PloS one 2017 10 2712(10) e0186616 doi 10.1371/journal.pone.0186616
Endometriosis is a multifactorial, oestrogen-dependent, inflammatory, gynaecological condition that can result in long-lasting visceral pelvic pain and infertility. Acupuncture could be an effective treatment for endometriosis and may relieve pain. Our aim in the present study was to determine the effectiveness of acupuncture as a treatment for endometriosis-related pain.
In December 2016, six databases were searched for randomised controlled trials that determined the effectiveness of acupuncture in the treatment of endometriosis-related pain. Ultimately, 10 studies involving 589 patients were included. The main outcomes assessed were variation in pain level, variation in peripheral blood CA-125 level, and clinical effective rate. All analyses were performed using comprehensive meta-analysis statistical software.
Of the 10 studies included, only one pilot study used a placebo control and assessed blinding; the rest used various controls (medications and herbs), which were impossible to blind. The sample sizes were small in all studies, ranging from 8 to 36 patients per arm. The mean difference (MD) in pain reduction (pre- minus post-interventional pain level-measured on a 0-10-point scale) between the acupuncture and control groups was 1.36 (95% confidence intervals [CI] = 1.01-1.72, P<0.0001). Acupuncture had a positive effect on peripheral blood CA-125 levels, as compared with the control groups (MD = 5.9, 95% CI = 1.56-10.25, P = 0.008). Similarly, the effect of acupuncture on clinical effective rate was positive, as compared with the control groups (odds ratio = 2.07; 95% CI = 1.24-3.44, P = 0.005). CONCLUSIONS
Few randomised, blinded clinical trials have addressed the efficacy of acupuncture in treating endometriosis-related pain. Nonetheless, the current literature suggests that acupuncture reduces pain and serum CA-125 levels, regardless of the control intervention used. To confirm these findings, additional, blinded studies with proper controls and adequate sample sizes are needed.