The aim of this review is to determine, among patients admitted with dyspnoea, the proportion of patients that can be excluded from having acute decompensated heart failure (ADHF) due to low N-terminal of the prohormone brain natriuretic peptide (NT-proBNP), below diagnostic cut-off, and to examine the diagnostic value of NT-proBNP in patients with high NT-proBNP levels, above cut-offs. Patients ≥40 years of age who were acutely admitted with dyspnoea were included. Of 654 patients, 194 (30%) had NT-proBNP below rule-out (<35 pmol/l ≈296 pg/ml), 157 (24%) had intermediate levels of NT-proBNP, and 303 (46%) had NT-proBNP above age-adjusted rule-in values. The negative predictive value of NT-proBNP below rule-out was 99.5% for ADHF. A severe non-cardiac condition was the primary complaint in 88 of the 157 patients with intermediate levels of NT-proBNP, and these were not considered to have ADHF. Thereby, 372 patients (56.9% of 654) were left for examination of ADHF. Certain ADHF was present in 166 (45%), plausible in 85 (23%), and no ADHF in 121 (32%). Receiver-operating characteristics (ROC) analysis for NT-proBNP to identify certain ADHF resulted in an area under the curve (AUC) of 0.69 (95% CI: 0.64-0.74, p<0.001). ROC analysis of patients with current atrial fibrillation (Afib)/flutter (AFL) resulted in AUC of 0.58 (95% CI: 0.48-0.68, p=0.8) to diagnose certain ADHF. In patients admitted with dyspnoea, low NT-proBNP levels will safely rule out ADHF in 30%. We conclude that rule-in values for NT-proBNP are inappropriate to diagnose ADHF, and in patients with Afib or AFL there is no added diagnostic value of NT-proBNP.
Please view the full content in the pdf above.