Safflower Injection for Chronic Kidney Disease: A Meta-analysis and Trial Sequential Analysis

OBJECTIVE To systematically evaluate the clinical efficacy of safflower injection in the treatment of diabetic nephropathy. METHODS Such databases as CBM, CNKI, WanFang, VIP, PubMed, The Cochrane Library, Embase were searched in order to collect randomized controlled trials about safflower injection combined with conventional medication for diabetic nephropathy on the deadline to July 2018. Two researchers independently screened articles according to the inclusion and exclusion criteria, and assessed the risk of bias of included studies. Meta-analysis was carried by using RevMan 5.3 software, Egger test was carried by using Stata 14.0 software to evaluate publication bias, and nonparametric trim and fill method was used to evaluate the impact of publication bias on the results. Trial sequential analysis was performed by using TSA v0.9 software. GRADE pro 3.6 software was used to grade the evidence quality of key outcome indicators. RESULTS A total of 20 studies were included, including 1 640 patients with diabetic nephropathy. Meta-analysis showed that the curative effect of safflower injection combined with normal treatment in the improvement of total effective rate, decrease of serum creatinine(SCr), decrease of blood urea nitrogen(BUN), decrease of 24 h proteinuria were better than normal treatment with statistically significant, however, the quality of evidence was required to be improved. The results of publication bias test showed that there were publication bias in efficiency and BUN. Nonparametric trim and fill method showed that the results of meta-analysis were stable, and the possible publication bias did not affect the results. Trial sequential analysis showed that the evidence of effective rate, SCr and BUN was definited. CONCLUSION Safflower injection combined with normal treatment can increase the total effective rate, decrease SCr, BUN and 24 h proteinuria with statistically significant and a few adverse reaction. There are some limitations due to the quality of included literatures, still need to be validated with higher quality literatures.