Skip to main content
Fig. 2 | Critical Care

Fig. 2

From: Predicting neurological outcome after out-of-hospital cardiac arrest with cumulative information; development and internal validation of an artificial neural network algorithm

Fig. 2

Prognostic performance. Receiver operating characteristic curves (ROC) and areas under the curves (AUROC) for all nine models including 95% confidence intervals. The figures present the capability of the models to discriminate patients with poor outcome (Cerebral Performance Category [CPC] 3–5) and those with good outcome (CPC 1–2) at six-months on an independent test set. The rows represent the different levels of added biomarkers to the available standard intensive care unit (ICU) observations from the TTM-trial: None (level A), clinically accessible biomarkers (level B), and research-grade biomarkers (level C). The columns represent the timeline after admission to the ICU. For each ROC curve, the 95% CI was calculated for specificity at different levels of sensitivity and displayed as a blue 95 CI band. TTM, target temperature management. CI, confidence interval

Back to article page

Critical Care

ISSN: 1364-8535