Machine learning to optimize use of natriuretic peptides in the diagnosis of acute heart failure

Dimitrios Doudesis

; Kuan Ken Lee

; Mohamed Anwar ; Adam J Singer ; Judd E Hollander ; Camille Chenevier-Gobeaux ; Yann-Erick Claessens ; Desiree Wussler

; Dominic Weil ; Nikola Kozhuharov ; +52 more... Ivo Strebel ; Zaid Sabti ; Christopher deFilippi

; Stephen Seliger ; Evandro Tinoco Mesquita ; Jan C Wiemer ; Martin Möckel ; Joel Coste ; Patrick Jourdain ; Komukai Kimiaki ; Michihiro Yoshimura ; Irwani Ibrahim ; Shirley Beng Suat Ooi ; Win Sen Kuan ; Alfons Gegenhuber ; Thomas Mueller ; Olivier Hanon ; Jean-Sébastien Vidal ; Peter Cameron ; Louisa Lam ; Ben Freedman

; Tommy Chung ; Sean P Collins ; Christopher J Lindsell ; David E Newby ; Alan G Japp ; Anoop SV Shah

; Humberto Villacorta ; A Mark Richards ; John JV McMurray ; Christian Mueller ; James L Januzzi ; Nicholas L Mills ; Gordon Moe ; Carlos Fernando ; Hanna K Gaggin ; Antoni Bayes-Genis ; Roland RJ van Kimmenade ; Yigal Pinto ; Joost HW Rutten ; Anton H van den Meiracker ; Luna Gargani ; Nicola R Pugliese ; Christopher Pemberton ; Michael Neumaier ; Michael Behnes ; Ibrahim Akin ; Michele Bombelli ; Guido Grassi ; Peiman Nazerian ; Giovanni Albano ; Philipp Bahrmann ; (2025) Machine learning to optimize use of natriuretic peptides in the diagnosis of acute heart failure. European heart journal. Acute cardiovascular care, 14. pp. 474-488. ISSN 2048-8726 DOI: 10.1093/ehjacc/zuaf051

Copy

Aims: B-type natriuretic peptide (BNP) and mid-regional pro-atrial natriuretic peptide (MR-proANP) testing are guideline-recommended to aid in the diagnosis of acute heart failure. Nevertheless, the diagnostic performance of these biomarkers is uncertain.

Methods and results: We performed a systematic review and individual patient-level data meta-analysis to evaluate the diagnostic performance of BNP and MR-proANP. We subsequently developed and externally validated a decision-support tool called CoDE-HF that combines natriuretic peptide concentrations with clinical variables using machine learning to report the probability of acute heart failure. Fourteen studies from 12 countries provided individual patient-level data in 8493 patients for BNP and 3899 patients for MR-proANP, in whom, 48.3% (4105/8493) and 41.3% (1611/3899) had an adjudicated diagnosis of acute heart failure, respectively. The negative predictive value (NPV) of guideline-recommended thresholds for BNP (100 pg/mL) and MR-proANP (120 pmol/L) was 93.6% (95% confidence interval 88.4–96.6%) and 95.6% (92.2–97.6%), respectively, whilst the positive predictive value (PPV) was 68.8% (62.9–74.2%) and 64.8% (56.3–72.5%). Significant heterogeneity in the performance of these thresholds was observed across important subgroups. CoDE-HF was well calibrated with excellent discrimination in those without prior acute heart failure for both BNP and MR-proANP [area under the curve of 0.914 (0.906–0.921) and 0.929 (0.919–0.939), and Brier scores of 0.110 and 0.094, respectively]. CoDE-HF with BNP and MR-proANP identified 30% and 48% as low-probability [NPV of 98.5% (97.1–99.3%) and 98.5% (97.7–99.0%)], and 30% and 28% as high-probability [PPV of 78.6% (70.4–85.0%) and 75.1% (70.9–78.9%)], respectively, and performed consistently across subgroups.

Conclusion: The diagnostic performance of guideline-recommended BNP and MR-proANP thresholds for acute heart failure varied significantly across patient subgroups. A decision-support tool that combines natriuretic peptides and clinical variables was more accurate and supports more individualized diagnosis.

Study registration: PROSPERO number, CRD42019159407.