Data sources

CMS publishes every organ procurement organization (OPO) cost report (form CMS-216-94) in one archive, OPO94-ALL-YEARS.zip at downloads.cms.gov, and refreshes it a few times a year. Each year inside it is three files: a numeric file (the amounts), an alpha file (the filer's own words, including every write-in label) and a report file (which report record belongs to which provider, for which fiscal period, with what status). A cell is keyed on report record number, worksheet, line and column. Anyone can pull the same files from the CMS cost report page.

Transplant hospitals file form CMS-2552-10. Worksheet D-4, one per organ program, carries the count of organs acquired from OPOs, the organ acquisition cost pool and the Medicare share. The form's own wording for each line is in that figure's citation. The same archive structure applies: numeric, alpha and report files, keyed the same way. Hospital fiscal years follow the same period-end rule as the OPO reports.

HRSA publishes the donation service area of every OPO by county. This site uses it for identity: which OPO a provider number belongs to and where it serves. No figure on this site is derived from geography.

Most OPOs file a Form 990. The figures here come from the IRS e-file XML, and each chip names the e-file object id and the XML element it was read from. Hospital-based OPOs file no return of their own. A return filed by a parent is marked as such on the row.

Every cost report in the loaded years, by report record number, with the provider it belongs to, its fiscal period and its status. The one marked in cohort for a period is the one the figures use.

Each source below is a published file. Where we hold the exact bytes, the sha256 is printed beside it, so the file you download can be checked against the one that was read. The date is when the link was last tested, not when the page was written.

This site carries money. It publishes what organ procurement organizations and transplant hospitals recorded on their cost reports and tax filings, and nothing else. It holds no measure of how well any organization performs and no outcome for any patient. Those are published, by others, and they are public.

Reading cost and outcome together is a separate piece of work. Cost on its own says what was spent. Outcome on its own says what happened. Neither answers what the other cost, and joining them means reconciling two record systems that count different things over different periods. That join is what Atlas does, also at 2460 Health Tech. Econ stays the money, so that a figure here can be checked against a filing without anything else having to be agreed first.