The identification of acute stroke: an analysis of emergency calls.

Background: Accurate dispatch of emergency medical services at the onset of acute stroke is vital in expediting assessment and treatment. We examined the relationship between callers' description of potential stroke symptoms to the emergency medical dispatcher and the subsequent classification and prioritisation of emergency medical services response.

Aim: To identify key 'indicator' words used by people making emergency calls for suspected stroke, comparing these with the subsequent category of response given by the emergency medical dispatcher.

Method: A retrospective chart review (hospital and emergency medical services) in North West England (October 1, 2006 to September 30, 2007) identified digitally recorded emergency medical services calls, which related to patients who had a diagnosis of suspected stroke at some point on the stroke pathway (from the emergency medical services call taker through to final medical diagnosis). Using content analysis, words used to describe stroke by the caller were recorded. A second researcher independently followed the same procedure in order to produce a list of 'indicator' words. Description of stroke-specific and nonstroke-specific problems reported by the caller was compared with subsequent emergency medical services dispatch coding and demographic features.

Results: Six hundred forty-three calls were made to emergency medical services of which 592 (92%) had complete emergency medical services and hospital data. The majority of callers were female (67%) and family members (55%). The most frequently reported problems first said by callers to the emergency medical dispatcher were collapse or fall (26%) and stroke (25%). Callers who identified that the patient was having a stroke were correct in 89% of cases. Calls were dispatched as stroke in 45% of cases, of which 83% had confirmed stroke. Of the first reported problems, Face Arm Speech Test stroke symptoms were mentioned in less than 5% of calls, with speech problems being the most common symptom. No callers mentioned all three Face Arm Speech Test symptoms.

Conclusion: Callers who contacted emergency medical services for suspected stroke and said stroke as the first reported problem were often correct. Calls categorised as stroke by the emergency medical dispatcher were commonly confirmed as stroke in the hospital. Speech problems were the most commonly reported element of the Face Arm Speech Test test to be reported by callers. Recognition of possible stroke diagnosis in fall and other presentations should be considered by emergency medical dispatchers. Further development and training are needed in the community to improve prehospital stroke recognition in order to expedite hyperacute stroke care.