Big Data’s Powerful Effect on Tiny Babies
By Michael Beyman
It's a familiar sound found throughout hospitals around the world. Medical monitors that gather information about patient's heart rates, pulse, breathing and other bodily functions are tools as commonplace as defibrillators and stethoscopes.They allow doctors to accurately check a patient's vitals and successfully chart progress over time.
But in 2009, Carolyn McGregor, the research chair of health informatics at the University of Ontario's Institute of Technology, believed those monitors had much more potential that could be tapped. A former banking and retail consultant, McGregor was an expert at analyzing large samples of data and finding patterns. McGregor wanted to know if there are patterns in the data the monitors collect.
A doctor "showed me what he uses when he went on rounds, a piece of paper that had one number every hour for the heart rate, a number every hour for the respiration and one number every hour for blood oxygen," McGregor recounts. "And I said 'but what about those numbers that are flashing now?' And I thought, with my background, with all these things we were doing to watch shopping behavior, why can't we do the same thing with these monitors. Surely there was a message we were missing."
McGregor and her team turned to the monitors that were keeping track of some of Canada's smallest patients: premature infants held in the neonatal intensive care unit at The Hospital for Sick Children in Toronto. Those babies are at high risk for infection, according to Dr. Andrew James, associate clinical director of the hospital's neonatal intensive care unit.
"Infection is a constant worry," Dr. James explains. "About 20 percent of low-birth-weight babies develop an infection, and of those babies about 18 percent actually pass away. So it's very serious—and very common." Using previous research as a guide, McGregor and James sought out to see if there were trends in premature infants' vital signs that could be an accurate predictor of infection.
Enter the Artemis Project. Named after the Greek goddess who protected babies and childbearing women, the project sought to synthesize the millions of data samples coming from the neonatal monitors to see if there were any notable patterns.
"We looked at the heart rate for trends; when babies become infected, the baseline heart rate tends to increase," Dr. James says. "What I've learned is that when you look at more data and you look at it in a more granular sense, not only is there more to be seen, but you actually see more. We're beginning to see abrupt changes in heart rate variability and that really makes us think of infection."
These changes in heart rate are usually a telltale sign of infection. "With inflammatory response, there are chemical signals to the brain that affect heart rate," Dr. James explains. "And—when there's an infection, heart rate variability decreases."
Today, Artemis allows doctors to see early warning signs of infection upward of 24 hours before the baby would have otherwise shown any other symptoms. "So we have a means potentially to intervene earlier," McGregor says, "so these babies aren't suffering the same impact of infection."
The Artemis Project's success at The Hospital for Sick Children has led other hospitals to join the study. At Women and Infants Hospital in Providence, R.I., Artemis is being used to help study how their neonatal intensive care unit has affected their premature patients' mental and neurological development.
"There are a wide range of problems that all of this data will help us to understand," says Dr. James Padbury, chief of pediatrics at women and infants. "We can use it to understand the clinical problems they're having, look at the best ways to intervene, and see if those interventions are best."
"The data are extraordinary," Dr. Padbury adds. "It really allows us to add to the kinds of neurological assessments that we're making with our hands and our eyes, it allows us to make a more comprehensive assessment of the child. The data should significantly improve the quality and the type of care we provide."
For McGregor, her research into improving the lives of premature infants hits close to home. "My own daughter was born premature. She had a very rare chromosome abnormality and she passed away," says McGregor. "And in my daughter's case, no matter what we discover and what we do with the type of research I'm doing now, it would never have saved her."
"But it made me realize that there's so many babies in units that like this that have the potential for a wonderful outcome, to go to school, to go and get married, and to go off and live their lives," McGregor says. "I wanted to do everything that I could as a mother, as a computer scientist to change the outcome."