57 - Temporal Patterns of Imaging Practices for Children Hospitalized with Community-Acquired Pneumonia

Publication Number: 3054.57

Michelle Polich, University of California, San Diego, San Diego, CA, United States; Melissa Cameron, Rady Children's Hospital San Diego, San, CA, United States; Elizabeth Mannino Avila, UCSD/Rady Children's Hospital San Diego, San Diego, CA, United States; Christiane Lenzen, University of California, San Diego School of Medicine, San Diego, CA, United States; Sarah Gray, University of California San Diego, Bethesda, MD, United States; Kyung E.. Rhee, UC San Diego School of Medicine, LA JOLLA, CA, United States; Ronald J. Teufel II, UC, San Diego/Rady Health, San Diego, CA, United States

Background: Chest radiographs (CXRs) are the recommended imaging modality for hospitalized children with community-acquired pneumonia (CAP) but have important limitations including variable interpretation, radiation exposure, and limited value for follow-up management. Lung ultrasound (US) has emerged as a highly sensitive and specific diagnostic tool for CAP and associated complications such as pleural effusion in children. Despite growing evidence supporting lung US, national trends in imaging utilization for inpatient pediatric CAP remain unclear.

Objective: To evaluate trends in CXR, computed tomography (CT), and US utilization for children hospitalized with CAP over a 9-year period.

Design/Methods: This multicenter study using the Pediatric Health Information System database, includes children aged 3 months-18 years hospitalized with CAP from 2016 through 2024. We excluded transfers from outside hospitals, neonatal intensive care admissions, deaths, and length of stay (LOS) >30 days. The unadjusted number of imaging studies per patient encounter was calculated. Generalized estimating equation models were used for each imaging modality to adjust for seasonality, COVID-19 pandemic, rates of ICU admissions, age, and clustering within each hospital. Given the clinical differences in management for those with simple vs. complicated CAP, we stratified analysis a priori by each CAP type.

Results: Among 327,287 encounters (median age 49 months, IQR 23-100), there were 617,705 CXRs, 19,427 chest ultrasounds, and 15,587 chest CTs performed. Median LOS was 3 days (IQR 2-6), with 30.7% admitted to the ICU, and 8.8% with complicated CAP. Unadjusted average imaging counts per 100 encounters were higher for complicated CAP compared to simple CAP (CXR: mean 511.0 (std 590.6) vs 157.7 (259.8), p< 0.001; US: 52.9 (82.7) vs 1.4 (12.8), p < 0.001; CT: 26.5 (52.2) vs 2.7 (17.0), p< 0.001). For simple CAP, adjusted CXR rates declined by 1.5% annually with no change in CT or US rates. However, for complicated CAP, CXR and CT scans increased by 1.8% and 4.1% respectively, while US rates did not significantly change.

Conclusion(s): From 2016-2024, CXR and CT imaging for children hospitalized with complicated pneumonia increased. In contrast, CXR and CT use for simple pneumonia improved and remained unchanged, respectively. Despite evidence promoting the use of chest US as an alternative to radiation-based modalities, US use remained low across all pneumonia admissions. Implementation of evidence-based imaging protocols and increased adoption of lung ultrasound have the potential to reduce radiation exposure among these children.

Table 1. Cohort Demographics, Overall and by Simple and Complicated Community-Acquired Pneumonia
Abbreviations: CAP, community-acquired pneumonia; IQR, interquartile range; NH, non-Hispanic.
*NH-Other = American Indian, Unknown, Other
^Other = Self Pay, Charity, Unknown

Table 2. Adjusted Percentage of Annual Change for All Imaging Types, Overall and for Simple and Complicated CAP
Abbreviations: CAP, community-acquired pneumonia; %Δ, estimated annual percent change, SE, standard error

Figure 1. Predicted Imaging Rates per 100 Patient Encounters From 2016 through 2024, for Simple and Complicated CAP