Twenty-seven RCTs were included in meta-analyses (n=1,205 patients). The median quality score of trials was 3 (ranging from 1 to 6).
Compared with controls, intrathecal morphine was associated with a significant reduction in intraoperative fentanyl or sufentanil consumption (WMD -145µg, 95% CI -181 to -109; nine RCTs) and postoperative 24 hours morphine consumption (WMD -16.9mg, 95% CI -23.7 to -10.1; 11 RCTs). Intrathecal morphine significantly reduced pain intensity at rest at four hours after surgery (WMD -1.9cm, 95% CI -2.9 to -0.8; five RCTs), at 12 hours after surgery (WMD -0.8cm, 95% CI -1.4 to -0.1; seven RCTs), and at 24 hours post surgery (WMD -1.0cm, 95% CI -1.7 to -0.4; eight RCTs). There were also significant reductions in pain intensity on movement at 12 hours after surgery (WMD -2.0cm, 95% CI -3.1 to -1.0; four RCTs) and at 24 hours after surgery (WMD -1.7cm, 95% CI -2.7 to -0.8; four RCTs). Intrathecal morphine was also associated with a significant reduction in the duration of hospital stay (WMD -0.49 day, 95% CI -0.89 to -0.09; eight RCTs).
Intrathecal morphine was associated with a significant increase in the incidence of respiratory depression (OR 7.86, 95% CI 1.54 to 40.3; NNH 84, 95% CI 47 to 409; 21 RCTs) and pruritus (OR 3.85, 95% CI 2.40 to 6.15; NNH 6, 95% CI 5 to 9; 18 RCTs).
Significant heterogeneity was only observed in the outcome of postoperative 24 hours morphine consumption (P<0.001), pruritus (p=0.04), and all the outcomes of postoperative pain intensity (p values not reported).
Subgroup analyses showed a significantly higher reduction in postoperative 24 hours morphine consumption after abdominal surgery, compared with after cardiac-thoracic surgery.
Sensitivity analyses did not materially affect the results. There was no evidence of a dose-response effect.