Derivation and Validation of a Prognostic Model for Acute Decompensated Cirrhosis Patients

BACKGROUND: Acute decompensated cirrhosis (AD) is related to high medical costs and high mortality. We recently proposed a new score model to predict the outcome of AD patients and compared it with the common score model (CTP, MELD and CLIF-C AD score) in the training and validation sets. MATERIALS AND METHODS: A total of 703 patients with AD were enrolled from The First Affiliated Hospital of Nanchang University between December 2018 and May 2021. These patients were randomly assigned to the training set (n=528) and validation set (n=175). Risk factors affecting prognosis were identified by Cox regression analysis and then used to establish a new score model. The prognostic value was determined by the area under the receiver operating characteristic curve (AUROC). RESULTS: A total of 192 (36.3%) patients in the training cohort and 51 (29.1%) patients in the validation cohort died over the course of 6 months. A new score model was developed with predictors including age, bilirubin, INR, WBC, albumin, ALT and BUN. The new prognostic score (0.022×Age + 0.003×TBil + 0.397×INR + 0.023×WBC- 0.07×albumin + 0.001×ALT + 0.038×BUN) for long-term mortality was superior to three other scores based on both training and internal validation studies. CONCLUSION: This new score model appears to be a valid tool for assessing the long-term survival of AD patients, improving the prognostic value compared with the CTP, MELD and CLIF-C AD scores.