Association Between Residential Rurality and Diagnostic Imaging Use in Pediatric Dog-Bite Encounters: A Cross-Sectional Multicenter Pediatric Health Information System Study (2021-2022)

TL;DR

Children in rural areas are more likely to get imaging for dog bites than those in urban areas, but the difference is small.

Contribution

This study identifies a rural-urban disparity in diagnostic imaging use for pediatric dog-bite injuries using a large national database.

Findings

Rural patients had a 5.32 percentage point higher imaging rate compared to urban patients.

The odds of imaging increased by approximately 5% per RUCA point, with a total difference of 8.61 percentage points between the most urban and rural groups.

Longer hospital stays were also associated with higher imaging use.

Abstract

Introduction: Imaging helps evaluate pediatric dog-bite injuries, but utilization may vary by rural-urban setting. We aimed to determine whether rural-urban commuting area (RUCA) classification of residence is associated with diagnostic imaging utilization in pediatric dog-bite encounters at US children’s hospitals.

Methods: We conducted a retrospective cross-sectional study using the Pediatric Health Information System (PHIS) across 45 children’s hospitals (January 2021 to December 2022). Patients aged ≤18 years with the International Classification of Diseases, 10th Revision (ICD-10) dog-bite codes (W54.0XXA; S01/S41/S61/S71 series) were included, and those with invalid RUCA codes were excluded. The exposure was RUCA code (2020 scale: 1 urban to 10 rural). The primary outcome was imaging during the encounter, defined as radiography, ultrasonography, computed tomography, or magnetic resonance imaging. Multivariable logistic regression modeled RUCA as a continuous predictor of imaging while adjusting for sex, payer, disposition, and length of stay (LOS). Model-predicted probabilities were summarized across one-unit RUCA bands, and linear regression was used to assess trends. PHIS data were de-identified.

Results: Among 13,901 encounters, the median age was seven years (interquartile range: 3-11), and 57.1% were male. Imaging occurred in 2,992 encounters (21.5%). Unadjusted imaging rates were urban 21.2% vs. rural 26.5%, an absolute difference of 5.32 percentage points (95% confidence interval (CI): 2.37-8.27; p < 0.001). The predicted probability of imaging increased stepwise with rurality; linear regression of RUCA bands showed a positive association (R² = 0.91; F(1, 8) = 81.97; p < 0.001). In the adjusted analysis, a higher RUCA code was associated with greater odds of imaging (β = 0.052 per RUCA point; odds ratio (OR) ≈ 1.05 per point; z = 3.37; p < 0.001), corresponding to approximately 8.61 percentage points higher absolute imaging between the most urban and most rural groups. A longer LOS was also associated with imaging (β = 0.341 per day; OR ≈ 1.41 per day; z = 9.37; p < 0.001). Most patients were discharged from the emergency department (83.0%), and 5.0% were admitted.

Conclusion: Rural residence was associated with higher imaging use in pediatric dog-bite encounters at children’s hospitals. These observational findings are hypothesis-generating; differences may reflect practice patterns, case-mix, or selection rather than causation. Absolute differences were modest, suggesting system-level relevance but limited impact for most individual encounters.

Figures