Table 1 Area under the precision-recall curves for various predictions

From: Reply: metrics to assess machine learning models

 

Hospital A

Hospital B

Inpatient Mortality, 24 h after admission, AUPRC

  

Deep learning model

0.41 (0.34–0.48)

0.42 (0.37–0.48)

Baseline (aEWS) model

0.24 (0.18–0.32)

0.25 (0.20–0.30)

Full feature, enhanced baseline

0.32 (0.25–0.39)

0.29 (0.25–0.35)

30 day unplanned readmission at discharge, AUPRC

  

Deep Learning Model

0.28 (0.26–0.30)

0.37 (0.36–0.40)

Baseline (mHospital) at discharge

0.20 (0.19–0.22)

0.28 (0.26–0.29)

Full feature, enhanced baseline

0.25 (0.24–0.28)

0.34 (0.33–0.36)

Length of stay at least 7 days, AUPRC

  

Deep learning model

0.67 (0.65–0.69)

0.66 (0.65–0.68)

Baseline (Liu) model

0.47 (0.44–0.49)

0.48 (0.46–0.50)

Full feature, enhanced baseline

0.63 (0.61–0.66)

0.63 (0.61–0.64)