A Retrospective Propensity‐Matched Analysis of Functional Outcomes for Carotid Puncture for Endovascular Thrombectomy

Abstract

Background Endovascular thrombectomy is a well‐established therapy for patients with large‐vessel occlusion stroke but its success largely depends on minimizing time to reperfusion. Direct carotid puncture (DCP) is a technique that avoids difficult aortic arch anatomy, which may cause intraprocedural delays, but because DCP is typically used as rescue access, the poorer outcomes that some have observed may reflect differences in patients chosen for DCP rather than the impact of DCP itself. We performed a propensity score–matched case‐control study to try to address potential confounding by indication to better evaluate whether the DCP procedure itself is responsible for differences in functional outcomes. Methods We identified all DCP cases and non‐DCP endovascular thrombectomy controls that were performed under general anesthesia at 2 academic medical centers from 2015 to 2021. Baseline characteristics, workflow time metrics, aortic arch measurements, and 90‐day modified Rankin scale were abstracted from the electronic health record. We then matched patients with controls based on propensity scores, then performed a linear regression to evaluate the relationship between DCP and 90‐day modified Rankin scale. Results We identified 13 patients with DCP and 67 control patients for our final analysis. The mean age for DCP was 85.4 years compared with 73.8 years in the control group ( P ≤0.001). We used sex, age, National Institutes of Health Stroke Scale, and last seen normal‐to‐groin times to generate a propensity score to estimate each patient's likelihood of receiving DCP. After matching, we found no significant difference in mean modified Rankin scale associated with DCP (β=−0.231, standard error 0.545; P =0.676). Conclusions After propensity‐score matching, we found that DCP was not associated with a difference in functional outcomes of endovascular thrombectomy for large‐vessel occlusion stroke compared with non‐DCP access. Advances in algorithms to identify patients highly likely to require DCP and in devices for safer carotid closure are needed before wider adoption of DCP.