| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of Cardiothoracic Surgery, University of Missouri-Columbia School of Medicine, Columbia, Missouri
b Division of Cardiothoracic Surgery, Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas
c The Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
d The Texas Heart Institute at St. Luke's Episcopal Hospital, Houston, Texas
e Department of Educational Psychology, University of Houston, Houston, Texas
Accepted for publication February 14, 2011.
* Address correspondence to Dr Gopaldas, Division of Cardiothoracic Surgery, University of Missouri-Columbia School of Medicine, One Hospital Dr, Ste MA312, Columbia, MO 65212 (Email: gopaldasr{at}health.missouri.edu).
Presented at the Fifty-seventh Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 3–6, 2010.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: From Nationwide Inpatient Sample database 1998 to 2007, we identified 6,153 (28.9%) patients who underwent CEA before or after CABG during the same hospital admission but not on the same day (STAGED) and 16,639 patients who underwent both procedures on the same day (SYNC). Hierarchic multivariable regression was used to assess the independent effect of operative strategy on mortality, neurologic and overall complications, and charges.
Results: Mean age (69.5 ± 9.0 years) and Charlson-Deyo score (4.6 ± 1.5) were similar for both groups. Mortality (4.2% vs 4.5%) or neurologic complications (3.5% vs 3.9%) were similar between the STAGED and SYNC groups (p > 0.7 for both). The STAGED patients had higher morbidity (48.4% vs 42.6%; odds ratio [OR] 1.8; 95% confidence interval [CI], 1.5 to 2.2; p <0.001) and more cardiac (OR, 1.5; 95% CI, 1.4 to 1.7; p <0.001), wound (OR, 2.1; 95% CI, 1.8 to 2.4; p < 0.001), respiratory (OR, 1.2; 95% CI, 1.1 to 1.3; p = 0.001), and renal complications (OR, 1.2; 95% CI, 1.03 to 1.3; p < 0.001). In SYNC patients, on-pump CABG increased stroke rates (OR, 1.6; 95% CI, 1.3 to 1.9; p < 0.001). The STAGED procedures were independently associated with higher hospital charges by $23,328 (p < 0.001).
Conclusions: We identified no significant difference in mortality or neurologic complications between STAGED and SYNC approaches. Staged procedures were associated with a greater risk of overall complications and higher hospital charges than SYNC. On-pump CABG was associated with higher stroke rates in SYNC patients.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
| 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.
|
The appropriate timing of surgical intervention for patients with coexisting carotid and coronary artery disease (CAD) has been controversial [1]. In clinical practice, the timing is typically determined from the extent of the vascular bed involved and by symptomatology. Strategies include a staged approach in which the more symptomatic vascular bed is intervened upon first and a synchronous approach in which both carotid endarterectomy (CEA) and coronary revascularization are performed during one general anesthetic session. Although carotid disease increases the stroke risk during coronary artery bypass grafting (CABG) and adds to the stroke risk associated with other factors, patients with existing CAD who undergo CEA are at high risk of developing perioperative myocardial infarction. However, combined and staged approaches have not been evaluated at a national level. The most recent study, conducted by Prasad and colleagues [2], was limited by the fact that it compared patients who underwent a dual procedure with patients who underwent a single procedure; there were no data regarding either the timing of the previous CEA or the symptomaticity of the carotid disease.
Our aim was to assess the nationwide use of two approaches to performing CEA and CABG: synchronous (ie, performed on the same day) and staged (ie, performed on different days during the same admission). We also evaluated differences in surgical outcomes and costs between the two operative approaches.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Variables available in the NIS database include patient and hospital demographics, payer information, treating and concomitant diagnoses, inpatient procedures, in-hospital mortality, and length of hospital stay. This database also captures up to 15 International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic and procedure codes per admission.
This study was exempted by the Institutional Review Board of Baylor College of Medicine because the data were nonidentifiable. The reported data conform to the data-use agreement for the NIS from the Healthcare Cost and Utilization Project.
Patient Selection
Specific procedural ICD-9 codes were used to select patients who underwent both CEA (ICD-9 code 38.12) and CABG (IDC 9 codes 36.10 to 36.16) on the same admission [4]. Only those with a primary diagnosis of carotid disease or CAD were selected. Using the methods described by Timaran and colleagues [5] and Gopaldas and colleagues [6], we excluded patients who underwent other cardiac procedures or carotid stenting during the same admission. Off-pump CABG procedures were identified by using methods reported by Chu and colleagues [7]. In the published literature, we and others have reported the feasibility of querying the NIS database for cardiac surgical procedures [8–10]. Using the corresponding procedure day code for each procedure code, we determined whether patients underwent CEA and CABG on the same day or on different days. The assumption was that if both procedures were performed on the same day, they were synchronous.
The Deyo Index [11] was used to compare preoperative morbidity between groups and for risk adjustment in statistical models. This index is a modification of the weighted Charlson comorbidity index and was designed specifically for use with administrative databases like the NIS database.
Study Endpoints
The primary outcomes of interest were in-hospital mortality, neurologic complications (eg, stroke), and combined death-stroke. The secondary endpoints were in-hospital morbidity, length of stay, and hospital charges.
All-cause in-hospital morbidity (Table 1) was recorded according to the ICD-9 diagnosis codes. Intraoperative-procedure-related complications were specifically defined as complications that occurred during a procedure or were directly related to technical aspects of the procedure.
|
2 statistics were computed to examine intergroup differences in age, race, sex, Deyo Index score, mortality, hospital length of stay, and discharge status. Risk-adjusted multivariable analysis with hierarchic linear and logistic regression methods were used to analyze the independent effect of staged versus synchronous procedures. We know that, because of the large sample size, our analyses could reveal differences that would be statistically significant but of minimal clinical significance. Therefore, 3 different effect size statistics were computed to assess the practical implications of our findings. Cohen's d was calculated for continuous data by using pooled standard deviations and was appropriately weighted for unequal sample size [12]. The phi (
) coefficient was computed for 2 tests for independence with 1 DF (degree of freedom). Cramer's V was computed for variables with more than 1 DF [13]. The following cutoff values were used to judge the computed effect sizes: less than 0.33 (small), 0.33 to 0.55 (medium), and greater than 0.55 (large). The sample groups were assumed to be similar for a particular characteristic if the effect sizes were small. |
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
|
Hospital Length of Stay and Charges
Hospital stays were longer for STAGED patients (14.7 ± 8.9 days) than for SYNC patients (11.64 ± 9.8 days; p < 0.001). Linear regression showed that the staged approach was independently associated with a longer hospital stay by 3.1 days (p < 0.001). Inflation-adjusted hospital charges were $118,801 ± 78,644 for STAGED patients and $98,106 ± 80,053 for SYNC patients (p < 0.001). Risk-adjusted models indicated that staged procedures were independently associated with a $23,328 higher hospital charge.
Subgroup Analysis: CEA Before CABG and CABG Before CEA
To further characterize the higher complication rates in the STAGED group, this group was split into two subgroups based on whether CEA or CABG was performed first. Among the STAGED patients, 585 (9.5%) underwent CABG first, and the other 5,568 (90.5%) underwent CEA first. As expected, cardiac complications were significantly more frequent among patients who underwent CEA first (838 of 5,568 [15.1%] vs 61 of 585 [10.4%]; p < 0.001), whereas neurologic complications were more frequent in patients who underwent CABG first (45 of 585 [7.7%] vs 173 of 5,568 [3.1%]; p < 0.001). Using SYNC as the reference group for risk-adjusted comparison, performing CEA first was associated with a 57% higher chance of cardiac complications than performing CABG first (OR 1.57; 95% CI, 1.41 to 1.75; p < 0.001).
Subgroup Analysis: On-pump Versus Off-pump Coronary Revascularization
A subgroup analysis comparing on-pump and off-pump CABG was performed within the STAGED and SYNC groups (Table 4). Off-pump CABG was performed in 5,280 patients (31.7%) in the SYNC group and in 2,004 patients (32.5%) in the STAGED group. Risk-adjusted regression models showed that performing CABG off-pump did not affect mortality, combined death-stroke, or overall complication rates in the SYNC and STAGED patients. However, on-pump CABG was associated with an elevated risk of stroke in SYNC patients (OR 1.55; 95% CI, 1.27 to 1.91; p < 0.001), but not in STAGED patients.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
In contemporary clinical practice, the synchronous approach is typically used when both vascular beds are equally symptomatic. However, "equally symptomatic" is hard to define quantitatively, especially if separate vascular or cardiac surgical teams are involved in evaluating the patient. The use of the synchronous approach was initially substantiated only by small randomized trials reported in the 1980s [16, 17], and most contemporary surgeons acknowledge that an ideal randomized study is still lacking [18]. An extensive systematic review of 97 published studies found no difference in the outcomes of staged and synchronous approaches, although the authors acknowledge the lack of comparable data for the two modalities [19]. The two main groups of patients in our study were more comparable because they each included patients who underwent both procedures during the same admission; the only difference was that the procedures were performed on the same day or on different days.
The synchronous approach is typically reserved for patients with severe symptoms involving both vascular beds. The study by Estes and colleagues [20] of combined CABG-CEA patients showed that the synchronous approach produced better outcomes in patients with asymptomatic carotid disease than in patients with symptoms in both vascular beds, thus supporting the use of the synchronous approach with asymptomatic carotid disease.
Our study supports the safety of the synchronous approach. Unlike many previous studies, in which historic controls were used, or in which patients who underwent combined procedures were compared with patients who underwent a single intervention, our study compared two closely matched groups of patients.
Although the effect sizes for all baseline differences were small, we corrected for these minor differences in the risk-adjusted model and found no between-group differences in mortality or stroke rate. In fact, the staged approach was associated with a higher overall complication rate that was mainly attributable to cardiac, wound, respiratory, and renal complications. Although the staged approach was independently associated with fewer intraoperative complications, this benefit was negated by the increase in postoperative complications associated with the staged approach. Furthermore, it is important to note that the staged approach was associated with higher hospital charges and longer hospital stays. Other single-center studies have also supported the economic benefits of synchronous CABG-CEA and the shorter hospital length of stay it requires [21, 22].
Because patients were admitted with a primary diagnosis of CAD or carotid disease, the assumption was that both lesions were severe enough to warrant surgical intervention. Our study used a more valid control group than other large database studies, thereby providing a more accurate assessment of mortality and complications. Our subgroup analysis of the STAGED group showed, as expected, that there were more cardiac complications when CEA was performed first and more neurologic complications when CABG was performed first. However, the nature of the NIS data did not allow us to conduct our analysis on an intention-to-treat basis. Thus, it is possible that some of our STAGED patients underwent the second procedure to treat a complication of the first. For example, a patient who was meant to undergo isolated CABG may have undergone subsequent CEA to treat a postoperative stroke. Although it is not possible to separate such patients from those whose staged procedures were planned, both types of patients nonetheless had lesions of significant severity in both vascular beds that met surgical criteria for intervention during the same admission, and it is around such patients that controversy centers regarding the timing of surgical interventions.
We found that in the SYNC group, off-pump CABG was associated with lower stroke risk than on-pump CABG, which is consistent with recent literature [23, 24]. The benefit of off-pump CABG probably results from avoiding the lower perfusion pressures associated with cardiopulmonary bypass. Off-pump CABG did not benefit STAGED patients, probably because the relief of carotid stenosis by CEA allowed unrestricted flow to the cerebral circulation and thus minimized the negative impact of lower perfusion pressures due to cardiopulmonary bypass.
One limitation of our study was that the exact severity of the carotid or coronary disease was unknown. Major adverse cardiovascular events, the gold standard of outcomes after vascular and cardiac procedures, could not be assessed in our analysis because postoperative myocardial infarction was not distinctly captured by the ICD-9 codes. The closest approximation to major adverse cardiovascular events in our study was the combined death-stroke rate. Additionally, our study focused only on in-hospital outcomes.
In addition, our study did not include patients for whom a staged approach was planned but who died or had major morbidity after the first procedure that precluded a second procedure. These patients could not be identified in the NIS database, which created a potential selection bias that could have skewed our results in favor of the staged approach. Similarly, patients who underwent CABG and CEA during separate but temporally close admissions should technically be considered to have undergone a staged procedure, but these patients were excluded from our analysis because the NIS separates the discharge records of these patients.
A significant limitation of our study is that for about half of the patients, data on the dates of admission, operation, or discharge were missing, which prevented us from classifying their operations as synchronous or staged. As a result, we were forced to exclude these patients from the final study sample. It is unknown whether this exclusion selectively biased the results one way or the other, but in the context of interpreting our results this limitation should be strongly taken into consideration. This study focused on a limited subset of patients who either were symptomatic from both lesions or had lesions severe enough to necessitate intervention on both vascular territories within the same admission.
The controversy regarding the timing of surgical interventions (CEA, CABG) is likely to continue. The availability of newer treatment modalities, such as carotid stenting, further complicates the issue by providing more choices. Also, the involvement of surgeons from two different specialties adds a political dimension to the choice between staged and synchronous treatment; each surgeon may be disinclined to operate on one vascular bed unless concomitant disease in the other bed has been corrected.
In summary, a synchronous approach to CABG and CEA is used in a significant proportion of patients when both procedures are performed during the same admission. Compared with the staged approach, the synchronous approach appears to be associated with equivalent mortality and death-stroke profiles, a lower overall complication rate, and lower hospital charges. Additionally, using the off-pump approach to CABG may reduce stroke rates in patients who undergo synchronous CEA and CABG.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
DR GOPALDAS: There is certainly an issue with surgeons trying to do this. Again, that is center dependent. This database captures the data from pretty much all participating hospitals in 37 states. It is not possible in the database to tease whether the carotid endarterectomy is done by a vascular surgeon or a cardiac surgeon. Typically, if a cardiac surgeon is performing both a carotid endarterectomy and a CABG [coronary artery bypass grafting], they are more likely to perform that in a synchronous approach. But I am sure most of you have experienced that if you are trying to get a vascular surgeon to do a carotid endarterectomy on somebody who has a left main, you are probably going to get nowhere because they are probably not going to fix it. There is always going to be so much fudging with the imaging studies and they may conclude that a patient may not need CEA [carotid endarterectomy] before CABG.
For the second question, we did consider doing a propensity match, and there are a couple of issues. One is you have to propensity match the 6,000 patients who underwent the staged approach. I have tried this with SAS and SPSS. It takes about seven to eight days to run the whole analysis unless you are tethering four computers. So that is one issue there from a technical standpoint. The second issue is, propensity matching will cut down our sample size, because we are going to be eliminating the remaining patients in the larger group. And for creating a propensity score, the first-level analysis to be done is actually a multivariable regression model to generate the propensity score. So our approach in the majority of the circumstances is if you are using the propensity approach, you have to be careful because you are still building that on a multivariable regression model, which is what we have actually used, although we didn't create a propensity score. So a propensity score is just an extended form of multivariable regression, and statisticians may argue one way or the other.
We could have certainly done that, but I think we were limited by the ability to use a very powerful computer. I was told that statisticians need to tie about four computers to get this done to revalidate the results. So that is why it wasn't done.
DR SAQIB MASROOR (Milwaukee, WI): Does the database tell you what procedure was done first in the sequential?
DR GOPALDAS: Yes, correct. We can find out if the carotid endarterectomy was done first or if the CABG was done first.
DR MASROOR: And does that make a difference?
DR GOPALDAS: I think it does make a difference. Actually if we can pull up my slides again, I do have a relevant slide. I did not include that in the presentation for time issues.
(Slide) So this is actually a subgroup analysis, which will be in the manuscript. The first one, 585 patients underwent CABG first, and I just highlighted the outcomes which are of relevance. Neurologic complications, as anticipated, were higher in patients who underwent CABG first, but then when we did the risk-adjusted analysis, the p value was not significant, although the odds ratio is 1.3. In patients who underwent carotid endarterectomy first, as anticipated, cardiac complications were higher, 15% compared to 10%, and then when we did a risk-adjusted analysis, we found a 57% higher chance of developing a cardiac complication in patients who underwent CEA first before CABG.
DR IKONOMIDIS: What do you do now at your institution?
DR GOPALDAS: Well, since I do not have privileges for carotid endarterectomies in my place, I have to request the assistance of the vascular surgeons. Sometimes you have to be firm in your stance, so that they can still do the operation for you.
DR IKONOMIDIS: Are you rescheduling your cases around your vascular surgery colleagues now so that you can get synchronous vascular and cardiac surgery performed?
DR GOPALDAS: Yes. But I think the bigger problem is actually to get them to even do it.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
N. Reduction in hospitalisation rates following simultaneous carotid endarterectomy and coronary artery bypass grafting; experience from a single centre Interact Cardiovasc Thorac Surg 2006;5:367-372.This article has been cited by other articles:
|
|
 |
|
  R. R. Gopaldas, T. K. Dao, S. A. LeMaire, J. Huh, and J. S. Coselli Endovascular versus open repair of ruptured descending thoracic aortic aneurysms: A nationwide risk-adjusted study of 923 patients J. Thorac. Cardiovasc. Surg., November 1, 2011; 142(5): 1010 - 1018. [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 |
