A website called CrowdMed is changing the way patients can be diagnosed. It uses a group of people to collectively analyse challenging medical cases through ‘crowd wisdom.’ A TedMed talk about the website by one of its co-founders can be found here.

Jessica Greenwalt, Jared Heyman, and Axel Setyanto founded CrowdMed in 2012 to help people who hadn’t been diagnosed through traditional medical routes. Since its launch, the website has handled over 900 patient cases. The people who have been diagnosed through the website have, on average, seen eight doctors, experienced symptoms for eight years, and spent over $50,000 on medical expenses. The idea has proved popular; over half of users agree that CrowdMed has helped them get closer to a diagnosis or cure, and the company now has users from twenty-one different countries worldwide. Cases from the website are also being used to teach students on a medical course at the University of California.

The website works using a combination of contributors (called ‘case-solvers’), incentives, a reputation system, online chat spaces, and complex algorithms. Patients post their stories on the website, and case-solvers, who are offered incentives for helpful suggestions, respond. The case-solvers are a diverse group, including medical professionals, researchers, and students. They are rated on a reputation system that gives more influence to those who have previously made good suggestions. The case-solvers can then discuss the case and can talk to the patient through online chat forums. An algorithm then ranks potential diagnoses to identify the most helpful answers. Two important caveats to the system are that patients should always go to doctors for diagnoses before asking CrowdMed and that the website should never be used in an emergency. However, for ‘un-diagnosable’ cases, it has been shown to be helpful.

One example of a successfully solved case, given by co-founder Jared Heyman in his TedMed talk, is of a boy called Joseph who, at age twelve, became seriously ill. He suffered from constant fatigue and body pain. After five years, fourteen doctors, and $75,000 in medical costs, the family felt as though they were out of options. After the family posted the case on CrowdMed 42 case-solvers worked together to diagnose Joseph with Lyme disease. Previous doctors had ruled out the condition based on a negative test, but CrowdMed participants claimed that the test was often inaccurate, and advised another one. This was positive for Lyme disease, and Joseph was finally treated.

As the BBC points out, the idea that crowd judgements can be better than those of any individual has a long history. One of the earliest people to make this suggestion was Charles Galton, the cousin of Charles Darwin and a famous statistician and psychologist in his own right. In 1907 he described how at the ‘guess the ox weight’ game at country fairs the combined average of the crowd’s guesses tended to be more accurate than anyone else’s, including those of the experts.

However, this is not always the case. A team of researchers in Zurich, Switzerland, found that each person’s guesses are influenced by other people’s guesses. A person who is giving their best estimate will tend towards what other people have already answered. This creates a consensus, but one that is often inaccurate. One member of the research team, Frank Schweitzer, carried out further studies. He concluded that the outcome depends on those involved; a group that has generally good judgement can work together to form more accurate conclusions, but a group of people who don’t will produce inaccurate answers. You can read more about the factors influencing crowd judgements here.

The complex factors influencing ‘crowd wisdom’ mean that the usefulness of the conclusions depends on the circumstances of the decision-making ‘crowd.’ Crowdsourcing answers can also be unreliable, less private, and highly dependent on the group offering diagnoses. However, the success stories from CrowdMed show that the website can be helpful for patients with difficult-to-solve symptoms. Jared Heyman said in his TedMed talk that he thinks, “crowdsourcing medical answers will be commonplace,” and a column by the New York Times that is based on the same principles supports its growing popularity. However, for the foreseeable future, doctors are likely to remain the primary source of information for the majority of people.

For more information about CrowdMed, you can visit their website, or click here to listen to a discussion about them.

What are your thoughts on CrowdMed? Share your stories, thoughts, and hopes, with the Patient Worthy community!

Are you searching for a diagnosis? Check out our partner group, the Rare & Undiagnosed Network (RUN)!