Thirty-six RCTs were included (n=1,908, sample size 20 to120). Study quality was good, with a median Jadad score of 4 out of 5 (range 2 to 5).
Intravenous morphine (13 RCTs, 22 comparisons)
Efficacy: A significant benefit for morphine was seen for at least one outcome in all three comparisons that used an inactive control intervention, but in only one of 19 comparisons that used active controls.
Side effects: Nausea, vomiting and sedation were significantly more common with morphine in 10 of 22 comparisons.
Epidural morphine (12 RCTs, 14 comparisons)
Efficacy: A significant benefit for morphine was seen for at least one outcome in both comparisons that used an inactive control intervention, but in only two of 12 comparisons that used active controls.
Side effects: Nausea and/or vomiting, sedation, pruritus and/or respiratory depression were significantly more common with morphine in seven of 14 comparisons.
Intramuscular morphine (eight RCTs, nine comparisons)
Efficacy: A significant benefit for morphine was seen for at least one outcome in both comparisons that used an inactive control intervention, but in only two of seven comparisons with that used active controls.
Side effects: Vomiting was significantly more common with morphine in four of nine comparisons.
Intrathecal morphine (two RCTs, three comparisons)
Efficacy: A significant benefit for morphine was seen for at least one outcome for all three comparisons, all of which used an inactive control intervention. Side effects: No significant difference was found between the groups in the incidence of side effects.
No association between morphine dose and efficacy was noted.