Skip to main content

Table 4 Outcome variance-analysis

From: 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

Characteristics

Stroke detected

Stroke not detected

Missing criteria

Stroke relevant criteria + “A” response

Stroke relevant criteria + no “A” response

Stroke nonrelevant criteria

N

Mean [Min.|Q1|Median|Q3|Max.]

N

Mean [Min.|Q1|Median|Q3|Max.]

N

Mean [Min.|Q1|Median|Q3|Max.]

N

Mean [Min.|Q1|Median|Q3|Max.]

Age

4773

3613

71.4 [18|63|73|81|104]

71.06 [19|62|72|81|104]

1200

588

69.87 [21|61|71|80|101]

69.94 [21|61|71|80|101]

1816

1134

71.08 [21|63|73|82|101]

69.96 [21|60|72|81|101]

1215

678

72.53 [19|63|74|83|102]

71.90 [19|62|73|83|102]

Time of day

4773

3613

12.82 [0|9|12|17|23]

12.93

[0|9|13|17|23]

1200

588

13.55 [0|10|14|17|23]

13.36

[0|10|13|17|23]

1816

1134

12.97 [0|9|13|17|23]

12.60

[0|9|12|16|23]

1215

678

13.83 [0|10|15|18|23]

13.56

[0|10|15|18|23]

Characteristics

Kruskal–Wallis test

Wilcoxon–Mann–Whitney (+ Bonferroni–Holm Correction)

Kruskal–Wallis-X2

Df

p-value

Age

29.226

13.944

3

2.00e−06

0.002983

 

Stroke Detected

Stroke relevant criteria + no “A” response

Stroke non-relevant criteria

Stroke relevant criteria + no “A” response

0.0017

0.1368

Stroke non-relevant criteria

1.000

0.8197

0.0055

1.0000

Missing criteria

0.0068

0.1551

8.7e−07

0.0052

0.0543

0.0273

Time of day

63.455

24.8

3

1.073e−13

1.7e−05

 

Stroke Detected

Stroke relevant criteria + no “A” response

Stroke non-relevant criteria

Stroke relevant criteria + no “A” response

2.8e−05

0.39660

Stroke non-relevant criteria

0.934

0.68081

0.019

0.04429

Missing criteria

7.1e−12

0.00037

0.055

0.62118

3.5e−07

2.5e−05

  1. “Italic” numbers indicate the subgroup analysis of stroke patients with a time-to-call < 4.5 h. “Bold” numbers indicate the difference to the initial analysis