What if patients could learn from each other?
What happens when patient data stops living in silos and thousands of health journeys can start learning from each other?
By Brian Rey · · what-if-patients-could-learn-from-each-other.md
The other day I found a really interesting company on X called Atlas Network.
They are working on health tools for people with rare and complex diseases, and one part of the idea stayed in my head.
Right now a typical patient has blood tests in one place and doctor notes in another. Smartwatch data lives somewhere else. Years of symptoms live only in their memory. Meanwhile someone with almost the same disease is going through something similar on the other side of the world. They try different treatments and get different results. And these two stories barely connect.
That feels like a huge waste.
What if AI could help put all of this together?
Your symptoms and lab results. Your treatments and your sleep. All the small things that change week to week. That could become one timeline. Then the system could show you what changed and what came before things got better or worse.
What is even more interesting is what happens when you connect many patients.
Imagine thousands of people tracking the same things over time. You might see that people with a certain biomarker pattern respond better to one treatment. Or that some symptom starts changing weeks before a lab result does.
This could also become very interesting for clinical trials.
A trial checks patients at fixed points. A lot happens between those visits. People sleep worse or eat differently. They start a new medication or stop one. Most of that never gets captured.
AI could make following people much easier. It could ask simple questions every day and summarize lab results. It could help researchers see how everything changes together over time.
Then instead of only asking whether a drug worked, we could start asking much better questions.
- Who did it work for?
- When did the change start?
- Which biomarkers moved first?
- What happened to people who did not respond well?
- Are there smaller groups of patients who seem to respond in similar ways?
This matters even more for rare diseases. Each hospital or country might only have a handful of patients. If everyone stays in their own silo we learn very slowly. If those journeys can be connected safely, every new patient adds a little to what we know.
Maybe one of the biggest opportunities in health right now is helping patients tell their own story better. Then letting thousands of those stories learn from each other.
We already have huge amounts of health data. Most of it is scattered or never used again. If AI can help connect those pieces across thousands of people, we might start understanding disease in a very different way.