Common methods of analyzing response data in clinical trials.

In evaluating published reports of clinical trials, look for clearly stated objectives; a well-defined patient population; avoidance of systematic selection or allocation of patients in a way that could bias results; an accounting for all enrolled patients; and definitions of the endpoints of interest that are both precise and standard across similar studies. In statistical terms, the outcomes observed in a trial are used to estimate treatment effects or test hypotheses about unknown and inherently unobservable outcomes among all patients with a given disease. Several measures are used to assess a therapy's effectiveness, but the simplest to analyze is the objective response of the disease measured immediately after completion of therapy. The authors describe methods used in the analysis of data, including techniques with versions for data with more than two response categories.