Beta-bloqueador - editorial

Perioperative [beta]-Blockade: How Best to Translate Evidence into Practice
[Editorial]
Fleisher, Lee A. MD

From the Department of Anesthesiology and Critical Care, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania.
Accepted for publication October 5, 2006.
Conflict of Interest: Dr. Fleisher is currently Chair, American College of Cardiology/American Heart Association Task Force Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery).
Address correspondence and reprint requests to Lee A. Fleisher, MD, Department of Anesthesia, University of Pennsylvania School of Medicine, 3400 Spruce St. Dulles 680, Philadelphia, PA 19104. Address e-mail to fleishel@uphs.upenn.edu.
During the 1980s, preoperative testing to identify patients with significant coronary artery disease and coronary revascularization was the mainstay of therapy to reduce the cardiac risk of noncardiac surgery (1). Beginning in the mid-1990s, several groups began to focus on postoperative monitoring and therapy, including the perioperative administration of [beta]-adrenergic blocking drugs (or [beta]-blockers) as a more effective approach (2,3). In 2002, Shojania et al. (4) published an evidence-based review funded by the Agency for Healthcare Research and Quality which identified perioperative [beta]-blockade for noncardiac surgery as a practice with the strongest basis in the literature. This followed a review of the literature on the use of perioperative [beta]-blockers that included clinical recommendations (5,6). These articles were followed by the establishment of perioperative quality of care measures by groups such as Leapfrog and the National Quality Forum which included perioperative [beta]-blockade. This in turn has led many groups of perioperative caregivers, including surgeons, anesthesiologists, cardiologists, and medical physicians, to debate the best protocols to accomplish this goal. In this issue of Anesthesia & Analgesia, the editors have chosen to publish a series of articles to help frame this debate, and better inform the clinician. By publishing the Guidelines from the American Heart Association/American College of Cardiology (ACC/AHA), a new meta-analysis, and a pro–con debate from leaders in the field, they hope to provide the individual practitioner with sufficient information to make his or her own informed decision and define the best protocol for their own interventions (7–10).

In trying to translate evidence into clinical practice, it is important to understand the different forms of evidence that frame this debate and how best to apply them (11). As all the articles published this month indicate, the strongest evidence for perioperative [beta] blockade comes from prospective randomized trials. Although several small randomized trials demonstrated a strong beneficial effect, others did not (12). Randomized trials offer the advantage of providing the strongest internal validity, but their external validity (i.e., ability to generalize the results) is less robust. In contrast, large cohort studies [e.g., administrative datasets used in the article by Lindenauer et al. (13)] offer insights into the efficacy of an intervention in routine clinical practice (i.e., external validity), but have much less internal validity. It is within this framework that the four articles are discussed.

The meta-analysis by Wiesbauer et al. (10) adds to a growing number of such analyses on this topic (12,14,15). The authors of the current meta-analysis focused on randomized controlled trials and included both published manuscripts and abstracts. By analytically combining these trials, the authors were unable to demonstrate an effect of [beta]-blockers on the hard end points of perioperative myocardial infarction or mortality. The clinician could therefore assume that either 1) [beta]-blockers are not effective, or 2) the studies included in the meta-analysis should not have been combined in the manner performed in the analysis because the populations or protocols used are different. If the latter is true, then the clinician should exert caution in specifics with regard to implementation of the protocols.

The pro–con debate (7,8) nicely illustrates this last point by outlining the issues related to interpretation of the data and how two groups of experts in this area choose to approach their own practice. On the pro side, the Dutch group led by Dr. Poldermans (8) clearly believes that many of the other trials did not control heart rate as tightly or provide perioperative [beta]-blockade for as long a duration as in their own studies. The importance of these comments is highlighted by two articles published by the group (16,17), after the editorial was accepted, which further demonstrate the beneficial effect of [beta]-blocker dosage and heart rate control on outcome. In contrast, Dr. London (7) presents a less expansive view with regard to the use of perioperative [beta]-blockade. He argues that the evidence is insufficient to generalize beyond the known literature, and outlines some of the deficiencies in both the evidence and the theoretical underpinnings of widespread treatment. As he notes in his conclusion, given the evidence, it is important to separate the mandatory use of [beta]-blockers as a quality assurance measure from their judicious use in the armamentarium to manage high risk patients.

It is in the context of this debate that the ACC/AHA produced a Focused Update on Perioperative Beta Blockade (9). Specifically, the American Medical Association Physician Consortium on Performance Improvement and the Surgical Care Improvement Project had both begun to evaluate the class and level of evidence to determine the appropriateness of developing performance measures based on continuation and initiation of perioperative [beta]-blockade in noncardiac surgery. It became increasingly important that the appropriate specialty societies weigh in and develop evidence-based guidelines upon which performance measures can be developed. As outlined in the introduction to the Guidelines, the American College of Cardiology has issued a formal position statement indicating that performance measures should be limited to Class I or Class III recommendations—those recommendations in which patients should or should not have the form of therapy—and that they should not include Class IIa or IIb recommendations, in which the evidence is less strong and for which opinion dictates the class of indications. The ACC/AHA mandates that only published trials be included in defining recommendations, and therefore, some of the literature discussed in the meta-analysis could not be included. Importantly, in developing Table 1 (see Ref. 9), we attempted to review the published literature and develop a schemata whereby recommendations for individual cohorts of patients can be easily changed to reflect new evidence.

So, how would I put it all together? Clearly, Class I recommendations should be followed, and therefore, patients receiving [beta]-blockers should be continued on [beta]-blockers, and patients with a positive stress test undergoing vascular surgery should be started on [beta]-blockers. There are large groups of patients currently not taking [beta]-blockers but who have Class I indications for [beta]-blockers independent of noncardiac surgery. For example, [beta]-blockers should be started and continued indefinitely in all patients who have myocardial infarction, acute coronary syndrome, or left ventricular dysfunction, with or without heart failure symptoms, unless contraindicated (18). As shown in multiple studies, many patients who are not taking [beta]-blockers present to vascular surgery with a history of a myocardial infarction (19,20). Therefore, there are patients who should be taking [beta]-blockers for long-term benefits, but for whom there is no evidence to demonstrate that acute administration in the perioperative period will impact outcome. For this reason, we considered such instances as Class IIa indications in the new guidelines, suggesting that they are likely beneficial but that this indication lacks evidence to mandate inclusion as a quality of care measure. In patients who do not have the above indications for [beta]-blockers independent of noncardiac surgery, there are now several trials and the meta-analysis that demonstrate no effect. The use of perioperative [beta]-blockers in the latter group thus represents a Class IIb indication. Finally, Lindenauer et al. (13) suggest harm in subpopulations of surgical patients without any coronary artery disease risk factors. If this harmful effect is shown in randomized trials, then this would qualify as a Class III indication.

The question remains regarding the best protocol to initiate perioperative [beta]-blockade. Ideally, these drugs should be started a week before surgery similar to the protocol by Poldermans et al. and titrated to heart rate-decreasing effect, but this is not always practical. Given emerging data suggesting that inadequate [beta]-blockade and heart rate control may be associated with worse outcomes, it is important to ensure that any protocol will yield the desired effect and not cause harm. Given the lack of data regarding the efficacy of starting [beta]-blockade the morning of surgery versus intraoperatively, the Surgical Care Improvement Project recently defined “appropriate” [beta]-blockade for patients who have not received this therapy before arrival at the operating room as initiating treatment before arriving in the postanesthesia care unit. This allows the caregivers to individualize therapy. In my opinion, further data are needed to understand the risks and benefits of starting [beta]-blockers in this group of patients, and that the results of the Perioperative Ischemic Evaluation (POISE) trial (21), a randomized controlled trial of metoprolol versus placebo in 10,000 patients undergoing noncardiac surgery, are eagerly awaited. The information and opinions in these four articles should allow clinicians to develop their own best approach to perioperative [beta]-blocker therapy in specific patient populations.