This is not legal advice, and it is not a substitute for a UK-registered signatory. It is a practical note from building PromoPack: most promotional material that stalls in medical-legal-regulatory review fails for the same handful of reasons, and almost none of them are “the designer used the wrong green.”
The ABPI Code of Practice, administered by the PMCPA, is the rulebook for promoting medicines to UK health professionals and, where relevant, the public. If you produce leave-pieces, digital banners, congress stands, or branded websites for UK use, you are in that system whether your agency’s deck says so or not.
Start from claims, not from layouts
A promotional item is a list of claims wearing a layout. Reviewers do not approve vibes. They approve (or reject) statements.
Before anyone opens Figma, write the claims down:
- What do we want to say?
- Is each statement a claim about the medicine, the disease, a service, or a competitor?
- What is the single best piece of evidence for each one?
- Is that evidence allowed in this channel, for this audience?
If you cannot complete that list, you do not have a brief. You have a hope. Agencies that start with “key visuals” and reverse-engineer claims in week three are why MLR calendars slip.
What “evidence” has to be
In UK medicines promotion, evidence is not “a paper we like.” Typical sources, in rough order of how often they survive contact with a signatory:
- The SmPC (or equivalent licensed particulars) for anything about the product
- Clinical publications that actually support the wording you used, not a related endpoint
- Guidelines, when you are describing the landscape rather than your brand
- Data on file, used sparingly and only when you can produce it on request
Common mismatches:
- Headline says “superior.” Footnote cites a non-inferiority study.
- Graph shows a subgroup. Copy talks about the ITT population.
- “Patients prefer” is hanging off a market-research slide that would not survive a complaint.
- A US claim set is pasted onto a UK item. Different licence, different code, different audience.
The test I use: if the PMCPA asked “justify this sentence,” could a named person email the exact source in ten minutes? If not, the reference pack is theatre.
The reference pack is the product
The leave-piece is what the HCP sees. The reference pack is what your company can defend. Mixing them up is how teams ship a beautiful PDF and a Zip file of unrelated papers.
A pack that is actually useful has:
- The final (or near-final) piece, page by page
- Every claim extracted, in order, in the words that appear
- A pointer from each claim to a source, with location (page, table, section)
- The source documents themselves, or a controlled way to retrieve them
- Versioning — which pack belongs to which item, on which date
If your “pack” is a folder called refs_final_v7_USE_THIS, you do not have a trail. You have archaeology.
That claim-to-evidence mapping is the artefact PromoPack is built to produce. You can do it in a spreadsheet. Software just makes it harder to “forget” a headline.
Failure modes that burn review cycles
Inherited US copy. US promotional culture is not a template for the UK. If the first draft was written for a different code, rewrite the claims. Do not localise adjectives.
Implied claims. Charts, annotations, and “before/after” layouts make claims even when the caption is timid. Review the pictures as if they were sentences.
Comparison without a fair basis. Comparative claims are where complaints cluster. If you cannot state the comparison fairly, do not imply it with a colour or a column.
Out-of-date particulars. The piece was approved against last year’s SmPC. The SmPC moved. Nobody re-ran the trail.
QR codes and landing pages. The item in the pack is clean; the destination is a global brand site with US claims. The code does not stop at the edge of the PDF.
Too many owners, no extractor. Brand writes, agency designs, medical comments in a PDF, legal comments in email, nobody owns the claim list. The list is the spine. Give it an owner.
A working sequence
- Audience and channel. HCP vs public vs internal. The code’s expectations change.
- Claim list from medical and brand, before design.
- Evidence map — one source per claim, or the claim dies.
- Design against the approved claim list, not the other way around.
- Extract again from the designed piece (layout creates new claims).
- Signatory review of piece and pack together.
- Archive the pair. Next version starts from that archive, not from memory.
None of this is glamorous. It is how you stop paying for three rounds of “can we just tweak the headline?”
Where PromoPack fits
PromoPack exists because I got tired of watching that sequence live in inboxes. It turns a piece of material into a reference pack with a visible claim-to-evidence trail. It does not replace a signatory, an agency, or reading the Code.
If you want the tool: promopack.app. If you want a second pair of eyes on how your team runs the workflow, Elliogen advisory is the human route — or just email me.