Crowdmed (https://www.crowdmed.com/), a company launched in April, is using the premise that crowdsourcing can be used to solve medical diagnoses, particularly those of rare conditions that have been missed by doctors. They also believe this can help reduce health care costs. As their website states, “You don’t need a medical degree to help save a life.”
There are two basic options for users, either sign up as a patient or as an “MD” (medical detective, no degree required). A patient can upload their symptoms and histories into the system. A medical detective can help solve others’ cases. Based on a proprietary algorithm, proposed diagnoses are ranked by likelihood. If a list of plausible diagnoses is generated, the patient pays $200 for the service. Medical detectives earn virtual points for qualifying diagnoses; these points are deposited in their accounts.
In another example, UCLA’s Ozcan Research Group has created Biogames (http://biogames.ee.ucla.edu/) a game to help identify cells infected with malaria, a disease that affects around half a billion people in the world every year. The puzzle the group created asks players to examine a number of red blood cells and to identify the ones that show signs of the infection. “We demonstrated that in the case of binary diagnostics decisions (e.g., infected vs. uninfected), using crowd-sourced games it is possible to approach the accuracy of medical experts in making such diagnoses…. within 1.25% of the diagnostic decisions made by a trained professional.”
Even the CDC is getting actively involved in social media (http://www.cdc.gov/socialmedia/). They’ve developed an iPhone app called “Solve the Outbreak”, available at https://itunes.apple.com/US/app/id592485067/. The free iPad app puts you in the shoes of a member of the Epidemic Intelligence Service. Since 1951, over 3,000 EIS officers have responded to requests for epidemiologic assistance within the United States and throughout the world.
The app isn’t used for true detective work, but simulates real-life situations met by EIS officers, and allows users to answer questions faced by EIS offices, e.g. Do you quarantine the village? Talk to the people who are sick? Ask for more lab results? The answers are scored, and the higher the score the more quickly lives are saved. This excellent app is a good way for the CDC to more effectively communicate with the public by getting them actively involved in learning more about disease outbreaks in an engaging way.
Groups such as Crowdmed must rely on what patients report when they input their information. Their users are left with potentially incomplete or incorrect information. Thus, these types of crowdsourcing will be used not primarily to make diagnoses, but to supplement the work of the medical team. The potential seems greater, at least in the short-term, for using them for more simple, e.g. infected-not infected, types of diagnoses.
For more information on social media in public health see: