Beta-Blockers as Primary Prevention of Coronary Disease

McLaughlin, Thomas

Abstract

Several large clinical trials have demonstrated that beta-blockers reduce the incidence of non-fatal myocardial infarctions and total mortality in patients who have already survived a myocardial infarction. Do beta-blockers confer the same sort of protection on hypertensive patients who do not have a history of coronary heart disease? This proposal describes a population-based, case-control study designed to answer this question. The setting is Group Health Cooperative (GHC). Using discharge abstracts and the death-registry files will enable us to identify 600 patients who have a history of pharmacologically treated hypertension and who present with first events of coronary heart disease (CHD), defined as sudden death, myocardial infarction, or unstable angina. Pharmacy records at GHC will serve as the source of a random sample of all patients who have received treatment with anti-hypertensive agents. From this list of potential controls, frequency matching will control for the possible confounding effects of age, sex, and year of presentation. Review of the out-patient medical records, performed by an assistant blind to the hypothesis, will ensure that all study subjects meet the same entrance criteria. This effort will also secure information about potential confounding factors. The computerized pharmacy records from the Boston Collaborative Drug Surveillance Project will provide a complete, unbiased history of exposure to beta-blockers for each study subject. Data analysis will involve the techniques of stratification and logistic regression. Such a study is practicable only at a few institutions in the USA other than GHC. Reviewing records from 1979-1984 at GHC will identify a number of cases adequate to demonstrate primary prevention if it exists (80% power to detect a relative risk of 0.7). Secondly, GHC is a closed system. The identification of cases there--hypertensive patients who present with first events of CHD--suggests a natural group of controls--other hypertensive patients at GHC. The use of population-based controls enhances the validity of any results from this study. Were this study to show that beta-blockers are effective in preventing CHD, these results would argue for replacing diuretics with beta-blockers as first-line agents in the treatment of hypertensive patients without a contra-indication to beta-blockade. In the long run, this shift in treatment patterns would reduce hospitalizations for CHD and, from the point of view of the individual, reduce both the morbidity and mortality attributable to hypertension.