A quick note before anything…
Professor Jason Arday died on Friday. He was 41.
There were serious allegations made about him. None of that was settled, it won’t be settled now in the way it should have been, and I’m not going to act like I can guess what a proper process would have decided.
What I can speak to is the shape of the last month. Watching someone be pulled to pieces in public, day after day, across every platform, until there was no version of him left that wasn’t “the story”. His family say he’d been living with sustained abuse for three years.
Scrutiny has a place. Academic work should be open to checks, and the people who did the checking had every right to ask what they asked. Still, the scale and pressure of what happened to him didn’t match anything I recognise as a process, and it’s hard not to notice who it happened to, someone held up for years as an exception, the youngest Black professor at Cambridge, autistic, the neat biography everyone wanted to repeat. Being turned into a symbol doesn’t keep you safe. It puts you under a harsher light. The more someone is raised up as evidence the system works, the harder the drop tends to be, and the fewer people there are between them and the impact.
I’ve spent this week writing about what institutions take from people, and what they never return. This feels close to that, though I haven’t worked out how to place it yet, and I’m not going to hammer it into a lesson while his family are still in the first week of living with it.
My condolences to them.
It feels strange, almost off, to be writing about anything else right now…
For context, I’m a British South Asian man. For the first twenty years of my life, people kept telling me I should be totally fine in heat like this. I wasn’t then, and I’m definitely not now. These last couple of weeks have turned into a daily back-and-forth with every fan in the house, a careful rationing of cold water like it’s going to run out, and me squinting at my laptop through this constant, heat-haze blur that I’m choosing, generously, to call “focus.”
Work still gets done. Just, slower, and with a lot more sweat involved.
Two things have been looping in my head all week, and I want to give them the time they deserve on the page. One is a small, easy-to-miss update tucked away on the European Commission’s HTA pages that seemed to pass most people by. The other is a conversation I had with Connie Lee Montgomery, and I haven’t been able to shake it since.
Building this is hard. Here is where we actually are.
I want to level with you this week, in a way I don’t think founders always give themselves permission to.
Unwritten Health is the most meaningful thing I’ve ever built. It’s also, right now, the toughest thing I’ve ever tried to carry. And if you’ve been reading this newsletter from the start, you deserve the real picture, not the cleaned-up, glossy one.
Here’s where we are: we’re actively working to close three clients. These aren’t vague leads, they’re live conversations with real organisations, and the potential is there. The pipeline feels more solid than it ever has. The message is finally clicking. The product is ready. Still, there’s that painfully familiar distance between “this is going somewhere” and “we’ve signed,” and anyone who’s built a company knows that distance doesn’t cover payroll.
So I’ve made a straightforward call. While we bring these deals over the line, I’m taking on some consulting work alongside building Unwritten Health. I’ve spent two decades advising health agencies, pharma, and CROs. That experience is worth something, and right now it makes sense to use it to bring in income while we get Unwritten Health to the point where those contracts actually land, for the business, and for the people who are counting on it.
I’m not ashamed of that. Pretending this tension isn’t part of the story feels like a lie, and lying isn’t why I write to you here.
Honestly, the roughest stretch of building something that’s truly different is that in-between period, when you’re sure you’re right, but you’re not getting paid for it yet. That’s exactly where we are. And we’ll move through it.
If you’re in clinical development, patient strategy, market access, or medical affairs, and you think it’s worth talking, about Equity Engine, about a consulting engagement, about anything in this space, reply to this email. I read every message.
What the EU's new eligibility checker actually tells you — and the question it cannot answer
A few weeks ago, the European Commission published three things on its Joint Clinical Assessment pages.
The first was an eligibility checker: answer a short set of questions, find out whether your product falls within scope for a JCA. The second was a downloadable tracker of ongoing joint clinical assessments, a live view of what is already in the machine, which comparators are being scoped, which 100-day windows are running. The third was a set of general principles on the use of AI in preparing the JCA dossier. And registration opened for an 18 September webinar for patients and clinical experts, the constituency whose input the process depends on and whose availability is, in practice, the bottleneck nobody schedules for.
The checker got the attention. It is the least consequential of the four.
Publishing an eligibility checker is not primarily an act of service. It is the Commission removing a category of excuse. Until now, we were not sure whether we were in scope was a survivable answer inside a sponsor organisation. It bought quarters. It let teams defer the comparator conversation until the clinical data matured. That answer is now gone. Anyone can settle the question in an afternoon, which means anyone in your organisation can ask why it has not been settled. Scope has moved from an open question to a matter of record. And once something is a matter of record, the accountability moves with it.
Here is the mechanic that actually matters. Once a product enters JCA, the scoping process gathers the populations, comparators, and outcomes that member states want assessed. Those requirements are not negotiated down to a convenient common denominator, the process accommodates genuine national divergence, which means the PICO set can be considerably wider than any single national submission would have demanded. Then the dossier window runs.
In 100 days, a competent team can write anything.
In 100 days, no team can generate anything.
That asymmetry is the whole of the risk. And it falls almost entirely on four categories of evidence that cannot be retrofitted inside the window.
Subgroup outcomes in populations your trial under-recruited. If your pivotal trial enrolled a socioeconomically or ethnically narrow sample — and most do — no analysis plan rescues you. You can report the gap. You cannot fill it.
Patient-reported outcomes validated in the populations being asked about. A PRO instrument validated in one population and administered in another produces numbers, not evidence. Validation is a study, and studies take longer than 100 days.
Patient-relevant endpoints that were never endpoints. If a member state asks about a burden your protocol treated as a secondary measure or did not measure at all, the answer is silence. The dossier can explain why. It cannot substitute.
Deliberative patient input with any standing. A convened, documented, methodologically defensible patient view — the kind that survives a challenge about who was in the room and how they were selected — takes months to constitute and longer to be worth citing. Assembled inside a submission window, it reads exactly like what it is.
Used well, the eligibility checker is not compliance triage. It is an evidence-debt audit. The run itself is trivial: put every asset in the relevant window through it and record the answer. The value is in the second question you ask about each yes.
If a national HTA body asks how this performs in the populations our trial recruited least well, do we have an answer, or do we have a paragraph?
Score every asset honestly on that question and you will get a distribution. A few assets will have real data. A larger number will have a paragraph. Paragraphs survive internal review. They do not survive a PICO set from a member state that has spent a decade building health inequality into its assessment framework.
The gap between the assets with data and the assets with a paragraph was created three to five years ago, by protocol decisions made when nobody in the room was thinking about a joint assessment. You cannot fix the assets already in the window. You can stop manufacturing the problem for the assets behind them.
The checker takes five minutes. The honest question after it takes considerably longer. And the evidence it reveals you are missing takes four to six quarters to build.
Sources: European Commission JCA implementation pages, August 2026. EMA draft reflection paper on patient experience data, 2026. Health Research Authority, 2026.
What Connie said. And what I cannot stop thinking about.
The technology failed on Wednesday. The stream dropped, the backup plan required a backup plan, and for a moment it looked like the conversation with Connie Lee Montgomery was not going to happen.
We moved to Zoom and kept going. I am glad we did.
I have been thinking about a question for a while. Whether engaging communities slides, over time, into extracting from them. Whether the difference is visible to the people doing the engaging, and whether it matters to them if it is not. I put the question to Connie, and she did not hesitate.
What she said was this:
It is reminiscent of what happened with slavery. Getting our skillsets, learning what it is we know, using the tools and the knowledge and the information that we have for your economic enhancement and to go forth, but not providing any value at all to us.
Connie is Gullah Geechee. Descended from people taken into the southeastern United States by force, singled out specifically because of what they already knew. Rice cultivation. How the land behaves, how to coax it into feeding people. Methods plantation owners did not have and could not recreate on their own. Their knowledge was not an add-on. It was the point. The value was not accidental. It was the motive. Take the expertise, convert it into profit, then close the door.
She is not reaching for a dramatic comparison to shock people when she says this. She is naming a shape she recognises. A pattern that has a historical form and a contemporary expression. And once she names it, you cannot unsee it.
When a research team steps into a community, gathers stories and insight and experience, turns it into a slide deck or a product pitch or a grant report, then evaporates, that is the pattern. When a clinical development team conducts a patient advisory board, files the output, and proceeds to make exactly the decisions they would have made without it — that is the pattern. When an engagement exercise ends and the community it was conducted with hears nothing further, receives no results, sees no change, that is the pattern.
We use the phrase hard to reach as if the problem lives on one side of the distance. As if communities are the ones holding the map upside down, failing to find their way toward the institutions that want to serve them. Connie's framing inverts this entirely. The distance is not stubbornness. It is not apathy. It is memory. A rational, accumulated memory of what happens when you hand over what you know to an institution that has a track record of showing up to take and then leaving with silence.
And here is what surprised me most about the conversation. Her answer was not do not engage. It was change the terms.
Be present before you need something. That is the first condition. Not showing up when you have a study to fill, a grant to justify, or a slide deck to populate. Showing up when there is nothing immediately at stake, and staying when there is. Connie called this being around. It sounds simple. It is the hardest thing for organisations to do, because it does not map onto a project timeline or a budget line.
Compensate people in a way that actually matches what you are asking of them. Not tokenistic payments. Not a gift voucher that signals exactly what you think someone's time is worth. Real compensation, proportionate to the knowledge being accessed.
Bring the results back. Close the loop. Let people see that their time mattered, that their voice did not disappear into an internal folder that nobody will ever open again. This is the part the industry gets most consistently wrong. Not because it is expensive or technically difficult. Because it requires treating the relationship as ongoing rather than transactional. And transactional is the default.
Co-creation, she said. Not as a checkbox.
I am keeping a lot of this in her words on purpose. It is her insight and it deserves to keep her cadence.
But I want to add something from my own position, because I think there is a harder version of this argument that the clinical research industry specifically needs to hear.
The extraction pattern is not primarily a problem of bad intent. Most of the teams who conduct extractive engagement genuinely believe they are doing meaningful work. The problem is structural. The incentives of the project, the timeline, the budget, the deliverable — all of them reward the taking and none of them reward the returning. There is no metric for whether the community felt the relationship was reciprocal. There is no KPI for whether the findings were shared back in a form people could use. There is no line in the grant report about whether trust was built or depleted.
Until those things get measured, they will not get resourced. And until they get resourced, co-creation will remain a commitment that sounds genuine in the planning phase and dissolves under pressure in the delivery phase.
The thin line between taking and building is not a moral line. It is a design line. And it can be moved.
If you work in research or patient engagement, I want to ask you Connie's question back. When have you felt the difference in your own work? The moment when something shifted from extraction to something genuinely reciprocal? I am genuinely asking. Reply to this email. I want to know what it looks like in practice.
Thanks for reading. This newsletter exists because I believe the right framing, in the right hands, changes decisions. If it did that for you this week, even a little, that is enough.
Ashish
