A Quantitative Approach for Benefit-Risk Assessment Using Stochastic Multi-Criteria Discriminatory Method

Benefit-risk (BR) assessment is important to ensure safety and effectiveness of medical projects in clinical development and post marketing surveillance. The evaluation of the balance between benefits and risks of a drug is fundamental to all stakeholders involved in the development, registration, and use of drugs. Many quantitative approaches have been developed that may draw together data from different sources to present them in ways that can aid decision-making. Among these quantitative approaches, multicriteria decision analysis (MCDA) is one of the most useful approaches, since it provides a framework for systematic and replicable analyses of complex decision problems involving value tradeoffs. In most real-life situations, decision makers are not able to get exact preference information (weights). To overcome this limitation of MCDA, a method called stochastic multicriteria acceptability analysis (SMAA) was proposed as a quantitative approach to BR assessment in drug development. The chief advantage of SMAA over most other MCDA methodologies is that it can be used with limited or no preference information and a data-driven approach can be used to obtain the inherent weighting among multicriteria. In this article, we propose the stochastic multicriteria discriminatory method (SMDM) that is based on the SMAA method with a focus on providing straightforward and informative assistance to decision making. We use SMDM to derive probability of better treatment, probability of significantly better treatment and the expected p-values of the evaluation results, on the basis of limited partial weight information or no weight information at all.