CrowdMed uses crowdsourcing to solve rare medical cases

By April 22, 2013
crowd forming a smiley

CrowdMed users are asked to trust in the ‘wisdom of crowds’ manifested in Internet feedback as an approach to diagnosing and treating illnesses that have been stumping the doctors.


There are thousands of different illnesses among human populations and new forms of existing maladies are appearing all the time. Sometimes a patient cannot obtain a diagnosis, sometimes s/he will be faced with contradictory diagnoses. Moreover, the profusion of new illnesses, new medication, coupled with new lifestyles, can add a whole series of unknowns to the process of analyzing a patient’s condition. These multiple parameters are complex and the way they interact is not always well understood, even by experts. This is what prompted San Francisco-based startup CrowdMed to tap into the ‘power of the crowd’, using the Internet. The basic approach is to gather a critical mass of data from Internet users on their illnesses, their symptoms and their experiences and to aggregate them in order to discern recurrent patterns, establish causal links and identify remedies. In short, CrowdMed aims to combine data from the greatest possible number of cases in order to understand rare conditions which are proving difficult for doctors to diagnose; and going to the crowd is the best way to obtain a broad response.

Collecting patient experiences from all over the Web

The CrowdMed site enables anyone and everyone to submit for further diagnosis or advice cases of rare illness where a doctor has not been able to make a diagnosis or where the patient feels unsure about the diagnosis. The ‘patient’ needs to supply comprehensive information on his/her symptoms, medical history and previous illnesses,  then the community of ‘Medical Detectives’ (MDs) registered on the site will suggest a diagnosis and try to work out the most suitable remedy. CrowdMed provides the platform for analysis based on aggregation of data fed back from the community. The results are then structured into suggestions for each patient. The service costs the ‘patient’ or requester $199 per case, but the fee is reimbursed if the patient is not satisfied with the suggestions offered.

Creating collective trust in the crowd

CrowdMed’s approach is thus one which requires users to trust in the collective intelligence of the crowd as a means of solving problems that people in positions of trust, but who are of course not infallible, have been unable to solve. The CrowdMed team point to three special advantages in this way of working. First, what the company calls ‘collective knowledge’; this is based on the idea that a group of people, even though less well-informed as individuals, together comprise a wide and valuable source of knowledge. Secondly, ‘collective intelligence’, which means that a large number of brains working together at the same time can process far more information much faster than one person alone. Last but not least, ‘unbiased opinions’; the argument here is that although everyone suffers from individual biases, these biases will, in a large and intellectually diverse group, tend to cancel each other out. CrowdMed is not however planning to compete head-on with medical practitioners. On the contrary, Jared Heyman, the leader of the CrowdMed team, sees the company potentially collaborating with doctors or medical institutions in future to try to create better diagnostic procedures.

Legal mentions © L’Atelier BNP Paribas