HOW TO DEAL WITH MODEL UNCERTAINTY AT THE PLANNING STAGE OF COST-EFFECTIVENESS ANALYSIS?- TIPS FROM GLOBAL EXPERIENCES
Author(s)
Joe Caputo, BSc., Vista Health Pte Ltd., Singapore, Singapore; Emiko Yoshida, MSc., Healthcare to All, Tokyo, Japan; George Papadopoulos, BSc(Hons), GradDipEpi, Lucid Health Consulting Pty. Ltd., NSW, Australia
Presentation Documents
ISSUE: At delivering Cost-Effectiveness Analysis (CEA), there are two uncertainties to consider. Parameter uncertainty has been addressed in various types of sensitivity analysis. On the other hand, there is no systematic approach to deal with model uncertainty. At the trial CEA in Japan, model uncertainty was found to be one of the biggest issues to be tackled before a broad introduction of CEA in Japan. At the CEA committee and at Chuikyo, to reduce discrepancy in CEA results, an agreement on analysis concepts between the CEA committee and the manufacturer prior to start an analysis was suggested. What would be the key points to agree on those concepts of a CEA? What is the successful agreement and how industry needs to prepare for that discussion?
OVERVIEW: Model structure is decided based on several key concepts of the analysis, such as perspective, time horizon, and comparator. Based on the concepts and model structure, data source including optimal measurement method of the outcomes are decided. A conceptual model and practical model could be different according to discrepancy in a set of reasonable parameters which are based on appropriateness in terms of the analysis concept, and data availability. From the manufacturer perspective, it is the most important to properly address the value of the new product. We would like to propose an Issue Panel, in which we could address issue in model uncertainty, and learn from over twenty years of global CEA experience, especially in the UK and Australia, also from the latest practices in Asian countries.
Conference/Value in Health Info
2018-09, ISPOR Asia Pacific 2018, Tokyo, Japan
Code
IP18
Topic
Economic Evaluation, Methodological & Statistical Research