Am Fam Physician. 2010;81(6):717-719
Author disclosure: Nothing to disclose.
Each month, three presenters review an interesting journal article in a conversational manner. These articles involve “hot topics” that affect family physicians or “bust” commonly held medical myths. The presenters give their opinions about the clinical value of the individual study discussed. The opinions reflect the views of the presenters, not those of AFP or the AAFP.
This Month's Article
Devereaux PJ, Yang H, Yusuf S, et al., for the POISE Study Group. Effects of extended-release metoprolol succinate in patients undergoing non-cardiac surgery (POISE trial): a randomised controlled trial. Lancet. 2008; 371(9627):1839–1847.
Should we abandon the practice of using beta blockers to prevent myocardial infarction (MI) in persons undergoing noncardiac surgery?
What does this article say?
Mark: In a study of 8,351 persons older than 45 years who were beta blocker naïve and who were undergoing noncardiac surgery, 4,174 were randomized to extended-release metoprolol (Toprol XL), 200 mg per day, and 4,177 were randomized to placebo. The first 100-mg dose of metoprolol was given two to four hours before surgery. Twelve hours after surgery, patients were given an additional 100 mg of metoprolol if they had a heart rate of at least 80 bpm or a systolic blood pressure of at least 100 mm Hg. Twelve hours after the first postoperative dose, the patients were started on 200 mg of metoprolol per day. They were treated for a total of 30 days.
Any patient meeting the inclusion criteria (Table 1) who had a heart rate of at least 50 bpm and a systolic blood pressure of at least 100 mm Hg was eligible. Exclusion criteria included a known contraindication to beta blockers. Interestingly, a history of congestive heart failure was not an exclusion criterion. The reason that this is important will be discussed later.
|Coronary artery disease|
|Peripheral vascular disease|
|Hospitalization for congestive heart failure in|
|the previous three years|
|History of major vascular surgery|
|Any three of the following: |
Outcome measures included a composite of cardiovascular death, nonfatal MI, and nonfatal cardiac arrest at 30 days after randomization. Data were analyzed as intention to treat. The drug was withheld from patients with a heart rate of less than 50 bpm or a systolic blood pressure of less than 100 mm Hg.
Here's what they found: there were fewer MIs in the metoprolol group (4.2 versus 5.7 percent; P < .05; number needed to treat = 67), but more deaths (3.1 versus 2.3 percent; P < .05; number needed to harm [NNH] = 125) and more strokes (1.0 versus 0.5 percent; P < .05; NNH = 200).
Should we believe this study?
Mark: Sort of—we can't dismiss this information, but it needs to be tempered a bit. Who in their right mind is going to give a patient who is beta blocker naïve, with a systolic blood pressure of 100 mm Hg and a heart rate of 50 bpm, a 100-mg dose of metoprolol two hours before surgery? And who is going to use 100 mg as a starting dose for someone with a history of congestive heart failure? This study seems like a setup for causing low-flow states and, as a result, more strokes (and falls, orthostatic hypotension, etc.).
In fact, there were more hypotension and bradycardia in the treatment group (P < .0001 for both outcomes) and more strokes in the group with perioperative hypotension. Could the protocol have been the cause of the additional strokes and deaths? We can't answer this question based on this study, but the choice of the protocol seems like a setup for finding adverse events with metoprolol.
Andrea: It would have been a more powerful study if it had involved a real-world scenario. Start the beta blocker approximately a week ahead of time, and shoot for a systolic blood pressure of 110 mm Hg and a heart rate of 60 bpm. Other studies that have used a lower dose of beta blockers have shown a cardiovascular benefit without any additional strokes.1,2
Mark: Another thing that is somewhat questionable is that, based on the inclusion criteria, the study authors would give a beta blocker to a 45-year-old male smoker who is hypertensive and undergoing urgent intra-abdominal surgery. Maybe this patient is a trauma patient. It certainly doesn't make intuitive sense to give him a beta blocker.
Andrea: It is good that the authors used an intention-to-treat analysis. This means that all of the patients were analyzed in the group to which they were assigned. For example, those who couldn't tolerate the beta blocker were still analyzed as part of the beta-blocker group. Often you will see a per-protocol analysis. Anytime you see this, head the other way. That means that the researchers can exclude any patient that didn't do well on the study drug.
Bob: There was a meta-analysis of beta blockers in noncardiac surgery published within a month of this study.3 If you dissect the meta-analysis and exclude the POISE study, there is no change in cardiovascular mortality (odds ratio [OR] = 0.78; 95% confidence interval [CI], 0.41 to 1.46), but there is a reduction in nonfatal MI (OR = 0.47; 95% CI, 0.30 to 0.74). This seems like a worthy goal. That said, the authors of the meta-analysis caution that the American College of Cardiology and American Heart Association guidelines committee “should soften their advocacy for this intervention until conclusive evidence is available.”
What should the family physician do?
Mark: I think the bottom line is that patients who need to be on a beta blocker (e.g., those with congestive heart failure or coronary artery disease) should start well before surgery and should have the dose stabilized before surgery. We know that patients with an indication for beta blockers benefit if we continue their beta blocker through the surgical period.4 In fact, the trials that were most positive were those that studied beta blockers in high-risk patients.3
Andrea: I agree. It doesn't seem as though this study should change our practice in toto. However, we should be a bit more circumspect and plan ahead for our patients who might need surgery.
Bob: For now, we should try to limit the use of perioperative beta blockers to high-risk patients, and we should use a reasonable dose. Start at least a week ahead and titrate to a reasonable heart rate and blood pressure.
addendum: We are prescient. The American Heart Association came out with a position paper after this was written suggesting that beta blockers should be continued in patients who are already on them, and that beta blockers titrated to heart rate and blood pressure are reasonable in patients in whom preoperative assessment identifies coronary artery disease or high cardiac risk (defined by the presence of more than one clinical risk factor) who are undergoing intermediate-risk surgery.5
|• The data on the use of beta blockers during noncardiac surgery remain inconclusive.|
|• Although the POISE study is an important contribution to the literature, it uses a protocol that is unrealistic. It should not change our practice on beta blockers wholesale.|
|• For now, we should try to limit the use of perioperative beta blockers to high-risk patients, and we should use a reasonable dose. Consider using the protocol in one of the Dutch studies and titrate the dose up slowly until you reach a reasonable heart rate and blood pressure.1,2|
|• Plan ahead. Get your patients with an indication for beta-blocker therapy on a stable, tolerable dose.|
|• Make sure that a study protocol makes sense. Don't assume that a study uses an appropriate dose or schedule of a drug.|
|• Look for an intention-to-treat analysis. This type of analysis evaluates all patients in the group to which they were assigned. Thus, persons who didn't tolerate the treatment (for example) are still analyzed in their original group. This reflects our real-world experience. Some of our patients are going to do badly or stop a medication. They still need to be included in our equation when we decide whether or not to use a drug.|
|• Avoid articles with a per-protocol analysis. This allows the researcher to throw out any data from patients who don't tolerate a drug (for example). So, a per-protocol analysis may not reflect the kind of results we will see in our practice and typically overestimates the net benefit of an intervention.|