Accuracy of Pre-Hospital Identification of Stroke (P04.059)

Abstract

Objective: Determine accuracy of emergency medical service (EMS) identification of stroke. Background Activation of stroke systems by EMS may expedite time to intervention for stroke patients. Gropen et al. reported on sensitivity and impact of EMS identification of stroke. We additionally explored specificity and predictive values (PV) of EMS and emergency department (ED) evaluations. Design/Methods: We reviewed records of all patients transported by EMS to the Loyola University ED between 10/1/2010 and 6/30/2011. EMS impressions, ED physician diagnosis (DX), and neurology discharge diagnosis (DDX) were compared to determine sensitivity, specificity, and PV of EMS and ED identification of stroke. For patients with EMS impression of stroke, or DDX of stroke, additional parameters explored included: age, iv tPA administration, Cincinnati Stroke Scale (CSS), NIHSS, and time from ED arrival to neurology consult. Results: 5300 EMS patient records were included: 84 had EMS impression of stroke, with 49 true positives (TP), 35 false positives (FP), 47 false negatives (FN), and 5169 true negatives (TN); sensitivity was 51.04%, specificity 99.33%, positive PV 58.33%, and negative PV 99.10% (p Conclusions: EMS identifies stroke with high specificity, though lesser sensitivity. Pre-hospital stroke team activation may therefore help in decreasing the time to treatment. Improving the sensitivity of EMS and ED identification of stroke is needed for early identification of additional possible stroke patients. Disclosure: Dr. Wild has nothing to disclose. Dr. Kim has nothing to disclose. Dr. Hyman has nothing to disclose. Dr. Venizelos has nothing to disclose. Dr. Cichon has nothing to disclose. Dr. Schneck has nothing to disclose.