Editorial policy

GMedAC Journal is the editorial publication of GMedAC — Global Medical Access & Care, a healthcare discovery and transactional marketplace that owns everything around the consultation: verification, jurisdiction-aware eligibility, booking, payment, reminders and payouts. GMedAC is not a clinic, not an EHR, not a medical messenger and not a video platform, and the consultation itself happens outside it. The journal inherits that boundary: it explains the machinery around a consultation, holds no clinical record and issues no clinical document.

What this publication publishes

Articles explaining how healthcare access works — what verification checks, how eligibility is decided, what a booking is and is not. Every article declares an articleType and a contentRisk of LOW, MODERATE, HIGH or CRITICAL, and that level governs how strong its sources must be and who must clear it.

There are no comments, no user profiles and no patient stories. A comment field puts unreviewed clinical claims under this publication's name and in front of a reader before anybody has read them; a patient story invites generalisation from one body to another. Neither is refused for tidiness.

No article describes the marketplace as live. It is at Phase 0 — no checkout, no live payment provider, no deployment — and writing an unbuilt flow in the present tense is a false statement about the company, not a rounding error.

No article asserts a legal conclusion about where GMedAC may operate. The Czech launch matrix stands at three gates RESEARCHED_PENDING_COUNSEL, seven COUNSEL_QUESTION_OPEN, two EVIDENCE_MISSING and zero approved. Excluded means unresearched; it never means permitted, and EVIDENCE_MISSING is not a clearance. The United States is never one corridor — it resolves state by state, and the corridor research that exists there is discovery rather than clearance: no state is launch-ready and none is declared so. On verification, this publication keeps the marketplace's reading: VERIFIED is not permission to practise or to book, it is a statement that somebody checked a document, published as a summary of what was checked and never as an endorsement. Whether that makes GMedAC an endorser rather than an intermediary is COUNSEL_REQUIRED and unresolved here.

The lifecycle

Every article carries a publicationStatus; the vocabulary is closed.

Status	Meaning
DRAFT	Checked by nobody but its author.
EDITORIAL_REVIEW	With an editor for structure, sourcing and honesty.
CLINICAL_REVIEW	With a reviewer whose qualifications cover the subject.
LEGAL_REVIEW	With a reviewer for jurisdictional and regulatory claims.
READY	Cleared by every review its risk level requires.
SCHEDULED	Cleared, with a future publication date set.
PUBLISHED	Live, first publication.
UPDATED	Live, materially changed since publication.
ARCHIVED	Taken down as superseded or no longer maintained.
WITHDRAWN	Taken down as wrong, unsafe or unsupportable.

An author moves an article to EDITORIAL_REVIEW and no further. Only an editor advances it into or out of a review stage. CLINICAL_REVIEW clears only on a reviewer named in reviewerIds who is not also in authorIds — an author reviewing their own article produces a second signature from the first pair of eyes. LEGAL_REVIEW clears only on a reviewer qualified in the jurisdiction the article speaks about. Publication is an editor's decision. Withdrawal is open to any editor or reviewer without waiting for agreement, because a slow withdrawal costs the reader.

None of that is enforced by a login: there is no database and no authentication. publicationStatus is a line in a file, and the control on who changes it is review of the change plus the checks over the result. Those checks reach part of the way: a published high-risk clinical article must name a reviewer whose recorded scope covers its section, and a jurisdiction-specific claim must be covered by a reviewer qualified in that jurisdiction. Nothing refuses an article whose reviewer is also its author, and anyone with write access can move an article to PUBLISHED alone. This is honest, not clean.

Only a published article is readable

PUBLISHED and UPDATED are the only statuses that produce a page, a sitemap entry, a feed item or a row in the search index. Every other status has no route, so there is no address for a crawler to reach or for a reader to be sent to. What that removes is the public address, not the text: the file sits in the repository, and anyone with access to it can still pass a draft around. robots and canonicalPath are declared per article, but they are the second line of defence: a directive asks a crawler to stay away; an absent page gives it nothing to ask about. A draft that is reachable is published.

Withdrawal ends circulation the same way, and today it ends the address with it. The intention is that the canonicalPath keeps a notice recording that an article was published there, that it was withdrawn, when and why. That notice is not built, and no check refuses a withdrawal that leaves nothing behind, so a reader arriving from an old link finds nothing and cannot tell whether the article never existed, moved, or was found to be wrong. The gap is known and unclosed.

Editorial and commercial

This journal is published by GMedAC and written and reviewed by people GMedAC pays, so no claim that its editorial decisions stand outside a commercial interest would be true; the conflicts of interest policy states that condition and what limits it. What is enforced sits elsewhere. In the marketplace, commission does not influence organic ranking and a test pins that. Here, GMedAC's account of itself enters the source vocabulary as COMPANY_STATEMENT, the weakest type in the list, which a HIGH or CRITICAL article may not rest on: the company cannot support its highest-stakes claims by describing itself. That no editorial decision is taken on the instruction of somebody with a stake in its outcome is a commitment by the people holding push rights, not a mechanism.

This publication carries no advertising, no sponsorship, no affiliate links and no advertising pixel today, and it never names an individual professional as a recommendation. Whether a clearly labelled paid surface may ever exist anywhere in GMedAC is undecided. Introducing one here would be a new decision, written into this policy before it appeared.

A disclaimer does not repair unsafe content

If a passage would lead a reader to act in a way that could harm them, a notice appended does not fix it. The passage is rewritten or cut. A disclaimer reaches the reader who already understood the limits of what they were reading; the unsafe sentence is aimed at the reader who did not. An article carrying emergencyInformation: true is held to this most strictly, and a stale one fails the build at any risk level.

A green build is not proof

The checks read structure. They confirm that a jurisdiction-specific claim does not rest on REASONED_INFERENCE, COUNSEL_CONFIRMATION_REQUIRED or UNKNOWN, that a HIGH or CRITICAL article rests only on sources at or above PEER_REVIEWED_STUDY, and that dateRetrieved and retrievalRoute are both present. None of that reads a sentence. No check knows whether a source says what the claim says it says, whether the retrievalRoute prose is true, or whether an article is honest about what it does not know. The checks prove shape; a reviewer proves substance.

The other policies

• Medical content policy — what a passage may not tell a reader to do.
• Legal and regulatory content policy — jurisdiction-specific claims and what the build refuses.
• Source policy — the source hierarchy, dateRetrieved and retrievalRoute.
• Corrections policy — previousClaim verbatim, and when a correction is material.
• Conflicts of interest policy — the conflict at the root of a journal published by its subject.
• Advertising and sponsorship policy — what is carried, what refuses it in code, what is open.
• AI assistance policy — where machine assistance is used and where it is disclosed.
• Author and reviewer policy — what a named reviewer attests to.
• Image rights policy — what every image must declare.
• Content retention policy — validityAsOf, nextReviewAt, and why low-risk staleness warns.
• Emergency content policy — why a stale emergency page fails the build at any risk level.
• Privacy for readers — no cookies, no analytics, search in the reader's browser.