Fifty-three RCTs (n=2,721) were included.
The median modified Oxford score for quality was 4 out of a possible 7 (range: 2 to 6).
Efficacy.
Pain intensity (10 RCTs): there was a statistically significant reduction in pain intensity at rest with ketamine compared with control at 6 hours (WMD -0.89 cm, 95% CI: -1.13, -0.65), 12 hours (WMD -0.42 cm, 95% CI: -0.72, -0.11), 24 hours (WMD -0.35 cm, 95% CI: -0.60, -0.09) and 48 hours (WMD -0.27 cm, 95% CI: -0.52, -0.02). There was no evidence of a relationship between the dose of ketamine and its analgesic efficacy.
Cumulative morphine consumption at 24 hours (5 RCTs): there was a statistically significant reduction in cumulative morphine consumption at 24 hours with ketamine compared with control (WMD -15.7 mg; 95% CI: -20.9, -10.5).
Opioid-related adverse effects (9 RCTs): there were no statistically significant differences between ketamine and control for any of the specified opioid-related adverse effects (nausea, vomiting, nausea or vomiting, pruritus, drowsiness or urinary retention).
Time to first request for analgesia (7 RCTs): the mean reduction in time to first request for analgesia was 15.7 minutes (95% CI: 12.5, 18.9) with ketamine compared with control.
Efficacy results for other ketamine regimens were also reported.
Safety.
Hallucinations (30 RCTs): there was a statistically significantly increased risk of hallucinations with ketamine (with or without benzodiazepines) in awake or sedated patients compared with controls: 7.4% versus 3.7%. The OR was 2.32 (95% CI: 1.09, 4.92) and the NNH was 21.
The risk of hallucinations with ketamine (regardless of use or not of benzodiazepines) in patients undergoing general anaesthesia was low: 0.8% versus 0.4%. The OR was 1.49 (95% CI: 0.18, 12.6) and the NNH was 286.