Pediatric asthma population risk stratification using k-means clustering.

Area of Science:

• Pediatric Health
• Social Epidemiology
• Health Disparities

Background:

• Asthma disproportionately affects children, with social determinants of health (SDOH) playing a significant role in exacerbations and outcomes.
• Identifying specific pediatric asthma patient groups based on SDOH is crucial for targeted interventions and reducing health disparities.

Purpose of the Study:

• To stratify pediatric asthma patients into distinct groups using SDOH.
• To identify high-risk groups for asthma exacerbations and inform tailored care strategies.

Main Methods:

• Utilized insurance claims and enrollment data for 22,169 children aged 5-18 with asthma in Arkansas.
• Employed Generalized Low-Rank Models and K-means clustering incorporating comorbid conditions, medication data, Child Opportunity Index, and rural-urban classification.
• Compared resulting clusters based on asthma-related emergency department visits and hospitalizations.

Main Results:

• K-means clustering identified six distinct pediatric asthma patient groups.
• Cluster 1, characterized by a high percentage of Black children and residence in low Child Opportunity Index neighborhoods, exhibited significantly higher asthma-related emergency department visits.
• This high-risk cluster demonstrated profound disparities in education, health, environment, and socioeconomic factors.

Conclusions:

• Pediatric asthma disparities necessitate interventions addressing social, economic, and environmental inequities.
• Population risk stratification, as demonstrated by clustering, can effectively identify high-risk groups, such as Black children in low opportunity areas, for tailored asthma management.
• The study highlights the potential of SDOH in understanding and mitigating pediatric asthma exacerbations.