Methodology. How every investigation is built.

Every Clinic Truth verdict — green or red — follows the same procedure. This page documents it in detail so readers can audit our work and replicate our findings.

Document version 1.0 · Last reviewed 22 April 2026 · Next scheduled review 22 July 2026

1. How we pick which clinics to investigate

Three inputs weight into the investigation queue:

• Patient demand. Questions we receive via WhatsApp and email, tallied weekly. If five patients in one week ask about Clinic X, Clinic X goes into the next-month queue.
• Search volume. Google keyword data for the clinic's own brand name + "reviews," "opinions," "gone wrong," in English, Italian, Spanish, and Greek markets. Clinics with >500 monthly brand searches enter the queue automatically.
• News / regulatory signal. If a clinic is mentioned in mainstream press in connection with a complaint, lawsuit, or regulatory action, it enters the queue regardless of volume.

We do not investigate clinics because they asked us to. We do not skip clinics because they are big, connected, or litigious.

2. Data we pull for every investigation

Source 1 — The clinic's own website

Full read of everything on the clinic's public website. We capture every specific claim: price ranges, clinician credentials, technique descriptions, "accreditation" badges, patient volumes, post-op guarantees. Screenshots are archived in our internal drive with a timestamp — this is how we cross-reference later if a clinic changes its claims.

Source 2 — Trustpilot (and Google) reviews at scale

We pull all available reviews from the clinic's Trustpilot profile using their public API. For clinics with more than 500 reviews, we take a stratified random sample of 200–500 reviews spread across all available years. Stratification ensures we don't over-represent recent reviews (which tend to be more positive for marketing-active clinics).

We categorise every sampled review under three axes:

The weighted score becomes one input into the final verdict. We never publish a verdict based on review score alone — score is necessary but not sufficient.

Source 3 — Regulatory register

For every named clinician mentioned by the clinic, we check the relevant national medical register:

• Turkey: Türk Tabipleri Birliği (TTB) — the Turkish Medical Association
• Albania: Ministry of Health registry of licensed clinicians
• Italy: FNOMCeO — the Federazione nazionale degli ordini dei medici chirurghi e degli odontoiatri
• Spain: Consejo General de Dentistas de España
• UK: GDC (dental), GMC (medical)
• Greece: Πανελλήνιος Ιατρικός Σύλλογος (Panhellenic Medical Association)

If a clinician named in marketing cannot be located in the relevant register, the investigation notes this explicitly. It does not automatically trigger a CAUTION stamp — registration failures can have administrative causes — but it does require a written response from the clinic before we proceed.

Source 4 — Independent references

At least two of: mainstream press coverage, documented patient forums (RealSelf, HairLossTalk, Reddit threads with verified patient credentials), academic papers on the clinic's claimed technique, court records where accessible. Forum content is only used when the patient's claimed treatment date and technique can be corroborated with independent data.

3. How we build the verdict

The verdict is one of two states. There is no "mostly green" or "yellow caution." Decisive output is part of the brand.

Green VERIFIED — four criteria, all four required

1. Named, registered clinicians. The clinic publishes specific doctor names with registration numbers that verify in the national register.
2. Written price transparency before travel. The clinic provides itemised quotes before the patient arrives — clinic fee, material cost, nursing, anaesthesia, interpreter if applicable.
3. Material traceability. The clinic provides lot numbers, manufacturers, and CE marking for implants, crowns, or any other medical device used.
4. Written clinical guarantee. Duration, scope, and transferability to the patient's home country in writing.

Red CAUTION — any of the following

1. Two or more of the VERIFIED criteria fail.
2. Trustpilot pattern analysis shows the weighted negative-outcome score above 15%.
3. Documented misleading marketing (claiming accreditations not held, inflated patient counts, fabricated before/after images).
4. Pattern of revision / second-surgery requests within 18 months above 10% of reviewed patients.
5. Evidence of patient harm beyond expected complication rates (hospitalisations, permanent injury, death).

4. What we do when a clinic responds

Before publishing, we send every clinic we investigate a pre-publication notice. It lists:

• Every specific factual claim we will publish about them
• The source for each claim
• A 7-working-day window for written response

If the clinic responds with evidence a specific fact is wrong, we correct the fact and record the correction on our public Corrections log. If the clinic responds with general disagreement but no counter-evidence, we publish the clinic's statement verbatim in the article as a response block — and publish the original investigation alongside it.

We never delay an investigation past 14 days because of a clinic response unless the clinic has provided concrete counter-evidence under review. Stalling responses are not a reason to hold a story.

5. What we cannot do — published limitations

Transparency about weaknesses matters as much as transparency about methods.

• We don't run clinical trials. We can't tell you if a specific technique will work for your specific anatomy. That's a clinician's job, not a publication's.
• We can't verify individual patient claims without corroboration. A single dramatic Trustpilot review isn't evidence; a pattern across hundreds is. That's why sample size matters more than any individual story.
• We have a bias toward Albania. We coordinate patients to Tirana clinics. That's declared openly. In comparison articles, we control for this by applying identical criteria to every clinic and publishing the scores, not just the verdict.
• Translation risk. Trustpilot reviews in Turkish, Italian, or Greek are translated. Translation inevitably changes tone. We sample in original language where possible and note when translation was involved.
• Data freshness. Clinic quality can change in a year. An investigation published in April 2026 describes the clinic as of April 2026. We flag when an investigation is older than 90 days and a refresh is pending.

6. Update + correction policy

• Scheduled reviews: every investigation re-read at 90 days and annually.
• Material changes: if we learn something that would change the verdict, we republish within 14 days with a clearly-flagged update.
• Corrections: every factual correction is logged on the Corrections page with original language, corrected language, source of correction, and date.
• Archive: every version of every investigation is archived. If you want the April 2026 version of an investigation that has since been updated, contact us at info@clinictruth.com.

7. Frequently asked

How many reviews do you read per clinic?

A random sample of 200–500 Trustpilot reviews per clinic, drawn from the full available history via stratified sampling. For clinics with fewer than 500 reviews, we read every review available.

How do you pick which clinics to investigate?

Patient demand signals via WhatsApp/email + UK/IT/ES/EL search volume + regulatory/press signal. The queue is public — email us if you want a clinic added.

Can a clinic pay to be reviewed or to change a verdict?

No. Clinics have never paid Clinic Truth and cannot. If a clinic contests a specific factual claim with evidence, we issue a public correction — we never remove a verdict.

What does a green VERIFIED stamp actually mean?

The clinic met our four published criteria: named registered clinician, price transparency before arrival, CE-marked material traceability, written guarantee. It is a procedural assessment. It is not a medical endorsement.

What does a red CAUTION stamp mean?

At least two of: failure on criteria, Trustpilot weighted-negative score above 15%, documented misleading marketing, revision-rate above 10%, evidence of patient harm. Every CAUTION article explains which factors triggered it.

How often are investigations updated?

Every investigation is scheduled for re-review at 90 days after publication and annually thereafter. If a clinic materially changes its practices between reviews, we update earlier with a public update note.

This page is version 1.0 and will be revised as the publication matures. All historical versions are archived on request.