1. Summary
- 65 of 14,690 certified nursing homes (0.44%) disclose a private-equity company in an ownership or control role in their own PECOS filing, and 73 (0.5%) disclose a REIT in such a role. Counting every disclosable party, including landlords and lenders, private equity appears for 97 facilities (0.66%) and a REIT for 577 (3.93%, 61,235 certified beds). Most REIT disclosures are landlords, not operators.
- These are disclosure rates, not ownership rates. GAO estimated that about 5% of nursing homes had a private-equity owner in 2022, and Chen and colleagues found the new CMS ownership data captured roughly a third of private-equity and under a fifth of REIT investments. The gap between those estimates and the 0.44% here is itself the finding: the disclosure system, three years after the 2023 rule, records a fraction of what researchers can see from other sources.
- Disclosure is concentrated. 13 states have at least one facility disclosing a private-equity owner; Pennsylvania alone has 24. The most-disclosed private-equity owners are small: EMSIR LLC is named by 7 facilities, and 28 facilities disclose a private-equity owner whose name CMS left blank in the public file.
- The facilities that disclose a private-equity owner look worse on every descriptive measure: an average overall rating of 2.30 against 2.99 for facilities with no such disclosure, 63.1% against 46.1% with a fine or payment denial, and 12.3% against 3.5% in the Special Focus Facility program. The group is small (65 facilities) and the comparison controls for nothing; it describes, it does not explain.
- 1,399 facilities (9.5%) changed hands in the 36 months before the vintage, on the PECOS change-of-ownership file. Recorded changes rose from 240 in 2016 to 873 in 2023. Care Compare's own twelve-month flag marks only 113 facilities.
- The federal files disagree with each other 2,209 times. 475 certified facilities match no PECOS enrollment at all; 293 disclose owners with no ownership percentage on any row; 1,437 current enrollments list no ownership-interest party (1,275 of them non-profit or government homes, where that is expected, and 162 for-profit). The register is public and per facility.
2. Why ownership, and why "as disclosed"
Who owns a nursing home shapes who is accountable for it. Federal payment rules, state licensing and the courts all reach the operator of record; the investors behind that operator, and the landlord who owns the building, are harder to see and harder to hold to account. Since 2023, CMS has required skilled nursing facilities to name on Form CMS-855A whether any owner or other disclosable party is a private-equity company or a real estate investment trust, and since late 2024 it has published those answers in the PECOS Skilled Nursing Facility All Owners file.
Two independent assessments frame what that file can and cannot show. The Government Accountability Office found that CMS's data "do not provide a means of readily identifying private equity firms" and that "there is no single, authoritative data source" on nursing home owners; it estimated that about 5% of homes had a private-equity owner in 2022. Chen, Skinner, Braun, Konetzka, Stevenson and Grabowski, writing in Health Affairs in 2024, matched the new CMS ownership data against independently identified investments and found it captured about one-third of private-equity and under one-fifth of REIT investments, with ownership percentages missing for most owners of the largest chains.
This report therefore counts one thing only: what each facility disclosed. A facility with no private-equity or REIT flag has not reported one. That is not a finding that it has none, and no sentence here should be read that way. The value of counting disclosures is that the count is reproducible, per facility, from a public federal file, and that its gaps are themselves measurable.
3. Data and method
Every number comes from release v2026.08 of The Care Ratings Open Nursing Home Data, built from the CMS August 2026 Provider Information, Penalties and Ownership files, the PECOS enrollment, All Owners and Change of Ownership files and the Nursing Home Chain Performance Measures file, with every source file's vintage and checksum recorded in the release manifest and the release archived at Zenodo. The tables were produced by one command, tcr-open-data ownership-study, and are committed as CSV in the public repository; section 11 of the methodology defines each one.
Owner versus party. PECOS lists every disclosable party with a role code. We call a party an owner when its role is an ownership or control role: direct or indirect ownership interest (codes 34, 35, 85, 86), partnership interest (38, 39) or operational or managerial control (43). Every other role, including the additional disclosable parties the 2023 rule created, is a party. A REIT that owns the building is a party; a REIT that holds an ownership interest in the operator is an owner. Headline figures use owner roles; the any-role figures are shown beside them.
Disclosure groups. Each facility falls into one group: discloses a private-equity owner; discloses a REIT owner; private equity or REIT among other parties only; no PECOS enrollment matched; no private-equity or REIT disclosure. Facilities with both a private-equity and a REIT owner are counted in the private-equity group.
Comparisons. Quality and enforcement figures are averages of Care Compare columns for each group. They are descriptive: no adjustment for case mix, size, region, chain membership or the reasons a facility changed hands. The private-equity group has 65 facilities, small enough that a handful of homes move its averages. We present the same comparison within for-profit facilities only, which removes the largest confounder, and we present no significance tests, because the question is what the disclosing facilities look like, not what private equity does to them.
4. How many facilities disclose a private-equity or REIT owner
Of 14,690 certified nursing homes, 14,215 match a PECOS enrollment whose ownership rows can be read. 65 (0.44%) name a private-equity company in an ownership or control role, holding 7,404 certified beds. 73 (0.5%) name a REIT in such a role.
The picture changes when every disclosable party counts. Private equity then appears for 97 facilities (0.66%), and a REIT for 577 (3.93%), with 61,235 beds. The difference is almost entirely landlords: 504 facilities disclose a REIT only as a party, which in these files means the owner of the real estate or a lender. A study that reported "3.93% of nursing homes are REIT-owned" would be wrong on its own terms; the file says 0.5% disclose a REIT as an owner.
| Flag | Owner or control role | Other parties only | Any role | Owner role, % of facilities | Any role, % of facilities |
|---|---|---|---|---|---|
| Private equity | 65 | 32 | 97 | 0.44% | 0.66% |
| REIT | 73 | 504 | 577 | 0.50% | 3.93% |
For context, 74% of facilities are for-profit under Care Compare's ownership type and 68.9% carry a CMS chain id. Private-equity and REIT disclosure sits inside the for-profit, chain-affiliated part of the sector: 100.0% of the facilities disclosing a private-equity owner are for-profit and 83.1% are in a CMS-identified chain.
5. Where they are
Disclosure of a private-equity owner is concentrated in a few states. 13 states have at least one such facility. Pennsylvania has 24 (3.66% of its facilities); Washington has 11 (5.70% of its facilities); South Dakota has 6 (6.25% of its facilities); California has 4 (0.34% of its facilities); Ohio has 4 (0.43% of its facilities); Oregon has 4 (3.13% of its facilities). The state map shows each layer; the table there carries every state.
REIT disclosure in any role is more widely spread and follows where the large landlords hold portfolios: District of Columbia 17.7%, Oregon 15.6%, Indiana 14.8%, Texas 12.2%, West Virginia 12.2%, Maine 11.5%. For-profit share runs from 6% in North Dakota to 91% in Arkansas, with Alaska and Minnesota at the low end and Oklahoma and California at the high end.
| State | Facilities | Disclose a PE owner | % of facilities | Disclose a REIT, any role | For-profit | Changed hands, 36 mo |
|---|---|---|---|---|---|---|
| Pennsylvania | 656 | 24 | 3.66% | 2.74% | 63.0% | 13.9% |
| Washington | 193 | 11 | 5.70% | 5.18% | 79.3% | 21.8% |
| South Dakota | 96 | 6 | 6.25% | 1.04% | 32.3% | 6.3% |
| California | 1,165 | 4 | 0.34% | 1.29% | 87.0% | 5.9% |
| Ohio | 922 | 4 | 0.43% | 2.06% | 84.9% | 6.3% |
| Oregon | 128 | 4 | 3.13% | 15.63% | 86.7% | 29.7% |
6. Who the disclosed owners are
The private-equity owners named in the file are not the national firms that dominate the policy debate. EMSIR LLC (7 facilities, PA); LADS AVENUE ASSOCIATES LLC (6 facilities, PA); 2020 GSR DYNASTY LLC (5 facilities, OH PA); AH DYNASTY LLC (5 facilities, OH PA); KJA PINNACLE LLC (4 facilities, PA). A further 28 facilities in 6 states disclose a private-equity owner whose organization name is blank in the public file; those rows can be identified only by associate id. Whether the large sponsors are absent because they hold no reportable interest, because facilities have not reported them, or because the flag is being applied narrowly cannot be told from the file, and we do not guess.
REITs are the opposite: a few large, well-known landlords. WELLTOWER INC appears for 113 facilities in 8 states (0 owner-role rows, 164 other-party rows); TRILOGY REAL ESTATE INVESTMENT TRUST appears for 72 facilities in 5 states (20 owner-role rows, 69 other-party rows); AMERICAN HEALTHCARE REIT INC appears for 68 facilities in 5 states (21 owner-role rows, 66 other-party rows); SABRA HEALTH CARE REIT INC appears for 21 facilities in 5 states (6 owner-role rows, 17 other-party rows); WELLTOWER OP, LLC appears for 16 facilities in 4 states (1 owner-role rows, 25 other-party rows). The owner-role rows are the cases in which a REIT also holds an interest in the operator; the other-party rows are the leases and loans.
Chains are where the two meet. Among the largest CMS-identified chains with any disclosure, the counts are: PACS GROUP, 274 facilities, 0 disclosing a private-equity owner and 3 a REIT in any role; CREATIVE SOLUTIONS IN HEALTHCARE, 149 facilities, 0 disclosing a private-equity owner and 1 a REIT in any role; SABER HEALTHCARE GROUP, 126 facilities, 0 disclosing a private-equity owner and 34 a REIT in any role; TRILOGY HEALTH SERVICES, 123 facilities, 4 disclosing a private-equity owner and 96 a REIT in any role; AVIR HEALTH GROUP, 116 facilities, 0 disclosing a private-equity owner and 93 a REIT in any role. Every chain has a profile listing its facilities and their disclosures.
7. Quality and enforcement by disclosure group
Table 3 describes each disclosure group on Care Compare's own measures. The facilities that disclose a private-equity owner have an average overall rating of 2.30, against 2.99 for facilities with no private-equity or REIT disclosure; 66.7% of them are rated one or two stars, against 40.7%. 63.1% have at least one fine or payment denial in the three-year lookback, against 46.1%, and their average fines are $52,110 against $30,872. 12.3% are Special Focus Facilities or candidates, against 3.5%, and 20.0% carry the abuse icon, against 9.5%. Reported nurse staffing is close: 3.84 against 3.86 hours per resident day.
Facilities that disclose a REIT owner look different again: an average rating of 3.22 and 42.0% with a penalty, slightly better than the no-disclosure group on both. Facilities with private equity or a REIT among other parties only sit between, at 2.74 and 53.9%.
Within for-profit facilities only, the private-equity group's average rating of 2.30 compares with 2.81, and 63.1% with a penalty compares with 48.4%. The gap narrows but does not close. Three cautions apply. The private-equity group is 65 facilities. It is concentrated in a few states and chains, so state survey practice and chain management travel with it. And the direction of cause is unknowable here: investors may buy troubled homes, or homes may become troubled after a purchase, and this file cannot tell those apart.
| Group | Facilities | Avg overall rating | Rated 1 or 2 stars | Avg staffing rating | Nurse hours / resident day | Nurse turnover | Any penalty | Avg fines | SFF or candidate | Abuse icon | For-profit |
|---|---|---|---|---|---|---|---|---|---|---|---|
| No PE or REIT disclosure | 13,552 | 2.99 | 40.7% | 2.92 | 3.86 | 45.9% | 46.1% | $30,872 | 3.5% | 9.5% | 74.2% |
| PE or REIT among other disclosable parties only | 529 | 2.74 | 49.4% | 2.48 | 3.56 | 47.7% | 53.9% | $39,569 | 3.8% | 13.6% | 86.2% |
| No PECOS enrollment matched | 475 | 3.07 | 38.6% | 3.30 | 4.23 | 40.8% | 37.7% | $25,552 | 4.6% | 11.2% | 49.1% |
| Discloses a REIT owner | 69 | 3.22 | 32.4% | 3.07 | 3.94 | 46.6% | 42.0% | $27,750 | 2.9% | 4.4% | 94.2% |
| Discloses a private-equity owner | 65 | 2.30 | 66.7% | 2.83 | 3.84 | 48.8% | 63.1% | $52,110 | 12.3% | 20.0% | 100.0% |
| Group | Facilities | Avg overall rating | Rated 1 or 2 stars | Avg staffing rating | Nurse hours / resident day | Nurse turnover | Any penalty | Avg fines | SFF or candidate | Abuse icon |
|---|---|---|---|---|---|---|---|---|---|---|
| No PE or REIT disclosure | 10,058 | 2.81 | 45.8% | 2.63 | 3.69 | 47.2% | 48.4% | $34,914 | 4.2% | 10.4% |
| PE or REIT among other disclosable parties only | 456 | 2.72 | 49.9% | 2.47 | 3.58 | 47.8% | 55.7% | $41,428 | 3.5% | 14.3% |
| No PECOS enrollment matched | 233 | 2.69 | 49.8% | 2.66 | 3.56 | 43.4% | 45.1% | $35,270 | 7.3% | 14.6% |
| Discloses a REIT owner | 65 | 3.19 | 34.4% | 3.10 | 3.93 | 47.1% | 43.1% | $29,124 | 3.1% | 4.6% |
| Discloses a private-equity owner | 65 | 2.30 | 66.7% | 2.83 | 3.84 | 48.8% | 63.1% | $52,110 | 12.3% | 20.0% |
8. Facilities changing hands
The PECOS Change of Ownership file records 240 changes with an effective date in 2016 and 873 in 2023; the file is cut in the summer, so the latest year is partial. 1,399 facilities (9.5%) changed hands in the 36 months before the vintage, 0 of them more than once.
Care Compare publishes its own flag, "changed ownership in the last 12 months", and it marks 113 facilities (0.77%). In the monthly history the share carrying that flag fell from 2.21% in 17 January 2019 to 0.77% in 26 August 2026, while the PECOS file shows more transactions, not fewer. The two measures are built differently and neither is wrong, but a reader who takes the Care Compare flag as the rate of ownership change will underestimate it several times over.
9. What the federal files leave out
Three gaps set the limits of everything above, and each is measured in the discrepancy register.
Facilities with no readable filing. 475 certified nursing homes (3.2%) match no Skilled Nursing Facility enrollment in the PECOS files, usually because the facility enrolled under a new number after a sale and the files have not caught up. Their disclosures, whatever they would be, cannot be read.
Owners without percentages. 293 facilities disclose ownership-interest owners with no percentage on any row, so the share each owner holds is unknown. A further 1,437 current enrollments list officers, directors, managing employees or other parties but no ownership interest at all; 1,275 of those are non-profit or government homes with no equity to report, which leaves 162 for-profit facilities whose filing names no owner.
A turnover series that measures reporting, not turnover. The Care Compare ownership file, the one on the public website, relabelled its owner roles in the December 2024, June 2025 and May 2026 vintages and gained new categories under the 2023 rule. Counted on the printed label, 143,990 owner relationships "first appear" in the 2026 files; 67% of them are rows in the added categories, 3% are the same name under a new label, and 24% are new names, most of them arriving in two vintages as fuller filings under the revised Form CMS-855A reached the file. We publish the decomposition so that no one, including us, reads that series as a wave of new owners.
| Check | Facilities |
|---|---|
| No PECOS enrollment matched to the CCN | 475 |
| PECOS enrollment lists no ownership-interest party | 1,437 |
| Ownership-interest owners disclosed without any ownership percentage | 293 |
| PE or REIT owner disclosed in PECOS, Care Compare lists individuals only | 3 |
| Chain id in Provider Information missing from the chain file | 0 |
| Ownership changed in the last 12 months per Care Compare, no PECOS change of ownership in 36 months | 1 |
10. What states are doing
Since January 2025, state legislatures have passed a wave of laws on health care ownership disclosure and transaction review, most of them aimed at hospitals and physician practices rather than nursing homes. Connecticut's Public Act 26-103, signed on June 4, 2026, is the one enacted law aimed squarely at nursing homes: from October 2026 Medicaid-contracted homes must report any investment entity, including a REIT, with a 5% beneficial interest, and from 2028 must attest that no such entity controls resident care and, where the state identifies a feasible instrument, carry a surety bond covering 90 days of operating costs. Virginia's SB 808, which would require nursing facilities to disclose related-party rent, was carried over to 2027. Illinois extended its Attorney General's pre-merger notice to private-equity companies that own health care facilities. The state law tracker lists each bill with the legislature's own record as its source.
11. Limitations
- Disclosure, not ownership. Every count is of what facilities reported. Independent research finds the reported share well below the actual share; nothing here narrows that gap.
- One vintage. The report describes the CMS August 2026 files. Enrollments, flags and chain ids change monthly; the release is versioned so the numbers can be reproduced, and later editions will show change over time.
- Descriptive comparisons. Group averages control for nothing and the private-equity group is small. They describe the facilities that disclose; they do not estimate an effect of ownership on care.
- Names as printed. Owner and chain names are the strings in the CMS files. We have not harmonized subsidiaries, funds or affiliates into parent companies, and CMS leaves some organization names blank.
- Chain identity is CMS's. The chain file and Provider Information are cut on different days, so a chain's facility count can differ by a few between them; both are shown.
- Percentages are often missing. Where the file has no ownership percentage we do not impute one, so no figure here is weighted by share of ownership.
12. Reproduce, cite, correct
Download release v2026.08, install the open-source build (github.com/ttiatun/careratings-open-data) and run tcr-open-data ownership-study --release releases/v2026.08 --history history --out analysis/ownership/v2026.08. The tables it writes are the ones committed in the repository and read by this page.
Cite this report: The Care Ratings Team (2026). Who owns America's nursing homes, as disclosed to CMS: Ownership study, first edition (version 1.0, release v2026.08). The Care Ratings. https://thecareratings.com/data/ownership/report-2026/ Data: https://doi.org/10.5281/zenodo.22780105
Authorship and review. Prepared by the Care Ratings Team under the program's editorial standards and reviewed before publication by the TCR Editorial Team. An external methodological reviewer is being engaged and will be named here in a later version; until then this edition has had no review outside The Care Ratings. The Care Ratings has no facility, chain, investor or trade-association funding. Errors are corrected in place and logged in the corrections log; report one through the contact page, naming the section and figure.
