Clinical research is expensive.
Running a proper clinical trial often costs millions—sometimes tens or even hundreds of millions—depending on the study. Because of this, researchers nearly always need outside funding. Governments, charities, and universities do fund some trials, but the biggest player is the pharmaceutical industry. For drug development that makes sense: a company funds the trial, a new therapy is discovered, and, if successful, that company brings a treatment to market. Patients benefit, and the company recoups its investment.
But there is a major issue: pharmaceutical companies need patent protection to justify those investments. Without a patent, a rival could quickly copy and sell the same drug at a fraction of the cost. That means industry money flows mainly toward treatments that can be patented, marketed, and sold profitably. Supplements, lifestyle changes, generic drugs, and other approaches that can’t be easily protected by intellectual property often fall through the cracks. They may be cheap, safe, and widely used, but they attract little research because there’s no clear financial return.
This leaves a dangerous gap. Many conditions that don’t have obvious patentable drug solutions—such as chronic fatigue, fibromyalgia, or long COVID—are left under-researched. Doctors searching for evidence-based guidance may find very little in the literature, and patients struggling with daily symptoms are left with few researched options.
Where do people turn when the research just isn’t there?
Increasingly, they look online. Reddit threads, Facebook groups, health forums, and patient communities are filled with people sharing their own experiences—what they’ve tried, what’s helped, what hasn’t. When I first looked into these online sources of conventional wisdom, I expected them to mostly be unreliable, but what I found suggested exactly the opposite. Hidden in subreddits and comment threads are thousands of stories about magnesium for sleep, antihistamines for long COVID, pacing strategies for fatigue, or yoga for pain. Some people report real improvements, others don’t, but together these posts represent an enormous, untapped pool of patient experience.
Of course, anecdote isn’t proof. What helps one person may not help another, and online reports are vulnerable to bias, exaggeration, or outright misinformation. But it’s also true that this is often the only evidence available in under-researched areas. Patients are already using these remedies. Clinicians are already being asked about them. What’s missing is a way to organise, filter, and make sense of this “folk evidence.”
That’s why I’m building Folk Evidence.
The idea is simple: create an open-source, online directory that gathers reports of non-prescription, non-patentable remedies from public forums and other sources. Instead of scattered anecdotes, we’ll have structured data. Instead of endless scrolling through Reddit, a GP could see:
- What remedies people are trying for a given condition.
- How many report improvements versus no change or harm.
- Which options come up most often, across the largest number of people.
- How reliable the underlying reports are.
The goal isn’t to replace clinical trials or formal evidence—it’s to fill the gap until those exist. A patient with long COVID or chronic migraine could come to their doctor, and the doctor could check Folk Evidence to see which remedies have the strongest anecdotal support. It won’t be a guarantee, but it could point them toward something safe and potentially useful, rather than leaving them with nothing at all.
Folk Evidence will have to address some big challenges.
- Data quality: separating genuine experience from spam, pseudoscience, or miracle-cure claims.
- Bias: people are more likely to post when something works, less so when it doesn’t.
- Safety: some remedies may interact with medications or carry risks that aren’t obvious in forum posts.
- Clinical trust: doctors will only use the platform if it’s transparent, cautious, and responsibly curated.
I plan to tackle these by combining automated tools (natural language processing to identify remedies and outcomes, filters for unreliable sources) with human moderation and, ideally, professional input. Remedies will be presented by percentage likelihood of benefit, number of supporting reports, and safety category (for example, clinician approved, or partially researched). Every entry will link back to its source, so readers can see the original context.
I want Folk Evidence to be open to everyone:
- Patients, who can see what others with their condition have tried.
- Clinicians, who can recommend options grounded in real-world experiences rather than guesswork.
- Researchers, who can identify promising remedies that deserve proper study.
This blog will document the journey of building Folk Evidence. I’ll share updates on the technical side—how to gather and process the data—as well as reflections on the ethical and medical questions that come with it. I’ll also highlight specific remedies or conditions as case studies, showing how online anecdotes can be transformed into usable knowledge.
Above all, I want to be transparent: this site will not give medical advice. Folk Evidence is not a replacement for professional care. It’s a way to gather and organise experiences so that doctors and patients can make better-informed choices together.
If this idea resonates with you, whether you’re a patient, a clinician, or just curious, I’d love for you to follow along.
Comment, share your thoughts, and help shape what Folk Evidence becomes. This is just the start, and the more perspectives involved, the better it will be.