黄芪注射液治疗脑梗死的Meta分析

目的：评价黄芪注射液治疗脑梗死的疗效与安全性。方法：计算机检索MEDLINE、EMbase、Cochrane图书馆、CNKI、VIP和万方数据库,按照纳入与排除标准由两位研究者独立筛选和提取资料,采用Cochrane协作网随机对照试验的质量评价标准评价纳入文献质量,使用Rev Man 5.2软件进行Meta分析。结果;共纳入10篇文献,864例患者。10篇文献报道均未见患者死亡情况,其病死率为0%。Met a分析结果显示：黄芪注射液组的临床疗效[RR=1.21,95%CI（1.13,1.29）]和神经功能缺损改善情况[MD=-3.28,95%CI（-5.27,-1.29）]均优于对照组,2组相比差异有统计学意义。纳入的10篇文献中,4篇文献报道未见明显不良反应,3篇未提及安全性或不良反应,2篇有不良反应报道,但具体组别未知,仅有1篇报道试验组出现轻度腹泻2例和身体局部瘙痒1例。结论：黄芪注射液治疗脑梗死,在提高临床疗效、改善神经功能缺损评分方面可能优于对照组药物,但由于本系统评价纳入研究质量低且样本量小,药物安全性不明确,尚需高质量随机对照试验进一步验证。

Meta Analysis of HuangQi Injection in Treating Cerebral Infarction

Objective： To evaluate curative effects and the safety of HuangQi injection in treating cerebral infarction. Methods：MEDLINE, EMbase, Cochrane Library, CNKI, VIP and WanFang database were searched by the computer, the information was screened and extracted independently by two researchers according to the including and excluding criteria. The quality of the included studies was evaluated based on quality assessment standard of the randomized controlled trials in Cochrane Library, Meta analysis was performed by Rev Man 5.2 software. Results：Ten articles were included, involving 864 patients. There was no death reported in ten pieces of the articles, and mortality rate was 0%. The results of Meta analysis showed：clinical efficacy [RR=1.21,95%CI（1.13, 1.29）] and the improvement of neurological impairment [MD=-3.28, 95%CI （-5.27,-1.29）] of HuangQi injection group were superior to these of the control group, the difference had statistical meaning. Among ten articles, there was no obvious adverse reaction reported in four articles, the safety and adverse reaction were not mentioned in three articles, there were adverse reaction reported in two articles but without the detailed grouping, only one article described two cases with slight diarrhea and one case with local itching in the experiment group. Conclusion：HuangQi injection in treating cerebral infarction might be superior to the drugs of the control group in improving clinical effects and neurological impairment scales, because of low quality of the included study with small sample size and unclear drug safety, it still needs to be further tested by high quality randomized controlled trials.