Corrections
Our Editorial Policy says we fix errors promptly and openly. This page is what makes that checkable. Every correction to something a reader could already have read appears here, permanently, with the old claim stated plainly enough to recognise.
| Date | Page | What we got wrong | What it says now | Found by |
|---|---|---|---|---|
| 2026-08-21 | GLP-1 pills: the three that are real, and what they cost — material | We reported that Eli Lilly published no self-pay price for Foundayo, its orforglipron tablet. The GLP-1 pill page showed “not published” in the price column, the cheapest-routes FAQ said we could not compare Foundayo on price, and five other pages called the $149 Wegovy tablet the cheapest published self-pay route to any approved GLP-1. | Lilly publishes a self-pay price for Foundayo from $149 a month at the 0.8 mg dose, rising to $199 at 2.5 mg and $299 above that, alongside a savings card bringing it to as little as $25 a month with commercial insurance and $50 a month on the Medicare GLP-1 Bridge. That $149 ties the Wegovy tablet rather than losing to it, so the two cheapest FDA-approved GLP-1 medications in the United States are both tablets at the same price. Every page that named a sole cheapest now says so. | Reading the pages that outrank us for cheapest-GLP-1 queries, to check we were not missing entities they covered. A retail pharmacy page referenced a Foundayo savings card our own record said did not exist; Lilly’s own page confirmed it, and the price had been published for some weeks while our figure sat at null with a July verification date. Nobody reported it. A price we record as absent is a claim like any other, and this one needed re-reading sooner. |
| 2026-08-21 | The cheapest way to get a GLP-1 — material | Four pages said the $149 Wegovy tablet undercut every compounded telehealth total we had been able to derive — the cheapest-routes comparison in a pull quote, the GLP-1 pill page, the compounded semaglutide guide, and the semaglutide cost page. | It undercuts all but one of them. Mochi Health publishes a $79 monthly membership and compounded semaglutide from $59, which totals $138 a month against the tablet's $149. The exception is now named on the semaglutide page along with the detail that narrows it: that programme bills per four-week supply, so a year takes thirteen of them rather than twelve. | A chart we built for the new semaglutide cost page, which derives its bars from the same cost model the tables use and drew a compounded bar shorter than the tablet bar while the prose above it said that could not happen. The sentence was written before Mochi's figures were last re-read and nobody checked it against the model afterwards — the arithmetic to catch it had been sitting in the codebase the whole time, unasked. |
| 2026-08-15 | What we cover, and what we don’t — material | Our coverage page listed “Titration schedules by product” among the subjects we have decided not to write at all, and told readers that the dosing guide “explains why doses climb slowly without publishing the numbers”. The entry for dosage charts said, in the same terms, that publishing a dose-by-dose chart was “the one thing the drug pages exist to refuse”. | Titration schedules are covered, by the titration tracker, which has published each product’s week-by-week ladder in milligrams since 12 August 2026 — three days before this page stopped saying otherwise. The refusal we described was aimed at charts assembled from other websites and still applies to those; the tracker reads each ladder from the product’s own FDA label and is built around the sentence those charts omit, that the early doses are initiation and escalation dosages rather than approved maintenance dosages. What remains deliberately absent is narrower and is now stated as such: converting syringe units into milligrams. | Our own review while planning the next piece of work. The cause was ours, and it is the second time this exact shape has appeared: we built a page and left the map that describes our coverage behind it, so the one page on this site whose entire purpose is to report our gaps honestly was itself describing a gap that had been filled. Nobody reported it. |
| 2026-08-13 | The cheapest way to get a GLP-1 — material | The cost comparison priced a route as 'Weight Care membership plus brand-name Wegovy' through Alloy at $99 membership plus $199 medication, and priced Alloy's compounded option as liraglutide at $80. It also carried Calibrate at roughly $1,500–$1,800 a year, Found's Zepbound at $349, Mochi's medication at $99, and Sesame's programmes 'from roughly $129/month'. | Every one of those figures now matches what the 11 August provider sweep actually read. The Alloy brand-name route has been removed entirely, because Alloy no longer publishes those prices and our own directory said so two days before this page stopped agreeing with it. Calibrate is $199/month on a three-month minimum, Found lists Wegovy and Zepbound from $650, Mochi's compounded semaglutide is from $59, and Success by Sesame is $249/month with medication included. | Our own review of the page while planning improvements to it. The cause was ours: the sweep updated lib/providers.ts and the directory, and left the cost model behind, so for two days two pages of this site disagreed about the same prices. A build-time guard now refuses to compile if the cost model falls behind the directory sweep. |
| 2026-08-12 | GLP-1 comparisons — material | We said head-to-head efficacy comparisons between individual GLP-1 products were missing from this site on purpose, because the trial figures 'come from separate trials' and a comparison table 'would imply a comparison nobody has actually run'. | SURMOUNT-5 is a randomised head-to-head. It put 751 adults with obesity on either tirzepatide or semaglutide for 72 weeks and reported −20.2% against −13.7%. It had already been published when we wrote that the comparison did not exist, and we had already built a page on it — the hub introducing that page was the one still saying otherwise. The hub now leads with the trial. The reasoning we gave still applies to lining up separate trials in a winner table, and that part stands. | Our own audit of what each page adds that competing pages do not, which flagged this hub as the thinnest on the site and turned up the contradiction. Nobody reported it. |
| 2026-08-11 | Find a provider — material | Our Calibrate listing linked to calibrate.com, and said the programme cost roughly $1,500–$1,800 for a 12-month commitment. | Calibrate's site is joincalibrate.com. The bare calibrate.com domain is an unrelated parked page — a bare list of cryptocurrency and NFT links, with nothing to do with weight loss or healthcare — so anyone who clicked through from us landed somewhere that had no connection to the provider we were describing. Calibrate's own pricing page now states $199/month with a three-month minimum, which it puts at $597 for the initial commitment. | Our own scheduled provider sweep on 11 August 2026, which reads each provider's published pricing page. Nobody reported it. |
| 2026-08-11 | Find a provider — material | Our Sesame listing said there was an optional 'Success by Sesame' subscription from $59/month, and that GLP-1 programmes were advertised from roughly $129/month. | Sesame advertises Success by Sesame at $249 per month with the medication included. Individual visits start at $34, the Wegovy pill is listed from $149/month, and a Costco-branded programme from $49/month. We cannot establish whether the price moved between our two checks or whether we mis-read Sesame's tiers the first time, and we are not going to guess in the flattering direction: either way, a reader who took our figure to Sesame would have found the bundled programme costing about twice what we said. | The same 11 August 2026 provider sweep. Nobody reported it. |
| 2026-08-11 | Find a provider — material | We said the public comment period on the FDA's proposal to leave semaglutide, tirzepatide and liraglutide off the 503B bulks list closed on 29 June 2026. | The original deadline was 30 June 2026, and it was extended to 30 July 2026 by a Federal Register notice published 26 June. The page now states the extended date and links to our update, which carries the full timeline and sources. | Our own check against the Federal Register while researching a separate page. Nobody reported it. |
What counts as a correction
A change to something a reader could already have read, where the previous version could have misled someone who acted on it. That is a deliberately strict test, and it cuts both ways.
- Logged. A wrong figure, a misstated approval, a price that was never what we said it was, a claim attributed to a source that does not support it.
- Logged, and marked material. Any of the above where the old version could have changed a health or money decision — a dosing implication, a safety claim, a cost comparison someone might have bought on.
- Not logged. Fixing a draft nobody saw, rewriting for clarity where the old wording was not wrong, adding new material, or updating a figure because the underlying fact changed. That last one is an update, not a correction, and the page’s own “last updated” date carries it.
The distinction matters because logging every tweak would bury the corrections that count under housekeeping, and logging none would make our Editorial Policy an empty claim.
How we handle one
- We check it against the primary source first — the FDA label, the trial publication, the provider’s own pricing page — not against another website that agrees with us.
- We fix the page and log it in the same release. The two are not separate tasks, so the log cannot fall behind the site.
- Entries are permanent. Nothing here is removed, and nothing is quietly reworded. A corrections log that can be rewritten is not a record.
- We say how we found out. Including when a reader told us, because a site that only ever catches its own errors is describing a process nobody outside it can see.
Found something wrong?
Tell us at corrections@glp1.healthcare. Corrections that could affect a reader’s health decisions are treated as urgent. You do not need to prove the error — pointing at the claim is enough, and we will do the checking against the source.