Fitness Files: The more the merrier, when it comes to a diagnosis

Jared Heyman’s sister, Carly, was desperately ill and confined to her bed. She sought help from two dozen doctors, resulting in $100,000 in medical expenses but no diagnosis. After two years, she was no closer to a cure.

Heyman decided to do something. A veteran of the successful start-up Infosurv, he put his talents to work helping patients like his sister toss out a wider lasso to pull in medical knowledge.

His answer: CrowdMed, a site where, according to Carrie Arnold of PBS.org last year, patients worldwide submit their cases, and medical detectives, some doctors, read the details of the patient’s symptoms together with previous diagnoses, treatments, test results and medications.

Medical detectives aren’t paid; they’re motivated by the satisfaction of solving medical mysteries. Sometimes patients offer about $200 for an accurate diagnosis, but many medical detectives donate this money to charities.

After a patient submits a case, “detectives” are able to ask for clarification with online chats. Through the process, patients, given pseudonyms for anonymity, may be presented with as many as a dozen potential diagnoses.

Here’s the interesting dynamic: The medical detectives earn points by betting — not for money — on the correct diagnosis. CrowdMed’s site explains that the detectives are “rewarded” by having the highest points, accumulated from frequent accurate diagnoses, therefore “earning recognition from their peers … but most importantly, helping patients … who have struggled for years … to get their lives back.”

According to The Economist in an article of a few months ago, “The crowd will see you now,” 80% of the diagnoses are correct. That percentage comes from the CrowdMed website. Although I’ve read differing percentages elsewhere, many quotes from the site ring true to me. Collective wisdom jogs doctors’ thinking and turns medicine into what Wired Magazine calls a “team sport.”

In one of my columns in August, I wrote doubtingly about online medical care that uses algorithms to aid diagnoses and prescribes medication without the patient being seen. CrowdMed is different. Patients pay to subscribe to CrowdMed but are not sold a treatment or medication.

What patients garner is an opinion, from the focused eyes of medical professionals, other patients and laymen together with the algorithmic probabilities of the CrowdMed site.

According to Terri Yablonsky Stat’s Chicago Tribune article from April, “A proprietary algorithm assigns probabilities to each possible diagnosis based on the medical detective’s prior performance. From this, patients get a detailed report with their most likely diagnosis. They have 30 days to discuss the list with their doctor for a final diagnosis and treatment plan.”

It’s important to me that patients take their CrowdMed diagnoses to their doctors’ consultation room. There, the physician has past records, the patient and the CrowdMed consultation all together to design future treatment.

Also of note is that crowdsourced diagnosis might save our healthcare system money.

Yablonsky Stat says, “Heyman hopes to partner with health systems, hospital groups and insurance providers to bring the concept to their members” in a way that quickly solves cases and costs less than the status quo. A quicker solution means less suffering, less work missed, less unnecessary testing and less surgery, and thus fewer dollars spent.

According to The Economist, CrowdMed is most beneficial for the 25 million sufferers of rare diseases who visit a doctor who has seen maybe one such case in a lifetime of practice and cannot possibly be expected to know what it is. Patients with rare diseases spend an average of 7 1/2 years searching for help. CrowdMed averages two months.

Heyman’s sister, Carly, eventually found that a rare genetic disorder was responsible for her illness. To test CrowdMed, Heyman plugged in his sister’s case. “We wanted to see retroactively if we could crowdsource the right answer,” he said.

With 100 medical detectives on the case, the CrowdMed community took just three days to correctly diagnose her.

For intractable illnesses, the wisdom of the crowd is worth a try.

Newport Beach resident CARRIE LUGER SLAYBACK is a retired teacher who, since turning 70, has ran the Los Angeles Marathon, placing first in her age group twice.