Fig. 2

Receiver operating characteristic (ROC) curve of the multiple biomarker model with fourteen peptides for prediction of clinical outcomes of HF in comparison with the BIOSTAT risk prediction model and the added value of the prediction model of fourteen peptides on top of the BIOSTAT risk prediction model. The red curve presents the BIOSTAT risk prediction model with an AUC value of 0.643 (Asymptotic 95% CI 0.530–0.757, p = 0.015) that risk scores were calculated using the online calculator available at: http://www.biostat-chf.eu (including age, HF hospitalisation last year, peripheral oedema, systolic blood pressure, NT-proBNP, haemoglobin, high-density lipoprotein, serum sodium and beta-blocker use at baseline). The green curve describes the multiple biomarker model with the fourteen peptides with an AUC of 0.817 (Asymptotic 95% CI 0.734–0.900, p = 0.0005). The blue curve displays the prediction model of fourteen peptides on top of the BIOSTAT risk prediction model with an AUC of 0.823 (Asymptotic 95% CI 0.743–0.904, p = 0.0005)