Skip to main content

Table 2 Performance metrics for the four base learners and the stacked ensemble. Median and 95%-confidence intervals are shown for the predictions in the test set in random repeated subsampling framework (see Methods)

From: Decision curve analysis confirms higher clinical utility of multi-domain versus single-domain prediction models in patients with open abdomen treatment for peritonitis

Method

Domain

AUROC

AUPRC

Brier Score

Logistic Regression

Demographics

0.64 (95%-CI: 0.58–0.71)

0.39 (95%-CI: 0.32–0.48)

0.19 (95%-CI: 0.18–0.20)

Physiological

0.81 (95%-CI: 0.75–0.86)

0.66 (95%-CI: 0.57–0.75)

0.14 (95%-CI: 0.12–0.16)

Surgical

0.70 (95%-CI: 0.63–0.76)

0.60 (95%-CI: 0.50–0.67)

0.15 (95%-CI: 0.14–0.18)

Multidomain

0.87 (95%-CI: 0.80–0.91)

0.75 (95%-CI: 0.65–0.84)

0.12 (95%-CI: 0.09–0.14)

Elastic Net

Demographics

0.65 (95%-CI: 0.58–0.72)

0.40 (95%-CI: 0.31–0.49)

0.19 (95%-CI: 0.18–0.19)

Physiological

0.81 (95%-CI: 0.76–0.86)

0.66 (95%-CI: 0.57–0.75)

0.14 (95%-CI: 0.12–0.16)

Surgical

0.70 (95%-CI: 0.63–0.76)

0.59 (95%-CI: 0.49–0.67)

0.15 (95%-CI: 0.14–0.17)

Multidomain

0.86 (95%-CI: 0.81–0.90)

0.74 (95%-CI: 0.63–0.81)

0.12 (95%-CI: 0.10–0.15)

Random Forest

Demographics

0.55 (95%-CI: 0.50–0.61)

0.30 (95%-CI: 0.26–0.38)

0.23 (95%-CI: 0.20–0.25)

Physiological

0.77 (95%-CI: 0.71–0.83)

0.60 (95%-CI: 0.50–0.70)

0.16 (95%-CI: 0.13–0.18)

Surgical

0.68 (95%-CI: 0.62–0.75)

0.57 (95%-CI: 0.49–0.67)

0.16 (95%-CI: 0.14–0.18)

Multidomain

0.85 (95%-CI: 0.79–0.89)

0.72 (95%-CI: 0.64–0.80)

0.12 (95%-CI: 0.11–0.15)

Gradient Boosting Machine

Demographics

0.61 (95%-CI: 0.53–0.68)

0.33 (95%-CI: 0.28–0.40)

0.19 (95%-CI: 0.18–0.22)

Physiological

0.79 (95%-CI: 0.73–0.85)

0.60 (95%-CI: 0.49–0.73)

0.15 (95%-CI: 0.13–0.18)

Surgical

0.70 (95%-CI: 0.65–0.77)

0.60 (95%-CI: 0.51–0.68)

0.15 (95%-CI: 0.13–0.17)

Multidomain

0.85 (95%-CI: 0.79–0.89)

0.71 (95%-CI: 0.60–0.80)

0.13 (95%-CI: 0.11–0.16)

Stacked Ensemble

Demographics

0.58 (95%-CI: 0.51–0.64)

0.32 (95%-CI: 0.26–0.40)

0.21 (95%-CI: 0.19–0.24)

Physiological

0.80 (95%-CI: 0.72–0.85)

0.62 (95%-CI: 0.54–0.72)

0.15 (95%-CI: 0.13–0.18)

Surgical

0.70 (95%-CI: 0.64–0.77)

0.58 (95%-CI: 0.48–0.66)

0.15 (95%-CI: 0.14–0.18)

Multidomain

0.86 (95%-CI: 0.80–0.90)

0.74 (95%-CI: 0.63–0.82)

0.12 (95%-CI: 0.10–0.16)

  1. AUROC: Area Under the Receiver Operating Characteristic Curve
  2. AUPRC: Area Under the Precision Recall Curve
Back to article page

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

Contact us