Aims: The study evaluated the use of age-related decision limits for N-terminal pro-B-type natriuretic peptide (NT-proBNP), for ruling out suspected systolic dysfunction in symptomatic patients in primary care, compared with the present standards.
Methods and results: Data were obtained from 5508 patients from 10 studies in the UK, New Zealand, Europe, and USA. All have had NT-proBNP analysis and echocardiography. The median age was 62 years (range 18-100 years) with a prevalence of reduced left ventricular systolic function (left ventricular ejection fraction < or =40%) of 18%. In a receiver operating characteristic curve analysis, overall area under the curve (AUC) was 0.89. When looking at different age groups, AUC was highest (0.95) for <50 years, intermediate (0.90) for 50-75 years, and lowest (0.82) for >75 years. Using optimized decision limits, sensitivity, specificity, and negative predictive values (NPVs) were: <50 years (50 ng/L): 99.2, 57.2, and 99.7%; 50-75 years (75 ng/L): 95.9, 51.0, and 96.8%; and >75 years (250 ng/L): 87.9, 53.7, and 92.4%, respectively. Using only a single decision value (125 ng/L for all ages) gave sensitivities of 89.1, 91.9, and 94.3%; specificities of 84.0, 69.1, and 29.3% and NPVs of 97.7, 97.6, and 93.4%. A decision value of 400 ng/L for all ages gave much lower sensitivities.
Conclusion: In a large population of patients in primary care, the use of age-stratified NT-proBNP decision limits considerably improves performance over current standards, with an excellent NPV for exclusion of reduced left ventricular systolic function.