Drug treatment of intermittent claudication: a critical analysis of the methods and findings of published clinical trials, 1965-1985.

1. All trials of drug therapy for intermittent claudication published in English during the period 1965-1985 were reviewed. A total of 75 trials had studied 33 different pharmacological agents. Treadmill exercise, the most reproducible method of evaluating symptoms in this condition, was used in 49% of trials. 2. Oxpentifylline was the drug that had been most frequently studied. In seven placebo-controlled trials the average response to oxpentifylline, compared with placebo and weighted for sample-size, was 65% improvement in claudication distance. There was, however, a significant negative relation between sample-size and response (rs = -0.79, P less than 0.05), suggesting that this estimate was likely to have been biased by non-publication of negative results. 3. One third of all trials were uncontrolled; 84% of these reported benefit from drug treatment, compared with 32% of placebo-controlled trials (P less than 0.001). Sample-sizes varied from seven to 227 patients; 31% of trials reported data from less than 20 patients and these were likely to have had insufficient statistical power. 4. Overall, 57 of the 75 trials (76%) had at least one of the following deficiencies: an uncontrolled design; not double-blind; failure to use treadmill exercise; less than 20 patients included in the analysis. Thus, a priori three-quarters of all trials were unlikely to have made a satisfactory assessment of drug efficacy. 5. The information available does not establish convincingly that any drug consistently improves exercise performance in intermittent claudication. In view of the deficiencies in previous trials, we propose guidelines for future studies with regard to trial design, sample-size and methods of exercise testing.