Artificial intelligence in Emergency Medical Services dispatching: assessing the potential impact of an automatic speech recognition software on stroke detection taking the Capital Region of Denmark as case in point

Table 3 Time-to-call in EMS of the Capital Region of Denmark 2016–2018

Category	Time-to-call
Category	Within 4.5 h (%)	Within 24 h (%)	Mean (Min.; Q1; Median; Q3; Max.)
Stroke detected	75.70	94.93	5.65 (0.00045; 0.312; 1.233; 4.296; 165.861)
Stroke relevant criteria + no “A” response	49.00	80.42	16.22 (0.0053; 1.019; 4.721; 18.914; 167.646)
Stroke non relevant criteria	60.93	85.06	14.07 (0.0016; 0.665; 2.451; 11.289; 167.377)
Missing criteria	55.80	83.62	14.16 (0.011; 0.995; 3.391; 12.514; 167.456)

Time of onset is based on the patients recall of symptom onset and thus might be no exact time of onset

ISSN: 1757-7241