Two Statistical Approaches to Incorporate Data Uncertainty into Multiple Criteria Decision Analysis (MCDA) for Benefit-Risk Assessment of Medical Products

• Activity Number: 680
• Type: Topic Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Shihua Wen*
• Companies: AbbVie
• Keywords: benefit-risk ; Multiple Criteria Decision Analysis

Abstract

PrOACT-URL framework and multiple criteria decision analysis (MCDA) have been recommended by the European Medicines Agency for structured benefit-risk assessment of medicinal products undergoing regulatory review. To incorporate the uncertainty from clinical data into the MCDA model, two statistical approaches, the delta-method approach and the Monte-Carlo approach, were proposed. The delta-method approach provides a closed form solution to quantify the variability of the overall benefit-risk score in the MCDA model; while the Monte-Carlo approach is more computationally intensive but can yield its true sampling distribution for further statistical inference. The obtained confidence interval of the overall benefit-risk score based on the MCDA model as well as other probabilistic measures from the two proposed approaches enhance the benefit-risk decision-making among different treatment options. In addition, a case study of comparing the benefit-risk profiles between a hypothetical drug and placebo was conducted to demonstrate the usage of the two approaches.