Abstract
While randomized clinical trials (RCTs) should provide the basis for evidence-based medicine, as currently designed and analyzed, they often mislead clinical decision-making. Comparative effectiveness evaluation of two treatments [Treatment 1 (T1) versus Treatment 2 (T2)] should not be determined by the statistical effect of treatments on individual measures of outcome (benefits and/or harms), but rather on the clinical effects of treatments on individual patients who can experience both benefits and harms. Such strategies for evaluation require both methods for statistical assessment of the rates of co-occurrence of such benefits and harms, and clinical assessment of their combined clinical impact on patients. The strategies discussed here are possible solutions to this dilemma. It is crucial to develop successful strategies to assess the effects of treatments on individual patients.