Eleven studies were excluded from the meta-analysis because their results could not be combined with the other studies accordingly.The remaining 22 trials were then grouped according to whether the pain being treated was acute or chronic, along with the style of acupuncture practiced and the other types of interventions used.
Data from the trials was extracted and analyzed based on the following outcomes: short-term effectiveness on pain, long-term effectiveness on pain, and short- and long-term effects on functional status and overall improvement. In addition, the quality of the trials was computed using two measurements: the Jadad quality score and the Cochrane Back Review Group quality score.
Short-term effectiveness on pain. Acupuncture was found to be "statistically significantly more effective" than sham acupuncture, sham transcutaneous electrical nerve stimulation (TENS) and no additional treatment. Compared to sham acupuncture, real acupuncture was 58 percent more effective in relieving pain, which the researchers equated to an improvement of 14.5 points on a 100-point visual analogue scale. Results comparing acupuncture to other active treatments were mixed, however. Acupuncture appeared more effective in three out of four studies using real TENS and two out of three studies comparing acupuncture and pain medications, but was less effective compared to massage, and "statistically significantly less effective" than spinal manipulation.
Long-term effectiveness on pain. Acupuncture was statistically significantly more effective than no additional treatment or sham TENS, and was an average of 61 percent more effective compared to sham acupuncture. Two studies found acupuncture to be more effective than TENS; one trial suggested that acupuncture was statistically significantly worse than massage.