Abstract 105: Geographic Disparities in Emergency Medical Services Calls for Stroke in Chicago

Introduction: Geographic Information Systems (GIS) is a methodology for matching clinical and spatial metrics to inform public health intervention and policy development. We evaluated statistical hot spots for emergency medical services (EMS) calls for stroke and compared these with stroke admission rates in Chicago. Methods: In a retrospective review, we compared stroke hospitalizations (ICD-9 code 430-436) for residents of Chicago using Illinois Hospital Association CompData to EMS stroke calls between May 15, 2012 and November 30, 2012. EMS stroke calls were identified through the patient tracking system of the Chicago Fire Department which also captured the scene coordinates (latitude and longitude). Using the Getis-Ord Gi* method, we identified hot spots for EMS stroke calls. Statistical hot spots are defined as areas where there is <1% chance of the case clusters occurring by chance. We then calculated a ratio of EMS calls to stroke admissions for each zip code in Chicago to identify areas where EMS usage was disproportionately low. Results: During the study period, there were 4,135 stroke admissions and 919 EMS stroke calls. Stroke admissions were more frequent in the south, west, and northwest sides of Chicago. While hot spot data for EMS stroke calls overlapped with areas of greatest stroke density and mortality, there were some areas with disproportionately low rates of EMS calls (figure). In particular, the regions in the northwest area of the city and downtown/south loop of Chicago appear to have low (4-11%) rates of EMS calls per stroke admissions. Conclusion: Discrete areas within Chicago were identified as hot spots for EMS-treated stroke which correspond generally with regions of greatest stroke density and stroke mortality. GIS mapping of EMS stroke calls may be valuable for targeted neighborhood educational outreach, operational deployment of EMS resources, and monitoring performance improvement initiatives.