CrowdMed is using crowds to solve difficult medical cases. Basically, it works like TopCoder for clinical diagnostics: On the one side, patients who have been sick for several years and have been to several doctors without much success, can upload their health history and create an anonymous „personal medical case“. On the other side, Medical Detectives (ranging from medical student, retired physician to regular people) can select active medical cases to help solving. After an collaborating phase between different Medical Detectives and clarifying communication with the patient, several diagnoses and treatments are suggested. The patient can use the CrowdMed´s ranking system to decide for a solution and present it to his personal physician. If the physician recognizes the suggestion as correct or helpful, the case is solved and the respective Medical Detective gets compensated.
Value Creation and value capturing
The value for the long-suffering patients is to finally solve their medical issue. As these case are rare and hard nuts to crack, the crowd-based model enables the patient to benefit from the knowledge and cross-communication of many medical advisors at the same time. So far they have successfully solved more than 50% of the cases submitted in the past years.
The medical advisors can use their existing medical knowledge beyond their professional occupation in an effective and flexible manner to help people anywhere to get the right treatment. While they can build up reputation within the community, they also earn a cash compensation when solving a case.
The patient can either use the limited free case submission option or purchase the Standard ($149/month) or Premium ($249/month) option to submit their personal medical case. CrowdMed keeps $49 of this fee and the rest is shared between the Medical Detectives who contributed to the chosen diagnosis. A case should be online a minimum of 2 months in order to get a sufficient number of medical detectives to participate. If after two months no helpful solutions is provided, the patient can get the money back.
In order to guarantee a high quality of diagnosis, CrowdMed uses a point-based system to track Medical Detective’s performance: Based on their confidence in their answer they can assign points to their suggestion and get rewards appropriately to it.
CrowdMed uses several metrics (confidence level, peer-ratings, past correct diagnosis) to determine the quality of a suggestion and to provide the patient and its personal physician with the necessary information to choose from the suggestions.
Additionally, this data is used in a „DetectiveRating“ in which every Medical Detective is rated form 1-10. Having a high ranking enables them to access more complex cases and get higher cash compensations.
While these mechanisms should increase platform engagement it also serves as a quality screening: Everyone, indifferent of their education, can sign up to become a Medical Detective. The idea is to judge suggestions not based on the person’s academic background but rather on performance to filter out bad suggestions and rank them properly.
While strong indirect network-effects are in place (the more Medical Investigators sign up, the higher the quality of diagnosis), direct network effects are not as strong. A referral system is existing (patients can invite others to CrowdMed), but other than a $10 discount on their monthly fee, they have no direct advantages of more patients joining CrowdMed (e.g. no communication between people suffering from similar conditions is fostered). The interesting aspect is that CrowdMed does encourage collaboration and rewards are shared between every contributor. While this can be a demotivating for some detectives as well, it overall encourages people to exchange ideas and contribute to each others suggestions.
CrowdMed is a very interesting application of the use of „wisdom of crowds“. The quality of diagnosis in a hospital often heavily depends on collaboration and knowledge exchange of medical professionals. CrowdMed enables all kinds of people to interact and share their thoughts on complex cases while having a smart peer-review system in place to guarantee quality. It will be interesting how this crowd-diagnosis concept develops and if it expands into the solely professional context as well (e.g. different hospitals discussing cases between each other).