O2113: Pitfalls of Applying Neonatal Sepsis Guidelines to Recently Vaccinated Neonates

Background: Various biomarkers have been employed to predict the risk of bacterial infections among neonates with fever, and thereby identify those warranting additional studies, hospital admission, and/or intravenous antibiotics. Although most guidelines now limit work-up recommendations to well-appearing and febrile infants < 60 days, institutions and clinicians often take a modified approach to those < 90 days. Routine vaccinations are administered to infants most commonly on/after 60 days, although they can be given as early as 42 days. Fevers post-vaccination occur in up to 40% of patients. To date, there are no published guidelines that standardize management of infants with fever following vaccination. We sought to evaluate how our institution performs at appropriately risk stratifying infants with fevers using old versus newly endorsed metrics.

Methods: Utilizing our institution’s electronic medical record, a retrospective review of infants 22-90 days old who presented to our emergency department (ED) from January 1, 2019 through June 30, 2021 with the chief complaint of “fever” was conducted. Infants were stratified into two groups based on whether or not they were presenting with fever < 24 hours following any vaccination. The primary outcome was presence of a bacterial infection. Additional data collected included their maximum measured temperature (Tmax), laboratory results, and diagnoses/outcomes. Diagnostic accuracy of Tmax, white blood cell (WBC) count, absolute neutrophil count (ANC), urinalysis, and procalcitonin for predicting bacterial infections was compared between groups.

Results: 106 patients met inclusion criteria and of those, almost half (43%) presented to the ED with fever in the context of recent vaccination. Rates of bacterial infections were significantly higher in those presenting without recent vaccination (P-value 0.004). The distribution of their diagnoses can be seen in Table 1. The WBC and ANC were poor predictors of infection for those with recent vaccination and accuracy of these metrics differed significantly when compared to those who had not been vaccinated (Table 2, P-value 0.027 and 0.001 respectively). The urinalysis was abnormal in the absence of culture growth in 16% of samples and similar between our groups.

Conclusion: Recent vaccination accounts for a high burden of ED evaluations related to neonatal fever, despite the comparably low rate of bacterial infections within this group. We identified a strong association between vaccination and elevated ANC. Our results suggest that non-infectious stressors may elevate the ANC and limit its utility as a biomarker particularly in this context. Our false positive rate for urinalyses was higher than anticipated, irrespective of vaccination status, and requires further investigation. High false positive rates introduce the potential for harm including unnecessary hospitalizations and antibiotics, and therefore assessments should be better balanced against the particular risk of serious infection. We recommend formal guidelines for approaching neonatal fever that include considerations for recent vaccination.