Categorizing Hospitals by Neonatal and Pediatric Diagnoses Treated

Area of Science:

• Pediatric hospital classification
• Healthcare quality research
• Inpatient care analysis

Background:

• Traditional hospital classification relies on physical attributes and branding, not patient populations.
• A novel approach is needed to group hospitals based on the types of pediatric patients they serve.

Purpose of the Study:

• To group hospitals based on patient case-mix, specifically neonatal services and pediatric diagnosis diversity.
• To move beyond traditional hospital classifications and create groups reflecting actual pediatric care patterns.

Main Methods:

• Retrospective analysis of 2.8 million pediatric hospitalizations (ages 0-20) from the 2019 Kids' Inpatient Database (3993 hospitals).
• K-means clustering was used to group hospitals based on neonatal case-mix and pediatric diagnosis diversity (DD).
• Low-volume hospitals (≤100 annual admissions) were stratified separately.

Main Results:

• Six distinct hospital groups emerged, characterized by varying levels of neonatal care severity and diagnosis diversity.
• Groups ranged from low-volume pediatric hospitals to comprehensive centers with high-severity neonatal care and broad diagnosis diversity.
• Specific group characteristics included hospitalization volume, neonatal service provision, and the number of unique pediatric diagnoses treated.

Conclusions:

• Children's inpatient care is delivered across six identifiable hospital groups, defined by neonatal case-mix and diagnosis diversity.
• This patient-centered classification provides a foundation for future research into healthcare utilization and outcomes.
• Further investigation is warranted to assess the utility of these groups for benchmarking and quality improvement initiatives.