Background. The availability of a selection of biomarkers that includes information about disease risk is very important in the treatment of sickle cell disease (SCD). We used the predictiveness curve (PC), which classifies diseased individuals according to low- and high-risk thresholds, for this purpose. Our aim was to define this new statistical method and to determine the biomarkers that predict vaso-occlusive crisis (VOC) in children with SCD to guide preventive treatment.

Methods. Thirty-eight pediatric patients with SCD were included in this feasibility study. Leucocytes (WBC), C-reactive protein (CRP), interleukin-6 (IL-6), tumor necrosis factor (TNF-α), and YKL-40 were studied in patients with VOC and without VOC. The patient group with a low or high risk of VOC was assessed using the PC. Risk prediction and classification performance were evaluated using the PC and receiver operating characteristic (ROC) curve.

Results. According to the PC, patients with a high risk of VOC could be detected via TNF-α, IL-6, and WBC, and TNF-α was the best risk prediction marker (TPF = 0.67).

Conclusions. The PC provides disease risk information by comparing more than one biomarker and can thereby help clinicians determine appropriate preventive treatments. This is the first study to evaluate biomarkers to predict VOC risk in SCD patients.

Keywords: classifying, predictiveness curve, risk prediction, risk threshold, vaso-occlusion crisis