Considerations in the selection of end points for AIDS clinical trials

An important aspect of the design of an efficacy clinical trial for persons infected with the human immunodeficiency virus is the selection of end points. Ideally, the choice of the primary end points for the trial should ensure that execution of the trial is feasible, that the data are directly interpretable with respect to the therapeutic question of interest, and that the results lead to a clear and implementable decision about the treatment of this patient population or about future trials. Seven considerations in the selection of an end point are observability, relevance, definition, time needed for observation, verifiability, variability, and implementability. We discuss these considerations and illustrate the principles with an example from a clinical trial conducted by the AIDS Clinical Trials Group.