Child-Turcotte-Pugh versus model for end stage liver disease score for predicting survival in hospitalized patients with decompensated cirrhosis.

ABSTRACT

BACKGROUND: Short term and medium term survival predictive value of the Child-Turcotte-Pugh (CTP) and Model for End Stage Liver Disease (MELD) score has been established but their usefulness in predicting survival of hospitalized patients with decompensated cirrhosis is lacking. We compared the survival predictive value of these scoring systems in hospitalized patients of decompensated cirrhosis and other associated factors. METHODS: A prospective, observational study in 216 consecutive cases of decompensated cirrhosis admitted in medical ward, were enrolled if the inclusion criteria were fulfilled. All cases were investigated and treated as per standard guidelines and clinician's judgment. CTP and MELD score were calculated for each case at the time of admission and followed throughout the hospital stay till discharge/death. The accuracy of the different score systems for predicting survival was evaluated through the area under ROC curve. RESULTS: CTP and MELD score were higher (12.44 ± 1.07, 31.91 ± 4.92) in expired cases than who improved and discharged (11.32 ± 1.28, 23.97 ± 5.36) respectively with significant p-value (<0.001). Area under ROC curve for serum creatinine, MELD score, blood urea and CTP score for predicting hospital survival were 0.887, 0.864, 0.836 and 0.738 respectively. CONCLUSIONS: MELD score is superior to CTP score in predicting survival at the time of discharge in hospitalized patients with decompensated cirrhosis. Renal failure in patients with decompensated cirrhosis carries poor prognosis and has a good outcome prognostic value, even superior to MELD/CTP scoring.