The Atlas is one worked example of a general idea: take fragmented public health data, combine it per person, and make hidden burden visible. The same engine points at new questions, new datasets, and far more people.
Each of these is the same map, asked a different question, powered by a different public dataset:
Map how burden concentrates across U.S. states and communities.
Attach the real economic weight that sits behind each burden group.
Track how prevalence shifted through the Long-COVID era.
Sleep and daily rhythms minute by minute, for a finer picture of fatigue.
As consented individual data matures, follow real people rather than snapshots.
Point the same approach at any invisible-illness area that needs to be counted.
The Atlas fuses several kinds of measurement on the same person only within a single survey. The lenses above are aggregate layers — a fair, honest use of each dataset, not a claim that we link individuals across federal systems.
Today the map holds about 3,919 people. Pooling more survey years, or adding larger datasets, pushes that into the hundreds of thousands and beyond. The map already renders a million people smoothly — you can try that live on the main page.
Under the hood it uses the same drawing technology behind today’s largest interactive data maps, so going from thousands to millions is a solved problem, not a rebuild. The method scales; only the number of dots changes.
Nothing about the Atlas is locked to us. It is designed to line up with the open standards health data already uses, so a future version can plug into records and public-health systems rather than sitting beside them. The goal is infrastructure the whole field can reuse, not a one-off demo.