State and regional estimates using seven cycles of pooled nationally representative HINTS data

Area of Science:

• Public Health
• Survey Methodology
• Cancer Research

Background:

• The Health Information National Trends Survey (HINTS) collects national data on public cancer beliefs and behaviors.
• HINTS previously lacked the capability for accurate state and regional estimates.
• Local health estimates are crucial for informing targeted cancer policies and community needs.

Purpose of the Study:

• To develop state- and regional-level estimates from the national HINTS survey.
• To enhance the utility of HINTS data for local public health research and policy.
• To address limitations in existing HINTS data for subnational analysis.

Main Methods:

• Pooled seven cycles of HINTS data (2012-2018).
• Applied standard calibration and raking methods to person-level weights.
• Harmonized and aggregated data across survey cycles to create pooled datasets.

Main Results:

• Generated subnational (state and regional) estimates for key cancer-related outcomes.
• Achieved estimates with small variances and minimal potential bias.
• Produced reliable survey weights for the pooled HINTS datasets.

Conclusions:

• The developed methods successfully create accurate state and regional estimates from HINTS data.
• Pooled HINTS data analysis offers opportunities for detailed local population examination.
• Challenges in pooled data analysis were identified and addressed.