Pediatric sepsis and septic shock in the era of sepsis-3

Abstract

Background and aims The definition of septic shock has recently been revised in adults. However, there are a lack of contemporary pediatric data to informsimilar approaches in children with severe sepsis. We assessed the performance of clinical variables at PICU admission to define sepsis severity and predict death in critically ill children. Methods Retrospective multicenter cohort study of children requiring intensive care in Australia and New Zealand from 2009-2015. Children <16 years with sepsis or septic shock on admission to PICU were included. We randomized the cohort into development (70%) and validation samples (30%) for risk score development. Multivariable logistic regression was used to estimate the risk of death for each admission. Results 2,607 children with sepsis were included, with a crude mortality of 7.4% (194). In the multivariate model, low systolic blood pressure, inotropes, lactate and base excess, mechanical ventilation and the ratio of PaO2/FiO2 at PICU admission were independently associated with an increased risk of mortality. Fixed dilated pupils, and immunosuppression were also associated with an increased risk of mortality. The model discriminated well in both the combined dataset and when applying the model derived from the development data to the validation subset (both AUROC=0.86). In comparison, PIM2 achieved an AUROC of 0.79 in-sample. Conclusions In view of the persistently high mortality of paediatric septic shock, improved outcome predictors are urgently needed to facilitate design of targeted interventions to high risk groups. The present study addresses a knowledge gap to inform Sepsis-3 translation to pediatric sepsis.L6138856862017-01-02 <br />