COVID-19 Research Q&A Series: Dr. David Fisman

July 6, 2020 | Author: Royal College Staff
2 MIN READ

Dr. David Fisman and team are using math and statistical modelling to gain a better understanding of the COVID-19 epidemic. Learn how other countries are using Canada’s COVID-19 modelling.

Dr. David Fisman, MD, FRCPC, is an epidemiologist and professor at the Dalla Lana School of Public Health, and a practising physician at the University Health Network in Toronto. Dr. Fisman received a grant from CIHR’s 2019 Novel Coronavirus (COVID-19) rapid research funding competition to gain a better understanding of the pandemic.

Dr. David Fisman

Dr. David Fisman

What is the context of your current work, and what did you set out to accomplish?

Our team of doctors, epidemiologists, public health professionals, and statisticians has deep experience responding to past outbreaks like SARS, H1N1 and Ebola. In this project, we’ve been using math and statistical modelling to do three things: forecast the near-term course of the epidemic; get a better understanding of the parts of the epidemic that are harder to pinpoint because they’re messy or noisy; and build simulations that can help guide Canadian health agencies as they try to control or limit the spread of COVID-19 in Canada.

Where are you in the research?

We published a near-term forecast in March that I think helped scare decision-makers into instituting a lock down. So, that was effective and certainly limited the spread of the virus. Now, we’re producing a forecast each morning that we send to colleagues in Ontario and also nationally. We operate outside government, so our fresh set of eyes on the numbers is appreciated by policy makers.

When you refer to “messy” data, what do you mean?

Nearly every aspect of a pandemic is messy. If cases surge in Ontario, we need to figure out why. Some interpretations will point to a lack of testing; others will say it’s because people got together in larger groups on Mother’s Day. Our job is to do a triangulation of data from different sources, looked at through different lenses, and use various math tricks to dig down into the numbers and figure out what’s driving them. For example, we do think that Mother’s Day get-togethers were important in puffing up Ontario’s reproduction numbers the week of May 25.

What are you learning from other countries’ experiences, and vice versa?

We have a web-based version of our model now, and we’re comparing notes with infectious disease epidemiologists in Belgium, Ireland and Korea. The Irish are using our model to examine their health system capacity, and I’ve spoken with the Prime Minister of South Korea about our forecasting models. We also had an international symposium with Deputy Prime Minister Chrystia Freeland, who asked excellent, highly intelligent questions. There’s lot to be proud of with Canada’s work on COVID-19.

What is around the corner with COVID-19?

First of all, if anyone tells you they’re certain about anything to do with COVID-19, they’re lying to you or to themselves. With that in mind, it’s pretty clear that we’re looking at doing things differently for a while. Even with better numbers in the summer, we’re confident that the R number will get juice in fall and winter. So, we’ll be walking through a minefield for a number of months.


Tags


Leave a comment. To reduce spam, comments are reviewed before they are visible on the post.

Your email address will not be published.

Submit

Lorna Schiralli | August 2, 2020
You say==>Our team of doctors, epidemiologists, public health professionals, and statisticians has deep experience responding to past outbreaks like SARS, H1N1 and Ebola. In this project, we’ve been using math and statistical modelling to do three things: forecast the near-term course of the epidemic; get a better understanding of the parts of the epidemic that are harder to pinpoint because they’re messy or noisy; and build simulations that can help guide Canadian health agencies as they try to control or limit the spread of COVID-19 in Canada.===> may I ask who are the statisticians behind the modelling and who are the programmers for the statisticians?