Twenty RCTs were included (n=1,054 patients). Median quality score was 4 out of 7 (range 1 to 6). Twelve studies were double blind, seven adequately described the randomisation method and four described allocation concealment.
Addition of clonidine was associated with a statistically significant increase in the duration of postoperative analgesia (WMD 123 minutes, 95% CI 74 to 169; 13 RCTs, 17 comparisons), sensory block (WMD 74 min, 95% CI 37 to 111; 10 RCTs, 13 comparisons) and motor block (WMD 141 min, 95% CI 82 to 199; seven RCTs, 11 comparisons). Significant heterogeneity was found in all three analyses (p<0.001).
Clonidine was associated with a significant decrease in time to onset of sensory block (WMD -2.2 minutes, 95% CI -4.1 to -0.4; eight studies; significant heterogeneity was found, p<0.001), but not in time to onset of motor block (four studies). There was no significant effect on incomplete anaesthetic block (16 studies).
Addition of clonidine was associated with a statistically significant increase in risk of arterial hypotension (OR 3.61, 95% CI 1.52 to 8.55; seven RCTs; NNH 11), orthostatic hypotension/faintness (OR 5.07, 95% CI 1.20 to 21.4; two RCTs; NNH 10), bradycardia (OR 3.09, 95% CI 1.10 to 8.64; seven RCTs; NNH 13) and sedation (OR 2.28, 95% CI 1.15 to 4.51; four RCTs; NNH 5). Data were homogeneous apart from sedation (p<0.04).
No evidence of dose responsiveness was found for any outcome.
Results of sensitivity analyses were reported.