ПИЛОТНАЯ ВАЛИДИЗАЦИЯ РАННИХ ПРЕДИКТОРОВ НЕОНАТАЛЬНОГО СЕПСИСА

Abstract

Purpose: pilot development of the system estimating the risk for sepsis in newborns and determining its discrimination power. Methods. Design: a retrospective observational single-center study. 124 newborns with sepsis are included; 34 of them died. Kullback’s measure was used to estimate the information value of the examined clinical and laboratory variables. Disease outcome (whether survived or dead) was the response function. The discrimination power was determined using ROC-analysis. Results . The information value of the examined clinical and laboratory variables in the newborns was analyzed as related to the risk for sepsis. Early neonatal sepsis predictors include platelet count, total protein content, body mass and neutrophil count. Discrimination power of the mentioned predictors was calculated. Result . Determination of the risk for sepsis in a newborn based on the estimation of platelet count, neutrophil count, total protein and body mass is of moderate value.