Study finds community-onset bacterial coinfection in children with critical COVID-19 is infrequent but empiric antibiotics are commonly prescribed

In a recent study published in Open Forum Infectious Diseases, researchers evaluated the use of empiric antibiotics to determine the prevalence rates of community-acquired bacterial coinfections among hospitalized pediatric critical coronavirus disease 2019 (COVID-19) patients and to identify opportunities for de-escalating antibiotic usage in case of no bacteria-caused sepsis among high-risk individuals, and those presenting with shock.

Study: Community-onset bacterial coinfection in children critically ill with SARS-CoV-2 infection. Image Credit: nokwalai/Shutterstock


Community-acquired bacterial coinfections among hospitalized adult coronavirus disease 2019 (COVID-19) patients are uncommon; however, empiric antibiotic usage is reportedly high. Data on empiric antibiotic usage and bacterial coinfections among pediatric individuals with critical severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections are limited.

The clinical manifestations of severe SARS-CoV-2 infections often include pulmonary distress and fever, findings that could be difficult to discriminate from serious bacterial infections, which might prompt the use of empiric antibiotics in the initial days of hospitalization, particularly among high-risk individuals.

About the study

In the present study, researchers investigated whether any radiographic, laboratory, or clinical features ascertainable during hospitalization were related to empiric antibiotic usage or were estimative of bacterial coinfections acquired in community settings.

The team evaluated individuals below 19.0 years and admitted to pediatric high-acuity units (HAU) or intensive care units (ICU) due to SARS-CoV-2 infections from March 2020 to December 2020. On the basis of microbiology reports from the initial 72 hours of hospitalization, the team adjudicated if patients had community-acquired bacterial coinfections.

Clinical and demographic variables of individuals with and without antibiotic prescriptions and bacterial coinfections in the initial days of hospitalization were compared. Poisson regression modeling was performed to assess factors related to the outcome, and the adjusted relative risk (aRR) values were calculated.

Data were obtained from patient electronic medical records and data from the nationwide overcoming COVID-19 population health active surveillance registry of patients hospitalized due to COVID-19-associated complications between 15 March 2020 and 31 December 2020 across >70.0 pediatric hospitals in 25 states.

COVID-19 diagnosis was confirmed using polymerase chain reaction (PCR). The team excluded multisystem inflammatory syndrome among children (MIS-C) patients diagnosed using the Centers for Disease Control and Prevention (CDC) criteria. Data were obtained on demographic parameters, clinical symptoms and signs, comorbidities, radiographical and laboratory investigations, and data on antibiotics prescribed at admission and the course of critical COVID-19, including clinical outcomes and hemodynamic and respiratory support needed.

The primary study outcome assessed was the prescriptions of empirical antimicrobials, for which enteral or intravenous antimicrobials administered in the initial two days of hospital admission were assessed. The second outcome evaluated community-acquired bacterial infection presence, for which relevant case report form (CRF) information from individuals with SARS-CoV-2-positive microbiological cultures, and PCR, were analyzed in the initial 72 hours of hospital admission.


Out of 532 individuals, 63.0% were administered empiric antibiotics; however, only seven percent developed bacterial coinfections, of which only three percent were respiratory-type. Empirical antibiotics had a greater likelihood of being prescribed to immunosuppressed individuals (aRR of 1.3), requiring non-mechanical ventilator-type respiratory aid (aRR of 1.4), or requiring invasive-type mechanical ventilators (aRR of 1.8), than no respiratory aid.

The most frequently prescribed antimicrobials were ceftriaxone (41%) and vancomycin (28%), followed by cefepime (20%). Most individuals were prescribed multiple antimicrobials, with 21%, 10%, and 18% receiving 2.0, 3.0, and ≥4.0 antibiotics in the initial two days of hospital admission. More than 33% of individuals received antibiotics for ≥5.0 days, despite no evidence of bacterial coinfections. The median social vulnerability index (SVI) values were significantly greater among those who received antibiotics than those who did not.

The median C-reactive protein (CRP) levels were greater among those who received antibiotics versus those who did not (4.6 mg per dL vs. 2.2 mg per dL), as were the median procalcitonin levels (0.4 ng per mL vs. 0.1 ng per mL). The median leukocyte counts showed no significant differences between the two groups. Antibiotic usage was related to COVID-19 severity, indicated by greater median values for PEdiatric Logistic Organ Dysfunction-2 (PELOD-2) scores at hospitalization among individuals who received antibiotics than those who did not.

Seven percent (n=38) of individuals had true community-onset bacterial coinfections, of which 13, 16, 8.0, and 4.0 were bloodstream infections, respiratory infections, urinary tract infections, and bacterial infections at other sites (peritonitis, colitis, meningitis, and pharyngitis), respectively.

No particular pathogenic organism predominantly caused bacterial coinfections, although most pulmonary coinfections were caused by Staphylococcus aureus and/or Pseudomonas aeruginosa. Greater PELOD-2 scores at admission were associated with bacterial coinfections (aRR of 1.2), in addition to age, sex, and pulmonary conditions other than asthma (aRR 2.3).


Overall, the study findings showed that community-onset bacterial coinfections among children with critical COVID-19 are not frequent; however, empiric antibiotics are usually prescribed. The study findings inform antibiotic use and underpin swift de-escalation in case assessments indicating that coinfections are not likely.