Acupuncture Analgesia: A Meta-epidemiological Study And Clinical Evidence Grading

Acupuncture as a treatment has a history of over 2000 years. The most important effect of acupuncture is analgesia. Until now there is no scientific conclusion on the mechanism of the analgesia effect of acupuncture; however, the clinical effect has widely recognized all over the world. With the increase of clinical studies on acupuncture analgesia in recent years, many problems during the designing and implementation have been exposed, and these problems exaggerated when these studies were included in the analyses of systematic reviews(SRs) and meta-analyses of acupuncture analgesia, thus affect the objective evaluation on the true effects of acupuncture and the generation of high-quality clinical evidence of acupuncture analgesia.SRs and meta-analyses are regarded as the highest level of evidence when evaluating whether a treatment is effective. In order to better assess the clinical evidence of acupuncture for pain, it is necessary to comprehensively understand and analyze SRs and meta-analyses. Whether these SRs and meta-analyses critically evaluated randomized controlled trials(RCTs) of acupuncture for pain; whether there are factors that will bias the results and conclusion of these SRs and meta-analyses; and what the grade of this clinical evidence. This project focuses on the above three questions, in the purpose to have a deep understanding of the current clinical evidence of acupuncture for pain.This present project has three parts. The first part systematically searched and analyzed the status of risk of bias(Ro B) evaluation in SRs of acupuncture for pain, in order to comprehensively evaluate the Ro B status in SRs of acupuncture for pain. The second part used the meta-epidemiological method to screen possible factors that may affect the effect size of meta-analyses of acupuncture for pain, such as the origin of the first author, single-center or multi-center, sample size, and Ro B of included studies. The third part uses the Grading of Recommendations Assessment, Development, and Evaluation(GRADE) to grade the clinical evidence provided by the SRs of acupuncture for pain, in order to screen high-quality clinical evidence of acupuncture for pain and analyze factors that contributed to the low- and very low-quality clinical evidence of acupuncture for pain.The first part included 91 SRs, in which 85 conducted Ro B evaluation. Over half(n=59, 64.8%) used standard tools such as Jadad score and Cochrane Ro B tools; over one-third(n=29, 34.1%) conducted domain-level assessment; and over half(n=48, 56.5%)combined the results of Ro B assessment into the data analysis of SRs. Those SRs that didnot combine the results of Ro B assessment into their data analysis have low reliability. In addition, those SRs used scales that based on the “study quality” may not truly reveal the methodological quality or fact of the studies.The third part included 23 meta-analyses. Forty-four clinical evidence were screened out, of which 16(36.3%) were of high quality, 11(25.0%) were of medium quality, 8(18.1%) were of low quality, and 9(20.6%) were of very low quality. In the high-quality evidence, 11 corresponded to pain degree, and 4 corresponded to response rate. For lowand very low-quality evidence, the factors that affect the grading of evidence are high or unclear Ro B, inconsistency and publication bias.This project has the following important findings:1. The Ro B evaluation is not well conducted in the SRs and meta-analyses of acupuncture for pain;2. Factors such as author origin, study design, sample size and Ro B affect the estimation of effect size of acupuncture treatment in the SRs and meta-analyses of acupuncture for pain;3. Although there are many problems need to be resolved in the design and conduct of RCTs of acupuncture for pain, high-quality clinical evidence is not rare in present.This project has forwarded the following suggestions for the future research of acupuncture for pain:1. The Ro B evaluation process should strictly follow the instructions in the Cochrane Handbook. The SR authors should try their best to search the study protocols and contact the study authors, in order to have a comprehensive understanding of the real status of the included studies.2. The results of SRs or meta-analyses should be judged with cautious. For the studies form Asian countries, single-center studies, studies of small sample size, and those has high or unclear Ro B, the effect size may be exaggerated when they are included into meta-analyses.3. The study design of acupuncture clinical research should not blindly pursue “large sample size”, “randomized”, or “double blinding”. These factors are not the guarantee of making high-quality clinical evidence.To sum up, this project for the first time used meta-epidemiological method to dig the Ro B factors in the SRs and meta-analyses of acupuncture for pain, and for the first time used GRADE to grade the clinical evidence in this field. The methods as well as the resultsof this project have guiding significance for the future clinical research of acupuncture.