This is the authors' abstract. We don't add key points for this paper.
Background/Aims: Absolute risk of acute-on-chronic liver failure (ACLF) can be overestimated when the hazards of competing events, such as liver transplantation (LT) or death without ACLF, are not incorporated. We developed and validated a novel competing-risk-based model integrating systemic inflammation to predict ACLF development. Methods: The derivation cohort comprised 1,146 patients with acute decompensation of cirrhosis from the prospective, multicenter Korean Acute-on-Chronic Liver Failure (KACLiF) cohort. External validation used a retrospective 1,104-patient cohort. The study followed the TRIPOD+AI reporting guidelines and applied a Fine-Gray subdistribution hazard model accounting for the competing risks of death or LT, with predictive performance evaluated using the C-index and time-dependent area under the receiver operating characteristic curve (AUROC). Results: The KACLiF-ACLF score integrated C-reactive protein, albumin, total bilirubin, international normalized ratio, and creatinine. In derivation, the competing-risk C-index (0.806 at 28 days; 0.791 at 90 days) significantly exceeded all eight comparators at 90 days, but not MELD or MELD-Na at 28 days. In external validation, it (0.892 and 0.889) significantly exceeded all comparators at both horizons. Time-dependent AUROCs at 28 days were 0.815 and 0.894, significantly exceeding all comparators at both horizons except MELD-Na at 28 days in derivation. Transported absolute risks were overpredicted, requiring recalibration. The model stratified patients into low-, intermediate-, and high-risk groups, correlating with cumulative ACLF incidence and transplant-free survival (P<0.001). Conclusions: The KACLiF-ACLF score discriminates ACLF development with high accuracy in external validation and supports risk stratification within a competing-risk framework. Recalibrated absolute risks require independent validation before clinical implementation.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Nam et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: