Association of empiric antibiotic selection and clinical outcomes in hospitalised children with severe orbital infections: a retrospective cohort study.

Objective: To determine the association of initial empiric antibiotic regimens with clinical outcomes in hospitalised children with severe orbital infections.

Design: Multi-centre observational cohort study using data from 2009 to 2018 clinical records.

Setting: Canadian children's hospitals (7) and community hospitals (3).

Patients: Children between 2 months and 18 years hospitalised for >24 hours with severe orbital infections.

Interventions: Empiric intravenous antibiotic regimen in the first 24 hours of hospitalisation.

Main Outcome Measures: Length of hospital stay and surgical intervention using multivariable median regression and multivariate logistic regression, with adjustment for covariates.

Results: Of 1421 patients, 60.0% were male and the median age was 5.5 years (IQR 2.4-9.9). Median length of stay was 86.4 hours (IQR 56.9-137.5) and 180 (12.7%) received surgical intervention. Patients receiving broad-spectrum empiric antibiotics had an increased median length of stay, ranging from an additional 13.8 hours (third generation cephalosporin and anaerobic coverage) to 19.5 hours (third generation cephalosporin, staphylococcal and anaerobic coverage). No antibiotic regimen was associated with a change in the odds of surgical intervention. These findings remained unchanged in sensitivity analyses restricted to more severely ill patients. There was a twofold increase in the percentage of patients receiving the broadest empiric antibiotic regimens containing both staphylococcal and anaerobic coverage from 17.8% in 2009 to 40.3% in 2018.

Conclusions: Empiric use of broad-spectrum antibiotics with staphylococci and anaerobic coverage was associated with longer length of stay and similar rates of surgery in children with orbital infections. There is an urgent need for comparative effectiveness studies of various antibiotic regimes.