There is a conference somewhere right now where someone is presenting a strategy for reaching underserved communities.
The slides will be well designed. The intentions will be genuine. And the strategy will almost certainly be built around the same assumption that has quietly shaped health equity work for the better part of thirty years: that the core problem is awareness and access, and that the solution is better messaging, more targeted outreach, and stronger community engagement.
I've sat in those rooms. I've commissioned those strategies. And I've watched, with increasing frustration, as they produce reports that get filed and participation rates that don't change.
The reason is simple, and it's the thing I've spent the last two years trying to articulate clearly enough that the right people actually hear it.
The communities we are trying to reach are not disengaged because they lack awareness. They are disengaged because they have learned, through decades of accumulated experience, to evaluate authenticity.
They have been consulted before. They have filled in surveys and attended town halls and sat on patient advisory boards. And they have watched as the decisions got made exactly the same way they would have been made without any of that input.
They are not disengaged. They are rational.
The wrong diagnosis
The health equity field has, for the most part, treated this as a communications and engagement problem. The logic goes: underserved communities don't participate in research, don't access healthcare services, don't engage with clinical trials, because they don't know about them, don't trust the organisations running them, or face logistical barriers to participation.
All of that is true. But it is downstream of a more fundamental problem.
The real issue is that the data infrastructure that drives decisions across the entire medicine lifecycle, from drug discovery through clinical development, market access, and real-world adoption, has never been built to include the lived experience of underserved communities as a structured, usable input.
Clinical data. Operational data. Social determinants of health data. These sit alongside each other in the systems where decisions get made.
Lived experience data does not. It gets collected, in interviews, in focus groups, in patient advisory boards, and then it gets qualitatively summarised, stripped of its structure, and filed somewhere it cannot influence a protocol, a site selection decision, a market access submission, or a reimbursement negotiation.
The problem isn't that we aren't listening. The problem is that there is no infrastructure to turn what we hear into something that changes decisions.
That is a data problem. And it requires a data solution.
What this costs
I want to be precise about what I mean when I say this is expensive, because I think the health equity field has sometimes made the mistake of framing the cost in purely moral terms when the commercial case is just as compelling.
The average cost of R&D per approved drug product, including failed trials, is estimated at $1.1 billion. A meaningful proportion of that cost comes from late-stage failures, and a meaningful proportion of late-stage failures are downstream consequences of protocol design decisions that didn't account for how diverse populations would actually experience the trial.
Visit schedules that don't work for people who can't take time off. Eligibility criteria that disproportionately exclude minority ethnic or working-class populations. Outcome measures developed fifty years ago on homogeneous populations that don't transfer to the patients now being prescribed the treatment.
The clopidogrel case is the most cited example in Europe, and it's worth sitting with. The UK's most widely prescribed anti-platelet drug may be ineffective for up to 57% of British Bangladeshi and Pakistani patients, because those patients were not represented in the approval trials, and because the pharmacogenomic differences that affect drug metabolism were not captured in the evidence base.
This is a drug being prescribed today, millions of times a year, to patients whose likelihood of response was never properly studied.
That is what a data gap looks like in practice.
Every month of clinical trial delay costs between $600,000 and $8 million. Most delays are foreseeable. Health equity risk is foreseeable. The question is whether you look for it before the protocol is signed, or after recruitment fails.
The regulatory environment is also shifting in ways that make this a commercial imperative rather than just an ethical one. The MHRA's mandatory Inclusion and Diversity Plans come into force in 2026. The FDA's Diversity Action Plans are binding for Phase III. HTA bodies are asking harder questions about whether trial populations reflect the patients who will actually use the medicine.
The window between aspiration and obligation is closing.
Why I know this
I want to tell you something about why this is personal, because I think the best work always is.
My dad came to the UK in the early seventies. My mum came in the late seventies. They had an arranged marriage — first time she saw him was the day they got married, a fact I have never quite been able to get my head around. They didn't have much when they started out. What they had was grit.
My mum worked in factories. My dad became a TV engineer. They worked 12-hour shifts, skipped holidays, saved everything. So I could have a shot at something different.
My dad was diagnosed with prostate cancer at 52. My mum with thyroid cancer at 48. Both too young. Both, in different ways, let down by a system that wasn't designed around how people like them interact with healthcare.
My dad knew something was wrong. He pushed for answers for over a year. They didn't listen. By the time they did, it was terminal. I was 17, sitting in a hospital room, when a consultant came in, said 'six months to live,' handed us a flyer, and left. No pathway. No options. No care. We walked out dazed and silent.
We didn't have advocates. We didn't know what questions to ask. We didn't even know what we didn't know.
My dad held on for six more years. He knew he wouldn't see me marry, so he wanted at least to see me graduate. He died six months before I did.
I've spent my career inside healthcare agencies — research and development, brand launches, market access, medical affairs, commercial strategy. I've seen what works. I've seen what doesn't. And I've seen, repeatedly, who gets left behind.
I didn't start Unwritten Health because I read a policy paper. I started it because I have lived the consequences of being unseen, unheard, and uncared for. And because after twenty years of trying to fix this from inside the system, I decided the system needed different infrastructure.
What different infrastructure looks like
Unwritten Health is a decision-grade data platform built on lived experience. That phrase, decision-grade, matters. It's the distinction between data that influences decisions and data that decorates them.
We work inside community spaces. Not presenting slides. Sitting in rooms, explaining governance, taking scrutiny, coming back. We capture lived experience with individual consent and community oversight. We transcribe it, code it thematically, aggregate it longitudinally. We turn it into structured data that can sit alongside clinical and operational data in the systems where decisions actually get made.
The dataset strengthens with every participation cycle. Insights compound over time. And because the data is gathered directly from individuals who consent to share it, it operates within GDPR rather than around it, which in continental Europe where institutional top-down approaches hit a wall, is a structural advantage that cannot be quickly or cheaply replicated.
We monetise insight, not individuals.
And we are built on the understanding that communities aren't hard to reach. They're hard to retain when every previous interaction taught them their input wouldn't matter. The job isn't to get communities to participate. It's to build infrastructure that makes participation worth their time.
The man in his seventies who stood up at a community town hall and said to me: 'People have been coming here since the 1960s promising change. Nothing has changed. Why should we trust you?' he wasn't rejecting engagement. He was asking whether this time, power would move.
That is the only question that matters. And it is the question that Unwritten Health is built to answer.
What comes next
Over the coming weeks in Unwritten Dispatches, I'll be going deeper on the evidence base, the regulatory shifts, the market gaps, the specific points in the development lifecycle where equity risk enters and what it costs when it's missed. I'll be writing from inside the community spaces where the platform is being built. And I'll be sharing the thinking behind a company that is trying to do something genuinely different.
If you're working in clinical development, market access, patient strategy, or health equity, this is written for you. Not to tell you things you don't know, but to give you the evidence and the framing to make the case internally for the things you already believe.
And if you're an investor who sees the infrastructure gap we're closing — I'd like to hear from you.
Until next week.
Ashish.
This week in data
57% of British Bangladeshi and Pakistani patients are intermediate or poor CYP2C19 metabolisers, potentially rendering clopidogrel, the UK's most prescribed anti-platelet drug, ineffective. This population was not represented in the approval trials.
26% of European vaccine trials between 2010 and 2020 even reported the ethnicity of their participants. In the UK the figure was 17%. In Germany, 19%. You cannot fix what you do not measure.