Outcomes of Stroke Thrombectomy Performed by Interventional Radiologists vs Neurointerventional Physicians.

Document Type


Publication Date



PURPOSE: To test the hypothesis that interventional radiologists (IR) have outcomes for endovascular stroke thrombectomy (EVT) similar to Neurointerventional (NI) physicians and could be used to improve availability of thrombectomy.

MATERIALS AND METHODS: Eight hospitals providing EVT performed by IR and NI in the same institution submitted sequential retrospective data limited to the era of modern devices. Good clinical outcome (90 day modified Rankin score 0-2) and successful revascularization (modified Thrombolysis in Cerebral Infarction score > 2b) were compared between specialties, adjusted for treating hospital, patient age, stroke severity, Alberta Stroke Program Early CT Score (ASPECTS), time from symptom onset to door, and clot location. Propensity score matching was used to compare outcomes. A total of 1009 patients were entered (622 treated by IR and 387 treated by NI).

RESULTS: Median stroke onset to puncture was 245 vs 253 minutes (p=.49), technically successful revascularization was 81.8% vs 82.4% (p=.81), and good clinical outcome was 45.5% vs 50.1% (p=.16), respectively. After adjusting, physician specialty was not a significant predictor of good clinical outcome (odds ratio 1.028 [95% CI 0.760-1.390]; p=.86). After matching, mRS 0-2 was 47.7% for IR and 51.1% for NI (p=0.366).

CONCLUSION: There was no significant difference in successful revascularization and good clinical outcomes between IR and NI physicians. Outcomes by IR were similar to NI outcomes from previously published trials and registries. This may be useful to address coverage and access to stroke interventions.

Publication Title

Journal of vascular and interventional radiology : JVIR


Online ahead of print