← Austin evidence ledger

Version 1.0 · July 2026

How a surgery-price claim earns the right to be published

The hard problem is not finding a dollar sign. It is preserving enough context to avoid comparing a surgeon-only teaser with an all-in episode price.

01

A listing is not an average

We publish a provider’s own statement as an individual observation. We do not convert a small or mismatched sample into a metro average.

02

Unknown means unknown

If a source does not clearly address anesthesia, facility, follow-up, prescriptions, testing, or revisions, the component remains unknown.

03

Comparability comes before arithmetic

A market range needs at least five independent organizations and a consistent procedure, intent, laterality, setting, and observation type.

04

Every public claim retains provenance

The canonical URL, publisher, retrieval time, exact evidence excerpt, and SHA-256 content fingerprint stay attached to the observation.

05

Directory records are leads, not evidence

NPPES and similar public datasets help build the provider census. They do not prove a price, inclusion, outcome, or quality claim.

06

Private documents follow a separate path

Patient documents are malware-screened, redacted, reviewed, and consent-controlled. Demo records can never be promoted into public evidence.

Corrections and source updates

Source agents can detect changes, but new or changed claims remain candidates until a named review. Published observations are versioned rather than silently overwritten. Providers can submit an attested correction through the provider portal.

Submit a provider package