Integration of B-type natriuretic peptide levels with clinical data and exercise testing for predicting coronary artery disease.

Natriuretic peptides have been shown to be high in patients with myocardial ischemia. We sought to create a diagnostic score using clinical data, stress testing, and B-type natriuretic peptide (BNP) levels to improve noninvasive prediction of coronary artery disease (CAD). Patients with stable angina pectoris and normal systolic left ventricular function were eligible for this prospective cohort study. Patients with arrhythmias, valvular heart disease, impaired left ventricular function, or renal dysfunction were excluded. All patients underwent clinical evaluation, bicycle stress testing, BNP testing, and coronary angiography. Then a diagnostic risk score was derived that combined cardiovascular risk factors, results of exercise testing, and BNP measurements and added 1 point for the presence of each of these variables. Seventy-one patients (52 years of age, range 31 to 61; 46 men) were included in the study. Prevalence of CAD, defined by 50% narrowing of > or =1 coronary artery on coronary angiography, was 45%. For 0 point in the risk score system, the negative predictive value was 93% with a negative likelihood ratio of 0.1 (95% confidence interval [CI] 0.02 to 0.38); for a score of 3 points, the positive predictive value was 93% with a positive likelihood ratio of 15.9 (95% CI 2.19 to 114.7). Serum BNP level >50 ng/L at rest was the best single diagnostic parameter, with 66% sensitivity and 97% specificity, and a positive likelihood ratio of 25.6 (95% CI 3.64 to 180) and a negative likelihood ratio of 0.35 (95% CI 0.22 to 0.57). In conclusion, a diagnostic score combining exercise testing, clinical data, and serum BNP values at rest can distinguish patients with CAD from those without CAD with high accuracy.