Chronic liver injury scores are superior prognostic indicators of outcomes in severe alcohol-related burns.

Acute and chronic alcohol abuse are common among burn patients and may be associated with chronic liver injury, a potential factor influencing outcomes. This study evaluates the predictive power of the blood alcohol concentration (BAC) and non-invasive liver fibrosis scores and their applicability in burn patients. A retrospective analysis was conducted on patients admitted to a high-volume supraregional burn center in Northern Germany between 2007 and 2024. Patients were categorized based on their BAC at admission: low (< 100 mg/dL) vs. high (≥ 100 mg/dL). Data collected included demographics, comorbidities, and outcomes. Non-invasive liver fibrosis markers such as the Fibrosis-4 (FIB-4) score, aspartate transaminase-to-platelet ratio index (APRI) and non-alcoholic fatty liver disease (NAFLD) fibrosis score were applied to both groups. Among 121 large-surface burn patients (mean total body surface area: 16.4%), no significant differences were observed between BAC groups in demographics, comorbidities, or ICU admission rates. The serum ethanol concentration showed no significant predictive value for mortality (AUC = 0.515). In contrast, the FIB-4 score (AUC = 0.781) and APRI (AUC = 0.736) demonstrated strong prognostic accuracy. In multivariate analysis, the Abbreviated Burn Severity Index (OR = 2.42; p = 0.001), serum albumin (OR = 0.29; p = 0.016), and FIB-4 score (OR = 1.50; p = 0.033) emerged as independent predictors of mortality. Propensity score matching analysis confirmed that BAC was not associated with increased mortality after adjustment for burn depth and extent. Non-invasive liver fibrosis markers, such as FIB-4 score, provide valuable prognostic insights in burn patients, independent of acute alcohol intoxication and should be considered a routine screening tool for large surface burn patients. Incorporating chronic liver dysfunction into existing burn severity models may enhance risk stratification and outcome prediction.