Generalizing the results of clinical trials to actual practice: the example of clopidogrel therapy for the prevention of vascular events

PURPOSE: An important element in translating the results obtained in clinical trials of a new treatment to clinical practice is the estimated event rate in patients who would be eligible to receive that treatment. We estimated the effect of clopidogrel, compared with aspirin, in actual practice using the relative risk reduction observed in the Clopidogrel versus Aspirin in Patients at Risk of Ischaemic Events (CAPRIE) trial. SUBJECTS AND METHODS: Ischemic event rates were estimated for 12,931 aspirin users drawn from the Saskatchewan Health population between 1990 and 1995 who had an index diagnosis of myocardial infarction, ischemic stroke, or peripheral arterial disease. To estimate the absolute risk reduction, the 8.7% relative risk reduction from clopidogrel compared with aspirin that was observed in CAPRIE was applied to these rates. RESULTS: The rates of ischemic events were greater in actual practice than among the control patients in the CAPRIE trial. In the Saskatchewan population, patients experienced an outcome event (myocardial infarction, stroke including intracranial hemorrhage, or death) at a rate of 15.9 per 100 patient-years, compared with only 6.9 per 100 patient-years in CAPRIE. If the same 8.7% relative risk reduction seen in the CAPRIE trial is also true for patients seen in routine clinical practice, the greater absolute risk in actual practice would reduce the number needed to treat to prevent one event from 200 patients to 70 patients. CONCLUSION: Absolute risk rates may be substantially greater in clinical practice than in the selected patients enrolled in randomized trials. As a result, similar reductions in relative risk, if true for clinical practice, may yield substantially more benefit in clinical practice than in randomized trials.