Prevalence and Descriptors of Aspirin Use as an Antiplatelet Agent in Two New England Communities

Anne L. Hume, University of Rhode Island
Marilyn M. Barbour, College of Pharmacy
Kate L. Lapane, Memorial Hospital of Rhode Island
Annlouise R. Assaf, The Warren Alpert Medical School
Richard A. Carleton, Memorial Hospital of Rhode Island


To determine the overall prevalence of aspirin use as an antiplatelet agent in a population-based sample of respondents between the ages of 18 and 64 years, and to describe their sociodemographic characteristics. Data were derived from five biennial cross-sectional household surveys conducted between 1981 and 1990 in two southeastern New England communities by the Pawtucket Heart Health Program. Individuals reporting aspirin use as an antiplatelet agent were identified using data from a structured medication interview. The prevalence of aspirin use per 1000 population was 2.5 (95 percent confidence interval [CI] 0.5,4.4), 5.0 (95 percent CI 2.4, 7.6), 7.8 (95 percent CI 4.6, 11.0), 7.5 (95 percent CI 4.3, 10.6), and 11.8 (95 percent CI 7.1, 16.5) in the first through the fifth survey periods, respectively (p=0.0002). The prevalence of aspirin use per 1000 men increased from 4.9 (95 percent CI 0.6, 9.1), 8.7 (95 percent CI 3.3, 14.1), 12.3 (95 percent CI 6.1, 18.5), 8.5 (95 percent CI 3.5, 13.5) to 23.2 (95 percent CI 13.4, 32.9) for the first through the fifth survey periods, respectively. The corresponding prevalence rates of aspirin use per 1000 women were 0.7 (95 percent CI 0, 2.1), 2.43 (95 percent CI 0.05, 4.80), 4.6 (95 percent CI 1.4, 7.8), 6.6 (95 percent CI 2.7, 10.6), and 2.7 (95 percent CI 0, 5.7). Both a gender trend (p=0.0002) and a survey trend (p=0.0001) were detected. A gender-survey interaction was not found. In this population, aspirin use was reported primarily by middle-aged men who had coexisting cardiovascular disease as indicated by concurrent medications. Aspirin use as antiplatelet therapy increased in the population-based samples of two New England communities over the past nine years, with slightly more than one percent of the respondents reporting its use in the fifth cross-sectional survey conducted in 1989–1990. This finding may reflect the publication of important clinical trials near the time the survey was performed and subsequent dissemination of the findings both in the professional and lay press. © 1993, SAGE Publications. All rights reserved.