Impact of the COVID-19 Pandemic on Low-Value Testing and Treatment of Bronchiolitis Free

This was a retrospective, cross-sectional study of the Pediatric Health Information Systems (PHIS) database. PHIS includes anonymized patient data for more than 50 tertiary care pediatric hospitals in the United States.¹⁴  We included children <2 years of age with an emergency department encounter with an International Classification of Diseases, 10th Revision, diagnosis for bronchiolitis (J21).¹⁵  We excluded children with complex chronic conditions (captured using the corresponding flag in PHIS) because these children may be treated differently.¹⁶  We also excluded children with an International Classification of Diseases, 10th Revision, diagnosis of asthma (J45) or croup (J05) in the same encounter. A priori, we selected 3 study periods: September 1, 2018, to February 28, 2020 (prepandemic), March 1, 2020, to August 31, 2022 (early pandemic), and September 1, 2022, to January 31, 2023 (late pandemic). We defined these study periods for several reasons: (1) to include a full, typical, prepandemic viral season before the beginning of the COVID-19 pandemic; (2) to encompass the irregularity of bronchiolitis incidences during the first seasons of the pandemic; and (3) to examine the late 2022 viral season as its own, separate entity, given the simultaneous reappearance of several viruses rarely seen earlier in the pandemic.

We described encounter characteristics for each study period, including demographic factors, insurance status, and hospital or ICU admission. Low-value care was the main study outcome, defined as 1 of the following: RSV viral testing, chest radiography, albuterol, and corticosteroids. Viral testing was split into several categories: RSV-only test, SARS-CoV-2–only test, combination SARS-CoV-2 test (a test for SARS-CoV-2 and at least 1 other virus), any SARS-CoV-2 test, and any viral test. Because it might impact treatment decisions and infection control guidance, we did not consider testing for SARS-CoV-2 as low-value care. We compared the percentage of patients with any and each low-value care element across the 3 time periods (main measure % difference and 95% confidence interval [CI]). This study was approved by our institution’s review board. Statistical analyses were conducted in R (version 4.2.1; R Foundation for Statistical Computing, Vienna, Austria).

There were 387 478 encounters with a bronchiolitis diagnosis during the entire study, and after exclusion of 32 727 with a complex chronic condition, 19 795 with asthma (8412 prepandemic, 7909 early pandemic, and 3474 postpandemic), and 3425 with croup, there remained 331 531 encounters in the analysis: 148 444 (45%) prepandemic, 117 697 (36%) early pandemic, and 65 390 (20%) late pandemic. There were no major differences in demographic factors between the cohorts (Table 1). Proportions of encounters resulting in hospitalization or ICU admission were similar in the three cohorts.

The percentage of patients with any low-value care was 45%, 47%, and 44% in the 3 time periods, respectively (Table 2). There was little variation across time periods in the use of albuterol (23% to 24%), chest radiography (25% to 28%), and RSV testing (7.4% to 8.2%). Systemic steroid use was seen in 7.3% prepandemic, 10% early pandemic (+3.1%; 95% CI, 2.8–3.3 compared with prepandemic), and 11% late pandemic (+3.7%; 95% CI, 3.4–4.0 compared with prepandemic) encounters. There was considerable increase in use of viral testing, from 36% prepandemic to 65% early pandemic (+29%; 95% CI, 29–29 vs. prepandemic) and 67% late pandemic (+31%; 95% CI, 30–31 vs. prepandemic) encounters. This increase in viral testing appeared to be driven by testing for SARS-CoV-2 (50% of early pandemic and 51% of late pandemic encounters). Combination testing, including SARS-CoV-2 and at least 1 other virus, increased from 21% in the early pandemic period to 33% in the late pandemic period, whereas single-virus testing for SARS-CoV-2 decreased from 32% to 19%, respectively (Fig 1).

Despite increased volumes of patients diagnosed with bronchiolitis seen in late 2022, we found no major increase in the frequency of low-value care for children with bronchiolitis during the COVID-19 pandemic. There was a small increase in the use of systemic corticosteroids, possibly because of their reported utility in patients with COVID-19.¹³  Because of SARS-CoV-2 testing, overall frequency of viral testing increased dramatically during the study. Although testing for SARS-CoV-2 was generally done as a single-virus test in the early-pandemic stages, by the conclusion of this study, it was most commonly done in combination with tests for other viruses.

These results provide an important update to the state of care provided to children with bronchiolitis at U.S. children’s hospitals. Despite significant changes to patient volumes and viral seasonality, there was not a clinically meaningful change in use of treatments such as albuterol or corticosteroids. This is important because these may not only have little to no clinical benefit, but also lead to patient harm due to their clinical side effects. Though viral testing does not lead to direct harm, previous guidelines¹  have suggested that such testing adds health care costs while not directly changing a child’s care. It is therefore quite notable that with the advent of SARS-CoV-2 testing, there was a near doubling of the percentage of bronchiolitis encounters with a viral test. Furthermore, the advent of combination test panels of SARS-CoV-2 with influenza and RSV has likely changed the standard pattern of testing for these viruses as well. As we transition out of the pandemic into a time in which COVID-19 may develop a similar seasonality to other respiratory viruses, we should also begin to question the utility of SARS-CoV-2 testing for children who do not require hospitalization and, therefore, would not be eligible for COVID-specific therapies.

High-quality bronchiolitis care revolves around avoidance of low-value testing and treatments. In avoiding low-value care, we educate families and caregivers on what is most beneficial for children with bronchiolitis and shape their expectation for management in the future, in addition to minimizing additional costs for both the family and the health care system. It is likely that family expectations of testing and treatment for diagnoses needing only supportive care have changed as a result of such frequent viral testing in the past few years. Changes to the availability of combination viral testing at an institutional level can help to mitigate the burden of low-value care, including thoughtful design of combination tests that identify viral disease eligible for antiviral therapies without unnecessary inclusion of other viruses. Future bronchiolitis guidelines should address viral testing recommendations to account for the rise in combination viral testing, and additional deimplementation and quality improvement efforts should be pursued.

Limitations of this study include an inability within the dataset to decipher timing of low-value care (emergency department vs. inpatient) and lack of knowledge of each contributing hospital’s viral testing options (especially early in the pandemic) and policies, each of which may have influenced the findings. Additionally, the lack of clinical data included in the PHIS database made assessment of clinical severity difficult; we attempted to account for this by reporting the frequency of hospital and/or ICU admission. Finally, although we were careful with exclusion criteria, there is a possibility of institutional variation in coding of bronchiolitis and asthma, which may impact our estimates of albuterol and steroid use.

Despite the changes in practice patterns seen during the COVID-19 pandemic, most low-value care for bronchiolitis remained similar to prepandemic frequency. In contrast, viral testing has increased tremendously due to SARS-CoV-2 testing, individually and as a combination test. This change in viral testing patterns may have unintended consequences for the care of children with bronchiolitis and should be addressed in future bronchiolitis guidelines.