Nason's picture
Upload README.md with huggingface_hub
4cb96fd verified
metadata
license: other
license_name: public-domain-with-cpt
license_link: LICENSE
task_categories:
  - tabular-classification
  - tabular-regression
tags:
  - medicare
  - cms
  - healthcare
  - physician
  - duckdb
  - government-data
  - medical-billing
pretty_name: CMS Medicare Physician & Other Supplier Database
size_categories:
  - 100M<n<1B

CMS Medicare Physician & Other Supplier Database

A clean, queryable DuckDB database built from the CMS Medicare Physician & Other Practitioners Public Use Files -- provider-level Medicare Part B claims data from CY2012 through CY2023.

121,707,609 rows across 3 tables covering what every physician billed, what Medicare paid, and how many services and beneficiaries per NPI per HCPCS code.

Built with cms-medicare-database.

Quick Start

DuckDB CLI

INSTALL httpfs;
LOAD httpfs;
ATTACH 'https://huggingface.co/datasets/Nason/cms-medicare-database/resolve/main/cms_medicare.duckdb' AS cms (READ_ONLY);

-- Total Medicare spending by year
SELECT year, ROUND(SUM(line_srvc_cnt * avg_medicare_payment_amt) / 1e9, 2) AS total_spending_billions
FROM cms.physician_services
GROUP BY year ORDER BY year;

Python

import duckdb
con = duckdb.connect()
con.sql("INSTALL httpfs; LOAD httpfs;")
con.sql("""
    ATTACH 'https://huggingface.co/datasets/Nason/cms-medicare-database/resolve/main/cms_medicare.duckdb'
    AS cms (READ_ONLY)
""")
con.sql("SELECT * FROM cms.physician_services LIMIT 5").show()

DuckDB uses HTTP range requests, so only the pages needed for your query are downloaded.

Tables

Table Description Rows
physician_services Provider-level Medicare Part B claims: one row per NPI per HCPCS code per place 106,515,734
physician_summary Provider-level aggregate summary: one row per NPI per year with total services, 12,232,194
geography_service Geographic aggregate: Medicare utilization and payment by state/national level, 2,959,681

Data Source

CMS Medicare Physician & Other Practitioners PUF -- maintained by CMS. Updated annually. Public domain U.S. government data. HCPCS descriptions include AMA CPT content used under CMS license.

License

Database build code: MIT. Underlying data: public domain (U.S. government work). Note: HCPCS descriptions contain AMA CPT content included as provided by CMS in the PUF.

GitHub

Full source code, build instructions, and data dictionary: github.com/ian-nason/cms-medicare-database