Navigating regulated promotional claims

# Navigating regulated promotional claims

In October 2021 the FDA told Owlet that its Smart Sock was a medical device. The hardware had been shipping for years. What triggered the letter was the sentences around it: the sock measured blood oxygen and pulse rate and alerted parents when the numbers drifted outside a set range. Measuring a physiological parameter in order to warn someone about a baby's health is a medical claim, and that claim needed a clearance the product did not have. Owlet stopped US sales.

What came back was close to the same sock with a different vocabulary. The Dream Sock sold sleep: sleep state, wakings, sleep quality. No oxygen alarms, no promise to catch anything. Owlet then did the slow work and returned with FDA clearances (a prescription pulse oximetry monitor, then health notifications on the consumer sock), earning the right to say what it had once said for free.

One piece of hardware, three claim regimes in three years. That is the craft: knowing what you may say today, what you would have to prove to say more, and how to stage the gap without freezing the launch.

The verb decides your status, not the product

Marketers assume the category determines the rules. In practice the verb does.

Wellness verbs: tracks, logs, records, shows trends, helps you understand, supports.

Medical verbs: detects, diagnoses, screens, monitors for, alerts, measures (a physiological parameter for a health purpose), treats, reduces the risk of.

The FDA's 2016 general wellness guidance leaves two safe zones. The first covers claims about maintaining a general state of health: sleep, fitness, relaxation, weight management. The second allows a link to a chronic condition only in the soft form, that healthy behaviour may help reduce the risk, and only where the connection is well established in public health literature. Name a specific disease, a specific patient population, or promise an intervention, and you have described a device, whatever the marketing deck calls it.

Apple works this line on purpose. The ECG app and irregular rhythm notifications went through FDA de novo authorization in 2018, and the consumer copy carries the limits out loud: not for users already diagnosed with atrial fibrillation, not for people under 22, not a replacement for clinical judgment. The blood oxygen app that shipped with Series 6 in 2020 stayed on the other side of the line, sold for fitness and wellness with an explicit statement that it is not intended for medical use. Same wrist, same sensor array, two vocabularies, because there were two different evidence packages behind them.

The failure mode is asymmetric. Overclaiming costs you the product. Underclaiming costs you conversion, and teams rarely measure that loss, so most wellness copy drifts vaguer than it needs to be. If you have a cleared indication, use its full width. Half the value of a clearance is the sentences it unlocks.

On-label, off-label, and the intended-use trap

On-label means promoting a product only for the uses, doses and populations in the approved labeling or the cleared indications-for-use statement. Off-label promotion means marketing an unapproved use. A physician may prescribe off-label; the company may not promote it. That asymmetry catches marketers who assume "doctors already do this" is permission.

The trap is that your own marketing is evidence of intended use. Regulators read the whole record, not just the ad you submitted for review. Common leaks:

• Search keywords bought against an indication you were never cleared for.
• An app store description or Amazon bullet written by a channel team that never saw the labeling.
• A patient testimonial where the patient says the thing you cannot say, in your video, on your site.
• A congress booth showing an investigational configuration without the required investigational labeling and a plain "not available for sale in the US" sign.

Recent promotional letters cite websites, sponsored posts and video far more often than television. The cheapest channels are the ones with the weakest review.

Anatomy of a warning letter

FDA warning letters are public and are one of the better free training resources in this field. You can browse them on the FDA Warning Letters database. OPDP publishes only a handful of promotional letters in a typical year, which is precisely why each one gets read across the industry.

Most cite one or more of these:

Omitting or minimizing risk. Efficacy in headline type, side effects in six-point grey. Regulators call the failure a lack of fair balance: benefit and risk need comparable prominence, not merely coexistence on the page.

Overstating efficacy. Saying "cures" for a product that manages, or implying superiority without head-to-head data.

Broadening the indication. Cleared for severe cases, advertised in a way that suggests mild ones. This is the most common accident, because it happens through imagery and testimonials rather than words.

Off-label claims. The highest-risk category.

23andMe shows how the whole arc plays out. In November 2013 the FDA ordered it to stop marketing the health reports of its Personal Genome Service, worrying about what a consumer might do with an unvalidated BRCA or drug-response result. The company sold ancestry only for roughly two years, then came back through the regulatory door: carrier status reports authorized in 2015, ten genetic health risk reports in 2017, three BRCA variants in 2018. That last authorization arrived welded to a limitation: the test covers three variants most common in people of Ashkenazi Jewish descent and does not detect the majority of BRCA variants. The caveat is now part of the claim, not the footnote, and the marketing has to carry it in the same breath as the benefit.

🎬 [VIDEO: "FDA Regulation of Prescription Drug Promotion" - youtube.com - an accessible overview of what OPDP reviews and how promotional rules work]

Building a claim that survives

Start from the cleared indication, not the brief

Before the creative concept, pull the indications-for-use statement and the evidence behind each sentence you want to write (the substantiation file the evidence lesson covers). If the trial measured symptom reduction at 12 weeks, you cannot imply a lifelong benefit through a photograph of someone at 70.

Separate the audiences

Copy for clinicians can be technical and comparative. Consumer-facing promotion faces stricter readability and risk-disclosure expectations, which is why the risk statement in a US television ad sounds the way it does. The same underlying evidence produces two different sentences, and reusing HCP language in a consumer asset is a routine cause of rework.

Build the claim in parts

A defensible claim usually pairs a benefit statement (on-label, substantiated), the citation, and the balancing risk or limitation, prominent rather than buried. Illustrative structure, not a real product: "In a 12-week study, Device X reduced [cleared endpoint] compared with standard care.¹ Not intended for patients under 18. See instructions for use." Notice what is absent: no "best in class", no unproven comparison, no hint of a population outside the indication.

Keep a claims library

Mature teams maintain a set of pre-approved claims with linked references and expiry dates, because references age out and a claim built on a superseded publication is an unsupported claim. Marketers then assemble campaigns from approved blocks instead of rewriting sentences and restarting the argument, which is the single biggest lever on review cycle time (the review workflow itself is the pre-launch compliance lesson's subject).

Sales messaging that stays in bounds

Warning letters often trace back to the field rather than the agency. A rep goes off-script in a clinic and there is no reviewer in the room.

Train reps to the cleared indication until they recognise drift mid-sentence. When a physician asks about an unapproved use, the safe move is to route the question to Medical Affairs, a separate scientific function that answers unsolicited requests through its own channels. Sales does not answer off-label questions. That firewall between commercial and medical keeps scientific exchange from turning into disguised promotion.

Reps present only approved visual aids and leave-behinds. No self-made slides, no annotated printouts, no "here is an interesting paper" that happens to describe an unapproved use, because handing over the reprint is itself a promotional act. The same discipline applies to speaker programs and advisory boards, which additionally carry the reportable transfers of value covered in the fair-treatment lesson and searchable by anyone in the Open Payments database.

Where the constraint pays

The rules push you to sell on evidence, in front of buyers who discount everything else. Apple's atrial fibrillation copy is more persuasive than any superlative would be, because the restrictions in it signal that someone checked. Owlet's second act sells the same reassurance the first one did, with the difference that it can now say so in the words parents actually want to hear.

Key Takeaways

• The verb sets the regime. Track, log and show trends stay in wellness; detect, alert, screen and monitor for describe a device and need clearance behind them.
• Use the full width of what you have cleared. Underclaiming is an unmeasured loss, and half the value of an authorization is the sentences it lets you write.
• Your marketing is evidence of intended use. Keywords, app store copy, testimonials and booth material are read as claims, and cheap channels get the weakest review.
• Limitations travel with the claim. When an authorization comes with a restriction, as 23andMe's BRCA reports did, the restriction belongs in the same breath as the benefit, not in the footnote.
• The field is your largest exposure. Train to the indication, route off-label questions to Medical Affairs, and let reps carry approved materials only.