Risk Factors for Serious Bacterial Infections Among Young Infants With Hypothermia: Protocol for a Multicenter, Retrospective Case-Control Study

Abstract

Background: Hypothermia in young infants presenting to the emergency department (ED) may indicate a serious bacterial infection (SBI) such as a urinary tract infection, bacteremia, or bacterial meningitis. Improved understanding of the epidemiology of SBI in infants with hypothermia and the development of prediction models can help avoid unnecessary invasive procedures and antimicrobial exposure. Objective: The aim of the study is to (1) describe the epidemiology of SBI and herpes simplex virus (HSV) among infants with hypothermia, (2) assess the role of biomarkers in predicting SBI, and (3) derive and internally validate a multivariable predictive model for SBI among infants with hypothermia. Methods: The study is being conducted through the Pediatric Emergency Medicine Collaborative Research Committee as a retrospective nested case-control study. We will include infants with hypothermia (rectal temperature<36.5 °C) presenting to 1 of 28 pediatric EDs in the United States between January 1, 2013, and December 31, 2022. Exclusion criteria will include (1) fever in the ED or prior to ED arrival, (2) transfer from another health care facility, (3) technology dependence, (4) trauma, (5) skin and soft tissue infections, and (6) presentation in cardiac arrest. The primary outcomes will be culture-confirmed SBI (objectives 1-3) and HSV-positivity (objective 1). The analytic approach for each objective will be (1) to use descriptive statistics to summarize the epidemiology of SBI and HSV and clinical outcomes, (2) to compare biomarker parameters of infants with and without SBI and determine optimal cutoff values using a receiver operator characteristic curve, and (3) to develop a multivariable predictive model using penalized mixed effects logistic regression within a 1:3 case-control sample among infants in the cohort who had a blood culture obtained. Results: Data collection for this study is ongoing, with a collection of data from 21 hospitals at the time of protocol submission. Of 16 sites evaluated in preliminary analysis (n=45,673), the median age is 25 (IQR 6-52) days, and 24,182 (52.9%) are male. In total, 340 (0.7%) infants have an SBI, of whom 103 had bacteremia or meningitis. Mortality occurred in 7 (2.1%; 95% CI 0.9%-4.3%). Conclusions: We will use a consensus-based outcome measure for SBI with an established data acquisition pathway. We will use a multicenter sample from US children’s hospitals, using a consensus-based outcome measure for SBI and a case-control approach to evaluate outcomes to improve the management of young infants with hypothermia in the ED.