Statistical Methodologies for Detecting Drug-Drug Interaction Using Spontaneous Reporting System

Abstract

Statistical methodologies of signal detection for drug-drug interactions (DDIs) have been developed and used. In this study, we review methodologies for DDI signal detection using the spontaneous reporting system. We describe the characteristics of the additive and multiplicative risk model, logistic regression model, Multi-item Gamma-Poisson Shrinker, <i>Ω</i> shrinkage measure model, and Chi-square statistics model. We discuss the advantages and disadvantages of each methodology along with examples, providing researchers with information to choose the most appropriate methodology for their study.