AF is a common complication of CABG, especially in the elderly [26, 27]. Although long-term sequelae of postoperative AF are unusual, it frequently results in an increased length and cost of hospitalization. Aranki and colleagues found that the length of hospitalization directly attributable to AF was 4.9 days . This translated into more than $10,000 in hospital charges per patient . Therefore, any intervention that would reduce the incidence of postoperative AF or AA would result in a tremendous economic benefit.
AA is a common complication of CABG and beta-blockers have been shown to decrease the incidence of postoperative AA. Although magnesium has proven effective in reducing ventricular tachyarrhythmias and early mortality in acute myocardial infarction , its role in suppressing POAT remains controversial. This meta-analysis showed that magnesium could not significantly decrease the incidence of postoperative AA after CABG in addition to beta-blocker. In order to diminish the extent of AA, we used a subgroup analysis to explore the influence of POAT, and showed the same result.
Although the meta-analysis by Miller et al.  concluded that magnesium administration is an effective prophylactic measure for the prevention of postoperative AF, only 4 of the 20 studies included in the analysis were clearly in favor of magnesium administration [12, 20, 29, 30], with 7 studies showing no reduction in AF with magnesium prophylaxis [19, 31–36]. There are 4 potential reasons for these different results, including: (i) the potential for β error secondary to small sample sizes; (ii) different definitions of AF; (iii) different doses of magnesium administered; and (iv) different use of concomitant beta-blocker . In order to avoid the above shortness to some extent, the 5 RCTs in this meta-analysis have same characteristics, including: (i) the total dose of magnesium beyond 35 mmol; (ii) the intervention group and the comparison group used the same type of beta-blocker; and (iii) AA was detected by continuous ECG monitoring. As a result of the overwhelming data supporting the benefit of beta-blockers and some encouraging studies on the use of magnesium, we sought to determine whether magnesium as a sole or adjuvant agent in the prevention of AA after CABG. As a result, the combination of magnesium and beta-blocker did not significantly reduce the incidence of postoperative AA compared with beta-blocker alone.
One issue with the use of antiarrhythmic agents to prevent postoperative AA is that the majority of patients does not develop postoperative AA after cardiac surgery but would still be exposed to potential side effects of prophylactic intervention. From our meta-analysis, the risk of postoperative adverse events was higher in the combination of magnesium and beta-blocker group than beta-blocker alone group. The majority of adverse events were bradycardia and hypotension. An explanation for this phenomenon could be that intravenous administration of magnesium prolongs sinoatrial node conduction time, atrioventricular nodal refractory period, and PR and atrial-His intervals, as shown in electrophysiologic studies in healthy human subjects . It is assumed that these effects are amplified when magnesium is combined with beta-blocker. Further specific studies are needed to evaluate this hypothesis.
It has been reported that sotalol is not only a beta-blocker but also a potent potassium channel blocker. Therefore, we conducted a subgroup analysis and subdivided trials into sotalol group and other beta-blockers group. Both sotalol group and other beta-blockers group showed the same result that magnesium in addition to beta-blocker could not significantly decrease the incidence of postoperative AA. At the same time, trials were subdivided into low and high dosages of magnesium, and showed the same result. Because there was no use of amiodarone in all the included trials, so we can exclude the impact of amiodarone on the results of this meta-analysis.
Since postoperative AA has been associated with LOS extended, greater risk of major morbidity and accompanying increases in hospital costs, we conducted subgroup analyses about LOS and mortality. From this meta-analysis, magnesium in addition to beta-blocker did not significantly reduce the LOS in this meta-analysis. Since only 3 studies provided data on LOS, additional studies or data are warranted. Three trials reported data on mortality, and the result showed that the overall mortality was not affected by magnesium administration in addition to beta-blocker.
This meta-analysis is limited by the lack of studies and complete availability of relevant data, particularly for LOS, mortality and adverse effects. Each included trial had different categories of adverse events. For example, the trial by Bert et al.  demonstrated postoperative adverse events that included myocardial infarction and ventricular ectopic activity, whereas the trial by Geertman et al.  only included serious bradyarrhythmias and nonsustained ventricular tachycardia. The research by Solomon et al.  showed the adverse events including bradycardia and hypotension. Due to the limitation in the number of included studies, we did not have subdivision of adverse events, such as bradycardia and hypotension which often occur in the studies. According to the Cochrane Handbook , tests for funnel plot asymmetry should be used only when there are at least 10 studies included in the meta-analysis, because when there are fewer studies the power of the tests is too low to distinguish chance from real asymmetry. Therefore, Because of 5 articles in this meta-analysis, the funnel plot was not done in this meta-analysis. In addition, the number of patients in most studies was small.