Objectives: Bronchiolitis is the most common disease of the lower respiratory tract in the first year of life. Treatment is controversial, with studies giving conflicting views on the benefits of bronchodilators and steroids. The objectives of this study were 1) to characterize the management of bronchiolitis in pediatric emergency departments (PEDs) in Canada, 2) to determine patient outcomes following emergency department (ED) visits, and 3) to provide descriptive data regarding bronchiolitis symptoms and family/personal medical history of these patients. Methods: A prospective consecutive cohort of children with bronchiolitis presenting to seven Canadian PEDs was enrolled during a seven‐to‐21‐day period. Standardized interviews with parents provided data regarding symptoms, previous treatment, and past history. Charts were reviewed for treatment, investigations, and disposition. Telephone follow‐up at two to three weeks collected information regarding duration of illness and return visits. Results: Two hundred thirty‐seven (91%) of 260 eligible patients were enrolled. One hundred eighty‐nine patients (80%) had both an interview and chart review, and 48 (20%) had only chart reviews; follow‐up was completed for 163 (69%) patients. One hundred fifteen (63%) had seen their primary care provider during their illness prior to the ED visit. Seventy‐three percent of patients (range per site 59–100%) were treated in the ED with bronchodilators (usually salbutamol or epinephrine) and 5% (range per site 0–14%) with oral steroids. Twenty‐four percent (58/237) were prescribed bronchodilators on discharge, 3% (7/237) inhaled steroids, and 2% (5/237) oral steroids. Chi‐square tests indicated significant practice variation by site in ED bronchodilator use (p < 0.001) and bronchodilator use at discharge (p = 0.0003). Admission rate was 31% (range by site 22–43%), 17% of patients had more than one ED visit, and 1% were admitted more than once. Admission rates were increased in younger children, children with comorbidities, and children with lower oxygen saturation. Viral studies were obtained in 53%, with 76% of these positive for respiratory syncytial virus (RSV). Median duration of cough was 12 days, poor sleeping and irritability eight days, and wheeze and poor feeding seven days. Conclusions: This study prospectively describes the treatment of bronchiolitis in the pediatric ED. The findings are consistent with the literature regarding the reported use of bronchodilators; however, use of steroids was found to be much lower than reported in other studies. Bronchodilator use in the ED and at discharge varied significantly by site. The results capture variation in treatment practices in Canadian PEDs, which may be the result of discordant randomized controlled trial evidence. Further research is needed to establish best practices.

Introduction: The first steps in goal-directed therapy for sepsis are early diagnosis followed by appropriate triage. These steps are usually left to the physician’s judgment, as there is no accepted biomarker available. We aimed to determine biomarker phenotypes that differentiate children with sepsis who require intensive care from those who do not. Methods: We conducted a prospective, observational nested cohort study at two pediatric intensive care units (PICUs) and one pediatric emergency department (ED). Children ages 2–17 years presenting to the PICU or ED with sepsis or presenting for procedural sedation to the ED were enrolled. We used the judgment of regional pediatric ED and PICU attending physicians as the standard to determine triage location (PICU or ED). We performed metabolic and inflammatory protein mediator profiling with serum and plasma samples, respectively, collected upon presentation, followed by multivariate statistical analysis. Results: Ninety-four PICU sepsis, 81 ED sepsis, and 63 ED control patients were included. Metabolomic profiling revealed clear separation of groups, differentiating PICU sepsis from ED sepsis with accuracy of 0.89, area under the receiver operating characteristic curve (AUROC) of 0.96 (standard deviation [SD] 0.01), and predictive ability (Q²) of 0.60. Protein mediator profiling also showed clear separation of the groups, differentiating PICU sepsis from ED sepsis with accuracy of 0.78 and AUROC of 0.88 (SD 0.03). Combining metabolomic and protein mediator profiling improved the model (Q² =0.62), differentiating PICU sepsis from ED sepsis with accuracy of 0.87 and AUROC of 0.95 (SD 0.01). Separation of PICU sepsis or ED sepsis from ED controls was even more accurate. Prespecified age subgroups (2–5 years old and 6–17 years old) improved model accuracy minimally. Seventeen metabolites or protein mediators accounted for separation of PICU sepsis and ED sepsis with 95 % confidence. Conclusions: In children ages 2–17 years, combining metabolomic and inflammatory protein mediator profiling early after presentation may differentiate children with sepsis requiring care in a PICU from children with or without sepsis safely cared for outside a PICU. This may aid in making triage decisions, particularly in an ED without pediatric expertise. This finding requires validation in an independent cohort.