Clinic Reviews: Six Signs They Were Bought
You are looking for a clinic abroad. You find a profile: 4.9 stars, eight hundred reviews, everyone delighted. The anxiety eases — it feels as though the decision has already been made for you. That is exactly the moment to stop.
How many are actually fake
This is not a hunch. It is a measured figure. The Trustpilot Trust Report 2025 states that in 2024 the platform removed 4.5 million fake reviews — roughly 7% of everything published. Nine in ten were caught automatically, by machine learning, neural networks and generative models.
Google reports numbers of the same order. In 2025 it blocked or removed more than 292 million policy-violating reviews from Maps, removed over 13 million fake Business Profiles, blocked 79 million inaccurate edits, and restricted more than 782,000 accounts.
It matters what that percentage means. Trustpilot's 7% is the share of fakes that were found. Nobody knows how many went unnoticed. The real share is higher than that figure, not lower.
Regulators have recognised the problem too. In August 2024 the US Federal Trade Commission adopted a rule that explicitly bans fake reviews — including AI-generated ones, reviews paid for on condition of a particular sentiment, undisclosed insider reviews, and the suppression of negative reviews through threats.
For healthcare specifically, no reliable statistic exists. The one figure that circulates — that around 20% of reviews of healthcare organisations on Google and Yelp show signs of suspicious activity — comes from the head of a company that sells fake-review detection, and dates from 2021. We cite it as an interested party's estimate, not as a measurement. The verifiable numbers are the ones above, from the platforms' own reports.
Why medical tourism is a particular target
Four conditions come together here. Individually each occurs in other industries; together they are rare.
It is a once-in-a-lifetime purchase. The patient will not be back next month for a second operation, so the ordinary mechanism of reputation — the fear of losing a repeat customer — barely applies.
The price is high. One patient acquired covers the cost of a hundred bought reviews, and that arithmetic suggests an obvious move to an unscrupulous intermediary.
The patient is in another country. They cannot walk in and look around, cannot ask a neighbour, cannot see what the admissions desk is like on a Thursday evening.
And there is the language barrier. A Russian-speaking patient does not read local Turkish sources, which means they see exactly the shop window that was prepared for them.
Six signals
1. Speed
Sort the reviews by date and look for clusters. Twenty or more reviews in forty-eight hours is not a flow of patients, it is a purchase. Genuine reviews are spread unevenly over time, but without spikes like that.
2. An unrealistic star distribution
Look at the histogram, not the average score. If more than 95% of ratings are top marks and there are almost no threes and fours, that is unnatural. Every working clinic has unhappy patients: someone waited longer than promised, someone disliked the food, someone was spoken to sharply at reception. A missing middle points to a filter, not to a perfect clinic.
3. Single-use accounts
Click the reviewer's name. If they have written exactly one review in their life and it is about this clinic, that is a signal. Check three or four reviewers in a row; the picture becomes clear quickly.
4. Short, generic text
Fewer than fifty words and not a single specific: "wonderful doctors, everything went perfectly". A real patient names things — the surgeon's name, how many days they stayed, what turned out to be harder than expected. Look for the detail, not the emotion.
5. Divergence between platforms
Find the same clinic in two or three independent places and compare. If the picture is flawless in one place and ordinary in another, reviews are being managed somewhere. The divergence itself tells you more than either rating on its own.
6. Linguistic traces
Read five reviews in a row. What should give you pause: flawless language where the author is presented as a foreigner; the same turns of phrase from different people; text that reads like a translation. The question is simple — does one hand show through?
What to do if you find a signal
There is one rule, and it matters more than all six signals. The conclusion is not "this clinic is bad". The conclusion is "these reviews cannot be trusted, I need another source". Inflated reviews tell you how a clinic sells itself, not how it operates. Those are different things, and conflating them is unfair — to yourself and to the clinic.
Where the real information is
Independent forums and patient communities, where people are not writing at a clinic's request and where you can ask them a follow-up question. One answer to a follow-up question is worth twenty glowing reviews.
People who have already been treated. Do not ask whether they liked it; ask the specifics. What was added to the bill beyond the quoted sum? How many days did it actually take? Who did you contact once you were home and a question came up? A bought review cannot answer any of those.
And the clinic itself. Ask for a written offer listing what the price includes and what it does not. The blanks in such a document are information too.
What Adamiani does about this
Adamiani is a platform where a patient describes their case once and clinics send offers directly. We do not take a percentage of the cost of treatment, so we have no reason to steer anyone towards a more expensive option. We do not replace reviews and we do not rate anyone's quality: we turn the offers received into a comparison — what the price includes, what is missing, what to ask about. The choice stays with the patient and their doctor.