| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Clinical Research Unit, Division of Cardiothoracic Surgery, Emory University, Atlanta, Georgia
b Department of Biostatistics, Rollins School of Public Health, Emory University, Atlanta, Georgia
Accepted for publication April 28, 2009.
* Address correspondence to Dr Puskas, Division of Cardiothoracic Surgery, Emory University, Emory Crawford Long Hospital, 550 Peachtree St, 6th Floor Medical Office Tower, Atlanta, GA 30308 (Email: john.puskas{at}emoryhealthcare.org).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| ADULT CARDIAC SURGERY:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: The Society of Thoracic Surgeons database was queried for all isolated, primary coronary artery bypass graft cases between January 1, 1997, and December 31, 2007, at a US academic center. The Society of Thoracic Surgeons Predicted Risk of Mortality (PROM) was calculated by a formula based on 30 preoperative risk factors. It was used in three ways to compare 30-day operative mortality between patients treated with OPCAB versus CPB. First, patients were divided into quartiles based on their PROM, and mortality rates were compared between OPCAB and CPB patients within each PROM quartile. Second, a logistic regression model tested for an interaction between surgery type and PROM; a significant interaction would indicate that the relative mortality risk of OPCAB versus CPB varied with different PROM levels. Finally, locally smoothed kernel regression curves were used to visually estimate a threshold PROM point at which mortality rates diverge for the surgery types.
Results: There were 14,766 consecutive patients, 7,083 OPCAB (48.0%) and 7,683 CPB (52.0%). There was no difference in operative mortality between OPCAB and CPB for patients in the lower two risk quartiles. In the higher risk quartiles there was a mortality benefit for OPCAB (odds ratio, 0.62 and 0.45 for OPCAB in the third and fourth risk quartiles). Logistic regression analysis confirmed a significant interaction between surgery type and PROM (p = 0.005) meaning that OPCAB is especially beneficial to patients with higher PROM. This benefit is most significant for patients with PROM values above 2.5% to 3%, where mortality curves sharply diverge.
Conclusions: Off-pump coronary artery bypass grafting is associated with lower operative mortality than coronary artery bypass grafting on CPB for higher risk patients. This mortality benefit increases with increasing PROM.
| Drs Puskas and Thourani disclose that they have financial relationships with Medtronic and Maquet; Dr Lattouf with Medtronic and Cardiogenesis.
|
In an effort to avoid morbidity and mortality that may be attributable to aortic manipulation, global myocardial ischemia, and systemic inflammatory response associated with cardiopulmonary bypass (CPB), some surgeons have embraced off-pump coronary artery bypass grafting (OPCAB). In 2007 20.4% of all primary coronary artery bypass grafting (CABG) cases in the National Adult Cardiac Surgery Database were performed with off-pump techniques [1]. Surgeons with different levels of experience may choose to perform OPCAB versus conventional CABG on CPB for different patient populations by highly variable criteria. Some of these criteria may be systematic, ie, guided by coronary anatomy [2] or hemodynamic data, whereas others may be idiosyncratic and inconsistent.
Various authors have suggested particular patient characteristics that may especially warrant the perceived extra effort of off-pump revascularization. Among these have been advanced age, female sex, peripheral vascular disease or aortic atherosclerosis, renal failure, and severe lung disease [3–8]. There exists no consensus on which patient subgroups may benefit most from OPCAB rather than CABG on CPB.
Moreover, numerous single-center prospective, randomized trials enrolling relatively small numbers of relatively healthy patients have been unable to demonstrate a statistically significant mortality benefit of OPCAB [9–11]. Although this is not at all surprising given the low incidence of operative mortality in this population, it has given rise to the misperception that there is no mortality benefit for OPCAB. In fact, several retrospective analyses of large databases have clearly shown reduced risk-adjusted mortality after OPCAB compared with CABG on CPB [12–15]. However, these reports have not led to an increase in adoption of OPCAB techniques nationwide. This may be in part because skeptical surgeons are unsure which patients will benefit most from these newer, more challenging off-pump techniques.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Interventions, Surgeons, and Surgical Technique
Each patient received a single surgical session performed at the discretion of any of 17 faculty surgeons. The median number of surgeries per surgeon was 578 (range, 1 to 2,841). The median number of OPCAB surgeries per surgeon was 93 (range, 0 to 2,135). The median number of CABG on CPB surgeries per surgeon was 443 (range, 0 to 1,358). Off-pump CABG was performed with commercially available cardiac positioning and coronary artery stabilizing devices. Conventional CABG on CPB was performed with standard techniques, using roller head pumps, membrane oxygenators, cardiotomy suction, arterial filters, cold antegrade and retrograde blood cardioplegia, and moderate systemic hypothermia (30° to 34°C).
Patients who were converted intraoperatively from OPCAB to CABG on CPB or from CABG on CPB to OPCAB were entered into the database according to the procedure they ultimately received. Intraoperative conversion began to be collected as part of the institutional database in 2002, when a data field for conversion was introduced into the STS National Adult Cardiac Database. This did not allow for a formal intention-to-treat analysis of this dataset; however, a separate substudy was conducted to measure the statistical impact of the conversions that were in the database (from 2002 forward).
Society of Thoracic Surgeons Predicted Risk of Mortality
The Emory institutional STS 30-day risk models were developed to provide clinicians and hospitals with a tool to evaluate risk-adjusted outcomes and to measure changes in quality improvements. The scores themselves are simply predicted probabilities (ranging from 0 to 1) calculated from a multivariable logistic regression model calibrated on STS data within a fixed time period. The 30 predictors in the model are known (Table 1) along with their associated odds ratios; however, the model intercept term is unreported, as are estimates for factors that were statistically insignificant predictors of mortality. Periodic updates of the model terms are undertaken in an effort to make the predictions commensurate with evolving technology and generally improved outcomes with time. The STS PROM score, most recently calibrated by Shroyer and colleagues [16], is known to discriminate well between survivors and nonsurvivors (c-index = 0.78) and has a high degree of agreement between predicted and observed mortality. The goal of this study is to compare the performance of two surgery types (CABG on CPB and OPCAB) against the national benchmarks provided by the STS PROM score.
|
The STS PROM was used in three ways to compare 30-day operative mortality between patients treated with OPCAB versus CPB. First, patients were divided into quartiles based on their PROM, and a logistic regression model was developed that related death as a function of OPCAB (yes or no) within each quartile. Second, for all of the data (across all quartiles) a logistic regression model was fit that related death as a function of OPCAB, PROM, and the interaction of OPCAB and PROM. A significant interaction would indicate that OPCAB is more beneficial at some levels of preoperative risk than it is at others. Finally, kernel estimation with local smoothing was used to visually estimate a threshold PROM point at which mortality rates diverge for OPCAB versus CABG on CPB. To this end, a plot was created in which the x axis was PROM and the y axis was observed mortality. Because observed mortality is dichotomous, the points on the plot at each PROM value are calculated by taking an average of the observed mortality (so-called uniform kernel smoothing) for a "neighborhood" around the PROM value. The neighborhood is the collection of points within a small range around the PROM value; thus, the point on the plot is the average mortality rate for that small region of PROM values. The width of the neighborhood varies locally according to a systematic algorithm; it is small on the low end of PROM where there was a lot of data and larger on the upper end (higher risk patients) where data were more sparse.
The data were managed and analyzed using SAS Version 9.1 (SAS Institute, Cary, NC). Unadjusted comparisons were performed with 
2 tests and two-sample Student's t tests for categorical and continuous predictors, respectively. All statistical tests were two-sided using an 
equal to 0.05 for the level of significance. No adjustments for multiple tests were made.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
To determine whether differences in outcomes among surgeons had confounded these results, surgeon identity was included as a covariate in the multivariable logistic regression model that assessed the interaction between surgery type and PROM with respect to observed mortality. When surgeon identity was included, its effect did not diminish the apparent benefit associated with OPCAB, nor the disproportionate magnitude of this benefit for higher risk patients.
This benefit is most significant for patients with PROM values above 0.03 (3%), for which locally smoothed kernel regression curves visually demonstrate a threshold PROM point at which mortality rates begin to sharply diverge for OPCAB versus CABG on CPB (Fig 1).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
These findings also explain in concrete statistical terms why prior small randomized controlled trials enrolling predominantly low-risk patients have failed to show a mortality benefit for OPCAB. No significant mortality benefit may exist for low-risk patients (although OPCAB has been clearly associated with lower myocardial enzyme release, fewer transfusions, and shorter length of stay, even among lower risk patients). The mortality benefit that has been repeatedly shown to exist for OPCAB in large retrospective analyses [12–15] is likely driven by a disproportionate benefit enjoyed by the higher-risk patients within these databases.
Further, these findings also corroborate the utility of the STS PROM model. If one accepts the notion that OPCAB imposes less physiologic stress (systemic inflammation, hemodilution, and so forth) on patients than CABG performed on CPB, then one would expect a valid model that predicted risk of mortality to behave as the STS PROM has in the present analyses. The STS PROM is provided free of charge for every patient by the STS database once preoperative risk factors and demographic data are entered.
Finally, the findings of the present study may provide guidance to surgeons who struggle to define a role for OPCAB within their practices. In particular, the finding that OPCAB and CABG on CPB mortality curves diverge sharply for patients with PROM in excess of 2.5% to 3% should guide surgeons to perform OPCAB in this patient population or alternatively to refer such patients to a surgeon experienced in OPCAB. Although we advocate a default policy of routine OPCAB for all coronary patients, surgeons who prefer a selective approach may use the STS PROM model to guide their patient selection for OPCAB.
The present single-institution, retrospective study is subject to all the limitations of such studies, despite sophisticated statistical analyses designed to control for biases in patient selection, potential differences among surgeons, and other possible confounders. It is clear that elimination of all possible sources of bias is impossible. Moreover, Emory University has a strong institutional interest and experience in OPCAB; it is possible that results obtained are not easily generalized to every institution. Further, the reported results use patient classifications according to the surgery type they ultimately received and does not take into account conversions in either direction. The Emory institutional STS database did not begin collecting information on conversions until 2002, about halfway through the study period. The STS National Cardiac Database adopted a data field for intraoperative conversions in 2004. In a separate analysis, the converted patients (a total of 137 of 7,120 [1.8%] patients were converted from planned OPCAB to CABG on CPB; among these, 30-day mortality was 4 of 137 patients [2.9%]) from 2002 forward were reclassified by the originally intended surgery type, and the study results were recalculated using the logistic model with the interaction term (see Results above). It was found that the parameter estimates of the model terms changed by less than 1% relative on average and the probability values were very similar. This is strong evidence that converted patients did not confound the comparisons performed in this study.
Off-pump CABG is associated with lower operative mortality than CABG on CPB in this single-institution retrospective study. This is driven by a disproportionate mortality benefit for higher risk patients, which increases with increasing STS PROM.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR PUSKAS: That's a good question. The STS predicted risk model does include presence of left main disease, it includes the number of diseased vessels, it includes ejection fraction, and it includes a variety of other important patient characteristics which are not included in the SYNTAX score, but it does not include a metric for diffuseness of coronary disease. And that is unlikely to change because it would require a good deal of data from each of the preoperative angiograms, which presently do not have data fields within the STS database.
DR MOHR: My question is in this direction; being an off-pump surgeon myself, there is a selection bias. If you have a patient with severe diffuse disease and maybe it's a redo patient, there is a greater likelihood that you do these patients on-pump. And of course, it does have an impact. And it is true for our institution and I would not think that this is different. Can you define that from your data analysis somehow?
DR PUSKAS: I don't think that the SYNTAX score will be easily replicated within a large database like this because details of angiography are simply not included. But there are some elements that are included that don't go into a SYNTAX score. Diabetes, of course, which is a marker for diffuse disease, is included in the STS predicted model. Number of diseased vessels is, and left main, as I mentioned.
The tendency among some surgeons to do more diffuse cases or those requiring more grafts on pump is real. That tendency exists and I think it is widespread. There are a few centers where that doesn't exist. There are a few surgeons for whom that tendency doesn't exist. But in general, I think that's a correct observation of present practice. And it could shunt patients who have increased risk into the on-pump group. But the STS predicted model includes cofactors that go along with those kinds of patients, such as diabetes, body mass index, and a description of coronary anatomy, that I think makes it still the best truly preoperative and widely available predicted risk adjustment methodology.
DR VALAVANUR SUBRAMANIAN (New York, NY): John, it's an excellent paper. I understand in the lower risk terciles it did not show any benefit of OPCAB, but again—
DR PUSKAS: No mortality benefit.
DR SUBRAMANIAN: No mortality benefit. So there is a 10-year period data, 1997 to 2007. The OPCAB maturation happened, I believe, the cutoff point is around 2003, when the exposed device or the cardiac position device came in. Can you comment on it? Because if it is in the earlier period, perhaps the potential benefit accrued from the OPCAB might have been neutralized because of the lack of maturity in OPCAB.
The reason I'm asking is because the initial analysis in the first paper from the New York Cardiac Surgery Data Bank, part of it as I can remember, and when we looked at it initially before presenting the data we did not consider the year. Once we considered the year, then the benefit was apparent for the OPCAB. So can you go back and look at the periods of the operation and then see perhaps there may be a difference?
DR PUSKAS: That would be an interesting analysis. Of course, this represents the institutional learning curve, includes all consecutive patients with no exclusions. And I think your comments are well taken. We do OPCAB differently now than we did before, but we also do on-pump surgery differently than we did before. And we think we do both better.
DR SUBRAMANIAN: That's correct. I think 2003 is sort of a breaking point for OPCAB, at least for us to do better, so perhaps the benefit of it being sort of neutralized by the lack of that technology improvement, so I'm just asking whether you can go back and look at it.
DR MICHAEL DENYER (Idaho Falls, ID): Dr Puskas, thank you for bringing us this important data. In recognition of the patients who convert intraoperatively from OPCAB to on-pump bypass, could you comment on how that was handled statistically in your analysis?
DR PUSKAS: It's a very important consideration. On January 1, 2004, the STS national database incorporated a data field to count intraoperative conversions from off-pump bypass to on-pump. We had created such a data field in our Emory institutional STS database two years earlier, in 2002. Prior to that, all the patients were entered into the database according to the operation they ultimately received.
Because this data set spans well before 2002, a minority of patients have that data field available. We did examine those particular patients for whom this data field was available and we found about a 2% intraoperative conversion rate. That did not change any of these analyses. So we did go back and look among those patients for whom we had this data, to do an intention-to-treat analysis, and it did not change any of the conclusions or meaningfully change any of the numbers. So rather than try to mix those two, we did this as a straight-up per treatment analysis, not an intention-to-treat analysis per se.
We look forward to repeating this with the national database, in which case we will truncate the data set beginning only in 2004 and do it by a formal intention-to-treat analysis.
DR MOHR: You just gave us some data about survival and mortality. And the stroke is our problem, more or less, in terms of comparison to PCI (percutaneous coronary intervention). Does your data set calculate some differences in terms of stroke rates?
DR PUSKAS: Interestingly, the STS does, in fact, have predicted risk models for stroke, prolonged length of stay, renal failure and death, or any mortality or any morbidity. And we have begun to look at, and to construct, similar curves for those other predicted risk models. They are surprisingly similar to this. In each case the off-pump group has outcomes favorable to those in the on-pump group. Mortality seems to be that which is most responsive to the predicted risk. As I say, the curves diverge and continue to diverge in an accelerating way with mortality, whereas for stroke and renal failure they tend to diverge in a more parallel linear way.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Kinoshita, T. Asai, T. Suzuki, S. Kuroyanagi, S. Hosoba, and N. Takashima Off-pump Bilateral Skeletonized Internal Thoracic Artery Grafting in Elderly Patients Ann. Thorac. Surg., February 1, 2012; 93(2): 531 - 536. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Y. Emmert, S. P. Salzberg, H. R. Cetina Biefer, S. H. Sundermann, B. Seifert, J. Grunenfelder, S. Jacobs, and V. Falk Total arterial off-pump surgery provides excellent outcomes and does not compromise complete revascularization Eur J Cardiothorac Surg, January 12, 2012; (2012) ezr225v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. D. Hillis, P. K. Smith, J. L. Anderson, J. A. Bittl, C. R. Bridges, J. G. Byrne, J. E. Cigarroa, V. J. DiSesa, L. F. Hiratzka, A. M. Hutter Jr, et al. 2011 ACCF/AHA Guideline for Coronary Artery Bypass Graft Surgery: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines Developed in Collaboration With the American Association for Thoracic Surgery, Society of Cardiovascular Anesthesiologists, and Society of Thoracic Surgeons J. Am. Coll. Cardiol., December 6, 2011; 58(24): e123 - e210. [Full Text] [PDF]
|
|
|
|
 |
|
 Writing Committee Members, L. D. Hillis, P. K. Smith, J. L. Anderson, J. A. Bittl, C. R. Bridges, J. G. Byrne, J. E. Cigarroa, V. J. DiSesa, L. F. Hiratzka, et al. 2011 ACCF/AHA Guideline for Coronary Artery Bypass Graft Surgery: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines Circulation, December 6, 2011; 124(23): e652 - e735. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Y. Emmert, B. Seifert, M. Wilhelm, J. Grunenfelder, V. Falk, and S. P. Salzberg Aortic no-touch technique makes the difference in off-pump coronary artery bypass grafting J. Thorac. Cardiovasc. Surg., December 1, 2011; 142(6): 1499 - 1506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P. Taggart and D. G. Altman Off-pump vs. on-pump CABG: are we any closer to a resolution? Eur. Heart J., November 7, 2011; (2011) ehr374v1. [Full Text] [PDF]
|
|
|
|
 |
|
 V. Falk and D. P. Taggart NICE guidance for off-pump CABG: turn off the pump Heart, November 1, 2011; 97(21): 1731 - 1733. [Full Text] [PDF]
|
|
|
|
 |
|
 N. H. Jun, J. K. Shim, J. C. Kim, and Y. L. Kwak Prognostic value of a tissue Doppler-derived index of left ventricular filling pressure on composite morbidity after off-pump coronary artery bypass surgery Br. J. Anaesth., October 1, 2011; 107(4): 519 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Kuss and J. Borgermann Do higher-risk patients benefit from off-pump coronary artery bypass grafting? Evidence from an ecologic analysis of randomized trials J. Thorac. Cardiovasc. Surg., September 1, 2011; 142(3): e117 - e122. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Blasberg, G. S. Schwartz, and S. K. Balaram The role of gender in coronary surgery Eur J Cardiothorac Surg, September 1, 2011; 40(3): 715 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Jacobs, F. H. Edwards, D. M. Shahian, R. L. Prager, C. D. Wright, J. D. Puskas, D. L. S. Morales, J. S. Gammie, J. A. Sanchez, C. K. Haan, et al. Successful Linking of The Society of Thoracic Surgeons Database to Social Security Data to Examine Survival After Cardiac Operations Ann. Thorac. Surg., July 1, 2011; 92(1): 32 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Y. Emmert, S. P. Salzberg, B. Seifert, H. Rodriguez, A. Plass, S. P. Hoerstrup, J. Grunenfelder, and V. Falk Is off-pump superior to conventional coronary artery bypass grafting in diabetic patients with multivessel disease? Eur J Cardiothorac Surg, July 1, 2011; 40(1): 233 - 239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. H. Moller, M. J. Perko, J. T. Lund, L. W. Andersen, H. Kelbaek, J. K. Madsen, P. Winkel, C. Gluud, and D. A. Steinbruchel Three-year follow-up in a subset of high-risk patients randomly assigned to off-pump versus on-pump coronary artery bypass surgery: the Best Bypass Surgery Trial Heart, June 1, 2011; 97(11): 907 - 913. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. N. Patel and G. D. Angelini Reply to the editor. J. Thorac. Cardiovasc. Surg., May 1, 2011; 141(5): 1331 - 1332.e1. [Full Text] [PDF]
|
|
|
|
|
|
D. J. LaPar, C. M. Bhamidipati, T. B. Reece, J. C. Cleveland, I. L. Kron, and G. Ailawadi Is off-pump coronary artery bypass grafting superior to conventional bypass in octogenarians? J. Thorac. Cardiovasc. Surg., January 1, 2011; 141(1): 81 - 90. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. N. Patel and G. D. Angelini Off-pump coronary artery bypass grafting: For the many or the few? J. Thorac. Cardiovasc. Surg., November 1, 2010; 140(5): 951 - 953.e1. [Full Text] [PDF]
|
|
|
|
|
|
A. Miceli, M. Glauber, and G. D. Angelini Letter by Miceli et al Regarding Article, "No Major Differences in 30-Day Outcomes in High-Risk Patients Randomized to Off-Pump Versus On-Pump Coronary Bypass Surgery: The Best Bypass Surgery Trial" Circulation, October 26, 2010; 122(17): e497 - e497. [Full Text] [PDF]
|
|
|
|
|
|
C. H. Moller, M. J. Perko, J. T. Lund, D. A. Steinbruchel, L. W. Andersen, H. Kelbaek, J. K. Madsen, P. Winkel, and C. Gluud Response to Letters Regarding Article, "No Major Differences in 30-Day Outcomes in High-Risk Patients Randomized to Off-Pump Versus On-Pump Coronary Bypass Surgery: The Best Bypass Surgery Trial" Circulation, October 26, 2010; 122(17): e499 - e499. [Full Text] [PDF]
|
|
|
|
|
|
W. Hueb, N. H. Lopes, A. C. Pereira, A. C. Hueb, P. R. Soares, D. Favarato, R. D'Oliveira Vieira, E. G. Lima, C. L. Garzillo, F. da Silva Paulitch, et al. Five-Year Follow-Up of a Randomized Comparison Between Off-Pump and On-Pump Stable Multivessel Coronary Artery Bypass Grafting. The MASS III Trial Circulation, September 14, 2010; 122(11_suppl_1): S48 - S52. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Marx, T. W. Jax, S. Perings, M. Schannwell, U. Sunderdiek, and H. Gulker Evaluation of early and late postoperative flow capacity of internal thoracic artery bypass by means of stress Doppler echocardiography Interact CardioVasc Thorac Surg, June 1, 2010; 10(6): 1003 - 1008. [Abstract] [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 |
