The instrument nobody built

She arrives with a folder. Printed pages, a spreadsheet of dates and symptoms she has kept for five months, a short list of questions written the night before so she will not forget them under pressure. She has rehearsed the order: sleep first, then the fatigue, then the strange fog in the second half of every month. Two minutes in, she can feel the clock in the room, and she begins to edit herself. The sleep gets one sentence. The fog gets cut entirely. She leaves with a test she has had twice before and a feeling she cannot name, somewhere between gratitude and defeat. The folder goes back in her bag, unopened.

She is a composite. I have to say that plainly, because I will not put a real woman's medical life on a public page. But I have met her many times, in many rooms, and I suspect you have too. You may have been her.

Here is what I want to tell her, and what took me most of a career to be able to say with authority: the problem in that room was not her, and it was not really the doctor either. The problem was decided years earlier, in rooms she will never see. I know, because I sat in them.

I did not come to women's health as a patient advocate. I came from the other side entirely. I spent years in pharmaceutical medicine and then in health economics, the discipline that decides, in effect, what counts. Which endpoints a trial should measure. Which outcomes justify a price. Which conditions represent a market worth entering and which do not. Health economics is where a health system's attention gets allocated, and attention, in medicine, is destiny. What gets measured gets studied. What gets studied gets funded. What gets funded generates evidence, and evidence is what a clinician is permitted to act on. The chain runs all the way from a spreadsheet in a strategy meeting to the ten minutes in that consulting room.

Now run the chain backwards for the woman with the folder. Her fog, her fatigue, her disrupted sleep are what the literature politely calls patient-reported outcomes, and for most of my career they sat at the bottom of the evidence hierarchy: soft data, subjective, difficult to standardise, easy to leave out of a trial design when the budget is tight. Symptoms that fluctuate across a menstrual cycle are harder still, because they add variance, and variance is expensive. So, for decades, the rational move at every link in the chain was to measure something else. Not out of malice. Out of method. The gaps in women's health are not failures of the machine. In the strict sense, they are its specifications. I have read the protocols. In some cases I helped cost them.

This is the part I most want to be understood, because it changes what the solution has to be. If the gaps were accidents, you could fix them with awareness campaigns and better bedside manner. But they are structural, which means the missing thing is not sympathy. The missing thing is an instrument. Her experience was never illegible because it lacked information. It was illegible because nothing was ever built to collect it, organise it, and read it on the timescales where it actually makes sense. A cycle is a month. A perimenopause is years. A ten-minute appointment samples that signal the way a single photograph samples a film, and then the system files the photograph as the whole story.

Meanwhile the information itself is not mysterious. The second-half-of-cycle fog she cut from her appointment has a plausible mechanism with a name: progesterone rises after ovulation, and its metabolite allopregnanolone acts on the same receptors as a sedative. Recovery that has quietly worsened tracks a stress system, the HPA axis, that can be watched. Ferritin can sit inside a reference range and still sit far from where she feels well. None of this is exotic. It is simply endocrinology, applied to a life, over time. What has been missing is the layer that does the applying: something that takes what a woman reports as seriously as a dataset, because that is precisely what it is.

That is what I left the other side of the table to build. Meia Lab is an interpreter. It conducts a proper intake, the unhurried conversation the clock never allows. It organises what she reports across five connected signals and watches how they move over weeks and months, because that is where the pattern lives. And when her next appointment comes, it hands her back her own evidence in a form the system is built to respect: structured, chronological, legible in two minutes. I am precise about the boundary, because I have seen what happens when products blur it. Meia Lab does not diagnose and does not treat. It does the thing the chain of evidence never did for her. It points the instrument at her experience and takes the reading.

I think often about the folder in her bag. Five months of careful observation, gathered by the person with the most complete access to the data and the strongest incentive to get it right, and the system had no port to plug it into. That is the absurdity I could not stop seeing, and in the end it was easier to build the port than to keep explaining the absence.

The feeling was always data. What was missing was someone to read it.