Multivariate analysis of short-term prognosis of liver transplantation in patients with end-stage liver disease

Objective To analyze many indicators during perioperative period of liver transplantation in patients with end-stage liver disease, only to seek related factors that can accurately predict short-term(≤three months)outcome. Methods We analyzed retrospectively clinical data of consecutive patients with end-stage liver diseases undergone liver transplantation in a single treatment center. Logistic regression analysis was used to analyze the perioperative indicators including recipient gender, age, body mass index, preoperative serum albumin level, serum sodium concentration, urea nitrogen level and donor-recipient blood group, et al. Correlated factors were analyzed by the method of multivariate logistic regression. Statistical processing package was SAS 9.1.3 soft. The difference was statistically significant with P<0.05. Results 18/165 patients died within 3 months after transplantation(mortality rate: 10.9 %). According to the result of univariate analysis, the indicators correlated with early mortality which were statistically significant were preoperative serum sodium, blood urea nitrogen, PT-INR, CTP score, MELD score and MELD-Na score. On the base of the result of Logistic multiple regression. However, only MELD-Na score was associated with 3 months prognosis(P=0.001, β=-2.510, OR=0.088, 95 % CI=0.037~0.349). Conclusions Preoperative MELD-Na score is an independent risk factor for short-term survival in patients with end-stage liver disease. Higher MELD-Na score is, the early mortality is higher. Key words: Liver transplantation; Prognosis; Risk factor; Multiplicity analysis