Pharma gave £2.6 billion to unnamed individuals over six years

Last time, I said the only way to start fixing a broken information system is to pick one clear hallway of power and go at it hard. Money flowing into the pockets of healthcare practitioners, medical researchers and education institutions is one such hallway. Over the next few weeks, I’m going to conduct deep technical investigations into how these payments are made, who makes them, and what is concealed.

But first, some background on how this all got started. Spurred on by the Physician Payments Sunshine Act (2010), the trade body representing the pharmaceutical industry in the UK wanted to show it was committed to transparency. If your doctor is being paid by the pharmaceutical industry, you should know, right? Most people agreed, and so… an intervention was needed!

Being a British intervention, it took years of debate. Feet were dragged, obligations were obfuscated, and codes were conduct were slowly created. Finally, six years behind America, the Association of the British Pharmaceutical Industry (ABPI) launched “Disclosure UK” : a public, searchable database implementing the rules which had taken years to agree on.

It’s meant to tell you which doctors and researchers take money from the pharmaceutical industry. To be fair, it does do that, but as you’ll see, there are...some caveats. Every year it announces a new transparency record about how successful the project has become:

“Today, we celebrate a new transparency record, with an estimated nine in ten (91.9 per cent) of healthcare professionals named [6]. This compares to 80.8 per cent for 2023 and 78.8 per cent for 2022.” ABPI blog on Disclosure UK

That sounds pretty good doesn’t it? You’re forgiven if you think this means 91.9% of all doctors and researchers are named in that database. It doesn’t mean that. You’re also forgiven if you think it means that 91.9% of all money paid by pharma is named in that database. It doesn’t mean that. It’s just meant to sound as though it means that.

Long time readers of The Digger will know to distrust every number a pharmaceutical adjacent company throws at you. This statement, designed to elicit a warm “transparency maxing” feeling, is incredibly misleading. It’s true within a very narrow and hidden scope, but taken at face value, it’s a classic magic trick of misdirection. The database, and the rules that make up what is recorded, are misleading by design.

You can imagine it like this: Across the money that’s tracked by this database, there's one pot tracking payments made to a person directly — fees, consultancy, event costs, travel. Of the doctors who were paid that way, an estimated 91% had their name published; the other 9% refused to be named and sit in an anonymous lump.

Fine. I guess…

But the largest channel of pharma money — research and development — isn't in that pot at all. R&D money flows to the institution, and it’s usually tied to named researchers, departments, and the studies they run. But, and it’s a big but, it is reported only as a faceless aggregate. No names are attached to it. Ever. These are the agreed rules.

“The names of these contracted individuals must be disclosed except in relation to payments in relation to research and development work, including clinical trials, as defined below, where disclosure should be on an aggregate basis.”

It’s not a huge leap of the imagination to see how these incredibly lucrative clinical trials and R&D contracts (perhaps granted for spurious reasons) might become a concerning conflict of interest. And here’s the thing: this pot of cash is the lion's share of the money in the database. It’s £2.6 billion over six years, on its own bigger than every named payment in the entire database combined.

How do I know this? Because I downloaded the whole database. Six years of it actually — including three years in which the database itself is no longer public. I had to recover those years from web-archive captures. I built it into a map where every single cash connection links back to the document that proves it. And then I added up the money that arrives with no name attached.

It comes to £2.6 billion.

That’s one category — “Research and Development” — over six years. Money the industry’s own rules say may be disclosed in a lump sum only, never against a person. For every £1 that carries a name in this system, £2.45 flows without one which means that seventy-one percent of all the money is nameless. As an example, I found that in 2024, a single company booked £114 million of research money without a single name on it. It might be fine...

One of the first threads I’m pulling at is the most obvious. Since researchers are (mostly) required to declare the money they take from industry in their own published papers, it should be possible to cross reference what was declared in their papers, with what’s declared in the Disclosure database.

The two don’t match.

Researchers declaring personal fees in their published work do not match with names in the database designed to hold exactly that. I’m not naming places or people today, every single case deserves its caveats laid out in full, but the gaps are real, and as I investigate them I’m sure I’ll find more.

I want to be precise over what I’ve built here. It isn’t a leak. Nothing here is stolen. Every number came from the industry’s own transparency spreadsheets. The whole point of the database is to allow the public to hold the industry to account, and that’s what I’m doing. It seems the transparency system, which took six years to build, has a blind spot exactly where the most important and voluminous money lies.

Over the next few weeks I’m going to investigate what’s on the other side of these payments. Who the money reaches, and how payments declared in published papers don’t seem to appear in the database that’s supposed to collect them. I’ll look at the funnel that educates doctors before they get their license, and I’ll look at the charities and independent businesses in receipt of pharma cash.

I’m doing it as a series, and I’m doing it a little differently from anything I’ve published before. Towards the end of this series, the full dataset will be made public. It’s not new data per se, but it’s data which hasn’t been collected and interrogated in this way before. Every claim will come with its source one click away. You’ll be able to check my work, and — when it’s ready — explore the map yourself.

The new shape of The Digger

A note on money, because I’d rather just say it. I’ve published in bursts and gone quiet, and that’s not fair on the people paying. And the plain truth is the Substack on its own doesn’t cover the kind of work I want to be able to do — this post sits on a database and a pile of tools that took weeks to build, and that’s only the first part.

There’s always a tension at the heart of content like this. It needs to be free or it just doesn’t reach anyone, but if it’s free, I can’t actually do the work. So here’s a model I think can help: the headlines stay free, and paid subscribers get the layer underneath — the names, the tables, the documents, detail and data — plus the lab notes as I dig. If you value this kind of work, this is how you can support it.

If you’ve been a free reader since the early days, this is the moment upgrading buys you something real: a seat in the room where the map gets made.

The moon is 238,000 miles away and we managed that. One hallway at a time.

---

Author’s note (the consulting floor — keep, it pays rent): I’ve spent years building AI tools — Pydantic AI, agent frameworks, scraping and extraction pipelines. This investigation runs on them. I’m now turning that toolkit toward deep investigations, and I’m opening a small amount of consulting time: a day or two to build you a tool or point you the right way. If that’s useful, reply to this email.

---