About
About the data finder
CMS publishes hundreds of free datasets across three websites. Finding the right one usually means knowing its official name. The data finder lets you describe what you need in everyday words and points you to the right file.
Runs in your browser. Nothing you type is uploaded or tracked.
What is in it
Every dataset listed in the public catalogs of these three sites, checked nightly:
- data.cms.gov: 159 datasets, mostly Medicare claims summaries, enrollment and provider lists.
- The Provider Data Catalog: 237 datasets behind Care Compare, such as hospital and nursing home ratings.
- data.medicaid.gov: 84 datasets on Medicaid and CHIP enrollment, drugs and quality.
Yearly or weekly files of the same data are grouped into one entry, with the older releases listed on its page. Test files and internal tables that show up in the catalog feeds are left out.
How search works
Search matches your words against dataset titles, summaries, keywords and column names. It also knows a list of everyday words and the CMS terms they map to, so "nursing homes" also finds "skilled nursing facility" and "drug costs" also finds "drug spending". If you name a state, results that have a state column move up, since most CMS datasets cover every state in one file.
Privacy
The search index is a file on this site. Your browser downloads it once and searches it locally, so nothing you type leaves your device. There are no analytics, no cookies and no third party scripts, and the page's security policy only lets it load files from this site. If you pick a light or dark theme, that choice is saved in your own browser.
Limitations
- It covers the public catalogs of data.cms.gov, the Provider Data Catalog and data.medicaid.gov only. Files posted only on cms.gov program pages, and restricted files such as LDS or RIF data, are not in it.
- The index is rebuilt once a night. CMS can post a new file between runs, so check the official page before you rely on a date.
- Search matches words, not meaning. It knows a hand-made list of everyday terms, but it can still miss a dataset that uses different wording. Try a second phrasing or browse by topic.
- Naming a state moves datasets with a state column up. It does not filter the rows; you filter the file or the CMS API yourself.
- Column meanings come from CMS's machine-readable dictionaries. The Provider Data Catalog publishes only column headers in that form, so for its datasets the full definitions are in the PDF dictionary on the official page.
- Short summaries for 58 widely used datasets were written by hand from CMS's own text. Everything else uses the first sentence of CMS's description.
- This is an unofficial tool, not affiliated with or endorsed by CMS. Cite CMS, not this site.
Open source
The code and the hand-written notes are free to reuse under the Apache License 2.0. The data comes from CMS and is in the public domain.
The data finder is part of the CMS data toolkit. More of Pradeep's work is at pradeeppeddineni.com.