After Jared Heyman's little sister spent three years with an unsolved medical condition — she'd seen two dozen specialists and racked up thousands in medical bills — he decided he needed to figure out how to help other patients get answers faster.
"I learned the medical system is well-equipped when it comes to helping people with common conditions," says Heyman. "But when it comes to less common things, the system is not equipped."
So Heyman launched CrowdMed, a website that essentially crowdsources potential diagnoses for patients' unsolved medical mysteries. An online community of "medical detectives" — from retired physicians to medical students and pretty much anyone else who wants to play Dr. House — can place bets online about the most likely solutions for patients with hard-to-diagnose medical conditions. So far, they've solved more than 900 cases, and the process usually takes a few weeks and costs about $200.
Still, CrowdMed has also attracted some criticism: some bioethicists have worried about privacy issues and sending patients down wasteful and wrong paths, and medical regulatory experts have wondered whether it's nothing more than the "unauthorized practice of medicine." We spoke to Heyman about his vision for the site and why he thinks CrowdMed can fill a gap that's been left by the medical profession.
Julia Belluz: So you're trying to harness the wisdom of crowds to solve difficult medical cases. What kinds of patients are using your site?
Jared Heyman: The people we help are patients with unresolved medical conditions. They are usually sick for several years — eight years on average. They've seen an average of eight doctors so far. They and their insurance company have incurred $60,000 in medical expenses [on average]. Yet their case is unresolved.
On our site, patients can submit their cases. We expose it to our community of thousands of medical detectives. They collaborate to solve the patients’ cases.
JB: What successes have the patients seen?
JH: Seventy percent of the patients tell us that [CrowdMed] led them to a diagnosis or cure. So we define success at the end of the process: we give patients a survey and ask them, "Did you get insights on our site that led you closer to a correct diagnosis or cure?" About 70 percent say yes. Zero percent of these patients had been cured through the medical system. So it’s working pretty well.
JB: Who are the medical detectives?
JH: We wanted to develop a performance-based rather than a credential-based medical system. In the traditional medical system, everything depends on what degree you got, where you went to medical school, all formal credentials. We believe that system is broken. On our site we don't pay a lot of attention to formal credentials. All the privileges are based upon actual performance. When you first join the site, you don't have a lot of influence, but as you start to improve yourself by making smart diagnostic suggestions, you get high ratings from your peers.
JB: What’s their incentive?
JH: There are cash and non-cash incentives. Cash rewards are up for grabs: if you prove helpful you can win some or all of a patient's cash reward offer, minus a 20 percent commission that we take. On the non-cash side, our detectives mention things like the intellectual challenge — the ability to put a detective hat on and try to solve difficult medical mysteries. They mention learning from their peers, and view CrowdMed as continuing medical education.
JB: Some people have raised concerns that your site can cost patients time, leading them down potentially wrong paths.
JH: On our site, patients can't get definitive diagnoses. Instead, what we try to produce is a short list of the most insightful diagnostic and solution suggestions for the patients to discuss with their doctor, but only their doctor can provide a definitive diagnosis and treatment plan.
But there are 7,000 to 8,000 known medical conditions out there, and most doctors haven't heard of the vast majority of them. We try to provide a short list of things the patients and their doctors hadn't considered yet.