|
|
||||||||
a Division of Cardiac Surgery, University of Ottawa Heart Institute, Ottawa, Ontario, Canada
b Division of Cardiology, University of Ottawa Heart Institute, Ottawa, Ontario, Canada
Accepted for publication March 23, 2009.
* Address correspondence to Dr Rubens, Division of Cardiac Surgery, University of Ottawa Heart Institute, 40 Ruskin St, Ottawa, Ontario, K1Y 4W7, Canada (Email: frubens{at}ottawaheart.ca).
| Abstract |
|---|
|
|
|---|
Methods: A link to an online survey was e-mailed to all practicing cardiac surgeons in Canada. Each surgeon was given a unique log-in identification number to complete the survey online through a secure web page.
Results: Surveys were sent to 166 surgeons; 119 completed surveys (72%) were returned. Only 58% of respondents routinely use β-blockade for prophylaxis. For nonusers, 44% are unconvinced of the evidence for this practice. The routine use of amiodarone among surgeons was 19%. Of the remainder, 43% cited a perceived increased risk of complications as the reason for not using this therapy. An additional 29% considered the therapy was excessively complicated or time consuming. Corticosteroids were routinely used by only one surgeon. Major barriers to use of steroids were unconvincing evidence (76%), a perceived increased risk of wound infection (38%), and hyperglycemia (30%).
Conclusions: Despite level 1 evidence, the use of β-blockers, amiodarone, and corticosteroids for prophylaxis of atrial fibrillation among Canadian surgeons remains less than expected. The results of this survey support the need for further clinical trials with robust and clinically relevant outcomes that may further influence surgeons to adopt this practice.
| Introduction |
|---|
|
|
|---|
There have been numerous trials evaluating strategies for AF prophylaxis after cardiac surgery. The most extensively studied agents include β-blockade, amiodarone, sotalol, magnesium, digoxin, and calcium-channel blockers. Of these, only β-blockade, sotalol, and amiodarone have proven consistently effective at decreasing the incidence of AF [14–16]. There is level 1 evidence supporting the efficacy and safety of β-blockers for this indication [14, 15]. Based on current data, the use of amiodarone has also emerged as an effective therapy for the prophylaxis of AF [14–20]. More recently, corticosteroids have been investigated as an attractive pharmacologic strategy for preventing the occurrence of AF. Published data demonstrate that corticosteroid treatment was associated with a lower incidence of postoperative AF as compared with placebo [21–25].
Despite the strength of the evidence for the use of these medications for the prophylactic treatment of postoperative AF, the authors hypothesized that these medications were being underutilized by Canadian cardiac surgeons. As such, we designed a survey to assess the frequency with which these medications are currently being applied by surgeons and to identify concerns and barriers to more widespread application of these prophylactic strategies.
| Material and Methods |
|---|
|
|
|---|
A list of all cardiac surgeons practicing in Canada was developed. The accuracy of the list was confirmed by cross referencing it to online databases such as CTS Net, university websites, and e-mails to program directors and division chiefs. Surgeons included in the study practiced adult cardiac surgery. Surgeons who were no longer in clinical practice and pediatric cardiac surgeons were excluded from participation. The final list consisted of 166 surgeons.
The survey was developed as an online tool in a user-friendly format. A link to an online survey was e-mailed to all practicing cardiac surgeons in Canada. Each surgeon was assigned a unique log-in identification for identification purposes. The survey was completed online through a secure web page. To maximize the response rate, nonresponders were e-mailed multiple times or contacted by fax or telephone. A total of 593 e-mails and 81 faxes were sent. After one e-mail, 16% of surgeons completed the survey, 22% completed it after two e-mails, and 8% after three e-mails. The remaining 54% received four or more e-mails and faxes.
| Results |
|---|
|
|
|---|
When asked if they initiated β-blockade for AF prophylaxis for patients who were not previously on β-blocker therapy, only 58% of respondents reported doing so routinely. The remaining 42% reported using β-blockers only sometimes, rarely, or never (Fig 1). For those who do not routinely use β-blockers, 44% are unconvinced of the evidence for this practice. A further 12% preferred an alternative therapy, and 7% did not use β-blockers owing to concern of increased side effects (Table 1). There was no difference in the routine usage of β-blockade between surgeons in the first 10 years of practice and more senior surgeons (p = 0.8).
|
|
Surgeons were also asked to estimate the absolute risk reduction for postoperative AF using amiodarone based on their knowledge of the literature. Seven percent of respondents said it was less than 10%, 24% thought it was 10%, 26% thought it was 20%, 19% thought it was 30%, and 10% of respondents thought the absolute risk reduction was greater than 30% with prophylactic amiodarone. The remaining 15% of surgeons reported that they were unaware of the data.
Corticosteroids were routinely used by one surgeon. The vast majority (92%) reported that they never used steroids (Fig 1). Major barriers to use of steroids were unconvincing evidence (75%), a perceived increased risk of wound infection (39%), and hyperglycemia (30%). A further 24% were concerned about other steroid-related complications, and 5% thought the use of steroids was too cumbersome or time consuming (Table 2). Surgeons were also asked to estimate the absolute risk reduction for postoperative AF using corticosteroids based on their knowledge of the literature. Of the respondents, 30% percent said it was less than 10%, 12% thought it was 10%, 10% thought it was 20%, 4% thought it was 30%, and 2% of respondents thought the absolute risk reduction was greater than 30%. Forty-three percent of surgeons reported that they were unaware of the data.
|
| Comment |
|---|
|
|
|---|
The current published evidence supports the use of β-blockers for AF prophylaxis. In this survey, we have observed that less than 60% of surgeons administered a β-blocker for AF prophylaxis for patients who were not already receiving one. Of those who did not start a β-blocker, 44% remained unconvinced of the evidence. There were two recent meta-analyses examining the use of β-blockade for this indication. The first demonstrated that β-blockers reduced the percentage of patients with AF from 33% in the control group to 19% in the β-blocker group (odds ratio [OR] 0.39; 95% confidence interval [CI]: 0.28 to 0.52). In a 2006 meta-analysis, β-blockers significantly reduced the incidence of postoperative AF (OR 0.36, 95% CI: 0.28 to 0.47) [15].
Despite the strong evidence, concerns about routine β-blocker use still persist. Trials of β-blocker prophylaxis frequently excluded patients with an ejection fraction less than 30%, bronchospasm, type 1 diabetes mellitus, atrioventricular block, or sick sinus syndrome [1]. In addition, the effects of β-blockade may have been overestimated because of trial heterogeneity and the spurious effect of β-blocker withdrawal [15].
Amiodarone is used by less than 20% of surgeons for AF prophylaxis. Of those who did not use amiodarone routinely, 33% remained unconvinced of the evidence. Several large randomized controlled trials [17–20] and three meta-analyses [14–16] have been published on the prophylactic use of this drug. All reported statistically significant reductions in AF in the intervention groups compared with placebo.
In general, however, amiodarone trials excluded patients with a low resting heart rate, second- or third-degree atrioventricular block, or New York Heart class III or IV congestive heart failure [1]. In addition, rates of concomitant β-blocker utilization in these studies were highly variable, ranging from 25% to100% [15].
A perceived increased risk of complications was cited as the reason for not using amiodarone therapy by 43%. In the PAPABEAR trial, complications were low with sustained postoperative ventricular tachyarrhythmias less frequent in the amiodarone-treated patients (p = 0.04) [20]. However, dosage reductions of the blinded therapy were more common among the amiodarone patients (11.4% versus 5.3%, p = 0.008), often because of bradycardia requiring temporary pacing (5.7% versus 2%, p = 0.02).
An additional 29% of surgeons thought amiodarone therapy was excessively complicated or time consuming. Numerous regimens have been published for the use of prophylactic amiodarone. Many study protocols, including the Prophylactic Oral Amiodarone for the Prevention of Arrhythmias That Begin Early After Revascularization, Valve Replacement, or Repair (PAPABEAR) trial, required initiating the drug preoperatively. That might make its administration cumbersome decreasing the general relevance and applicability of the therapy.
In an attempt to gauge surgeon's familiarity with the current evidence, they were asked to estimate the magnitude of absolute risk reduction for postoperative AF with amiodarone. The largest randomized controlled trial reported atrial tachyarrhythmias in 16.1% of amiodarone-treated patients compared with 29.5% of placebo patients (hazard ratio [HR] 0.52; 95% CI: 0.34 to 0.69), an absolute risk reduction of 13.4% [20]. Only 24% of survey respondents reported the absolute risk reduction to be approximately 10%. An additional 26% believed it was approximately 20%.
Corticosteroids were the least frequently used modality for AF prophylaxis. Only one Canadian cardiac surgeon reported routine use of steroids for this indication. Canadian surgeons were largely unconvinced of the evidence for this practice. Using a combination therapy of methylprednisolone and dexamethasone for 24 hours, Prasongsukarn and colleagues [22] reported an occurrence of postoperative AF of 21% in the steroid group compared with 51% in the placebo group (p = 0.003). In another trial carried out by our group, we demonstrated that 1 g methylprednisolone administered before the institution of cardiopulmonary bypass significantly inhibited the incidence of atrial fibrillation (p = 0.02) [23]. In the trial by Halonen and coworkers [24] patients randomly assigned to hydrocortisone were significantly less likely to have AF than were patients randomly assigned to placebo (HR 0.54; 95% CI: 0.36 to 0.82). Finally, a meta-analysis concluded that corticosteroid treatment was associated with a lower incidence of postoperative AF (relative risk 0.67; 95% CI: 0.54 to 0.84; p = 0.001) [24].
The other major barriers to routine steroid use determined by this survey were hyperglycemia (30%) and a perceived increased risk of wound infection (39%). Hyperglycemia is observed more frequently in patients receiving steroids, which may contribute to the perceived increased risk of infection [26]. Their use, however, has not been conclusively demonstrated to be associated with postoperative wound infection [27]. In the trial by Prasongsukarn and colleagues [22], although there was a statistically significant difference in minor complications between the steroid group and the placebo group, there was no difference in major and overall complications. We showed that intraoperative insulin requirements were significantly increased in the groups receiving intravenous methylprednisolone (p = 0.044) [23]. Despite this, the incidence of postoperative infection was not different between the groups.
In an attempt to gauge surgeons' familiarity with the current evidence, they were asked to estimate the magnitude of absolute risk reduction for postoperative AF with steroids. The largest and most recent randomized controlled trial demonstrated that atrial tachyarrhythmias occurred in 30% of steroid-treated patients compared with 48% of placebo patients (HR 0.54; 95% CI: 0.36 to 0.82), an absolute risk reduction of 18% [24]. The Prasongsukarn trial [22] demonstrated an absolute risk reduction of 30%. Only 10% of survey respondents reported the absolute risk reduction to be approximately 20%, and 3% believed it to be 30%.
Nearly 60% of surgeons expressed a willingness to try steroid therapy if the absolute risk reduction was 30% or more. Assuming steroid therapy was proven safe and efficacious, 93% of surgeons said a single perioperative dose would make steroid prophylaxis more appealing for routine use. Our survey results suggest that a trial testing a clinically relevant and simple strategy of single-dose corticosteroid, utilizing a study design with adequate power and AF as a primary outcome, would be useful to Canadian cardiac surgeons. The trial would need to include practical proactive glucose management and appropriate cotherapy with β-blockade with appropriate power to robustly address secondary safety endpoints and, in particular, wound complications.
Recommendations
There is a wealth of safety and efficacy data for the routine use of β-blockade therapy for postoperative cardiac surgical patients. The results of more than 30 randomized trials and three meta-analyses have been summarized above. The majority of more recent trial protocols involve the use of a cardioselective β-blockade. Dosages vary widely between trials based on combinations of body size and left ventricular function, and many titrated to achieve a target resting heart rate. We currently prescribe metoprolol to nearly all postoperative patients, with dose titration based on body size and left ventricular function. Metoprolol is continued upon discharge until the first postoperative visit 4 to 6 weeks later, when the dose is reevaluated.
Amiodarone is effective for the prophylaxis of postoperative AF based on the results of nearly 20 randomized trials and three meta-analyses. This approach is employed rarely and selectively in our patients. Particular attention is paid to bradycardia, which can be significant with this agent. Although protocols and dosages varied widely in the literature, if feasible, we employ the PAPABEAR trial protocol as described by Mitchell and associates [20]. That involves 6 days preoperatively and 6 days postoperatively of oral amiodarone therapy at a dose of 10 mg/kg. Concomitant cardioselective β-blockade is routinely administered.
The use of steroids for prophylaxis of AF is promising. Medications used in the trials discussed above included methylprednisolone, dexamethasone, and hydrocortisone. The protocol of Prasongsukarn and coworkers [22] utilized 1 g methylprednisolone (equivalent dose 5,000 mg hydrocortisone) at operation followed by four doses of 4 mg dexamethasone (equivalent dose 150 mg hydrocortisone) at intervals of 6 hours. Halonen and associates [24] administered 100 mg hydrocortisone every 6 hours for the first 3 postoperative days. In our trial of corticosteroids, 1 g methylprednisolone (equivalent dose 5,000 mg hydrocortisone) was given at induction of anesthesia [23]. While we believe there is a great deal of potential benefit, at present steroids are not routinely used for AF prophylaxis in our institution.
Study Limitations
These data represent the results of survey information and are subject to a number of limitations. Responses were based on the recollection of the respondent and were not verified for accuracy. While every effort was made to maximize the response rate, 28% of Canadian cardiac surgeons did not participate in the survey. It could be that the practice pattern of these surgeons is significantly different from the majority of respondents. The results of this survey apply only to the practice of coronary and aortic valve surgery, and the conclusions cannot necessarily be extrapolated to other types of cardiac surgery.
In summary, despite level 1 evidence from meta-analyses and randomized controlled trials, the use of β-blockers, amiodarone, and corticosteroids for prophylaxis of postoperative atrial fibrillation among Canadian cardiac surgeons remains less than expected. Although some of our findings could be attributed to a lack of awareness of published literature and guidelines, many of the explanations for inconsistent use of prophylactic medications are rooted in realistic concerns about the safety of the medications. An attempt is needed to refine current therapies to address the concerns highlighted by this study. Corticosteroids in particular are not utilized by Canadian surgeons despite excellent evidence or their efficacy and safety. A well-designed, randomized controlled trial is required.
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
S. Bramer, A. H. M. van Straten, M. A. Soliman Hamad, K. C. van den Broek, J. G. Maessen, and E. Berreklouw New-Onset Postoperative Atrial Fibrillation Predicts Late Mortality After Mitral Valve Surgery Ann. Thorac. Surg., December 1, 2011; 92(6): 2091 - 2096. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Halonen, P. Loponen, O. Jarvinen, J. Karjalainen, I. Parviainen, P. Halonen, J. Magga, A. Turpeinen, M. Hippelainen, J. Hartikainen, et al. Metoprolol Versus Amiodarone in the Prevention of Atrial Fibrillation After Cardiac Surgery: A Randomized Trial Ann Intern Med, December 7, 2010; 153(11): 703 - 709. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Joyce Invited Commentary Ann. Thorac. Surg., July 1, 2009; 88(1): 110 - 111. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |