Specialty facilities

Medicare-certified specialty facilities from CMS Care Compare: outpatient dialysis clinics from dim_dialysis_facility and fct_dialysis_quality, inpatient rehabilitation facilities from dim_irf, and long-term care hospitals from dim_ltch. All three populations sit outside the acute- care dim_hospital universe — see the caveats for the CCN-namespace details. Every view below reads a pre-aggregated source query; the raw facility and measure grains never leave the warehouse.

Dialysis facilities

7,490

Inpatient rehab facilities

1,222

Long-term care hospitals

311

Avg dialysis five-star

2.87

Dialysis facilities

Reach and services

7,490 Medicare- certified outpatient dialysis facilities operate across 56 states and territories, running roughly 132,735 dialysis stations combined. In-center hemodialysis is nearly universal; peritoneal dialysis is offered by a bit over half of clinics, home hemodialysis training by roughly a third, and a late shift by fewer than one in six.
Loading...

The chain landscape

Two operators — DaVita and Fresenius Medical Care — run roughly three-quarters of every Medicare-certified dialysis facility in the country between them. Independent clinics (no chain affiliation) are the third-largest cohort, and everything else is a long tail. Both DaVita's and Fresenius's chain-averaged five-star rating sits within a tenth of a point of the national dialysis mean; the Kaiser Permanente-run clinics are a small (22-facility) outlier, averaging nearly a full star higher.

Loading...
No Results

Five-star ratings

CMS assigns each dialysis facility a 1–5 star rating in whole-star increments based on a composite of clinical measures. The overall distribution centers on three stars, with about

7,490 minus 6,999 facilities unrated in the current vintage (typically new or low-volume clinics).
Loading...

Top states

No Results

Standardized-ratio measures

The five standardized-ratio measures in fct_dialysis_quality compare each clinic's observed outcome (mortality, hospitalization, infection, readmission, ED visits) against a case-mix-adjusted expected value. CMS bins each facility's ratio into "Better than Expected", "As Expected", or "Worse than Expected" using confidence-interval overlap with 1.0; facilities without enough denominator to compute a ratio appear as "Not Available". The infection ratio (SIR) stands out for its lopsided tail: about 2,800 facilities land "Better than Expected" against only 47 "Worse"; the other four ratios sit "as expected" for the large majority of reporting facilities.

Loading...
No Results

Selected percent-of-patient measures

Percent-family dialysis measures publish an unweighted average across reporting facilities. Adherence measures (adequacy of dialysis — the Kt/V targets) run above 90%, long-term-catheter use averages about 19%, and the two anemia measures are asymmetric (hemoglobin under 10 is common, over 12 is rare). Healthcare-worker COVID-19 vaccination adherence stands out as very low in this vintage — mean about 4% at the roughly 3,900 reporting facilities.

No Results

Inpatient rehabilitation facilities

Freestanding vs hospital unit

1,222 Medicare- certified IRFs operate across 52 states and territories. IRFs split into two structurally different populations that both bill under IRF PPS: freestanding rehabilitation hospitals (their own CCN and physical plant) and hospital-based rehab units carved out of an acute-care hospital. Roughly two-thirds are hospital units; freestandings are more concentrated in the largest states — Texas alone has 77 of the 410 freestanding IRFs nationally.
Loading...
No Results

Top states

Loading...

Long-term care hospitals

Population and ownership

311 LTCHs operate across 47 states. Unlike the near-parity ownership mix among IRFs, LTCHs are dominated by for-profit operators (about 70% of the population), with non-profit the runner-up. Bed capacity is populated for 305 of 311 LTCHs and totals 21,765 certified beds nationwide.
Loading...

Bed capacity

Most LTCHs are small — half sit under 50 beds — but a long tail of larger facilities (100+ beds) drives most of the aggregate capacity. The single largest LTCH in the file reports 760 beds.

Loading...

Top states

No Results

Caveats

  • Snapshot vintage. Dialysis directory and quality data come from the Care Compare snapshot as of June 16, 2026 , IRFs as of August 12, 2026 , and LTCHs as of August 12, 2026 . Every figure above is a point-in-time snapshot — the marts don't carry facility-level history.
  • Dialysis measure suppression. fct_dialysis_quality publishes 202,230 rows at the (ccn, measure_code) grain across 27 measures, but not every facility reports every measure. CMS carries an availability code on each row; `is_reported` is true when the score is present (`availability_code = '001'`) and non-reported rows have `NULL` in `score_numeric` (for the ratio measures, CMS labels them with the literal category `Not Available`, not `NULL`). Every reporting count above filters to `is_reported = true`; percent-family averages exclude suppressed rows, and the ratio "Not Available" bucket is the suppressed count itself.
  • CCN namespaces. These three specialty populations are almost entirely disjoint from the acute-care hospital universe. Zero of the 7,490 dialysis facilities share a CCN with dim_hospital; only 5 LTCH CCNs overlap (dual certifications). All 410 freestanding IRFs sit outside dim_hospital; 762 of the 812 hospital-unit IRFs can be traced back to a parent acute-care hospital via the CCN position-3 → '0' rewrite the source file uses, but the remaining 50 units don't reconcile. Cross-mart joins should treat specialty CCNs as their own space.
  • County coverage is FFM-scoped. dim_county is the Federally- Facilitated Marketplace's rating-area geography (30 FFM states), so specialty-facility rows for the 20 non-FFM states carry a county name but no dim_county match. Only about half of dialysis clinics, ~62% of IRFs, and ~64% of LTCHs land in an FFM county.
  • State code counts include territories. The 56 -state dialysis footprint, 52 -state IRF footprint, and 47 -state LTCH footprint count U.S. territories (Puerto Rico, Guam, USVI, etc.) and DC alongside the 50 states, matching the raw CMS files.
  • Unweighted averages. Every star / score / bed average above weights each facility equally, regardless of station count, bed count, or patient volume.