Assessing Efficacy Data

Abstract

A drug must demonstrate efficacy to be given marketing approval. That is, it must provide therapeutic benefit to patients with the disease or condition of clinical concern. Efficacy is assessed in two ways: there must be compelling evidence of both statistically significant efficacy and clinically significant efficacy. The former is addressed via formalized hypothesis testing involving the creation of a research question, a research hypothesis, and the null hypothesis. The latter is addressed via the placement of confidence intervals around the drug’s treatment effect. While important, the demonstration of statistically significant efficacy alone does not equate to the demonstration of a biologically and medically important degree of efficacy. Statistical significance addresses the reliability of the treatment effect, while clinical significance addresses the magnitude of the treatment effect, i.e., its degree of therapeutic benefit.