Researchers are confident they can put sensors in a home; train them to pick up on light, sound and temperature clues that precipitate difficult behavior in someone with dementia; and give a heads-up to caregivers that agitation is coming.
What they don’t know is whether giving suggestions to caregivers on how best to keep their relatives calm will relieve the stress of caring for a spouse or parent who is unable to communicate their distress.
“We might reduce agitation but still not reduce the burden of care giving,” said Martha Anderson, a professor at Virginia Tech Carilion School of Medicine and one of the investigators on a three-university research project that is looking at developing an in-home product to ease the stress for families caring for a relative with dementia.
“We are looking at caregiver self-efficacy and caregiver depression. What you find, because you follow someone for 60 days and this disease is progressive, is that even though the caregiver said, ‘Yes, I like the equipment; yes, I liked the recommendations; or yes, I was happy to have this trial in the home,’ they are still care giving,” she said. “And they are still burdened by that. We are not at the point of rescuing anyone.”
Electrical engineers from the University of Virginia and data-crunchers from North Carolina A&T State University have been working with Anderson and other gerontologists at Carilion Clinic’s Center for Healthy Aging to develop simple-to-use smartwatches, tablets and sensors that can be deployed to stop bouts of agitation that make it difficult to care for people with dementia.
Since agitation is one of the top reasons families place relatives in assisted living or nursing homes, the researchers are hoping to find a way to intervene before an episode escalates. They are now looking for current and former caregivers to meet with them, look at the technology and make suggestions to guide further development.
“One of the main research questions of this is how would we come up with a good intervention suggestion. Based on the data we have, what is the appropriate suggestion to notify the caregiver? So we are hoping to get some feedback from the focus group,” Nutta Homdee said.
Homdee is a graduate student at the University of Virginia and the lead author of an article recently published in Computer, a journal of the Institute of Electrical and Electronics Engineers, to share the results of the Behavioral and Environmental Sensing and Intervention system, or BESI. (It’s pronounced “Bessie,” as in an elderly aunt’s name.) The report said ambient environmental data can be used to predict upcoming dementia agitation with 81% accuracy.
BESI consists of sensors that are attached to walls in a home that monitor light, temperature, sound and air pressure. The system does not have cameras nor does it record spoken words, although it does monitor the volume and pitch of voices and other sounds. The person with dementia wears a watch that picks up on movements and also allows the caregiver to push a button when agitation occurs. The caregiver communicates with BESI through a tablet app.
During the first 30 days, the machine learns the conditions in the home, and the caregiver flags when agitation occurs. BESI then looks at the conditions in the minutes prior to agitation, and during the second 30 days, it flags similar changes to predict oncoming agitation.
“We start sending alerts to the caregiver to say, ‘Is this agitation or could this be agitation?’ And they can touch any button on the watch and say ‘yes,’ because we don’t want to distract them from taking care of the person,” Anderson said. “Sometimes they would say ‘I don’t know,’ and they would go to the tablet and read instructions.”
The tablet will offer suggestions based on what the caregiver has told the Carilion team and BESI.
“A simple recommendation would be, if the caregiver had said they like country music, is to try playing some soft country music. Or if we had down [that] the person really liked to go for a walk, we might send a recommendation that said, ‘Would this be a good time for a walk?’ ” she said. “We just try to use what was good for them in general so they felt someone was experiencing it with them.”
Anderson said they are still interviewing caregivers who participated in the trial to gather suggestions about what was helpful and what was not. The trial was small, with 17 families participating.
She shared what they’ve learned so far at the Alzheimer’s Association International Conference in July but said they need to talk with more caregivers to determine what is helpful.
They have funding from the Virginia Alzheimer’s Disease and Related Disorders Commission to host focus groups and recruit current and former home caretakers to evaluate the technology.
Homdee said the engineers hope to advance the technology so that it can predict agitation much sooner.
“We have confidence in detecting early agitation. The next step would be to actually predict it,” he said. “What we are doing right now is maybe five minutes before the agitation is really happening.” He said the goal would be to know the triggers for each person and be able to predict 30 minutes in advance that agitation will occur, which could allow for more effective intervention.
This article was written with the support of a journalism fellowship from the Gerontological Society of America, Journalists Network on Generations and the Retirement Research Foundation.
