Abstract Body: Background: Previous findings have shown high mobile stroke unit (MSU) call density scattered across a coverage area. We sought to differentiate the geographic pattern of MSU calls carrying an accurate diagnosis of an acute cerebrovascular condition using a geospatial software analysis.
Methods: The Mobile Stroke Treatment Unit of metropolitan Columbus, Ohio, provides emergency stroke care coverage for 325 square miles spanning 47 zip codes. Our MSU is operational from 0700-1900 and is dispatched for all 911 calls designated as probable or definite strokes. We retrospectively reviewed all MSU calls from June 2019 to December 2023. We retrospectively reviewed charts to confirm the final diagnosis and included only patients with an acute cerebrovascular condition. The calls were organized into 3 time epochs (0700-1100, 1101-1500, 1501-1900). Each call address was geocoded and entered into the ESRI/ArcGIS Online geospatial analysis software. Call addresses were then clustered into hexagonal grids with a distance band of 1 km for hotspot analysis. Statistically significant hotspots were qualitatively defined as clustered grids with high activity and quantitatively defined as having a Gi Bin score > 2, which correlates to statistical confidence > 95%.
Resuts: We received 7442 calls during our study period. Of these calls, 738 were retrospectively confirmed as a stroke. From 0700-1100, 4 regions were considered hotpots (Blue). From 1101-1500, 3 regions were considered hotspots (Red). From 1501-1900, 1 region was considered a hotspot (Purple). Two particular regions showed proximity over consecutive time epochs.
Conclusion: Our study showed that MSU actual stroke call density may have less variation compared to previous undifferentiated studies. Geographic analysis shows relative clustering of these hotspots, which can aid in strategic MSU deployment.