• Shanxia Luo, Youlin Long, Wenzhe Xiao, Xin Wang, Rui Chen, Qiong Guo, Jia Liu, Ruochen Shao, Liang Du, and Min Chen.
• Department of Mental Health Center, West China Hospital of Sichuan University, Chengdu, P.R. China.
• J Evid Based Med. 2020 Feb 1; 13 (1): 25-33.

ObjectivesTo assess the use of risk of bias (ROB) assessment tools and the reporting quality of ROB assessment results in systematic reviews (SRs) of acupuncture for depression, as well as to evaluate the ROB of depression-related randomized controlled trials (RCT).MethodsEmbase, Medline, Chinese Journal Full-Text Database (CJFD), VIP Chinese Technology Periodical Database, and WanFang Data Resource System of Digital Periodicals were searched from their inception to 24 November 2017. SRs of RCTs concerning acupuncture on depression were included. General characteristics and the information related to risk of bias in SRs were extracted. A descriptive analysis was used.ResultsThirty-nine SRs were included. Of these, two (5%) did not perform a ROB assessment, 18.9% did not report the ROB assessment results, and 62.2% did not report the assessment results of each ROB item. Text descriptions and tables were commonly used in reporting forms. Only 32.4% of SRs reported support for judgment. The reporting rate of ROB assessment results was low in all items (13.5%-35.1%). Regarding RCTs, 59.7% used adequate randomization methods, 13.1% performed adequate allocation concealment, 12.5% performed adequate blinding of participants and personnel, 27.3% performed adequate blinding of the assessment outcomes, and 41.5% and 49.3% had a low ROB in terms of incomplete outcome data and selective outcome reporting, respectively.ConclusionFor the SRs of acupuncture for depression, the selection of ROB assessment tools needs to be optimized. The reporting quality is poor, and the overall ROB of RCTs is high. Therefore, the results may not be reliable.© 2020 Chinese Cochrane Center, West China Hospital of Sichuan University and John Wiley & Sons Australia, Ltd.

19 keywords...

hide…