| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;77:1514-1524
© 2004 The Society of Thoracic Surgeons
a Department of Thoracic and Cardiovascular Surgery, Cleveland, Ohio USA
b Department of Biostatistics and Epidemiology, Cleveland Clinic Foundation, Cleveland, Ohio, USA
Accepted for publication September 5, 2003.
* Address reprint requests to Dr McCarthy, Department of Thoracic and Cardiovascular Surgery, 9500 Euclid Ave, F25, Cleveland, OH, USA 44195
e-mail: mccartp{at}ccf.org
Presented at the Thirty-ninth Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31–Feb 2, 2003.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
METHODS: From 1972 to 2000, 46,984 patients underwent primary isolated coronary artery bypass grafting; 451 (0.96% prevalence) had electrocardiogram-documented preoperative atrial fibrillation (n = 411) or flutter (n = 40). Characteristics of patients with and without atrial fibrillation were contrasted by multivariable logistic regression to form a propensity score. With this, comparable groups with and without atrial fibrillation were formed by pairwise propensity-matching to assess survival.
RESULTS: Patients with preoperative atrial fibrillation were older (67 ± 9.0 versus 59 ± 9.8 years, p < 0.0001), had more left ventricular dysfunction (66% versus 52%, p < 0.0001) and hypertension (73% versus 59%, p < 0.0001), but less severe angina (39% moderate or severe versus 49%, p < 0.0001). Many of these factors are themselves predictors of increased time-related mortality. In propensity-matched patients, survival at 30 days and at 5 and 10 years for patients with versus without atrial fibrillation was 97% versus 99%, 68% versus 85%, and 42% versus 66%, respectively, a survival difference at 10 years of 24%. Median survival in patients with atrial fibrillation was 8.7 years versus 14 years for those without it.
CONCLUSIONS: Atrial fibrillation in patients undergoing coronary artery bypass grafting is a marker for high-risk patients; in addition, atrial fibrillation itself substantially reduces long-term survival. Thus, if patients in atrial fibrillation require surgical revascularization, it is appropriate to consider performing a concomitant surgical ablation procedure.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
Because it is impossible to randomize patients to having preoperative atrial fibrillation or not, direct causal inference can never be made. However, special statistical methods have been devised that are claimed to allow causal, or at least close to causal, inferences to be made from nonrandomized comparisons [2, 3]. Therefore, the purposes of this study were to (1) identify factors associated with increased prevalence of preoperative atrial fibrillation in a large group of patients undergoing CABG, and (2) use this information to identify two similar groups of patients, one with and the other without atrial fibrillation, and compare their early and late survival.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
Demographics, clinical signs and symptoms, comorbidities, preoperative electrocardiographic diagnoses (including atrial fibrillation and flutter), previous myocardial infarction and qualitative angiographic assessment of left ventricular function, details of preoperative coronary angiography, operative details, and hospital outcomes have been recorded prospectively and concurrently with patient care in a Cardiovascular Information Registry. Data from this ongoing registry have been approved for use in research by the Institutional Review Board.
Follow-up
The first 1,000 patients undergoing primary isolated CABG have been actively followed every 5 years to 25 years. In addition, for this study all patients who were not part of this "first 1,000" cohort were passively followed using the Social Security Death Index, which we have found among these patients to be both highly specific and sensitive [4]. Because not all patients in the Cardiovascular Information Registry have recorded social security numbers, either because they were foreign or did not have this information available in hospital records, time-related survival was obtainable in 34,025 patients. Average follow-up was 12.6 ± 7.3 years, and a total of 428,476 patient-years of data were available for analysis.
Data analysis
Descriptive statistics are summarized as mean and standard deviation for continuous variables and as frequencies and percentages for categorical variables.
Strategy
Because factors associated with increased prevalence of atrial fibrillation are reported to increase mortality early and late after CABG, there is confounding between atrial fibrillation and these other factors when assessing risk factors for outcome. To address this, we used propensity score methods to balance comparison groups across potential confounding factors [2, 3]. Specifically, using the propensity score, we pairwise matched patients [5] as closely as possible with respect to all known preoperative and operative variables and then compared survival.
Factors associated with atrial fibrillation
Two multivariable logistic regression models were developed for factors associated with preoperative atrial fibrillation, (1) a parsimonious explanatory model and, subsequently, (2) a propensity model. Details of creating these models are provided in the Appendix.
Comparison of mortality
Nonparametric estimates of survival, both of unmatched and propensity-matched patients stratified according to presence or absence of preoperative atrial fibrillation, were obtained by the Kaplan–Meier method [6]. These were compared using the log-rank test. A parametric method was used to resolve the number of phases of instantaneous risk of death (hazard function) and to estimate its shaping parameters, both overall and according to presence or absence of atrial fibrillation [7]. (For additional details, see http://www.clevelandclinic.org/heartcenter/hazard.)
Presentation
Confidence limits of proportions and survival estimates are equivalent to one standard error (68% confidence limits). However, confidence limits for survival differences, hazard ratios, and difference in lifetime (area between survival curves) are 90% because these have been shown for differences and ratios to have equivalent information content as nonoverlapping 68% confidence limits of independent proportions and hazards [8]. Tables of risk factors are presented with their regression coefficients rather than odds ratios because data transformations make the latter difficult to interpret.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
|
|
|
|
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
In some ways our finding of higher mortality in patients with atrial fibrillation was expected. Atrial fibrillation occurs in older patients and in those with more advanced cardiac disease, and therefore would be expected to increase risk for perioperative complications and to decrease late survival. A novel finding of this analysis, however, similar to population-based studies, is that atrial fibrillation itself decreases postoperative survival. The Framingham study [9] examined the influence of atrial fibrillation on survival and also showed both statistically and clinically significant reduction of survival in patients with atrial fibrillation. Mortality at 30 days and at 1 and 10 years for patients with versus without atrial fibrillation was 15% versus 0.3%, 18% versus 5.9%, and 61.5% versus 30%, respectively. Our survival curves for propensity-matched patients resemble those of the Framingham study.
Other nonmatched surgical studies have also indicated increased operative mortality or decreased late survival in patients with preoperative atrial fibrillation [1, 10, 11], although this has not been identified in all studies [12]. Atrial fibrillation creates impaired hemodynamics from reduced ventricular filling [13], renders patients susceptible to stroke, and when treated with anticoagulants, increases the risk of anticoagulant-related bleeding and complications [14–16]. Patients with atrial fibrillation, even after adjusting for risk factors for stroke, have an increased stroke risk of 2.6 to 4.5 times [14, 17]. Anticoagulation reduces the stroke risk, but carries a 1% to 5% per year risk of major bleeding [18].
The cause of late increased mortality in patients with atrial fibrillation is not clear. Certainly, the increased risk of stroke and bleeding accounts for some of the difference, but also several studies indicate that chronic tachycardia also correlates with late mortality [19–24].
Clinical implications
Can the surgeon intervene to reduce the risk of mortality from atrial fibrillation by returning the patient to sinus rhythm at the time of CABG? Results from the Maze procedure indicate a low perioperative risk, excellent late freedom from atrial fibrillation, improved quality of life, and remarkably low risk of late stroke [25–32]. However, the Maze procedure is not widely applied because the operation can be demanding, it increases cardiopulmonary bypass and aortic clamp times, and its technical aspects are not widely disseminated in the cardiac surgery community. We and others have recently been using pulmonary vein isolation with left atrial appendage closure in patients with preoperative atrial fibrillation who are undergoing cardiac surgery [33–36]. Our current philosophy is that any patient undergoing cardiac surgery who has a history of atrial fibrillation should undergo a procedure for atrial fibrillation, usually consisting of pulmonary vein isolation and closure of the left atrial appendage. Goals of this combined procedure are to (1) increase the chances of returning to sinus rhythm and (2) reduce the risk of stroke, because most thrombus originates from the left atrial appendage [37, 38]. We expect that more-limited pulmonary vein isolation lesions will not be as effective as the complete Maze procedure, but early results with both percutaneous and surgical pulmonary vein isolation [29, 39, 40] indicate an effectiveness of approximately 60% to 85%.
As new technologies emerge to create lesions quickly and easily, we expect that these interventions will become more widespread [29, 41, 42]. The technology ideally will allow for fast, simple operations that can be widely applied without adding important risk or cost. The goal of these surgical interventions concomitant with CABG would be to improve survival, hopefully so that long-term survival of patients with preoperative atrial fibrillation is similar to that of patients without preoperative atrial fibrillation. Demonstrating improved survival in patients treated this way will require either a randomized trial or future propensity-matched studies, using statistical techniques such as those used in this study.
Limitations
Our study likely underestimated the prevalence of atrial fibrillation. We elected to use only Cleveland Clinic Foundation immediate preoperative electrocardiography-documented atrial fibrillation to define the prevalence. Therefore, patients with paroxysmal atrial fibrillation, or atrial fibrillation that had been cardioverted before this electrocardiogram was obtained, would be counted in the non–atrial fibrillation group. In this registry study, we did not have history regarding duration of atrial fibrillation, medications used to treat atrial fibrillation before or after surgery, or the ventricular rate associated with preoperative atrial fibrillation. Information on thromboembolic events after hospital discharge was not known. Furthermore, mortality was not cause-specific (and cannot be with any degree of reliability [4]), and, therefore, we do not know whether atrial fibrillation directly contributed to deaths.
Conclusion
In conclusion, preoperative atrial fibrillation is a marker for a high-risk group of patients undergoing primary CABG surgery. Furthermore, the prevalence of atrial fibrillation is increasing with time, in part secondary to an increase in age of patients undergoing surgery. Risk factors predictive of preoperative atrial fibrillation included older age, impaired left ventricular function, hypertension, diabetes, peripheral vascular disease, and presence of left main disease, all of which have been identified as independent risk factors for death early and late after CABG. However, as demonstrated by propensity-matching, atrial fibrillation was also an independent risk factor for early and late death after CABG. Therefore, a safe, simple, effective means of treating atrial fibrillation, and perhaps occluding the left atrial appendage, should be considered and investigated as a way to improve survival after CABG in patients with atrial fibrillation.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
DR QUADER: If we are not planning to do any mitral valve procedure and are not opening the left atrium, we would limit ourselves to doing the pulmonary vein isolation without opening the left atrial appendage.
DR SMITH: So you would perform the isolation epicardially in some fashion? Or did I misunderstand you?
DR QUADER: Yes, we do it epicardially.
DR RALPH DAMIANO (St. Louis, MO): First of all, congratulations on looking at I think a very important question in a beautifully presented study. I have a couple of questions for you.
First of all, in 1994 the Mayo Clinic presented a large series of mitral valve patients in which they looked at the exact same issue, and the group at the Mayo Clinic found absolutely no difference in 5-year survival from patients with atrial fibrillation and without atrial fibrillation in mitral valve disease. I wonder if you could speculate why in that population, or at least in their study, it did not seem to make a difference, but in your study it was quite significantly different?
The other thing is I think you need to be a little cautious with this propensity analysis in that when you have such absolutely widely divergent groups that are very different in their preoperative characteristics, if you happen to miss one of the variables that were associated with mortality, it may have still skewed your conclusions. Obviously this is a weaker conclusion than we would have if we either had some type of a prospective, randomized trial, or you could show that if you treated the atrial fibrillation, you could improve survival.
DR QUADER: Thank you for the comments. Let me answer the first question about the study design. We have identified, as well as the literature has supported, that patients who have atrial fibrillation are a high-risk group. It is impossible to randomize the patients who have atrial fibrillation and who do not have atrial fibrillation and report the study in a longitudinal way. So the closest that we could do with this kind of clinical situation is use the best available statistical tools we have, and we used the propensity score.
The second question about the atrial fibrillation in patients who have mitral valve disease. We know that patients with mitral valve disease have higher mortality compared with patients who are coming in for a primary coronary revascularization, which are the index patients for our study group. So, maybe the difference in the two groups of patients could account for the different outcomes in two studies.
DR DAMIANO: Then also have you looked at the cause of death in those patients, and if you have, which I think would be very important, was the cause of death, if it was mainly cardiac disease, you would think perhaps not related to atrial fibrillation, but was excess mortality because of stroke, or why are you saying to recommend left atrial appendage ligation? Was that whole difference in mortality related to cerebrovascular events?
DR QUADER: That is a very important question, and I am sure I thought somebody would bring that one up. It is very hard to define cause of death, especially if you are looking at 47,000 patients' follow-up. So we definitely do not have cause-specific mortality in this study population. However, the available data that we have for the hospital mortality as well as the stroke incidence of the patients while they were in the hospital show that even after matching for the two groups, the patients with atrial fibrillation had a stroke incidence of about 3% compared with 1.5% in the patients without atrial fibrillation. I am not saying that stroke is the reason that these patients died, but certainly stroke did have a significant comorbidity in these patients.
DR GERALD LAWRIE (Houston, TX): We have been very interested in atrial fibrillation for a number of years. I did want to ask you about your choice of surgical procedure for this population. Unlike the valve patients we see, these people tend to have chronic, well-established atrial fibrillation, and the pulmonary vein isolation is not a very curative procedure in that population as opposed to perhaps the mitral valve population in which we see a lot of people who have been having paroxysmal atrial fibrillation and we can expect a reasonably high cure rate from pulmonary vein isolation. If we are operating to cure atrial fibrillation in the chronic, well-established atrial fibrillation population, I am just wondering why you are not applying the Maze procedure, which is the procedure that does have close to 100% success rate?
I think the availability of the radiofrequency technologies that make it very easy to perform pulmonary vein isolation should not necessarily lead us away from the cure for these people, because I think there is going to be an epidemic of unsuccessful attempts to ablate atrial fibrillation during the next 12 to 18 months.
Thank you.
DR QUADER: Thanks for the comments and the questions. I thank and respect Doctor Jim Cox for giving us this valuable information, that the majority of patients who are in chronic atrial fibrillation, the focus of fibrillation is in and around the pulmonary veins. With the evolving technology, we have a wide range of energy sources to use to create the Maze set of lesions.
When we think of performing an atrial arrhythmia surgery on patients, we take into consideration a few things. First and foremost, the age of the patients and how long they have been in atrial fibrillation. Whether or not they are symptomatic, have they had any significant history of stroke in the past? Also, most importantly, what kind of cardiac surgery procedures are we performing?
I will give an example. If there is a patient who is 40 years old, coming in for mitral valve surgery who also happened to have atrial fibrillation of long-standing duration, we would certainly embark on performing the gold standard Cox-Maze III procedure. However, if there is a 70-year-old patient who is coming in for two valves plus coronary artery bypass grafting and who also happened to have atrial fibrillation, we would like to limit the amount of cross-clamping time as well as the pump time, and we will try to start with the pulmonary vein isolation. We have learned from the limited data available from the pulmonary vein isolation that the success rate in restoring sinus rhythm is close to 80%, not as good as the standard Cox-Maze procedure but better than no procedure to treat atrial fibrillation.
DR JAMES L. COX (Naples, FL): I would like to comment on what you just said. This is a beautiful paper and I think it is an important one, but it is very easy to misinterpret your recommendations. The statement that Doctor Lawrie just made is absolutely correct, and it is the key to what your recommendation should be. Pulmonary vein isolation is not a good operation for continuous atrial fibrillation. Furthermore, we have never had any evidence that the length of time that a patient has had atrial fibrillation has any relationship whatsoever to the ability to cure it.
I would also like to clarify one other matter. Haissaguerre, not I, showed that most intermittent atrial fibrillation originates from the pulmonary veins. Thus, if a patient has intermittent atrial fibrillation, the proper treatment for that patient is to encircle the pulmonary veins, which yields about a 90% cure rate for intermittent atrial fibrillation. However, if the patient is in continuous atrial fibrillation, pulmonary vein isolation should not be expected to cure that patient. You may inadvertently cure 10% of these patients with continuous atrial fibrillation by simply encircling the pulmonary veins, but that is not the proper operation and should not be recommended for such patients.
Thank you.
DR JOHN A. ODELL (Jacksonville, FL): Working with Joe Blackshear, our cardiologist, we reviewed the Coronary Artery Surgery Study data, which is public domain, and we found similar findings that were expressed today. A difference in survival exists with a higher risk of stroke in patients who had preoperative atrial fibrillation.
The other comment I would make is to support ligation of the left atrial appendage in patients with atrial fibrillation. We did an analysis of all published papers in which the site of thrombus was noted at either autopsy or echocardiography, and in 91% of patients with nonrheumatic atrial fibrillation, the thrombus is found within the atrial appendage.
Now, whether ligation or obliteration of the atrial appendage will influence stroke down the road is difficult to prove, but the implications of thrombus within the atrial appendage and possible embolic stroke are strong, and I strongly urge obliteration of the appendage. This can be done simply at the time of operation.
|
Appendix |
|---|
Variable screening
Initial screening of variables possibly related to atrial fibrillation used 
2 tests for contingency tables and student's t tests for continuous variables. Potential predisposing factors were organized for analysis as in Appendix Table 1. Continuous and ordinal variables were assessed univariably by decile risk analysis to suggest transformations of scale to incorporate into the analyses to assure that the relation of these variables to outcome was well calibrated with respect to logistic model assumptions. Because the study spanned nearly 30 years, not all variables were available throughout; therefore, indicator variables were formed to identify blocks of missing values before values being collected. The mean from the era in which values were collected was imputed for the variable during the missing era, and the pair (indicator and variable) was incorporated into the model. Sporadic missing values were treated by noninformative imputation of mean values.
Variable selection
Variable selection used the algorithmic method of bootstrap bagging [43, 44]. For this, 200 datasets obtained by sampling with replacement from the original data set were used, each the same size as the original. For each, automated stepwise variable selection was used with a p = 0.05 criterion for retaining variables in the models. The models were then summarized, and variables retained in 50% or more of the models were identified as risk factors of a parsimonious model.
Propensity model
To the parsimonious model described above were added at least one representative of each category of variable listed in Appendix Table 1, no matter what the significance level. In all, 30 variables were incorporated into the propensity model.
Propensity-matched pairs
The propensity model was solved to generate a unique propensity score for each patient, representing the patient-specific probability of having atrial fibrillation, whether or not the patient experienced it. The propensity score was used as the sole criterion for identifying propensity-matched pairs of patients for comparison by greedy matching [5].
|
Appendix Table 1 |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
|
LAD = left anterior descending coronary artery;
LCx = left circumflex coronary artery;
LMT = left main trunk;
RCA = right coronary artery.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentDiscussionAppendix Table 1References |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Banach, G. Mariscalco, M. Ugurlucan, D. P. Mikhailidis, M. Barylski, and J. Rysz The significance of preoperative atrial fibrillation in patients undergoing cardiac surgery: preoperative atrial fibrillation--still underestimated opponent Europace, November 1, 2008; 10(11): 1266 - 1270. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Ngaage, A. R.J. Cale, M. E. Cowen, S. Griffin, and L. Guvendik Early and Late Survival After Surgical Revascularization for Ischemic Ventricular Fibrillation/Tachycardia Ann. Thorac. Surg., April 1, 2008; 85(4): 1278 - 1281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Groh, O. A. Binns, H. G. Burton III, S. W. Ely, and A. M. Johnson Ultrasonic Cardiac Ablation for Atrial Fibrillation During Concomitant Cardiac Surgery: Long-Term Clinical Outcomes Ann. Thorac. Surg., December 1, 2007; 84(6): 1978 - 1983. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. C. Austin Propensity-score matching in the cardiovascular surgery literature from 2004 to 2006: a systematic review and suggestions for improvement. J. Thorac. Cardiovasc. Surg., November 1, 2007; 134(5): 1128 - 1135. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Ngaage, H. V. Schaff, C. J. Mullany, S. Barnes, J. A. Dearani, R. C. Daly, T. A. Orszulak, and T. M. Sundt III Influence of Preoperative Atrial Fibrillation on Late Results of Mitral Repair: Is Concomitant Ablation Justified? Ann. Thorac. Surg., August 1, 2007; 84(2): 434 - 443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Balasubramanian, T. Theologou, and I. Birdi Microwave surgical ablation for atrial fibrillation during off-pump coronary artery surgery using total arterial-Y-grafts: an early experience Interactive CardioVascular and Thoracic Surgery, August 1, 2007; 6(4): 447 - 450. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Calkins, J. Brugada, D. L. Packer, R. Cappato, S.-A. Chen, H. J.G. Crijns, R. J. Damiano Jr, D. W. Davies, D. E. Haines, M. Haissaguerre, et al. HRS/EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation: Recommendations for Personnel, Policy, Procedures and Follow-Up: A report of the Heart Rhythm Society (HRS) Task Force on Catheter and Surgical Ablation of Atrial Fibrillation Developed in partnership with the European Heart Rhythm Association (EHRA) and the European Cardiac Arrhythmia Society (ECAS); in collaboration with the American College of Cardiology (ACC), American Heart Association (AHA), and the Society of Thoracic Surgeons (STS). Endorsed and Approved by the governing bodies of the American College of Cardiology, the American Heart Association, the European Cardiac Arrhythmia Society, the European Heart Rhythm Association, the Society of Thoracic Surgeons, and the Heart Rhythm Society. Europace, June 1, 2007; 9(6): 335 - 379. [Full Text] [PDF]
|
|
|
|
|
|
D. L. Ngaage, H. V. Schaff, C. J. Mullany, T. M. Sundt III, J. A. Dearani, S. Barnes, R. C. Daly, and T. A. Orszulak Does preoperative atrial fibrillation influence early and late outcomes of coronary artery bypass grafting? J. Thorac. Cardiovasc. Surg., January 1, 2007; 133(1): 182 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. C. Burgess, M. J. Kilborn, and A. C. Keech Interventions for prevention of post-operative atrial fibrillation and its complications after cardiac surgery: a meta-analysis Eur. Heart J., December 1, 2006; 27(23): 2846 - 2857. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W.W. Wong and K.-H. Mak Impact of Maze and Concomitant Mitral Valve Surgery on Clinical Outcomes Ann. Thorac. Surg., November 1, 2006; 82(5): 1938 - 1947. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Ngaage, H. V. Schaff, S. A. Barnes, T. M. Sundt III, C. J. Mullany, J. A. Dearani, R. C. Daly, and T. A. Orszulak Prognostic implications of preoperative atrial fibrillation in patients undergoing aortic valve replacement: is there an argument for concomitant arrhythmia surgery? Ann. Thorac. Surg., October 1, 2006; 82(4): 1392 - 1399. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Rogers, G. D. Angelini, L. A. Culliford, R. Capoun, and R. Ascione Coronary surgery in patients with preexisting chronic atrial fibrillation: early and midterm clinical outcome. Ann. Thorac. Surg., May 1, 2006; 81(5): 1676 - 1682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Akpinar, I. Sanisoglu, M. Guden, E. Sagbas, B. Caynak, and Z. Bayramoglu Combined Off-Pump Coronary Artery Bypass Grafting Surgery and Ablative Therapy for Atrial Fibrillation: Early and Mid-Term Results Ann. Thorac. Surg., April 1, 2006; 81(4): 1332 - 1337. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Knaut, S. M. Tugtekin, S. G. Spitzer, F. Jung, and K. Matschke Intraoperative endocardial microwave ablation for treatment of permanent atrial fibrillation during coronary artery bypass surgery: 1-year follow-up. Europace, January 1, 2006; 8(1): 16 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Jones The Year in Cardiovascular Surgery J. Am. Coll. Cardiol., May 3, 2005; 45(9): 1517 - 1528. |
