What a Clinic Listing Is Worth Seventeen Days Later

2026 ·8 minute read · By Sumesh Peringeth

In short

A sweep on 8 August 2026 left seventeen clinic and diagnostics rows queued for the record. Before any were published, every operator's own site was read again on 25 August, seventeen days later. Twelve were recorded. One of those twelve, Medinstitut Berlin, went in at a corrected price: the sweep had recorded about EUR 4,000, which turned out to be the top of a published range starting near EUR 650. One row was reversed, because the operator's own site describes general multi-department outpatient care and does not carry the opening date the record had relied on. One lost the price it was recorded with, because the operator no longer publishes a price. One could not be read on either day, and an unreadable site is not evidence of anything. Two of the seventeen were already in the record from an earlier pass.

A directory entry is not a fact. It is a claim about a business on a particular day, and the business does not know you made it.

On 8 August 2026 a verification sweep left seventeen clinic and diagnostics rows queued for the record. Most had cleared the delivery bar on their operators’ own sites. A few were already sitting on the line. The obvious next step was to work through them and publish what passed.

They were read again first. On 25 August, seventeen days after the sweep, every one of the seventeen operators’ own sites was opened a second time and judged fresh, row by row, against the bar rather than against the first pass’s verdict.

Twelve were recorded. Three came out somewhere other than where they went in. A fourth could not be read on either day, which is a different kind of answer and the most useful one here.

The one that went in at the wrong price

Medinstitut Berlin is a check-up practice in Berlin-Steglitz selling graded examinations to private individuals, ending in a summarizing internal-medicine assessment. It cleared the bar in August and cleared it again on the re-read. It is in the record.

What changed was the number. The sweep had recorded its price as about EUR 4,000. On the second reading the operator’s own page describes a starter program from about EUR 650, with packages running between roughly EUR 1,200 and EUR 4,000 depending on what is added.

EUR 4,000 was the top of a range. Published as the price, it would have told a reader that the cheapest way into this clinic costs roughly six times what the operator says it costs. The row went in as a range instead, with the low figure as the entry point.

That is the error worth dwelling on, because it is not a factual mistake about the world. Every number in the sweep was on the operator’s page. The mistake was reading one number out of a set of numbers and calling it the answer. A single figure is the format a directory wants. It is often not the format a clinic sells in.

The one that was reversed

A Tokyo clinic had passed in August on an announcement page. The announcement described preventive care with new endoscopy, MRI and CT equipment, and gave an opening date earlier in 2026, which was what took it past the rule against recording something that has not opened.

On the second read the operator’s own site describes outpatient care across internal medicine, gastroenterology, orthopedics, ophthalmology and gynecology, with health checkups alongside. It states no results consultation. The opening date the first pass had relied on is not on the site and could not be verified from it.

So the verdict was reversed and the row was not recorded. Nothing here says the clinic is anything other than what it says it is. A general multi-department outpatient clinic is a perfectly good clinic. It is simply not the thing this record is a record of, and the bar the Atlas applies is whether the person who pays receives a result and a reading of that result.

The instructive part is the shape of the error. The first read was of a page written before the clinic opened. The second was of a page written after. Both were the operator’s own site. An announcement is a plan, and a plan is the most perishable thing a business publishes.

The one whose price disappeared

A London full-body scanning operator had been recorded in August with a published price of GBP 1,650, taken from its own page. On the re-read that page publishes no price at all, only instalment financing. No location is stated either.

The row could not be filled completely, so it was not filled. It is held.

Worth being precise about what this is and is not. It is not a price increase, a price cut or a business in trouble. It is an operator that used to publish a number on a page and now does not, which is a thing operators do all the time and owe nobody an explanation for. The only party with a problem is the directory that wrote the number down and moved on.

The one nobody could read

The fourth was not a change at all. It was a failure to look.

A major Korean hospital’s health promotion center has a site that returns a successful response carrying almost nothing: a few hundred bytes whose entire visible text is the center’s name. The page assembles itself in the browser, and neither of the two automated transports tried could see what it assembles. That was true in August and true again in the re-read.

Nothing was recorded about it, and nothing will be until a person opens it in a real browser. This is a standing rule and it matters more than it sounds. A page that will not load is a transport failure, not a finding. Treating an unreadable site as evidence that a service is absent is how a directory manufactures facts out of its own network problems, and it is very easy to do by accident, because a blank result and a genuine absence look identical in a spreadsheet.

It also means this pass cannot report a clean rate of change, and does not. One of the seventeen is a row nobody has read. A rate calculated over a denominator that includes it would be counting an unknown as a pass or as a failure, and it is neither.

Two that were never new

Two of the seventeen were already in the record, entered through an earlier country pass, and were not duplicated.

That is the least dramatic finding here and the one that most changes the arithmetic. A batch of seventeen candidate rows was never seventeen new listings. A count taken at the front of a pipeline is not a count of anything that shipped, and the gap between those two numbers is invisible unless somebody writes both of them down.

What this actually argues

The instinct after a pass like this is to promise a better first check. That is the wrong lesson and it is not available anyway. Nothing about the August reads was careless. Every figure came from the operator’s own page. The pages changed, or the pages were of a different kind than they looked, or the pages did not load.

A clinic listing is a perishable good. Prices move, pages get rewritten, a pre-opening announcement becomes an operating site, and a business quietly stops publishing a number it used to publish. Seventeen days is not a long time, and seventeen days was enough.

Which leads to the thing worth taking away. A directory that shows you a listing is showing you a claim with no date on it, and a claim with no date on it is asking you to assume somebody checked recently. Usually nobody knows. The record’s answer is to carry the date each row was last checked against its own sources, on the row, drawn from the data rather than from whenever the site was last rebuilt. The methodology page and the data page both report how fresh the record is from those dates.

That date is not a footnote. It is most of what a record is worth. Anyone can copy a price off a clinic’s website. The question a reader actually needs answered is when, and a directory that does not say when is telling you less than it appears to.

What this pass did not do

It did not re-read anything already published. Re-reading these seventeen was possible because they were candidates rather than rows on the site, and the standing rule is that a published entity is not quietly re-litigated.

It did not check any operator against anything except that operator’s own site. No press release, no aggregator and no third-party listing decided any of these outcomes.

It did not test whether these changes are typical. Seventeen rows in one sweep, read twice seventeen days apart, is one observation about one batch. It is not a decay curve and nothing here should be read as one.

Three of the four operators above are named only by country and type. They are candidate rows and not entries in the record, and a candidate is a working note about whether a business belongs here, not a published finding about that business. The one named is named because it is recorded, and because the correction is to our number rather than to anything it did.

And none of this says anything about the quality of care anywhere. Out of scope is a statement about this record’s boundaries. It is never a statement about a business.

The Atlas of the Healthspan Economy is a neutral record of organizations working on healthspan. It does not recommend any provider and it does not give medical advice. Each recorded organization carries the date it was last checked against its own sources. Read the methodology, see the data page, and for the scope rule these judgments apply see What Counts as a Longevity Clinic.

A note on what this piece does not claim. No operator described here is said to have misled anyone. A price that moves, a page that is rewritten and a site that will not load for an automated reader are ordinary things. Where a row was not recorded, that is a decision about this record’s scope and completeness rules and not a judgment about the operator’s business or its care.

Related insights

What Abu Dhabi's Longevity Clinic Standard Actually Requires
The Healthy Longevity Medicine Clinic Standard is the rare case of a regulator writing down what a longevity clinic has to do. Its mandatory diagnostic list is conventional internal medicine. Telomere length, liquid biopsies and senescent cell counts sit under Not Recommended outside of research protocol. IV NAD, IV stem cells, gene therapy, exosomes, plasmapheresis and hyperbaric oxygen are all coded Not Currently Applicable / Needs further data.
Half of what calls itself a longevity clinic is not one
When a longevity-branded batch is checked against what each operator actually delivers, a large share turns out to be wellness or aesthetics, not longevity medicine. The problem is scope, not honesty, and one regulator now draws the same line.
Italy Has No Longevity Clinic Category, and the Reason Is Structural
Two Italian instruments explain the shape of the country's private longevity market better than any commercial account of it. Article 8-ter types healthcare structures by delivery mode, so a longevity category has nowhere to sit. A 2015 ministerial decree confines non-routine advanced therapy medicinal products to research institutes, university hospital trusts and public hospitals, so a private clinic cannot lawfully supply them. Both are structural, and neither was written with longevity in mind.
Four Countries, Four Regulators, and No Longevity Clinic Category
The longevity clinic category exists commercially and, in these four markets, not legally. Hong Kong regulates the premises under Cap. 633. Canada regulates whether a service may be billed to a patient. Thailand regulates what a facility may say about itself. Vietnam regulates the practitioner, on a five year cycle with a competency assessment phasing in. Compliance in one predicts nothing about the next.

Questions

Frequently asked questions

How long is a clinic listing good for?

There is no general answer, which is the point. In this pass, seventeen rows read on 8 August 2026 were read again on 25 August, and three of them came out somewhere other than where they went in. A fourth could not be read on either day. The honest thing a directory can do is publish the date it last looked, so a reader can judge the age of the claim rather than assume it.

Does a re-read mean the first check was careless?

No. Each change came from something the operator's own site said on the second day and had not said, or had said differently, on the first. A price page changed. A pre-opening announcement was replaced by an operating page. A verification pass records a business as it publishes itself, and businesses republish themselves.

What happens when an operator's own site cannot be read?

Nothing is recorded. A page that returns an error, a challenge or an empty shell is a transport failure, not a finding, and the Atlas does not treat it as evidence that a service is absent. The row is held and joins a queue to be read by a person in a real browser.

Where can I see when a record was last checked?

Each recorded organization carries the date it was last checked against its sources, and that date is data on the row rather than the date the site was last rebuilt. The [methodology](/methodology) page and the [data](/data) page both report the freshness of the record from those dates.