Methodology

Every number on this site is traceable to a hospital's published file. The comparison rules are public, and rows excluded from comparison carry a reason. This page summarizes the method; the sub-pages document the rules.

• What the price types mean — list price, cash price, negotiated rate, and why they are never ranked against each other.
• How comparison works — the comparison key, the 3-hospital floor, and all 11 blocker reasons.
• How scores work — the data usability score, its five components, and the exact band thresholds.

Where the numbers come from

1. Hospitals publish machine-readable standard-charges files, as required by federal price transparency rules. Every hospital page links to the hospital's own source file.
2. We ingest and normalize those files through an open data pipeline: structural parsing first, then semantic normalization (code systems, payer names, amount classification, service context).
3. Comparison rules are applied in the data layer, not in this website. The site reads pre-computed, documented tables; it cannot introduce new comparisons or reclassify anything. Every rule described here is enforced and tested upstream.
4. This site presents the results with the comparison status, hospital count, and confidence attached to every number.

The commitments behind every page

• Corpus-bounded claims. Every statistic covers only the named corpus (currently Nashville metro). Nothing here is a regional or national benchmark.
• One price type at a time. List prices, cash prices, and insurer negotiated rates are never ranked against each other.
• The 3-hospital floor. No market statistic is computed from fewer than 3 hospitals reporting the exact same service context.
• Blocked, never hidden. Rows that fail a comparison rule stay visible with a named reason. See data quality.
• Usability, not compliance. Hospital scores measure how usable a published file is for comparison — they are not legal-compliance findings and say nothing about care quality.
• Traceability. Every page links to definitions (glossary), the data behind it (downloads), and the hospital's own source file.

Verifying our work

The full data dictionary and every table this site uses are on the downloads page, including a build identifier that ties the published numbers to an exact version of the open-source pipeline. If you believe a number is wrong, the about page explains how to reach us and what we will check.