sinaiem.org

Dinali did a stellar job this month at Journal Club; not only was her talk thought-provoking and clinically relevant, but I learned more about the history of of our world’s 8th most abundant element than I would have ever thought possible.

This month’s JC was primarily about Davey and Teubner’s AEM paper on using Magnesium sulfate as an adjunct to “usual care” for rate control in atrial fibrillation (AEM Vol 45, No 4, April ‘05, p347-353). We also touched upon some data from an all-Greek study in the Int’l Journal of Cardiology on Mg alone vs diltiazem alone in A-fib (Chilidakis, IJC 79 2001 p287-291).

Davey and Taubner looked at 199 Australians with rapid afib in this prospective, randomized, double-blinded, placebo-controlled trial. They got “standard rate-reduction agents” plus placebo, or te same agents plus 20 mEq of magnesium sulfate infused over 20 minutes plus another 20 mEq infused over 2 hours.

Problems with this setup were immediately apparent: First, Australia is a wondrous land, but different from our country. They use digoxin a lot over there, because hey, who really knows the LV function of our patients in the ED? Truthfully, that’s commendable, but Americans still reach for trusty beta-blockers in most cases; these Australians used dig 80% of the time. Thus, a trial comparing standard care +/- mag is going to be hard to apply to a land with such a different standard.

Also, they failed to enroll consecutive afib patients; it was at the EM attending’s discretion which patients were included. Maybe the really unstable were excluded (as per their methods, the pharmacy had to be notified to deliver “Solution A” or “B” to co-administer it with the standard rate reduction agent — if the patient was really sick, the doc may skip that step and just push diltiazem.)

Also, followup was only 150 minutes. Their outcomes also don’t include embolic stroke or reversion to afib.

Despite that, their results show the 102 magnesium-receivers were about twice as likely as the 99 placebos to have their rate fall below 100 beats / min (63 vs. 32), and more than twice as likely to sinus convert (25 vs. 11). You could say, then, that the risk of NOT controlling rate was about 66% in standard+placebo, but just 33% in standard+magnesium. So the number needed to treat (NNT is 1/ARR) is about 3, which is excellent.

NNT for sinus conversion was about 8. But hey, wouldn’t sinus conversion actually be dangerous if we’re dealing with a longterm afibber who’s not anticoagulated? Maybe, but probably not — these converters often go back and forth all the time… still, it would’ve been nice for the authors to comment on this. (For the record, Tintinalli recommends TEE to rule out thrombi, or 1-3 weeks of anticoagulation, before rhythm converting any patient who’s been in afib for more than two days. Emcrit.org cites a 2002 Mayo paper reporting 1.7-4.7% risk of embolism after DC cardioversion in those without acute afib.)

It’s worth noting that Davey et al did fail to achieve thedesired goal of getting a 15 bpm difference between the magnesium group and the control group. The adverse events they did notice with magnesium were local pain, a flushing sensation, and hypotension to less than 100 mmHgsystolic, and bradycardia. These side effects, seen in about a dozen altogether, easily outweighed the flush-free placebo group, but they don’t say how bad the hypotension was, and it’s hardly that surprising or especially worrisome.

So, all and all, we decided it was difficult to apply these results to our population, which usually gets a beta-blocker or diltiazem instead of digoxin as a first-line agent. But magnesium looks to be an attractive adjunct, especially in patients that are drunk, pregnant, asthmatic, or just plain low on mag. In fact, the Greek study cited above found mag alone to be at least as good as diltiazem alone, in reducing ventricular rate and converting rhythm. If a second agent is going to be needed for your afib patient who’s refractory to initial treatment, magnesium may be a more palatable alternative than mixing beta- and calcium-channel blockers. But of course, that statement is not evidence based… yet.

Posted on Sunday, August 20th, 2006 at 10:15 pm by Nick. Filed under Arrhythmias, Journal Club.
You may post a comment.