W.Va.-Based Study Aims To Predict Coronavirus Exposure Early
A typical day at Ruby Memorial Hospital in Morgantown has its stresses.
Curtis Ash, nursing manager in the hospital’s emergency department, knows that well. He supports about 160 staff, who themselves are caring for sick patients.
But in mid-March, the reality of being a frontline health care worker at Ruby, and across the United States, shifted as fears of the coronavirus and cases began to trickle in.
“We’re just like many other hospitals around the world,” he said.
For Ash, that meant his staff was suddenly putting themselves at risk daily. It also meant stress levels among hospital staff were running high. To help protect themselves from contracting the virus, health care workers are donning extra layers of personal protective equipment, such as gowns, masks and gloves.
That was when researchers at nearby West Virginia University wondered if there was a way to use the physiological experiences of health care workers to help determine if they had contracted the coronavirus, possibly before they themselves knew it.
Scientists at WVU’s Rockefeller Neuroscience Institute partnered with smart ring maker Oura Health to craft a new study that aims to do just that.
About 150 health care workers in West Virginia were recruited to participate. They are wearing a smart ring – it looks like a wedding band – that collects data such as sleep patterns, oxygen levels and pulse rate. Participants also do twice-daily motor tests using a phone app. That data is fed into an artificial intelligence computer model developed by the Rockefeller Neuroscience Institute.
Dr. Ali Rezai, who serves as chair of the institute, said preliminary results show the model can predict when a health care worker is infected with the virus up to 24 hours before they show symptoms. Knowing who has the virus and when is essential to prevent future outbreaks, he said.
“We’re looking at a way that we can have an earlier way to detect if somebody is contagious with the COVID-19 virus, but is not symptomatic,” Rezai said. “The sooner we know this, the sooner we can forecast or predict who’s going to be exhibiting the symptoms, the more we can help in terms of limiting the spread.”
For the past two years, researchers at the institute have fed the data from 30,000 individuals into the predictive, AI model being used in this COVID-19 study. Broadly, the model looks to predict the “recovery, readiness and resilience” of humans in terms of their jobs or their current experience,” Rezai said.
For example, the model has been used to help predict if a person struggling with an opioid addiction would be more or less likely to have cravings or relapse.
Using the physiological data collected by the smart rings and daily motor tests, the model has been retooled to help predict the contraction of a virus such as COVID-19.
Researchers are working to expand the study to include hundreds of frontline health care workers across the country. Rezai said the study focused on this group of people not just because of they’re increased risk of getting the virus, but because of the personal toll this is taking on health care workers.
“Health care workers are under enormous stress and anxiety about exposure to a virus that there’s no treatment, there’s no vaccine for and very limited personal protective equipment,” Rezai said. “And they don’t know the risk themselves or to their families.”
Peace of mind for both her family and her patients was one reason Donna Tennant, admissions marketing director at the Sundale Nursing Home in Morgantown, decided to participate. Sundale has been hit hard by the coronavirus. Data from the West Virginia Department of Health and Human Resources shows as of May 10, 39 residents and 15 staff have tested positive. Five people from the facility have died.
“We're working it every day,” she said. “So, being able to see your levels [and] what's happening, it gives you a little bit of a comfort knowing that you know, ‘yes, I'm okay right now.’”
Appalachia Health News is a project of West Virginia Public Broadcasting, with support from Marshall Health and Charleston Area Medical Center.