When we get sick, doctors capture all kinds of data about us in their office or at the hospital: bloodwork, scans, interviews, probes. Then there's a follow-up appointment to see how we're doing. Maybe if it's serious, you get a follow-up call from a nurse. That leaves the other 99.9% of our time spent not in the clinic, just living alone with our ailments. How do we measure our progress in the in-between time? How do we measure our distance from health?
The answer to this question has become an obsession for data-driven scientist Paul Wicks. The neuropsychologist from the U.K. (and a 2013 TED Fellow) is research director for PatientsLikeMe, a social network that is producing some of the most compelling clinical data the health care industry has ever seen.
PatientsLikeMe was launched in 2004 by family members of an architect who had contracted ALS five years earlier at the age of 29. They had raised millions of dollars in a failed effort to find a cure for ALS and also ended up creating a data-sharing patient social network to go with it. Early on they brought Wicks over to Boston from London where he was moderating an ALS patient group and doing neuroimaging research out of King’s College. Wicks ended up being employee number 6 and never left.
PatientsLikeMe now has more than 200,000 patients on the platform and is tracking 1,800 diseases. The majority of them have neurological diseases such as ALS, multiple sclerosis and Parkinson’s, but it has been moving into AIDS and mood disorders. The Big One, cancer, is still a bit too complicated to tackle, with subtypes upon subtypes each bearing their own numbers and measures, but Wicks thinks cancer won’t be an impossible challenge forever. For the diseases it does track it is one of the more powerful tools out there. ALS patients, for example, are extremely rare in the wild (more rare than one would think given the name-recognition of the disease). But, says Wicks, “I can push a button and survey 200 ALS patients and get results in two weeks,” says Wicks.
Like a lot of people I thought patient Web groups were merely chat boards and self-help news distribution. “That’s the Internet of the 1990s,” says Wicks, dismissing those sites as “just-so stories.” Quantitative data is what he wants. “Qualitative data gets hijacked by people with an agenda, or extroverts or people who write well.”
PatientsLikeMe tries to get patients to quantify whatever they can about the diseases they have, scoring side effects like nausea and mood not usually captured as a number. It has designed 20 detailed questionnaires it circulates regularly to its members. Epileptics can enter their seizure information into a seizure monitor. It created a contrast sensitivity test with the Massachusetts Eye and Ear Hospital for people with Parkinson’s and the hallucinations that come with mood disorders. It’s working with a startup run by another TED Fellow Max Little who created a diagnostic test for Parkinson’s that involves calling a phone number and talking into a recording device to measure their voice’s shakiness. See his TED talk here. It has a survey tool to measure how closely patients adhere to their treatment regimen, but also scans language in the chat boards for words like “needles” and “no way I will ever do that again."
“I want an electric sensor on your head, in your toilet, on your pancreas, on your steering wheel and in your Google Glasses,” said Wicks. “Our goal is to build a learning health care system.”
PatientsLikeMe is also a strong advocate for unlocking patients' own data. In a new initiative being funded by the Robert Wood Johnson Foundation, PatientsLikeMe is building an open platform for anyone who wants to develop new outcome measures with patients. Those measures will be made freely available to anyone measuring the disease (including drug companies). With approximately $25 million in venture funding raised to date, there is plenty of capacity for the platform to grow, especially outside the U.S., which makes up 85% of its 200,000-patient pool. “There’s so much that is not being captured by medicine. We’re trying to build an ontology for what a disease can be."
