Women with chest pain: Is exercise testing worthwhile?

Objective - To determine the diagnostic value of the exercise tolerance test (ETT) in women presenting with chest pain. Design - Prospective study of all women presenting to a centre with chest pain between 1987 and 1993 who were assessed by an ETT and coronary angiography. Setting - The outpatient clinic of one consultant cardiologist in a tertiary referral centre. Patients - All women referred to this outpatient clinic with chest pain were screened. For inclusion, patients had to perform ETT and undergo coronary angiography. Of the 347 referred during this period, 142 were excluded because they were unable to perform ETT or because of Q waves or other abnormalities on their resting electrocardiogram. Results - Overall the sensitivity of the ETT was 68% and the specificity was 61%, with a positive predictive value of 0.61 and a negative predictive value of 0.68. There were 42 false positive and 31 false negative ETT results (36% of the study group). The predictive value of a negative test was higher in younger women (< 52 years) than in the older group (greater than or equal to 52 years) (P = 0.004), but the positive predictive value in the two groups was not significantly different. The predictive value of a negative test was also higher in those with two or fewer risk factors than in those with three or more risk factors (P = 0.001). The negative predictive value for those women above 52 years with three or more risk factors (24% of the study group) was only 0.25. Lack of chest pain during ETT was associated with a higher negative predictive value in the younger group than in the older women (P = 0.006). Conclusions - In women with chest pain use of the ETT was a misleading predictor of the presence or absence of coronary disease in 36% of these patients. In particular, a negative test in older women with three or more risk factors had a very low predictive value. The inclusion of risk factors and division by age can, however, be used to identify a population at intermediate risk for coronary artery disease in whom the ETT result has the highest diagnostic utility.