Methodology

Sources

Hospital directory data is the CMS Hospital General Information dataset (the same data behind Medicare Care Compare), refreshed when CMS republishes it. Price data comes from each hospital's own machine-readable file of standard charges, published under 45 CFR Part 180. We find the current file for each hospital through the cms-hpt.txt index the hospital is required to keep at the root of its website, falling back to a community-maintained registry of file addresses.

How prices are extracted

Hospitals publish in the CMS template layouts (CSV wide, CSV tall, or JSON, versions 2.x and 3.x). Our reader scans each file for a tracked list of common procedure codes (CPT/HCPCS, plus MS-DRG bundles for childbirth) and records: gross charge, discounted cash price, the file's stated minimum and maximum negotiated charges, and every payer-specific negotiated dollar rate (or the file's estimated allowed amount for that plan, where the hospital provides one). The "typical insurance price" is the median of those per-plan rates, and the negotiated range combines the file's stated minimum and maximum with the lowest and highest per-plan rates found. Obvious data errors (zero or negative amounts, and values above a per-procedure sanity ceiling) are discarded.

Limitations