Albumin For Patients With Cirrhosis And Ascites:a Systematic Review

Objective:The clinical efficacy and safety of human serum albumin for patients with cirrhosis and ascites were systematically assessed. The recommendations were made by GRADE.Methods:We searched Ovid Medline, Embase, the Cochrane Central Register of Controlled Trials, Pubmed, CBM, VIP, WanFang Date and CNKI. The above databases were searched up to March2014. Relevant journals were also manually searched up to March2014. Two reviewers independently identified studies, performed quality assessment, extracted the data, made qualitative or quantitative analysis, graded the quality of evidence by GRADE, and made the conclusions as well as the recommendations.Results:In protocol I,8randomized controlled trails met the inclusion criteria involving440participants with cirrhosis and ascites, which were published from1999to2013. There are3studies with high quality and5studies with low quality. Statistical analysis showed that human serum albumin combination with short-time therapy can significantly improved the efficiency of treatment [RR=3.43,95%CI (1.84,6.38), P=0.0001], shortened disappeared time of ascites [MD=-4.04,95%CI (-4.58,-3.50), P<0.0001], increased urinary sodium excretion [MD=10.32,95%CI (0.14,20.49), P=0.05] and decreased serum creatinine concentration [MD=-3.32,95%CI (-6.60,-0.05), P=0.05]. The incidence of adverse reactions [RR=0.68,95%CI (0.30,1.54), P=0.35] and24-h urinary volume were no significant differences between two groups. The quality of evidence was low. In protocol Ⅱ,4randomized controlled trails met the inclusion criteria involving372participants with cirrhosis and ascites, which were published from1962to2006.There are 3studies with high quality and1studies with low quality. Statistical analysis showed that human serum albumin combination with long-term therapy can significantly reduced the incidence of complications [RR=0.35,95%CI (0.18,0.69), P=0.002], readmission rate [RR=0.22,95%CI (0.12,0.42), P<0.0001] and ascites recurrence rate [RR=0.11,95%CI (0.06,0.20), P<0.0001]. There were no significant differences in mortality [RR=1.06,95%CI (0.58,1.92), P=0.86], human serum albumin related adverse reaction and well-being rate. The quality of evidence was moderate.Conclusion:Short-term co-administration of albumin and diuretics shows some benefit with moderate quality of evidence. Strength of recommendation was weak. Large-scale randomized studies with high quality are needed to delineate the role of this strategy. Long-term combination therapy reduce the incidence of complications and re-admissions, but no survival benefit. Considering the high cost of human serum albumin, recommendation was weak.