Abstract Background and aims Rapid interhospital transfer is critical for patients with large vessel occlusion (LVO) presenting to primary stroke centers (PSCs). Although MRI provides high diagnostic certainty, its routine use prior to transfer may compromise workflow efficiency. We evaluated how combined CT and MRI protocols impact transfer delays and outcomes. Methods We analyzed 209 LVO patients transferred from PSCs to three thrombectomy-capable stroke centers (TCSC) (2019–2022). Patients were stratified into Rapid (156 min, n=104) and Delayed (≥156 min, n=105) groups based on the median PSC imaging-to-TCSC arrival time. Multivariable logistic regression identified independent predictors of delay among imaging protocols (CT-only, MRI-only, CT+MRI). Results The median transfer time was more than double in the Delayed group (216 min IQR 180–324) compared to the Rapid group (104.5 min IQR 73–126; P0.001). The CT+MRI protocol was a strong independent predictor of delay (adjusted OR 3.56; 95% CI 1.67–7.59; P=0.001), whereas the MRI-only protocol was not associated with significant delays (P=0.54). Three-month functional outcomes did not differ between groups (P=0.46). Conclusions Redundant CT+MRI evaluation at PSCs is a major modifiable driver of interhospital transfer delays, doubling the time to TCSC arrival. The comparable outcomes despite significant delays may suggest a survivorship bias (late window paradox), wherein the delayed group primarily comprised slow progressors with robust collaterals who could tolerate the workflow inefficiency, while fast progressors may have been excluded from thrombectomy. To prevent avoidable exclusion of vulnerable patients, systemic implementation of CT-first, non-redundant transfer protocols is mostly warranted. Conflict of interest Nothing to disclose
Kim et al. (Fri,) studied this question.