Fig. 5: Comparison of outcome predictions when complementing demographics and comorbidities with biomarker levels. | Cell Death & Differentiation

Fig. 5: Comparison of outcome predictions when complementing demographics and comorbidities with biomarker levels.

From: Potential of biomarker-based enrichment strategies to identify critically ill patients for emerging cell death interventions

Fig. 5

A ROC curves for survival prediction comparing two models: one including demographics data and comorbidities (red), and an optimized model with additional variables, specifically maximal values during the first three days after ICU admission of a subset of biomarkers (green). B ROC curves for acute kidney injury (AKI) prediction comparing a model including demographics data and comorbidities (red), and an optimized model with additional variables, specifically the level on the first day in the ICU of a subset of biomarkers (green). C ROC curves for septic shock prediction comparing a model including demographics data and comorbidities (red), and an optimized model with additional variables, specifically the level on the first day in the ICU of a subset of biomarkers (green). D Observed SOFA score vs. predicted SOFA score. A linear model using demographics and comorbidities (red) is compared with an optimized model with additional variables, specifically the level on the first day in the ICU of a subset of biomarkers (green).

Back to article page