In Britain, five leading groups of experts have issued joint treatment recommendations in patients at risk for heart disease. These recommendations include the initiation of aspirin and antihypertensive therapy when the risk of heart disease is 15 percent over 10 years. The use of statin drugs is also recommended when coronary risk is 15 percent over 10 years.

These and other recommendations depend on an ability to estimate risk in individual patients on a routine basis and to accurately identify the target group of patients with 15 percent risk over 10 years. The Sheffield table, which is based on the Framingham risk function, is one of the tools used to estimate coronary risk for primary prevention.

Wallis and colleagues modified the Sheffield table by basing it on the ratio of total cholesterol to high-density lipoprotein (HDL) cholesterol and age, sex, smoking, diabetes and hypertension (see the accompanying table on page 1146). They tested this version of the table in a random sample of 1,000 adults living in Scotland.

The sample was selected to be representative of the Scottish adult population. Risk was assessed for each participant by two physicians. For each participant, the risk of cardiovascular disease was calculated using data from the Framingham studies to provide a gold standard. The accuracy of the Sheffield tables in predicting coronary risk was estimated by comparison with the Framingham estimates.

The study group included 562 (56.2 percent) women, 299 (29.9 percent) smokers and 16 (1.6 percent) patients with diabetes. The mean age was 49 years, mean blood pressure was 132/75 mm Hg and the mean 10-year coronary risk function by Framingham function was 7.2 percent. The Sheffield table had a 97 percent sensitivity and 95 percent specificity for coronary risk of 15 percent or greater over 10 years. Negative tests had a predictive value of 99.5 percent; positive tests had a predictive value of 73 percent. The false-positive results all had risks in the range of 10 to 15 percent.

The authors conclude that the simple table correctly identified about 97 percent of patients most likely to benefit from aspirin and statin drugs in primary prevention of coronary disease. While the Sheffield table identified patients who definitely should be offered treatment, the authors point out that it should not be used to deny treatment to people close to treatment thresholds.

editor's note: This article reflects a growing trend in British general practice literature to provide practical advice and tools for practitioners to use in primary prevention of the major causes of morbidity and mortality. This issue of BMJ contains 10 articles on such topics and several examples of charts and techniques to calculate level of risk. The Sheffield tables are easy to use and, as shown by the study, can accurately identify asymptomatic candidates for intervention. Interestingly, the British authors place emphasis on avoiding the expense of unnecessary testing in low-risk patients and identifying the high-risk group. Besides helping to strategize ways to avoid coronary disease, using the tables with patients might also have a powerful educational effect. The “big three”—hypertension, smoking and diabetes—dominate the tables. Patients may quickly appreciate how avoiding or controlling certain risk factors directly and substantially reduces the risk of heart disease.—a.d.w.