Problems in the Detection and Quantification of Adverse Drug Reactions

Abstract

In the situation where drugs are uncommonly used, and events are rare, case reports have utility in elucidating cause and effect relationships. In other situations, however, case reports can reflect coincidence, or bias, rather than cause and effect. In addition, even when valid, they provide no data as to incidence. In the situation where drugs are commonly used, and when events are also common, case reports have little or no utility, and the use of algorithms may simply serve to reinforce preexisting biases. Only rigorous epidemiologic research can document either adversity or safety.