of PRO data within EMRs for the purposes of enhancing patient care and informing CER studies.


5:00 PM - 6:00 PM
Room: Room 520ad (5th Floor)

W19: CHOICE DEFINES VALUE: TURNING PREFERENCES OF MULTIPLE STAKEHOLDERS INTO EVIDENCE FOR HEALTH CARE DECISION MAKING

Discussion Leaders:

Axel C. Mühlbacher, PhD, MBA, Professor of Health Economics and Health Care Management, IGM Institute Health Economics and Health Care Management, Hochschule Neubrandenburg, Neubrandenburg, Germany

Benjamin M Craig, PhD, Associate Member, Health Outcomes & Behavior, Moffitt Cancer Center, Tampa, FL, USA

Juan Marcos González, PhD, Senior Research Economist, Health Preference Assessment, RTI Health Solutions, Research Triangle Park, NC, USA

PURPOSE:

This workshop will show how discrete choice experiments (DCEs) can guide regulatory decisions in light of multiple stakeholders and reflects increased interest in transparency and inclusiveness in health care decision making. By integrating stakeholder preferences, DCEs enhance the capabilities of regulatory bodies to systematically inform and communicate their decisions.

DESCRIPTION:

While patients are considered the primary stakeholders in health care, regulators determine the use of limited resources to satisfy competing health care needs (i.e., infrastructure development, payments, and investments in innovation.) Regulators’ decisions rely informally on more than just outcomes data. Regulators weigh the expected benefits and harms of treatments as represented by the clinical evidence at hand. To achieve transparent weighing of evidence and allocative efficiency, knowing stakeholders’ preferences is critical, particularly when regulators make decisions regarding the adoption of new technologies when little information about the value of their impact to stakeholders is available. This workshop is devoted to how DCEs can inform budgetary and regulatory decisions by systematically measuring stakeholder preferences using real-world examples. Specifically, this workshop discusses the following questions: 1) What are the patient-relevant decision criteria? How can these endpoints be identified?; 2) How do these decision criteria contribute to patients’ treatment benefits? How should these criteria be weighted when making health care decisions?; 3) What is equivalent to the patients’ value or benefit? How can a unique welfare metric be derived based on multiple decision criteria? The workshop will include an interactive discussion with attendees integrating three well-established approaches (health valuation, conjoint, and multi-criteria decision analyses). Attendees will receive a worksheet presenting how stated choices can be used to collect evidence on stakeholder’s preferences that can aid and support difficult health care decisions. Through this introduction, attendees will be able to investigate applications of DCEs to answer challenging questions in budgetary and regulatory decision making.

Use of Real World Data

5:00 PM - 6:00 PM
Room: Room 520be (5th Floor)

W20: EXPLORING AND VALIDATING ALTERNATIVE METHODS OF USING MULTIPLE DATABASES TO ANSWER COMPLEX LONGITUDINAL RESEARCH QUESTIONS: IS LINKING DATABASES THE ONLY ANSWER?

Discussion Leaders:

Rolin L Wade, RPh, MS, Principal, Health Economics and Outcomes Research, Real World Evidence Solutions, IMS Health, Parsippany, NJ, USA

David Macarios, MBA, Vice President, Health Economics and Outcomes Research, LifeCell Corporation, Bridgewater, NJ, USA

PURPOSE:

The purpose of this workshop is to engage the audience in exploring innovative methods of utilizing multiple retrospective databases to answer research questions where the use of a single database may fall short to provide complete answers.  This workshop will focus on the implementation and validation of alternative techniques for answering research questions utilizing multiple retrospective databases.

DESCRIPTION:

Considerable excitement has been generated over the last few years regarding the strength of linking disparate databases with one another, such as linking electronic medical records with claims data. While these methods are proving to be valuable in many situations, they bring a set of limitations that is unique to these linked data assets. There are methodological alternatives when choosing to combine the power of multiple data sets, and this workshop will focus on implementing and comparing two distinct methods. The first method will implement direct linking of patients from one database to the other; the second will involve matching patients through multiple variables across both databases. As a case study, we will utilize these two methods using an academic hospital based patient registry along with a large administrative claims database, and will explore a surgical intervention where clinical detail regarding the hospitalization as well as the longitudinal post-discharge follow-up is important to measuring important outcomes. As the second method (population matched on multiple variables) serves as a proxy to the directly linked population, we will explore with the audience methods of validating the precision of the estimates in the matched population using the linked data as a reference.


5:00 PM - 6:00 PM
Room: Room 520cf (5th Floor)

W21: DOES CONFOUNDING IN COMPARATIVE EFFECTIVENESS ASSESSMENT ALWAYS MATTER FOR PAYERS’ DECISIONS?

Discussion Leaders:

Billy Amzal, X-Eng, MSc, MPA, PhD, Senior Scientific Vice President, LASER Analytica, London, UK

Lamiae Grimaldi, PharmD, MSc, PhD, Senior Scientific Vice President, LASER Research, Paris, France

Yola Moride, PhD, Professor of Pharmacoepidemiology, Faculty of Pharmacy, Université de Montréal, Montreal, QC, Canada

Lucien Abenhaim, MD, PhD, Honorary Professor, London School of Hygiene and Tropical Medicine, London, UK

PURPOSE:

To assess and discuss to what extent and when adjusting for confounding factors is relevant (or not) in the context of comparative effectiveness research to support payers and reimbursement decisions.

DESCRIPTION:

Comparative effectiveness research has become in most western countries a critical step to support pricing and reimbursement decisions. Policy makers, payers, and care providers are looking into pragmatic measures of the health and economic impact of a drug using data collected from real-life practice. Pragmatic trials and observational studies are often being perceived as lower-quality information, as many real-world conditions may confound the results such as patient risk factors, concomitant prescriptions or patient compliance may be unbalanced initially or over time across comparators, potentially modifying or interacting with the theoretical effect of a therapy as estimated by clinical trials. Methods are frequently proposed to eliminate these biases (such as propensity scores or modelling), increasing very significantly the internal validity of real life studies and their acceptability. However, the paradigm of payers’ evaluation is in essence very different from regulators’ or researchers’ ones. Payers’ evaluations of drugs are seeking for pragmatic information that can support their decisions while regulators and drug developers generate evidence to establish knowledge on the relative effect of a therapy versus others. A very frequent issue is to understand how the effect varies across populations and is influenced by so-called ‘real life’ factors. As a consequence, controlling or not for confounding and real life factors may or may not be relevant, depending on the decision at hand. Through decisive case studies of comparative effectiveness research, the relevance and methods of adjusting for confounding will be reviewed and discussed. Implications on real-world evidence generation for market access and the related decisions will be drawn.

Wednesday, June 4, 2014
1:45 PM - 2:45 PM
WORKSHOPS - SESSION IV
Clinical Outcomes Research

1:45 PM - 2:45 PM
Room: Room 518abc (5th Floor)

W22: METHODOLOGICAL CHOICES FOR PROPENSITY SCORE MATCHING AND THEIR CONSEQUENCES

Discussion Leaders:

Samuel Aballéa, PhD, Vice President, Health Economics & Outcomes Research, Creativ-Ceutical, Paris, France

Firas M. Dabbous, MS, PhD candidate, Public Health and Epidemiology, University of Illinois at Chicago, Chicago, IL, USA

Katia Thokagevistk, MSc, Senior Analyst, Health Economics & Outcomes Research, Creativ-Ceutical USA, Chicago, IL, USA

PURPOSE:

Propensity score matching (PSM) is now a widely used method in comparative effectiveness research to reduce or eliminate confounding bias in observational studies. The purpose of this workshop is to provide some guidance on how to implement PSM approaches. Methodological choices that have to be made when using this approach will be reviewed and innovative methods for implementing PSM will be presented.

DESCRIPTION:

The workshop will start with some reminders about the PSM method, and when it should be used rather than regression analysis, simple matching methods, or other methods based on propensity scores, such as stratification or inverse probability weighting. Logistic regression is generally used to develop propensity scores, but other methods such as boosted classification and regression trees (CART) have been proposed. Another important consideration for developing the scores is the selection of variables. Common matching algorithms will be presented and compared, i.e. the nearest-neighbor matching, the Radius matching (i.e. ‘caliper matching’), and the Kernel matching, and the choice between matching with or without replacement will be discussed. Most applications use a one-to-one matching ratio, but increasing the matching ratio may improve precision at small cost in bias. When using one-to-many matching approaches, the analyst has to choose between parallel and sequential selection, and fixed or variable ratio. Techniques for diagnosing and correcting covariates balance will be introduced. Finally, features of PSM will be put in perspective with the non-parametric method of Genetic Matching. Throughout the workshop, simulated datasets will be used to illustrate the impact of different methodological choices on covariate balance, mean bias, and mean variance, and stimulate discussion with the audience.

Economic Outcomes Research

1:45 PM - 2:45 PM
Room: Room 519ab (5th Floor)

W23: AN ALTERNATIVE TO HAZARDOUS HAZARDS ASSUMPTIONS: EVIDENCE SYNTHESIS FOR SURVIVAL OUTCOMES IN AN ECONOMIC EVALUATION OF AN ONCOLOGY DRUG

Discussion Leaders:

Jeroen P Jansen, PhD, Director, Redwood Outcomes, San Francisco, CA, USA

Jeffrey S. Hoch, PhD, Associate Professor, Department of Health Policy, Management and Evaluation, Faculty of Medicine, University of Toronto, Toronto, ON, Canada

Christopher O'Regan, MSc, Head of Health Technology Assessment & Outcomes, Merck Sharp & Dohme Limited, Hertfordshire, UK

PURPOSE:

This workshop will demonstrate a novel technique to combine evidence from survival analyses used in an economic evaluation.  We will illustrate that the strength of this technique lies in the fact that it does not rely on the proportional hazards assumption and will connect this advantage with economic evaluation, noting that expected survival and QALY estimates are less likely to be over- or under-estimated than those obtained with the standard approach, resulting in more reliable estimates of cost effectiveness.

DESCRIPTION:

The proportional hazards assumption that underlies current approaches of evidence synthesis of survival outcomes is not only often implausible, but can have a huge impact on decisions based on cost-effectiveness analysis. Common practice is to assume a parametric survival function for the baseline intervention (e.g. Weibull) and apply the treatment specific constant hazard ratio obtained with the (network) meta-analysis to calculate a corresponding survival function enabling comparisons of expected survival. Since the tail of the survival function has a major impact on the expected survival, violations of the constant hazard ratio can lead to severely biased estimates of incremental effectiveness and cost-effectiveness. Hence, the proportional hazards assumption has become a source of concern in the creation of evidence to support drug reimbursement decisions. The workshop starts with a short overview of the standard practice of evidence synthesis and cost-effectiveness analysis for interventions that aim to improve survival, followed by an introduction of the alternative approach including an oncology drug case study. We will illustrate that digitally scanned published Kaplan-Meier curves can be synthesized with a (network) meta-analysis model and the resulting pooled (progression free) survival curves can directly be integrated with your Bugs or Excel-based economic model. The workshop will be concluded by considering the value of the new methods from a HTA decision-maker’s perspective.

Health Care Policy Development Using Outcomes Research

1:45 PM - 2:45 PM
Room: Room 520a (5th Floor)

W24: THE ROLE OF PATIENT-REPORTED OUTCOMES DATA IN HEALTH CARE DECISION MAKING IN RARE DISEASES

Discussion Leaders:

Margaret K. Vernon, PhD, Senior Research Scientist & EU Director, Outcomes Research, Evidera, London, UK

Philip Ruff, PhD, Director, Global Market Access, Shire, Lexington, MA, USA

Isabel Kalofonos, MBA, Director, Global Market Access, Shire, Lexington, MA, USA

Asha Hareendran, PhD, Senior Research Scientist, Outcomes Research, Evidera, London, UK

PURPOSE:

The objective of this workshop is to discuss the role of data that reflect the perspectives of patients in drug development programs for treatments for rare diseases to ensure optimization health care decision making.

DESCRIPTION:

Rare diseases are an expanding area of clinical development; the number of targeted treatments in the pipeline for rare diseases has nearly tripled compared with decade ago.  While both regulatory agencies (through accelerated approval opportunities) and payers (with reduced emphasis on traditional QALY approaches) recognize that these populations have unique unmet needs, the hurdles for undertaking clinical development in rare diseases are often challenging.  As most new treatments for rare diseases are costly, payers require a good understanding of the impact of diseases on patients’ lives and may like to see a comprehensive evaluation of the benefits and harms associated with these treatments to make coverage decisions, including benefits important to patients. Collecting these data in rare diseases is challenging given the heterogeneity of these conditions, limited information regarding impact of disease on patients’ lives, and dearth of well validated instruments to measure meaningful outcomes. Workshop leaders will provide examples of strategies and methods used to address these challenges in rare diseases like hereditary angioedema, cystic fibrosis, Hunter syndrome, Sanfilippo A and B, and metachromatic leukodystrophy. Examples will be shared about data that have influenced favorable regulatory and payer review. This includes use of patient data for primary efficacy endpoints for product registration, understanding of burden of illness for patients and caregivers, and methods for developing data on burden of illness and unmet need in the population. The audience will be involved in a discussion of these examples and be asked to share other examples for using PROs in registration and reimbursement, given the rapidly developing knowledge in the field of rare disease research.

Patient-Reported Outcomes & Patient Preference Research

1:45 PM - 2:45 PM
Room: Room 520b (5th Floor)

W25: GOOD PRACTICE GUIDELINES FOR QALYs: ISPOR'S DANGEROUS OMISSION?

Discussion Leaders:

Christopher McCabe, MSc, PhD, Professor, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada

John Brazier, PhD, Professor, Health Economics and Decision Science, School of Health and Related Research (ScHARR), University of Sheffield, Sheffield, UK

David M Meads, MSc, Lecturer in Health Economics, Academic Unit of Health Economics, University of Leeds, Leeds, UK

PURPOSE:

To debate the value of ISPOR producing a Good Practice Guide for the development and utilisation of QALY measures (preference based health related quality of life measures) in health outcomes research. 

DESCRIPTION:

The workshop will consist of three presentations followed by a facilitated discussion aimed at developing a proposal to ISPOR for a Good Practice Guide: 1) Role of QALYs in international HTA and evidence of variation in practice establishing the need for consensus on practice in the development and use of QALYs in HTA; 2) Key Issues in development and validation of QALY measures -  a survey key methodological challenges leading to priorities for achieving consensus in current practice and for future research; and 3) Key issues in the use of QALYs in CEA/HTA - a survey key methodological challenges in the use of utility data in HTA submissions leading to priorities for consensus on current practice and future research. The presentations will be followed by an open discussion aimed at outlining a work programme for the production of Good Practice Guide for the development and use of QALYs in HTA, with the support of key opinion leaders in the Pharmacoeconomics and Outcomes Research Community.

Use of Real World Data

1:45 PM - 2:45 PM
Room: Room 520c (5th Floor)

W26: BIG DATA AND LITTLE DISEASES: MEETING THE CHALLENGES FOR RARE DISEASE OUTCOMES RESEARCH

Discussion Leaders:

Alaa Hamed, MD, MPH, MBA, Director Rare Diseases, Genzyme POME, Genzyme, a Sanofi Company, Cambridge, MA, USA

Alexandra Ward, PhD, MRPharmS, Senior Research Scientist, Evidera, Lexington, MA, USA

Sumeet Panjabi, PhD, Director, Global Health Economics, Onyx Pharmaceuticals, Inc., an Amgen subsidiary, South San Francisco, CA, USA

Kathleen W. Wyrwich, PhD, Senior Research Leader, Evidera, Bethesda, MD, USA

PURPOSE:

This workshop will explore the challenges of health outcomes research in the rare disease arena, and present several creative solutions using medical databases and multimedia opportunities.

DESCRIPTION:

Strong retrospective and prospective evidence is necessary in all phases of clinical outcomes research, including the demonstration of the clinical, humanistic, and economic burden of disease; the determination of treatment patterns and associated adverse events; and the validation of clinical outcome assessments. By definition, however, a rare disease is shared by few individuals; and for many rare diseases, knowledgeable treatment is provided by a scarce number of qualified clinicians. Moreover, these rare conditions often do not have a specific code for accurate case identification in large medical databases; yet the accurate identification improves opportunities for greater understanding of the condition, symptoms, treatments, and ultimately outcomes. Although electronic medical records and the promise of expanded ICD-10 codes offer improved capabilities to find specific rare disease needles in large database haystacks, the process remains challenging. Enhanced social and multimedia networking, which connects specific rare disease sufferers and their clinicians across the globe to learn from and with each other, offer the promise of addressing some of these challenges, and play an important role in shared knowledge about these conditions and the development of sound and relevant rare disease endpoints. However, difficulties exist for making these key connections to meaningfully improve measurement and treatment in rare diseases. Embracing these opportunities and challenges, the speakers will interactively share their novel processes and solutions for improved rare disease patient identification in large medical claims databases, an understanding of patient-level treatment trends in key markets, and the use of multimedia in the development and validation of clinical outcome assessments for a rare disease. After these presentations, the audience will be invited to ask questions and exchange other useful strategies.

3:00 PM - 4:00 PM
WORKSHOPS - SESSION V
Economic Outcomes Research

3:00 PM - 4:00 PM
Room: Room 520c (5th Floor)

W27: EXPECTED VALUE OF PERFECT INFORMATION: ACTIVE LEARNING THROUGH USER-FRIENDLY COMPUTATIONS, DISPLAYS, AND APPLICATION DISCUSSIONS

Discussion Leaders:

R. Brett McQueen, PhD, Post Doctoral Fellow, School of Pharmacy, University of Colorado Anschutz Medical Campus, Aurora, CO, USA

Eldon Spackman, PhD, Research Fellow, Centre for Health Economics, The University of York, Heslington, York, UK

Jonathan D Campbell, PhD, Assistant Professor, School of Pharmacy, University of Colorado Anschutz Medical Campus, Aurora, CO, USA

PURPOSE:

Health economic studies that perform simulation modeling are designed to inform medical decision making and health care resource allocation.  When an aim of the study is to aid decision makers about acquiring additional information to reduce uncertainty, The ISPOR-SMDM Modeling Good Research Practices Task Force recommends the results be presented in terms of expected value of perfect information (EVPI).  EVPI is a measure of uncertainty used to inform decision makers on the expected value of gaining additional evidence to support a decision.  While the use and presentation of EVPI in the literature base has steadily increased, the concept remains complex, especially to consumers of health economics research.  Gaining hands-on experience and learning by doing via a practical example will help consumers better understand how to apply and interpret EVPI.

DESCRIPTION:

This workshop will offer participants a practical introduction to understanding the calculation, presentation, and policy implications of an EVPI analysis.  Discussion leaders will draw from prior educational experiences with EVPI to conduct this workshop.  Through active learning, we will engage the audience by providing a one-page simplified exercise of calculating EVPI.  Step-by-step, discussion leaders will walk the participants through the one-page exercise using visual aids.  Using Excel®, we will expand on this simplified exercise by demonstrating the calculation and graphical presentation of an EVPI curve from a published cost-effectiveness model with probabilistic sensitivity analyses.  Additionally, we will provide an introduction to how EVPI is and can be used to inform reimbursement and research funding decisions.  The target audience for this workshop includes outcomes researchers, clinicians, decision makers, and other health economic research consumers interested in gaining an understanding of the concept of EVPI beyond the definition.

Health Care Policy Development Using Outcomes Research

3:00 PM - 4:00 PM
Room: Room 519ab (5th Floor)

W28: HOW TO ADDRESS UNCERTAINTY RELATED TO TRANSFERABILITY OF CLINICAL TRIAL FINDINGS TO CLINICAL PRACTICE?

Discussion Leaders:

Mondher Toumi, MD, PhD, Professor, Market Access, University Claude Bernard Lyon 1, Lyon, France

Asa Kornfeld, MSc, Director, Pricing and Market Access, Creativ-Ceutical, Paris, France

Pascal Auquier, MD, PhD, Professor & Hospital Practitioner, Department of Public Health, Timone University Hospital, Marseille, France

Isaac AO Odeyemi, DVM, PhD, MBA, MSc, Senior Director, Health Economics & Outcomes Research, Astellas Pharma Europe Ltd, Chertsey, UK

PURPOSE:

As prices of new drugs increase, payers are increasingly concerned about uncertainty surrounding their effects. Uncertainty has become a primary reason for non-reimbursement / non-recommendation of newly approved drugs. Among all sources of uncertainty, the transferability of findings from randomized clinical trials (RCT) to real life is the primary source of uncertainty for payers. As RCT design decisions are made at early stage, future payers’ expectations are usually not given enough consideration. When payers’ evidence requirements are integrated in decision making, it is often too late for revising RCT designs. The objective of this workshop is to raise awareness about this market access issue, and discuss methods to minimize uncertainty related to transferability of RCT findings to clinical practice.

DESCRIPTION:

We will review the different sources of uncertainty related to the transferability of RCT, and present case studies. This will include inclusion/exclusion criteria, disease management, concomitant condition, geography of investigational centres, discontinuation/censoring effect, composite and surrogate endpoints, drug regimens, heterogeneity/homogeneity of RCT population, etc. For each situation we will propose recommendations to minimize the risk of non-transferability of RCT in real life clinical practice. Some available options relate to RCT design adjustments, such as follow-up of drop out/censored patients and collecting background patient information prior to trial inclusion. In other situations, it is necessary to conduct new studies, such as pragmatic phase IIIb studies, use of registries, or databases. This will be illustrated by case studies that were positively appraised by health technology assessment (HTA) bodies. Three speakers will represent different perspectives: an HTA committee (an experienced HTA committee member), industry, and academia. The moderator will discuss proposed recommendations and seek comments from the audience.


3:00 PM - 4:00 PM
Room: Room 520a (5th Floor)

W29: THE NEXT FRONTIER: EXPLORING ACCESS EVOLUTION IN THE MIDDLE EAST GULF COOPERATION COUNCIL MARKETS – HEALTH TECHNOLOGY ASSESSMENT, REGIONAL PROCUREMENT, AND VALUE FOR MONEY

Discussion Leaders:

Meghan Gallagher, Director, Global Health Economics and Outcomes Research, Oncology, Sanofi, Cambridge, MA, USA

Cyrus A. Chowdhury, Chief Executive Officer, CBPartners, New York, NY, USA

Husein Reka, MSc, Manager, National Health Authority, State of Qatar, Doha, Qatar

Ola Ghaleb Al Ahdab, PhD, Pharmaceutical Advisor, Registration and Drug Control Department, Ministry of Health, Abu Dhabi, United Arab Emirates

PURPOSE:

The purpose of this workshop is to explore recent activities undertaken by key authorities within the GCC region as part of the gradual evolution towards a value-for-money access-based environment. 

DESCRIPTION:

During the past few years, key markets within the region, including the United Arab Emirates, Qatar, and Saudi Arabia, have taken major steps forward toward establishing a more rigorous system for health care decision making, with an increasing focus on value for money assessment.  This workshop will focus on reviewing some of these activities, and identifying the areas that remain to be developed, including data aggregation and formal HTA capacity building.  The potential for regional collaboration across GCC markets is a high likelihood, and will be explored as a potential future scenario to increase purchasing power, negotiation leverage, and regional efficiency.  Although the prospect of a formal HTA is still a few years away, the steps made today will set the groundwork for a value-assessment process that is seemingly inevitable for the region.  Participants will come away with a solid understanding of the current environmental landscape and evolution in the GCC  markets, and how policies being implemented today will establish the value-for-money infrastructure of their future.

Patient-Reported Outcomes & Patient Preference Research

3:00 PM - 4:00 PM
Room: Room 520b (5th Floor)

W30: QUANTITATIVE CHALLENGES FACING PATIENT-CENTERED OUTCOMES RESEARCH

Discussion Leaders:

Wen-Hung Chen, PhD, Director of Psychometrics, Patient-Reported Outcomes, RTI Health Solutions, Research Triangle Park, NC, USA

Lauren Moore Nelson, PhD, Director of Psychometrics, Patient-Reported Outcomes, RTI Health Solutions, Research Triangle Park, NC, USA

Lori Davis McLeod, PhD, Head of Psychometrics, Patient-Reported Outcomes, RTI Health Solutions, Research Triangle Park, NC, USA

Maria Orlando Edelen, PhD, Senior Behavioral Scientist, RAND Corporation, Boston, MA, USA

PURPOSE:

Patient-centered outcomes researchers collect data from patients and caregivers that can be used to guide health care decisions and improve health care delivery and outcomes. This workshop presents challenges associated with conducting quantitative data analysis of patient-centered data. At the end of the workshop, participants will have a better understanding of these analytical challenges and available approaches to successfully overcome them.

DESCRIPTION:

Topic 1: Heterogeneity in patient-centered outcomes often translates into multidimensionality in data analysis. Different languages and cultures also contribute to heterogeneity of patient-centered outcomes that may lead to bias results. Methods to explore dimensionality and differential item functioning (DIF) are presented, including available software and programs.

Topic 2: Insufficient sample size may lead to large measurement errors or nonconvergent models. Too large of a sample size overpowers tests of significance. Recommendations for sample size when evaluating patient-centered measures are discussed. Rules of thumb for commonly used psychometric analyses, to ensure appropriate statistical inferences, are presented.

Topic 3: Missing data are inevitable, but non-random missing or skip-pattern questions by design may lead to bias and incorrect results. Patterns of missing data should be investigated with respect to demographics, disease severity, and study arms. A case study of a pattern-mixture model is used to identify groups of subjects with similar missing data patterns.

Topic 4: Low response rates are typically non-random and can threaten generalizability of results. Options for maximizing response rate and minimizing respondent burden, including use of IRT-based tools (short forms and computer adaptive tests) and multiple assessment platforms (hand-held devices such as tablets and smart phones) will be discussed.

While the entire workshop will involve a very interactive exchange of ideas, including examples from published work, the last 15 minutes will be dedicated to a discussion of the topics presented and any others relevant to the challenges.

Use of Real World Data

3:00 PM - 4:00 PM
Room: Room 518abc (5th Floor)

W31: EVALUATION OF THE BARRIERS AND OPPORTUNITIES OF BIG DATA IN HEALTH OUTCOMES RESEARCH

Discussion Leaders:

Diana Brixner, PhD, RPh, FAMCP, Professor & Chair, Department of Pharmacotherapy, University of Utah, Salt Lake City, UT, USA

Carl V Asche, PhD, Director, Center for Outcomes Research & Research Professor, Department of Medicine, University of Illinois College of Medicine, Peoria, IL, USA

Maggie Gunter, PhD, Director of Medical Outcomes Research, Lovelace Respiratory Research Institute, Albuquerque, NM, USA

Manuel Prado, BA, President & CEO, Real Health Data, Santa Cruz, CA, USA

PURPOSE:

The purpose of this workshop is to review sources of Big Data and their advantages and disadvantages vs. standard databases used for health outcomes research.   An interactive discussion on the future of Big Data in health outcomes research will be facilitated.

DESCRIPTION:

The discussion leaders will present different approaches to gathering and analyzing Big Data.  Dr. Brixner will introduce general concepts of Big Data, how the data is brought together and some of the challenges in using Big Data in health outcomes research.  Dr. Asche will present some of the potential opportunities for Big Data and some of the opportunities and challenges faced when proposing Big Data as a data source for research projects.  This will be followed by a presentation by Mr. Manuel Prado of Real Health Data, who will provide specific examples highlighting how Big Data can help us understand not simply that an event happened but more importantly why it happened.  Dr. Maggie Gunter will close out the presentations by describing how the HMO Research Network, a coalition of research organizations affiliated with large health plans, come together to share data in a common format to conduct multi-site research, and providing examples of how this data is used in research and what some of the challenges are.  Dr. Brixner will facilitate a discussion between the workshop presenters and the audience on how Big Data is being used today, how it might be used in the future, and what obstacles need to be overcome in order to maximize its value.

Contact ISPOR @ info@ispor.org | View Legal Disclaimer
© 2018 International Society for Pharmacoeconomics and Outcomes Research.
All rights reserved under International and Pan-American Copyright Conventions.