LAUSR

Abstract Background Community-onset bacterial coinfection in adults hospitalized with coronavirus disease 2019 (COVID-19) is reportedly uncommon, though empiric antibiotic use has been high. However, data regarding empiric antibiotic use and bacterial coinfection in children with critical illness from COVID-19 are scarce. Methods We evaluated children and adolescents aged <19 years admitted to a pediatric intensive care or high-acuity unit for COVID-19 between March and December 2020. Based on qualifying microbiology results from the first 3 days of admission, we adjudicated whether patients had community-onset bacterial coinfection. We compared demographic and clinical characteristics of those who did and did not (1) receive antibiotics and (2) have bacterial coinfection early in admission. Using Poisson regression models, we assessed factors associated with these outcomes. Results Of the 532 patients, 63.3% received empiric antibiotics, but only 7.1% had bacterial coinfection, and only 3.0% had respiratory bacterial coinfection. In multivariable analyses, empiric antibiotics were more likely to be prescribed for immunocompromised patients (adjusted relative risk [aRR], 1.34 [95% confidence interval {CI}, 1.01–1.79]), those requiring any respiratory support except mechanical ventilation (aRR, 1.41 [95% CI, 1.05–1.90]), or those requiring invasive mechanical ventilation (aRR, 1.83 [95% CI, 1.36–2.47]) (compared with no respiratory support). The presence of a pulmonary comorbidity other than asthma (aRR, 2.31 [95% CI, 1.15–4.62]) was associated with bacterial coinfection. Conclusions Community-onset bacterial coinfection in children with critical COVID-19 is infrequent, but empiric antibiotics are commonly prescribed. These findings inform antimicrobial use and support rapid de-escalation when evaluation shows coinfection is unlikely.