Sensors increase ability to predict senior citizen falls
The Center for Eldercare and Rehabilitation Technology, led by Marjorie Skubic from the University of Missouri College of Engineering and Marilyn Rantz of the Sinclair School of Nursing, has provided several breakthroughs as part of its mission to help senior citizens stay in their homes longer via cutting-edge ways of monitoring their health.
The latest collaborative research developed sensor systems that can predict increased risk of falls by measuring gait speed and stride length, allowing health professionals to potentially intervene before a fall occurs. These findings were chronicled in the paper “Using embedded sensors in independent living to predict gate changes and falls,” which recently appeared in the Western Journal of Nursing Research.
Alongside Skubic and Rantz, the multidisciplinary team was made up of MU researchers Lorraine Phillips, Chelsea DeRoche, Gregory Alexander, Laurel Despins, Carmen Abbott, Bradford Harris, Colleen Galambos and Richelle Koopman.
A total of 23 residents of Tiger Place — an aging-in-place, 54-unit retirement residence operated by Americare and the Sinclair School of Nursing — participated in the study in intervals ranging from 3 to 48 months. The team used Microsoft Kinect sensors, which logged baseline gait speed and stride length data via a depth sensor, then alerted staff if those variables began to decrease. Researchers also noted falls, and analyzed the association between pre-fall deviations in gait speed and stride length and known falls.
“We have developed a non-wearable sensor system that can measure walking patterns in the home, including gait speed and stride length,” said Skubic, director of the Center and professor of electrical and computer engineering at MU. “Assessment of these functions through the use of sensor technology is improving coordinated health care for older adults.”
The data was collected over a period of 10 years, and what it revealed is that within three weeks of a one-week decline in walking speed of 2.54 centimeters per second, residents were four times more likely to experience a fall than a resident with no cumulative gait change or a minimal cumulative gait change. A gait speed decline of 5 centimeters per second was associated with an 86.3 percent chance of falling within three weeks, while an overall decrease in stride length was associated with a 50.6 percent chance of falling within the same time period.
What the National Institutes of Health-funded study indicates is that including these depth sensors in homes and connecting them to health care providers can help senior citizens stay in their homes longer, as health professionals would be able to intervene before a fall occurs. Falls are a serious risk for senior citizens, in certain cases proving deadly.
“Aging should not mean that an adult suddenly loses his or her independence,” said Rantz, Curators’ Professor Emerita of Nursing. “However, for many older adults, the risk of falling impacts how long seniors can remain independent. Being able to predict that a person is at risk of falling will allow caretakers to intervene with the necessary care to help seniors remain independent as long as possible.”
Using the Kinect as a depth camera allows depth images to be captured unobtrusively. In Tiger Place, the devices were typically located on a shelf above the door, with the computer that records the data located in a cabinet, out of sight. This unobtrusive technology also used algorithms to account for and exclude data from visitors and other residents to ensure accurate data for a single individual.
Additionally, the team recently received an award from Mather LifeWays Institute on Aging for sensor-based research. This research, published last year in Nursing Outlook, illustrated that using an integrated approach combining care coordination and sensors at Tiger Place, seniors are able to live independently much longer.
“When combined with care coordination, using the sensors with health alerts at TigerPlace increases length of stay in independent senior housing an additional 2.5 years from the national average (1.8) to 4.3 years,” Skubic said.
“The care coordination by nurses and social worker, even without the sensors allows people to stay there longer. So these people are already getting really, really good care compared to most,” Skubic said. “And the sensors add something on top of that pretty significantly.”