Assessing health insurance coverage in Florida using the behavioral risk factor surveillance system

Area of Science:

• Biostatistics
• Public Health
• Health Services Research

Background:

• Large-scale surveys like the Behavioral Risk Factor Surveillance System (BRFSS) offer state-level health insurance coverage data.
• These surveys are not designed for reliable small-area or subpopulation inferences, creating an information gap.
• Accurate health insurance coverage data is crucial for public health planning and resource allocation.

Purpose of the Study:

• To investigate health insurance coverage in Florida at the county level and for specific demographic subpopulations (age, gender, race).
• To address the limitations of traditional survey data for small geographical regions and subpopulations.
• To employ advanced statistical methods for more precise and reliable local-level estimates.

Main Methods:

• Utilized Bayesian predictive inference for finite population quantities, avoiding necessary approximations in other methods.
• Conducted rigorous model diagnostics, including residual checks, cross-validation, and formal concordance tests.
• Investigated potential selection bias and the role of survey weights in improving inferences.

Main Results:

• Generated county-level health insurance coverage estimates for Florida, including variations across subpopulations.
• Visualized results using choropleth maps to illustrate spatial patterns and map variation.
• Compared generated estimates with those from BRFSS and the US Census Bureau's SAHIE program.

Conclusions:

• Bayesian predictive inference provides a robust approach for estimating health insurance coverage at granular levels.
• The method overcomes limitations of traditional surveys for small-area and subpopulation analysis.
• Findings offer valuable data for targeted public health interventions and policy development in Florida.