Mathematical modelling is essential to develop a quantitative and predictive theory of the COVID-19 epidemic. Our team has been generating new understanding and making predictions based on rigorous mathematical modelling since March 1st. The majority of our work to date has been in response to urgent questions passed down to us from the provincial response committee, and as such has not been made public except in a highly summarized form. We believe that our insights and models will be extremely valuable to modellers working in other jurisdictions and that it is time to publish our work. Simultaneously, we believe that it is important to raise public awareness of mathematical modelling and its role in the epidemic response. We therefore also request funding for public outreach efforts in this direction.
Our models are parameterized via direct fitting to BC epidemic data (public data and additional information held at BCCDC), using international data (especially to understand the time course of disease in an infected individual) and additional information, e.g. around the effects of social distancing in BC such as transit usage, cell phone provider location data and Google location summaries.
To date we have developed two distinct models of the epidemic process in BC:
(1) An extended susceptible-exposed-infectious-recovered (SEIR) differential equation model with explicit social distancing. After careful parameterization, we have found that this model has similar predictive power to model (1), but it is far less computationally intensive, so we are now using it as the workhorse for short-to-medium term prediction.
(2) A large-scale, age- and activity-structured SEIR-type differential equation model, developed from a legacy model of pandemic influenza in BC, but significantly extended to reflect social distancing in the context of COVID-19. The main purpose of this model is to explore scenarios around the end of social distancing in BC.
Group Lead: Daniel Coombs (Mathematics)
Project Outcomes and Target Dates:
We intend to write manuscripts for publication in high-end medical / epidemiological journals. We expect to make the submission of project (1) within the next two weeks. Project (2) will also be developed into a journal article; the goal of a draft within 2-3 weeks and submission around May 31 seems achievable at present. Preprints will be posted at the PWIAS web site before publication.
We also wish to generate public-facing outputs from this work, and other work that may arise from our covid-19-modelling team. Specific outputs with dates are envisaged as follows. All items will be hosted by PWIAS web site:
(1) illustrated lay summaries of the two model papers in the form of blog posts (project 1: approximately May 15; project 2: approximately May 31)
(2) brief (10min) summary podcast for each paper to be released with the blog posts.
(3) illustrated public-accessible blog posts summarizing other outputs from this group.
Are you interested in receiving the Peter Wall Institute monthly newsletter? If so, click here.