Nt-Probnp In Patients With Acute Dyspnoea Admitted To The Emergency Department: Which Cutoff Should We Use?

The NT-proBNP determination plays a key role in the emergency setting to rule out acute heart failure (AHF) in patients presenting with acute dyspnoea. Particularly, in the PRIDE study using 450 ng/L as cutoff the marker showed a negative predictive value (NPV) of 99%. In this retrospective study, we evaluated 498 emergency department (ED) patients with acute dyspnoea (248 males; mean age, 77 +/- 12 years) by NT-proBNP determination (Roche Elecsys 2010) at admission. The final diagnosis of AHF was determined on the basis of clinical history, physical examination, electrocardiogram and chest X-ray, but without knowledge of NT-proBNP results. 306 patients (61.4%) were diagnosed as AHF and 192 as having dyspnoea of noncardiac origin. NT-proBNP was >450 ng/L in 295 AHF patients (sensitivity, 96.4%), while 102 patients without AHF had NT-proBNP <450 ng/L (specificity, 53.1%). NPV and positive predictive value (PPV) were 89.4% and 76.3%, respectively. As the obtained NPV was lower than that reported in the PRIDE study, we also evaluated the diagnostic accuracy of NT-proBNP using the cutoff of 300 ng/L, as suggested by the ICON study. Using this lower cutoff, NT-proBNP was positive in 300 AHF patients (sensitivity, 98.0%) and negative in 81 patients without AHF (specificity, 42.2%). NPV and PPV were 92.0% and 72.7%, respectively. While sensitivities obtained at two different cutoffs were not significantly different (P=0.32), the specificity using 450 ng/L was significantly higher (P=0.04). In conclusion, we confirmed that, using NT-proBNP for ruling out AHF in ED, 450 ng/L as cutoff performs better than 300 ng/L, showing similar NPV but improved specificity.