Abstract TP40: Magnetic Resonance Versus Computed Tomography As Preferred Screening Of Suspected Acute Stroke: A Retrospective Comparative Cohort Analysis

Background: In published cohorts, MRI screened patients tend to have better outcomes than CT-screened, even though MRI takes longer to acquire and may delay initiation of thrombolysis. However, overall differences in stroke workflow and clinical outcomes in stroke centers that prefer MRI over CT have not been reported. Methods: Between 2019 and 2021 three stroke centers in our network, switched their code-stroke imaging protocol from CT-first preferred to MR-first preferred. Under the MR-first protocol, CT was performed for MRI contraindications, patient instability, or MR not immediately available during a code stroke. Data from one year prior to the switch (CT-first) was compared to the year following the switch (MR-first). Patient demographics, quality stroke metrics and outcomes were obtained from a prospectively maintained patient database. Results: A total of 1027 (CT-first) and 1056 (MR-first) cases were evaluated (Table). Ischemic stroke was more frequent and TIA less frequent under the MR-first protocol. Cross-over scanning occurred more frequently under the MR-first protocol (p<0.0001). Rate of thrombolysis was 11.8% of all cases under CT-first and 13.3% under MR-first. Door-to-treatment for thrombolytic and thrombectomy was longer under MR-first but fewer stroke mimics were treated under MR-first (0.7% vs 9.2%, p=0.0014). Clinical outcomes did not differ for thrombolysis, but discharge disposition to home vs hospice or death after thrombectomy was favored under the MR-first protocol (p =0.0176). Conclusion: Under an MR-first stroke protocol there was a greater rate of cross-over imaging and delayed workflow and treatment times, but less frequent thrombolytic treatment of stroke mimics and better outcomes in thrombectomy patients. Workflow metrics and outcomes for patients as imaged will be presented at the conference.