Eight RCTs were included in the review. Several trials compared magnesium sulfate with more than one other treatment. Two trials compared magnesium sulfate with placebo (92 in the magnesium sulfate group and 99 in the untreated group). Four trials compared magnesium sulfate with ritodrine (182 in the magnesium sulfate group and 166 in the ritodrine group). Six trials compared magnesium sulfate with a beta-agonist agent including ritodrine (222 in the magnesium sulfate group and 261 in the beta-agonist group).
The primary reviewers initially agreed on 78.5% of the items in the trial evaluation. The kappa-statistics ranged from 0.13 to 1.0 for the various items assessed, with the lowest scores observed for "'intention to treat analysis' and 'blinded randomisation'.
In the 2 trials of magnesium sulfate versus placebo, the results were not significant for either the delay of delivery (greater than 48 hours and greater than 7 days) or delivery after 37 weeks; inter-study heterogeneity was also insignificant.
In the trials of magnesium sulfate versus ritodrine, there was no significant difference between treatments for either the delay of delivery (greater than 48 hours and greater than 7 days) or delivery after 37 weeks; inter-study heterogeneity was also insignificant. Major adverse drug effects were uncommon and the treatments did not differ in terms of the occurrence of these events. The mean latency between treatment groups was 0.2 days (95% CI: -4.86, 5.26), showing no substantial difference.
In the trials of magnesium sulfate versus beta-agonists, there was no significant difference between treatments for either the delay of delivery (greater than 48 hours and greater than 7 days) or delivery after 37 weeks. There was a significant improvement in the mean latency in the beta-agonist group relative to magnesium sulfate, but there was considerable heterogeneity between 3 of the 6 trials for this outcome. Major adverse drug effects were uncommon and the treatments did not differ in terms of the occurrence of these events. However, compared with beta-agonists, there was a significant reduction in the odds of magnesium sulfate being discontinued because of side-effects.