Long-term care staff is faced with challenges when it comes to the care of residents with severe dementia. One of these is pain detection, assessment and monitoring, which can be quite difficult, depending on the resident. One study aims to change all that by helping long-term care staff with pain assessments through technology. "Pain is highly prevalent in old age," notes Dr. Thomas Hadjistavropoulos, co-lead of the AGE-WELL research project along with Dr. Babak Taati. "According to some estimates, as many as 80% of those who live in long-term care have persistent pain." And those with dementia are often under treated for their pain or treated incorrectly. "Seniors with advanced dementia often have limitations in their ability to communicate," says Dr. Hadjistavropoulos. "This includes limited ability to express pain verbally." Because they are under assessed, staff members sometimes doesn’t even know there is a problem and severe problems such as abscessed teeth and even fractures can go undetected for days or even weeks. "One study showed [dementia patients] were six times less likely to receive an analgesic medication," points out Dr. Hadjistavropoulos. "But more critically, when people with dementia have pain and cannot communicate, they often end up showing behavioural disturbance such as aggression." What happens then, Dr. Hadjistavropoulos notes, is that staff attributes that aggression to a psychiatric issue rather than pain. "So instead of treating that problem with analgesics, they treat it with psychotropic medication."

The Study "My work for many years, before AGE-WELL, focused on the evaluation of pain behaviours like specific facial gestures," says Dr. Hadjistavropoulos, a health psychologist who holds a Research Chair in Aging and Health at the University of Regina. "Over time we developed various methods for evaluation of pain in people with severe dementia including an easy to use checklist for clinical staff who can then record the number of pain-related behaviours and produce an estimate based on those about how severe the person's pain is." Dr. Hadjistavropoulos was then turned on to the idea that this same pain assessment could be aided by the use of technology. The study started with a CIHR grant and then they were funded by AGE-WELL which Dr. Hadjistavropoulos notes "expanded the scope substantially." "The aim of the study is to develop an easy-to-use, relatively inexpensive computer vision system that could be installed in nursing homes and would be able to detect and monitor pain behaviours in the residents as they go about their daily routines," says Dr. Hadjistavropoulos. How the system works is revolutionary. "The system will be able to detect pain behaviours as [residents] go about their daily routines," notes Dr. Hadjistavropoulos. "When a certain threshold of pain behaviours has crossed, the system will then be notifying the nurses’ station so that they know which resident to follow up with."What the study researchers are working on now is making this technology affordable and easy to use in the long-term care environment. "The technology is already possible," Dr. Hadjistavropoulos points out. "But it only works under ideal circumstances; with high quality, HD cameras facing the person directly. And it's not clear whether it works on an older face with wrinkles." Instead, they want to be able to make it work so that pain behaviours can be captured while residents go about their daily business. This means developing an algorithm that not only takes an older face into account but side angles of the camera as well. This technology is aimed at helping staff with their jobs. "What this allows us to do is it makes it automatic and it addresses a very serious problem which is a resource issue in long-term care," says Dr. Hadjistavropoulos. "It's not that nurses can't do a [pain assessment] themselves. They can but assessments cannot be done often enough because of resource constraints." So how is this technology going to be achieved? It takes a lot of research. "One of the things we've done over the years is consulted literature that has identified facial responses that are most likely to occur during pain," notes Dr. Hadjistavropoulos. "It's important to differentiate between a pain state versus a non-pain state. But certain behaviours are very likely to occur during pain and if you have enough behaviours occur at the same time, you have a very high likelihood that pain is present." In terms of progress, Dr. Hadjistavropoulos notes that his team in Regina have collected all the clinical data they need from nursing homes. This meant staring at video screens, and then manually coding different non-verbal pain behaviours, thousands of frames worth. The data collected in Regina is now being used to develop an algorithm. This is the job of Dr. Taati’s engineering and computer vision team at the Toronto Rehabilitation Institute-University Health Network. Once a prototype is built, the plan is to evaluate it in two long-term care homes. And the research team has taken things like privacy into account. "We're very sensitive on the privacy concern," Dr. Hadjistavropoulos says. "The cameras will not be recording videos. They will be feeding the computer with what's going on but nothing will be recorded other than whether a pain expression has occurred." So what does Dr. Hadjistavropoulos and his team hope to achieve? "The ideal is that we will be able to detect pain and monitor pain in a way that allows for early, effective and appropriate treatment," states Dr. Hadjistavropoulos. "The result would be less pain-related suffering and better quality of life for residents. We are also hoping to see substantial cost-savings for the health care system through earlier detection and treatment of pain-related problems." He also notes that it would help with staff stress. "We did a study that showed that when pain is assessed on a regular basis, staff stress goes down, possibly because residents with dementia become less likely to show behavioural disturbance." This technology will positively impact long-term care in many ways but most importantly, will ensure that residents with severe dementia will be able to be properly assessed and treated for pain.

SIDEBAR AGE-WELL is a federally-funded Network of Centres of Excellence established in 2015 to support Canadian research and innovation in the area of technology and aging. AGE-WELL is dedicated to the creation of technologies and services that benefit older adults and caregivers. Its aim is to help older Canadians to maintain their independence, health, and quality of life through technologies and services that increase their safety and security, support their independent living, and enhance their social participation.