METHODS: We recruited ACLF patients between 2009 and 2020 from APASL-ACLF Research Consortium (AARC). Their clinical data, investigations and organ involvement were serially noted for 90-days and utilized for AI modelling. Data were split randomly into train and validation sets. Multiple AI models, MELD and AARC-Model, were created/optimized on train set. Outcome prediction abilities were evaluated on validation sets through area under the curve (AUC), accuracy, sensitivity, specificity and class precision.
RESULTS: Among 2481 ACLF patients, 1501 in train set and 980 in validation set, the extreme gradient boost-cross-validated model (XGB-CV) demonstrated the highest AUC in train (0.999), validation (0.907) and overall sets (0.976) for predicting 30-day outcomes. The AUC and accuracy of the XGB-CV model (%Δ) were 7.0% and 6.9% higher than the standard day-7 AARC model (p
METHODS: ACLF patients recruited from the APASL-ACLF Research Consortium (AARC) were followed up till 30 days, death or transplantation, whichever earlier. Clinical details, including dynamic grades of HE and laboratory data, including ammonia levels, were serially noted.
RESULTS: Of the 3009 ACLF patients, 1315 (43.7%) had HE at presentation; grades I-II in 981 (74.6%) and grades III-IV in 334 (25.4%) patients. The independent predictors of HE at baseline were higher age, systemic inflammatory response, elevated ammonia levels, serum protein, sepsis and MELD score (p
METHODS: Prospectively collected data from the AARC database were analyzed.
RESULTS: Of the 1249 AH patients, (aged 43.8 ± 10.6 years, 96.9% male, AARC score 9.2 ± 1.9), 38.8% died on a 90 day follow-up. Of these, 150 (12.0%) had mild-moderate AH (MAH), 65 (5.2%) had SAH and 1034 (82.8%) had ACLF. Two hundred and eleven (16.9%) patients received CS, of which 101 (47.87%) were steroid responders by day 7 of Lille's model, which was associated with improved survival [Hazard ratio (HR) 0.15, 95% CI 0.12-0.19]. AARC-ACLF grade 3 [OR 0.28, 0.14-0.55] was an independent predictor of steroid non-response and mortality [HR 3.29, 2.63-4.11]. Complications increased with degree of liver failure [AARC grade III vs. II vs I], bacterial infections [48.6% vs. 37% vs. 34.7%; p
METHODS: Patients with AIH-ACLF without baseline infection/hepatic encephalopathy were identified from APASL ACLF research consortium (AARC) database. Diagnosis of AIH-ACLF was based mainly on histology. Those treated with steroids were assessed for non-response (defined as death or liver transplant at 90 days for present study). Laboratory parameters, AARC, and model for end-stage liver disease (MELD) scores were assessed at baseline and day 3 to identify early non-response. Utility of dynamic SURFASA score [- 6.80 + 1.92*(D0-INR) + 1.94*(∆%3-INR) + 1.64*(∆%3-bilirubin)] was also evaluated. The performance of early predictors was compared with changes in MELD score at 2 weeks.
RESULTS: Fifty-five out of one hundred and sixty-five patients (age-38.2 ± 15.0 years, 67.2% females) with AIH-ACLF [median MELD 24 (IQR: 22-27); median AARC score 7 (6-9)] given oral prednisolone 40 (20-40) mg per day were analyzed. The 90 day transplant-free survival in this cohort was 45.7% with worse outcomes in those with incident infections (56% vs 28.0%, p = 0.03). The AUROC of pre-therapy AARC score [0.842 (95% CI 0.754-0.93)], MELD [0.837 (95% CI 0.733-0.94)] score and SURFASA score [0.795 (95% CI 0.678-0.911)] were as accurate as ∆MELD at 2 weeks [0.770 (95% CI 0.687-0.845), p = 0.526] and better than ∆MELD at 3 days [0.541 (95% CI 0.395, 0.687), p 6, MELD score > 24 with SURFASA score ≥ - 1.2, could identify non-responders at day 3 (concomitant- 75% vs either - 42%, p
METHODS: Prospectively collected data of ACLF patients from APASL-ACLF Research Consortium (AARC) was analyzed for 30-day outcomes. The models evaluated at days 0, 4, and 7 of presentation for 30-day mortality were: AARC (model and score), CLIF-C (ACLF score, and OF score), NACSELD-ACLF (model and binary), SOFA, APACHE-II, MELD, MELD-Lactate, and CTP. Evaluation parameters were discrimination (c-indices), calibration [accuracy, sensitivity, specificity, and positive/negative predictive values (PPV/NPV)], Akaike/Bayesian Information Criteria (AIC/BIC), Nagelkerke-R2, relative prediction errors, and odds ratios.
RESULTS: Thirty-day survival of the cohort (n = 2864) was 64.9% and was lowest for final-AARC-grade-III (32.8%) ACLF. Performance parameters of all models were best at day 7 than at day 4 or day 0 (p 12 had the lowest 30-day survival (5.7%).
CONCLUSIONS: APASL-ACLF is often a progressive disease, and models assessed up to day 7 of presentation reliably predict 30-day mortality. Day-7 AARC model is a statistically robust tool for classifying risk of death and accurately predicting 30-day outcomes with relatively lower prediction errors. Day-7 AARC score > 12 may be used as a futility criterion in APASL-ACLF patients.