Twenty-five publications reporting on 28 RCTs were included.
The quality scale ratings ranged from 2 to 5 (median 3; 15 of the 25 comparisons had scores of at least 3) and the validity scores ranged from 2 to 11 (median 8).
Sensitive trials
Intra-articular morphine versus placebo, immediate period (0 to 2 hours): Fifteen comparisons were sensitive (7 positive, 8 negative). Of the eight negative comparisons, seven were of 1mg and one of 4mg. Of the seven positive comparisons, one was of 1mg, one of 3mg and five were of 5mg.
Intra-articular morphine versus placebo, early period (2 to 6 hours): Twelve comparisons were sensitive (8 positive, 4 negative). Of the four negative comparisons, three were of 1mg and one of 4mg. Of the eight positive comparisons, two were of 1mg, one of 3mg and five were of 5mg.
Intra-articular morphine versus placebo, late period (6 to 30 hours): Thirteen comparisons were sensitive (10 positive, 3 negative). All three negative comparisons were of 1 mg. Of the ten positive comparisons, four were of 1mg, one of 3mg, one of 4mg and four of 5mg.
Intra-articular morphine versus placebo, PCA outcome trials: Two comparisons were sensitive. In both studies, statistically more PCA morphine was used in the placebo groups than in the intra-articular morphine (2mg and 5mg) groups.
Dose-response trials: One comparison was sensitive; 5mg intra-articular morphine was shown to be statistically superior to 1mg intra-articular morphine.
Cross-route comparison: One cointersmparison was sensitive for immediate and late outcomes; no statistical difference was shown between 5mg of intra-articular morphine and 5mg intramuscular morphine.
Insensitive trials
Immediate period (0 to 2 hours): All 10 insensitive trials were negative.
Early period (2 to 6 hours): Nine of the 10 insensitive trials were negative.
Late period (6 to 30 hours): Ten of the 11 insensitive trials were negative.
Dose-response: None of the three dose response comparisons showed a difference between the groups.
Cross-route comparisons; none of the two comparisons showed a difference between intra-articular and intramuscular administration, or intra-articular and subcutaneous administration.