Effectiveness of Carotid Intervention for Long-term Stroke Prevention: A Cohort Study in a Large Integrated Health System

The results of prospective trials comparing the effectiveness of carotid intervention (carotid endarterectomy [CEA] or carotid artery stenting [CAS]) versus standard medical therapy without intervention for long-term stroke prevention in patients with asymptomatic carotid stenosis (ACS) will not be available for several years. However, this question can also be explored using modern causal inference methods combined with machine learning to emulate a two-arm randomized experiment using observational data on a large and contemporary cohort of patients. This was a retrospective cohort study of patients with 70% to 99% ACS (no neurologic symptom within 6 months) with no prior ipsilateral intervention. Patients were followed from their first imaging study between 2008 and 2012 through the earliest of ipsilateral ischemic stroke (event of interest), death, health plan disenrollment or December 2017. We compared the event-free survival over 96 months (8 years) between two treatment strategies: intervention (CEA or CAS without concomitant coronary artery bypass grafting) within 12 months from cohort entry versus no intervention during the entire follow-up. To account for both baseline and time-dependent confounding, we used inverse probability weighting estimation to derive adjusted hazard ratios and cumulative risk differences based on two logistic marginal structural models for counterfactual hazards. Propensity scores were data-adaptively estimated using super learning. The cohort included 3824 eligible patients with ACS (mean age, 74; 57.9% male; 12.3% active smokers; 21.7% diabetic), of which 1467 underwent CEA, 43 underwent CAS, and 2297 never underwent either intervention. Median follow-up was 68 months (interquartile range, 31-96 months). There were 1760 participants (46%) who died, 445 were lost to follow-up and 158 patients experienced ipsilateral ischemic stroke. The adjusted hazard ratio (0.81; 95% CI, 0.68-0.93) and cumulative risk differences for each year of follow-up showed a protective effect of intervention starting in year 2 persisting to year 8 compared with patients not receiving either intervention (Table). In this contemporary cohort study of patients with asymptomatic carotid stenosis using rigorous analytic methodology, intervention appears to have a small but statistically significant effect on stroke prevention out to 8 years. Further study is needed to appropriately select the subset of patients most likely to benefit from intervention. Given the large proportion of patients who died during follow-up, this finding also requires further study to assess the overall health impact of intervention in this cohort.TableAdjusted cumulative risk differences by year after study entry between the intervention (carotid endarterectomy [CEA] or carotid artery stenting [CAS]) and control groupsYear after study entryPoint estimate of the risk differencea95% CI1−0.2%−0.02% to 0.5%2−1.1%− 0.5% to −1.7%3−2.0%−1.1% to −2.9%4−3.0%−1.8% to −4.1%5−3.1%−1.9% to −4.4%6−2.3%−4.3% to −2.7%7−2.2%−4.4% to −0.02%8−2.4%−4.7% to −0.2%CI, Confidence interval.aA negative value represents a lower risk of ipsilateral stroke among patients who underwent a carotid intervention compared with patients who did not undergo an intervention. For instance, in year 5, carotid intervention is estimated to have decreased the risk of ipsilateral stroke by 3.1% (95% CI, 1.9%-4.4%) compared to patients who received medical therapy alone. Open table in a new tab