Abstract
OBJECTIVES: Model for End-Stage Liver Disease (MELD) alone and with sodium (MELD-Na) have decreasing predictive capacity as trends in liver disease evolve. We sought to combine transient elastography (TE) with MELD-Na to improve its predictive ability. METHODS: This is a retrospective cohort study comparing the use of TE, MELD-Na, and composite MELD-Na-TE to predict liver transplantation and all-cause mortality, with hepatic decompensation as a secondary outcome. Cox proportional hazards regression was used to measure predictive ability and control for confounders. RESULTS: Of the 214 patients, the mean age was 53 years with 35% being female and 76% being Caucasian. Hepatitis C (59%) and nonalcoholic fatty liver disease (22%) were the most frequent liver disease etiologies. On univariable analysis, MELD-Na [hazard ratio (HR) 1.12, 95% confidence interval (CI) 1.06-1.2, P < 0.001], TE (HR 1.04, 95% CI 1.03-1.06, P < 0.001) and composite MELD-Na-TE (HR 1.13, 95% CI 1.08-1.19, P < 0.001) were associated with death or transplant. On multivariable analysis, MELD-Na was no longer significant (HR 1.08, 95% CI 0.95-1.22, P = 0.27) after adjusting for TE (HR 1.05, 95% CI 1.03-1.07, P < 0.001) while composite MELD-Na-TE remained significant (HR 1.16, 95% CI 1.09-1.24, P < 0.001). Composite MELD-Na-TE predicts mortality or liver transplant with the highest C-statistic of 0.81. Age (HR 1.05, 95% CI 1-1.09, P = 0.04), TE (HR 1.04, 95% CI 1.03-1.06, P < 0.001) and composite MELD-Na-TE (HR 1.11, 95% CI 1.06-1.15, P < 0.001) were significantly associated with hepatic decompensation. CONCLUSION: Composite MELD-Na-TE better predicts liver transplantation, death, and hepatic decompensation compared to MELD/MELD-Na or TE alone.