Can dropout and other noncompliance be minimized in a clinical trial? Report from the Veterans Administrative National Heart, Lung and Blood Institute cooperative study on antihypertensive therapy: mild hypertension.

Long-term clinical trials must face the problem of participants who drop out before the study is closed or who in other ways do not comply with the protocol. In a joint Veterans Administration-National Heart, Lung and Blood Institute study of mild hypertension, 1,012 men and women, 21 to 50 years of age with diastolic pressure from 85 to 105 mm Hg, were randomized into two double-blind treatment groups and followed for up to 30 months. The data were analyzed for factors related to dropout and to medication and visit noncompliance. Large differences in dropout and compliance rates were found by clinic and age but not by treatment allocations (active vs. placebo) nor race-sex. In addition, noncompliance was 78%-118% higher in participants who subsequently chose to drop out of the study. Indications were that the attitude of a clinic towards participant complaints, willingness to temporarily reduce medication and vigorous pursuit of those who failed to keep appointments were important factors in reducing dropout. Strategies for minimizing dropout and noncompliance must be part of the study protocol, clinic personnel should be trained in using such strategies, and both clinic and participant compliance should be centrally monitored.