WORKSHOPS – SESSION ITuesday, May 6, 2008
2:30PM-3:30PM

ECONOMIC OUTCOMES RESEARCH

HEALTH CARE POLICY DEVELOPMENT USING OUTCOMES RESEARCH

PAYING FOR PILLS BY RESULT: PERFORMANCE-BASED REWARDS FOR INNOVATION
Discussion Leaders: Louis P. Garrison PhD, Professor of Pharmacy, University of Washington, Department of Pharmacy, Seattle, WA, USA; Sean D Sullivan PhD, RPh, Professor of Pharmacy and Public Health and Director, University of Washington, Pharmaceutical Outcomes Research and Policy Program, Seattle, WA, USA; Peter J. Neumann ScD, Professor, Tufts-New England Medical Center, Institute for Clinical Research and Health Policy Studies, Boston, MA, USA; Adrian Towse MPhil, Director, Office of Health Economics, London, UK
PURPOSE: The purpose of the workshop is to define and discuss options for basing reimbursement for drugs, devices, and diagnostics on ex post outcomes, such as quality-adjusted life years gained, tumor response, or percent reaching a target biomarker level.
DESCRIPTION: Recently, there has been growing interest and discussion of new models for paying pharmaceutical and other medical technology manufacturers for innovative products. Payers would reward them with robust reimbursement for drugs or tests based on an acceptable threshold level of evidence of clinical utility in exchange for additional information and post-market tracking and analysis to determine whether the drug or test “performed” as expected. These types of arrangements—sometimes called “risk-sharing”-- have been a topic of discussion for many years. To date, the most prominent example has been a special reimbursement system for some multiple sclerosis drugs covered by the UK National Health Service. In a major and highly visible proposal, the UK Office of Fair Trade recently issued a report calling for valued-based pricing to replace the Pharmaceutical Price Regulation Scheme. Other recent examples have drawn attention, such as measuring tumor response with Velcade in the UK, and a notable US reimbursement arrangement between United Healthcare and Genomic Health for its Oncotype Dx test, designed to predict the risk of recurrence among women who have had surgery for breast cancer.
Measuring real-world performance is a costly scientific and operational challenge. The limited number of examples is testament to these challenges, as well to the understandable reluctance of technology manufacturers to take full responsibility for actual performance in a real-world setting. This growing discussion around both value-based pricing and value-based reimbursement raises a host of questions about economic incentives, market trends, and potential government interventions of critical importance to medical technological manufacturers, payers, and policymakers. The discussion leaders in this workshop are co-investigators in a study to examine these new reimbursement models, analyze the associated incentives, and assess their feasibility and generalizability. These and other recent examples will be discussed, and the implications for practice and policy will be explored.

PATIENT-REPORTED OUTCOMES RESEARCH

DEVELOPMENT AND USE OF DECISION BOARDS FOR DETERMINATION OF THE GENERAL PUBLIC'S PREFERENCE IN WILINGNESS-TO-PAY ANALYSIS
Discussion Leaders: Michael Iskedjian BPharm, MSc, President, PharmIdeas, Buffalo, NY, USA; Olivier Desjardins BSc, Senior Research Analyst, PharmIdeas Research and Consulting Inc, Ottawa, ON, Canada; Thomas Einarson PhD, VP Scientific Affairs, PharmIdeas Research and Consulting Inc, Oakville, ON, Canada
PURPOSE: The purpose of this workshop it to examine the value of modified decision boards, as well as their development and use for determination of preference in the general public in Willingness-to-Pay (WTP) analysis.
DESCRIPTION: WTP analysis is used in contingent valuation to estimate the maximum dollar amount that an individual is hypothetically willing to pay for a preferred good or service. WTP is applied in healthcare to assess the willingness to pay for a novel drug or intervention, in comparison to and existing alternative or standard of care. WTP analysis first relies on the preference for an intervention before assessment of the maximum willingness to pay. Decision boards have been used by physicians to transfer clinical information such as effectiveness and safety of treatment options, as well as the probabilities associated with the benefits and adverse events of the different treatment options available. Once the clinical information has been presented to survey participants they can be asked to indicate which treatment option they prefer or whether they are indifferent between the options. This workshop will examine all the aspects of developing modified decision boards, as originally suggested by Gafni (1991) and later by O'Brien and Gafni (1996). Various applications will be reviewed from the literature with discussions of issues, and suggestions for their solution. Issues of interest will cover data sources, data synthesis and validation of summarized texts and illustrations, design and presentation formats, as well as practical issues of administration of the decision boards. A discussion will follow on potential bias at each step, and various ways of addressing it.

USE OF REAL WORLD DATA

RECOMMENDATIONS FOR THE USE OF PATIENT REGISTRY DATA AS A COMPLEMENT FOR RANDOMIZED CLINICAL TRIALS
Discussion Leaders: Steven K Takemoto PhD, Associate Professor, Saint Louis University, Center for Outcomes Research, Saint Louis, MO, USA; Nancy A. Dreyer PhD, Chief of Scientific Affairs, Outcome, Cambridge, MA, USA; Claudio Faria PharmD, MPH, Associate Director of Clinical Research, UMass Medical School, Charlestown, MA, USA; Fang Wang MD, PhD, Director, GlaxoSmithKline, Global Health Outcomes, King of Prussia, PA, USA
PURPOSE: To present and develop a foundation to firmly establish patient registry data in the hierarchy of medical evidence and present recommendations for the use of registry data as a complement for randomized controlled clinical trials in health care decision making. This workshop will be of interest to both individuals who perform outcomes research as well as those who apply outcomes research for health care policy decision making.
DESCRIPTION: Although randomized controlled trials have traditionally been used to assess the efficacy of medical interventions, patient registries are increasingly used to complement these data by providing evidence of effectiveness.
The ISPOR Patient Registry Special Interest Group's Communicating Patient Registry Information for Health Care Decision-Making (CIPR-HCDM) Working Group has established a team to develop recommendations for the use of patient registry data as a complement to randomized clinical trial data in health care decision-making. This workshop will present the Patient Registry & Randomized Clinical Trial Data (PR & RCT Data) Project Team's work to date. The workshop consists of three sections.  The first section shall address: 1) randomized clinical trial strengths and weaknesses; 2) patient registry strengths and weaknesses, and 3) the relative contributions of trials and registries in the hierarchy of evidence.  The second section will present results of the team's survey of payers and policy makers to assess the Usefulness of Observational and Registry Data for Health Care (Coverage and Reimbursement) Decisions.  The third section will involve audience participation in the development of recommendations for the use of registry data for complementing randomized controlled trials. * Discussion leaders represent the ISPOR Patient Registry SIG-The Communicating Patient Registry Information for Health Care Decision-Making Working Group - PR & RCT Data Project Team Members: Nancy Dreyer PhD, MPH; Claudio Faria PharmD, MPH; Trisha Hutzul MSPH, Gerlinde Kaempf MA; Pat McCollam PharmD, RPh, BS; Steven Takemoto PhD; Fang Wang MD, PhD; Jaroslaw Wechowski PhD, MD, BA; Jasmanda Wu PhD, MPH.

CLINICAL OUTCOMES RESEARCH

ECONOMIC OUTCOMES RESEARCH

COUNTRY-TO-COUNTRY ADAPTATION OF PHARMACOECONOMIC RESEARCH: METHODOLOGIC CHALLENGES & POTENTIAL SOLUTIONS
Discussion Leaders: David Thompson PhD, Vice President, i3 Innovus, Medford, MA, USA; Amy K O'Sullivan PhD, Associate Director, i3 Innovus, Medford, MA, USA; Debbie L Becker MSc, Director, i3 Innovus, Burlington, ON, Canada
PURPOSE: To discuss methodological challenges and solutions in adapting pharmacoeconomic research projects initiated in one country to another that has different population, institutional, and health-care financing characteristics. The workshop will be of interest to health economics managers in the pharmaceutical industry and applied researchers in academic and consulting environments.
DESCRIPTION: There now exists a well-established tradition for conducting pharmacoeconomic research in the world's major pharmaceutical markets in North America, Europe, and the Asia/Pacific region. As the body of work has grown, so too have opportunities to adapt scientifically rigorous research conducted in one country to other countries not included in the original study. The concept of “adaptability” is similar yet distinct from that of “transferability”, which has been the subject of an ongoing ISPOR Task Force. Strictly speaking, research transfer refers to the application of research findings generated in one country to a foreign setting—without any modifications made to account for differences in patient populations, health-care delivery systems, or financing mechanisms. In contrast, adapting research involves at least some—and possibly extensive—modification of the research to account for these differences and make the findings more applicable to the new setting(s). However, the process of adaptation usually meets several challenges, principal among them data limitations, lack of generalizability of analytic structures, and differences in decision-maker information needs. Researchers therefore find themselves making difficult judgment calls related to how extensive research needs to be modified before it will be admissible in another country. This workshop will describe these and other methodological hurdles in adapting each of the major types of pharmacoeconomic evaluation—including modeling, trial-based evaluations, and retrospective database analyses—from one country to another and provide a forum for structured discussion of how these hurdles can be overcome in a cost-effective manner. Attention will be focused on obstacles presented by differences in health-care financing and delivery systems, drug funding decision making, treatment patterns, and medical-care costs. Examples from the discussion leaders' own experience as well as the literature will be used and workshop participants will be encouraged to share their thoughts and experiences in country-to-country adaptation of pharmacoeconomic research.

HEALTH CARE POLICY DEVELOPMENT USING OUTCOMES RESEARCH

ARE WE BETTER OFF OR WORSE OFF WITH VALUE-BASED PURCHASING (VBP)?
Discussion Leaders: Peter K. Wong PhD, MBA, MS, RVP, Quality, Clinical Effectiveness & Chief Pharmacy Officer, Clincial Effectiveness Mercy Health Partners, Southwest Ohio, Cincinnati, OH, USA; Dennis J Hanseman PhD, Senior Health Researcher, Clincial Effectiveness Mercy Health Partners, Southwest Ohio, Cincinnati, OH, USA; Alan H Mutnick PharmD, FASHP, Director, Clincial Effectiveness Mercy Health Partners, Southwest Ohio, Cincinnati, OH, USA
PURPOSE: Discuss the implementation of VBP and the potential financial and outcomes consequences to a hospital system.
DESCRIPTION: The implementation of value-based purchasing (VBP) by the Center for Medicare and Medicaid (CMS) beginning in fiscal year 2009 will have a huge impact on providers, CMS, and patients. The purpose of the program is to promote cost-effective care while potentially eliminating defects that are part of today's health care delivery systems. The report to congress sent by the U.S. Department of Health and Human Services entitled “Plan to Implement a Medicare Hospital Value-Based Program” has detailed the road map and has described how the program will function. In its contents, there are several alerts, which include: 1) the department will not pay for “Never Events”, e.g. foreign object left in the body cavity and Hospital-Acquired Conditions, e.g. hospital-acquired infections: 2) the department will pay hospital providers based on their quality performance and patient experiences; 3) hospital providers must submit with the UB-92 claims submissions all the conditions that are present on admission; and 4) more data submission to the department in the future. While the concept of paying for good performance is attractive and theoretically sound, the true financial impacts have not been assessed. Despite the tremendous efforts to improve quality in the past, elements of risk when using the healthcare system such as bed sores, patient falls with injuries, etc cannot be totally eliminated. Under the proposed payment scenarios, all of the associated risks of hospitalization will be transferred from the payers to the providers. Our Quality Improvement team has devised a method to estimate and assess the financial impact for our hospital system under the proposed VBP.

CONSIDERATION OF CLINICAL AND ECONOMIC VALUE IN U.S. HEALTH TECHNOLOGY ASSESSMENT: A MULTIPLE STAKEHOLDER VIEW
Discussion Leaders: Daniel A. Ollendorf MPH, Chief Review Officer, Institute for Clinical & Economic Review, Boston, MA, USA; Steven D. Pearson MD, MSc, Senior Fellow, America's Health Insurance Plans, Washington, DC, USA; Amy Knudsen PhD, Senior Scientist, Massachusetts General Hospital, Institute for Technology Assessment, Boston, MA, USA
PURPOSE: The purpose of this workshop is to review and understand the key issues with current approaches to health technology assessment in the U.S., and to introduce a novel method for consideration of intervention value from a multiple stakeholder perspective.
DESCRIPTION: Policy makers and proponents of health care system reform argue that meeting the goals of sustained innovation, cost control, and improved quality within the health care system require systematic appraisal of the clinical effectiveness and comparative value of new or emerging interventions, yet few current approaches to health technology assessment (HTA) regularly integrate these considerations. This workshop will involve an overview of current HTA efforts in the U.S., with a focus on: 1) mission and funding source(s); 2) breadth of assessment types; 3) presence/absence of an explicit process for topic consideration/prioritization; 4) assessment design and key activities; 5) presence/absence of consideration of cost-effectiveness; and 6) key assessment outputs (e.g., standalone reports, recommendations, action plans) Following this review, a novel process and format for HTA with a focus on multiple stakeholder input and more actionable results will be introduced. To support this presentation, a case study will be described, based on a recent appraisal of CT colonography vs. colonoscopy for routine colorectal cancer screening. The case study will include a detailed discussion of the assessment process, collection of input from patients, providers, payers, and researchers, specifics of the analysis performed, and the ultimate research conclusions. An interactive exercise will then be conducted to allow the audience to convert the research results into potential reimbursement and/or policy changes. Active audience participation through informal discussion will also be encouraged. This workshop will be of interest to those involved in technology assessment, comparative effectiveness research, and policy implementation and evaluation.

PATIENT-REPORTED OUTCOMES RESEARCH

USE OF REAL WORLD DATA

ECONOMIC DATABASES FOR PHARMACOECONOMIC EVALUATIONS IN CANADA: OVERVIEW AND USE
Discussion Leaders: Carl V. Asche PhD, Research Associate Professor, University of Utah College of Pharmacy, Outcomes Research Center, Department of Pharmacotherapy, Salt Lake City, UT, USA; Philip Jacobs DPhil, Director, Institute of Health Economics, Edmonton, AB, Canada; Rita Yim MA, MHSA, Research Fellow, Institute of Health Economics, Edmonton, AB, Canada; Joanne Kingston PhD, Senior Economist, Institute of Health Economics, Edmonton, AB, Canada
PURPOSE: The aim of this workshop is to allow participants to understand the availability of databases associated with evaluations in Canada. We will use the ISPOR international classification so that international participants will be able to better relate the material to their own national contexts. The purpose of this workshop is to increase the awareness of the economic databases available in Canada that are applicable to conducting pharmacoeconomic analyses. The participants will gain an understanding of Canadian healthcare databases and how to use them.
DESCRIPTION: The first part of the workshop will deal with the uses of economic databases in economic evaluation studies. The second part of the workshop will deal with an overview of the concepts used in economic evaluations, including the study perspective (governmental, societal), components of the health care system (focusing on the Canadian system) and key elements of a data base. Data will also be summarized in relation to the ISPOR Digest of International Databases. The third part of the workshop will provide a detailed introduction to Canadian databases and their uses, including the hospital discharge database, the physician fee-for-service database, the pharmaceutical database, and the ambulatory care database. In the fourth part of the workshop we will present examples of how these databases have been used in economic studies. Strengths and weaknesses of the Canadian databases will be highlighted.

Discussion Leaders: George A. Goldberg MD, Medical Director, i3 Innovus, Santa Monica, CA, USA; Matt Moore MHA, Director Health Economics, Ethicon Endo-Surgery, Inc, Health Economics, Cincinnati, OH, USA; Michael Dutro PharmD, Director, Pfizer Inc, Albuquerque, NM, USA
PURPOSE: The purpose of this workshop is to describe the common mistakes, misconceptions and misinterpretations related to the use of procedure and diagnosis codes in claims data analyses, and to provide insight into accurately using claims data to identify study cohorts, disease, utilization and costs for research studies.
DESCRIPTION: Retrospective claims data are a valuable resource for health economic and outcomes research projects, providing access to large numbers of patients and extensive cost and utilization data for relatively low cost. Accepting the inevitable losses in translation moving from medical record entries to claim forms, these studies rely on diagnosis and procedure codes to accurately identify patients with disease, and to attribute health care utilization and costs to certain diseases and treatments. Limitations of claims data mean that careful consideration of how to work with claims data, and coding in particular, is needed to avoid serious errors in identification and attribution. The workshop will describe the benefits and limitations inherent in using codes in claims data to identify disease and attribute health care resource utilization and costs. We will focus on describing the options, with their pros and cons, of: 1) using codes to identify a disease cohort, 2) using codes to attribute services and costs to a specific disease, and 3) addressing other challenging aspects of diagnosis and procedure coding Examples from retrospective claims data studies will be used to illustrate the issues and solutions. One example is a study of minimally invasive procedures covering a variety of disease areas, procedures, co morbid conditions and outcomes. Another example will focus on a large claims data disease categorization and reporting system.

Clinical Outcomes Research

ADVANCES IN META ANALYSIS: TECHNIQUES FOR INCLUDING MULTIPLE STUDY DESIGNS, MULTIPLE ENDPOINTS, AND MULTIPLE TREATMENTS

Discussion Leaders: Jeroen P Jansen PhD, Associate Research Director, Mapi Values, Boston, MA, USA; Melvin Olson PhD, Senior Biostatician, Novartis Pharma AG, Basel, Switzerland; Chris Evans PhD, MPH, Director of Economics and Outcomes, Mapi Values, Boston, MA, USA
PURPOSE: Participants will be introduced to meta-analysis methods that enable a greater proportion of the available evidence to be used: 1) meta-analysis with multiple study designs (i.e. parallel randomized designs, cross-over designs, and uncontrolled trials); 2) multiple endpoints; and 3) multiple treatments. Focus is on familiarizing participants with new or seldom used methods and why a Bayesian approach provides the information needed for decision-making.
DESCRIPTION: Regularly only results from randomized controlled trials (RCTs) with a parallel design are combined in meta-analyses. Cross-over trials are often ignored, whereas these designs may contribute evidence in a fifth of systematic reviews (Lathyris et al.2007). Furthermore, single arm trials, while not providing information on relative effects, can provide valuable evidence on absolute effects. Especially for rare diseases, RCTs account for the minority of information, and it can be of interest to draw together both randomized and single arm studies. In order to consider the totality of the evidence base, results from studies with different designs can be analyzed simultaneously by explicitly acknowledging the differences in designs. Despite an interest in multiple endpoints, meta-analyses are often performed by endpoint. Analyzing multiple outcomes simultaneously has several advantages. For example by documenting relationships between endpoints which may be of clinical or scientific interest (Arends, 2006) Traditional meta-analysis can also be extended by including multiple different pair-wise comparisons across a range of different interventions (Lu & Ades, 2004). This allows for indirect comparisons in the absence of head-to-head data and can strengthen the evidence base as well. Traditionally meta-analysis is performed with a frequenters approach resulting in a pooled estimate along with a confidence interval. One of the drawbacks of this approach is the non-intuitive interpretation from a decision-making perspective, especially when there is no statistically significant finding. By contrast a Bayesian approach delivers a joint posterior distribution of the parameters of interest with a natural interpretation that can be directly fed into a decision framework to support clinical or reimbursement decision-making. In this workshop we provide an overview of the more advanced methods of meta-analysis and we will interactively discuss the value of the use of additional evidence and the use of a Bayesian approach for decision-making.

AN UNBIASED OVERVIEW AND UNDERSTANDING OF THE USE OF PROPENSITY SCORING IN PHARMACOECONOMIC AND PHARMACOEPIDEMIOLOGY RESEARCH
Discussion Leaders: Matthew W. Reynolds PhD, Managing Director, Epidemiology and Database Services, United BioSource Corporation, Medford, MA, USA; Christopher Hollenbeak PhD, Assistant Professor, Penn State College of Medicine, Health Evaluation Sciences, Hershey, PA, USA; David J. Vanness PhD, Assistant Professor, University of Wisconsin-Madison, Department of Population Health Sciences School, Madison, WI, USA
PURPOSE: This symposium is designed for the researchers who are not “experts” in the area of propensity score methodology. It is intended to be a discussion on propensity scores, their appropriate use in pharmacoeconomics and pharmacoepidemiology, and their benefits and limitations.
DESCRIPTION: Since the prevalence of the use of propensity scores is rapidly increasing, it is important to better understand this methodology and how, when, and why to use or not use them in our research methodologies.
The following topics will be addressed and then there will be several examples presented for open discussion and audience participation regarding appropriate use of and interpretation of propensity scoring methodology:1) an introduction to define the term, “propensity score”, and to discuss the multitude of ways that propensity scores are implemented in pharmaceutical research methods; 2) how do propensity scores compare to other matching and covariate-control methods? This section will also include a discussion of when are the most and least appropriate situations to employ propensity scoring; 3) he pros and cons of propensity scores. This presentation will discuss the benefits and potential limitations to using propensity scores in pharmacoeconomic and pharmacoepidemiological research; 4) several interesting examples of propensity scoring and corresponding results will be presented for discussion. Audience participation will be encouraged through active discussion of the use of propensity scores. We will present and discuss multiple examples and evaluate the strengths and limitations as well as appropriate research use of propensity scores. This workshop should be of interest to any researcher who has worked with large datasets and needs to address the issues of patient selection bias, patient channeling, confounding by indication, and other reasons for consideration of employing propensity scores.

ECONOMIC OUTCOMES RESEARCH

EDUCATION/COMMUNICATIONS IN OUTCOMES RESEARCH

HEALTH BEHAVIOR CHANGE: LEADING MODELS AND THEIR PRACTICAL APPLICATION
Discussion Leaders: Vernon F Schabert PhD, Senior Director, IMS Consulting, Health Economics and Outcomes Research, Santa Barbara, CA, USA; Alexandra Drane BA, President, Eliza Corporation, Beverly, MA, USA
PURPOSE: The purpose of this workshop is to highlight fundamentals of health behavior change, using examples from speech recognition telephonic interventions to illustrate their practical application.
DESCRIPTION: Increased attention in outcomes research to topics such as treatment adherence and the distinction between efficacy and effectiveness highlight how patient behavior mediates the relationship between treatments and outcomes. Outcomes researchers find themselves increasingly charged with, or partnered with those charged with, influencing patient behavior to optimize treatment effectiveness. The purpose of this workshop is to highlight fundamentals of health behavior change, using examples from speech recognition telephonic interventions to illustrate their practical application. Discussions will include a survey of major health behavior change models from the perspective of a stakeholder responsible for health promotion. These models will be supplemented with general theories of persuasion and behavior change to address gaps in the application of health behavior models. The workshop will include examples of how the development of telephonic, automated voice programs for patient education and reminders consider and apply these behavioral change models. Authors will present examples from their own research and from findings in the published literature. The workshop will engage audience members by listening to recorded audio from actual intervention calls and discussing the behavioral theories that influenced scripting decisions. The audience will also participate in a team exercise to identify incentives and barriers at various stages of behavior change in specific health conditions.

HEALTH CARE POLICY DEVELOPMENT USING OUTCOMES RESEARCH

TRANSPARENT AND QUANTIFIABLE APPROACHES TO HEALTH CARE DECISION MAKING: CHALLENGES AND OPPORTUNITIES FOR MULTI CRITERIA DECISION ANALYSIS (MCDA)
Discussion Leaders: Mireille M Goetghebeur PhD, VP Operations, BioMedCom Consultants Inc, Montreal, QC, Canada; Louis Niessen MD, PhD, Senior Researcher, Erasmus Medical Center, Institute for Medical Technology Assessment, Rotterdam, The Netherlands; Lonny J Erickson PhD, Senior Associate, Health Technology Assessment, BioMedCom Consultants Inc, Montreal, QC, Canada; Hanane Khoury PhD, Senior Research Application Associate, BioMedCom Consultants Inc, Montreal, QC, Canada
PURPOSE: Introduce participants to the Multi Criteria Decision Analysis approach, applying this method to drug formulary decision making
DESCRIPTION: Internationally, there seems to be a shift towards more comprehensive criteria in drug reimbursement decisions (NICE, Netherlands, Australia). Multi Criteria Decision Analysis (MCDA) is an established method that is widely used in various disciplines. This quantifiable approach facilitates decision making by structuring information and allowing scoring of relevant components. Comprehensive MCDA approaches have recently been employed to aid decision making in the context of healthcare, and are relevant to current debates regarding evidence and healthcare decisions. This workshop aims to raise awareness of MCDA approaches and to facilitate collaborative work in this area among healthcare stakeholders at the international level. In particular, we will apply this specific method in a relevant area of healthcare planning. Participants will leave this workshop with basic knowledge of the MCDA method and hands-on experience. The exercise should stimulate future ideas for exploration and application of quantifiable approaches to decision support in participants own environments. The workshop will be structured to give an overview of the MCDA method from theory to context and application in a logical manner, combining short presentations followed by a practical exercise: 1)An overview of the MCDA method and its potential to support healthcare decision making by an author of key innovative work in this area1 (LW Niessen); 2) An overview of the EVIDEM framework (Evidence and Value : Impact on DEcisionMaking), a practical and objective method applying MCDA to segregate and quantify components of healthcare decision making (MM Goetghebeur); and 3) Hands-on exercise for workshop participants involving presentation of a MCDA framework to support drug formulary decision-making, and application of part of the framework to evaluation of a candidate drug (LJ Erickson, H Khoury, M Wagner).

PATIENT-REPORTED OUTCOMES RESEARCH

USE OF REAL WORLD DATA

APPLES, ORANGES, AND PEARS: SURVIVAL ANALYSIS OF MULTIPLE ENDPOINTS
Discussion Leaders: Nicole M. Engel-Nitz PhD, Senior Researcher, i3 Innovus, an Ingenix Company, Eden Prairie, MN, USA; Xin (Sam) Ye MS, PhD, Senior Researcher, i3 Innovus, an Ingenix Company, Eden Prairie, MN, USA
PURPOSE: The purpose of this workshop is to demonstrate and discuss the use of survival analyses of multiple endpoints in health care outcomes studies.
DESCRIPTION: Outcomes researchers often use survival analysis techniques such as Cox Proportional Hazards models to assess outcomes such as medication discontinuation. Multinomial logistic regression is used to assess multiple outcomes, but the field has made less use of survival techniques to assess multiple outcomes while accounting for variable follow-up time (e.g. medication discontinuation, dose increase, or dose decrease). This workshop will focus on competing risk models. We will present examples using real-world data, along with SAS and Stata code for the examples. Comparisons will be provided of outcomes assessed using single-event survival analysis and multiple event survival analysis techniques. Audience participation will be encouraged through the discussion of the application of these techniques to specific health outcomes. This workshop will benefit researchers with an interest in time-to-event analyses involving multiple events.

CLINICAL OUTCOMES RESEARCH

ECONOMIC OUTCOMES RESEARCH

ASSESSING PATIENT COSTS FOR CANCER IN THE U.S: HOW, WHO, WHEN, AND WHAT
Discussion Leaders: Michael T. Halpern MD, PhD, Strategic Director, American Cancer Society, Health Services Research, Atlanta, GA, USA; K. Robin Yabroff PhD, Epidemiologist, National Cancer Institute, Applied Research Program, Bethesda, MD, USA; Ya-Chen Tina Shih PhD, Associate Professor, University of Texas M.D. Anderson Cancer Center, Department of Biostatistics and Applied Mathematics, Houston, TX, USA
PURPOSE: To discuss issues and practices in evaluating patient costs for individuals with cancer in the U.S. and the impacts of selecting different methods on patient cost results.
DESCRIPTION: Beyond having substantial impacts on morbidity and mortality, cancer can lead to tremendous economic burdens. With the development of newer and very expensive biologic agents for cancer treatment, there has been increased concern on the patient cost burden for individuals with cancer. However, there is considerable variability in how patient costs are defined and assessed. This workshop will discuss issues related to assessment of U.S. cancer patient costs, including: types of data sources to be used; criteria used in defining the study patient population; time period and/or phase of care for patient costs; types of services and conditions to be included; differences in measures of patient costs; and imputation of missing patient costs. Examples will be drawn from the presenters' own research experiences as well as from the published literature. The implications of different methodology choices on patient cost results will be discussed. Audience participation will be encouraged through active discussion and opportunities for audience members to relate their experience. This workshop will be of interest to health economic and health services researchers as well as government policy makers and patient advocates.

EXTRACTING PATIENT INFORMATION FROM NATIONAL DATABASES: INCLUDING NIS AND MEPS
Discussion Leaders: Patricia B Cerrito PhD, Professor, University of Louisville, Mathematics, Louisville, KY, USA; John C Cerrito PharmD, Pharmacist, Kroger Pharmacy, Louisville, KY, USA
PURPOSE: The purpose of this workshop is to discuss the necessary preprocessing requirements to use statistical methods to analyze large, national databases. In particular, the datasets must be filtered to specific diagnoses or procedures; often these procedure codes must be compressed. SAS code instructions for preprocessing the data will be provided.
DESCRIPTION: The National Inpatient Sample (NIS) contains 15 columns of diagnosis codes and another 15 columns of procedure codes either in CCS format or ICD9 format. The Medical Expenditure Panel Survey (MEPS) contains a separate file for patient conditions, and that file can have multiple records (unlimited) for each patient, with one diagnosis code for each record. All other databases will have patient information in one of these two forms-multiple columns or multiple records. Before any analysis can take place, it is often necessary to filter down to patients with specific conditions and/or co-morbidities. It is important to ensure that the filtering is performed accurately; otherwise, the subsequent analysis will yield incorrect conclusions. For example, the ICD9 code, ‘042' represents HIV. However, it can also occur in ‘25042' for Type II diabetes with renal manifestations. Care must be taken to extract only those patients with the relevant codes. In this workshop, we will demonstrate several different coding methods that can be used to extract the relevant information on patient diagnosis and procedures codes, and to demonstrate how the extraction can be validated. Also, risk adjustment methods typically require the use of indicator functions (0=absence of diagnosis, 1=presence of diagnosis) to define risk adjustment and patient severity indices. We show how these functions can be defined and validated using preprocessing techniques in the statistical software, SAS (SAS Institute, Inc; Cary, NC). The methods will be demonstrated using data from both the NIS and MEPS. Novel data mining techniques will also be presented. Audience participation will include discussion of the relevant problems associated with the choice of indicator functions; different choices will result in different risk adjustment measures, and subsequently, different rankings of healthcare providers.

EDUCATION/COMMUNICATIONS IN OUTCOMES RESEARCH

HEALTH CARE POLICY DEVELOPMENT USING OUTCOMES RESEARCH

DIFFERENCES IN PHARMACOECONOMIC DATA SUBMISSION GUIDELINES FOR THE US, CANADA, AND MEXICO: IMPLICATIONS FOR MULTI-COUNTRY HEALTH-ECONOMICS PROGRAMS
Discussion Leaders: David Thompson PhD, Vice President, i3 Innovus, Medford, MA, USA; Shawn J Barry MA, Associate Director, Analytics, i3 Innovus, Burlington, ON, Canada; Jf Mould-Quevedo PhD, MSc, MBA, Pharmacoeconomics Manager, Pfizer Mexico, Pharmacoeconomics Department, Mexico City, Mexico
PURPOSE: To discuss differences in pharmacoeconomic data submission guidelines and practices in the US, Canada, and Mexico, and what these imply for international health-economics programs designed to support product launch in these three North American markets.
DESCRIPTION: Guidelines for inclusion of pharmacoeconomic data to support formulary submissions were first developed in Australia in the early 1990s. Since then, numerous other countries have followed suit, including the US, Canada, and Mexico. The existence of these guidelines and the widespread practice of submitting pharmacoeconomic data to support formulary submissions continues to fuel the growth of this research; however, international differences in guidelines complicate matters. In the US, pharmacoeconomic data are not formally required by any governmental body, yet the Academy of Managed Care Pharmacy (AMCP) has included economic evaluation as a key component of its "format" for product dossier submissions to health plans, and this has influenced not only the assembly of this information but the underlying methodology as well, fueling the recent rise in budgetary impact assessment, for example. In Canada, submission requirements for new drugs to be reimbursed within the provincial and federal publicly-funded health programs are well developed, rigorous, and extensive. The Canadian Agency for Drugs and Technologies in Health (CADTH) acts as a clearing house for submissions and central authority for the Common Drug Review process. Analyses regarding value (cost-effectiveness and cost-utility) and affordability (budgetary impact) are prepared by or on behalf of manufacturers and reviewed separately. In Mexico, pharmacoeconomic data are formally required for drug inclusion on national formularies. Manufacturers submit research (primarily cost-effectiveness and cost-utility analyses) to gain access and facilitate pricing negotiations with health-care decision makers. Mexican Pharmacoeconomics Guidelines are currently in the process of being finalized by local stakeholders (health-care institutions, academia, and manufacturers). This workshop will use these three North American markets to illustrate commonalities and differences in submission guidelines and associated implications for the design and implementation of multi-country health-economics programs. Examples from the experience of the workshop leaders will be used as a foundation for group discussions.

USE OF REAL WORLD DATA