Network Analysis to Define Pediatric Acute Care Regions in Wisconsin

Area of Science:

• Health Services Research
• Network Analysis in Healthcare
• Pediatric Healthcare Systems

Background:

• Defining optimal geographic regions for healthcare delivery is crucial for resource allocation and outcome assessment.
• Existing healthcare regions are often based on adult utilization patterns, potentially not reflecting pediatric needs.
• A data-driven approach is needed to delineate pediatric-specific service areas.

Purpose of the Study:

• To pilot a novel system for deriving pediatric healthcare regions.
• To compare the fit of these derived pediatric regions with existing adult-based regions using pediatric utilization data.
• To evaluate the effectiveness of network analysis in defining pediatric emergency service areas (PESAs) and pediatric emergency referral regions (PERRs).

Main Methods:

• Cross-sectional study of acute care encounters for children under 16 in Wisconsin (2021-2022) using Healthcare Cost and Utilization Project (HCUP) data.
• Network analysis, specifically community detection, was applied to ZIP code-level patient-hospital connections to form PESAs.
• A second community detection iteration on inter-PESA referrals defined PERRs, with modularity used as the primary fit measure.

Main Results:

• Analysis of 587,886 encounters resulted in 24 PESAs and 4 PERRs.
• PERRs demonstrated superior fit for interhospital pediatric referral patterns (modularity 0.53) compared to hospital referral regions (HRRs) (0.46).
• PERRs were significantly larger, encompassed more children, and included more hospitals than HRRs and other benchmark regions.

Conclusions:

• A network analysis approach successfully derived pediatric-specific acute care regions (PERRs) with strong alignment to pediatric utilization data in Wisconsin.
• These PERRs offer a better fit for pediatric referral patterns than traditional adult-based regions.
• Further validation across the United States is recommended to enable robust comparisons of costs and outcomes between pediatric regions.