Docket Article
CMS PECOS Ownership Data: Finding the Ultimate Owner
A 2026 guide to CMS PECOS ownership data: reconcile NPI and CCN identities, reported owners, parent flags and CHOW events with a provider-type source matrix and an auditable ownership worksheet.

Summary
- 01PECOS ownership rows identify reported relationships, but a defensible ultimate-parent conclusion requires the correct provider identity, a supported ownership path, a stated parent definition, and compatible dates.
- 02Provider-type coverage differs: hospital guidance leaves the ultimate parent unidentified, while the SNF dictionary includes a reported parent-company flag that requires corroboration. Blank flags do not establish a negative.
- 03Preserve equity, managerial, property, and accounting relationships separately. Calculate indirect ownership only along supported paths; totals across ownership tiers are not a partition of equity.
- 04Keep source periods, ownership-event dates, collection dates, and decision dates separate. Retrieved documentation establishes available fields, while live CMS ownership releases and update cadence remain unverified in the report.
- 05Record supported group or chain membership separately from ultimate-parent status. Missing parents and unresolved records do not establish independence, and even a verified parent does not establish willingness to sell.
Inside this article
- 01Executive Summary
- 02Introduction and Background
- 03Key Changes
- 04Resolve Provider Identity Before Ownership
- 05Choose Provider-Type Files and Interpret Their Rows
- 06Implementation Considerations and Process Changes
- 07PECOS Ownership Worksheet and Decision Outcomes
- 08Reconcile the Graph and Adjudicate the Target
- 09Data Analysis and Evidence
- 10Implications and Future Directions
- 11Frequently Asked Questions (FAQs)
- 12Conclusion
Executive Summary
CMS PECOS ownership data can identify reported owners and support a healthcare ownership screen, but they do not automatically establish the ultimate owner. The Provider Enrollment, Chain, and Ownership System (PECOS) supports Medicare enrollment; a provider identifier, an enrollment, a facility, and a branded platform are different research units. National Provider Identifier (NPI) issuance does not itself establish Medicare enrollment or credentials. [1] The September 2024 hospital guidance states that the ultimate parent is not identified in those data. [2]
The provider-type distinction matters. The retrieved skilled nursing facility (SNF) dictionary includes PARENT COMPANY - OWNER, a reported ultimate-parent indicator, alongside owner roles and percentages. That is a useful assertion to corroborate, rather than a complete ownership graph. [3] Reporting thresholds also require qualification: the current Medicare enrollment definition generally includes owners with 5% or greater direct or indirect ownership, but includes any partnership interest. Separate disclosures can capture other relationships. [4] [5] A blank or absent record therefore cannot establish independence.
A Government Accountability Office (GAO) study estimated private equity ownership of 5% of Medicare-enrolled nursing homes in 2022, with a 95% interval of 2% to 8%. Such population estimates cannot classify an individual target. [6]
A defensible screen resolves identifiers, selects the applicable files, preserves owner roles, records ownership events separately, and corroborates the proposed parent. Legal Entity Identifier (LEI) parent data concern accounting consolidation; Internal Revenue Service (IRS) related-organization filings cover their own reporting periods; state entity searches assist legal-name resolution. Their scopes should remain visible. [7] [8] [9] The worksheet below records supported group or chain membership separately from an auditable ultimate-parent decision: resolved ultimate parent, reported owners with the ultimate owner unresolved, conflicting evidence, or unknown. The research cutoff is October 4, 2026. Accessible dictionaries were checked, but live CMS ownership release metadata was not independently verified for this report; no latest-release or cadence claim is made.
Introduction and Background
For private equity origination, “who owns this healthcare company?” is usually a screening question with operational consequences. The answer may determine whether a location belongs to an existing platform, whether several records should be researched together, or whether ownership remains unresolved. Healthcare ownership due diligence starts with the exact entity being evaluated: the licensed facility, enrolled provider, operating company, holding company, or branded group.
PECOS lookup is an imprecise search term because enrollment access and public ownership research serve different purposes. CMS's provider resource page directs users to NPPES for NPI information and to PECOS for Medicare enrollment updates. NPPES is the National Plan and Provider Enumeration System. Neither an identifier search nor enrollment-system navigation reconstructs a corporate hierarchy. [10]
The public ownership material is more useful when treated as a collection of dated, relational evidence. A reported organization can own an interest, manage a provider, or occupy another disclosed role. Meanwhile, an external parent record can describe consolidation rather than the equity chain. GLEIF, the Global Legal Entity Identifier Foundation, expressly frames its parent relationships as accounting relationships. [7] The practical task is reconciliation across definitions, entities, and dates.
This report concentrates on that reconciliation, rather than general succession research or a broad directory of company-data vendors. Its proposed worksheet and decision rules are analytical recommendations. The distinction between evidence quantity and evidence quality has a useful precedent in Public Company Accounting Oversight Board (PCAOB) audit guidance, but applying it here does not make origination research a financial audit. [11]
Docket describes Triage, Scout, and Audit as separate stages for identity checks, research, and evidence review. The relevant application is a reviewable provider-level ownership decision. A platform label, a source count, or an organization name should never substitute for the underlying relationship evidence.
Key Changes
Ultimate-parent coverage differs by provider type
The first change in analytical practice is to stop applying one historical limitation to every current file. The retrieved SNF dictionary provides a reported parent-company flag. Preserve that field as an assertion with its source version, rather than upgrading it automatically to a verified current ultimate parent. [3]
The same dictionary distinguishes reported private equity and real estate investment trust (REIT) flags. Their documented values use Y if reported and blank if not reported. A blank is not equivalent to a verified “no.” Screening logic should retain those original values and reserve a separate field for the analyst's adjudicated conclusion. [3]
A practical improvement is to separate “CMS-reported parent” from “corroborated ultimate owner.” This prevents the addition of a field from being mistaken for proof of completeness. It also preserves the difference between an affirmative source statement and the research team's assessment of that statement.
Equity owners are only part of the disclosed network
SNF disclosure rules extend beyond equity stakes. The current enrollment definitions include a party leasing or subleasing real property to an SNF among additional disclosable parties. Such a party can be relevant to the operating structure without being the parent of the operator. [4]
The threshold should also remain attached to the applicable rule and role. The Federal Register explanation specifies an interest in a single provider; separate stakes below the threshold are not aggregated across different providers to reach it. Its partnership exception also prevents the shorthand “owners below 5% never appear” from being accurate. [5]
For a healthcare platform ownership structure, retain distinct relationship types for equity ownership, managerial control, property arrangements, and accounting consolidation. GLEIF's ontology describes consolidation as presenting a parent and subsidiaries as a single economic entity. That does not supply each legal intermediate or ownership percentage. [12]
Event evidence and current status need separate clocks
An ownership change is an event with an effective date. A current parent determination is a dated conclusion about a structure. These should not occupy the same worksheet field.
The retrieved SNF change-of-ownership (CHOW) guidance identifies transactions on or after January 1, 2016, for currently approved enrollments, and describes active enrollment coverage rather than a complete historical register. That scope matters when a provider is absent or an older transaction is missing. [13]
External evidence has its own clock. IRS Schedule R considers related organizations present at any time during the tax year. A relationship on that return is not automatically a statement about the publication date. [8] The recommended record therefore keeps source period, event date, collection date, and decision date separately.
Resolve Provider Identity Before Ownership
Define the unit and keep identifiers separate
Start with a written question such as “identify the parent of the operating entity associated with this enrolled facility.” Avoid starting with a brand name and assuming that every location, license, and enrollment belongs to the same company.
An NPI is a 10-digit provider identifier, not an ownership code. CMS distinguishes Type 1 individuals from Type 2 organizations, and organizational enumeration can involve multiple NPIs. It also cautions that obtaining an NPI does not itself establish credentials or enrollment. [1] Keep the CMS Certification Number (CCN), enrollment identifier, and PECOS Associate Control identifier (PAC ID) in separate columns when available.
Use an identity checklist before assigning any parent:
- Research unit: State whether the record represents a facility, enrolled provider, legal entity, or platform.
- Legal name: Retain the source spelling and a separate normalized search version.
- Trade name: Preserve doing-business-as (DBA) and assumed-name variants.
- Location: Distinguish the facility address from an owner's mailing address.
- Provider type: Confirm which ownership file applies before joining records.
- Identifier crosswalk: Store each identifier with its source and entity scope.
- Match decision: Record the reason a provider record is linked to a corporate entity.
State sources help with the name crosswalk. New York's database includes assumed-name filings for corporations, limited liability companies (LLCs), and limited partnerships. [9] California recommends distinctive keywords and explains that exact searches depend on spaces and punctuation. A search miss after a single name variant is weak negative evidence. [14]
Resolve conflicts before building the graph
Provider identity and facility licensing should be compared before ownership labels are transferred. California's Cal Health Find exposes facility ownership alongside licensing and certification status, providing a useful facility-level corroboration source. Its information remains bounded by the program's maintained records. [15]
An identifier's continuity does not settle equity ownership. CMS describes NPI as a lasting identifier, while recognizing organizational circumstances that can require a new NPI. The appropriate inference is limited: an unchanged identifier does not prove unchanged ownership, and a changed identifier alone does not prove an acquisition. [16]
Make unresolved identity a stopping condition for the ownership screen. A plausible parent attached to the wrong operating entity is less useful than an explicit unknown attached to the correct provider.
A plausible parent attached to the wrong operating entity is less useful than an explicit unknown attached to the correct provider.
Choose Provider-Type Files and Interpret Their Rows
Use a source matrix with an explicit verification boundary
Table 1 summarizes the official documentation actually retrieved for this report. Documentation availability is not live dataset freshness. The latest ownership release, update cadence, and functioning download or application programming interface (API) endpoint remain unverified for these rows. The matrix should guide file selection, not imply a newly downloaded October 2026 ownership census.
| Provider type | Verified documentation and joins | Ownership and event treatment | Release and cadence status |
|---|---|---|---|
| Hospitals, including critical access and rural emergency hospitals | Hospital guidance links enrollment and owner records through ENROLLMENT ID; use owner associate ID and role. | Ultimate-parent absence is stated in this hospital guidance; resolve any proposed parent externally. [2] | Live release and cadence not verified. |
| Skilled nursing facilities | Retrieved owner dictionary includes owner type, role, association date, percentage, and parent-company flag. | Preserve the reported parent flag; CHOW material is a separate event source. [3] | Live release and cadence not verified. |
| Home health agencies (HHAs) | Guidance links enrollment and owners and supplies additional NPI/address files. | Shared identifiers can generate chain candidates; a possible ultimate owner requires corroboration. [17] | Live release and cadence not verified. |
| Hospices | Guidance describes enrollment, owners, additional NPIs, and additional addresses. | Owner-file coverage includes active ownership interest or managing control. [18] | Live release and cadence not verified. |
| Federally qualified health centers (FQHCs) | Dictionary distinguishes enrollment, provider PAC ID, and owner PAC ID. | Ownership percentage pertains to the FQHC enrollment; avoid assigning it to a branded network. [19] | Live release and cadence not verified. |
The matrix makes an important limitation actionable: the same-looking owner name can occur in different file structures. The presence of an owner dictionary establishes documented fields, not complete current coverage of every facility. For a production screen, first retrieve the actual release, validate its dictionary, and record its source date. If that cannot be done, leave the freshness requirement unresolved.
Classify the row before reading the percentage
Retain the source's owner type and relationship role. An organization name in an owner file is not enough to establish an equity edge. Officer, director, partner, manager, direct owner, and indirect owner should remain distinct categories wherever the source provides them.
Apply the following row-level rules:
- Reported role: Preserve the original role text and code before normalizing it.
- Entity type: Separate organizational owners from individuals.
- Ownership percentage: Retain a missing value as missing; do not replace it with zero.
- Association date: Treat it as the source-defined association date, not a closing date.
- Source version: Keep the exact dictionary and data release together.
- Parent flag: Store the reported value separately from the verified-parent conclusion.
- Event relationship: Attach a CHOW owner to the buyer or seller side specified by the file.
For example, SNF CHOW owner information links to buyer or seller enrollment IDs and distinguishes owning or managing roles. Event-side attribution is essential: combining those rows into one undated ownership list can confuse predecessor and successor structures. [13]
For FQHCs, the association-date definition concerns when the owner became associated with the enrollment. It is a useful field, but it should not silently become the date a sponsor acquired a platform. [19]
- Hospital guidance links enrollment and owner records through ENROLLMENT ID; preserve the owner associate ID and role.
- The September 2024 hospital guidance leaves the ultimate parent unidentified; resolve a proposed parent externally.
- The retrieved dictionary includes a reported parent-company flag. Preserve it with its source version and corroborate the proposed ultimate parent.
- Private equity and REIT flags use Y if reported and blank if not reported. A blank does not establish a verified negative.
Retrieved documentation establishes documented fields. Live ownership release and cadence were not verified in the report.
Implementation Considerations and Process Changes
Build a typed graph rather than a flat parent column
An ownership graph should contain entities as nodes and explicitly typed relationships as edges. A provider, legal operating entity, holding company, individual, and property owner can all be nodes, but they are not interchangeable.
A recommended graph legend is:
- CMS-reported edge: A relationship stated in the selected public file, with role and release preserved.
- Externally reported edge: A relationship stated in a filing, registry, or dated company document.
- Inferred chain link: A grouping supported by shared ownership evidence but not yet a proven legal intermediate.
- Candidate parent: An entity proposed as the endpoint of the hierarchy.
- Adjudicated parent: The endpoint supported under the team's stated definition and evidence rule.
- Unresolved link: A missing intermediate, incompatible date, or relationship that has not been reconciled.
These categories are proposed analytical controls, not CMS field definitions. Their purpose is to make derivation visible. The World Wide Web Consortium (W3C) PROV data model treats provenance in terms of entities, activities, and responsibility, and defines derivation as a transformation from existing information. A reconciled parent record can follow that pattern by pointing back to its input rows and review activity. [20]
“Ultimate owner” also needs a stated definition. The Federal Trade Commission (FTC) defines an ultimate parent entity for its Hart-Scott-Rodino (HSR) framework in terms of absence of control by another entity. Its voting-security control test is one rule-specific test, not a universal translation of CMS percentages. [21]
Calculate indirect interests only along supported paths
The current Medicare enrollment definition determines indirect ownership by multiplying percentages through the ownership chain. Its examples distinguish a reportable 8% indirect interest from a 4% interest below the general threshold. This is a path calculation, not a sum of every percentage in a provider's owner file. [4]
A hypothetical example based on the regulation's illustration shows the worksheet logic: if an intermediate owns 80% of the provider and a parent owns 10% of that intermediate, the path implies 8%. The calculation is useful only if both links, dates, and percentage bases are supported. If the upper relationship is merely a shared name or an inferred platform link, retain the path as a hypothesis. [4]
Direct and indirect totals require different tests. ASPE's historical SNF methodology explains that indirect interests can exceed 100% in aggregate because different tiers refer to the same underlying ownership. A total across tiers is not a partition of equity. [22]
Avoid deriving a voting-control conclusion from economic interest alone. The FTC's control definition and GLEIF's closest fully consolidating parent relationship demonstrate why control, ownership, and accounting parent are separate questions. [21] [12]
PECOS Ownership Worksheet and Decision Outcomes
- 01Resolve provider identity
Define the facility, enrolled provider, legal entity, or platform being researched. Resolve identity before assigning ownership labels.
- 02Select and date the release
Retrieve the applicable provider-type release, validate its dictionary, and record the source date. Leave freshness unresolved when the release cannot be verified.
- 03Preserve the reported role
Retain source owner types and relationship roles. Keep officers, directors, partners, managers, direct owners, and indirect owners distinct wherever the source permits.
- 04Build supported graph paths
Use entities as nodes and typed relationships as edges. Calculate indirect interests only when both links, dates, and percentage bases are supported.
- 05Reconcile entity and period
Compare evidence for the same legal entity and period. Check relationship definitions and source scope before resolving disagreement.
- 06Adjudicate the endpoint
Record group or chain membership separately, then assign resolved ultimate parent, reported owners with an unresolved endpoint, conflicting evidence, or unknown.
Resolve the ultimate parent only when evidence supports the path and endpoint under the stated parent definition and compatible dates.
Retain unresolved or conflicting outcomes when identity, intermediate links, endpoint evidence, or compatible periods are missing.
Make the worksheet the review interface
Table 2 sets out a PECOS ownership reconciliation worksheet. It is a proposed schema, designed to retain both observations and conclusions. It extends the provenance distinction between source material and derived records rather than asserting a new regulatory requirement. [20]
| Field group | Required worksheet fields | Review purpose |
|---|---|---|
| Provider identity | Research unit; provider type; NPI; CCN when applicable; enrollment ID; PAC ID; legal name; address | Confirm the exact provider before transferring ownership evidence. |
| Reported owner | Owner name; individual/organization; owner identifier; original role; percentage; association date | Preserve the source row without converting management into equity ownership. |
| Source provenance | File name; release/version; dictionary; source URL; collection date; source period | Make the evidence reproducible and its time scope visible. |
| Graph interpretation | Inferred tier; edge type; supported group/chain membership; candidate parent; parent definition; supported intermediate links; endpoint evidence | Distinguish group membership from a resolved ultimate parent and reported relationships from analytical reconstruction. |
| Corroboration | Filing or registry URL; supporting passage; document date; entity-match reason | Link each external statement to the correct entity and period. |
| Conflict and adjudication | Conflict type; competing observations; selected status; reason; reviewer; decision date; next review trigger | Preserve unresolved evidence and the basis of the screening decision. |
The value of this worksheet is not its column count. It is the ability to inspect the complete path from provider identity to parent conclusion. PCAOB guidance distinguishes evidence sufficiency from appropriateness; more evidence of the same poor quality does not resolve the underlying weakness. Used as an analytical analogy, that principle supports checking each edge rather than counting references. [11]
Apply fixed decision outcomes
CMS's hospital guidance leaves the ultimate parent unidentified in its data. [2] ASPE's hospital analysis distinguishes correctly delineating a chain from identifying its ultimate owner. [23]
For this worksheet, record supported group or chain membership in a separate field when the correct provider is linked to the named group with compatible relationship evidence. Then adjudicate the ultimate-parent question using these proposed outcomes:
- Resolved ultimate parent: Evidence supports the path from the correct provider to the named endpoint under the stated parent definition and compatible dates, including why that entity is the endpoint rather than an intermediate. Group membership alone does not meet this worksheet rule.
- Reported owners, ultimate owner unresolved: Owner rows or supported group or chain membership are available, but the hierarchy or evidence establishing its endpoint remains incomplete.
- Conflicting sources: Comparable evidence for the same entity and period remains incompatible after scope checks.
- Unknown: The required identity or ownership evidence is unavailable or insufficient.
Supported group or chain membership can coexist with an unresolved ultimate owner; retain both findings rather than treating membership as endpoint resolution.
An independent classification requires affirmative evidence under the mandate's definition. “No parent found” should remain a search result with a checked-source list, not become a corporate fact. GLEIF exceptions and state-record limitations make that distinction particularly important. [24] [9]
Reconcile the Graph and Adjudicate the Target
Compare evidence sources on the relationship they establish
Table 3 compares ownership evidence routes, including research services that organize and review findings. Each route retains its own scope; a research conclusion does not become an official enrollment record.
| Evidence route | Relationship or function | Useful proof and boundary |
|---|---|---|
| CMS ownership and CHOW files | Reported provider relationships and ownership events | Preserve enrollment scope, role, and date; corroborate any inferred hierarchy. [13] |
| SEC subsidiary exhibits | Public registrant's subsidiary list | Useful positive entity linkage; permitted omissions limit negative conclusions. [25] |
| IRS tax-exempt filings | Related organizations and direct controlling entities | Match legal name, address, and employer identification number (EIN); retain the tax-year scope. [8] |
| State corporate and facility records | Entity identity, filed roles, and facility-level information | Resolve names and licenses; a registered agent or search abstract is not a complete ownership ledger. [26] |
| GLEIF Level 2 records | Direct and ultimate accounting consolidating parents | Useful corroboration under the accounting definition; read exceptions when a parent is absent. [24] |
| Docket | Self-serve or managed target research through Triage, Scout, and Audit | Its documented research records retain a source, supporting excerpt, and collection date; the workflow does not replace the original provider records. |
Compare the same legal entity and period across routes. Agreement across different definitions needs interpretation; disagreement should trigger a scope check before selecting an answer.
Use filings positively and qualify negative searches
For a public parent, current Securities and Exchange Commission (SEC) exhibit rules require subsidiary information and permit specified omissions. An exhibit match can support group membership; failure to find a provider name is not definitive evidence that it is outside the group. [25]
For a nonprofit structure, IRS Tax Exempt Organization Search exposes returns and determination information. Schedule R can then identify the directly controlling entity for related organizations. Start with entity matching, rather than assuming similarly named foundations, hospitals, and operating organizations are interchangeable. [27] [8]
State corporate records have different boundaries. California says operating agreements and bylaws are maintained by the entity rather than filed with the Secretary of State. Its search guidance describes filed-information abstracts, while New York warns that completeness and accuracy depend on submitted information. Record what the public search actually discloses. [26] [14] [9]
For LEI records, distinguish a missing relationship from a verified absence of a parent. GLEIF reporting exceptions encompass different reasons, and non-public relationship information is not required for registration or renewal. An exception code deserves its own worksheet cell. [24]
Seek independent corroboration where the decision is material. The PCAOB evidence standard generally favors knowledgeable independent sources and original documents. Here, that is a practical reason to trace a repeated transaction statement to its originating filing or announcement, rather than treating copies as independent confirmations. [11]
Keep announcement status and current structure distinct
Record an announcement's parties, entity, status, and date as an event. This proposed rule reserves the current-parent field for evidence of the present relationship.
Periods matter. An IRS related-organization return can cover relationships held at any time during a tax year; a GLEIF parent is an accounting relationship; a state filing may expose a role rather than an equity stake. Resolve those differences explicitly instead of treating all three as interchangeable confirmations. [8] [12] [26]
Supported group or chain membership can coexist with an unresolved ultimate owner; retain both findings rather than treating membership as endpoint resolution.
Data Analysis and Evidence
Historical study populations and methods
The related historical SNF analysis used 14,864 facility observations, identified by CCN. Its affiliation methodology included network relationships extending beyond a simple equity hierarchy. The relevant lesson is methodological: a facility count, an associated-entity count, and an inferred chain count measure different units. [22]
GAO's September 2023 study used a random sample of 200 from 14,815 Medicare-enrolled nursing homes as of December 30, 2022. Its 5% private equity ownership estimate and 2% to 8% interval concern that population and method; they are not a current ownership label for a facility. [6]
Ownership interests and consolidation counts answer different questions
A neutral public-company example reinforces the distinction. Tenet's results for June 30, 2026 stated that United Surgical Partners International (USPI) had interests in 538 ambulatory surgery centers, with 405 consolidated. Those two counts describe different measures, not contradictory ownership totals. [28]
Its subsidiary exhibit is dated to December 31, 2025 and lists entities with formation jurisdictions. Neither the exhibit's period nor the later operating count supplies every facility's complete October 2026 legal hierarchy. Record both the relationship definition and the source period. [29] [25]
Calculate integrity measures with explicit denominators
The following are recommended worksheet diagnostics, not measurements from a newly downloaded CMS file:
- Missing owner names: Count applicable owner rows without a usable individual or organizational name.
- Missing percentages: Divide applicable equity-owner rows with missing shares by all applicable equity-owner rows.
- Missing direct owners: Count provider records with indirect equity rows but no identified direct equity row.
- Excess direct totals: Flag direct-owner totals above 100% for the same enrollment and compatible period.
- Indirect totals: Retain tier and path; never interpret their unrestricted sum as mutually exclusive equity.
- Ultimate-parent resolution rate: Divide providers classified as resolved ultimate parent by all correctly identified providers requiring an ultimate-parent decision. Report supported group or chain membership separately.
- Release lag: Calculate elapsed time from the documented source period, and distinguish it from collection time.
A total test is an exception queue, not an automatic correction. Roles, duplicate rows, dates, and percentage bases need review before a value is discarded. The indirect-interest calculation in the Medicare definition and ASPE's cross-tier example support treating direct and indirect tests differently. [4] [22]
Report unresolved identity and unresolved ownership separately. Do not calculate “independent share” by assigning every unresolved record to the independent category. The denominator and classification rule should travel with every screening metric.
Implications and Future Directions
Ownership research benefits from a maintained evidence model rather than a one-time parent-name export. Store the public source row as an observation, the graph link as a typed relationship, and the ultimate-owner decision as a derived record. W3C's provenance vocabulary provides a useful foundation for that separation, without prescribing a particular database or origination product. [20]
Updates should be source-specific. The fetched NPPES download page states that Version 1 files ceased support effective March 3, 2026, and that Version 2 expands name fields. Its listed monthly release was September 14, 2026. This verifies the download page's listing, not inspection of the ZIP contents or ownership-file freshness. [30]
A change to an identity source can therefore require a parser or matching review without changing the ownership conclusion. Conversely, a parent relationship can change while familiar names and provider identifiers persist. Match rules should retain their own version, and derived decisions should identify which rules were used.
Provider-directory participation belongs in a separate field. The Office of the National Coordinator for Health Information Technology describes its Trusted Exchange Framework and Common Agreement (TEFCA) participation map as drawing on an organizational exchange directory. Exchange participation can support directory context; it is not itself a statement of equity ownership. [31]
For healthcare-services origination teams, the recommended operational changes are:
- Refresh triggers: Revisit records when a relevant ownership event or incompatible filing appears.
- Source boundaries: Keep provider, corporate, property, and accounting evidence distinguishable.
- Review priority: Escalate missing intermediates and conflicting periods before accepting a parent.
- Decision history: Preserve prior observations when the current classification changes.
- Mandate fit: Evaluate independence, institutional backing, and sale intent as separate questions.
A new disclosure field can improve candidate generation without resolving every hierarchy. A stronger review record can make uncertainty actionable without removing it. Evidence quality depends on relevance and reliability, not solely on how many pages agree. [11]
Frequently Asked Questions (FAQs)
Is PECOS an ultimate-owner database?
Use its ownership rows as reported evidence, then establish the endpoint under a stated parent definition. Accounting-consolidation parent relationships provide a distinct corroboration route, rather than an interchangeable equity hierarchy. [7]
Can an NPI lookup prove who owns a healthcare company?
An identifier resolves a provider record. It does not convert the record into equity evidence. Confirm corporate ownership through the applicable owner roles and independent relationship sources. State assumed-name records can assist the legal-name crosswalk. [9]
Why can indirect ownership percentages exceed 100%?
Different tiers can refer to the same underlying provider interest. Preserve the tier and calculate supported paths; do not add those rows as though each were a separate equity slice. [22]
Does an absent parent establish independence?
A missing LEI relationship can reflect a reporting exception, and a public entity database has its own completeness limits. Preserve “unknown” or “ultimate owner unresolved” until the independence standard is met. [24] [9]
Conclusion
Identifying an ultimate owner from CMS PECOS data requires reconciliation. Begin with the enrolled provider or operating entity, retain its identifiers, select the applicable ownership documentation and release, and distinguish equity relationships from managerial or other disclosed roles. A name shared across records can generate a candidate chain; a parent conclusion requires a supported relationship and a compatible period.
The most useful output for an origination team is a worksheet that exposes the entire reasoning path. It should show the source owner row, any externally supported intermediate, the parent definition, and the adjudication. It should also show what could not be verified. This allows a research lead to compare like entities and direct additional work toward the links that actually determine the screen.
Apply percentage checks according to relationship type. Keep events separate from current structure, preserve missing values, and do not convert blank flags into negatives. When a source has incomplete coverage or an inaccessible release record, state that boundary rather than replacing it with an assumed current fact.
The final ownership screen should record supported group or chain membership separately from the ultimate-parent status: resolved ultimate parent, reported owners with an unresolved endpoint, conflicting evidence, or unknown. Independence requires its own affirmative support. None of these statuses, including a verified parent, establishes willingness to sell. The objective is a reproducible ownership decision that remains intelligible when another analyst opens the evidence, checks the dates, and reaches the same conclusion or identifies exactly why it must change.
External Sources (31)
Published by
Docket Research
Docket provides deal-origination research software and managed research for private equity teams, screening companies against a mandate and showing the evidence behind each finding.
Editorial note
Researched and written by Docket Research with AI-assisted research tools; sources are cited inline and checked before publication. Report an error: https://docket.capital/evidence-standards#corrections.