Three RCTs were included in the review (n=166 patients). The methodological quality of included trials was poor. None reported methods of randomisation, details of allocation concealment, described patient or assessor blinding, or gave sufficient details of drop-outs and withdrawals. The degree of confidence that acupuncture was applied appropriately ranged from 50 to 90%.
Two RCTs were combined in meta-analysis and showed no statistically significant difference in effect between electro-acupuncture and conventional drug therapy, with no evidence of statistical heterogeneity.
Individual trials had mixed outcomes relating to the superiority of acupuncture or other therapies. Meta-analysis of two RCTs showed that drug therapy was superior to acupuncture for activities of daily living (WMD -1.29, 95% CI -1.77 to -0.80; n=72 patients). There was no statistically significant heterogeneity. In two trials comparing electro-acupuncture alone or electro-acupuncture plus drug therapy versus drug therapy alone, there were no statistically significant effects in favour of acupuncture in terms of response rates.