거짓침을 대조군으로 사용한 국내 침 임상시험에 대한 체계적 고찰과 메타분석
김정은, 강경원, 김태훈 et al. · 2011
This study aimed to review randomized controlled trials of acupuncture performed in South Korea that used sham acupuncture as a control group. Methods : The following databases were searched through the end of September 2011: Koreanstudies information service system (KISS), Korean medical database (KMbase), national discovery OR science leaders (NDSL), oriental medicine advance searching integrated system (OASIS), and research information service system (RISS). The following search terms were used: acupuncture AND (sham or placebo). The reference lists of searched articles and Korea institute of oriental medicine (KIOM) reports(2005~2009) were identified. The following data were extracted: year/first author, disease, number of participants, blinding, intervention, outcome, and result. Where appropriate, we performed meta-analysis. The methodological quality was assessed according to the Jadad scale and ``risk of bias`` by Cochrane Handbook procedure. Results: Twenty-nine studies were included in this review. In eighteen studies, penetrating sham controls were used as the control intervention, whereas the remaining eleven studies adopted non-penetrating sham controls such as the Park Sham Device or blunt auricular acupuncture. Nine studies showed statistically significant difference in outcomes. Twelve studies concerning insomnia after stroke, chronic tension-type headache, idiopathic Parkinson`s disease, Hwa-Byung, and smoking cessation were included in meta-analysis. A meta-analysis of insomnia after stroke only found significant difference(MD -4.31, 95% Cl -6.19 to -2.42, p<0.00001). In general, all of the studies showed low methodological quality(Jadad score: mean 2.1). Risk of bias by Cochrane Handbook procedure varied. Conclusions: The results of this study could not suggest conclusive evidence that acupuncture is more effective than sham acupuncture in several diseases. In the future, more studies with rigorous acupuncture trials using sham controls should be conducted.
Cited by 3 works.