Page 13 of 13 · CFMG–Wellpath California · Prepared · revised · Record current through · record checked

Sources, Methodology, Evidence Grades, and Corrections

This investigation is built around a simple rule:

A public conclusion must be independently supportable from public evidence.

Every proposition is drawn from sources that readers can independently inspect, and every page states what its sources prove and what they do not.

That rule governs everything from source selection to corrections.

1. Public sources only

The investigation uses only material a reader can retrieve: statutes and regulatory guidance; filed contracts; County Board, procurement and budget records; court dockets and orders; labor-board records; bankruptcy filings; and public institutional statements.

No proposition rests on a document that cannot be reached from a cited public source.

Where a public record concerns an identifiable individual, it is used only for the institutional fact it establishes. Public material about individuals that adds nothing to the institutional question is left out.

The standard for every fact is:

Is this institutional fact necessary, independently verifiable from a cited public source, and stated no more broadly than the source supports?

2. What sources qualify

The investigation prioritizes sources in this order.

Grade A — Primary government or judicial record

Examples:

  • statutes;
  • court orders;
  • filed contracts;
  • County Board records;
  • procurement records;
  • NLRB case records;
  • bankruptcy filings and orders;
  • official regulatory guidance;
  • government audit or monitoring reports.

These sources receive the greatest evidentiary weight.

Grade B — Direct institutional source

Examples:

  • Wellpath press releases;
  • H.I.G. Capital transaction announcements;
  • official company reports;
  • public corporate statements.

These are strong evidence of what the institution itself represented.

They are not automatically independent verification of the representation.

Grade C — Public repository reproducing a primary record

Examples:

  • a public-interest archive hosting an identifiable court-filed contract;
  • a docket mirror reproducing a filed order.

These may be used when the underlying document is clear and identifiable.

Where possible, the investigation links directly to the government or court source.

Grade D — Secondary reporting

Examples:

  • newspapers;
  • trade publications;
  • commentary;
  • advocacy reports.

Secondary material is most useful for:

  • identifying leads;
  • explaining public controversy;
  • locating primary documents.

Core structural conclusions should not rest solely on secondary reporting when primary sources are available.

3. Evidence type is separate from source quality

A high-quality source can still contain only an allegation.

For example, a federal court docket is a Grade A source, but a complaint filed in that docket remains a party allegation unless proved or adopted by the court.

Every important proposition should therefore be classified by evidence type.

JUDICIAL FINDING

A court actually decided the proposition.

GOVERNMENT RECORD

A government body generated the record.

This does not mean every characterization inside the record is an adjudicated fact.

SWORN TESTIMONY

A witness made the statement under oath.

Sworn testimony is evidence, but it can be disputed or contradicted.

PARTY STIPULATION / FORMAL REPRESENTATION

The parties agreed to or formally represented a proposition.

A stipulation may be binding within a case without becoming a universal corporate-law conclusion.

PARTY ALLEGATION

A complaint, motion, arbitration demand, regulatory accusation, or other adversarial assertion.

It is reported as an allegation unless independently established.

ADMINISTRATIVE METADATA

A portal label, case caption, employer field, or agency database entry.

Metadata can be authentic evidence of how a system classified something without deciding the underlying legal question.

COMPANY STATEMENT

A public statement made by Wellpath, CFMG, an investor, or another institution.

It proves the company made the representation.

ANALYSIS

The project’s reasoned synthesis of underlying sources.

Analysis is labeled as such and should not be presented as if it were a source fact.

OPEN

The available public record does not responsibly establish the answer.

“Open” is a valid investigative conclusion.

4. The “what it does not prove” rule

Every consequential source is tested twice.

First:

What does this source prove?

Then:

What does it not prove?

Examples:

A County contract naming CFMG proves the County contracted with CFMG.

It does not automatically prove:

  • who administered HR;
  • who controlled every clinical decision;
  • who owned CFMG shares;
  • or that Wellpath had no role.

A County document using “CFMG dba Wellpath” proves public operating nomenclature.

It does not automatically prove a legal merger.

A bankruptcy motion describing a “friendly PC” model proves the debtor made that formal representation.

It does not prove every professional corporation had identical contracts.

A court-approved case stipulation that CFMG and Wellpath are treated as the same entity for that lawsuit may be highly significant within that case.

It does not automatically establish statewide legal identity.

This limiting discipline is one of the project’s principal safeguards against overstatement.

5. The functional-control method

The investigation does not ask only:

“Who controls CFMG?”

That question is too broad.

Instead, it separates functions such as:

  • County contracting;
  • physician employment;
  • HR administration;
  • physician workload;
  • referrals;
  • utilization management;
  • credentialing;
  • privileging;
  • peer review;
  • clinical policy;
  • medical records;
  • IT/EHR administration;
  • finance;
  • insurance;
  • claims;
  • litigation defense.

For each function, the preferred reconstruction is:

origin → recommendation → professional review → approval or veto → implementation

The person who implements the result is not automatically the person who made the substantive decision.

6. The practical-veto test

Formal contracts can allocate authority.

Actual disagreement is often more probative.

The project therefore places special value on events where:

  • management recommends one result;
  • physician authority rejects or modifies it;
  • the final result shows whose decision controlled.

A verified event in which CFMG physician leadership rejected a management recommendation and management complied would be strong evidence of practical professional independence in that domain.

A verified event in which management proceeded despite contrary CFMG physician direction would support the opposite inference.

The investigation will publish either result.

7. Corporate identity rules

The project does not use the word “Wellpath” as a substitute for every related legal entity.

Where identity matters, the site attempts to distinguish:

  • California Forensic Medical Group, Inc.;
  • Wellpath LLC;
  • Wellpath Management, Inc.;
  • professional corporations in other states;
  • predecessor management entities;
  • the Wellpath Liquidating Trust;
  • County clients;
  • individual clinicians.

Likewise, terms such as:

  • “affiliate”;
  • “dba”;
  • “same entity”;
  • “subsidiary”;
  • “formerly”;
  • “separate and distinct”

are quoted with their source context rather than treated as interchangeable legal conclusions.

8. County authority is kept separate

Correctional healthcare is a government service delivered in a custody environment.

Counties can:

  • select the contractor;
  • set staffing requirements;
  • establish service levels;
  • audit performance;
  • approve costs;
  • control security and transportation;
  • impose contractual remedies;
  • replace vendors.

Court orders or remedial plans can impose additional requirements.

Those are independent sources of authority.

The site therefore does not attribute every operational requirement at a County facility to CFMG or Wellpath.

9. Clinical governance terms are not collapsed

The site distinguishes:

Credentialing

Verification of qualifications and professional information.

Privileging

Authorization to perform particular clinical functions.

Peer review

Professional review of competence or conduct under an applicable process.

Quality improvement

Systematic measurement and improvement of care processes.

§805 reporting

A specific California reporting mechanism triggered by defined peer-review actions and conditions.

Technical access

Ability to activate, deactivate, or administer a computer system.

These categories can interact.

They are not synonyms.

10. Bankruptcy terminology is handled narrowly

Bankruptcy often uses specialized terms:

  • debtor;
  • nondebtor;
  • estate;
  • stay;
  • claim;
  • discharge;
  • injunction;
  • Liquidating Trust;
  • assumption;
  • confirmation.

A bankruptcy result concerning one Wellpath entity does not automatically determine:

  • a nondebtor CFMG obligation;
  • employer identity;
  • professional authority;
  • a California statutory claim;
  • or an insured claim.

Every bankruptcy proposition is therefore tied to the precise entity and legal issue.

11. Current-status verification

Litigation and contracts change.

Whenever a page states that a case is:

  • pending;
  • dismissed;
  • settled;
  • on appeal;
  • under enforcement;
  • or otherwise active,

the page should carry a last verified date.

For matters where public docket aggregators are incomplete, the site should say so.

PACER or the official court docket should be used for procedural precision when available.

12. Quotations

The site uses direct quotations sparingly.

A quotation should be included only when the exact wording materially matters.

Long passages should be summarized and linked to the source.

Where a phrase such as:

  • “separate and distinct”;
  • “same entity”;
  • “dba Wellpath”;
  • “stock transfer restriction agreements”

is legally or analytically important, the precise wording may be retained with source context.

13. Corrections policy

Corrections are part of the investigation, not an embarrassment to be hidden.

When a material factual proposition changes, the site should record:

  1. the prior statement;
  2. the new source;
  3. why the conclusion changed;
  4. the correction date.

Corrections should not silently erase the existence of the earlier position if the change is material to understanding the investigation.

Minor typographical and formatting corrections need not receive a public correction entry.

14. Contradictions policy

Apparently contradictory sources are not automatically reconciled by choosing one.

They should first be classified.

For example:

  • a County procurement record may use a brand;
  • a federal court may address juridical identity;
  • an NLRB record may use labor-employer nomenclature;
  • a bankruptcy filing may describe an enterprise relationship.

Those records can all be accurate within different domains.

A contradiction is genuine only when the sources address the same proposition and cannot reasonably coexist.

15. Individuals in the public record

Public records name people: litigants, clinicians, county staff, corporate officers, witnesses.

The investigation names an individual only where the public record itself makes the name necessary to the institutional point — for example, a party whose sworn testimony a court relied on.

No private individual is profiled, and combinations of details that would single out a person without advancing the institutional analysis are not published.

16. Publication workflow

Every major new page should proceed through the same sequence:

Step 1 — Question

Define the institutional question.

Step 2 — Public-source collection

Gather primary sources: government, court, regulatory, labor-board and bankruptcy records, and institutional statements.

Step 3 — Public Fact Ledger

Record each proposed factual statement with:

  • source;
  • evidence type;
  • what it proves;
  • what it does not prove;
  • date verified;
  • publication status.

Step 4 — Draft

Write from the public ledger and primary sources.

Step 5 — Adversarial review

Ask:

  • strongest competing explanation?
  • strongest contrary evidence?
  • what evidence would falsify the conclusion?

Step 6 — Individuals check

Apply section 15: name a person only where the public record makes it necessary.

Step 7 — Status check

Refresh current litigation/contract status where relevant.

Step 8 — Publish

Include source links and last-updated date.

Step 9 — Correct

Maintain a visible material-corrections record.

17. Public Fact Ledger template

Each material factual proposition should be trackable using fields such as:

Field Purpose
Public Fact ID Stable identifier
Page Publication location
Proposition Exact factual statement
Entity Exact legal/entity reference
Date range Relevant period
County/state Geography
Function HR, clinical, contract, claims, etc.
Source Exact source
Source URL/docket Reader access
Source grade A/B/C/D
Evidence type finding/testimony/stipulation/etc.
What it proves Narrow supported proposition
What it does not prove Mandatory limitation
Last verified Freshness
Correction history Audit trail

18. Editorial standard for conclusions

The investigation should prefer conclusions that remain accurate even if the reader disagrees with the broader thesis.

For example:

Strong:

Public contracts establish CFMG as the formal contractor in multiple California counties while other public records use Wellpath as an operating identity.

Weak:

Counties secretly contract with Wellpath through CFMG.

Strong:

The management agreement assigns extensive administrative functions to the manager while expressly reserving identified professional functions to CFMG.

Weak:

The contract proves Wellpath controls the medical practice.

Strong:

The current public record does not establish the terms of CFMG’s stock-transfer restriction agreements.

Weak:

Wellpath controls CFMG ownership.

The goal is not cautious writing for its own sake.

It is writing that can survive adversarial scrutiny.

19. What this investigation is — and is not

This project is:

  • a public-record investigation;
  • an entity and authority map;
  • a contracting and litigation atlas;
  • a governance analysis;
  • a source archive;
  • a living corrections project.

It is not:

  • a court;
  • a regulator;
  • an accusation that every integrated management function is unlawful;
  • a claim that corporate separateness means operational independence;
  • a substitute for reading the underlying records.

20. Current synthesis

The first 12 substantive pages support a stable public formulation:

CFMG remains a legally significant California professional corporation and government contractor within a highly integrated Wellpath management architecture. Public contracts and litigation establish both legal separateness and substantial operational integration. California law makes the practical boundary important where management functions intersect physician-reserved authority. The public record documents broad Wellpath administration and formal CFMG professional reservations but does not yet provide a complete statewide map of practical physician veto in every protected domain. The most valuable unresolved evidence concerns ownership succession, manager replacement, clinical-governance authority, and real disagreement events showing whose decision controlled.

That is the standard against which future evidence will be tested.


Core public-source families

  • California statutes and Medical Board guidance
  • CFMG Management Services Agreement and 2019 assignment
  • County contracts, budgets, procurement records, and monitoring reports
  • California federal and state litigation
  • NLRB employer records
  • Wellpath Chapter 11 filings and orders
  • Wellpath/H.I.G. institutional statements
  • public quality, accreditation, and oversight records

End of methodology

Kanwar Partap Singh Gill, MD
Family Medicine Physician · Fresno, California, USA

Original KPSGILL public-record investigation · public sources only · labelled as such · never official-government data · record current through 15 September 2026 · sources checked 15 September 2026 · prepared 15 September 2026 · revised 16 September 2026 · this revision pending review by Kanwar Partap Singh Gill, MD.