Sepsis Prediction: Biomarkers Combined in a Bayesian Approach

Background Sepsis is a serious public health problem. The soluble triggering receptor expressed on myeloid cells-1 is a marker of inflammatory and infectious processes that has the potential to become a useful tool for predicting the evolution of sepsis. To build a prediction model for sepsis by combining soluble triggering receptors expressed on myeloid cells-1, C-Reactive Protein, and leukogram using a Bayesian network. Methods This is an exploratory translational study, was carried out with 32 children with congenital heart disease who had undergone surgical correction in a public referral hospital in the northeast of Brazil. The TRIPOD recommendations were used. Results We found that in the postoperative period, the mean soluble triggering receptors expressed on myeloid cells-1 was higher among patients diagnosed with sepsis than among those without the diagnosis, respectively (soluble triggering receptors expressed on myeloid cells-1 post = 394.58 pg/mL versus 239.93 pg/mL p < 0.001). Analysis of the ROC curve for soluble triggering receptors expressed on myeloid cells-1 and sepsis showed that the area under the curve was 0.761 with a 95% CI (0.587–0.935) and p = 0.013. We found, with the Bayesian model, a 100% probability of sepsis related to post-operative blood concentrations of C-Reactive Protein above 71 mg/dL, leukogram values above 14,000 cells/µL, and soluble triggering receptors expressed on myeloid cells-1concentrations above the cut-off point (283.53 pg/mL). Conclusions The model proposed using the Bayesian Network approach proved to be effective for diagnosing sepsis, with adequate sensitivity and specificity.