Prognostic Value of HALP and AHEAD Scores for Predicting 1-Month Heart Failure Following Myocardial Infarction

Le résumé fourni par la source

Background: Heart failure (HF) remains a major early complication following myocardial infarction (MI), contributing significantly to morbidity and adverse clinical outcomes. Reliable early risk stratification is essential for optimizing post-MI management. This study aimed to evaluate the prognostic performance and incremental value of the HALP (Hemoglobin–Albumin–Lymphocyte–Platelet) score and the AHEAD score in predicting 1-month HF after MI. Methods: This retrospective cohort study included 3205 consecutive patients with MI. The primary endpoint was the development of HF within one month. Three multivariable logistic regression models were constructed: a baseline clinical model (Model 1), a HALP-integrated model (Model 2), and an AHEAD-integrated model (Model 3), with component variables excluded to avoid collinearity. Model performance was assessed using odds ratios (ORs), 95% confidence intervals (CIs), and discrimination metrics (AUC). Incremental predictive value was further evaluated using net reclassification improvement (NRI). Internal validation was performed using bootstrapping and 5-fold cross-validation. A predefined subgroup analysis was conducted in patients with preserved ejection fraction (EF ≥ 40%), excluding EF from the models. Results: In the full cohort, all models demonstrated high discriminative ability for 1-month HF (AUC range: 0.950–0.954), with minimal differences between models. The AHEAD-based model showed the highest point estimate (AUC = 0.954, 95% CI: 0.944–0.963), but ROC curves were largely overlapping. Despite limited changes in AUC, the AHEAD score provided moderate improvement in risk reclassification (NRI = 0.287), whereas the HALP score showed minimal incremental value (NRI = 0.152) and was not independently associated with HF in multivariable analysis. In the EF ≥ 40 subgroup, HF incidence was lower (1.9%), and model performance was attenuated but remained robust (AUC range: 0.839–0.882), with the AHEAD score retaining strong independent predictive value. Peak CKMB and creatinine were consistently associated with increased HF risk. Although the odds ratio for CKMB appeared close to unity, this reflects unit scaling, and clinically meaningful increases corresponded to substantial risk increments. A clear dose–response relationship between AHEAD score and HF probability was observed. Conclusions: While both HALP and AHEAD scores are associated with post-MI HF risk, only the AHEAD score provides consistent independent and incremental prognostic value beyond established clinical predictors. Its simplicity and ability to capture comorbidity burden make it a practical adjunct for early risk stratification, particularly in patients with preserved EF. However, given the minimal differences in discrimination metrics and lack of external validation, these findings should be interpreted cautiously and considered hypothesis-generating.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.