| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2005;129:997-1005
© 2005 The American Association for Thoracic Surgery
General Thoracic Surgery |
a Royal College of Surgeons of England, London, United Kingdom
b Health Services Research Unit, Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, London, United Kingdom
c Cardiothoracic Unit, Guy’s Hospital, London, United Kingdom
d Department of Medicine and Division of Health Policy, Yale University School of Medicine, New Haven, Conn.
Received for publication June 17, 2004; revisions received July 27, 2004; accepted for publication July 28, 2004. * Address for reprints: Artyom Sedrakyan, MD, PhD, Clinical Effectiveness Unit, Royal College of Surgeons, 35-43 Lincoln’s Inn Fields, London WC2A 3PE, United Kingdom (E-mail: asedrakyan{at}rcseng.ac.uk).
 |
Abstract |
|---|
 Top Abstract MethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
METHODS: MEDLINE, EMBASE, Cochrane Database of clinical trials (1980–2003), and reference lists of relevant articles were searched for randomized controlled trials with placebo control, general thoracic patients, and noncombined and prophylactic use of the medications. Search, data abstraction, and analyses were performed and confirmed by at least 2 authors. A fixed-effects model was used to perform meta-analyses.
RESULTS: There were 11 unique trials (total n = 1294) that met the inclusion criteria. Calcium-channel blockers and ß-blockers reduced the risk of atrial tachyarrhythmia in 4 and 2 trials, respectively (relative risk of 0.50 and 95% confidence interval of 0.34–0.73; relative risk of 0.40 and 95% confidence interval of 0.17–0.95, respectively). However, ß-blockers tended to increase the risk of pulmonary edema (relative risk, 2.15; 95% confidence interval, 0.74–6.23). Magnesium tested in one unblinded trial also reduced the risk of atrial tachyarrhythmia (relative risk, 0.4; 95% confidence interval, 0.21–0.78). On the other hand, digitalis preparations were found to be harmful because they increased the risk of atrial tachyarrhythmia in 3 trials (relative risk, 1.51; 95% confidence interval, 1.00–2.28). Finally, 2 other medications, flecainide and amiodarone, were each tested in a single small trial, and their effects were associated with great uncertainty.
CONCLUSIONS: Calcium-channel blockers and ß-blockers are effective in reducing postoperative atrial tachyarrhythmia. The use of these medications should be individualized, and possible adverse events of ß-blockers should be taken into account. Randomized clinical trials do not support the use of digitalis in general thoracic surgery. The value of magnesium as a supplement to a main prophylactic regimen should be explored.
Only a handful of studies were conducted to determine whether antiarrhythmic medications could reduce the incidence of AT in GTS. Although one recent review addressed the usefulness of ß-blockers in noncardiac thoracic surgery13,14 and the results of the study were suggested to be applicable to GTS patients, the study mostly included trials enrolling patients undergoing vascular surgery.
It would be inappropriate to apply the evidence found in vascular surgical studies in the setting of GTS without systematically reviewing the evidence of the effectiveness of ß-blockers, as well as other pharmacologic interventions. Most patients in GTS practice undergo resections for lung cancer and have a long history of smoking and advanced chronic obstructive pulmonary disease (COPD). Thus the safety of the ß-blockers should also be determined because pulmonary morbidity aggravated by large resections of the lung (pneumonectomies and lobectomies) might interact with operative stress and is likely to worsen pulmonary dysfunction that might potentially be associated with ß-blockers.15,16 Similarly, the safety profiles of other medications that are used for arrhythmia prophylaxis are unknown in GTS.
Therefore we sought to determine whether the use of ß-blockers and other antiarrhythmic medications before or during surgical intervention is associated with a reduced rate of postoperative AT in GTS by systematically reviewing available evidence from randomized clinical trials (RCTs). We also sought to determine whether the use of antiarrhythmic medications is associated with hypotension, bradycardia, severe pulmonary complications, myocardial ischemia, or mortality.
|
Methods |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
For each trial, reported data on arrhythmia, bradycardia, hypotension, severe pulmonary complication, myocardial ischemia, and mortality were collected in the intervention and control groups.
Search strategy for identification of studies
RCTs were identified by searching MEDLINE, EMBASE, and the Cochrane Controlled Trials Register from 1980 through 2003. Only studies in human subjects were considered, and no language restriction was applied. In addition, we searched the reference lists of RCTs and reviews to look for additional studies.
A standard RCT filter designed by the Cochrane Collaboration for identifying RCTs was used for MEDLINE and EMBASE. These search algorithms are available at the Scottish Intercollegiate Guideline Network Web site (http://www.sign.ac.uk/methodology/filters.html). The strategy (with MeSH and text terms) used to search the databases is outlined in Appendix 1.
Search of the above databases identified 262 unique abstracts. These were independently reviewed for relevance for review, which was independently performed by 2 authors (A.S. and J.V.M.). Independent review of these abstracts identified 16 unique studies after excluding multiple publications from the same trials.17–19 On detailed review, 5 of the 16 studies did not meet the predefined inclusion criteria. One of the rejected studies randomized patients undergoing heart-lung transplantation,20 2 studies were found to be nonrandomized trials,8,21 1 study enrolled cardiac surgical patients,22 and 1 study related to short-acting ß-blocker (esmolol) use during anesthesia to measure responses to laryngoscopy and did not report information on any of the outcomes of interest.23 The remaining 11 unique studies4,11,17,24–32 met the inclusion criteria.
For each trial, abstracted data included the frequency of the events in the medication and placebo groups, as well as the numbers of patients randomized to the treatment groups. Other abstracted data included mean age, sex, publication date, blinded allocation of the study drug, and number of dropouts. One of the authors (A.S.) abstracted the data, and another coauthor (J.B.) independently confirmed the counts of events and the number of patients randomized to treatment groups. All authors participated in adjudication of any discrepancies. Corresponding authors of all selected studies were contacted to learn about missing information. Reports-descriptions such as "no major complications were observed in the study" were not considered as zero events. Only explicit description of outcome events was accepted as a valid count. If the primary investigator did not reply to personal communication within a 4-week period and if the article did not contain information on the adverse event of interest, then the end point was left as a missing value.
AT was evaluated on the basis of reports of "supra-ventricular tachycardia," "atrial fibrillation," "atrial flutter," and "atrial tachyarrhythmia." Hypotension and bradycardia incidences were based on the reported incidences of these conditions. The criteria for severe pulmonary complications were reports of "pulmonary edema" and "adult respiratory distress syndrome (ARDS)." Myocardial ischemia measurement was based on the reports of "myocardial ischemia" or "myocardial infarction." Mortality mostly included in-hospital clinical diagnosis of death. The methods of ascertainment of the events differed in the trials. However, within each trial, they were applied equally to the 2 treatment groups.
|
Statistical analyses |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
2 statistic, we did not use a random-effects model. The standard error of the pooled risk ratio (RR) estimate was used to calculate 95% confidence intervals (CIs). Numbers needed to treat and their CIs were also calculated by risk difference (RD) analysis. RD statistics were particularly important in the trials with 0 events, in which RR was not estimable. In these instances RD statistics still provided estimates of uncertainty around 0. On the basis of number-needed-to-treat statistics, the numbers of events averted-induced were calculated per 100 procedures. Because each class of medication was considered qualitatively different from another, only subgroup analyses were performed for each class of medication, and no single estimate for all medications was calculated. RevMan 4.2, developed by the Cochrane Collaboration, was used in all analyses.
The methodologic quality of the included studies was evaluated by using the Jadad criteria,33 which are based on the following: (1) "randomized" study description, (2) description of correct randomization procedure, (3) double-blinding study description, (4) description of correct double blinding, and (5) dropouts and adequate description of the end points of interest.
The overall quality score was based on the number of criteria met (score range, 1–5).
|
Results |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
|
|
|
Magnesium
There was one trial enrolling 194 patients. The overall incidence was 10.7% in the medication group and 26.7% in the placebo group (RR, 0.40; 95% CI, 0.21–0.78), and 16 episodes of AT were estimated to be averted per 100 patients administered magnesium (95% CI, –27 to –5).
Digitalis
AT was assessed in all 3 trials of digitalis, which included 285 patients (Figure 1 and Table 2). In 2 of the trials the incidence of "dysrhythmia" was reported,28,29 whereas in the third trial the incidence of AT was determined. The overall incidence was 30.2% in the digitalis groups and 19.9% in the placebo groups (RR, 1.51; 95% CI, 1.00–2.28).
Flecainide
There was one trial enrolling 30 patients. The overall incidence was 0% in the medication group and 18.7% in the placebo group (RR, 0.16; 95% CI, 0.01–2.89).
Amiodarone
There was one trial enrolling 62 patients. The overall incidence was 3.1% in the medication group and 21.9% in the placebo group (RR, 0.14; 95% CI, 0.02–1.10).
Hypotension and bradycardia
Calcium-channel blockers
Four trials (total n = 618) reported the incidence of hypotension, and 3 trials (total n = 290) reported the incidence of bradycardia (Table 2). The overall incidence of hypotension was 7.4% in the medication groups and 0.6% in the placebo groups (RR, 9.29; 7 additional events per 100 patients treated). Bradycardia occurred in 6.2% of the medication groups and in 0% of the placebo groups (RR, 18.81; 6 additional events per 100 patients treated).
ß-blockers
One trial24 (n = 99) reported information on ß-blockers (Table 2). The incidence of hypotension was 49.0% in the medication group and 26.0% in the placebo group (RR, 1.88; 23 additional events per 100 patients treated). The incidence of bradycardia was 24.5% in the medication group and 4.0% in the placebo group (RR, 6.12; 20 additional events per 100 patients treated).
Magnesium
There were 0 events of hypotension and bradycardia reported in the trial (n = 194; RR not estimable with 0 additional events).
Digitalis
No information on hypotension was available in the studies. Bradycardia was assessed in 2 trials, which included 209 patients (Table 2). The overall incidence was 9.8% in the digitalis groups and 1.9% in the placebo groups (RR, 4.46; 8 additional incidences of bradycardia per 100 patients treated).
Flecainide
The overall incidence of hypotension was 57.1% in the medication group and 0% in the placebo group (per investigator report; RR, 19.27; 57 additional events of hypotension per 100 patients treated). There were no events of bradycardia reported in 30 enrolled patients.
Amiodarone
The incidence of hypotension was 3.1% in the medication group and 3.1% in the placebo group. Bradycardia occurred in 3.1% of the medication group and 0% of the placebo group (Table 2).
Severe pulmonary complications
Calcium-channel blockers
All 4 trials (total n = 618) reported the incidence of ARDS (Table 3). The overall incidence was 2.6% in the medication groups and 2.0% in the placebo groups (RR, 1.31; with a tendency of one additional event per 100 patients treated; Table 3.
|
Magnesium
There was no information on the incidence of this outcome in the study.
Digitalis
No information on the incidence of this outcome was available in the studies.
Flecainide
There were 0 events reported in the trial (n = 194).
Amiodarone
The overall incidence reported was for ARDS, and it occurred in 9.4% of the medication group and 0% of the placebo group (RR, 7.00; with a tendency toward 9 additional events per 100 patients treated).
Myocardial ischemia and mortality
There were few outcomes of interest reported in all trials. Given the rare incidence of these outcomes, the safety of the medications was more reliably estimable for calcium-channel blockers because there were 359 patient-based reports for myocardial ischemia and 618 patient-based reports for mortality. The estimates for other medications were associated with much greater uncertainty (Table 3).
|
Discussion |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
Calcium-channel blockers and ß-blockers were associated with a reduction in AT to less than half of the incidence seen in the control patients. Our estimates were robust and show 11 to 14 AT reductions per 100 patients administered these medications. These medications were reported to be associated with increased risk of hypotension and bradycardia. The severity and definitions of these events were not reported in the trials, but accepting definitions provided by the authors, 7 cases of hypotension and 6 cases of bradycardia were associated with calcium-channel blockers, and 23 cases of hypotension and 20 cases of bradycardia were associated with ß-blockers per 100 treated patients. The higher incidence of these events in the ß-blocker group might be related to fewer trials and different definitions rather than qualitative differences between these 2 classes of medications. In all trials hypotension and bradycardia were temporary events and disappeared after discontinuation of the medication. Thus these findings should be interpreted as a note of caution in the GTS population and should not influence the policy of arrhythmia prophylaxis.
In our study ß-blockers tended to be associated with pulmonary edema in one trial.24 Although personal communication with the investigators revealed no ARDS (a measure of the severe pulmonary dysfunction in calcium-channel blocker trials), the Cochrane review of ß-blocker use in the population with COPD also reported a tendency toward a higher risk of pulmonary dysfunction.15,16 Thus there is a theoretic possibility that ß-blockers might interact with operative stress in the GTS setting and worsen preexisting pulmonary dysfunction in patients with lung cancer who have long-term smoking histories and COPD.
Calcium-channel blockers, on the other hand, are known to reduce pulmonary vascular resistance and right ventricular pressure in addition to their antiarrhythmic effects.10,27 Thus they are less likely to cause pulmonary complications. Moreover, their beneficial effect on pulmonary vascular resistance might be valuable in lobectomies and pneumonectomies, in which there is a concern of increased pulmonary vascular resistance. Thus one might argue that for larger resections, advanced COPD, and intolerance of ß-blockers, calcium-channel blockers should be the first choice. The costs of the calcium-channel blockers should be considered, and generic brands should be preferred. The evidence for calcium-channel blockers is limited to verapamil and diltiazem intravenous formulations, and there appears to be no difference between these medications.
The effectiveness of ß-blockers and calcium-channel blockers in cardiac surgery was recently summarized. The evidence of halved risk of atrial fibrillation associated with ß-blockers found in the review by Crystal and associates34 in cardiac surgery and Auerbach and Goldman13 in noncardiac surgery (non-GTS) is being rapidly popularized. However, a similar review by Wijeysundera and colleagues35 that also found comparable benefits associated with calcium-channel blockers in cardiac surgery had not received as much attention. Moreover, the latter study reported a more than 40% reduction of postoperative myocardial infarction and a more than 45% reduction of myocardial ischemia associated with calcium-channel blockers, evidence that was not found for ß-blockers.
Because ß-blockers were found to be very useful in reducing mortality after myocardial infarction, there seems to be a trend toward recommending this group of medications in most settings without extensive research on other alternatives or specifics of patient populations. We believe that GTS is the area in which calcium-channel blockers should be considered of equal value if not a first choice.
Magnesium was tested in one large unblinded trial of 200 patients. It was shown to be associated with a 2.5 times reduction in the risk of AT and 16 event reductions in 100 patients administered magnesium. One limitation of this trial was digitalis use in some control patients. Furthermore, RCTs conducted on patients undergoing heart surgery report controversial findings. Although Wollert and coworkers22 also reported reduced risk of AT in cardiac surgery, a number of other investigators did not find any difference between magnesium and placebo.36,37 On the other hand, Wilkes and associates38 found a reduced risk of AT when magnesium was supplemented after measuring levels of ionized plasma magnesium. The value of magnesium as a supplement to the main antiarrhythmia regimen warrants further investigation.
Three RCTs of digitalis reported dysrhythmia incidence,28,29 and we found a 51% higher risk of arrhythmia for digitalis compared with that for placebo (RR, 1.51; 95% CI, 1.00–2.28). Moreover, digitalis was associated with a more than 4 times higher risk of bradycardia compared with placebo. In 2 other RCTs comparing digoxin with diltiazem39 and flecainide,40 the former was shown to be inferior. Digitalization is still being used quite often in GTS according to one recent review.3 Thus this finding should help clinicians to discontinue the use of this antiarrhythmic medication in GTS.
Flecainide and amiodarone were each tested in only one trial, and reduction of arrhythmic events in the intervention groups was reported. Borgeat and coworkers25 have shown overall reduction of arrhythmic events (6 in control and 0 in placebo), but the study was too small (n = 30) to address our end point of AT (3 in control and 0 in placebo). Personal communication with the investigator revealed a high incidence of hypotension in the medication group. Although van Mieghem and associates31 reported a significant reduction of AT for amiodarone (RR, 0.16) in a sample of 62 patients (unblinded study), they were urged to stop the trial because of 3 cases of ARDS (2 deaths) observed in the amiodarone arm. In another study conducted in patients undergoing cardiac surgery, amiodarone was associated with a longer duration of cardiovascular instability and a longer need for intensive care.37 Thus we believe that there is inconclusive evidence to support the use of these medications.
Although we had some limitations, such as the inability to fully explore the safety of medications regarding mortality and myocardial ischemia or to perform further subgroup analyses (by type of procedure or medication), we believe that these limitations are not substantial. Furthermore, these findings have important implications in the current practice of GTS. Given the high incidence of AT, the use of proper antiarrhythmic prophylaxis can potentially prevent more than 1000 events in the United Kingdom and more than 7800 events in the Unites States (halving the baseline incidence rate of approximately 20% in 10,000 and 78,000 procedures, respectively). Moreover, antiarrhythmic medications are generally inexpensive, and substantial costs savings can also be expected because AT is associated with up to $10,000 in additional costs per incidence41 and an additional 2 to 5 days of hospital stay in the studies of coronary artery bypass grafting surgery.41,42
|
Conclusions |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
|
Appendix 1 |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
{1 exp Anti-Arrhythmia Agents/ (152477); 2 ((antiarrhythm$ or anti-arrhythm$ or arrhythm$) adj2 (drug$ or agent$ or therap$)).tw. (10205); 3 (tachycardia or antiarrhythm$ or anti-arrhythm$ or arrhythm$).tw. (63738); 4 exp Arrhythmia/ (101335); 5 (tachycardia or antiarrhythm$ or anti-arrhythm$ or arrhythm$).tw. (63738); 6 ((atrial or atrium or auricular) adj6 (flutter$ or fibrillat$)).tw. (14810); 7 or/4–6 (126560); 8 exp drug therapy/ (249078); 9 (dt or pc or tu).fs. (1644964); 10 or/8–9 (1715378); 11 7 and 10 (37925); 12 or/1–2,11 (172231); 13 exp Thoracic Surgery/ (4215); 14 thoracic surgical procedures/ or exp pulmonary surgical procedures/ or thoracoplasty/ or thoracostomy/ or thoracotomy/ or thymectomy/ or tracheostomy/ or tracheotomy/ (47756); 15 ((thoracic or lung or pulmon$ or bronch$) adj6 (surgery or surgical or resect$)).tw. (29046); 16 (lobectom$ or pneumonectom$ or segmentectom$ or esophagotom$ or thoracotom$).mp. (26886); 17 or/13–16 (81182); 18 (Randomized controlled trials/ or Randomized controlled trial.pt. or Random allocation/ or Double blind method/ or Single blind method/ or Clinical trial.pt. or exp clinical trials/ or (clinic$ adj trial$1).tw. or ((singl$ or doubl$ or treb$ or tripl$) adj (blind$3 or mask$3)).tw. or Placebos/ or Placebo$.tw. or Randomly allocated.tw. or (allocated adj2 random).tw.) not (Case report.tw. or (Letter or Historical article or Review of reported cases or Review, multicase).pt.) (487507); 19 Animal/ not (Animal/ and Human/) (2680505); 20 (12 and 17 and 18) not 19 (126); 21 from 20 keep 1–10 (10); 22 from 20 keep 1–126 (126)}.
|
EMBASE search strategy |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractMethodsStatistical analysesResultsDiscussionConclusionsAppendix 1EMBASE search strategyReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. C. Fernando, M. T. Jaklitsch, G. L. Walsh, J. E. Tisdale, C. D. Bridges, J. D. Mitchell, and J. B. Shrager The Society of Thoracic Surgeons Practice Guideline on the Prophylaxis and Management of Atrial Fibrillation Associated With General Thoracic Surgery: Executive Summary Ann. Thorac. Surg., September 1, 2011; 92(3): 1144 - 1152. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Saran, G. D. Perkins, M. A. Javed, V. Annam, L. Leong, F. Gao, and R. Stedman Does the prophylactic administration of magnesium sulphate to patients undergoing thoracotomy prevent postoperative supraventricular arrhythmias? A randomized controlled trial Br. J. Anaesth., June 1, 2011; 106(6): 785 - 791. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. N. Wilkinson, S. H. Pennefather, and R. A. McCahon Post-operative arrhythmias OSH Thoracic Anaesthesia, January 1, 2011; 1(1): med-9780199563098-div1-25 - med-9780199563098-div1-25. [Full Text]
|
|
|
|
 |
|
 S. G. Memtsoudis, P. Rosenberger, and J. M. Walz Critical Care Issues in the Patient After Major Joint Replacement J Intensive Care Med, March 1, 2007; 22(2): 92 - 104. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Bobbio, D. Caporale, E. Internullo, L. Ampollini, S. Bettati, E. Rossini, P. Carbognani, and M. Rusca Postoperative outcome of patients undergoing lung resection presenting with new-onset atrial fibrillation managed by amiodarone or diltiazem Eur J Cardiothorac Surg, January 1, 2007; 31(1): 70 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Amar, A. Goenka, H. Zhang, B. Park, and H. T. Thaler Leukocytosis and increased risk of atrial fibrillation after general thoracic surgery. Ann. Thorac. Surg., September 1, 2006; 82(3): 1057 - 1061. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Auerbach and L. Goldman Assessing and Reducing the Cardiac Risk of Noncardiac Surgery Circulation, March 14, 2006; 113(10): 1361 - 1376. [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 |
