Methodology

Overview

This page documents how WellChina collects, verifies, and updates the information published on this platform. It is the technical companion to two other documents:

• Editorial Policy — the rules governing what we publish, who reviews it, and how we handle corrections and conflicts of interest
• About — who we are, why we built this, and our business model

If you want to know what is published, read About. If you want to know how content is reviewed and signed, read Editorial Policy. If you want to know how we know what we say is true — including hospital verification, pricing methodology, source tiering, and the boundary between AI drafting and human verification — read this page.

This methodology is a commitment document, not a statement of current practice. Where the commitment exceeds today's operational maturity, it creates an obligation we work toward and revise via the Version History when standards change.

How We Research Hospitals

Hospitals enter the WellChina directory only after passing a triple filter:

1. JCI accreditation or equivalent international standard. We rely on the public JCI accredited organizations registry as the primary source. For hospitals accredited by equivalent national standards (e.g., Australian Council on Healthcare Standards International), we record the specific accreditation, the issuing body, and the validity dates.

2. Demonstrated English-language clinical capability. We verify by direct contact with the hospital's international patient department: we confirm the existence of English-speaking attending physicians for the relevant specialties, English-language consent forms, and English discharge summaries. Self-reported "English service" without specific staff or document evidence does not pass this filter.

3. Named international patient coordinator. A specific person or named role at the hospital responsible for international patients. Generic email addresses (info@hospital.com) without a coordinator name do not satisfy this requirement.

Hospitals that pass all three filters appear in the directory with their accreditation status, verified English-capability claims, and coordinator contact pathways. Most hospitals in China do not pass this filter; the curated subset is the value of the directory.

Hospital Inclusion Criteria

Hard requirements (must be true at publication time):

• Valid JCI accreditation or equivalent, with documented effective dates
• At least one named English-speaking attending physician in each specialty we list for that hospital
• A named international patient coordinator (person or role)
• Hospital-side written consent to be listed (we will remove any hospital that requests removal)

Soft signals (improve placement and depth, not gate inclusion):

• Specialty depth indicators: Fudan ranking position, hospital-published annual case volume, Chinese MOH grade (3A and equivalent)
• Existing relationships with international insurers (direct billing networks)
• Multilingual website / patient education materials beyond Chinese and English

Reasons we exclude:

• JCI accreditation lapsed or never granted, with no equivalent accreditation
• English-language clinical capacity could not be verified through direct contact
• Hospital declined to confirm coordinator information
• Active regulatory action against the hospital from a national health authority
• Documented pattern of medical-tourism complaints unresolved through the hospital's complaint channel

Pricing Methodology

WellChina maintains a 10-country pricing reference table covering the United States, United Kingdom, Germany, Japan, South Korea, Russia, Indonesia, Vietnam, Thailand, and India for each procedure we list.

Source rules:

• China-side prices: hospital-published price ranges for international patients, verified by direct contact with the hospital's international patient billing department, with effective-from dates recorded
• Origin-country prices: published rates from national health authority sources, major insurance carrier reimbursement schedules, and at least two independent second-source publications per country per procedure
• We do not source prices from individual patient testimonials, marketing aggregators, or referral platforms

Currency conversion:

• Each price is stored in its native currency with the effective-from date
• Display conversion uses our cached exchange rate, refreshed daily, with the conversion date shown alongside the converted figure
• Where the conversion date differs from the source effective date by more than 30 days, the converted figure is shown with a "rate-of-record" flag

What "price effective on YYYY-MM-DD" means:

• The price was published or directly confirmed by the source on that date
• The price has not been re-verified since that date — it is a snapshot, not a guarantee
• We always prompt users to confirm current pricing with the hospital before making travel decisions

Verification Process

We commit to the following minimum verification cadence. We may verify more often when material changes are reported.

Annual baseline: every published hospital, procedure, and price record is reviewed at least once per calendar year. The review confirms institutional facts (JCI status, coordinator name, English capability) and re-checks the source for material price changes.

Triggered re-verification: any of the following triggers an immediate re-verification within 7 calendar days:

• A user submits a corrections request through Editorial Policy §corrections flagging a specific record
• A hospital publicly announces a status change (JCI lapse, specialty closure, coordinator change)
• A national health authority issues a regulatory action against the hospital
• An exchange rate movement exceeds 5% relative to the rate of record

What we ask during verification: JCI status and validity dates, named English-speaking attending physicians by specialty, current international coordinator name and contact, current price ranges for the specific procedures we list, and any hospital-side changes the institution wants reflected.

What we record: verification date, person contacted, contact channel (phone, email, in-person), summary of confirmed and changed items, and where applicable a non-personally-identifying audit log entry retained for at least three years.

Source Tiers

We classify every source claim into one of four tiers. Tier determines what kind of statement the source can support.

Tier 1 — Authoritative: government health authorities (NHC of China, US HHS, UK NHS, etc.), JCI accredited-organization registry, peer-reviewed medical journals indexed in PubMed or equivalent. Tier 1 sources can support any factual claim, including medical descriptions on YMYL pages.

Tier 2 — Hospital official: hospital-published websites, hospital-issued press releases, direct verifications from named hospital staff. Tier 2 sources can support institutional facts (services offered, capacity, coordinator name) but cannot alone support medical claims about treatment outcomes.

Tier 3 — Industry secondary: industry trade publications, established medical-tourism trade associations (Patients Beyond Borders, IMTJ, MTA). Tier 3 sources can corroborate Tier 1 / Tier 2 claims but cannot alone support a primary publication claim.

Tier 4 — Self-reported / unverified: patient testimonials, social media posts, hospital marketing pages without named author or date. Tier 4 sources are not used for any factual claims on the platform; they may be referenced as illustrative experience only with explicit labeling.

A claim that we cannot trace to at least one Tier 1 or Tier 2 source is not published.

Claim Verification Cadence

Different categories of facts require different verification frequencies. Pages display a "last reviewed" date corresponding to the strictest applicable class.

Class A — Volatile (24-hour fact-check window): prices, exchange rates, JCI accreditation status changes, regulatory actions. When a change is reported, we re-check the underlying source within 24 hours and either update or flag the page.

Class B — Institutional (7-day window): hospital coordinator names, specialty rosters, address and contact details. Material changes are reflected within 7 calendar days of confirmation.

Class C — Procedural (annual review window): medical procedure descriptions, recovery timelines, prerequisites, post-care guidance. These are reviewed annually as part of the verification baseline. Earlier review is triggered by guideline changes from the relevant medical society or significant peer-reviewed publication.

The schema dateModified field reflects the most recent re-verification across any applicable class.

AI vs Human Boundaries

We are explicit about which content categories AI is allowed to draft and which require human-only handling. This boundary is informed by the founder's machine-learning and clinical-AI background (About §team).

AI may draft: non-medical content (city guides, visa overviews, payment overviews, currency calculators), translations of canonical English source content into the seven other supported languages, schema metadata, and SEO meta tags.

Human must verify before publication: hospital institutional facts (JCI status, coordinator name, English capability), pricing data (sources and effective dates), procedure descriptions (compliance with the medical reviewer process per Editorial Policy), legal/financial content (visa rules, insurance coverage interpretation).

Human only — AI does not touch: medical reviewer sign-off decisions, conflict-of-interest disclosures, corrections handling, the determination of whether a fact is verifiable.

When AI is used in any publication path, the AI involvement is logged internally with the model used, the prompt template, and the human reviewer who approved the output.

What We Track and Update

Every published page exposes the following timestamps and metadata:

• datePublished in JSON-LD schema: when the page first went live
• dateModified in JSON-LD schema: most recent material content change
• "Last reviewed" header on each page: most recent verification per the Claim Verification Cadence
• Refresh-age badge on hospital and pricing records: visual indicator when the underlying source has not been re-verified within the class-appropriate window

Internal change tracking covers content edits (commit log), price-source updates (verification logs), and AI involvement (prompt + reviewer log).

Feedback & Corrections

If you believe any claim on the platform is inaccurate, outdated, or misleading, please file a correction through Editorial Policy §corrections-policy. The corrections SLA, response windows, and escalation path are documented there.

If you believe this Methodology document itself contains a process gap, contradiction, or commitment we are failing to honor, please email contact@wellchina.top with subject line "Methodology bug". Methodology bug reports are reviewed at least quarterly and feed into the next Version History increment.

Version History

• v1.0 — 2026-04-27 — Initial publication. Triple-filter hospital research, four-tier source classification, three-class verification cadence, AI/human boundary documentation, annual + triggered re-verification commitment.

Next scheduled review: 2027-04-27, or earlier if material practice changes occur.