REAL WORLD DATA METHODS

Retrospective Database Analysis – Econometric Methods

Faculty: William H. Crown PhD, President, i3 Innovus, Auburndale, MA, USA; Henry Henk PhD, Associate Director, i3Innovus, Eden Prairie, MN, USA

Course Description: Large administrative claims databases provide a unique opportunity to examine retrospectively the effects of drug use on clinical and economic outcomes in "real world" settings. This course will cover a discussion of the ISPOR Checklist for Retrospective Database Studies - Report of the ISPOR Task Force on Retrospective Databases and selected topics related to estimators and sampling distributions, properties of sampling distributions (unbiasedness, efficiency, mean square error), and ordinary least squares (OLS) regression. OLS model assumptions and the implications of violations (e.g., heteroscedasticity, multicollinearity, autocorrelation) will also be discussed. More complex topics beginning with the problem of endogeneity, identification, instrumental variables, sample selection models, propensity score models, maximum likelihood methods and the estimation of limited dependent variables models including logit, multinomial logit, count models, and survival models will be discussed. This course will assume participants have knowledge of statistical methods through OLS regression and experience in the analysis of administrative claims databases.

REAL WORLD DATA METHODS

Propensity Scores and Comorbidity Risk Adjustment

Faculty: Fadia Shaya MPH, PhD, Associate Professor and Associate Director, Center on Drugs and Public Policy, University of Maryland School of Pharmacy, Baltimore, MD, USA

Course Description: A large part of the evidence about the effectiveness of different treatments is based on retrospective studies. Issues of bias and confounding relate to the non-random assignment of subjects and co-morbidity burden. This course will outline the concerns about bias and explain the methods for causal inference in observational studies, where researchers have no control over the treatment assignment. A lack of balance in the covariates between the treatment and control groups can produce biased estimates of the treatment effects. We will explain how propensity scores can be used to reduce bias, through stratification, matching or regression. Confounding and the pros and cons of standard adjustment, propensity scoring methodology (sub classification on one confounding variable, overlap in treatment groups, variable selection) will be discussed. In the second part, we will elaborate on risk adjustment models, focusing on morbidity indices, e.g the Charlson Comorbidity Index, and Chronic Disease Scores. Examples using a step by step approach will be presented. This is an introductory course, designed for those with little experience with this methodology but some knowledge of observational databases.

MODELING METHODS

Bayesian Analysis: Advanced

Faculty: Bryan Luce MBA, PhD, Senior Vice President, Science Policy, United BioSource Corporation, Bethesda, MD, USA; Keith R. Abrams PhD, Professor of Medical Statistics, Department of Health Sciences, University of Leicester, Leicester, UK; Christopher S. Hollenbeak PhD, Assistant Professor, Surgery and Health Evaluation Sciences, Penn State College of Medicine, Hershey, PA, USA; David Vanness PhD, Assistant Professor of Population Health Sciences, University of Wisconsin Medical School, Madison, WI, USA.

Course Description: This course introduces the use of Bayesian methods in evidence synthesis (including meta-analysis) and allows participants to gain hands on experience using such modeling techniques within WinBUGS. Methodological issues considered in the course include; fixed and random effects models, choice of prior distributions, subgroups, meta-regression and adjusting for baseline risk, together with indirect and mixed treatment comparisons. Further meta-analysis topics for which a Bayesian approach can be of benefit will also be highlighted.
Participants will be expected to be familiar with the use of WinBUGS and will be responsible for bringing a laptop with the latest, unrestricted version of WinBUGS pre-installed. This course is a follow-up to the short course: Bayesian Analysis-Overview and Applications. Basic knowledge of the Bayesian approach and use of WinBUGS (equivalent to attendance at Bayesian Analysis-Overview and Applications) will be assumed.

Discrete Event Simulation for Economic Analyses

Faculty: J. Jaime Caro MDCM, FCRPC, FACP, Adjunct Professor of Medicine, Adjunct Professor of Epidemiology and Biostatistics, McGill University, Montreal PQ & Senior Vice-President of Health Economics, United BioSource Corporation, Concord, MA, USA; Jörgen Möller MSc Mech Eng, Simulation Specialist, United BioSource Corporation, Eslov, Sweden.

Course Description: This course will provide a basic understanding of the key concepts of discrete event simulation (DES). The focus will be on the use of these simulation models to address pharmacoeconomic (and device-related) problems. The course will be structured around practical exercises. Topics to be covered are: Why DES? Dynamic simulation as a tool; Components of a DES; How do you build a model? Modeling of processes and resource use; Modeling of variables and decisions. If time permits, simple animation will be demonstrated. We will use ARENA to build simple models. Instructors will distribute training versions of Arena. This course is designed for those with some experience with modeling.

QUALITY OF LIFE / PATIENT-REPORTED OUTCOMES / PREFERENCE-BASED METHODS

Patient-Reported Outcomes - Item Response Theory

Faculty: Lori McLeod PhD, Director, Psychometrics, RTI Health Solutions, Research Triangle Park, NC, USA; Cheryl Hill PhD, Director, Psychometrics, RTI Health Solutions, Research Triangle Park, NC, USA

Course Description: There is a great need in health outcomes research to develop instruments that accurately measure a person's health status with minimal response burden. This need for psychometrically sound and clinically meaningful measures calls for better analytical tools beyond the methods available from traditional measurement theory. Applications of item response theory (IRT) modeling have increased considerably because of its utility for instrument development and evaluation, assessment of measurement equivalence, instrument linking, and computerized adaptive testing. IRT models the relationship, in probabilistic terms, between a person's response to a survey question and their standing on a health construct such as fatigue or depression. This information allows instrument developers to develop reliable and efficient quality of life measures tailored for an individual or group. This introductory workshop will discuss the basics of IRT models and applications of these models to improve health outcomes measurement. Illustrations will be used throughout the presentation that focus on measuring key health-related quality of life domains in different disease populations. This introductory course is designed for those with none to little experience with IRT.

USE OF PHARMACOECONOMICS / ECONOMIC / OUTCOMES RESEARCH INFORMATION

Case Studies in Pharmaceutical/Biotech Pricing II - Advanced

Faculty: Jack Mycka, President and Partner, MME LLC, Montclair, NJ, USA; Renato Dellamano PhD, President, ValueVector (Value Added Business Strategies), Milan, Italy

Course Description: Case studies will be employed to lead participants through the key steps of new product pricing, with focus on the need to thoroughly analyze the business environment and its constraints and opportunities and the need to closely integrate the pricing, reimbursement and pharmacoeconomic strategy for the new product with the clinical development and marketing strategies. Practical exercises will allow participants to consolidate the concepts delivered in the “Elements” introductory session and expanded here. Areas covered will include the post-launch issues of reimbursement and pricing maintenance as a part of life-cycle management in a global environment. This course is for individuals who have completed Elements of Pharmaceutical Pricing I – Introduction or are familiar with both the key determinants of pharmaceutical pricing and the main international health systems.  Enrollment for this course is limited.

PHARMACOECONOMIC / ECONOMIC METHODS

Statistical Considerations in Health Economic Evaluations

Faculty: Henry Glick PhD, Associate Professor of Medicine, Division of Internal Medicine, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Jalpa Doshi PhD, Research Assistant Professor of Medicine, Division of Internal Medicine, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Course Description: The adoption and diffusion of new medical treatments depend increasingly on robust analysis of costs and cost-effectiveness. During this course, the following statistical considerations in economic evaluations will be discussed: affect of distributional assumptions, analyzing univariate and multivariable analysis data, analyzing censored data, sample size and power calculations, sampling uncertainty, point estimates for variables, net monetary benefit, and confidence intervals for cost-effectiveness ratios. During this course, study examples will be provided to illustrate concepts. Participants should have some knowledge of basic economic evaluations and statistics.

SUNDAY, MAY 4, 2008 (Afternoon Courses)
1:00 PM - 5:00 PM

PHARMACOECONOMIC / ECONOMIC METHODS

Applications of Statistical Considerations in Health Economic Evaluations

Faculty: Henry Glick PhD, Associate Professor of Medicine, Division of Internal Medicine, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Jalpa Doshi PhD, Research Assistant Professor of Medicine, Division of Internal Medicine, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Course Description: This course will provide applications of statistical considerations in economic analysis. Specific exercises will be conducted to illustrate affect of distributional assumptions, univariate & multivariable analysis of costs, the effect of sample size & power calculations on economic evaluations and point estimates for cost-effectiveness ratios. Participants are encouraged to have hands-on experience and bring their laptops. STATA trial software will be distributed if not already installed and used in this course. The publication “Economic Evaluation in Clinical Trials” (Oxford: OUP, 2007) is suggested as recommended reading for this course.  The course, Statistical Considerations in Economic Evaluations, is a strong prerequisite for this course.

REAL WORLD DATA METHODS

Patient Registries

Faculty: Chris Pashos PhD, Vice President and Executive Director, HERQuLES Health Economic Research and Quality of Life Evaluation Services Abt Bio-Pharma Solutions, Inc., Lexington, MA, USA

Course Description:This course is designed to provide an overview of patient registries and their applications in identifying 'real world' clinical, safety, and patient-perspective issues. The pros and cons of registry data compared to other ‘real world’ and clinical trial data collection will be presented.  How registry information can be used to support other health economics /outcomes research initiatives and health care decision-making will be addressed.  Registry strategy, design, operations and measures of program success will be discussed.  In addition, regulatory trends and requirements, including the Agency for Healthcare Research & Quality’s (AHRQ) May 2007 publication: “Registries for Evaluating Patient Outcomes: A User's Guide”, will be examined.  This course is designed for those with little experience with patient registries.

QUALITY OF LIFE / PATIENT-REPORTED OUTCOMES / PREFERENCE-BASED METHODS

Utility Measures

Faculty: F. Reed Johnson PhD, Senior Fellow and Principal Economist, RTI Health Solutions, Research Triangle Park, NC, USA; A. Brett Hauber PhD, Head, Health Preference Assessment, RTI Health Solutions, Research Triangle Park, NC, USA

Course Description: Course participants will learn the conceptual and empirical features of various health-utility measures and their uses for informing health care decision-making. Cost-utility analysis (CUA), risk-benefit analysis (RBA), and cost-benefit analysis (CBA) are often used to evaluate new health-care technologies. These methods are useful for informing decision-makers about the relative benefits of an intervention to individual patients and to society as a whole. CUA employs health-state utilities based on cardinal utility theory to define quality-adjusted life years (QALYs) for different health states. RBA employs utility measures to place both risks and benefits in comparable units. CBA estimates take the form of ordinal utility values expressed as money-equivalent values (often called ‘willingness to pay’). This course will review the theory and application of utility estimation in health economics and risk-benefit analysis. This course is designed for those with some experience with psychometric measures.

Real World Data Method

Outcomes Research for Medical Devices & Diagnostics

Faculty: Seema Sonnad PhD, Associate Professor, Department of Surgery, University of Pennsylvania, Philadelphia, PA, USA; Stacey Ackerman, MSE, PhD,  Vice President, Covance Market Access Services, San Diego, CA, USA

Course Description: This course will present outcomes research practices that are specifically tailored for the fast-paced medical device and diagnostics technology environment and address issues related to these health technology assessment methodologies. Outcomes research including clinical outcomes, economic outcomes, and patient-reported outcomes will be discussed. Outcomes research for medical devices & diagnostics will be differentiated from other health care interventions such as drugs. The evidence hierarchy for medical devices and diagnostic procedures including ‘real world’ outcomes research information in coverage and reimbursement decisions will be discussed. This course is designed for those with little experience with outcomes research for medical devices and diagnostic technologies.

USE OF PHARMACOECONOMICS / ECONOMIC / OUTCOMES RESEARCH INFORMATION

Introduction to Risk/Benefit Management in Health Care

Faculty: Dennis W. Raisch, PhD, Associate Center Director, Scientific Affairs, VA Cooperative Studies Program, Clinical Research Pharmacy, Albuquerque, NM, USA; Anthony Lockett, MD, PhD, MBA, Medical Director, MEDQP Ldt., Leeds, UK; Suellen Curkendall, PhD, Principal Investigator, Cerner Health Insights, Vienna, VA, USA

Course Description: This course will provide an overview of risk/benefit management for pharmaceuticals and devices. The risk/benefit assessment process will be described in regards to stage of product development, from pre-marketing through post-marketing. Risk mitigation includes the various strategies employed by manufacturers, regulators, and health care providers, with an emphasis on international differences in risk mitigation and decision-making. Risk/benefit communication processes will be described, focusing on how decisions regarding risks and benefits of pharmaceuticals and devices are communicated to health care providers and the public. This includes direct mailing, direct-to-consumer marketing, and labeling. Real world exercises will allow participants to discuss key topics and propose implementation strategies for risk management. This course is designed for those with a basic understanding of pharmacoepidemiology principles.

MODELING METHODS

Advanced Decision Modeling for Health Economic Evaluations

Faculty: Andrew Briggs PhD, Lindsay Chair in Health Policy and Economic Evaluation Section of Public Health, University of Glasgow, Glasgow, Scotland; Mark Sculpher PhD, MSc, Professor, University of York, Centre for Health Economics, Heslington York, UK

Course Description: During this course, the key aspects and new developments of decision modeling for economic analysis will be considered. How models can be made probabilistic to capture parameter uncertainty (including rationale, choosing parameter distributions, & types of uncertainty) will be covered. How to analyze and present the results of probabilistic models will be presented. How the results of probabilistic decision modeling should be interpreted and how decisions should be made (including decisions with uncertainty, and expected value of perfect information [EVPI]), will be presented. Specific examples including Excel programming will be used to illustrate concepts.  The publication “Decision Modeling for Health Economic Evaluation” (Oxford, 2006) is recommended reading for this course.  This is an advanced course.  Participants should have a basic understanding of decision analysis. The course, Modeling: Design and Structure of a Model, is a strong prerequisite for this course.