About AskPublicHealth

AskPublicHealth is a natural-language query interface for US public health data. Ask a question in plain English and get a precise, cited answer drawn directly from authoritative government datasets — no guessing, no hallucination.

Why not just use ChatGPT?

General-purpose AI chatbots are trained on text scraped from the web. When you ask ChatGPT about county-level diabetes rates, it draws on whatever health statistics happened to appear in its training corpus — which may be years out of date, aggregated imprecisely, or simply made up. AskPublicHealth is different in three fundamental ways:

• Live government data, not training memory. Every answer is computed at query time from the actual CDC, NCHS, or County Health Rankings release files — not recalled from a model's compressed memory of the internet.
• County- and tract-level precision. ChatGPT cannot tell you the diabetes prevalence in Travis County, TX across its 289 census tracts. AskPublicHealth can, because it queries the underlying microdata directly.
• Verifiable, cited results. Every answer includes the dataset it came from and the year of the release. The SQL query used to compute the result is available so you can audit or reproduce it independently.

Datasets

County Health Rankings (2025)

County-level health outcomes and social determinants for all US counties — poverty, income, education, smoking, obesity, premature death, uninsured rate, air quality, violent crime, food access, and more. countyhealthrankings.org

CDC PLACES (2024)

Census-tract-level prevalence estimates for 29 chronic disease and prevention measures across all US census tracts — diabetes, obesity, depression, heart disease, smoking, physical inactivity, and more. cdc.gov/places

BRFSS (2024)

State-level Behavioral Risk Factor Surveillance System data — health behaviors and self-reported outcomes including smoking, alcohol use, exercise, diet, and preventive care utilization. cdc.gov/brfss

NCHS Teen Birth Trends (2019)

State-level teen birth counts and rates by age group (15–17, 18–19, 15–19 combined) for all 50 states and DC, covering 1990–2019. cdc.gov/nchs

Limitations

• Data reflects the most recent available release; there may be a lag of 1–2 years from current conditions.
• AI-generated queries may occasionally produce incorrect results. Always verify important findings against the primary source.
• Not for clinical diagnosis, treatment planning, or real-time surveillance.

Built by Murwan Abdelgafour.