A revised scope in different prognostic models in cirrhotic patients: Current and future perspectives, an Egyptian experience.

BACKGROUND AND STUDY AIM

The prognosis of cirrhosis is of great interest for liver transplantation and new therapies of related complications. Traditional prognostic models such as Child-Pugh (CP) and Model for End-stage Liver Disease (MELD) were developed to predict mortality in decompensated cirrhosis, but lack parameter(s) related to complications. Recently, new models such as creatinine-modified Child-Turcotte-Pugh (CrCTP) and sodium-based MELD variants were developed to improve prognostic accuracy and enhance outcome predictive capability. Our aim was to investigate the prognostic ability of these models and their relation to complications among Egyptian cirrhotic patients to determine the best one and to assess adding new variables to improve the prognostic ability of that model.

PATIENTS AND METHODS

A total of 1000 cirrhotic patients were enrolled in a retrospective study; traditional and new prognostic models such as CP, MELD, CrCTP, integrated MELD (iMELD), MELD plus sodium (MELD-Na, MELDNa) and MELD:sodium ratio (MESO) were calculated. The predictive abilities of prognostic models were compared using the area under receiver operating characteristic curve (AUC) and 1-year survival rates were evaluated by Kaplan-Meier survival analysis. An index of cirrhosis-related complications was added to reveal the best prognostic model.

RESULTS

Using AUC, MELD and its sodium variants was significantly better than CP and CrCTP scores in predicting risk of 1-year mortality, where MELD-sodium (MELD-Na) had the highest AUC (0.743). Adding an index of cirrhosis-related complications (C) to MELD-Na creating a new scoring system (MELD-Na-C) improved its prognostic accuracy (AUC 0.753). Kaplan-Meier survival curves predicted increased mortality with higher prognostic scores.

CONCLUSIONS

All prognostic models were good predictors of 1-year mortality in patients with decompensated cirrhosis; however, MELD-Na was the best for outcome prediction. MELD-Na-C was a new model enhancing the predictive accuracy in assessing cirrhotic patients with related complications.