We thank Dora, Kramer, and Canani (1) for their interest in the Standards of Medical Care in Diabetes—2008 (2) and their critique of the recommendations for aspirin for primary prevention of cardiovascular disease (CVD). The American Diabetes Association's “A”-level evidence also encompasses supportive evidence from a meta-analysis incorporating quality ratings. Supportive evidence can derive from studies in people without diabetes but at similar CVD risk.
The systematic review of evidence for the U.S. Preventive Services Task Force estimated that aspirin reduced the risk for nonfatal and fatal myocardial infarction (odds ratio 0.72 [95% CI 0.60–0.87]). The review acknowledged the low numbers of diabetic subjects in most trials but concluded that subset analyses and a single trial in diabetic patients suggested that the estimates extended to those with diabetes (3). The task force stated that the risk-to-benefit ratio favors aspirin use when 5-year coronary heart disease risk equals or exceeds 3% and suggested that aspirin therapy be considered for men >40 years of age, postmenopausal women, and younger people with coronary heart disease risk factors (including diabetes) (4).
In Dora, Kramer, and Canani's described population, all but the low-risk patients (in total 75–86% of their population) would seem to fall into a 5-year risk level of ≥3%. Additionally, their estimates of CVD risk in a Brazilian population may differ from those of other populations. Bax et al. cited an overall annual mortality rate of 1–3% in U.S. type 2 diabetic patients and 5-year myocardial infarction rates of 7–8% in several studies of newly diagnosed type 2 diabetic patients (5).
Clinical Practice Recommendations are based on studies of populations of patients, which include ranges of baseline risk and of estimated probabilities of benefit and harm from an intervention. There will always be some patients in whom baseline risk of the outcome is low enough to tip the harm-benefit ratio of the recommendation to unfavorable; therefore, guidelines must be individualized.
The Professional Practice Committee annually examines the positions of the American Diabetes Association and considers changing recommendations (or levels of evidence) based on new evidence or on feedback from providers and will include Dora, Kramer, and Canani's critique in this year's deliberations. We will also consider emerging evidence suggesting that low-dose aspirin may be less beneficial for people with diabetes than those without (6), which would certainly impact estimates of the risk-to-benefit ratio of aspirin prophylaxis in people with diabetes.
