Background: In patients with TIA/stroke, early assessment is critical.

Aim: To describe patients who were not directly transported to hospital by ambulance after prehospital assessment.

Methods: Patients hospitalised with TIA/stroke in Gothenburg, Data were obtained from the EMS and hospital case record system.

Results: There were 7,812 patients with TIA/stroke, of which 4,853 (62%) were candidates for EMS transport. Among them, 176 (3.6%) were not directly transported to hospital by ambulance. In 45% of them, delay from symptom onset to calling for EMS was ≤ 24 hours. On EMS arrival, common symptom was dizziness (28%), followed by weakness in arm or leg (21%), loss of sensibility (13%), speech disturbances (7%) and facial numbness (4%). The modified National Institute of Health Stroke Score (mNIHSS) was 0 in 80% and > 1 in two per cent. The NIHSS at the emergency department was 1-4 in 39% and 5-15 in six per cent. The EMS clinician made the decision not to transport the patient to hospital by the EMS in 84%, the dispatcher in 12% and the patient or relatives in four per cent. Patients were involved in the decision in 51%. Final diagnosis was stroke in 74% and the proportion who were independent in normal daily activities at hospital discharge decreased by 15% compared with before event.

Conclusion: About 3-4% of patients with TIA/stroke were not directly transported to hospital by EMS after prehospital assessment. The most common symptom was dizziness. Decision-support tools for EMS to identify time-sensitive conditions are required.