When the industry backs off, we lean forward

Three days at Patients as Partners. One question nobody answered.

Patients as Partners is one of the best conferences in this space. I mean that. Brilliant people, genuine intent, substantive conversations. I left with more energy than I arrived with, which is not something I can say about most industry events.

But something sat with me on the flight home that I want to try to articulate properly here, because I think it matters more than any single session I attended.

Across three days, I heard a lot of introspection. The industry knows what the problem is. It knows that patient engagement fails when it is bolted on rather than built in. It knows that sustaining community trust requires operational and financial commitment, not just good intentions. It knows that health literate communication is about designing for real people rather than producing more materials. It knows that data without humanity is just numbers.

The problem is correctly diagnosed. Sadly, the solution put forward keeps missing the mark.

And the reason, I think, is that the industry is trying to solve an infrastructure problem with a communications solution. It keeps reaching for better messaging when what it needs is different architecture.

Every conversation I had in Boston came back to some version of the same question: how do you reach the communities that clinical research has consistently missed?

The answers varied. More community outreach. Better health literacy. Stronger patient advocacy partnerships. AI-assisted engagement at scale.

All of those things have merit. But they share an assumption I want to push back on, gently but directly: that the primary challenge is reaching communities. Getting the message out. Bringing people in.

It isn't.

Ray Oldenburg, a sociologist, described what he called the third place. Home is the first. Work is the second. The third place is where community actually lives. The barber shop. The mosque. The community centre. The corner cafe. Spaces people trust because they belong to them, not to any institution.

His argument was that third places are essential to democracy and civic life. I would extend it: they are essential to health equity research. Because the intelligence that matters most, the hesitation, the trust, the lived reality behind the words, does not live in a database or an EHR. It lives in those spaces. In conversations that happen when no researcher is in the room.

The problem is not that communities are hard to reach. The problem is that the industry keeps trying to reach them on its own terms, in its own spaces, through its own channels. And then wondering why the response is lukewarm.

You cannot communicate your way into trust with communities that have decades of evidence you won't follow through. The answer is not better messaging. It is different infrastructure. Go to where trust already exists and build from there.

I need to be honest about something I observed in Boston, because I think it is the most important thing I am bringing back from this trip and it did not happen in any session.

On the panels, the language was careful and constructive. Representation matters. Diverse trial populations produce better science. Patient engagement is a strategic priority. All true. All said with conviction.

Then I walked off the stage and had conversations on the floor.

Diversity Action Plans scrapped. Not paused. Scrapped. Because, I was told, it is no longer a priority. DEI job titles quietly changed or eliminated. Company positioning shifted to avoid language that has become politically uncomfortable in the US following the executive orders rolling back federal DEI commitments. The same organisations that were nodding along in the sessions were, in the corridors, telling me they had been instructed to stand down.

I expressed disappointment on my panel. Genuine disappointment. Because I have sat in community spaces with people who had spent decades deciding whether or not to trust the healthcare system, and who were starting, slowly and conditionally, to open that door. And what is being communicated to them now, not in words but in decisions, is that the industry's commitment to their inclusion was contingent on it being politically convenient.

That is not a small thing to walk back from.

I understand the legal and commercial pressures that US organisations are navigating right now. I am not naive about them. But there is a version of this retreat that will take years to recover from in the communities that matter most. Trust is not a dial you turn up and down based on the regulatory climate. It is built slowly and destroyed quickly. And right now, some very large organisations are destroying it quickly.

Here is my position, and I want to say it plainly: when the industry backs off, I lean forward.

Not because it is commercially convenient. Not because the regulatory pressure in the UK and Europe gives me cover to do so. But because the communities this work is for have not changed. The disease burden has not changed. The data gap has not changed. The only thing that has changed is the appetite of some organisations to be seen caring about it.

Unwritten Health is not going anywhere. The work continues. And I would rather be building this infrastructure now, when others are retreating, than scrambling to rebuild it when the political tide turns again and everyone suddenly rediscovers their commitment to representation.

One of the sharpest moments in Boston came from a discussion about AI and patient engagement. The observation on the table was that it sounds oxymoronic at first. Technology and human connection feel like opposites.

I sat in that session thinking: the oxymoron dissolves the moment you ask what the technology is actually for.

There is a version of AI in patient engagement that is genuinely dangerous. It optimises for efficiency, strips out nuance, and produces insights that are technically rigorous but built on assumptions about what matters rather than what actually does. Most insight tools in this space were built by very smart people writing surveys that were never co-created with the communities they are trying to understand. The output looks like data. It is not. It is the researcher's assumptions, processed at scale.

But there is another version. One that helps you reach more people, understand more stories, and act on that understanding faster, without sacrificing the depth that makes the understanding real.

What the session could not answer, and what I think represents the genuine frontier in this space, is how you capture what does not get written down. The hesitation before an answer. The way someone's energy changes when a question touches something real. The trust, or its absence, in how someone holds their own experience.

That is what we are building the AI component of our platform to do. Not surveys at scale. Not text analysis on pre-written questions. Non-verbal signals from video, mapped to validated constructs around engagement and trust. Built with explicit consent. Built with bias monitoring across age, gender, skin tone, and body type. Because if this is for underserved communities, it has to work for everyone.

The insight that changes a protocol is not the average. It is the signal underneath the average that no standard tool was designed to catch.

My most important moment in Boston did not happen on the main stage.

It happened when I met Misha Mehta, a scientist, a patient advocate, and a mother. Her son Neev passed away from a brain tumour. Everything she does is for him.

We were talking about underserved communities in research. And she said something I have not been able to put down since.

Paediatrics. The most underserved of all.

We talk about representation gaps as if we have the full picture. As if the communities we know are missing are the only ones missing. Misha's work is a reminder of how incomplete that picture is. Children are not small adults. Their biology is different. Their experience is different. Their needs are different. And yet they are nearly invisible in the evidence base that is supposed to serve them.

The underrepresentation problem in clinical research is not one problem. It is many, layered on top of each other. And every time I think I understand the full shape of it, a conversation like the one with Misha reminds me I am still only seeing part of it.

That is not a comfortable thing to sit with. But I think it is necessary. The moment we believe we have adequately mapped who is missing, we stop looking. And the people still invisible in our data pay the price.

Boston also gave me the clearest external read I have had on where Unwritten Health sits relative to the field. So I want to use the end of this quarter to say plainly where we are.

Three thousand people on the platform. Not users in the passive sense. People who chose to share their lived experience, with consent, with transparency about how it will be used, and with the expectation that it will change something. That expectation is one we take seriously. It is also what makes the data worth anything.

Pilots running with one major pharma company, one global CRO, two NHS trusts, and one site network. These are not proof of concepts. They are working relationships generating real insight. Q2 is about converting them to paid contracts. That is the commercial priority and I am not going to dress it up as anything softer.

We are building the AI component of the platform in partnership with a university. We are on the investment journey, pitching, taking feedback, improving. The conversations are live.

Q1 was a building quarter. The platform is real. The pilots are real. The community is real.

Q2 is where we find out whether the market is ready to pay for what we have built.

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's enough.

Ashish.