How This Site Works
Where the numbers come from and what we will not publish.
Where the numbers come from and what we will not publish.
First-year all-in cost
One dataset behind every page
Every table, ranking, comparison and calculator on this site reads the same file: 32 programmes, all-in cost at six dose tiers, first-year totals, fee structure, commitment terms, care model, pharmacy disclosure, evidence status and source, each with a capture date. Change the file and every page changes with it. Nothing is assembled by hand.
That has one consequence worth stating plainly: we cannot quietly move a programme up a table, because there is no table to edit. There is a dataset and a sort order, and both are published.
What we do before publishing a price
Read it on the provider's own pricing page and record the date; or take it from the manufacturer; or take it from a named editorial source that states its own verification date. Anything else is labelled as unconfirmed, and where two sources disagree we show both rather than choosing.
Why some rows are blank
Because a blank is more accurate than a guess. A programme that reveals cost only after an intake, or that uses split pricing without publishing a medication cost, gets a row with no figure and a page explaining what to ask. Filling that in with an estimate would corrupt the only thing this site sells.
How to check us in five minutes
Download the dataset, sort it by first-year cost, and compare against our cheapest ranking. They should match. Then open any provider's own pricing page from the link on its review and check our figure against theirs. If they differ, ours is wrong and we want to know.
How this page fits the rest of the site
Everything here rests on one dataset: all-in monthly cost at six dose tiers for every tracked programme, with a capture date and an evidence status on each record. Rankings sort that dataset on a stated criterion; comparisons read the same rows; the calculators run against the same file. Nothing on this site is assembled by hand, which is why a policy page can describe a rule rather than an intention.
What we publish when we are not sure
The absence. A programme that does not publish a price gets a page saying so and listing what to ask before handing over a medical history. A figure taken from a third-party round-up is labelled as unconfirmed rather than presented alongside verified prices as though the two were equivalent. Where two sources disagree we show both.
That produces a shorter table than competitors publish. It also produces one that survives being checked, which is the only durable advantage a comparison site has.
The disclosures that matter more than any policy
Compounded preparations are not FDA-approved and are not reviewed by FDA for safety, effectiveness or quality before marketing. Tirzepatide is a prescription medicine and nothing here is medical advice. No programme pays for placement or position, and rankings are computed from the published dataset rather than assigned.
Tell us when we are wrong
Prices move, terms change and programmes leave the market. If a figure here differs from a programme's own page, ours is wrong. Send the URL through contact and the change is logged in corrections with the date it was made.
The standard this site holds itself to
Publish what can be verified, label what cannot, and record what changed. Those three rules generate everything else: why 11 of 32 programmes carry a verified tag and the rest do not, why some rows are blank, and why corrections appear on the page where the error was rather than in a log nobody reads.
They also explain what is missing. There are no ratings out of ten, no star scores and no aggregate quality judgements, because we cannot measure service quality from outside and inventing a number for it would corrupt the figures that are real.
Why the dataset is published rather than described
Every table on this site is generated from one file: 32 programmes with all-in cost at six dose tiers, first-year totals, fee structure, commitment terms, care model, pharmacy disclosure, evidence status and source, each with a capture date.
That file is downloadable. Sort it by first-year cost and it should reproduce our cheapest ranking exactly. If it does not, that is a bug or a lie, and you can find it without our help. A comparison site that will not publish its own data is asking to be trusted rather than checked.
The commercial conflict, named
Comparison sites in health are funded by the companies they rank. That creates an obvious incentive to rank the highest-paying option first and describe the arrangement vaguely, and it is the reason readers are right to be sceptical of every site in this category.
The defences available are boring and checkable: compute rankings from a published dataset on a stated sort key, publish the dataset, log corrections publicly, and record negative facts about highly ranked programmes. None of that proves good faith. It makes bad faith detectable, which is the most any publisher can honestly offer.
What good looks like in this market
A programme that publishes its price at every dose tier, names the pharmacy that compounds the medicine, states its cancellation terms before you pay, and can decline to prescribe. Those four together are rarer than they should be, and they cost a programme nothing except the ability to advertise a number that is not the number.
The cheapest tracked route currently runs $215 a month all-in at a 10 mg maintenance dose, about $2,580 for a first year. That figure is what a well-informed buyer should be measuring every other offer against.
Where this sits against the dataset
The figures behind this page come from one file: 32 programmes, all-in cost at six dose tiers, terms, care model, pharmacy disclosure, evidence status and source, each with a capture date. All-in cost at a maintenance dose runs $215 to $548 across it.
Change the file and every table, ranking and calculator changes with it. There is no separate editorial layer to adjust, which is the structural reason a policy page here can describe a rule rather than an intention.
What it would take to prove us wrong
Download the dataset, sort it on the criterion named in any ranking title, and check whether our published order reproduces. Then open any provider link and compare our figure to theirs. Both checks take minutes and neither requires trusting us.
If they disagree, that is a bug or a lie and we would rather hear about it than have it found later by someone with less reason to be generous.
Open these rather than taking our word for it. Every one is a regulator, a trial registry, a label, an accreditor or the manufacturer.