Predicting short-term and long-term mortality of hospitalized Portuguese patients with alcoholic hepatitis

Alcohol abuse can result in a spectrum of liver injury that ranges from mild fatty infiltration to alcoholic hepatitis (AH), cirrhosis, and hepatocellular carcinoma. The present study aimed to evaluate current scoring systems in predicting short-term and long-term mortality because of AH.

Records of 170 consecutive patients with AH admitted to a tertiary center between January 2005 and October 2015 were reviewed. Clinical and biochemical parameters were retrieved for the assessment of AH scores for the day of admission (D1) and for the seventh day of hospitalization (D7). Endpoints included admission to the ICU, and 30-day, 90-day, and 1-year mortality.

The Maddrey discriminant function and the Model of End-Stage Liver Disease (MELD) were modest predictors of the need for ICU admission. In-hospital, 30-day, 90-day, and 1-year mortality were 15.9, 18.2, 21.8, and 30.0%, respectively. There was a numerical, albeit nonsignificant, trend for higher accuracy using D7 scores, especially the MELD, in predicting 30-day and 1-year mortality. Overall, all scores showed high negative predictive values (30 day: 91.2-98.7% and 1 year: 78.8-93.7%), but modest positive predictive values (30 day: 30.6-70.8% and 1 year: 42.1-61.2%). Survival rates were the highest among patients showing a decrease in the MELD, Glasgow Alcoholic Hepatitis Score, and Age, serum Bilirubin, International normalized ratio, and serum Creatinine score over the first week of admission.

AH scores were comparable in identifying patients at low risk of mortality up to 1 year following admission. Reassessment of the MELD, Glasgow Alcoholic Hepatitis Score, and Age, serum Bilirubin, International normalized ratio, and serum Creatinine score scores after 1 week further improved mortality prediction.