Retrospective and prospective validation of model for end-stage liver disease (MELD) score in predicting mortality in patients of alcoholic liver disease.

Abstract

Patients with alcoholic hepatitis and alcoholic cirrhosis have significant morbidity and mortality. The model for end-stage liver disease (MELD) score is used to predict mortality in patients with liver disease. Our study aimed to validate the utility of MELD score versus Child Pugh Turcotte (CPT) and modified CPT (CPT+creatinine; CPTC) scores in predicting mortality among alcoholic liver disease (ALD) patients. We examineda retrospective (n=110; mean age 44.5years; 109 men) and a prospective (n=96; mean age 45 years; 92 men) cohort of patients suffering from ALD, with varying disease severity, and compared the ability of MELD, CTP and CPTC scores to predict the 1-, 3- and 6-month mortality using concordance (c-) statistics. The median MELD, CPT and CPTC scores were 21, 10 and 11 in the retrospective group and 21, 10 and 10 in prospective groups, respectively. The 1-month, 3-month and 6-month mortality was 41 (37.3%), 59 (53.6%) and 81 (73.6%) in the retrospective group and 27 (28.1%), 41 (42.7%) and 65 (67.7%) in the prospective group. The predictive ability of MELD, CPT and CPTC was 0.72, 0.67, 0.68 for 1-month mortality, 0.73, 0.70, 0.68 for 3-month mortality and 0.83, 0.75, 0.73 for 6-month mortality in retrospective group and 0.86, 0.56, 0.63 for 1-month mortality, 0.80, 0.57, and 0.60 for 3-month mortality, and 0.88, 0.51, and 0.64 for 6-month mortality in the prospective cohort, respectively. The predictive accuracy of MELD was significantly better than CPT and CPTC scores in the prospective group. MELD is a valid prognostic score compared to CPT and CPTC scores for predicting short-term mortality in patients of alcoholic liver disease.