Would you trust the "wisdom of the crowd" over your own doctor? CrowdMed thinks you might. The San Francisco start-up has an audacious plan to use crowdsourcing techniques to tap the "collective wisdom" of strangers to help diagnose patients - particularly those who've bounced from doctor to doctor for years trying to understand uncommon symptoms.
While many may worry that healthcare is too important to trust to strangers, I think this is awesome.
After all, crowdsourcing is already used to help find missing persons, track down terrorists, answer life's vexing questions, pick stocks - and to select our President. SETI uses crowdsourcing to search for extraterrestrial life. Why not employ crowdsourcing to help our multi-trillion-dollar healthcare industry?
CrowdMed recently received $1.1 million in seed financing from some of Silicon Valley's top venture capital firms, including NEA, Greylock Partners, Y Combinator and Andreessen Horowitz.
Ask Your Doctor? No. Ask the Crowd.
CrowdMed works like this: Patients pay a $199 fee to list their case on CrowdMed. They fill out a "patient questionnaire" that details their symptoms, case history and personal information. Though CrowdMed founder Jared Heyman declined to say exactly how many patients have enrolled so far, he claimed that there has been "pretty strong demand." Without the fee, Heyman explained, the site would be overwhelmed with patients who might not get diagnosed.
Once a case is posted, the crowd, what CrowdMed somewhat coyly terms "MDs" - for "medical detectives" - can review the patient's information and offer up what they believe is the correct - or most likely - diagnosis.
According to Heyman, "close to 3,000 people have signed up as medical detectives." He said CrowdMed's "MDs" include doctors, residents and "regular people that like solving medical mysteries." Why sign up to be a medical detective? First, there's the chance to help patients. Second, CrowdMed awards its detectives "points" for the diagnoses they correctly predict.
CrowdMed utilizes a so-called prediction market methodology to help glean the correct diagnosis. For example, when a detective selects a case to review, they use up some of their points. They use up still more when they suggest a diagnosis or vote up (or down) other suggested diagnoses. Essentially, it "costs" to play. The more accurate their predictions, however, the more points they are ultimately awarded.
Points do not have any cash value, however. For now, they can be exchanged only for donations to Watsi, an organization that helps fund medical treatments in the developing world. Heyman did not say how much CrowdMed is donating.
While it's true that CrowdMed's detectives may not always correctly diagnose a particular patient, if they can narrow the likelihood of someone's illness to, say, two or three likely options - those that garner the most points, for example - that could speed up decision making and help point to which tests should be perfomed.
In Crowd We Trust?
The obvious question: Can a crowd of strangers with unknown amounts of medical expertise be trusted to safely and correctly diagnose baffling medical problems? CrowdMed claims that after "four years of development" it possess a patented "unique technology" specifically designed to optimize group intelligence for medical diagnostic purposes. From its site:
Groups hold far more knowledge collectively than any individual member, no matter how brilliant. With hundreds of minds working in parallel, groups can process information much faster than individuals.
Heyman told me that his sister suffered for three years from a rare disease. Once it was finally correctly diagnosed, doctors were able to significantly ease her symptoms. CrowdMed used her case to help validate its model - Heyman says it accurately diagnosed her within days.
What Do Real MDs Think?
The first rule of medicine is primum non nocere, Latin for "first, do no harm." It does not necessarily apply to the crowd. Not surprisingly, the CrowdMed approach bothers many real doctors.
Dr. Hubert Chen, the Associate Medical Director for biotech pioneer Genentech, said, "I want to be enthusiastic, but I have concerns about it." Dr. Chen's primary concern was the potential for numerous "false positives" that CrowdMed's "detectives" might generate: "I've seen many patients misled by the Web. Doctors often have to un-educate them."
Dr. Aaron Roland, wo runs a family practice in northern California and is an associate clinical professor at UC San Francisco, had different concerns. "I wouldn't pay $200," Rolan said. He also wondered whether CrowdMed could attract the scale it needs. "Crowdsourcing is good when there's a lot of people in the crowd," he said, "but until you get that crowd, I'm suspicious."
Not surprisingly, Martorana was very positive about the concept. There are many "experts," she said, not necessarily doctors, who may have suffered from a particular disease, or have a family member who has suffered, and whom can now contribute to the site.
She hopes to "reach out" to staffers - not just doctors - at medical research, counseling and support organizations that concentrate on specific issues - think, autism, for example, or Parkinson's dioease - and encourage them to participate in CrowdMed.
Martorana also suggested crowdsourcing diagnoses could be a boon for health insurance companies: "If you are insured and going to multiple specialists, but not getting relief, that costs a lot of money - you, your employer, your insurer all must bear those costs. At some point, there probably will be a pretty significant revenue stream for CrowdMed coming from insurance companies. Right now, their cost numbers are staggering."
The relatively paltry $1.1 million CrowdMed has raised so far suggest that investors remain unsure of the idea's potential risks and rewards. But connecting patients with chronic medical symptoms to experts, regardless of their titles, clearly holds massive disruptive potential. CrowdMed's ambitious, even inspiring idea is to use connectivity, collaboration and collective intelligence to help people avoid needless suffering. Despite the risks, it seems like it's a worth a try to me.
- Social Revolution: Crowdsourcing For Change
- The Problem With Crowdsourcing Crime Reporting In The Mexican Drug War
- The Key To Crowdsourcing: Smarter Crowds