Every year in Canada, approximately 28,000 babies are born preterm, before they reach 37 weeks of gestation. According to the Canadian Institute for Health Information1, preterm births are the leading cause of infant mortality in developed countries. Since more than one-third of brain growth and significant lung development occur during the final weeks of a normal gestation period, preterm births result in increased rates of respiratory distress, temperature instability, hospital readmissions and neurocognitive problems. It’s not surprising, then, that preemies pose a serious concern for our public health system — not to mention emotional distress for the parents.
But these parents will soon rest a little easier, thanks to the ongoing research of computer scientist Carolyn McGregor and her team at the University of Ontario Institute of Technology (UOIT), in Oshawa, Ont.
Known as the Artemis Project, McGregor’s research is helping to make sense of the reams of data collected from critically ill premature babies so that doctors in the future can better recognize subtle changes in the preemies’ vital signs. With this information, medical staff will be able to intervene earlier in the hope of reducing mortality rates.
The Artemis software takes in all the biomedical data gathered continually from monitoring equipment, such as heart rate, respiration and blood pressure, and aggregates the information in real time so that it can be analyzed and applied in a timely and clinically meaningful way. This is a significant advancement, because there is currently no way doctors can effectively make use of all the data that are largely now recorded on paper and read as discrete blocks of information.
Artemis has the ability to show trends and minute changes in physiological data, which can be particularly important for these fragile young patients. “We are specifically targeting the most life-threatening and frequent conditions that premature babies can develop,” says McGregor, who is also a Canada Research Chair in Health Informatics. These conditions include hospital-acquired infections. McGregor’s research shows that compared with current indicators, Artemis could detect subtle changes in heart-rate behaviour up to 24 hours sooner, which could make the difference between life and death for some preemies.
Earlier detection of illness can also mean a shorter hospital stay, thereby reducing health-care costs.
McGregor’s informatics work began in her native Australia, but it was in Canada that she was able to begin applying her informatics research toward eventual practice in a hospital environment. “The funding I received from the CFI has been paramount in enabling my innovative ideas to become reality,” she says. “I have been able to establish a truly internationally unique research environment.”
Employing a number of medical devices used in neonatal intensive care units (NICUs), McGregor and her team are developing new solutions for the analysis of real-time, high-frequency physiological data. In addition, McGregor has partnered with the Durham College/UOIT Simulation Lab to use its simulated newborn mannequin for testing.
McGregor’s groundbreaking research earned her an IBM First of a Kind Research Award in 2008, allowing her to merge her health-informatics endeavours with IBM’s stream-computing platform to create the Artemis Project. Artemis was first piloted in August 2009 at The Hospital for Sick Children in Toronto, where her team collected approximately 90 million points of data per patient every day.
With CFI funding, McGregor has also been able to purchase equipment to design a cloud-computing version of Artemis. Through that initiative, her team receives secure data from the Women & Infants Hospital of Rhode Island, in Providence, for processing in real time at UOIT. The Artemis cloud presents great potential for rural and remote populations, where specialists, such as neonatologists, are not always available. “Many of these premature babies need to be supported remotely via telephone until the transport team arrives and moves them to a NICU at another location,” explains McGregor. “Babies who are transported have higher mortality rates. If we can provide more advanced approaches to care and more information to the remote neonatologist, we have the potential to delay or even, in some cases, eliminate the need to move these babies.”
McGregor believes that we are “at the cusp of a new wave of health care and health-care research” and that the advanced high-frequency analysis of physiological data has the potential to revolutionize the health-care sector. As more patient data are collected over longer periods, she says, more progress will be made. Plus, the technology can be applied in other hospital environments as well as in patients’ homes. “From the foundation I have been able to establish here in Canada, we are in a great position to lead this voyage of discovery.”
Her passion for applying informatics to health care stems not only from the work itself but from personal experience. Thirteen years ago, McGregor had a premature daughter who did not survive, due to a rare chromosomal abnormality. “My husband and I lived the journey of grief of the loss of a child,” she says. “While my daughter’s life was not viable, there are many other premature babies who have a chance to survive. So beyond finding it an interesting research problem, I am personally motivated to use my skills to change the way information is used in the NICU for better outcomes for babies and their families. If we can save just one life, it will be worthwhile. I think we have the potential to save many more.”
1Canadian Institute for Health Information, Too Early, Too Small: A Profile of Small Babies Across Canada (Ottawa: CIHI, 2009).