Publication

Abstract : Methods: In this multicenter, prospective study spanning 2 years, clinical, biochemical, and 90-day survival data were collected from 625 patients with AD (European Association for the Study of the Liver criteria) across North, South, and East India. We divided the cohort into a Derivation cohort (DC: 318 patients) and a Validation cohort (VC: 307 patients). Predictive models for pre-ACLF were derived, validated, and compared with established scores such as model for end-stage liver disease (MELD) 3.0 and chronic liver failure Consortium acute decompensation. Results: Of 625 patients (mean age 49 years, 83% male, 77.5% with alcohol-related liver disease), 32.2% progressed to ACLF. Patients progressing to ACLF showed significantly higher bilirubin (10.9 vs 8.1 mg/dL), leukocyte counts (9,400 vs 8,000 per mm 3 ), international normalized ratio (1.9 vs 1.8), and MELD 3.0 (28 vs 25) but lower sodium (131 vs 134 mEq/L) and survival (62% vs 86%) compared with those without progression ( P < 0.05) in the DC. Consistent results were noted with alcohol-associated hepatitis, infection and hepatic encephalopathy as additional risk factors in VC. Liver failure at presentation (odds ratio: 2.4 [in DC], 6.9 [in VC]) and the 7-day trajectories of bilirubin, international normalized ratio, and MELD 3.0 significantly predicted ACLF progression ( P < 0.001). A new pre-ACLF model showed superior predictive capability (area under the curve of 0.71 in DC and 0.82 in VC) compared with MELD 3.0 and chronic liver failure Consortium acute decompensation scores ( P < 0.05). Discussion: Approximately one-third of AD patients in this Indian cohort rapidly progressed to ACLF, resulting in high mortality. Early identification of patients at risk can guide targeted interventions to prevent ACLF.