Abstract: Objective To investigate the predictive efficacy of serum prealbumin （PA） combined with model of end-stage liver disease（MELD） score in the prognosis of patients with decompensated hepatitis B-induced liver cirrhosis. Methods 231 patients with decompensated liver cirrhosis and hepatitis B were recruited in our hospital between December 2015 and December 2016,and all of them were followed-up for six months. Serum PA levels were collected and the MELD score and CTP score were calculated,and MELD/PA model was established. The area under curve of receiver operating characteristic （AUC） was applied to estimate the efficacy of prognosis. Results A total of 83 patients died during the 6-month follow-up period;serum PA content in the dead group was（32.2±9.3） mg/L,significantly lower than[（47.3±26.4）mg/L,P<0.05] in the survival group;the MELD score was（24.1±5.6）,significantly higher than[（18.0±6.7）,P<0.05] in the survival;the CTP score was （11.8±1.2）,significantly higher than [（9.0±2.0）,P<0.05] in the survival;the ROC analysis showed that the AUC of MELD score was 0.868 （95%CI:0.823-0.912）,much higher than[0.753（95%CI:0.690-0.816）,P<0.05] by CTP score,or [0.675（95%CI:0.606-0.743）,P<0.05] by serum PA;further comparison analysis result showed that the AUC of MELD/PA score was 0.896（95%CI:0.857-0.935）,significantly higher than by MELD（P<0.05）. Conclusion The application of serum PA levels combined with MELD score in predicting the 6 month prognosis of patients with decompensated hepatitis B liver cirrhosis is feasible,which needs further investigation.

Key words: Liver cirrhosis, Prealbumin, Model of end-stage liver disease, Predictive efficacy