Survival gain after liver transplantation for patients with alcoholic liver disease: a comparison across models and centers

BACKGROUND: Prognostic models, which estimate survival probabilities, enable the survival benefit from a procedure to be assessed and allow the comparison of outcomes between centers. We therefore compared three models that estimate survival in the absence of transplantation for patients with alcoholic liver disease (ALD) to assess survival benefit after transplantation and to compare the outcomes in the six liver-transplant centers in England. METHODS: The Béclére and Birmingham models and the Model for End-Stage Liver Disease (MELD) were used to estimate survival in the absence of transplantation in a cohort of 82 patients with end-stage ALD. Posttransplant survival in the same cohort of patients was calculated using conventional survival techniques. Each individual's short-term survival gain after liver transplantation was also calculated and compared across UK liver transplant centers. RESULTS: The expected gain in survival differed substantially depending on the model used. Over the 4-year study period, the survival gain for all ALD patients was 1.70 years (95% confidence interval [CI] 1.37-2.03) using the Béclére model, 0.95 years (CI 0.60-1.30) using MELD, and 0.08 years (CI -0.31-0.47) using the Birmingham model. Two centers consistently had greater estimated survival gains up to 4 years postliver transplant regardless of the model used to estimate nontransplant survival. CONCLUSIONS: These findings suggest that although liver transplantation is associated with an improvement in survival, the gain over 4 years is modest. The three models are poorly correlated and should be applied with caution. Survival gain does, however, appear to vary between centers.