Unearth the power of declared data
Insights That Work
Brand & Retailer tickets for all IIeX events now start at just $99! Get or give one today!

My Experience Using Crowdsourcing Via CrowdMed To Diagnose A Rare Medical Ailment

CrowdMed is another example of the efficacy of several important trends in insight generation: crowdsourcing, prediction markets, gamification, communities and behavioral economics.



According to Medscape

There are no published statistics on the number of undiagnosed “rare disease” patients currently in the United States. The accepted definition of a rare disease is one that affects fewer than 200,000 Americans at any given time. According to the National Institutes of Health (NIH), there are between 6000 and 7000 rare diseases affecting from 25 to 30 million Americans, making it likely that most, if not all, healthcare professionals have these patients in their practice. The average time to diagnosis for a rare disease is about 7.2 years. Patients with disorders of unknown etiology (undiagnosed rare disease patients) sometimes never receive a diagnosis or appropriate treatment.

As someone going through the roller coaster of trying to obtain a diagnosis for an uncommon disease (I previously wrote about it here), I can attest to how frustrating, confusing, frightening, and expensive this process can be. I’m 4 years into my own journey to find a diagnosis for what appears to be some variety of uncommon neuromuscular disorder, and although I’ve collected tons of diagnostic, empirical and experiential evidence and seen many doctors, I am no closer to a definitive diagnosis (let alone prognosis or treatment plan!) than I was when my initial symptoms first manifested.

And that brings us to CrowdMed.

CrowdMed utilizes a crowdsourcing model to solve difficult medical cases quickly and accurately online. It should be specified that the platform was not designed as a substitute for medical professionals, but aims to provide help to provide a second opinion, narrowing the field of possibilities and proposing suggestions that physicians can consider.




I became aware of CrowdMed several years ago as a new venture from Jared Heyman, the founder and CEO (now Chairman) of Atlanta-based market research supplier InfoSurv. Infosurv was an early (and successful) proponent of prediction markets as a research method, and that focus on harnessing the “wisdom of the crowd” has been a driving force behind the CrowdMed model as well.

Jared almost lost his own sister to a medical mystery that took nearly three years, almost two dozen doctors, and more than $100,000 in medical expenses to diagnose. Her medical mystery was eventually solved by a large, interdisciplinary team of top-notch medical experts who all collaborated and worked together, and remained focused on the case for as long as it took to reach a consensus-based diagnosis. Jared speaks movingly about his own experience trying to help his sister in the video below.



CrowdMed takes this concept to the next level, helping anyone connect to a world of medical experts, all working together to solve medical mysteries. CrowdMed describes their “Medical Detectives” as a resource to deliver real results for people like me who are struggling with a not-easily diagnosed health condition. As to the makeup of this resource, on their website they say:

There are so many people in the world with exceptional knowledge on a range of medical conditions who actually are not doctors. While most of our “crowd” is comprised of people who are active in medicine and healthcare, many others are not medical professionals but have first-hand experience with various medical conditions.

After asking Jared to speak at IIeX Health earlier this year, I shared with him some of my own struggle and he offered a two month “pilot” for free to check out the platform, and then share my experience publicly. I agreed, and you’re now reading the result.

The technical process is relatively straight forward:

  1. Complete an online patient questionnaire and upload all relevant medical information (test results, imaging studies, medical records, etc..), all of which can be done anonymously. I selected the pseudonym “Randal”
  2. Decide how long you want your case to be online and how you wish to reward the volunteers who help solve it. I went with 60 days and allocated $500 as a reward value.
  3. CrowdMed’s patented prediction market technology then collects and filters diagnostic and solution suggestions from the Medical Detectives who participate in the case.
  4. You can invite their entire community to help, or limit participation to their top case solvers.
  5. At the end of the process, you receive a detailed report containing the best suggestions to discuss with your doctor.
  6. Payment is a snap, and if CrowdMed results aren’t accurate you get your money back.




So how was the experience? In a nutshell, I’d describe it as “immensely gratifying”, which is pretty strong praise indeed for folks like me who come to CrowdMed with a decidedly negative experience of the health care system.

First, the engagement by the Medical Detectives was surprisingly consistent, friendly and supportive. Each used the Chat function on the site to ask me clarifying questions and offer their own experience and messages of support. That is no small consideration: for me at least dealing with medical issues can be an isolating experience for a variety of reasons, so it’s very affirming to make these connections with folks who are interested in helping. And not only were the individual chats helpful, but the chats between the Medical Detectives as they shared information with each other to reach consensus on recommendations were informative and helpful.

Each user/case has their own online dashboard to easily provide updates, engage with the Medical Detectives and manage your account. The UX was simple, clean, and easy to navigate.




After just a few weeks about twenty folks were working on my case and about a dozen suggested diagnoses had been tendered.  Based on my own research, some of these seemed to be better candidates for exploration than others, but all of them had a sound basis for inclusion based on the information I had shared with the Medical Detectives and were worth consideration as possible diagnoses.

I could see each of them on my dashboard. Here is a screenshot of the final consideration set put forward in my case:





The Medical Detectives aren’t just doing this from the goodness of their heart, although everyone I interacted with struck me as genuinely wanting to help. They place “bets’ on the accuracy of their proposed diagnoses and are paid once patients can confirm whether they are right. The screenshots below show the leaderboard for how much the leading Medical Detectives have earned for their services:




The end result is a very comprehensive report that captures not only all of the information I had provided, but also detailed transcripts of each chat, email, and rationale behind the proposed diagnoses.




The big question is “How accurate were the suggestions?”.

Unfortunately in my case, that is the part of the CrowdMed experience I cannot speak to. Despite having the ability to  invite my medical team to engage directly with  CrowdMed, mine did not do so. When I shared some of the suggestions with my Neurologist during an office visit, he was dismissive of the idea. I am in a tightly managed HMO insurance plan and have few options to visit specialists without the approval of a Primary Care Physician, so discussing these results with others who may have more experience or openness is challenging. That is simply the tradeoff of this type of plan for someone like me vs. the excellent routine care available for the rest of my family. “It is what it is” right now, but hopefully at some point in the future I’ll be able to use the information provided by CrowdMed to help get to a diagnosis.

Setting aside my personal conundrum, according to the case studies available on the CrowdMed site, personal recommendations of other users,  and  the stats of “cases solved” on the leaderboard there is every indication that the crowdsourcing model works.  And I can personally attest to having a great experience using the site and interacting with the Medical Detectives.  In short, I’m a believer and would recommend to anyone else struggling  with a hard to diagnose ailment to check out CrowdMed.

I’d be remiss if I didn’t thank Jared and his team, as well as the Medical detectives who engaged in my case, for the opportunity to try out the platform and share my experience with others. It was a great experience overall, and I am grateful for it.

So what does all this have to do with market research? CrowdMed is another example of the efficacy of several important trends in insight generation: crowdsourcing, prediction markets, gamification, communities and behavioral economics. It’s a synthesis of some of the best thinking from each of these approaches, wrapped up in a great user-focused design. This model could easily be applied to a variety of business issues market research is generally applied to, from ideation to testing and, of course joins the already significant body of evidence of the predictive value of “wisdom of the crowds” based approaches. Companies such as Spigit, KL Communications, Intengo and Consensus Point offer research specific variation of this model already, and of course BrainJuicer and Jon Puleston of Lightspeed-GMI have been talking about the positive impact of  prediction markets for many years.

In my case we might not be able to measure the impact of this approach from a validation perspective, but I think it’s safe to say that as an industry we can take significant inspiration from CrowdMed and perhaps even treat some of our own ailments by following their example.

Please share...

3 responses to “My Experience Using Crowdsourcing Via CrowdMed To Diagnose A Rare Medical Ailment

  1. Very interesting Lenny, I wish I knew it about it years ago. I was misdiagnosed for 6 years, gone through cardiologists, gastroenterologists, then acupuncturists, naturopaths, chiropractors, dietitiants … then ended up in a coma as a consequence of a hypertensive crisis. The aftermath showed a pheochromocytoma, a rare tumor at the adrenal gland. Mistaken for stress … sigh.
    Alas, in medical schools they teach “if you hear hoofsteps, think horses not zebras”, which is fair enough, but leave all of us zebras in trouble. This could be a great way to harness the wisdom of the crowd. Thanks for sharing.

    1. Thanks Robbie, and sorry to hear about your own experience. I’m glad an answer was eventually found though! I am a big fan of the “democratization” of information occurring now, and this is one more example of that in action that has the potential to create a real difference in people’s lives. There is much to be learned here for sure!

  2. Hi Lenny,

    I hope you are well. We met at the Iiex here in Atl in June. Kimberly Nasief (my employer and friend) mentioned you and suggested I use the CrowdMed to possibly help me discover the reason for the medical issues I’ve been dealing with since Aug 2014. I actually think I may have found the underlying issue with a random Google search of my symptoms I decided to try this morning, but I am going to try your technique and see what might come up there as well. I am so sick and tired of being sick and tired. Tired of dealing with all of these doctors who only want to treat the symptoms and not take the time to find the root cause. I need a solution before I lose my mind.

    Thank you in advance for this:-)

Join the conversation