Outcomes after index hospital admission for decompensated alcohol-related cirrhosis versus severe alcoholic hepatitis: a retrospective cohort study

Abstract

BackgroundWe have previously described outcomes in severe alcoholic hepatitis, identifying abstinence after hospital discharge as the only independent predictor of long-term survival. The outcome of patients presenting with other forms of decompensated alcohol-related liver disease is unknown. Our aim was to assess predictors of inpatient and long-term survival in patients admitted with decompensated alcohol-related cirrhosis compared with those with severe alcoholic hepatitis. MethodsA retrospective case-note review was performed for all index hospital admissions with alcoholic-related liver disease between April 1, 2005, and Dec 31, 2011. Patients with decompensated alcohol-related cirrhosis (Child-Pugh score ≥7 and bilirubin <80 mmol/L) were compared with a previously studied cohort with severe alcoholic hepatitis admitted over the same time period (discriminant function ≥32, corroborated by biopsy results where available). Predictors of inpatient survival were determined using binary logistic regression, and the Kaplan-Meier method was used to estimate long-term survival. Findings177 patients were studied (73 decompensated alcohol-related cirrhosis, 104 severe alcoholic hepatitis). In decompensated alcohol-related cirrhosis, the independent determinants of inpatient mortality were hospital-acquired infection (odds ratio 20·2, 95% CI 3·5–118; p=0·001) and hepatic encephalopathy (9·3, 1·8–47; p=0·007). Overall median survival was similar in both groups (26 months [IQR 12·8–39·2] in decompensated alcohol-related cirrhosis vs 23 [10·6–35·4] in severe alcoholic hepatitis, p=0·872) with similar abstinence rates at last follow-up (41·1% vs 34·6%, p=0·431). In severe alcoholic hepatitis, abstinence at last follow-up was the only predictor of survival (3 year survival 75·2% in abstainers vs 33·2% in continued drinkers, p=0·012). However, in decompensated alcohol-related cirrhosis, abstinence did not predict long-term survival (median survival 36 months [IQR 22·3–49·7] in abstainers vs 38 [9·9–66·1] in continued drinkers, p=0·949). InterpretationInpatient mortality in decompensated alcohol-related liver disease was predicted by hospital-acquired infection and hepatic encephalopathy. Long-term survival in decompensated alcohol-related cirrhosis did not significantly improve with abstinence, which is in contrast to severe alcoholic hepatitis, and this suggests a reduced scope for disease reversibility thereby mandating earlier assessment for liver transplantation. FundingNone.