Table 3

Classification task

GP (4 h)

EuroSCORE (4 h)

EuroSCORE vs. GP

Nurses (6 h)

Nurses vs. GP

ICU physicians (6 h)

ICU physicians vs. GP


Validation cohort, n = 499

aROC

0.758

0.726

p = 0.286

X

X

X

X

Brier Score

0.179

0.324

p < 0.001

X

X

X

X

Brier Score Scaled

11%

0%

p < 0.001

X

X

X

X

Hosmer Lemeshow p-value

0.382

< 0.001

X

X

X

X

X

Nurses answer ≤ 6 h, n = 396

aROC

0.769

0.726

p = 0.124

0.695

p = 0.018

X

X

Brier Score

0.177

0.326

p < 0.001

0.245

p < 0.001

X

X

Brier Score Scaled

13%

0%

p < 0.001

1.35%

p < 0.001

X

X

Hosmer Lemeshow p-value

0.405

< 0.001

X

< 0.001

X

X

X

Physicians answer ≤ 6 h, n = 159

aROC

0.777

0.726

p = 0.334

X

X

0.758

p = 0.719

Brier Score

0.166

0.328

p < 0.001

X

X

0.216

p = 0.055

Brier Score Scaled

14.2%

0%

p < 0.001

X

X

12.5%

p < 0.001

Hosmer Lemeshow p-value

0.696

< 0.001

X

X

X

< 0.001

X


The interpretation of the different validation measures can be found in the methodology section.

Meyfroidt et al. BMC Medical Informatics and Decision Making 2011 11:64   doi:10.1186/1472-6947-11-64

Open Data