Promotional claims substantiation done before MLR review, every claim tied to the page that supports it
Every claim in a promotional draft tied to the page that supports it, before it reaches the review committee.
See one case, screen by screen ↓An agency sends a draft — an HCP email, a detail aid, a banner, a social post. Before it can go to the committee, someone has to find every claim in it, including the ones hiding in a subject line or a headline, check each against the approved claims library, find the page and sentence in the Prescribing Information or the study that supports it, and annotate the references for the reviewers.
Most of what reviewers send back is not a judgement call. It is overstated wording, a library claim used without its qualifier, an exploratory endpoint dressed up as a benefit, a comparison with no head-to-head study, or safety information from the label before last. Each one costs a round, and each round waits for the next committee slot.
Estimated split for a typical promotional piece, from agency draft to committee approval, by hand and with the solution.
Five specialist agents read each draft and check every claim against the library, its references and the label; the review coordinator decides each flag, and a sixth agent builds the pack for the committee.
At 07:52 Harborline Health sends PC-2144 “VELA-2 two-year results”, an HTML email to 18,400 US dermatologists. Thursday’s committee cut-off is 17:00 today. Here is what happens next, screen by screen, in the working solution.
Lena’s home screen leads with the MLR committee: Thu Oct 8 at 10:00, cut-off today at 17:00, 7 h 48 min left, 3 slots open. Five pieces wait for her, 7 flags the agents could not settle, 199 claims checked this week — and one new draft, not checked yet.
“Before a piece reaches MLR, agents link every claim to the claims library and the exact page of its reference. You decide what goes to the committee.”
Lena opens the new draft and runs the pre-check — about 40 seconds, and nothing changes in the source file. The piece reader takes apart the HTML email: subject line, body copy, images, safety block, links. The claim extractor finds 12 claims — 3 efficacy, 2 dosing, 2 safety, 2 access, 1 durability, 1 symptoms, 1 comparative. The library matcher links 11 to approved claims, the reference anchorer ties all 12 to 5 references and 9 pages, and the label and fair-balance checker flags 5 claims and 2 safety-text items. Readiness: 58 of 100.
“Agents pre-check only. You decide each flag; the MLR committee keeps the review and signs.”
“Clear skin that lasts: VELA-2 two-year data” reads as complete, permanent clearance for every patient. At Week 16, 56% reached PASI 100, and the two-year data come from an open-label extension — 52% PASI 100 at Week 104, as-observed. The checker names the rule, highlights the passage on page 6 of the VLT-305 extension, and offers approved wording CL-0118 with its qualifier.
Brand rule 3.2 · no absolute efficacy words · overstating effectiveness (21 CFR 202.1(e)(6))
Itch at Week 1 came from an exploratory analysis with a nominal p-value, not adjusted for multiplicity and not in the Prescribing Information; the VELA-2 publication calls it hypothesis-generating. The library has no approved claim for it. The nearest one is CL-0109, the key secondary endpoint: significant itch improvement at Week 16, 71% vs 19% placebo.
“Higher PASI 90 rates than other IL-23 and IL-17 treatments” has nothing behind it. VELA-2 compared veltrimab with placebo only, and its own discussion on page 9 warns against comparing across separate trials. Nothing in the 142 approved claims says this, so the suggestion is to remove it.
“Comparative claims need an adequate head-to-head study.”
“Suitable for patients with a history of tuberculosis” conflicts with Warnings and Precautions 5.3: evaluate for TB before treatment, never give veltrimab with active TB, and consider anti-TB therapy first when past treatment cannot be confirmed. Here there is no “keep with reason” — only remove the claim or ask the agency.
“A conflict with the label cannot be kept as written.”
The email carries ISI v3 from June 2025. The approved ISI is v4 of Sep 18, 2026, which adds the hepatotoxicity paragraph, and the “Full Prescribing Information” link still opens v3. The boxed warning, safety prominence and the unsubscribe and privacy links all match.
Lena uses the library wording for the subject line and the itch claim, adds the required qualifier to the Week 104 claim, removes the comparison and the TB line, replaces the ISI with v4 and points the link at Prescribing Information v4. Readiness goes to 100; 7 of 7 flags decided, each with her name and the time.
Send to MLR shows the committee — Thu Oct 8, 10:00, slot 8 of 10 — readiness 100 of 100 and 5 claim decisions on record. The reviewers are Dr. Maya Chen (medical), Omar Haddad (regulatory) and Dana Okafor (legal). Below the readiness threshold, or with a label conflict or old safety text still open, the piece cannot be sent.
The pack builder numbers the 10 claims left in the piece and annotates the supporting passage in each reference with the same number — 4 references, 6 pages. The pack is built for the committee and downloads as a PDF.
“PC-2144 is on Thursday’s committee · slot 8 · reference pack ready”
The same draft, seen by the people who carry it — what their week looked like, and what it looks like now.
Five specialist agents read each draft and check every claim against the library, its references and the label; the review coordinator decides each flag, and a sixth agent builds the pack for the committee.
Opens the draft in any format — PDF, DOCX, PPTX, HTML email, banner bundle, video script — and separates the copy, footnotes, safety text, links and image text, in reading order.
Finds every claim — benefit, number, comparison, dosing, safety, access and savings — including the implied ones in headlines, subject lines and supers, and labels each with its kind.
Matches each claim to the approved claims library — word for word, same meaning, or new wording — and carries over the qualifier the library requires.
Anchors every claim to the page and sentence, or table row, of the reference that supports it — Prescribing Information, published study or data on file — and says so when the passage says less than the claim.
Applies the readiness rules: consistency with the label, absolute or overstated wording, comparisons without head-to-head evidence, exploratory endpoints, the safety information version, boxed warning placement and links. It names the rule and proposes the approved wording when there is one.
Builds the annotated reference pack, with every claim numbered and the matching passage in each reference marked with the same number, and drafts the note back to the agency.
Decides each flag; reviewers sign. The agents propose; a named person decides.
The coordinator can ask about any piece, claim or reference — or tell it to change a rule. It answers from the claims library, the references and the pre-check results, with the page behind every answer.
What is blocking the VELA-2 email from MLR?
Readiness 58 of 100 (send at 90). Open items: the tuberculosis line conflicts with the label; “Higher PASI 90 rates than other IL-23 and IL-17 treatments” has no head-to-head study; the subject line is overstated; itch relief at Week 1 is not supported; the Week 104 claim needs its qualifier; the Important Safety Information and the Prescribing Information link are v3. Each has a suggested fix from the claims library, and accepting them takes readiness to 100.
Which pieces use the claim retired by the label update?
CL-0127 “No routine laboratory monitoring required” was retired on Sep 18 when Prescribing Information v4 added liver tests before treatment. Three pieces use it. The live HCP dosing page matters most — it needs a revision before anything else. The replacement claim is CL-0128 “Check liver enzymes and bilirubin before starting treatment”.
Add a rule: hold data on file older than 6 months
Done. New rule: data on file must be newer than 6 months. Over the pieces in flight it finds one: the payer slide PC-2135 cites formulary data from Jan 9, 2026, nine months old. Newer data on file exists from Apr 14, 2026 (91% of commercial lives), and that one is due for refresh on Oct 14. The rule is live in Settings and on the audit trail.
Draft the agency note for the Corventa social post
To Ana Lima, North Pier Creative, cc Leo Grant: three changes before Thursday’s committee. The headline “Itch-free nights start here” overstates the result — please use library claim CL-0208. “For anyone 6 and older” is outside the label — Corventa is approved from 12; please use CL-0201. Move the “Serious side effects” headline and boxed warning link into the first lines of the caption. Due Oct 9.
13 screens from the working solution, on its sample data. Pick one to see it large.
The rules the agents apply, who decides and where pieces come from are settings, not a project.
| Setting | Default | Choose from |
|---|---|---|
| Send to MLR at readiness | 90 | Up or down in steps of 5 |
| No absolute efficacy words | On | On · Off |
| Library claims keep their qualifier | On | On · Off |
| Exploratory endpoints cannot carry a claim | On | On · Off |
| Comparisons need a head-to-head study | On | On · Off |
| Safety text must match the current approved version | On | On · Off |
| Data on file must be newer than 6 months | Off | On · Off |
| Second look on “keep with reason” | Omar Haddad | Nobody · Lena Ortiz · Dr. Maya Chen · Omar Haddad · Dana Okafor · Priya Raman |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published medical-legal-regulatory review benchmarks (2024–2025: about 29 business days per piece). Regulations named (21 CFR 202.1, Form FDA 2253) are context only. People, brands, studies and agencies named on this page are characters and sample data in the working solution.
A check of a promotional draft before it reaches the medical-legal-regulatory (MLR) committee: every claim is found, matched to the approved claims library and tied to the page and passage of the reference that supports it. Overstated, unsupported or off-label claims and out-of-date safety text are flagged for the review coordinator to decide.
The piece reader opens the draft in any format — PDF, DOCX, PPTX, HTML email, banner bundle or video script — and separates copy, footnotes, safety text, links and image text. The claim extractor then treats any statement a reviewer would need a reference for as a claim, including headlines, subject lines and supers.
Yes. It checks that the ISI, boxed warning and indication statement match the current approved version, that safety is as prominent as the benefits, and that Prescribing Information links point to the current label.
A label watcher checks for new label versions every morning. When Veltrimab Prescribing Information v4 added liver tests, claim CL-0127 was retired, its replacement CL-0128 approved, and every piece still using the retired claim was listed — including a live web page that needs a revision.
No. The agents never edit the source file and only propose. The coordinator decides each flag, and the MLR committee keeps the review and signs. A conflict with the label cannot be kept as written.
Agency producers can upload drafts and see the findings for their own pieces. When a piece goes back, the coordinator sends a numbered note — what, why and the approved wording to use — drafted from the open flags.
Yes. The readiness score needed to send to MLR, each readiness rule, who makes pre-check, medical, regulatory and legal decisions, and who gives a second look on a kept claim are all settings.
The Agentic Solution Engine builds and deploys it from your requirements — your claims library, reference library, brand style and ISI rules and a sample of past pieces — and it goes live once every quality gate has passed. We will walk you through it on your own drafts first.
The quarterly risk narrative, drafted in minutes with every figure from a live query.
1-min filmField PlaybooksEvery rep opens the right playbook for their market, built 70% faster.
1-min filmClaims ReviewEvery chargeback claim read first by agents, each verdict with its rule and evidence.
1-min filmWe’ll run the Claims Substantiation Pre-Checker on a sample of your own past drafts and your claims library.