Uptake and influential factors of randomized trials in Cochrane systematic reviews of acupuncture

Article type
Authors
Wang LQ1, Jin XY2, Shang YX3, Lu CL3, Liu XH3, Tian ZY3, Zhu SJ3, Liang N3, Fei YT3, Zhang Y3, Liu JP3
1School of Acupuncture-Moxibustion and Tuina, Beijing University of Chinese Medicine
2Third Affiliated Hospital, Beijing University of Chinese Medicine
3Center for Evidence-Based Chinese Medicine, Beijing University of Chinese Medicine
Abstract
Background: increasing numbers of randomized clinical trials (RCTs) have been published internationally to evaluate the effect and safety of acupuncture. Cochrane Reviews are an important source for both guidelines and practice.

Objectives: to evaluate the uptake and influencing factors of RCTs in Cochrane Reviews on acupuncture.

Methods: we systematically searched Cochrane Reviews of acupuncture in the Cochrane Library from inception to January 2018. We extracted included and excluded RCTs in the reviews to analyze the uptake and factors for which acupuncture RCTs failed to be included.

Results: we identified 14 Cochrane Reviews of acupuncture, as well as their included and excluded RCTs. Five reviews (35.7%) were published before 2012, seven reviews (50%) had Chinese authors and four reviews did not search Chinese databases. Among 926 articles (representing 870 RCTs) at the full-text screening, 167 (18.03%) articles (127 RCTs) were finally included in the 14 reviews, 718 (77.54%) articles (706 RCTs) were excluded, 13 (1.40%) articles (11 RCTs) were in the waiting list, and 28 (3.02%) articles or protocols (26 RCTs) were in the ongoing list. The uptake rate (included in reviews at the full-text screening) of RCTs in Chinese was 10.03% (59/588), in English it was 30.19% (64/212), and in other languages it was 23.53% (4/17), respectively. We could not determine the language of pubicatio for 50 RCTs (24 had no list of references and 26 were protocols). Among 127 included RCTs, 68 RCTs (53.54%) were considered as high risk of bias (one item is high risk bias) which had no difference in language, 47 RCTs (37.1%) as unclear risk of bias, and 12 RCTs (9.45%) as low risk of bias (all items were low risk bias), which were all published in English. Besides, 11 (78.57%) reviews had uncertain conclusions due to high risk of bias of the RCTs. The excluded rate of RCTs was 89.63% (527/588) in Chinese, 65.57% (139/212) in English, and 76.47% (13/17) in other languages. The reasons of exclusion for 718 articles from 14 reviews were inadequate reporting or without recognized diagnosis criterion of participants (296, 41.23%), inappropriate interventions and controls (206, 28.69%), lack of core outcomes (48, 6.69%), unclear methods for random sequence generation (62, 8.63%), other methodological issues (5, 0.70%), obviously irrelevant to the research topic (82, 11.42%), duplication (39, 5.43%), and other reasons (18, 2.51%), such as published in abstract, unavailable full text, language barriers, and possible plagiarism.

Conclusions: acupuncture RCTs in Cochrane Reviews, especially published in Chinese, have a low uptake rate. Internationally recognized diagnostic criteria, appropriate controls, core outcome, high-quality design and reporting in accordance with CONSORT and STRICTA are encouraged to improve the usage of acupuncture RCTs in future.