How advertising rules split promotional and non-promotional content in pharma
A medical science liaison hands a hospital consultant a slide showing survival data from a trial that sits outside the approved label. Legal, in some circumstances. The same slide, printed and left behind by a sales representative at the end of a detailing call, is a breach in every major market. Same data, same doctor, same afternoon. What changes is who asked, who is paid for what, and whether anyone went looking.
That is the boundary this lesson works on: disease awareness material, medical affairs communication and congress content, and the tests regulators use to decide which side of the line a piece of content has landed on. The wider regulatory setup, the market-by-market rules on speaking to patients, the risk-information requirements in patient creative and the sign-off gate itself are all handled by their own lessons in this module. Here they are assumed.
The definition that does the work
Article 86 of Directive 2001/83/EC defines advertising of medicinal products as "any form of door-to-door information, canvassing activity or inducement designed to promote the prescription, supply, sale or consumption of medicinal products". Intent, not format. A tweet, a symposium, a printed leaflet and a slide deck are all judged by the same question.
Article 86(2) then lists what falls outside: labelling and package leaflets, correspondence needed to answer a specific question about a particular medicine, factual announcements and reference material about things like pack changes or adverse reaction warnings, trade catalogues and price lists with no product claims, and information relating to human health or diseases "provided there is no reference, even indirect, to medicinal products".
That last phrase, *no reference, even indirect*, carries more weight than any other in European pharma marketing. It is what turns a link, a colour palette or a mechanism-of-action animation into a problem. The US has no equivalent single statutory definition; the FDA reasons from intended use and consistency with the approved labelling, supported by its 2018 guidance on communications consistent with FDA-required labelling and its 2011 draft guidance on responding to unsolicited requests for off-label information.
The tests regulators actually apply
- Who paid, and can the audience tell? Sponsorship is a starting presumption rather than a verdict. In Damgaard (C-421/07, 2009) the Court of Justice held that a third party writing about a medicine on his own initiative, with no connection to the manufacturer, can still fall inside the advertising rules.
- Push or pull. In MSD Sharp & Dohme v Merckle (C-316/09, 2011) the Court found that publishing an unaltered, complete reproduction of the summary of product characteristics on a website, reachable only by someone who goes looking for it, is not advertising. Email that same file to a list unprompted and it changes character.
- Proactive or reactive. An unprompted, specific question from a health professional can be answered with off-label data, privately, by medical staff, with the exchange documented. If a representative planted the question, the request stopped being unsolicited before it was asked.
- Does the content identify a product without naming it? An awareness page for a condition with exactly one licensed treatment identifies that treatment, whatever the copy says.
- What is the communicator paid for? GSK announced in December 2013 that it would stop paying doctors to speak about its medicines and stop tying sales representative bonuses to prescription volumes, completing the change in 2016. Incentive structure is evidence of intent, and regulators read it that way.
Boundary test (apply in order)
1. Is a specific product identifiable, even indirectly (name, logo,
livery, MOA, single-treatment indication)? -> branded
2. Did the recipient request it, specifically and unprompted?
-> reactive, medical
3. Was it pushed to an audience that did not ask? -> promotional
4. Who is accountable for it, and how are they paid?
Commercial line + volume-linked pay -> promotional
5. Does the audience gate match the content tier
(public / HCP / payor)? -> if no, stopDisease awareness: where unbranded stops being unbranded
Say a Sanofi team wants a public page about undiagnosed type 2 diabetes. Under the ABPI Code, material like this must not raise unfounded hopes of successful treatment, must not encourage the public to ask their doctor for a particular medicine, must tell readers to consult a health professional, and must declare which company funded it. Nothing there stops a good campaign. What kills campaigns is the drift that follows.
The common failure modes are dull and repeatable. A screening quiz ends with a button through to the branded product site, which supplies the indirect reference. Copy describes "a new once-daily option" in a category where only one once-daily product exists. The awareness site uses the brand's typeface and colour, so the visual system does the naming. Prevalence figures get stretched to make the market look bigger than the epidemiology supports.
The sharpest edge case is the narrow indication. In a rare disease with a single licensed therapy, there is no honest unbranded space at all: any awareness activity promotes the only product that treats the condition, and regulators treat it accordingly. Orphan portfolios feel this most, which is why rare disease awareness work usually runs through patient organisations with genuine editorial independence rather than through brand teams.
There is a commercial consequence teams underrate. Unbranded work builds category demand that any competitor in the class can collect, and it cannot be measured the way branded work is. The temptation is then to close the loop with data: retarget the awareness audience with branded creative, or reuse the same pixel audience across both. That is one of the quickest ways to demonstrate, in an audit trail, that the two assets were always one campaign.
Medical affairs and the reactive-only rule
Medical affairs exists because doctors need answers that the label does not contain. The function only holds up if it is structurally separate: a reporting line that does not end at commercial, compensation with no volume component, standard response letters pre-approved by medical rather than written on the day, and a log of who asked what and when.
Payor communication sits slightly outside this. In the US, the 21st Century Cures Act amended section 502(a) of the Food, Drug and Cosmetic Act in 2016 to give health care economic information addressed to payors and formulary committees more latitude than the same claim made to a prescriber, provided it relates to the approved indication and rests on competent and reliable evidence. The same budget-impact model, sent to a clinician, has no such protection.
Two collapse patterns recur. A liaison is used as a warm-up visit before the representative arrives, which makes the whole sequence commercial. And field insight reports, written to capture clinical questions, get circulated to brand teams as targeting intelligence.
Congress season: the hardest room to control
At a large congress, one company occupies four content tiers at once. The booth is promotional and confined to the local licence. The satellite symposium is company-organised and must be labelled as such, held outside the official scientific programme. Posters and abstracts belong to their authors. The medical information counter is reactive only, and is staffed accordingly.
A team like Takeda's, presenting gastroenterology or rare disease data across several markets, has to hold all four apart in the same hall, in front of an audience whose national licences differ. UK rules make a narrow allowance here: material about a medicine not licensed in the UK can be made available at a genuinely international meeting if the licensing status is stated clearly.
Social media dissolves the separation fastest. The PMCPA has ruled that an employee liking or sharing a post on LinkedIn amounts to proactively disseminating that material to their whole network, including members of the public, which pulls congress data into public promotion in one tap.
Knowledge check
1. Why do regulators subject promotional pharma content to a higher evidentiary bar than ordinary corporate communications?
2. According to the FDA's core rule enforced by OPDP, what determines whether promotional material is compliant?
3. A pharma company publishes a full prescribing label on its website with dosing tables and adverse event lists. Why is this page least likely to trigger a compliance fight compared to a marketing banner or HCP slide deck?
4. Select ALL correct answers describing characteristics that would likely classify pharma content as 'promotional' under regulatory frameworks.
Select all the correct answers.
5. Select ALL correct answers about why off-label promotion is heavily scrutinized by regulators like the FDA.
Select all the correct answers.
What misclassification costs
The PMCPA publishes every completed case with the company named and the clauses breached. A ruling under the clause on bringing discredit to the industry is the sector's public censure, and sanctions escalate from a published reprimand to an audit with pre-vetting of material, up to suspension or expulsion from the ABPI: Novo Nordisk was suspended for two years in March 2023. In the US, GSK's 2012 settlement with the Department of Justice, three billion dollars, covered promotion of Paxil for under-18s and Wellbutrin for unapproved uses among other conduct.
Warning letter volume is a poor guide to exposure. OPDP has issued single digits of letters in recent years against dozens a decade earlier, so each one carries more attention, not less. The direct bill is rarely the fine: it is withdrawing material from the field, corrective letters to health professionals, retraining, and months of pre-approval on every asset after an audit.
The structural cost is the one leaders miss. Two separate content operations, two review paths and two sets of people are expensive, and when budgets tighten the first proposal is always to have one team produce both. That is how the boundary goes, quietly, before anyone files a complaint.
🎬 [VIDEO: "FDA Regulation of Prescription Drug Advertising" - https://www.youtube.com/results?search_query=FDA+regulation+of+prescription+drug+advertising - search results overview of how OPDP evaluates promotional pharma content, useful as a primer before digging into actual warning letters]
Key Takeaways
- The EU test is intent, written into Article 86 of Directive 2001/83/EC, with health information exempt only if there is no reference, even indirect, to a medicine.
- Push versus pull and proactive versus reactive decide most borderline cases: the same document is lawful when pulled or requested and promotional when sent unprompted.
- Disease awareness has no honest unbranded space where only one licensed treatment exists, and click-throughs, brand livery or shared retargetingretargetingShowing ads to users who have previously visited your site or interacted with your brand, to bring them back and drive conversion.View full definition → audiences supply the indirect reference that removes the exemption.
- Medical affairs holds only if it is structurally separate, with no volume-linked pay and a documented log of unsolicited requests; payor communication gets more latitude than the same claim to a prescriber.
- Congresses run four content tiers in one hall, and an employee sharing a poster on LinkedIn can convert scientific data into public promotion instantly, as PMCPA rulings have found.