RCTs of needle acupuncture compared with placebo, Western medication or no treatment for participants with a chief complaint of insomnia were eligible for inclusion in the review. Studies of other types of acupuncture (such as auricular, acupressure and laser) were excluded.
Participants in the included studies varied widely in clinical characteristics (for example, pregnancy, stroke, spondylosis), duration of insomnia (range four days to 28 years, where reported) and mean age (27 to 71 years, where reported). Methods of diagnosing insomnia varied; Chinese Classification of Mental Disorder criteria were frequently used. Treatment was usually daily, using either a standardised or an individualised regimen. Treatments varied widely in acupoints used, number of main acupoints treated (range four to 16) and duration (range two to 56 days). Stimulation, where used, was manual or electrical (where reported). In most studies the comparator was a benzodiazepine; other control interventions included sleep hygiene counselling and sham acupuncture. Few studies used a standardised outcomes measurement tool. Efficacy outcomes included effectiveness rate (proportion of participants with improved symptoms). Sleep quality scores such as insomnia severity index (ISI). The review also reported adverse events. All studies assessed only short-term outcomes. Most studies were set in China.
Two reviewers independently selected the studies. Disagreements were resolved by discussion.