Magnesium as an adjuvant to postoperative analgesia: a systematic review of randomized trials.

BACKGROUND

Randomized trials have reached different conclusions as to whether magnesium is a useful adjuvant to postoperative analgesia.

METHODS

We performed a comprehensive search (electronic databases, bibliographies, all languages, to 4.2006) for randomized comparisons of magnesium and placebo in the surgical setting. Information on postoperative pain intensity and analgesic requirements was extracted from the trials and compared qualitatively. Dichotomous data on adverse effects were combined using classic methods of meta-analysis.

RESULTS

Fourteen randomized trials (778 patients, 404 received magnesium) tested magnesium laevulinate, gluconate or sulfate. With magnesium, postoperative pain intensity was significantly decreased in four (29%) trials, was no different from placebo in seven (50%), and was increased in one (7%); two trials (14%) did not report on pain intensity. With magnesium, postoperative analgesic requirements were significantly reduced in eight (57%) trials, were no different from placebo in five (36%), and were increased in one (7%). Magnesium-treated patients had less postoperative shivering (relative risk 0.38, 95% confidence interval 0.17-0.88, number-needed-to-treat 14). Seven trials reported on magnesium serum levels. In all, serum levels were increased in patients who received magnesium; in six, serum levels were decreased in those who received placebo.

CONCLUSIONS

These trials do not provide convincing evidence that perioperative magnesium may have favorable effects on postoperative pain intensity and analgesic requirements. Perioperative magnesium supplementation prevents postoperative hypomagnesemia and decreases the incidence of postoperative shivering. It may be worthwhile to further study the role of magnesium as a supplement to postoperative analgesia, since this relatively harmless molecule is inexpensive, and the biological basis for its potential antinociceptive effect is promising.