Discoveries
A new way to app-raise seniors' health
Using "apps," applications created for smartphones, it's possible to count calories, interpret a baby's cry, or whistle for a dog. Now, DMS researchers are developing an app to monitor the health of senior citizens.
The standard way to evaluate physical and mental well-being is through surveys. But self-evaluations are subject to bias from patients' imperfect memories and from their desire to give favorable responses, says epidemiologist Ethan Berke, M.D. So he set out to create a better way.
In 2005, while at the University of Washington, Berke teamed up with Tanzeem Choudhury, Ph.D., a specialist in designing machines that measure physical and social behaviors. Their first attempt resulted in a fanny pack filled with sensors and batteries. By 2009, Berke and Choudhury, by then both at DMS, had developed a less obtrusive device—a two-inch-by-two-inch mobile sensor—that they tested in a pilot study.
Sensors: For 10 days, eight residents of a local retirement community wore sensors clipped to their waists. The sensors recorded subjects' activity level, including the time they spent walking or running, as well as their social interactions, including the number of conversations they engaged in and the pitch and volume of the participants' voices (the actual conversations were not recorded). Then the researchers compared the data gathered by the sensors to the subjects' responses on surveys of their physical and mental health.
Social: The results, published in the Annals of Family Medicine, were compelling. On measures of social and mental well-being, the data from the sensors correlated closely with the survey results. Subjects who spent the most time talking to others—as measured by the sensors—reported on the surveys being happier and more socially connected.
It runs on Android phones and was 80% accurate in a pilot.
On measures of physical activity, the sensor data didn't correlate quite as closely with the survey results. A likely explanation, Berke says, is that subjects had trouble recalling past physical activity.
But with the general relationship established, Berke and Choudhury began seeking a way to incorporate the sensors into smartphones. With help from Dartmouth computer scientist Andrew Campbell, Ph.D., they created an app called BeWell that monitors physical activity, voice acoustics, and sleep patterns. It runs on Android phones and gives real-time "well-being scores." The app proved 80% accurate in a pilot study. Before making it available to the public, Berke plans to test it on 100 older adults at a health center in Manchester, N.H.
Avert: He envisions a day when the app can help avert physical and mental health problems. Often, he says, patients are given mental-health surveys only after they show symptoms of severe depression. "If people were wearing these devices, we might have the ability to detect changes in mental-health status," he says, "which could result in earlier diagnosis and . . . treatment."
If you'd like to offer feedback about this article, we'd welcome getting your comments at DartMed@Dartmouth.edu.
This article may not be reproduced or reposted without permission. To inquire about permission, contact DartMed@Dartmouth.edu.