Originally published in the Chicago Tribune
After testing positive for COVID-19, Angela Mitchell initially felt OK. As an employee of UI Health in Chicago, she went on home quarantine, but agreed to let a nurse monitor her vital signs remotely. Mitchell wore a small sensor that looks like a bandage to track her heart, breathing and activity rates.
Looking back at that decision, she said, “I think it probably saved my life.”
The remote monitoring system that helped save Mitchell is now the subject of a clinical trial that aims to test the device on 1,600 patients. Doctors believe it could be an early-warning system to catch COVID-19 symptoms before they become severe, so they can be treated much more quickly and successfully.
The study pokes at the basic mystery of COVID-19, which is why are some patients fine and others die.
The sensor is sensitive enough to tell if someone is sleeping and on what side, said Karen Larimer, lead investigator for Chicago biotech firm physIQ Inc. By using machine learning, the researchers develop an algorithm to find which combination of variables sounds the alarm that the patient is getting sicker.
“It’s figuring out what is the fingerprint of COVID that begins to show us something is going wrong,” Larimer said.
Once the study results are published, the data will be available for public research on a National Institutes of Health database.
In the case of Mitchell, the 58-year-old grandmother has asthma, diabetes and high blood pressure, which put her at higher risk of severe complications from the coronavirus. By the third day after her diagnosis, she was coughing, sneezing and having trouble breathing. But she didn’t think she was in need of immediate medical attention.
In the middle of the night, she got a call from the nurse, who said the sensor showed her heart rate was alarmingly high at 200. The nurse urged her to get medical help. Mitchell went to see her doctor that morning, and was immediately admitted to Northwestern Memorial Hospital.
She was treated with intravenous medication and oxygen, but got worse. Doctors warned she may have to go on a ventilator and had her sign papers to allow that in case she crashed, since her family could not be with her.
“I was very afraid,” she said. “I’ve never felt so out of control of myself. I’ve never even thought what would happen, how close can I be to losing my life?”
She was also angry, because she got sick despite wearing a mask, washing her hands and keeping her distance.
After a doctor administered a drug cocktail that knocked her out for about 24 hours, she awoke and started feeling better. She was released after seven days, but it took her five months to recover enough to go back to work. She still gets headaches, clouded memory and other lingering symptoms, and does exercise therapy for her heart on bicycles and treadmills.
Her three adult daughters live with Mitchell and her husband in south suburban Lansing. Her oldest daughter lost her job due to pandemic closures, so her whole clan of six grandchildren shares the house too.
“The house has been turned upside down by COVID,” she said. “We’re family, so we’re handling it.”
Based on experiences like Mitchell’s, the remote sensing program is very promising in the eyes of Dr. Terry Vanden Hoek, chief medical officer at University of Illinois Hospital in Chicago. The sensors seemed to catch COVID-19 when vital signs were in danger but before patients felt seriously ill, so they could be warned to see a doctor right away. With early treatment, the patients usually recovered much faster, Vanden Hoek said.
The sensors are held in wireless, flexible stickers like bandages affixed to the chest, and transmit to a supplied cellphone, which uploads to a cloud database monitored by an advanced practice nurse. Mitchell said her sensor didn’t bother her.
As head of the emergency department, Vanden Hoek said he often sees patients who come in with unbelievably low numbers like 70% oxygen saturation, but weren’t even aware that they were seriously ill.
“If we can figure out how to identify this early, we could prevent hospitalization, intensive care or the use of ventilators,” Vanden Hoek said.
Hospitals often ask COVID-19 patients to monitor their blood oxygen levels at home after being discharged. This study asks patients to report their oxygen levels with a pulse oximeter, but researchers believe the heart, respiratory and activity rates are better indicators of illness. The resulting data will be kept anonymous and used to explore differences in how COVID affects different ages, genders, races and ethnic groups.
Late last year, the hospital enrolled about 400 COVID-19 patients in an initial study of the sensors. The National Institutes of Health is sponsoring a trial to enroll another 1,200 in a phase two study called DECODE. Because COVID cases are down, the study may be expanded to other hospitals in multiple states.
Patients 18 and over are monitored for 28 days after a COVID-19 diagnosis or after they are discharged from a hospital with COVID-19.
The goal is to figure out which combination of vital signs at what levels are the best warning signs to predict the course of the disease. About 10% of the initial patients had complications severe enough to warrant an alert.
Vanden Hoek hopes that someday the technology might be used to monitor people with other medical conditions, such as those at risk for heart failure.
“It seems like a very promising technology,” he said. “Usually when we see people getting into trouble, they’re much sicker than they realize.”
Nearly half the participants so far are Latino, and about one-third are Black. Including its Mile Square Health Center clinics on the South and West sides, UI Health serves some of the most vulnerable communities in the state. Sponsors hope to enroll more Latino and Black participants to find ways to predict downturns, and prevent the pandemic from affecting them so severely.