Abstract WP340: Is There a Ceiling Effect For EMS Stroke Diagnostic Accuracy?

Background: The accuracy of paramedic diagnosis of stroke is essential to the American Heart Association’s Stroke Chain of Survival. The highest rate of stroke diagnostic accuracy reported in the literature is 79% (Wojner-Alexandrov et al, 2005) and resulted from continuous feedback education. We aimed to see whether an EMS Report Card intervention could achieve similar results in a rural community hospital setting. Methods: Consecutive EMS-diagnosed stroke patients were evaluated by the Stroke Team. An EMS Report Card was developed to provide feedback on all suspected stroke patients and formal education was provided based on trends in EMS reports to support improved diagnostic accuracy. Results: A total of 198 EMS diagnosed Stroke Alerts were analyzed; 41 were admitted in 2009, 83 in 2010, 74 in 2011. The rate of stroke mimic transports for EMS was 29% in 2009 (sensitivity=71%), and improved to 26% in 2010 (sensitivity=73%), and improved to 23% by 2011 (sensitivity=77%). Specificity was not calculated. Table 1 shows the frequency of stroke mimic diagnoses for this 3 year period. The number of EMS admissions treated with intravenous tPA was 8 for 2009 (door to bolus time 55.6 minutes + 21.7, median 60.5 minutes, range 25-84 minutes), 22 for 2010 (door to bolus time 52.1 minutes + 14.5, median 50.5 minutes, range 20-90 minutes), and 17 for 2011 (door to bolus time 50.8 minutes + 18.3, median 51 minutes, range 28-68 minutes). The number of intravenous tPA treatments, and door to bolus time were not significantly different over the 3 year period. There was a trend towards increased tPA treated patients by EMS squad (Chi Square=23.3; p=0.078). Conclusion: Accuracy of stroke paramedic diagnosis is important. While our intervention did not demonstrate further improvement, this may indicate a ceiling effect for paramedic stroke diagnostic accuracy above 77-79% that should be considered acceptable.