SEPSIS-INDUCED COAGULOPATHY ATTENUATES THE PROGNOSTIC VALUE OF THE HEMOGLOBIN–ALBUMIN–LYMPHOCYTE–PLATELET (HALP) SCORE IN SEPSIS

BACKGROUND: The hemoglobin-albumin-lymphocyte-platelet (HALP) score predicts outcomes in sepsis, but its platelet denominator is also a core component of sepsis-induced coagulopathy (SIC), and may distort its prognostic meaning. We examined whether SIC modifies HALP's association with 28-day mortality. METHODS: We analyzed 6,218 adults meeting Sepsis-3 criteria in MIMIC-IV. HALP was calculated from first ICU measurements; SIC was defined by the 2017 ISTH criteria (score ≥4). Analyses included multivariable regression, interaction testing, propensity-score methods, and platelet-independent coagulopathy definitions, and evaluation of modified HALP (m-HALP; platelets in the numerator). RESULTS: Twenty-eight-day mortality was 25.6%; SIC incidence was 26.1%. HALP showed a J-shaped association with mortality (P for nonlinearity <0.001). The lowest quartile independently predicted death (adjusted OR 1.43, 95% CI 1.20-1.70; HR 1.24, 95% CI 1.10-1.39). This association persisted in non-SIC patients (OR 1.51, 95% CI 1.28-1.79) but was attenuated and non-significant in SIC patients (OR 1.14, 95% CI 0.82-1.60; interaction P=0.006). Albumin carried HALP's discrimination in non-SIC patients, whereas platelets dominated in SIC (OR 0.51 per SD, P=0.001). When coagulopathy was redefined without platelets, the modification was no longer significant (P=0.295), indicating it partly reflects the shared platelet component. m-HALP was inversely associated with mortality in both strata (interaction P=0.691). Results were robust to weighting, 90-day mortality, and alternative Cox models. CONCLUSIONS: HALP is an independent prognostic marker mainly in septic patients without coagulopathy; once SIC develops, thrombocytopenia both raises mortality risk and arithmetically inflates HALP. Coagulation status should be assessed before HALP-based risk stratification; m-HALP may offer a platelet-consistent alternative.