Abstract / Summary
Sepsis-associated liver injury (SALI) is an independent predictor of mortality in adult sepsis, but its association with outcomes in children has not been systematically synthesized. We pooled SALI-associated mortality and SALI proportions and summarized candidate biomarker performance. In this systematic review and meta-analysis (PROSPERO CRD420261387360), PubMed, Scopus, MEDLINE (via EBSCO), Cochrane, and ClinicalTrials.gov were searched from inception to 21 May 2026 without language restrictions; Embase was not searched. Eligible studies enrolled children aged one month to 18 years with consensus-defined sepsis and SALI-stratified outcomes. Two reviewers independently screened records and extracted data. Risk of bias was assessed using the prespecified Newcastle–Ottawa Scale and the Quality in Prognosis Studies tool. Mortality odds ratios were pooled using Mantel–Haenszel random effects, and SALI proportions using a logistic-normal generalized linear mixed model. Certainty of evidence was assessed with GRADE. Ten studies were included. SALI was associated with higher unadjusted mortality odds (five studies; in-hospital mortality in three, PICU mortality in one, and unspecified timing in one; 1,717 children; OR 3.88, 95% CI 2.21–6.80; prediction interval 0.63–23.80; Hartung–Knapp 95% CI 1.65–9.12). Adjusted mortality analyses were reported in two studies. Saini et al. reported that SALI remained associated with mortality after adjustment for selected clinical covariates, but the reported mortality model did not include pSOFA or another validated global severity score. Of the two studies, only Koca et al. adjusted hepatic markers for validated pediatric severity scores (PRISM III and PELOD-2); neither ALT nor total bilirubin remained independently associated with mortality. The pooled SALI proportion was 29% (seven studies; 1,876 children; 95% CI 17–46%). APRI showed AUROCs of 0.83–0.89 in two cohorts, but thresholds differed substantially and overlap between the AST-based index and an ALT-defined outcome limits interpretation. SALI was associated with higher unadjusted mortality odds in children with sepsis, but evidence certainty was very low and the prediction interval crossed unity. Whether hepatic dysfunction adds prognostic information beyond validated illness-severity measures remains uncertain. Heterogeneity in SALI proportions supports standardized pediatric definitions and ascertainment schedules. Candidate biomarkers require prospective multicenter validation before clinical use. PROSPERO CRD420261387360. Not applicable.