Usefulness of Model for End-stage Liver Disease Score for Predicting Mortality after Intra-abdominal Surgery in Patients with Liver Cirrhosis in a Single Hospital

BACKGROUND/AIMS Recent studies have suggested that the model for end-stage liver disease (MELD) score is superior to the Child-Turcotte-Pugh (CTP) score as a predictor of postoperative mortality, especially up to 90 days. This study aimed to determine whether MELD score can predict the postoperative outcome of patients with liver cirrhosis in Korea. METHODS We reviewed the medical records of 98 patients with liver cirrhosis who underwent intra-abdominal surgery under generalized anesthesia between March 2003 and December 2008 at Kangbuk Samsung Hospital. Univariate and multivariate cox proportional hazards analyses were performed to determine the correlation between risk factors and mortality. RESULTS Eighty-two percent of patients (n=80) were male. Mean MELD score was 10.82 ± 3.84. Common causes of liver cirrhosis were hepatitis B (57.2%) and alcohol (22.4%). Ninety-day mortality ranged from 2.1% (MELD score, ≤ 9) to 25% (MELD score, ≥ 17). By multivariate analysis, MELD score > 9 (HR 2.490; [95% CI 1.116-5.554; p=.026]) and American Society of Anesthesiologists Class ≥ IV (HR 2.433; [95% CI 1.039-5.695; p=.041]) predicted mortality at 30 days after surgery. Only MELD score was a predictor of prognosis at 90 days (HR 2.446; [95% CI 1.118-5.352; p=.025]). Etiology of cirrhosis and CTP score were not predictors of mortality. CONCLUSIONS MELD score was a useful predictive parameter of postoperative mortality at 30 days and 90 days, independent of the etiology of cirrhosis.