Comparison of HALP and DECAF in predicting 30 day in hospital mortality in a retrospective cohort study of acute exacerbation of COPD

Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are associated with substantial short term mortality and healthcare utilization, yet the prognostic utility of the immunonutrition based hemoglobin, albumin, lymphocyte, and platelet (HALP) score in this setting remains unclear, particularly in comparison with the well validated dyspnoea, eosinopenia, consolidation, acidaemia, atrial fibrillation (DECAF) score. We conducted a retrospective, single centre observational cohort study of adults presenting to the emergency department (ED) with AECOPD between June 2022 and June 2025. DECAF and HALP scores were calculated at admission using routinely collected clinical and laboratory data, and the primary outcome was 30 day in hospital mortality. Discrimination was assessed using receiver operating characteristic (ROC) analysis, with Youden-derived, data-driven cut-off values, and we further evaluated previously validated DECAF thresholds and a categorical DECAF model. Associations with 30 day mortality were examined using univariable and multivariable logistic regression models adjusted for baseline clinical variables. Model calibration was evaluated using Brier scores and the Hosmer–Lemeshow test, and 30 day survival was analysed with Kaplan–Meier. A total of 499 patients were included in the study, 63 (12.6%) died within 30 days, and 76% of the cohort were male. The HALP score showed acceptable predictive performance (area under the curve (AUC) 0.724), at a cut-off of 16.84, sensitivity was 57.1%, specificity 75.0% and negative predictive value (NPV) 92.4% for 30 day in hospital mortality. The DECAF score demonstrated excellent discrimination for 30 day in hospital mortality (AUC 0.932), and showed significantly better discrimination than HALP (DeLong p < 0.001), using a cut-off of ≥ 4, sensitivity was 76.2%, specificity 95.6% and NPV 96.5%. In ED patients with AECOPD, the HALP score provides only limited prognostic information for 30 day in hospital mortality, whereas the DECAF score exhibits excellent discriminative performance and retains its value as a robust, clinically applicable predictor of short term outcomes in acute care.