Sample Size and Postmarketing Surveillance

Abstract

The detection of rare or unusual adverse effects by postmarketing surveillance is a topic of discussion among health professions, industry, the FDA, and Congress. While postmarketing surveillance studies have been performed and results published, one of the criticisms surrounding these studies is the sample size necessary to elicit meaningful results. Using binomial probability, we offer a table that includes levels of probability and occurrence rates and explain its use using a non-statistical analogy.