The positive predictive value of an ambulance prealert for stroke and transient ischaemic attack

Abstract

Objective: Therapeutic options for ischaemic stroke, such as thrombolysis or thrombectomy, are time sensitive. Multiple innovations have been established to reduce the symptom-to-needle time. One such innovation is the prealerting of emergency department (ED) or stroke unit staff by prehospital personnel of suspected stroke patients. The diagnosis of stroke can sometimes be difficult, with stroke mimics being a recognized issue. The prealert mobilizes ED, stroke and imaging personnel, which, for a true-positive, improves door-to-needle times. However, there are a proportion of false-positive prealerts (nonstrokes) that have a significant resource activation implication. The aim of this study was to evaluate the positive predictive value of a prealert for stroke and transient ischaemic attack (TIA).

Methods: Ambulance service prealert forms for stroke and TIA collated by the ED were compared with the Scottish Stroke Audit database findings, ED electronic notes and imaging reports to establish whether the prealert was a true-positive or a false-positive.

Results: A prealert was obtained for 77 patients as query stroke/TIA. The true-positive rate was 52 and the false-positive rate was 25. The positive predictive value was 0.675. The median symptom-to-arrival time for prealerted patients was 97 min and the door-to-needle time for thrombolysis (n=17 patients) was 38 min.

Conclusion: The diagnosis of true-positive stroke can be difficult in the prehospital environment. Although prealert has been shown to improve the patient's journey in terms of door-to-thrombolysis times, we have identified that the prealert has a significant false-positive rate that has important resource allocation and activation consequences. Further analysis of this may inform paramedic training and improve protocols for information handover.