Potential impact of exclusion criteria on results of hypertension trials.

Recent trials of antihypertensive therapy, including the Veterans Administration trials, the Hypertension Detection and Follow-up Program, the Multiple Risk Factor Intervention Trial, the Australian Mild Hypertension Trial, and the British Medical Research Council Trial, are reviewed with a particular emphasis on the criteria leading to the exclusion of potentially eligible participants. The observation of all-cause and cause-specific mortality rates in the group ultimately selected to participate in the trial is suggested as an index to the general applicability of trial results. Because end-point rates are fundamental for determining sample size, substantial reduction in these end-point rates by patient exclusion should be taken into account by the trial design. Some recent trials may have generated end-point-event rates so low that the power of the trial to detect reasonable treatment effects was substantially reduced. Future trials should attempt to take this important factor into account at the design stage.