MELD scores with incorporation of serum sodium and death prediction in cirrhotic patients on the waiting list for liver transplantation: a single center experience in southern Brazil.

ABSTRACT

To compare the accuracy of standard model for end-stage liver disease (MELD) score with that of four MELD-based scores incorporating serum sodium (SNa) to predict three- and six-month mortality in cirrhotic patients after their placement on the waiting list for liver transplantation (LT). A cohort study was performed. Receiver operating characteristic (ROC) curves were generated for MELD, MELD incorporating SNa (MELD-Na, MELD-Na2), integrated MELD (iMELD), and MELD to SNa ratio (MESO) index to assess the predictive accuracy of these scores to determine three- and six-month mortality. The c-statistic (area under the ROC curve [AUC]) was used to determine predictive power and the Cox proportional-hazard ratio to estimate death risk. We studied 558 patients. There was a statistically significant difference in the predictive accuracy of scores at three months (AUCs MELD = 0.79 [95% CI = 0.72-0.87]; MELD-Na = 0.84 [95% CI = 0.78-0.90]; MELD-Na2 = 0.85 [95% CI = 0.80-0.91]; iMELD = 0.85 [95% CI = 0.80-0.90]; MESO = 0.81 [95% CI = 0.80-0.91]) and at six months (MELD = 0.73 [95% CI = 0.67-0.80]; MELD-Na = 0.79 [95% CI = 0.73-0.84]; MELD-Na2 = 0.80 [95% CI = 0.74-0.85]; iMELD = 0.80 [95% CI = 0.75-0.85]; MESO = 0.75 [95% CI = 0.69-0.81]) (p < 0.001). Death risk was independent of age and sex. Sodium-modified MELD scores are able to more accurately predict three- and six-month mortality among cirrhotic patients awaiting LT.