Risk Stratification for Adverse Events in End-Stage Liver Disease

Study Goal The goal of this multicenter, prospective observational cohort study is to develop and validate a multimodal predictive model that integrates portal hypertension, inflammation, and liver function indicators to predict the risk of adverse clinical outcomes (e.g., further decompensation, mortality, recompensation, and liver failure) in patients with decompensated cirrhosis. Target Population The study will enroll adults aged 18-80 years with chronic liver disease of any etiology who have experienced at least one decompensating event (e.g., ascites, variceal bleeding, overt hepatic encephalopathy, jaundice, or infection requiring treatment) at or prior to enrollment. Main Questions It Aims to Answer 1. How accurately can the multimodal model predict the risk of further decompensation (e.g., recurrent ascites, rebleeding, spontaneous bacterial peritonitis, or hepatorenal syndrome-AKI) within 1 year of enrollment? 2. Can the model reliably predict 6-week, 1-year, and 2-year liver-related mortality, as well as the incidence of liver failure and the probability of recompensation (per Baveno VII criteria), and does it outperform traditional scoring systems? Comparison Group Researchers will compare the performance of the newly developed multimodal model against established traditional prognostic tools, including the MELD, MELD-Na, and Child-Pugh scores, to assess whether the new model provides superior predictive accuracy, better risk reclassification, and greater clinical net benefit. Participants Will * Provide written informed consent before any study-related data collection. * Undergo baseline and routine 3-month follow-up assessments (aligned with their standard clinical care) over a total of 3 years, during which clinical data, laboratory tests, and imaging results will be recorded. * Voluntarily provide additional biospecimens (peripheral blood 10 mL, 3 stool tubes, and 10 mL urine) at baseline and during return visits for exploratory biomarker and multi-omics analysis (participants who provide samples receive a ¥200 nutritional subsidy; those undergoing clinically indicated MRI also receive free MRE testing).