AI Infrastructure for Quicker Health Policy Decision-Making Earns Prof. New Funding
By Françoise Makanda, Communications Officer at DLSPH
A DLSPH researcher is using artificial intelligence to greatly speed up knowledge synthesis in health policy – allowing decision-makers high-quality data in just days, rather than many months.
Prof. Andrea Tricco has just received funding from the Canada Foundation for Innovation’s (CFI) John R. Evans Leaders Fund (JEFL) for this work, which will help health systems to react to change in real-time.
“Knowledge synthesis provides evidence. The totality of evidence can be incredibly informative for decision-making. But to do it at a very high quality takes 12 months. That’s not enough time for a decision-maker,” says Associate Prof. Tricco who is also a researcher at Unity Health Toronto and Tier 2 Canadian Research Chairholder.
Knowledge synthesis pools a large body of research together and conceptualizes information for decision-making. What would take a year can be brought down to 14 days, says Tricco, with the addition of AI within the dashboard.
Her team had successfully brought down the process to three months but with AI and funding from the CFI, the “CAL-Synthesis.SR” Dashboard can take the time down to two weeks— and it will be available for all researchers. Ambitiously, Tricco would like to shave off another four days.
Quicker decision-making means fewer resources spent on essential lifesaving decisions for Canadians. And, rapid evidence will enable better prevention and treatment decisions for health leaders, contribute to sustaining high-quality healthcare and promote high-value innovations in Canada.
“Your typical knowledge synthesis takes months to years,” she says. “And this reflects literally millions of research studies that search hundreds of studies examined by human reviewers. COVID-19 has really shown us the importance of urgency.”
Tricco is the nominated principal investigator on a pan-Canadian team that has worked on providing information to decision-makers. The team has supported the Public Health Agency of Canada, The WHO and Health Canada among many others.
“The feedback that we receive is that the results have been useful and relevant within their decision-making window of opportunity,” she says. “Otherwise, the information will come too late, and they wouldn’t be able to make those decisions.”
The DLSPH alumna has been working in knowledge synthesis for the past decade. Her goal is to ensure this toolbox is used widely among health researchers. Her current work sources data from existing and large databases like PubMed and Embase to expedite knowledge synthesis.
“We’ve been using some semi-automation for some of our COVID work,” she says. “But we want to develop this tool even further. And so that’s where that CFI comes in. And that’s where we want to use our toolbox with AI and make this available for all researchers so that we can continue providing evidence that’s helpful for decision-making in a very timely way.”
Tricco has had a string of awards this year. In addition to shortening turnaround times, she would like to include patient partners as part of the synthesis to further the evidence’s quality.
“We’re hoping that this will produce valid and meaningful evidence for Canadian decision-makers and patient partners. So even with those ten-day timelines, we have patient partners who join the research team, which I think is pretty innovative and also leads to more impact.”