Sample Size Considerations for Pharmacoepidemiologic Studies

A Guide for Assessing Pharmacoepidemiologic Studies

“…An RCT would require more than 5555 patient-years of exposure to provide a 95% probability of observing just one additional serious gastrointestinal hemorrhage among treated patients, and a substantially larger sample size (approximately 75,000 patient-years in the exposed group and a similar number of controls) to test the causation hypothesis Statistically. 12 It is for this reason that a cohort study was more feasible to assess the suggested association.ll An association between diethylstilbestrol (DES) ingestion by pregnant women and reproductive abnormalities in their daughters many years later has been d e~c r i b e d . '~ A prospective study designed to test this relationship would require 20 years from the time the association was first suspected to completion of the study.…”