A latent class model to identify city/town chronic disease patterns

Area of Science:

• Public Health
• Health Services Research
• Biostatistics

Background:

• Health departments face challenges in allocating resources for chronic disease management.
• Effective resource allocation requires understanding localized disease patterns.

Purpose of the Study:

• To adapt a state-level model for customized city/town chronic disease profiling.
• To identify distinct patterns of chronic disease across 39 Rhode Island municipalities.

Main Methods:

• Utilized four data sources to define 20 indicators across demographics, health behaviors, healthcare access, and hospitalizations.
• Employed a latent class model to group municipalities based on indicator patterns.
• Analyzed data in 2014.

Main Results:

• The latent class model identified three distinct classes of cities/towns.
• These classes represented varying levels of socioeconomic and health indicators.

Conclusions:

• A simplified, adaptable model can identify chronic disease patterns in local communities.
• Findings enable targeted resource allocation for public health programs, potentially uncovering overlooked needs.