The pandemic’s toll on our healthcare system has been huge. Surgeons have performed 560,000 fewer procedures than usual. Nursing staff are facing burnout from working overtime to manage surges, and 77 percent of those who have dealt directly with COVID-19 patients reported their mental health has worsened.
As the rate of Omicron infections ebbs, vaccine maker Moderna says it expects COVID to become endemic sometime next year. But public-health experts have warned that the virus will still be deadly, with unpredictable waves of outbreaks. Even once the virus becomes endemic, we’ll need better ways to contain it and bring relief to healthcare staff.
Since we’ll have to live with this virus, the driving question is clear: how do we ensure that our healthcare system can cope in the era of ever-present COVID? New technologies are providing some clues.
“The two tools that proved to be powerful in response to COVID-19 were general surveillance — so sequencing of the viral genome — and rapid international data sharing,” says Marc Fiume, CEO of DNAstack, a Toronto health technology company. “You need both and, at the start of the pandemic, we didn’t have infrastructure for either.”
Scientists look at mutations in the virus’s genome, particularly on the infamous spike protein, to identify new variants of concern. During the pandemic, countries including Canada rapidly expanded their facilities for mass-sequencing of COVID samples, and that capacity will need to be retained to some degree. But Fiume says scientists still need more efficient ways to share data — it sometimes takes months for sequences to be uploaded into global repositories.
DNAstack has created Viral AI, a system for international data sharing, in which each contributor retains ownership and control of their data sets, but partners can query the whole of it. “It’s a faster, more scalable model for how we might share genomics and health data,” says Fiume. A better picture of what’s happening globally should enable faster detection of variants and give vaccine manufacturers a better sense of what to target.
If data-sharing platforms were paired with monitoring of wastewater and airborne pathogens, it could catch signs of danger much earlier across a range of diseases. It could be, Fiume says, “a sort of digital immune system to keep us safe going forward.”
In normal times, between 10 and 12 percent of patients in the general internal medicine unit at Toronto’s St. Michael’s Hospital are considered at high risk of dying or needing intensive care. With COVID, that number doubled, putting pressure on staff to quickly identify those in greatest danger.
“Often, by the time we realize that somebody is crashing, we only have a few hours to react and that’s not a lot,” says Muhammad Mamdani, vice president of data science and advanced analytics for Unity Health.
In October 2020, St. Michael’s started using an algorithm to monitor patients’ health data. The algorithm can assist doctors to predict who will need intensive care with about 15 percent more accuracy than clinician judgment alone. The resulting foresight allows at-risk patients to be monitored more closely for warning signs like developing sepsis. Mamdani says preliminary data suggest this has resulted in 15 percent fewer deaths among high-risk general internal medicine patients.
If future COVID peaks coincide with busy periods like winter flu season, greater use of algorithms to flag critical information could save vital time for overstretched doctors and nurses. And better forecasting of prognosis would enable hospitals to allocate acute-care patients more evenly among nurses, which can go a long way toward preventing burnout.
While the algorithm is currently only operating at St. Michael’s, Mamdani says other hospitals have expressed interest.
The global scientific community has been forced to learn about COVID at incredible speed — and there is no slowdown in sight. In February, more than 100 potential COVID treatments were in the late stages of clinical trials. And there will likely be more to come as researchers investigate the causes of a multitude of long-COVID symptoms.
“Over the next number of years, we’re going to have new variants to think about and new information about how different patients are reacting to different kinds of medication,” says Sachin Aggarwal, CEO of medical-technology company Think Research. “We’ve got to keep on top of that.”
Think Research has created software that incorporates the latest scientific evidence into decision-support tools for healthcare providers so that best practices can be accessed while treating patients. The software is integrated with a hospital’s medical-record system, and walks a nurse or doctor through a checklist of steps related to lab tests, diagnostic imaging and treatment.
Aggarwal says tools like this will be needed so that clinicians don’t spend time second-guessing how to treat a patient or ordering unnecessary tests in cases where the science is settled. This will be particularly important as COVID drops from the headlines and new evidence becomes less present in the media — and yet, getting that information into the hands of clinicians quickly will still be as important as ever. Without some help, there’s a risk of patients slipping through the cracks.
Aggarwal says information-sharing technologies like these create the infrastructure to deal with the next wave or next pandemic. “Because we can do better next time.”
Canada is experiencing its greatest innovation boom ever. MaRS sits at the fulcrum of that progress. Read our latest to Impact Report to learn more.