Background: Recent studies have focused on improving prehospital stroke assessment tools, but specificity and sensitivity have been insufficient to reliably detect stroke in the prehospital setting. To assess the ability of emergency medical services (EMS) personnel to identify acute stroke in the field, we compared EMS stroke recognition with receiving medical center discharge diagnosis from a large community-based stroke dataset, the San Diego County Stroke Registry. The registry was founded in 2010 after San Diego County established 16 diverse stroke receiving centers. EMS uses the Cincinnati Prehospital Stroke Scale for screening.
Methods: We captured all EMS transports in San Diego County from 2013 to 2015. Accuracy of stroke detection by the EMS providers was analyzed by: a) coding of stroke related provider impression (PI) by EMS; b) “stroke” recorded as the reason the transport destination was selected. All patients with diagnosis stroke on hospital discharge were considered confirmed stroke, and separated by: Acute Ischemic Stroke (AIS), Subarachnoid Hemorrhage (SAH), Transient Ischemic Attack (TIA) or Intracranial Hemorrhage (ICH).
Results: Between 2013 and 2015, we identified 577,643 EMS transports, 7,425 (1.3%) were diagnosed as stroke by the treating facility (68.2% AIS, 14.4% TIA, 13.6% ICH, and 3.9% SAH). a) Of these 7,418 (99.9%) had a coded PI. Stroke related PI was positive in 53.8% (AIS: 55.9%; SAH: 18.1%; TIA: 60.5%; ICH 46.4%). b) A recorded reason for destination was found in 6,813 (91.8%) of all stroke patients. Stroke was the coded reason in 16.4% (AIS 16.4%; SAH 8.0%, TIA 18.0%, ICH 17.3%).
Conclusion: In a large community EMS system, using routine stroke screening, 53.8% of all stroke patients were identified by EMS. Stroke was the coded reason for the selected destination in only 16.4% of EMS transported stroke cases. This emphasizes the need for better prehospital stroke detection to improve triage and direct patient care.