Skip to main content

Table 4 The outcome prevalence (Pre.), Area under the curve (AUC), and the Brier scores evaluated for the Null model, the Logistic prediction and the discrete event simulation

From: Prediction of future labour market outcome in a cohort of long-term sick- listed Danes

Macro-economic period

State

Null model

Logistic regression

Discrete event simulation

  

Pre.

AUC

Brier score

Pre

AUC

Brier score

Pre.

AUC

Brier Score

Growth

W

44.1%

50.0%

0.2466

43.7%

61.9%

0.2347

43.4%

62.2%

0.2344

 

SA

7.8%

50.0%

0.0719

7.8%

64.3%

0.0699

6.6%

64.0%

0.0704

 

LTS

4.5%

50.0%

0.0434

4.6%

58.7%

0.0432

5.6%

57.4%

0.0434

 

U

5.3%

50.0%

0.0502

5.5%

72.1%

0.0477

5.0%

71.6%

0.0477

 

D

0.3%

50.0%

0.0034

0.3%

73.4%

0.0034

0.7%

73.8%

0.0034

Recession

W

41.9%

50.0%

0.2435

42.5%

61.0%

0.2333

42.5%

61.5%

0.2321

 

SA

9.0%

50.0%

0.0819

8.9%

61.9%

0.0806

6.7%

61.2%

0.0813

 

LTS

4.0%

50.0%

0.0383

4.0%

58.6%

0.0381

4.7%

56.2%

0.0383

 

U

7.5%

50.0%

0.0695

7.4%

71.0%

0.0664

7.8%

70.4%

0.0665

 

D

0.1%

50.0%

0.0014

0.2%

76.6%

0.0014

0.5%

75.1%

0.0015

  1. The estimates are shown for each of the macroeconomic periods and the prediction of being in each of the six states work (W), sickness absence (SA), long-term sick-listing (LTS), unemployment (U), and disability pension (D).