A Critique of Methodology in Studies of Anticoagulant Therapy for Acute Myocardial Infarction

Abstract

Of 32 reports of anticoagulant therapy for acute myocardial infarction, appraised to ascertain the maintenance of eight methodologic standards that affect the scientific comparability of treated patients and controls, only 25 per cent described precise criteria for diagnosis of myocardial infarction; 59 per cent were performed as surveys of medical records rather than experimental trials, and 28 per cent did not have concurrent control groups. In most "Cooperative" studies, ancillary therapy at different hospitals was not co-ordinated. Treatment was allocated randomly in four studies,. assessed with appropriate prognostic stratifications in 17 and applied with double-blind technics in only one. No diagnostic criteria were described for thromboembolic phenomena. Anticoagulant therapy was ranked superior to none more often in reports that failed to fulfill the methodologic standards than in the reports that did. Although generally regarded as a problem in statistics, the current controversy about anticoagulants seems to have more basic scientific roots in the selection and classification of the patients who are the "experimental material.".