Objective To assess the efficacy and safety of treating post-stroke depression with acupuncture compared to western medicine systematically. Methods Databases were retrieved such as Cochrane Library, PubMed, Web of Sicence, EMbase, CBM, CNKI and WanFang Data so as to look up randomized controlled trials （RCT） of treating post-stroke depression with acupuncture and western medicine, and the time limit for the retrieval spanned from the date of database established to September 2013. By extracting data and evaluating methodological quality of included studies according to inclusion and exclusion criteria, RevMan 5.2 software was applied for Meta-analysis and evidence quality was assessed by adopting the GRADE system. Results Atotal of 23 RCTs （845 patients with post-stroke depression） were included in this study. It was revealed by Meta-analysis that the differences of the two groups were statistically significant in terms of HAMD scale scores measured at the end of the treatment by comparing the treatment of acupuncture and western medicine [SMD=0.26, 95% CI （0.11, 0.40）], but via susceptibility analysis （excluding low quality studies）, the differences were not statistically significant [SMD=-0.06, 95% CI （-0.37, 0.25）], and the reliability of the results was low; in terms of adverse events, the differences were statistically significant [RR= 0.32, 95% Cl （0.19, 0.53）], and the risk of adverse events reduced by 68% in the acupuncture group. From the assessment on evidence quality grade based on GRADE system, it was revealed that HAMD scores and adverse events were evidences with fairly low quality. Conclusion Although this system assessment showed that treating post-stroke depression with acupuncture was more effective compared with western medicine, the result was less reliable and quality of evidences was poor. The above-mentioned results need more high- quality randomized controlled trials for further verification.
