Keywords: benefit-risk, patient preferences
Motivated by PDUFA VI, the 21st Century Cures Act and ICH’s revised expectations for drug submissions, Janssen R&D routinely implements structured, patient-focused benefit-risk (B-R) assessments and patient preference studies for medicines development programs. This talk reviews pragmatic process, methodological, and statistical considerations of structured benefit-risk assessment and demonstrates these points with a case study. Structured B-R approaches are applied to the selection and analysis of key endpoints in the development programs. The approach typically begins with the creation of a value tree and ends with a formal assessment to inform regulatory submissions. While selecting endpoints may be straight forward, there are situations where statistical issues require careful consideration when collectively considering benefits and risks. For example, the populations examined (ITT vs. Safety populations), the potential for double-counting of events, and the unit of analysis (rates vs. proportions vs. hazard ratios) must utilize best practices to avoid misleading or less effective B-R assessment. These complexities often lead to the need for a B-R statistical analysis plan, which Janssen implements on a regular basis. Examples of benefit-risk assessment from inception through submission, including important best practices relating to statistical considerations and collaboration will be shared. How and when to include patient preference studies to provide context about acceptable benefit-risk tradeoffs will also be described.