Predictive models for sepsis in children with Staphylococcus aureus bloodstream infections: a retrospective cohort study.

ABSTRACT

BACKGROUND: The presence of Staphylococcus aureus in the bloodstream can lead to the development of sepsis; however, the severity and risk factors of the systemic inflammatory response to Staphylococcus aureus bloodstream infections were unclear. This study is aimed to build a model to predict the risk of sepsis in children with Staphylococcus aureus bloodstream infections. METHODS: A retrospective analysis of hospitalized pediatric patients diagnosed with Staphylococcus aureus bloodstream infections was performed between January 2013 and December 2019. Each patient was assessed using the pediatric version of the Sequential Organ Failure Assessment score (pSOFA) within 24 h of blood culture collection. A nomogram based on logistic regression models was constructed to predict the risk factors for sepsis in children with Staphylococcus aureus bloodstream infections. It was validated using the area under the receiver-operating characteristic curve (AUC). RESULTS: Of the 94 patients included in the study, 35 cases (37.2%) developed sepsis. The pSOFA scores ranged from 0 to 8, with 35 patients having a pSOFA score of ≥ 2. Six children (6.4%) died within 30 days, who were all from the sepsis group and had different pSOFA scores. The most common organs involved in sepsis in children with staphylococcal bloodstream infections were the neurologic system (68.6%), respiratory system (48.6%), and coagulation system (45.7%). Hospital-acquired infections (adjusted odds ratio [aOR], 3.0; 95% confidence interval [CI], 1.3-7.2), implanted catheters (aOR, 10.4; 95% CI, 3.8-28.4), procalcitonin level ≥ 1.7 ng/mL (aOR, 15.4; 95% CI, 2.7-87.1), and underlying diseases, especially gastrointestinal malformations (aOR, 14.0; 95% CI, 2.9-66.7) were associated with Staphylococcus aureus sepsis. However, methicillin-resistant Staphylococcus aureus infection was not a risk factor for sepsis. The nomogram had high predictive accuracy for the estimation of sepsis risk, with an AUC of 0.85. CONCLUSIONS: We developed a predictive model for sepsis in children with Staphylococcus aureus infection.