Timing of Carotid Revascularization Procedures After Ischemic Stroke
Background and Purpose—In 2006, the American Heart Association recommended that carotid revascularization generally occurs within 2 weeks of stroke based on data from 2 trials of carotid endarterectomy (CEA). We aimed to determine whether the time between stroke and CEA or carotid artery stenting (CAS) has decreased and whether the proportion of procedures occurring within 14 days has increased.
Methods—Using validated International Classification of Diseases, Ninth Revision, Clinical Modification codes and administrative claims data from nonfederal hospitals in CA, FL, and NY, we identified patients with ischemic stroke who underwent CEA or CAS within 90 days of an ischemic stroke from 2005 to 2013. Our outcomes were the number of days between stroke and CEA/CAS and the proportion of patients undergoing CEA/CAS within the recommended 14-day period. We assessed temporal trends using nonparametric correlation, the χ2 test for trend, and logistic regression.
Results—We identified 16 298 patients with ischemic stroke who underwent CEA/CAS within 90 days. The time from stroke to CEA/CAS decreased from 25 days (interquartile range, 5–48 days) in 2005 to 6 days (interquartile range, 3–17 days) in 2013 (P<0.001). The proportion of patients who underwent CEA/CAS within 14 days of stroke increased from 40% (95% confidence interval, 37%–43%) in 2005 to 73% (95% confidence interval, 71%–76%) in 2013 (P<0.001). These temporal trends remained significant after adjustment for patient demographics and comorbidities.
Conclusions—Since 2005, revascularization for symptomatic carotid disease has been occurring progressively sooner after ischemic stroke.
Randomized controlled trials have shown that carotid endarterectomy (CEA) and carotid artery stenting (CAS) reduce the risk of stroke in patients with symptomatic carotid stenosis.1–3 A pooled analysis of these trials concluded that surgical revascularization is most beneficial when it occurs within 14 days of a stroke.4 As a result, the American Heart Association recommended in 2006 that revascularization for symptomatic carotid artery stenosis should ideally occur within 14 days of an ischemic event.5 Using a large, population-based cohort, we aimed to determine whether the time between stroke and carotid revascularization and the proportion of patients undergoing these procedures within 14 days has changed since the release of these guidelines.
We performed a population-based cohort study using administrative claims data from the Healthcare Cost and Utilization Project on all discharges from nonfederal acute care hospitals in CA from 2005 to 2011, FL from 2005 to 2013, and NY from 2006 to 2013. Further details about the data can be found in the online-only Data Supplement.
We identified all patients hospitalized with ischemic stroke using a validated International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code algorithm (online-only Data Supplement) that previous investigations have shown has a sensitivity of 88% and specificity of 86%.6 We used validated International Classification of Diseases, Ninth Revision, Clinical Modification procedure codes to identify patients who underwent CEA or CAS (online-only Data Supplement for details of these procedure codes). We excluded any patients who did not undergo one of these procedures within 90 days of admission for their index stroke. When an index stroke and a carotid revascularization procedure occurred during the same hospitalization, we included only those cases where the stroke was coded as present on admission, so as to exclude patients who were previously asymptomatic and then suffered a periprocedural stroke.
Our primary outcome was the time between index stroke occurrence and CEA/CAS, calculated as the number of days between the day of admission and the procedure if both occurred during the same hospitalization, or the number of days between initial admission and subsequent procedure if they occurred on separate hospitalizations.
To account for changes in patient characteristics that might have affected trends in the timing of revascularization, we extracted data on age, sex, race, insurance status, state of residence, the Elixhauser Comorbidity Index, and vascular risk factors and comorbidities.
We assessed trends using the χ2 test, ordinal logistic regression, and a modified form of multiple logistic regression. We performed several sensitivity analyses to test the robustness of our findings (online-only Data Supplement).
We identified 16 298 patients who underwent CEA/CAS within 90 days of an ischemic stroke. Clinical characteristics were generally similar between patients who had carotid revascularization procedures within 14 days and those who did not (Table), except for certain comorbidities that were associated with more timely revascularization (online-only Data Supplement).
The median time from stroke to CEA/CAS progressively decreased from 25 days (interquartile range, 5–48 days) in 2005 to 6 days (interquartile range, 3–17 days) in 2013 (P<0.001; Figure 1). This decrease in the time between stroke and revascularization was significant even after adjustment for patient characteristics and state of residence (global odds ratio per year, 0.88; 95% confidence interval [CI], 0.87–0.89).
The proportion of patients who underwent CEA/CAS within 2 weeks of stroke for the entire period from 2005 to 2013 was 58% (95% CI, 57%–59%). This proportion increased from 40% (95% CI, 37%–43%) in 2005 to 73% (95% CI, 71%–76%) in 2013 (P<0.001; Figure 2). These temporal trends remained significant after adjustment for patient demographics, state of residence, and comorbidities (risk ratio per year, 1.06; 95% CI, 1.05–1.07).
In a demographically heterogeneous, population-based cohort of patients who underwent carotid revascularization after ischemic stroke, we found that the time between stroke and revascularization decreased substantially from 2005 to 2013 in CA, FL, and NY. During this period, the median time from stroke to carotid revascularization was cut by 76%; in addition, nearly three fourths of patients in 2013 received CEA or CAS in the 14-day time period recommended by the American Heart Association, nearly twice as often as in 2005. Although this trend is encouraging, efforts to further promote earlier revascularization in the United States may still be beneficial because there remains a high risk of recurrent stroke even within the recommended 14-day window.7
Our study has several noteworthy limitations. First, because we relied on International Classification of Diseases, Ninth Revision, Clinical Modification codes and administrative claims data to identify patients, there may have been some misclassification of our exposure and outcome. However, any such misclassification error would likely be nondifferential and therefore not significantly affect the robust temporal trend we found. Furthermore, we used previously validated algorithms for ischemic stroke, and the codes we used for CEA/CAS seem to be highly reliable. Importantly, however, our study was specifically limited to patients who underwent revascularization after an ischemic stroke. Second, by using the day of admission for the index stroke as the day of stroke onset, we likely underestimated the time interval to carotid revascularization in some patients. Given the trends we found in our study, it is possible that the effects of such underestimation may have become progressively reduced because these same trends likely also led to earlier referral and admission for revascularization after symptom onset. Third, our results may not be generalizable nationwide, though CA, FL, and NY comprise about 25% of the US population and contain diverse patient populations and practice settings. Fourth, we lacked data on certain characteristics that can affect the timing of carotid revascularization, such as stroke severity and size, the degree and morphology of culprit carotid stenosis (including laterality relative to previous stroke), and the experience and clinical volume of treating physicians and hospitals. As a result, there may be factors other than the American Heart Association recommendations contributing to the temporal changes we found in our study. Finally, we were unable to assess the relationship between revascularization timing and recurrent stroke for several reasons: because our data did not allow identification of recurrent strokes within the same hospitalization, because we could not distinguish whether subsequent hospitalizations with a diagnosis code for stroke represented recurrent events or simply carryover of chronic codes, and because any such analyses would be limited by immortal time bias given that patients have to survive longer to be able to undergo a delayed revascularization procedure.
Since 2006 when the American Heart Association released its recommendation to perform carotid revascularization procedures within 14 days of a stroke, progressively more of these procedures are being performed in a relatively timely manner among patients who present with ischemic stroke in CA, FL, and NY.
Sources of Funding
Dr Kamel is supported by National Institutes of Health (NIH) grants K23NS082367 and R01NS097443, and the Michael Goldberg Research Fund. Dr Gupta is supported by NIH grant KL2TR000458. Dr Navi is supported by NIH grant K23NS091395 and the Florence Gould Endowment for Discovery in Stroke. Dr Gialdini receives funding from the Feil Family Foundation.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.015766/-/DC1.
- Received October 15, 2016.
- Revision received October 15, 2016.
- Accepted November 8, 2016.
- © 2016 American Heart Association, Inc.
- Sacco RL,
- Adams R,
- Albers G,
- Alberts MJ,
- Benavente O,
- Furie K,
- et al
- Tirschwell DL,
- Longstreth WT Jr
- Johansson E,
- Cuadrado-Godia E,
- Hayden D,
- Bjellerup J,
- Ois A,
- Roquer J,
- et al