Ten RCTs (n=515) were included in the review: five placebo-controlled and five comparing magnesium with alternative anti-arrhythmic drugs.
Study quality scores ranged from 2 to 5. All trials were randomised, but only the 5 placebo-controlled studies were double-blinded. All except one of the trials conducted an intention-to-treat analysis. The proportion of patients lost to follow-up was less than 10% in all trials.
Rhythm conversion (10 RCTs): 5 placebo-controlled trials showed no benefit of adding magnesium to digoxin or ibutilide (OR 1.22, 95% CI: 0.56, 2.65, p=0.61). There was little evidence of heterogeneity (I2=25.8%; p=0.24). There was also no difference in rhythm conversion in 5 trials that compared magnesium with alternative anti-arrhythmic drugs (OR 2.82, 95% CI: 0.64, 12.43, p=0.17). There was strong evidence of heterogeneity (I2=66.7%, p=0.02).
Ventricular response rate (5 RCTs): 3 placebo-controlled trials reported a beneficial effect of adding magnesium to digoxin (OR 3.23, 95% CI: 1.93, 5.42). However, 2 trials that compared magnesium with alternative anti-arrhythmic drugs found that magnesium was less effective than veraparril (OR 0.19, 95% CI: 0.09, 0.44). There was no evidence of heterogeneity for either meta-analysis (I2=0%, p=0.47 and p=0.99).
Risk of significant bradycardia (7 RCTs): 3 placebo-controlled trials showed no benefit of adding magnesium to digoxin (OR 3.47, 95% CI: 0.54, 22.22, p=0.19). Four trials that compared magnesium with alternative anti-arrhythmic drugs found that magnesium was less effective (OR 0.13, 95% CI: 0.02, 0.76, p=0.02). There was no evidence of heterogeneity for either meta-analysis (I2=0%, p=0.74 and p=0.96).
Risk of atrioventricular block (5 RCTs): 2 placebo-controlled trials showed no benefit of adding magnesium to digoxin (OR 3.92, 95% CI: 0.43, 35.69, p=0.23). Four trials that compared magnesium with alternative anti-arrhythmic drugs found that magnesium was less effective (OR 0.09, 95% CI: 0.01, 0.77, p=0.02). There was little evidence of heterogeneity (I2=0%, p=0.05).
Incidence of flushing, tingling and dizziness (4 RCTs): magnesium was associated with an increased risk of minor adverse effects in 3 placebo-controlled trials and one trial comparing magnesium with diltiazem (OR 14.52, 95% CI: 3.72, 56.72). There was no evidence of heterogeneity (I2=0%, p=0.63).
Proportion of patients requiring rescue anti-arrhythmic drugs (number of studies unclear): there was no difference in the proportion of patients requiring rescue anti-arrhythmic drugs among placebo-controlled trials or among those that compared magnesium with alternative anti-arrhythmic drugs (OR 0.84, 95% CI: 0.25, 2.81).
The funnel plot showed slight asymmetry.