A Cochrane review included 14 studies on acute stroke, with a total of 1280 subjects. Ten trials included patients with only ischaemic stroke. Compared with sham acupuncture or open control, there was a borderline significant trend towards fewer patients being dead or dependent with acupuncture (Odds ratio [OR] 0.66, 95% confidence interval [CI] 0.43 to 0.99). After three months or more, significantly fewer were dead or needed institutional care (OR 0.58, 95% CI 0.35 to 0.96) in the acupuncture group. There was also a significant difference favouring acupuncture in the mean change of global neurological deficit score during the treatment period (standardized mean difference (SMD) 1.17, 95% CI 0.30 to 2.04). Comparison of acupuncture with sham acupuncture only showed a statistically significant difference on death or requiring institutional care (OR 0.49, 95% CI 0.25 to 0.96), but not on death or dependency (OR 0.67, 95% CI 0.40 to 1.12), or change of global neurological deficit score (SMD 0.01, 95% CI -0.55 to 0.57).
Another Cochrane review on acupuncture for stroke in the subacute or chronic stage (abstract , review ) included 5 studies with a total of 368 subjects. Methodological quality was considered inadequate in all trials. The overall estimate suggested the odds of improvement in global neurological deficit to be higher in the acupuncture group compared with the control group (OR 6.55, 95% CI 1.89 to 22.76; 4 trials), but there was substantial heterogeneity (I2 = 68%). One trial showed no significant improvement of motor function between the real acupuncture group and the sham acupuncture group (OR 9.00, 95% CI 0.40 to 203.30), but the confidence interval was wide and included clinically significant effects in both directions. No data on death, dependency, institutional care, change of neurological deficit score, quality of life or adverse events were available.