Honam Research Center, Medifarm Hospital, YangYul-gil 306, Suncheon 540-300, Republic of Korea; Department of Oriental Medicine, Medifarm Hospital, YangYul-gil 306, Suncheon 540-300, Republic of Korea.
Honam Research Center, Medifarm Hospital, YangYul-gil 306, Suncheon 540-300, Republic of Korea; Department of Rehabilitation, Medifarm Hospital, YangYul-gil 306, Suncheon 540-300, Republic of Korea.
J Clin Epidemiol. 2014 Aug;67(8):858-69. doi: 10.1016/j.jclinepi.2014.02.021. Epub 2014 Apr 26.
To investigate the nocebo effect using randomized acupuncture trials that include sham and no-treatment groups.
Electronic databases were searched to retrieve eligible trials from their inception until June 2013. Risk differences were then calculated using the acupuncture and sham groups to determine the treatment effect and the sham and no-treatment groups to determine the nocebo effect.
In total, 58 eligible trials were analyzed. On the basis of the rate of patients with any adverse event in 31 trials reporting available data, the treatment effect was 0.012 (95% confidence interval [CI]: 0.003, 0.021), with a number needed to harm (NNH) of 83 (95% CI: 48, 333). The nocebo effect was 0.049 (95% CI: 0.012, 0.086), with an NNH of 20 (95% CI: 12, 83). By contrast, the rate of dropouts due to adverse events in 39 trials reporting available data showed no differences for both effects. In addition, nearly 70% of the trials reported zero dropouts in the sham and no-treatment groups.
Our findings suggest that (1) the nocebo effect of acupuncture is clinically meaningful and (2) the rate of patients with any adverse event may be a more appropriate indicator of the nocebo effect.
通过纳入假针和无治疗组的随机针刺试验来研究非安慰剂效应。
检索电子数据库，检索从成立到 2013 年 6 月期间符合条件的试验。然后使用针刺组和假针组计算治疗效果，使用假针组和无治疗组计算非安慰剂效应，以计算风险差异。
共分析了 58 项符合条件的试验。基于 31 项报告了可用数据的试验中出现任何不良事件的患者比例，治疗效果为 0.012（95%置信区间：0.003，0.021），需要治疗的患者数（NNH）为 83（95%置信区间：48，333）。非安慰剂效应为 0.049（95%置信区间：0.012，0.086），NNH 为 20（95%置信区间：12，83）。相比之下，在报告了可用数据的 39 项试验中，因不良事件而退出的比例对于这两种效应均无差异。此外，近 70%的试验报告假针和无治疗组的退出率为零。
我们的研究结果表明，（1）针刺的非安慰剂效应具有临床意义，（2）出现任何不良事件的患者比例可能是评估非安慰剂效应的更合适指标。
