Model for end-stage liver disease combined with serum prealbumin to predict the prognosis of patients with decompensated liver cirrhosis

OBJECTIVE: To evaluate the prognostic value of model for end-stage liver disease (MELD) combined with serum prealbumin (PA) in patients with decompensated liver cirrhosis. METHODS: A total of 252 patients were enrolled in the study and followed 1 year. PA was measured and MELD score was calculated on the first day of admission. Analysis of variance (anova) was used to assess correlation between PA level and MELD score. Multivariable Cox proportional hazards model was used to screen the prognosis related factors. Kaplan-Meier survival curves were drawn. RESULTS: Of the 252 patients, 28 died within 3 months, 58 within 6 months and 91 within 1 year. Serum PA level in dead patients was significantly lower than that in survival patients (P < 0.005) and decreasing with increasing of MELD score. Cox analysis showed that MELD score > 18 (RR = 2.749) and PA < 70 mg/L (RR = 2.412) were independent prognosis risk factors. The risk ratio of MELD score combined with PA level (1.854, P < 0.01) was higher than that of MELD score alone (1.054, P < 0.05). Kaplan-Meier survival curve analysis showed that MELD score < 18 combined with PA >= 70 mg/L could clearly discriminate patients who would survive or die within 6 month and 1 year follow up. CONCLUSION: MELD score < 18 combined with PA >= 70 mg/L could predict the 6-month and 1-year prognosis of patients with decompensated liver cirrhosis, and was superior to that of MELD score alone.