Crowdsourcing as a Medical Tool: Interview with CrowdMed CEO Jared Heyman.
CrowdMed is an online medical crowdsourcing platform where people submit medical cases with information about symptoms, medical history, family history, and other pertinent data. Subsequently, the community of “medical detectives” suggests diagnoses and places point bets on the outcomes they think are most likely. CrowdMed’s patented prediction market algorithms aggregate the medical differential set by the crowd and distills it down to a probable list of diagnostic suggestions for each patient.
Jared Heyman is the Founder and CEO of CrowdMed, where he is actively furthering the strategic direction of the product and vision of the company. I asked him about CrowdMed and his thoughts about its function as a novel medical tool.
Tom Fowler, Medgadget: Before founding CrowdMed, you spent 2 years on an around-the-world traveling sabbatical. What is the one thing you do differently now because of that experience?
Jared Heyman: My sabbatical was instrumental to the creation of CrowdMed. It gave me a chance to step back and really think deeply about what I wanted to do next as an entrepreneur. I came to the conclusion that I should do something at the intersection of my greatest passions, which include technology, crowd wisdom, and medical diagnosis. My experience at Infosurv (the previous company I founded) fueled my passion for the first two, and my sister’s 3 year diagnostic struggle fueled my passion for the latter.
Medgadget: Explain the concept of “wisdom of crowds” and how it works in this medical setting.
Jared Heyman: “The wisdom of crowds” is the notion that large groups of non-experts can collectively be much wiser than individual experts. It’s counter-intuitive, but has been proven to be true, most famously in the book “The Wisdom of Crowds” by James Surowiecki and less famously in academic papers by Justin Wolfers, Robin Hanson and others. I spent several years studying crowd wisdom and prediction markets at Infosurv.
Given my deep appreciation of crowd wisdom, I was surprised that no one was yet harnessing it for medical diagnosis and treatment. If we can crowdsource something as expert-driven as writing an encyclopedia (e.g., Wikipedia), why not crowdsource something as expert-driven as medicine? I knew that it was theoretically possible to do so, but only with the right mechanism in place. Our mission at CrowdMed is to build this mechanism. You can see how CrowdMed works here: https://www.crowdmed.com/how-it-works
Medgadget: You were partly inspired to found CrowdMed from the experience of your younger sister who only after 3 years, $100k+, and 16 doctors, received the correct diagnosis of her rare genetic disorder. How did CrowdMed do with her case?
Jared Heyman: Once we’d developed an early prototype of CrowdMed, we wanted to test it with a difficult real-world medical case. Of course, my sister’s case came immediately to mind. I wasn’t very optimistic that we’d solve it given how long it had stumped her doctors, but figured it was worth a try just to see where our system would break down. Much to my amazement, the crowd’s consensus on her most likely correct diagnosis was “Fragile X-associated primary ovarian insufficiency” — which was exactly what she had.
Medgadget: Would you consider CrowdMed a “medical device” that could eventually be routinely ordered in a clinical setting?
Jared Heyman: I wouldn’t classify us as a “medical device” but rather a “medical informational resource”. We’re much closer to Wikipedia, Quora, and WebMD than we are to a physical diagnostic test. That said, I do believe that CrowdMed has eventual clinical applications. Though the majority of our cases are currently submitted by patients, we welcome physician-submitted cases as well and have seen several on the site.
Medgadget: Certainly there have been some critics on the accuracy of CrowdMed diagnoses. What criticism have you heard the most, and what is your best response?
Jared Heyman: I’ve seen many more criticisms of CrowdMed from people unfamiliar with it than our actual users. About 80% of our patients have reported their top diagnostic suggestions to be accurate — and bear in mind that our average patient has been sick for 6 years, seen 8 doctors, and incurred over $50,000 in medical expenses by the time they submit their case to our site. We also get rave reviews from our Medical Detectives, who find the case-solving process to be fun, engaging, intellectually stimulating, and sometimes even financially rewarding.
The criticisms usually come from people who have heard about our concept for the first time and just can’t get their heads around a crowdsourced approach to medicine. I’d imagine that Jimmy Wales heard the same sorts of criticisms when he first told people that he was going to crowdsource an encyclopedia.
Medgadget: What are some tips you can share for others that are leading ventures that aim to change fundamental understandings in the healthcare industry today?
Jared Heyman: Persevere! The healthcare industry is probably more resistant to change than any other in our economy, but that’s also why innovative entrepreneurs must keep trying to change it.
# # #
Tom Fowler worked as a programmer in the healthcare IT industry before settling back in school to learn how to become a doctor. He likes to dabble with biotech startups, write postcards to his relatives, and play his ukulele. He was a TEDMED ’13 scholar, has published research in biomedical informatics, and continues to advocate for international maternal and child health. Currently in the SELECT MD leadership program at USF Health Morsani College of Medicine.
This article was originally produced for medGadget.