Lactate/albumin ratio and subsequent in-hospital septic shock among adults with necrotizing fasciitis: a single-center retrospective cohort with a supportive multicenter analysis using eICU-CRD

Necrotizing fasciitis is a rapidly progressive soft-tissue infection in which early identification of patients at risk of septic shock may inform monitoring intensity, intensive care unit triage, and perioperative management. The lactate/albumin ratio has attracted interest as a simple composite marker of acute illness severity, but its association with subsequent septic shock in confirmed necrotizing fasciitis and, importantly, whether the albumin component contributes prognostic information beyond lactate alone remain unclear. We conducted a retrospective single-center cohort study of adults with surgically and/or pathologically confirmed necrotizing fasciitis treated between January 2015 and January 2025. The primary exposure was admission lactate/albumin ratio (LAR), calculated from the initial laboratory assessment. The primary local outcome was incident clinically adjudicated in-hospital septic shock occurring after hospital admission. Multivariable logistic regression adjusted for age, sex, diabetes history, and American Society of Anesthesiologists class was used to assess the local association. Discrimination was evaluated using receiver operating characteristic curves. Component-based models using lactate alone, albumin alone, and lactate plus albumin were fitted to examine whether the ratio provided information beyond its individual components. A supportive multicenter analysis was conducted in a biomarker-available eICU Collaborative Research Database subcohort to assess the directional consistency of the observed association. The local cohort included 91 patients, of whom 62 developed in-hospital septic shock after admission. Admission lactate/albumin ratio was higher in patients who developed septic shock than in those who did not. Each doubling of admission lactate/albumin ratio was independently associated with higher odds of septic shock after adjustment for prespecified covariates (odds ratio 1.87, 95% confidence interval 1.20–2.91; P = 0.006). The lactate/albumin ratio showed moderate discrimination in the local cohort (area under the curve 0.738, 95% confidence interval 0.635–0.841), similar to lactate alone (area under the curve 0.731, 95% confidence interval 0.625–0.836). Component-based analyses showed that lactate alone remained independently associated with septic shock, whereas albumin showed no clear independent association, and adding albumin to lactate did not materially improve model performance. In a biomarker-available eICU-CRD subcohort of 84 patients from 42 hospitals, early intensive care unit lactate/albumin ratio was positively associated with diagnosis-based septic shock (odds ratio 2.50, 95% confidence interval 1.42–4.41; P = 0.001). Admission LAR was independently associated with subsequent in-hospital septic shock in adults with necrotizing fasciitis, but its discriminatory performance was similar to that of lactate alone. Component-based analyses indicated that the prognostic signal was largely driven by lactate, with no clear incremental contribution from albumin. These findings do not support preferential use of LAR over lactate for early risk stratification in this setting.