Editorial policy

Doseworth covers the cost of prescription weight-loss and metabolic treatment in the United States: what telehealth providers and pharmacies charge, what the charge actually resolves to over a year, and what the companies doing the charging have said about themselves on the public record.

We sell nothing. There is no product, no plan, no consultation and no subscription available on this site. We are funded by affiliate commission on outbound links, which is set out in full — including what it does not buy — on the disclosure page.

A price is only publishable here if it came off the provider's own pricing or checkout page. Not an aggregator, not a press release, not a comparison site, not another publisher's table, and not a figure someone remembers being true. First-party or it does not run.

Every published price row carries two things it cannot be published without:

• A capture date. The day the number was read. It appears next to the price on the page, because a price without a date is a claim nobody can check or age out.
• A source URL. The specific provider page it was read from, so you can go and look at what we looked at.

This is enforced by the data layer rather than by anyone remembering. A price row missing its capture date or its source fails validation when the site is built, so it is not possible to ship an undated or unsourced number by accident.

The advertised number is almost never the number you pay. The headline figure on a provider's page is usually a prepay rate — the monthly equivalent if you commit to three, six or twelve months up front — or a first-month promotional rate that expires. Two terms do the work of separating that from reality:

Walk-away price

What one cycle actually costs if you stop after it. No prepay commitment, no promotional first month, no bundle. It is the price of not being locked in.

Year at this price

The walk-away price multiplied by the number of times you are actually billed in a year — which depends on the billing cycle, not on the calendar.

Billing cycle is where the arithmetic bites. A provider billing monthly charges 12 times a year. A provider billing every four weeks charges 13 times — four-weekly is not monthly, and the thirteenth charge is a full extra payment that never appears in the “$X/mo” headline. We always annualise on the cycle the provider actually bills on, and we label which one it is.

The walk-away price is a required field on every price row in our data, with no default and no fallback anywhere. A content file that supplies only an advertised price fails to build. That is deliberate: the failure mode this site exists to prevent is a page that quietly shows the prepay number alone, so it is made structurally impossible rather than left to editorial discipline.

Providers run split tests. Two visitors loading the same page minutes apart can be shown different prices, different intro offers, or a different structure entirely — and neither of them is a mistake on the provider's side.

When we hit one, the procedure is fixed:

1. Re-load the page in a clean session — no stored cookies, no prior visit, nothing carried over from a previous capture.
2. If a single price is stable across clean loads, that is the price, captured and dated normally.
3. If two prices persist, we publish the less favorable one and record the condition in the row's note. Publishing the better arm of someone else's experiment would mean printing a number most readers will never be shown.
4. Where the difference is a condition rather than a coin flip — a promotional code, a prepay tier, a first-order rate — we publish the unconditional price and state the condition alongside it rather than folding the discount silently into the headline.

Two kinds of board, two rules, and each board prints its own. These are the same strings the boards render, so this page cannot say one thing and a board another:

On a /best board, companies that pay us a commission are listed in a first group and everyone else follows; within each group the order is computed from the price rows. On a /cheapest board there is no partner group — every row is in price order. What commission never touches is the price: each figure is read from the company's own page, and the list of partners cannot write to a price row. Within a group, and on every /cheapest board, the order is re-derived from those prices on every render, so pre-sorting the input has no effect either.

The commercial half of this — what we are paid for, and what the money does not touch — is on the disclosure page.

Doseworth maintains a record of FDA warning letters issued to telehealth companies and pharmacies in this market. This is the part of the site with the most potential to be misread in both directions, so the policy is stated plainly.

A warning letter is FDA writing to a company to say that, in the agency's view, something it saw is out of compliance, and asking the company to respond. Nearly every letter in this market concerns marketing and labeling — claims the company made about a product in its own advertising, cited under sections 502(a) and 502(bb) of the Federal Food, Drug, and Cosmetic Act. That is the same category of letter a long-established manufacturer receives for overstated ad copy. It is a statement about what was said, not a finding about what was made.

Which means, explicitly, that a warning letter is not:

• a recall;
• a fine or a penalty;
• a criminal charge;
• an injunction or a seizure;
• a court finding that the company broke the law;
• evidence that the product is unsafe, contaminated or badly made — unless the letter itself concerns product quality, which we label separately when it does.

Our rule: we disclose letters, we do not delist providers for them. A company that has received a warning letter keeps its review, keeps its rows on the boards, and keeps its place in price order. What it gains is a visible notice, dated, naming the drugs at issue, quoting what FDA actually wrote, and linking to the letter on fda.gov so you can read the primary document rather than our summary of it. The only thing that removes a company from this site is ceasing to operate — because at that point there is nothing left for a reader to buy.

Getting this wrong in either direction causes harm. Treating a marketing letter as proof of a dangerous product defames a company over something it did not do. Hiding a letter because the company is a commercial partner withholds a public fact from a reader who is about to spend money. Neither is acceptable, and the middle position — publish it, quote it, link it, do not inflate it — is the policy.

Practically: every letter we publish is fetched from its own fda.gov URL before it is entered, never reconstructed from a summary or an index listing. What we describe as alleged is built from the letter's own quoted language, never from the boilerplate paragraph that appears identically in every letter of a given wave. Each entry records the letter's current status and the date we last checked it, because a letter shown as open long after FDA closed it out is a false, checkable statement about a real company. You can read the whole record at our FDA warning-letter index.

Outside pricing, a medical or efficacy claim on this site has to be tied to a primary source — peer-reviewed research or official product labeling — and cited, not paraphrased from a secondary write-up. Links are checked mechanically before the site builds, so a citation pointing nowhere fails rather than ships.

Where the evidence is weak, we say so instead of rounding it up. Where a question is unsettled, we say that too. “We could not confirm this” is a publishable result here, and it appears on the site as such.

Sites in this category tend to advertise credentials they do not possess, so here is the honest inventory of ours:

• No medical review board. Not a single article on this site has been reviewed by a clinician on our behalf, and none carries a “medically reviewed” badge, because it would be false.
• No laboratory and no product testing. We have never assayed a vial. Any page claiming otherwise would be fabricated.
• No certification, accreditation or audit. Nobody has certified this site, and we display no seal suggesting anyone has.
• No mystery-shopping of clinical care. We check prices and terms on public pages; we do not evaluate the quality of a consultation.

If a named clinical reviewer is ever engaged, they will be named, with their credentials, on the pages they actually reviewed — and not a page more.

Doseworth is edited by Marla Whitfield, Pricing Editor. Marla tracks what GLP-1 and peptide treatment actually costs — cash, insured, and compounded — and checks every published price against the provider's own checkout before it runs on Doseworth.

That byline claims a competence in pricing and billing structures and claims no medical credential, deliberately. The judgment being exercised on this site is financial: whether a number is real, what it becomes over a year, and what a provider's terms do to it. Clinical judgment is not on offer here — see the medical disclaimer.

No provider previews a page before it publishes, approves a page, or is given the opportunity to change one. Nobody buys a review, a placement, a rewrite, or the removal of an unflattering fact. An unflattering finding about a company that pays us stays in the page exactly as it would if they did not — a four-weekly billing cycle, a prepay requirement, a hard cancellation term or an FDA letter is reported the same either way.

If a provider disputes something we published, the route is the same one open to any reader, and it deals in facts rather than conclusions: the corrections page.

Prices go wrong. Providers restructure plans, our capture misses a condition, an annualisation uses the wrong cycle. When we get one wrong we fix it and we say that we fixed it, with what changed and when. The procedure, and what to send us, is on the corrections page. Reports go to editor@doseworth.com.