Abstract DP387: Accuracy of Mobile Stroke Treatment Unit Cancellations by On Scene Emergency Medical Services

Intro: Mobile stroke treatment units (MSTUs) enable faster acute stroke treatment compared to conventional Emergency Medical Services (EMS) transportation. However, they can only be activated for one patient at a time. High cancellation rates of a MSTU may reflect effective triage by on scene EMS or missed treatment opportunities. Our group noticed a high cancellation rate therefore, our group wanted to evaluate the accuracy of prehospital MSTU activation as well as rate and reasons for cancellations en route to ensure best use of the MSTU. Methods: We retrospectively reviewed all MSTU activations from 6/3/25-8/3/25 from a single site (ID #2819). Data collected included MSTU cancellation reason as reported by on scene EMS, patient demographics, final discharge diagnosis, whether an Emergency Department (ED) stroke alert was activated after MSTU cancellation, and if acute stroke therapy was administered at the hospital. Dispatch software Priority Dispatch triages suspected strokes using BEFAST criteria, with MSTU to EMS communication en route to confirm symptoms. Accuracy metrics (sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV)) were calculated for identifying treatment-eligible strokes. Results: Over the study period, 208 MSTU activations occured, of which 149 (71%) were cancelled prior to stroke team evaluation. Of these, 57 were cancelled by on scene EMS, with complete outcome data available for 25 patients (Figure 1). No patients (0/25) had treatment-eligible stroke, though 3 were later diagnosed with ischemic stroke on MRI (all outside treatment window or with resolved deficits). In the non-cancelled group, 59 admitted patients included 13 treatment-eligible strokes (10 TNK, 3 EVT individual treatments) and 46 non-eligible cases. Sensitivity was 100% (95%CI 77.2-100), Specificity was 35.2% (95% CI, 25.1-46.8), PPV was 22.0% (95% CI, 13.4-34.1) and NPV was 100% (95%CI, 86.7-100). Discussion: In this early experience, EMS and dispatch decision to cancel MSTU activation achieved high prehospital accuracy in sensitivity and NPV effectively excluding patients unlikely to benefit from acute stroke intervention. This preserves the MSTU’s availability for potentially treatable strokes. While specificity and PPV were lower, possibly due to high stroke mimics admission, BEFAST based triage, coupled with direct MSTU to on scene EMS communication, prevented unnecessary deployment. Larger studies are warranted to validate accuracy.