Six trials (n=699 participants, range 40 to 283) were included in the review. Study quality varied: two trials scored 5 and single trials scored 4, 3, 2 and 1. Follow-up, where reported, ranged from three to 12 months.
Compared with placebo, magnesium infusion significantly reduced the risk of death or dependency (RR 0.62, 95% CI 0.46 to 0.83; five studies), but did not significantly reduce the risk of delayed cerebral ischaemia.
Based on high-quality studies only, magnesium infusion significantly reduced the risk of death or dependency (RR 0.67, 95% CI 0.49 to 0.93; five studies) and delayed cerebral ischaemia (RR 0.64, 95% CI 0.44 to 0.94; three studies). The withdrawal rate for adverse effects was significantly greater for magnesium treatment compared to placebo (RR 9.98, 95% CI 3.04 to 32.74; three studies).
Significant heterogeneity was absent for these comparisons.