Health Services Use for SARS-CoV-2–Infected Children With Croup or Bronchiolitis

We conducted a retrospective cohort study using electronic health record (EHR) data from PEDSnet institutions. PEDSnet (pedsnet.org) is a multiinstitutional clinical research network that aggregates EHR data from several of the nation’s largest children’shealth care organizations. Participating institutions included Children’s Hospital of Philadelphia, Cincinnati Children’s Hospital Medical Center, Children’s Hospital Colorado, Ann & Robert H. Lurie Children’s Hospital of Chicago, Nationwide Children’s Hospital, Nemours Children’s Health System (a Delaware and Florida health system), Seattle Children’s Hospital, and Stanford Children’s Health. PEDSnet standardizes data to the Observational Medical Outcomes Partnership common data model and has accrued data for >8 million children with at least 1 visit from 2009 to 2021 from inpatient and outpatient clinical settings. Data were extracted from the PEDSnet coronavirus disease 2019 database version May 19, 2022. The Children’s Hospital of Philadelphia’s institutional review board designated this study as nonhuman subjects research and the need for consent was waived.

We retrieved EHR data from children aged 30 days to 14 years who underwent SARS-CoV-2 polymerase chain reaction testing within 72 hours of a diagnosis of croup or bronchiolitis made in outpatient, inpatient, and emergency department settings. The study period spanned from March 1, 2020, to May 31, 2022. We used the International Classification of Diseases, Ninth Revision, Clinical Modification/SNOMED codes to identify subjects with bronchiolitis and croup (Supplemental Tables 3 and 4). We formed 2 study cohorts: Children with croup (aged between 6 months and 14 years) and children with bronchiolitis (aged between 30 days and 2 years). The entrance was the day of the first positive or first negative test (Supplemental Table 5).

The primary outcome was hospitalization with croup or bronchiolitis. Secondary outcomes included all-cause readmissions within 30 days of initial diagnosis with bronchiolitis or croup, hospital length of stay, and admission to the ICU. Variant periods were defined as the timing in which a variant represented >50% of cases in the United States. The predelta period was from March 2020 to June 2021, the delta period was from July to December 2021, and the omicron period was from January to May 2022. Patient-level variables included age, sex, race, ethnicity, insurance status, and complex conditions. Complex conditions were defined as using the Pediatric Medical Complexity Algorithm classification system.⁶ 

χ² or Mann-Whitney tests were used to assess bivariable relationships between children with SARS-CoV-2–positive and negative croup and bronchiolitis with respect to demographic variables (age, sex, race/ethnicity, insurance, site, chronic conditions), and clinical outcomes (hospitalization, hospital length of stay, ICU admission, readmission). Multivariable mixed effects logistic regression models with random intercept for site were used to identify associations between SARS-CoV-2 positivity and clinical outcomes, adjusting for baseline characteristics including age, sex, race, ethnicity, insurance status, and complex conditions. We considered all factors that were statistically and clinically significant on a bivariable level for inclusion in the model. Multivariable mixed effect log gamma regression was used for the secondary outcome of length of stay. Odds ratios and 95% confidence intervals were estimated, and results with P < .05 were considered statistically significant. Analyses were conducted using R version 4.1.2 (November 1, 2021).

The number of emergency department visits and hospitalizations for croup and bronchiolitis among SARS-CoV-2 cases by month is shown in Fig 1. The peak of SARS-CoV-2 croup and bronchiolitis was in January 2022 during the omicron period. Across all time periods, more than half (59.6%) of SARS-CoV-2 croup cases presented during the omicron period. SARS-CoV-2 bronchiolitis cases were more evenly distributed across the delta (42.6%) and omicron (44.3%) time periods.

Among children with croup, the median age was 33 months (interquartile range, 18–56 months), and most (89.9%) were aged <6 years. Among children with croup, 1835 children were SARS-CoV-2 positive (Table 1). SARS-CoV-2–positive children were younger (median age 24 months) compared with SARS-CoV-2–negative children (median age 33 months). A greater proportion of children with SARS-CoV-2 were Hispanic/Latinx (24.2%) compared with children without SARS-CoV-2 (15.0%; P ≤ .001). A greater proportion of children with SARS-CoV-2 were seen during the omicron time frame (59.6%), compared with predelta (10.8%) and delta (29.8%; P ≤ .001) time frames.

Among subjects with bronchiolitis, the median age was 8.3 months (interquartile range, 4–15 months); 652 subjects were SARS-CoV-2 positive. SARS-CoV-2–positive children were younger (median age 6 months) compared with SARS-CoV-2–negative children (median age 9 months). A greater proportion of children with SARS-CoV-2 infection were Medicaid/Children’s Health Insurance Program beneficiaries (56.0%) compared with children without SARS-CoV-2 (50.9%; P ≤ .001). A greater proportion of children with SARS- CoV-2 were Hispanic/Latinx (21.3%) compared with children without SARS-CoV-2 (18.7%; P = .001). A greater proportion of children with SARS-CoV-2 were white race (50.9%) compared with children without SARS-CoV-2 (49.3%; P ≤ .0001). A greater proportion of children with SARS-CoV-2 were seen during the omicron time frame (44.3%), compared with pre delta (13.0%) and delta (42.6%; P ≤ .001) time frames.

The results of the multivariable regression analyses are shown in Table 2. After controlling for patient-level variables and clustering within hospital, for both bronchiolitis and croup, we found no difference in the adjusted odds of clinical outcomes (hospitalization, hospital length of stay, ICU admission,/readmission) comparing children with SARS-CoV-2 to those without.

In this multiinstitutional, retrospective study evaluating children with molecular testing for SARS-CoV-2 with croup or bronchiolitis during a time of universal SARS-CoV-2 testing for hospitalized children in the United States, we found no differences in clinical outcomes when comparing children with evidence of SARS-CoV-2 infection to those without. Overall, SARS-CoV-2 positivity was low among croup and bronchiolitis cases but accounted for a higher proportion of croup cases than bronchiolitis cases. The positivity rates of SARS-CoV-2 in cases of croup and bronchiolitis align with findings from similar studies.^3,7,8  Across all time frames, most SARS-CoV-2 croup cases were seen during the omicron time frame, whereas SARS-CoV-2 bronchiolitis was seen across both the delta and omicron time frames. This difference may result from younger children’s increased susceptibility during the omicron period, with smaller airway caliber and differing pathophysiology between delta and omicron variants, specifically omicron’s rapid replication rate in upper airways.⁴  Although we did not find significant differences in admission rates comparing children with croup who tested positive for SARS-CoV-2 to those who tested negative across the time periods, we did note that, across all periods, 59.6% of the admissions for SARS-CoV-2–associated croup occurred during the omicron period. This finding is in line with a previous observational study that noted that, among children with SARS-CoV-2, there were higher rates of hospitalization for SARS-CoV-2–associated croup during omicron period versus pre-omicron.² 

Differences in demographics (age, race, ethnicity), complex chronic conditions, and insurance coverage were noted between SARS-CoV-2–positive and negative patient groups in our study. Previous studies have found disparities in infection rates and outcomes by race, ethnicity, and socioeconomic status.⁹  Our study was not able to determine the underlying factors associated with these differences in SARS-CoV-2 infection, which likely reflect unmeasured factors (eg, overrepresentation of disadvantaged populations considered essential workers).

The study has several limitations that need to be acknowledged. First, the data collection period for the study did not include data from June 2022 onwards, which limits the temporal scope of the findings. Additionally, the study did not evaluate treatments or oxygen requirements, which could have provided additional insights into patient severity and outcomes. Another limitation is that the use of hospitalization and ICU admission as surrogate markers for severity may not fully capture the complexity of the disease progression, and different thresholds for ICU admission may exist across PEDSnet sites. The SARS-CoV-2–negative children represent a heterogeneous group, and other viral results were not available for all SARS-CoV-2–negative patients, so we were unable to compare specific pathogens to SARS-CoV-2. Lastly, important variables such as vaccination data and treatment information, which could have impacted the outcomes, were not included in the study.

The findings of this study suggest that pediatric patients with croup and bronchiolitis associated with SARS-CoV-2 infection do not experience higher rates of hospital and ICU admissions or readmissions compared with patients without SARS-CoV-2. The study contributes to our understanding of the impact of SARS-CoV-2 on respiratory illnesses in children. These findings have implications for capacity and workforce planning and public health recommendations, and highlight the need for continued monitoring of variant-specific effects. Further research is necessary to investigate potential treatments and other factors that may influence disease severity in pediatric patients with croup and bronchiolitis caused by SARS-CoV-2.

We thank the PEDSnet Data Coordinating Center at Children’s Hospital Philadelphia and the clinical informatics team at Children’s Hospital Colorado for their assistance with obtaining and preparing the data set for the study.: