A Guide for Assessing Pharmacoepidemiologic Studies

Abstract

Pharmacoepidemiologic studies are appearing with increasing frequency in clinical journals, describing associations between commonly used drugs and rare adverse clinical events. Before practitioners accept conclusions of causality and alter patterns of practice based on these studies, they should perform two steps: apply a critical appraisal filter to the evidence to determine whether the claims are justified, and explore the consequences of not using a drug associated with an undesired effect. This guide for critically appraising the literature on adverse drug effects emphasizes the need to evaluate the potential biases associated with studies that use nonexperimental designs.