Admission NT-proBNP provides stronger prognostic discrimination than the AHEAD score for 1-year mortality in hospitalized acute heart failure: A retrospective cohort study

Abstract

Background Both admission N-terminal pro-B-type natriuretic peptide (NT-proBNP) and the AHEAD score predict prognosis in acute heart failure, but their comparative and complementary value for admission risk stratification remains uncertain. Methods We screened 512 consecutive adult hospitalizations for acute heart failure; 478 records had sufficient baseline data, and 430 patients had ascertainable 1-year vital-status follow-up and constituted the analytic cohort. We compared admission NT-proBNP (log-transformed) with the AHEAD score for 1-year all-cause mortality using Cox models, Harrell C-index, apparent calibration, and reclassification (continuous net reclassification improvement [NRI] and integrated discrimination improvement [IDI]). A combined model of both markers and a combined AHEAD x NT-proBNP stratification were also evaluated. Results During 1 year, 84 deaths (19.5%) occurred. ln(NT-proBNP) was strongly associated with mortality (adjusted HR 2.63, 95% CI 2.05–3.37 per 1-unit increase; approximately HR 1.95 per doubling). AHEAD categories were associated with mortality in univariable analysis (HR 1.95 for score 2 and 3.61 for score ≥ 3 vs 0−1), but were attenuated after adjustment for ln(NT-proBNP) and admission covariates (adjusted HR 1.03 and 1.81). ln(NT-proBNP) showed higher discrimination than AHEAD categories (Harrell C-index 0.758 vs 0.608). The combined model improved discrimination and reclassification compared with AHEAD alone (C-index 0.757; Delta C-index 0.150; continuous NRI 0.840; IDI 0.136), but not compared with ln(NT-proBNP) alone (Delta C-index 0.000). In combined stratification, the highest NT-proBNP tertile (T3; > 6,385 pg/mL) identified high-risk groups regardless of AHEAD category. Conclusions In hospitalized acute heart failure, admission NT-proBNP provided stronger prognostic discrimination than AHEAD categories for 1-year mortality in this cohort. AHEAD may still provide complementary clinical context, but adding AHEAD to ln(NT-proBNP) did not materially improve discrimination beyond NT-proBNP alone. External validation is warranted.