Medicare Cost Report Data (HCRIS)

The Skilled Nursing Facility Medicare Cost Report data is available to researchers in machine-readable format through the Healthcare Cost Report Information System (HCRIS), which is maintained by the Centers for Medicare and Medicaid Services (CMS). Historically these files could be purchased from CMS in a user-friendly format. Now, however, the data are only available in a very user-unfriendly "variable length" format.

CMS organizes the data by federal fiscal year, i.e., cost report periods beginning during a specific federal fiscal year. Each year's data are updated quarterly, usually in the month following the end of the federal fiscal year quarter.

We routinely download and archive the skilled nursing facility (SNF) Medicare Cost Report data from the HCRIS system. To identify which fields are captured by HCRIS, download the Medicare Cost Report Form CMS 2540-96. Fields marked with an "x" are captured in the HCRIS system. Variables are defined by unique combinations of schedule (or worksheet), row, and column. The schedules and worksheets from which the data are drawn can be loosely categorized by letter as follows:

S - provider information and statistical data such as days, discharges, length of stay, staffing, and RUGs
A - cost data from the provider's general ledger
B - overhead cost allocation to the revenue cost centers
C - ratio of cost to charges
D - computations of routine cost
E - computations of reimbursement settlements
G - balance sheets
H - home health agency data
I - RHC and FQHC services
J - outpatient rehab cost
K - hospice data

The standard extract file includes the 6 digit CMS provider number, cost report status (as-submitted, desk reviewed, audited, re-opened, etc.) and all of the available fields from Worksheets A, B, B1, C, D, E, S, S-2, S-3, and S7-IV in the Medicare Cost Report Form CMS 2540-96.

Data from other schedules can be added on a custom basis and we can also link to other datasets such as OSCAR. Beware that some of the worksheets/fields are not widely used and therefore have very little data (i.e., are very sparsely populated).

Please contact us for a price quote or if you have any questions.