Author + information
- Received September 30, 2016
- Revision received November 15, 2016
- Accepted November 17, 2016
- Published online February 6, 2017.
- Dmitriy N. Feldman, MDa,∗ (, )
- Rajesh V. Swaminathan, MDb,
- Joshua D. Geleris, MDa,
- Peter Okin, MDa,
- Robert M. Minutello, MDa,
- Udhay Krishnan, MDa,
- Daniel J. McCormick, DOc,
- Geoffrey Bergman, MDa,
- Harsimran Singh, MDa,
- S. Chiu Wong, MDa and
- Luke K. Kim, MDa
- aDivision of Cardiology, Weill Cornell Medical College, New York Presbyterian Hospital, New York, New York
- bDuke University Medical Center, Duke Clinical Research Institute, Durham, North Carolina
- cDepartment of Cardiovascular Medicine, Pennsylvania Hospital–University of Pennsylvania Health System, Philadelphia, Pennsylvania
- ↵∗Address for correspondence:
Dr. Dmitriy N. Feldman, Division of Cardiology/Department of Medicine, Interventional Cardiology and Endovascular Laboratory, Weill Cornell Medical College/New York Presbyterian Hospital, 520 East 70th Street, New York, New York 10021.
Objectives The aim of this study was to compare trends and outcomes of 3 approaches to carotid revascularization in the coronary artery bypass graft (CABG) population when performed during the same hospitalization.
Background The optimal approach to managing coexisting severe carotid and coronary disease remains controversial. Carotid endarterectomy (CEA) or carotid artery stenting (CAS) are used to decrease the risk of stroke in patients with carotid disease undergoing CABG surgery.
Methods The authors conducted a serial, cross-sectional study with time trends of 3 revascularization groups during the same hospital admission: 1) combined CEA+CABG; 2) staged CEA+CABG; and 3) staged CAS+CABG from the Nationwide Inpatient Sample database 2004 to 2012. The primary composite endpoints were in-hospital all-cause death, stroke, and death/stroke.
Results During the 9-year period, 22,501 concurrent carotid revascularizations and CABG surgeries during the same hospitalization were performed. Of these, 15,402 (68.4%) underwent combined CEA+CABG, 6,297 (28.0%) underwent staged CEA+CABG, and 802 (3.6%) underwent staged CAS+CABG. The overall rate of CEA+CABG decreased by 16.1% (ptrend = 0.03) from 2004 to 2012, whereas the rate of CAS+CABG did not significantly change during these years (ptrend = 0.10). The adjusted risk of death was greater, whereas risk of stroke was lower with both combined CEA+CABG (death odds ratio [OR]: 2.08, 95% confidence interval [CI]: 1.08 to 3.97; p = 0.03; stroke OR: 0.65, 95% CI: 0.42 to 1.01; p = 0.06) and staged CEA+CABG (death OR: 2.40, 95% CI: 1.43 to 4.05; p = 0.001; stroke OR: 0.50, 95% CI: 0.31 to 0.80; p = 0.004) approaches compared with CAS+CABG. The adjusted risk of death or stroke was similar in the 3 groups.
Conclusions In patients with concomitant carotid and coronary disease undergoing combined revascularization, combined CEA+CABG is utilized most frequently, followed by staged CEA+CABG and staged CAS+CABG strategies. The staged CAS+CABG strategy was associated with lower risk of mortality, but higher risk of stroke. Future studies are needed to examine the risks/benefits of different carotid revascularization strategies for high-risk patients requiring concurrent CABG.
Carotid artery disease is an important cause of ischemic stroke and often coexists with coronary artery disease. Optimal treatment of carotid artery disease remains controversial and is therefore a focus of ongoing clinical trials (1). The benefit of carotid endarterectomy (CEA) over medical therapy in both symptomatic and asymptomatic patients has been established in randomized trials performed in the early 1990s (2,3). Carotid artery stenting (CAS) was introduced as a minimally invasive endovascular alternative to CEA, particularly in patients considered to be poor candidates for CEA because of advanced age, high-risk anatomic features, or medical comorbidities (4,5). Recently published results from the CREST trial (Carotid Revascularization Endarterectomy Versus Stenting Trial) and the ACT I trial (Asymptomatic Carotid Trial I) have demonstrated similar long-term outcomes between CAS and CEA strategies with respect to the risk of stroke, myocardial infarction (MI), or death (6,7).
The prevalence of significant carotid disease among patients undergoing coronary artery bypass graft surgery (CABG) is estimated to be 6% to 14% (8,9). Severe internal carotid artery disease, particularly in recently symptomatic patients, is an important risk factor for developing stroke after CABG (10,11). Therefore, CEA either before (staged) or simultaneous with (combined) cardiac surgery are proposed revascularization strategies to prevent death/stroke following CABG (12). However, the reported incidence of adverse events after either approach remains high (10% to 12%) (13). CAS has been proposed as a minimally invasive alternative to traditional carotid endarterectomy for CABG patients in consideration of their high risk of post-operative adverse events (14–22). Yet, in the absence of randomized clinical trials, the best approach to the management of concomitant severe carotid and coronary disease in the CABG population remains controversial. Using the Nationwide Inpatient Sample (NIS) database 2004 to 2012, we examined temporal U.S. trends in CEA and CAS use when performed during the same hospitalization as CABG and compared risk-adjusted in-hospital outcomes between the 3 strategies: 1) combined CEA+CABG; 2) staged CEA+CABG; and 3) staged CAS+CABG procedures. Furthermore, we examined in-hospital outcomes in key subgroups of age, sex, and asymptomatic/symptomatic carotid artery disease status, as well as changes in outcomes over time.
Data source, study population, and endpoints
Data were obtained from the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project NIS files between 2004 and 2012 (23). The NIS is a 20% stratified sample of all nonfederal U.S. hospitals and, in 2012, contained deidentified information on 36,484,846 discharges from 4,378 hospitals and 44 states. The values reported in the analysis represent 5 times the number of examined NIS discharges, corresponding to 20% sampling of NIS records. Discharges are weighted based on the sampling scheme to permit inferences for a nationally representative population (23). Each record in the NIS includes all procedure and diagnosis International Classification of Diseases (ICD) codes recorded for each patient’s hospital discharge.
From January 2004 through December 2012, hospitalizations leading to CABG were selected by searching for the International Classification of Diseases-Ninth Revision-Clinical Modification (ICD-9-CM) procedure codes 36.10, 36.11, 36.13, 36.14, 36.15, 36.16, 36.17, or 36.19 in any of the 15 procedure fields in the database. The absence of ICD-9-CM codes 39.61 and 39.62 identified off-pump CABG surgery. Concomitant valve surgeries were identified by the ICD-9-CM codes 35.2. CEA and CAS procedures were selected using the ICD-9-CM codes 38.12 and 00.63, respectively, in any of the 15 procedure fields during the same hospitalization. Any patients who underwent CEA or CAS after CABG were excluded; patients undergoing CAS and CABG on the same day were excluded. Three carotid revascularization strategies were examined: CEA before CABG (staged CEA+CABG), combined CEA and CABG (performed on the same day), and CAS before CABG (staged CAS+CABG). Using the corresponding procedure day code for each procedure code, we determined whether patients underwent CEA/CAS and CABG on the same day or on different days. The assumption was that if CEA and CABG procedures were performed on the same day, they were synchronous.
Patient-level and hospital-level variables were included as baseline characteristics. Hospital-level data elements were derived from the American Hospital Association Annual Survey Database. The Agency for Healthcare Research and Quality comorbidity measures based on the Elixhauser methods were used to identify comorbid conditions (24). To distinguish between asymptomatic and symptomatic carotid artery disease, symptomatic status was identified by the ICD-9 codes 362.31, 368.12, 781.4, 433.11, 435, and 434. The primary outcome measures were in-hospital all-cause mortality, stroke, and mortality/stroke. Post-procedural stroke was identified by ICD-9 code 997.02.
Trends in the annual rates of concurrent CABG and carotid revascularization procedures were assessed using linear regression modeling. For descriptive analyses, we compared baseline patient and hospital characteristics for: 1) combined CEA+CABG; 2) staged CEA+CABG; and 3) staged CAS+CABG procedures. Continuous variables are presented as medians; categorical variables are expressed as frequencies (percentages). To compare baseline characteristics and in-hospital care patterns with respect to undergoing combined CEA+CABG, staged CEA+CABG, and staged CAS+CABG, either Mann-Whitney Wilcoxon nonparametric tests or analysis of variance tests were used for continuous variables, and Pearson chi-square tests were used for categorical variables. All statistical tests were 2-sided, and a p value of <0.05 was set a priori to be statistically significant.
Unadjusted in-hospital outcome rates were calculated for each of the 3 revascularization strategies. Multivariable logistic regression analyses were used to compare outcomes between the 2 strategies of CEA with CABG versus staged CAS+CABG, adjusting for other potential predictors that had significant univariate association with outcomes (p < 0.01). The models were adjusted for the following demographic, hospital-level, and procedural variables: age, sex, race, hospital bed size, hospital teaching status, region, payer, symptom status from carotid disease, anemia, collagen vascular disease, congestive heart failure, chronic pulmonary disease, diabetes mellitus, coagulopathy, hypertension, liver disease, obesity, peripheral vascular disorders, chronic renal failure, concomitant valve surgery, on-pump versus off-pump CABG, prior CABG, and prior percutaneous coronary intervention (PCI). For all regression analyses, the Taylor linearization method “with replacement” design was used to compute variances. To further adjust for potential revascularization strategy selection bias, a logistic propensity score was calculated using a multilevel model including all covariates used in the primary analysis. The multilevel model propensity score was then incorporated as an additional variable in the weighted logistic regression models (25).
In order to assess the effect of a learning curve on the outcomes of concurrent carotid revascularization and CABG, the cohorts were divided into 2 time periods (2004 to 2008 vs. 2009 to 2012) using 2009 as the cutoff. This cutoff year was selected as the midpoint of our study period, which also coincides with the completion of the CREST trial enrollment (5). The effects of age (age ≥80 years vs. age <80 years), sex, and clinical indication (symptomatic vs. asymptomatic) on the relationship between treatment strategy and outcomes were assessed by including interaction terms between procedure performed and the age groups, sex, and symptom status in the propensity score-adjusted models. All statistical tests were 2-sided, and a p value of <0.05 was set a priori to be statistically significant. All multivariate regression analyses were conducted using SAS, version 9.2 (SAS Institute, Cary, North Carolina) and SPSS, version 20 (IBM Corporation, Armonk, New York).
Study population and trends
Of 351,367,791 discharge records analyzed between 2004 and 2012, 2,303,088 patients underwent CABG. Of those, 22,501 patients also had a carotid revascularization during the same hospitalization, with 15,402 (68.4%) undergoing combined CEA+CABG, 6,297 (28.0%) undergoing staged CEA+CABG, and 802 (3.6%) undergoing staged CAS+CABG approaches. The annual rate of CEA and CABG during the same hospitalization decreased by 16% from 2,443 procedures in 2004 to 2,050 procedures in 2012 (ptrend = 0.03). However, the annual rate of staged CEA+CABG increased from 542 to 680 procedures between 2004 and 2012, whereas the annual rate of combined CEA+CABG decreased from 1,901 to 1,370 procedures. The frequency of staged CAS+CABG within the same hospitalization (51 procedures in 2005 and 70 procedures in 2012, ptrend = 0.10) did not change significantly during the study period (Figure 1).
Table 1 compares baseline characteristics of patients undergoing the 3 carotid revascularization strategies relative to CABG within the same hospitalization. Patients undergoing CAS+CABG were less likely to be white, more likely to have a history of hypertension, diabetes with complications, anemia, chronic pulmonary disease, obesity, chronic renal failure, and previous PCI, and less likely to have their procedures during an elective admission. The indication for CEA was predominantly asymptomatic carotid artery disease (96.4% of patients). CABG with concomitant valve repair/replacement or on-pump CABG surgeries was performed more commonly in those undergoing CEA+CABG.
Outcomes after carotid artery revascularization and CABG
The unadjusted incidence of in-hospital death for combined CEA+CABG was 4.4%, for staged CEA+CABG was 3.8%, and for staged CAS+CABG was 1.9% (p < 0.01). The rates of post-procedural stroke were not significantly different between the 3 strategies (p = 0.37) (Figure 2). The death or stroke rate for combined CEA+CABG was 6.8%, for staged CEA+CABG was 5.4%, and staged CAS+CABG was 4.2% (p < 0.01).
Both combined and staged CEA+CABG were associated with significantly increased odds of mortality in both univariate and multivariate propensity score-adjusted logistic regression analyses: adjusted odds ratio (OR): 2.08; 95% confidence interval (CI): 1.08 to 3.97 (p = 0.03) for combined CEA+CABG; adjusted OR: 2.40; 95% CI: 1.43 to 4.05 (p = 0.001) for staged CEA+CABG when compared to CAS+CABG (Table 2). The adjusted risk of stroke was lower with both combined CEA+CABG (OR: 0.65; 95% CI: 0.42 to 1.01; p = 0.06) and, particularly with staged CEA+CABG (OR: 0.50; 95% CI: 0.31 to 0.80; p = 0.004) approach compared with CAS+CABG. The adjusted risk of death or stroke was similar in both CEA groups and staged CAS+CABG group.
In order to assess the effect of operator learning curve on the outcomes after concurrent carotid revascularization and CABG, the cohorts were divided into 2 time periods, 2004 to 2008 and 2009 to 2012. In the staged CEA+CABG group, there were no significant differences in the rates of death (4.1% vs. 3.6%; p = 0.32), stroke (2.0% vs. 1.7%; p = 0.31), and death or stroke (5.8% vs. 4.9%; p = 0.12) when comparing the 2004 to 2008 cohort to the 2009 to 2012 cohort (Figure 3A). However, in the combined CEA+CABG group, there was significant improvement in death (4.7% vs. 3.8%; p < 0.01) and death or stroke (7.3% vs. 6.1%; p < 0.01) with a trend towards improvement in stroke (3.0% vs. 2.5%; p = 0.07) during later years (Figure 3B). In the staged CAS+CABG group, there was significant improvement in outcomes from the 2004 to 2008 cohort to the 2009 to 2012 cohort: death (3.5% vs. 1.0%; p = 0.02), stroke (5.2% vs. 1.7%; p < 0.01), and death or stroke (7.0% vs. 2.7%; p < 0.01) (Figure 3C). After adjustment for baseline differences using propensity score-adjusted multivariate logistic analysis, there was no improvement in the later period (2009 to 2012) in adjusted odds of death, stroke, and death/stroke in the staged CEA+CABG group (Table 3). In the combined CEA+CABG group, there was an improvement in mortality (OR: 0.75, 95% CI: 0.88 to 0.90; p = 0.002) and death/stroke (OR: 0.80, 95% CI: 0.68 to 0.92; p = 0.002) over time. The greatest improvement in outcomes over time was seen in the CAS+CABG group; the adjusted odds ratio of death was 0.19 (95% CI: 0.04 to 0.92; p = 0.04) and death/stroke was 0.33 (95% CI: 0.14 to 0.79; p = 0.01) in 2009 to 2012 compared with 2004 to 2008.
Outcomes after carotid artery revascularization and CABG in key subgroups
In octogenarians (age ≥80 years), the adjusted risk of death was significantly greater in the staged CEA+CABG strategy (OR: 6.67, 95% CI: 3.35 to 13.35; p < 0.001) and in the combined CEA+CABG group (OR: 3.17, 95% CI: 1.82 to 5.54; p < 0.001) when compared with staged CAS+CABG approach, whereas the risk of stroke was similar between 3 groups (Table 4). In both women and men, the combined CEA+CABG approach was associated with a greater risk of death and death/stroke compared with the CAS+CABG approach, with that risk being higher in women. In women, the staged CEA+CABG approach was associated with a greater risk of death (adjusted OR: 2.44, 95% CI: 1.14 to 5.20; p = 0.02), decreased risk of stroke (adjusted OR: 0.34, 95% CI: 0.17 to 0.68; p = 0.003), and overall similar risk of death or stroke. In symptomatic patients, the CAS+CABG strategy was associated with better outcomes compared with the other 2 strategies. The risk of death was ∼3-fold greater (adjusted OR: 3.06, 95% CI: 1.10 to 8.52; p = 0.03) and risk of stroke was ∼4 fold greater (adjusted OR 4.07, 95% CI: 2.21 to 7.47; p < 0.001) in the combined CEA+CABG strategy. The risk of stroke was much greater in the staged CEA+CABG group (adjusted OR: 4.70, 95% CI: 2.60 to 8.48; p < 0.001), whereas the risk of death was similar to the CAS+CABG strategy.
There are several important findings in this large, nationally representative sample of U.S. hospital discharge records comparing 3 strategies of carotid intervention in patients undergoing CABG during the same admission between the years 2004 and 2012. First, CEA and CABG was the predominant strategy, performed in ∼96% of patients undergoing carotid revascularization and CABG during the same hospitalization (68% combined CEA+CABG and 28% staged CEA+CABG). The overall frequency of CEA+CABG decreased by 16% from 2004 to 2012, with a decrease in combined CEA+CABG partially offset by an increase in staged CEA+CABG procedures. The frequency of CAS+CABG procedures remained low and relatively unchanged over this time period. Second, as expected due to reimbursement restrictions imposed by Centers for Medicare & Medicaid Services, CAS was generally performed in a higher risk cohort, which was older, more often with symptomatic carotid disease and with more cardiovascular comorbidities. Third, despite being performed in a higher-risk cohort, the strategy of CAS+CABG was associated with a significantly lower adjusted risk of mortality while carrying a higher risk of stroke. Fourth, patients undergoing CAS+CABG during the later years of 2009 to 2012 had a markedly lower risk of death or stroke compared with the earlier years 2004 to 2008. Finally, in symptomatic carotid disease, both CEA+CABG strategies had markedly increased risk of stroke and death/stroke when compared with the CAS+CABG strategy.
Our results are consistent with prior reports evaluating U.S. utilization trends in carotid interventions, though few have specifically looked at carotid interventions before CABG. As previously demonstrated, there had been a decreasing trend in the annual utilization rate of all carotid artery revascularization procedures in the United States between 2001 and 2010 (26). Overall, CAS rates nationally have remained low and unchanged between 2006 and 2010. On the other hand, whereas overall CEA rates decreased by ∼40% between 2001 and 2010, Timaran et al. (27) demonstrated that between 2001 and 2004, the rates of combined CEA+CABG were not changing significantly. In this analysis, the rates of CEA+CABG procedures during the same hospitalization have declined by ∼16% from 2004 to 2012, with a shift toward more CEA procedures being staged before CABG. Part of the decline in CEA+CABG procedures is likely attributable to controversy in the medical, surgical, and neurological communities about the value of carotid revascularization in asymptomatic patients, including those undergoing CABG, and the optimal strategy for managing carotid disease. In the United States, >85% of patients undergo CEA for asymptomatic carotid artery disease (26), whereas >96% of patients in our cohort underwent CEA+CABG for asymptomatic disease. Given potentially important improvements in medical management of carotid disease since the 1990s, as well as concerns about risks/benefits of revascularization, particularly in high-risk asymptomatic elderly patients, further trials are needed to address this controversy. The ongoing CREST-2 trial aims to address the value of carotid revascularization (CEA and CAS) and intensive medical management versus medical management alone in patients with asymptomatic high-grade carotid stenosis (28), though it will not address peri-CABG stroke, which may have a different pathophysiology. Furthermore, decreasing rates of same-hospitalization CEA and CABG surgeries may be reflective of decreasing prevalence of severe carotid disease nationwide due to improvements in primary and secondary prevention efforts and improved medical management of cerebrovascular and cardiovascular risk factors. Finally, the benefit of prophylactic carotid artery screening before cardiac surgery is uncertain (29), and rates of preoperative carotid stenosis screening may be decreasing over time.
Stroke is one of the major noncardiac complications following CABG. Several observational studies have suggested that the risk of stroke associated with CABG is ∼2% in the overall CABG population without significant carotid disease (30) and ∼3% in patients with asymptomatic, severe carotid stenosis (11). These figures increase to ∼5% in those with bilateral carotid stenosis and to 7% to 11% in those with carotid occlusion (11). Despite the fact that only ∼50% of strokes are ipsilateral to pre-existing carotid artery stenosis, concerns about perioperative cerebral hypoperfusion made CEA an attractive strategy to reduce the excessive stroke and death rates in those with combined carotid and coronary disease. Controversy remains regarding the preferred timing of carotid revascularization with CEA. Early reviews and meta-analyses that have assessed perioperative outcomes of staged and simultaneous CEA and concurrent CABG demonstrated that staged procedures generally were associated with lower stroke and death rates than simultaneous ones (31). More recently, Gopaldas et al. (32) found no significant difference for in-hospital death and stroke rates between the 2 CEA-based strategies using the NIS 1998 to 2007 database. On the other hand, a single-center study by Shishehbor et al. (15) suggested worse 30-day outcomes with a staged CEA+CABG approach, driven by a higher number of interstage MIs. Given the ongoing controversy regarding the timing of CEA, we examined both staged and combined CEA+CABG surgeries performed during the same hospitalization in our analysis in comparison to staged CAS+CABG. In our analysis, the staged CEA+CABG strategy had similar rates for in-hospital death (3.8% vs. 4.4%), stroke (1.9% vs. 2.8%), and death/stroke (5.4% vs. 6.8%) compared with the combined CEA+CABG strategy. Our in-hospital mortality rates for both strategies are in line with the previously reported 4.2% mortality for staged CEA+CABG and 4.5% for combined CEA+CABG mortality rates by Gopaldas et al. (32). Our in-hospital stroke rates for both strategies were slightly lower than 3.5% staged CEA+CABG stroke rate and 3.9% combined CEA+CABG stroke rates seen in this 1998 to 2007 NIS report (32). Improved stroke rates over time after CEA and CABG may be due to better patient selection as well as evolution of medical and surgical therapies for coronary and carotid disease.
Earlier observational studies have examined outcomes of staged CAS followed by CABG, with low periprocedural complication rates (33,34). Overall, most prior studies comparing CAS and CEA before CABG, tend to show better outcomes with the CAS strategy. Ziada et al. (35) compared 30-day outcomes between staged CAS+CABG (n = 56) and concomitant CEA-CABG (n = 112) strategies and reported a trend toward fewer strokes or MIs with CAS+CABG strategy after propensity adjustment (p = 0.06), whereas no difference was seen for combined death, MI, or stroke (p = 0.12). Timaran et al. (27) compared CAS+CABG (n = 887) and CEA+CABG (n = 26,197) approaches in the 2000 to 2004 NIS registry (without differentiating between staged and concomitant procedures) and demonstrated that patients undergoing CAS+CABG experienced a lower incidence of post-operative stroke (2.4% vs. 3.9%), even after risk stratification, and combined stroke and death (6.9% vs. 8.6%) than the combined CEA+CABG strategy (p < 0.001), whereas in-hospital death rates were similar. This is in contrast to our findings of a higher risk of post-procedural strokes between the 2 strategies, but lower risk of mortality with the CAS+CABG strategy. This difference may be partly due to the more contemporary nature of our study and more patients with high-risk clinical and anatomic features for CEA undergoing CAS in later years (26). Even though CAS patients had more cardiovascular comorbidities in our report, the strategy of CAS+CABG was still associated with lowest in-hospital mortality, with a mortality of only 1.0% in the years 2009 to 2012. Despite this, only ∼4% of U.S. patients requiring both carotid and coronary revascularization currently undergo concurrent CAS+CABG. Greater emphasis may need to be placed on patient preference, given data demonstrating that a significant portion of patients still prefer CAS over CEA even when being quoted a slightly higher periprocedural risk in comparison with CEA (36).
Shishehbor et al. (15) retrospectively examined 3 strategies of carotid revascularization before open-heart surgery (OHS): staged CEA+OHS (n = 45), combined CEA+OHS (n = 195), and staged CAS+OHS (n = 110). At 30-days, stroke rates were similar (p = 0.11) in CAS+OHS (2%), staged CEA+OHS (2%), and numerically higher in the combined CEA+OHS group (7%); with statistically higher rates of MIs in the staged CEA+OHS group. The association between CAS+CABG and decreased mortality in our study may be partly due to fewer cardiovascular and cerebrovascular post-operative adverse events, but may also be due to a selection bias for a particular strategy and the urgent need to revascularize severe coronary artery disease with a combined CEA+CABG approach. Furthermore, in our analysis in those with symptomatic carotid disease, both CEA+CABG strategies had markedly increased risk of stroke and death/stroke when compared with the CAS+CABG approach. Prior series suggested that perhaps in a cohort of symptomatic carotid patients with a small minority undergoing CABG, there might be a higher risk of stroke associated with a CAS+CABG strategy (27). Importantly, our study provides reassuring data that the risk of stroke may actually be ∼4-fold less in symptomatic patients with CAS+CABG versus either CEA+CABG strategy, although only a small number of symptomatic patients were examined.
Our study suggests that the risk of adverse outcomes after CAS+CABG and combined CEA+CABG procedures has declined during the course of 9 years. Given >25 years of experience with CEA, despite the decreasing number of these procedures, the most likely explanation for improving outcomes after CEA+CABG over time may be attributable to better case selection for CEA based on clinical and anatomic high-risk criteria. Prior studies have suggested a strong relationship between operator learning curve and outcomes after CAS (37–39). Improved outcomes over time with concurrent CAS+CABG procedures are likely due to better patient selection before CAS, technological advances, and a steep learning curve of operator procedural skills.
Although the present study is large, contemporary, and based on the entire spectrum of carotid revascularization and CABG experience in the United States, several limitations of this study should be acknowledged. First, this is a retrospective study based on data from NIS, with the sample designed to approximate the national distribution of key hospital characteristics. Our estimates for CEA+CABG and CAS+CABG were derived from a 20% sample of U.S. hospitals, and it is possible that either cohort were either under-represented or over-represented by this sample. However, the NIS has been used extensively to examine national health care trends, and its sampling design has been validated in numerous publications (26). Second, miscoded and missing data can occur in large administrative datasets; however, Healthcare Cost and Utilization Project quality control procedures are routinely performed to confirm that NIS data values are valid, consistent, and reliable (40). Third, unmeasured confounders could not be accounted for despite our best efforts to comprehensively adjust for all clinical variables. Data regarding antiplatelet therapy in CAS patients were not available. The NIS does not include detailed information about patient clinical characteristics, such as the degree of carotid stenosis and the presence of bilateral critical carotid stenosis, frailty, coronary anatomy, angina class, heart failure class, left ventricular function, or medications. Given that currently, only patients with high surgical risk are covered by the Centers for Medicare & Medicaid Services, it could be presumed, but not proven, that the aforementioned factors are more commonly present in the CAS cohort. Furthermore, referral bias has been demonstrated in a previous analysis of the National Cardiovascular Data Registry–Carotid Artery Revascularization and Endarterectomy, with markedly different clinical characteristics in patients referred for CAS versus those referred for CEA (41). Given imbalance of patient characteristics undergoing CAS versus CEA and the small size of the CAS+CABG group, it is unlikely that the use of NIS database could control for treatment selection bias, even with propensity score matching analysis, thus limiting the comparative effectiveness of this analysis. Furthermore, given lack of validation of the coding algorithms for prior strokes/transient ischemic attacks, the NIS database cannot be used reliably to differentiate between symptomatic and asymptomatic patients. Fourth, the NIS database provides only in-hospital outcomes, and our findings do not reflect 30-day and long-term outcomes after CAS and CEA, commonly examined in randomized trials. However, given that most adverse events after carotid revascularization and CABG occur soon after the procedures, this analysis provides important insights regarding the comparative safety of CAS and CEA with concurrent CABG. Furthermore, we were unable to include post-procedural MIs as part of the primary endpoint, which are difficult to separate from pre-procedural MIs in the NIS dataset. This may contribute to a falsely lower combined major adverse event rate compared with randomized trials. The post-procedural stroke, identified by ICD-9 code 997.02, has not been validated in prior studies. Fifth, the retrospective nature and inability to perform an intention-to-treat analysis in patients undergoing staged procedures may bias the results, because patients who suffer a stroke/death with the staged carotid revascularization strategy (CEA or CAS) are unlikely to undergo subsequent CABG. Finally, cardiac biomarkers are routinely monitored in most carotid registries/trials in addition to protocol mandated, comprehensive neurological examination after CAS, which may detect more strokes post-operatively. Therefore, the ascertainment bias may further contribute to an increase in adverse outcomes seen in CAS-treated patients.
In U.S. hospitals between 2004 and 2012, combined CEA+CABG is utilized most frequently, followed by staged CEA+CABG and staged CAS+CABG strategies. The staged CAS+CABG strategy was used in higher-risk patients requiring concurrent carotid revascularization and CABG, and was associated with lower adjusted risk of mortality and higher risk of stroke. Outcomes in patients undergoing CAS before CABG have improved markedly during the 9-year course of the study. Further studies are needed to validate these findings and to examine the risks and benefits of different carotid revascularization options for high-risk patients requiring concurrent CABG.
WHAT IS KNOWN? CEA or CAS are used to decrease the risk of stroke in patients with carotid disease undergoing coronary artery bypass graft surgery. However, the optimal approach to managing coexisting severe carotid and coronary disease remains controversial.
WHAT IS NEW? In US hospitals between 2004 and 2012, combined CEA+CABG strategy is utilized most frequently, followed by staged CEA+CABG and staged CAS+CABG strategies. Staged CAS+CABG strategy was associated with lower adjusted risk of mortality and higher risk of stroke.
WHAT IS NEXT? Further studies are needed to examine the risks and benefits of different carotid revascularization options for high-risk patients requiring concurrent CABG.
This work was supported by grants from the Michael Wolk Heart Foundation and the New York Cardiac Center, Inc. The Michael Wolk Heart Foundation, Clinical Translation Science Center, and the New York Cardiac Center, Inc. had no role in the design and conduct of the study, in the collection, analysis, and interpretation of the data, or in the preparation, review, or approval of the manuscript. Dr. Feldman is a consultant/speaker bureau member for Abbott Vascular, Medtronic, and St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery bypass graft surgery
- carotid artery stenting
- carotid endarterectomy
- confidence interval
- International Classification of Diseases-Ninth Revision-Clinical Modification
- myocardial infarction
- Nationwide Inpatient Sample
- open-heart surgery
- odds ratio
- percutaneous coronary intervention
- Received September 30, 2016.
- Revision received November 15, 2016.
- Accepted November 17, 2016.
- American College of Cardiology Foundation
- Cutlip D.E.,
- Pinto D.S.
- Barrera J.G.,
- Rojas K.E.,
- Balestrini C.,
- et al.
- Shishehbor M.H.,
- Venkatachalam S.,
- Sun Z.,
- et al.
- Kassaian S.E.,
- Abbasi K.,
- Hakki Kazazi E.,
- et al.
- Versaci F.,
- Reimers B.,
- Del Giudice C.,
- et al.
- Mićović S.,
- Bošković S.,
- Sagić D.,
- et al.
- Chiariello L.,
- Nardi P.,
- Pellegrino A.,
- et al.
- ↵Healthcare Cost and Utilization Project (HCUP). HCUP Nationwide Inpatient Sample (NIS). 2003-2011. Rockville, MD: Agency for Healthcare Research and Quality. Available at: http://www.hcup-us.ahrq.gov/nisoverview.jsp. Accessed January 5, 2016.
- Dugoff E.H.,
- Schuler M.,
- Stuart E.A.
- Kim L.K.,
- Yang D.,
- Swaminathan R.V.,
- et al.
- Chaturvedi S.,
- Sacco R.L.
- Masabni K.,
- Raza S.,
- Blackstone E.H.,
- Gornik H.L.,
- Sabik J.F.
- Van der Heyden J.,
- Suttorp M.J.,
- Bal E.T.,
- et al.
- Kramer J.,
- Abraham J.,
- Jones P.A.
- Calvet D.,
- Mas J.L.,
- Algra A.,
- et al.
- Gray W.A.,
- Rosenfield K.A.,
- Jaff M.R.,
- Chaturvedi S.,
- Peng L.,
- Verta P.,
- CAPTURE 2 Investigators and Executive Committee
- ↵Healthcare Cost and Utilization Project. HCUP Quality Control Procedures. Available at: http://www.hcup-us.ahrq.gov/db/quality.jsp. Accessed January 5, 2016.
- Longmore R.B.,
- Yeh R.W.,
- Kennedy K.F.,
- et al.