Skip to main content
OriginalsNila Originals

Framework note · Nila Original

Human-centred evidence: start with the person, not the snapshot

Menopause can't be understood in a snapshot. You can only understand it over time, through the person living it. What the research actually says about single readings, trajectories and self-report, and the standard Nila is built on.

Erin Beattie, founder of Hello NilaOctober 22, 20268 min

Most of the answers I was offered in menopause came from a single moment.

A blood result from one morning, in a body whose ovaries were already gone, read as if one number could tell the whole story. A sleep score from one night, from a gadget that was never tested on a body like mine. A short appointment, where I tried to squeeze a year of symptoms into the time it took to sit down. A pink box that assumed I was 51, cisgender and easing in slowly, when I'd actually arrived overnight, after cancer, by surgery.

Each of those answers was a snapshot. Some were accurate, as far as they went. None of them held the whole picture, and a few were used as if they did. That second part is where the harm lives.

I'm giving the alternative a name in this piece, because I think midlife care needs one: human-centred evidence. It starts with the person living in the body. Not as a source of stories that a "real" measurement will later confirm or correct, but as the centre that everything else has to answer to.

The guidelines already work this way

This isn't a fringe idea. It's how menopause is supposed to be recognized.

NICE's guideline on menopause (NG23, updated November 2024) says that for otherwise healthy people aged 45 or over, perimenopause and menopause are recognized without lab tests. Perimenopause is spotted from hot flashes or night sweats plus irregular periods. Menopause is spotted from symptoms in people without a uterus. That last part is for people whose uterus is gone but whose ovaries remain, so there are no periods to count. When the ovaries are removed too, there's nothing to work out: menopause starts with the surgery itself.

NICE also explains why a hormone blood test doesn't help here. FSH, the hormone most tests look at, "fluctuates considerably over short periods of time during the years leading up to menopause," so a blood level is "not a helpful addition to what is a clinical diagnosis." In plain terms: a doctor or specialist listening to your symptoms tells them more than one blood number does.

Doctors in Norway say the same thing in their Choosing Wisely advice. The two hormones the brain sends to the ovaries, FSH and LH, swing up and down through the cycle and "provide only a snapshot." For people over 45 who still have periods, symptoms are the most reliable guide.

So the system already agrees, on paper, that what you notice over time matters more than a single reading. What it hasn't built is a place for that account to live, in a form that survives the trip into the appointment.

Change over time is how this stage actually works

The best long-term research we have says menopause doesn't move in a straight line, and it doesn't move the same way for any two people.

A large US study called SWAN followed women of many backgrounds through menopause for years. It found that frequent hot flashes and night sweats lasted a middle value of 7.4 years, longer for some groups than others (Avis et al., 2015). People also fell into four different patterns:

  • symptoms that started early and eased after the last period
  • symptoms that peaked around the last period
  • few or no symptoms at all
  • symptoms that started well before the last period and carried on for years after, a group the researchers nicknamed the "super flashers"

Hormone levels showed the same thing: several different paths, not one shared slope.

A word on why this section says "women." That's who SWAN set out to study, and it's who the research we have mostly counts. Like most menopause research, it wasn't designed to include trans men, nonbinary and other gender-expansive people who go through menopause, or people whose menopause came from surgery or cancer treatment. I'm using the researchers' word so I don't claim more than they found. The gap is real, and it's part of why your own record matters: when the studies didn't count you, the pattern you keep may be the best evidence there is about your body.

Researchers have a map for this called STRAW+10. It describes ten stages, not the three boxes (peri, meno, post) most medical records offer (Harlow et al., 2012). One blood sample can't tell you where you sit on a ten-stage journey that takes years and looks different in every body. A record you keep over time at least has a chance of showing which pattern is yours.

That's the first half of human-centred evidence: it's kept over time, because the thing it's tracking changes over time.

Your own record counts as evidence, and it changes the appointment

A lot of health tech treats what you say about yourself as the weak option, something to collect until a sensor or a lab can replace it. US regulators don't see it that way. The FDA's guidance on patient-reported outcomes describes these tools as making possible "the systematic collection of how patients feel, function and survive as valid scientific evidence." It names symptom diaries as the way to capture things like pain and anxiety that nobody else can see from the outside.

The key word is systematic. A feeling noted once is a story. The same feeling, noted the same way for weeks, next to your sleep, pain, medication and bleeding, is a pattern. And a pattern is evidence.

There's newer research on what happens when you bring that pattern to an appointment. Sillence and colleagues interviewed 29 women in the UK (again, that's who the study included) about using menopause apps (DIGITAL HEALTH, 2025). They used their notes to understand what was happening to them, then took them to appointments. In the researchers' words, "'armed' with evidence, the apps were used as a tool to combat epistemic injustice often experienced by women in healthcare settings." Put simply, the apps helped them push back against not being believed. They felt more sure of their own experience, better listened to, taken more seriously and more able to make treatment decisions together with their doctor or specialist.

The same research is honest about the limits, and I want to be too, because they tell us how to build things. Not everyone kept using an app, so we don't yet know the long-term effect. And doctors don't always know what to do with an app report. A pile of screenshots isn't something anyone can use. A tidy one-page summary is different: built on a recognized questionnaire (for us, the Menopause Rating Scale, Heinemann et al., 2004), with dates and symptom areas someone can read in under a minute. That's the version that has a chance in a ten-minute visit.

That's the second half: your evidence has to travel. Into the room, onto one page, in an order someone else can follow.

What sits beside your record, and what each part is for

Human-centred evidence doesn't throw out research, devices or tests. It gives each one a seat, with its job and its limits said out loud.

Research sits beside your record, graded A to D on this site, dated, with the original source linked and the people it studied named, including who it left out. A study of natural menopause at 51 is real evidence. It isn't automatically evidence about surgical menopause at 36, and the grade alone won't tell you that. Who was studied has to travel with the finding.

Device numbers sit beside your record as the rough counts they are. Watches and rings guess your sleep from movement, heart rate and temperature. They don't measure your brain. Studies comparing them with proper sleep lab tests find they're best at telling asleep from awake, worse at counting time spent awake in the night, and worst at sorting sleep into stages. They get less accurate again for people with a sleep problem (Scientific Reports, 2025; Sensors, 2024; Frontiers in Sleep, 2024). We set our rules for this in Your ring doesn't know how last night felt: counts, not verdicts, no "readiness" or "recovery" scores, and how you felt wins whenever the two disagree.

Tests come in when the result would change a decision. The Menopause Society's 2022 hormone therapy statement calls spit and urine tests used to set doses of custom-mixed hormones "unreliable." The FDA says there's no scientific basis for using spit tests to adjust hormone levels. NICE keeps FSH blood tests for specific situations, mainly people aged 40 to 45 with symptoms and changing periods, and people under 40 where early menopause is a possibility. Can a spit test tell you you're in perimenopause? goes through that evidence in full. A test that answers a real question, at the right time, with someone who's caring for you, is very different from a stage label sold by mail.

Whose words, whose record

The record side of this has its own evidence, and I wrote it up in The patient at the centre of the record doesn't exist. Three of its rules belong here, because they decide whose account of a body gets believed.

Nothing is guessed. UBC's Centre for Gender & Sexual Health Equity, in its Research Equity Toolkit, is clear about how to fix records that guess: let people describe themselves, never assume, never fill things in for them. Inside Nila, your identity, the cause of your menopause and your hormone history come only from what you've told us. Never from a page you clicked or something you asked in a chat.

Records have room for a real life. Sex and gender are separate questions, because the health records standard from the HL7 Gender Harmony Project treats them separately, each with its own medical job. And the cause of menopause can be three true things at once. It was for me. A September 2026 review in the journal Menopause found that real health systems rarely recorded menopause status at all. Instead they guessed it from age and prescriptions, and often left out the stage entirely (Ashton et al., 2026). We've taken three of these record fixes to The Persons Project as proposals, because a fix that only lives in one app doesn't fix the system.

It stays yours. You can download it in one click and take it with you if you leave. It's never sold, never used to train AI and never shown to an employer. What happened to everyone else's data looks at thirteen health-app privacy cases, from GoodRx to BetterHelp to Premom, comparing what was promised with what happened. A promise is just a paragraph. So in Nila those rules are built into the app and checked regularly, not only written on a policy page.

What I'm not claiming

I'm not saying what you notice about yourself is always right. Memory slips, there are gaps, and some days nobody has a minute to log anything. That's why we use one-minute check-ins, a recognized questionnaire and device counts beside your notes, instead of a long form months later. I'm not saying this replaces a doctor or specialist. It helps you have a better conversation with one. And I'm not saying a hormone test, a ring or a study is worthless. Each has a job. None of them is you.

An app built by settlers hasn't decolonized anything, and I won't pretend it has. What good design can do, in small, specific places, is shift who holds the say: who writes the record, whose words describe a body, whose body counts as "normal," and who owns what's learned from it. Human-centred evidence shifts all four toward the person. That's the claim, and it's no bigger than that.

When hormones shift, mental health shifts with them shows why this matters. Mental health risk gathers around the moments of change: perimenopause, after giving birth, surgical menopause, the weeks when a treatment stops. Those are exactly the moments a snapshot is most likely to miss and a pattern is most likely to catch, if someone was keeping it, in their own words, before they ever needed to prove anything.

Start with the person, not the snapshot. Build the evidence around them, over time. Then let the tests, the devices and the appointments do their proper job: helping make sense of a life the person already knows better than any single reading ever will.

References

The framework · Human-centred evidence · The six pieces below each carry one rule of this standard.

Companion pieces: The patient at the centre of the record doesn't exist · Menopause isn't pink · Your ring doesn't know how last night felt · Can a spit test tell you you're in perimenopause? · What happened to everyone else's data · When hormones shift, mental health shifts with them

More originals

Next read

Read: how we grade evidence