The Reason Smartphones Are Providing Better Insights For Medical Researchers

People are far more comfortable responding to medical survey requests by smartphone than by traditional methods.

This has led to more data of better quality (broader demographics, fewer inherent biases, more detail) which means the knowledge gained is more broadly applicable, more accurate - and for providers, more profitable. JL

Charles Wallace reports in the Wall Street Journal:

Studies confirmed that surprisingly large numbers of people can be recruited into long-term research studies on mobile phones; that their consent to participate can be obtained more easily than in conventional studies; and that the medical data obtained can be made anonymous, collected and analyzed by advanced algorithms in ways never before imagined. (An) app used an algorithm (that) can detect Parkinson’s from a digital voice file with 98% accuracy.
When smartphones first began collecting health data such as users’ heart rates and number of steps walked, doctors were dubious about the medical value of information gathered by a phone.
Three years later, doctors have changed their minds, thanks to a series of pioneering medical studies that demonstrated the efficacy of cellphone-based medical research.
The studies have confirmed that surprisingly large numbers of people can be recruited into long-term research studies on mobile phones; that their consent to participate can be obtained much more easily than in conventional studies; and that the medical data obtained can be made safely anonymous, collected and analyzed by advanced algorithms in ways never before imagined.
Indeed, so much data can be collected automatically and accurately via mobile phones—without participants or lab workers having to log it—that some scientists believe it will be easier to conduct and monitor many trials involving drugs or exercise in larger populations than have been examined up to now in conventional studies.What’s more, doctors believe it will be possible to give participants feedback not only about their own health but also about the population at large much faster than is possible with conventional medical studies, which often appear in scientific journals years after they’re conducted. Participants could quickly see, for example, the beneficial effects of exercise or diet and adjust their behavior accordingly.
Most of the studies were made possible by an open-source smartphone platform called ResearchKit, which was released free to researchers by Apple Inc. in 2015. Doctors and scientists can use it to create iPhone apps that collect data such as weight and blood pressure that are stored in the Health app on iPhones.
Because many more cellphones, especially in less affluent areas and in developing countries, use the Android operating system, an Android version of the platform called ResearchStack was created by scientists at Cornell University. Sensing the utility of these devices, Alphabet Inc.’s Verily Life Sciences subsidiary recently unveiled a watch, called the Study Watch, that will collect data on users’ heart rate and movement for studies of a number of diseases.
Initially, five studies were launched in 2015 using ResearchKit, covering asthma, breast cancer, cardiovascular disease, diabetes and Parkinson’s. Studies involving epilepsy, autism and moles were added later, and a new Harvard Medical School study to monitor the health of National Football League players is just getting under way.
Researchers have long worried that clinical studies in general are skewed because most are limited to areas around research hospitals, so the study group isn’t a representative cross-section of the population. An even greater concern is that, except for the controlled conditions of laboratory research studies, most people have been at best lax about reporting health issues or downright dishonest about, say, how much exercise they get. The smartphone, on the other hand, can track activity and other health-related data passively and dispassionately.
Following publicity about the studies, anyone could download the apps associated with the research. In nearly all the studies, thousands of people responded, many times more than the numbers for traditional studies. In all, the initial five studies involved a total of more than 50,000 people, including much larger cohorts from all 50 states than traditional research studies.
Participants were asked a series of questions that helped determine their health status, which is similar to conventional studies; gave formal consent to being studied and sharing their data on the phone; and began using the phone’s sensors to record things like how much activity they had each day. The data was collected and distributed to researchers by Sage Bionetworks, a Seattle health-tech firm that developed the sharing and security protocols for the studies.
Following publication of the research studies this year, a number of researchers said they received inquiries from pharmaceutical firms interested in how they could use smartphones for clinical trials for drugs. “There is huge interest in the studies from pharma, who want us to help them deploy this kind of technology for clinical-trial populations,” says Eric Schadt, director of the Icahn Institute for Genomics and Multiscale Biology at Mount Sinai Hospital in New York, who worked on the asthma study.
In fact, British drugmaker GlaxoSmithKline has become the first pharmaceutical company to carry out a medical study using ResearchKit, tracking 300 rheumatoid-arthritis patients’ fatigue, mood and joint pain over three months with data from iPhones. It hasn’t yet released the results.
Ray Dorsey, a Parkinson’s disease researcher at the Center for Human Experimental Therapeutics at the University of Rochester, says the Parkinson’s study conducted by iPhone was a breakthrough not only because the researchers obtained consent from more than 9,500 patients over the phone, but also because the study’s iPhone app, mPower, used four innovative tests to gauge the extent of the disease: It required the Parkinson’s patients to tap the phone’s screen as fast as they could; walk 20 steps, turn around, wait 30 seconds and take 20 steps again; take a memory test; and speak into the phone. Among other analytical tools, the app used an algorithm developed at Oxford University that can detect Parkinson’s from a digital voice file with 98% accuracy.
“We’ve come up with a scale that’s objective, continuous and can be done by anyone, anywhere and can provide a wealth of information on how people’s symptoms vary, and use that to help manage their condition,” Dr. Dorsey says.
While all the phone-based studies were initially limited to the U.S., a number have reached overseas, such as the cardiovascular study, which is being rolled out in the U.K., Hong Kong and the Netherlands. One aspect of phone-based studies that researchers are trying to improve on: Many people drop out. The studies still end up with many more participants than traditional studies, but researchers would like to find a way to retain more of the people who initially sign on.
Euan Ashley, a researcher at Stanford University who worked on a cardiovascular study called MyHeart Counts, says about 5,000 people did the six-minute walk the app called for, the standard for measuring heart health. That was more than five times the number of people ever studied in a single group taking that test before. But more than 40,000 people who signed up for the app dropped out of the study before they got to the walk.
“The bar for entry in these studies is very low,” Dr. Ashley says. “So there is less adherence than in a clinical setting.”
Researchers say they are looking at ways to make the apps more engaging over time, such as giving users not only their own data, but also comparisons to other users in the same age or demographic group.