Analysis of predictive scores for mortality after the first year of hospitalization for decompensated cirrhotic patients

Abstract

Patients with liver cirrhosis can progress to decompensation, reflecting in high mortality rates. Scores predictors of prognosis and mortality are useful tools in the management of patients with liver disease. The present study investigated the potential of scores and models to predict mortality after the first year of hospitalization in cirrhotic patients treated at a tertiary hospital in the Federal District, Brazil. This is a prospective cohort study that included adult individuals with liver cirrhosis, admitted for acute decompensation of cirrhosis, between November 2018 and May 2021. The criteria for the end of the prospection was one of the following outcomes: end of the period above, death or transplantation. Patients (male and female) with cirrhosis admitted at the hospital for more than one day for treatment of acute decompensation were included according to the previous criteria of the study. The following variables were analyzed: sex, age group, ethnicity, underlying cause of cirrhosis, reason for admission, presence of ascites, presence of encephalopathy, values ​​of biochemical tests on admission (creatinine, sodium, total bilirubin, INR, albumin) and severity of liver disease by CLIF-SOFA, MELD, MELD-Na and Child-Pugh scoring system models, measured in the admission to predict mortality after one year. Of the 115 patients with liver cirrhosis initially included, 63 were excluded for cancer during follow-up, therefore, 52 completed the study, 65% males. Alcoholic hepatitis was the main underlying cause of cirrhosis (46%), and intestinal bleeding was the main cause of hospitalization in decompensation (35%). Among the isolated biochemical parameters, only albumin on admission was significantly associated with the mortality outcome in one year. Among the models evaluated, the CLIF-SOFA showed a significant capacity as a predictor related to mortality in this period.