Health Products Compliance Guidance
U.S. Federal Trade CommissionSupports: The need for truthful, non-misleading health-product advertising and competent substantiation.
Open primary source ↗Loading…
Free review beta
Wellness Claims & Citation Checker
Catch wording that deserves a harder look before it reaches a website, social post, email, or staff script. The checker flags; a qualified human decides.
5
flag classes
4
review classes
Free
no account, no sign-in
The tool
Run the Claims & Citation Checker
Paste a draft — a service page, a social caption, an email, a line the front desk says out loud — and the screen marks the clauses a reviewer will stop on. It flags; it never clears.
Preparing Claims & Citation Checker…
This automated screen identifies language and citation-review issues; it does not determine whether a claim is lawful, medically accurate, adequately substantiated, or appropriate for a specific audience. It is not legal or medical advice and does not approve publication. Health claims require qualified human review.
Sample walkthrough
A worked example: one draft sentence, three flags
Fictional draft copy: “Our cold plunge protocol treats chronic inflammation and members see 40% faster recovery — results guaranteed.” Here is how the screen reads it.
The claim units detected
- “treats chronic inflammation”
Treatment language
- “40% faster recovery”
Quantified outcome, no mapped citation
- “results guaranteed”
Guarantee
What the screen returns
- Treatment language
Blocker — disease-treatment phrasing routes straight to qualified human review
- Quantified outcome
High review — a numeric claim with no source link; citation coverage reads 0 of 2 review-worthy units
- Guarantee
Blocker — outcome guarantees are one of the named high-risk patterns
- Rewrite prompts
Neutral alternatives are suggested for the reviewer to consider — never an approval badge
- Review packet
Draft, flags, links, and warnings export together for the human who decides
The sample sentence is fictional. A clean screen is not approval: accuracy, substantiation, audience fit, disclosures, and legal compliance still require qualified human review.
What you get
What the checker flags — and what it never approves
Clause-level flags for treatment language, guarantees, causal or superiority claims, quantified outcomes, and implied-typical testimonials.
A citation-presence score showing which claim units have a mapped source link.
Neutral rewrite prompts and a reviewer checklist—never an approval badge.
A downloadable review packet containing the draft, flags, links, and warnings.
Keep going
The riskiest claims are the ones nobody typed
The fieldbook writes down what the front desk says at visits one, two and three, which is where a guarantee gets made out loud and never reaches a screen.
PDF · 6 pages
The First Three Visits Fieldbook
A practical journey map, staff conversations and a 14-day experiment to turn a good first visit into a clear next step.
Download the fieldbookRecovery Studio Pro — The Weekly Brief
One email a week for people running physical recovery and wellness studios, written by the team building Praxium.
Free. Unsubscribe any time. We store your address for the brief, notify our team of the signup and hash it for ad measurement: Privacy policy
Methodology
How claim flags are triggered
Method 0.9.0-beta, published 2026-07-22. Next review: 2026-08-22. This review is overdue.
- 01The beta uses deterministic phrase and pattern rules, not an automatic legal or medical judgment.
- 02Citation coverage measures whether a review-worthy claim is linked to a source; it does not judge whether the source is adequate or the claim is lawful.
- 03Blocker and high-review labels route copy to a human. The tool never publishes or promotes a claim.
Source ledger
What supports the checker
Each source is tied to the part it supports; a link is not treated as blanket approval for every output.
General Wellness Policy for Low Risk Devices
U.S. Food and Drug AdministrationSupports: The distinction between general-wellness positioning and disease-related claims; January 2026 guidance.
Open primary source ↗Toolkit
Pairs with the rest of the toolkit
Every tool keeps its assumptions and limits visible, so a result from one is safe to carry into another.
Live tool
Launch & Expansion Planner
Build a dependency-aware, phase-ordered launch checklist for a new recovery studio, a new location, or an added wellness modality.
Open the Launch & Expansion PlannerPreview — not a finished report
State of Recovery Studios
Preview the measurement framework, data dictionary, exclusions, and release standard for Praxium's future State of Recovery Studios dataset.
Open the State of Recovery StudiosLive tool
Studio Economics Calculator
Model capacity, session economics, fixed and variable costs, debt service, break-even utilization, and payback from your own recovery-studio assumptions.
Open the Studio Economics CalculatorPreview — not a finished report
Market Gap Explorer
Explore modality representation in a dated, source-backed recovery-studio research sample—with denominators, missingness, and confidence shown.
Open the Market Gap ExplorerThe Weekly Brief
Recovery Studio Pro
Where the method changes get published: refreshed data freezes, what moved in the inclusion rules, and the reading behind the tools on this page.
See what the brief coversClaims & Citation Checker
Frequently asked questions
Does a green or neutral result mean my claim is compliant?
No. It only means the current deterministic rules did not find one of their named patterns. Accuracy, substantiation, audience fit, disclosures, and legal compliance still require qualified review.
Does the checker validate research papers?
No. It records whether a claim unit has a mapped citation link and prompts a reviewer to evaluate relevance, quality, product fit, and wording.
Where is my draft processed?
In a frame served by app.sprinter.ai, a separate origin from this page — so the text you paste leaves your browser and reaches that service. The screen is free and needs no account or sign-in, but it is not a local-only tool: do not paste health records, member names, or anything else you would not send to a third-party service.
Is my draft sent to an AI model?
No. The beta runs deterministic phrase and pattern rules, not a language model. Every flag it raises is one of the named patterns set out in the method below.
Can I say cold plunge reduces inflammation in my marketing?
The checker cannot answer that — by design. It will flag the phrasing as a health-outcome claim that needs a mapped citation and human review. Whether any version is appropriate depends on your evidence, wording, audience, and jurisdiction; the FTC health-products guidance in the source ledger is the right starting point for that conversation.
What counts as a citation for the coverage score?
A source link mapped to the specific claim unit. Coverage reports presence only — it does not judge study quality, relevance to your service, or whether the evidence level matches the strength of the wording. That judgment stays with the reviewer.
Will the rules catch every risky claim?
No. The beta screens for named deterministic patterns; novel phrasing, implication, imagery, and context can all carry risk the rules do not see. Treat a neutral result as 'no named pattern found', not as clearance.
A flag is only useful if it reaches the person who decides.
Praxium keeps the draft, the flags and the reviewer's call on one record, so the same sentence does not get re-argued the next time it is used.