In recent years, there has been increasing level of interest and publications in methodologies to carry out structured quantitative Benefit Risk assessments to facilitate the interpretation and decision making process in clinical trials.
Often, the interest is to identify a sub-group of population that optimize benefit-risk index. Currently the approach is to carry out in a univariate approach by focusing either on primary efficacy or safety endpoint. This approach does not consider the totality of information at the same time. In this presentation, we introduce an algorithm which allows calculating patient level contribution to benefit risk index. Subsequently, one's method of choice for sub-group identification, such as tree based methods can be implemented. Example for the proof of concept will be provided utilizing Multi-Criteria Decision Analysis (MCDA) benefit-risk index.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.