Theobio main page

Post last updated: 7 Dec 2021

Context

Every three weeks, our group provides COVID-19 infection report forecasts for the province of Ontario to the Ontario Modelling Consensus Table (MCT), a partner of the Ontario Science Advisory Table (SAT), which presents this information to the Health Coordination Table of the Ontario Ministry of Health.1 Other modelling groups also provide forecasts to the MCT. After reviewing all forecasts provided, the MCT provides consensus projections to the SAT.

Our forecasts are based on a compartmental epidemic model implemented in our publicly available McMasterPandemic R package, and involve statistical fits to Ontario’s latest infection report data.

Forecast overview

The forecast presented here was made on 20 Sep 2021 and is an updated version of the forecast submitted to the MCT for inclusion in the SAT’s update on COVID-19 projections posted on 28 Sep 2021.

This set of projections explores the potential effect of schools reopening the week of 7 Sep, as it is still too soon to glean any effect that this event may have had on transmission from the infection report data. We describe the transmission scenarios we considered in detail below.

We have updated, and will continue to update, this post in order to show the latest infection reports, but the forecasts themselves have not, and will not, be modified after the forecast date.

Forecast results

The following figure gives the infection report forecasts (curves), with 95% confidence intervals as bands around each forecast curve. Observed infection reports to which the model was fit are plotted with solid points, while observations after the fact are included as hollow points.

There is a high degree of uncertainty in these projections as they are based on a transmission rate calibrated during a relatively short period (three weeks in September), and this estimate is itself uncertain: there is a wide range of transmission rates that could reproduce the pattern in infection reports observed at the tail end of the time series.

The status quo scenario suggests a plateau in infection reports at a relatively-low-to-moderate level through the fall, though a further increase in transmission (due to children attending schools or increases in riskier behaviours) could provoke a steady climb in infections. If transmission decreases despite school reopening (e.g. if individuals curtail riskier activities to more than compensate for any transmission in schools), infection reports would be driven even further down from recent counts.

Forecast and model details

Transmission

We assume that the transmission rate is piecewise constant through the fit period, selecting re-estimation dates for this rate based on large-scale changes in public health measures (e.g. a stay-at-home order or a major reopening step) or when it appears individual behaviour may be changing. By assuming that the transmission rate is constant between re-estimation dates, we capture the average effect of any changes in transmission over the entire period between these dates. Recent re-estimation dates used in this forecast include the step 1-3 reopening dates.

Our forecast scenarios focus on exploring the potential effect of schools beginning to reopen on 7 Sep 2021. Since only 13 days had elapsed since school started reopening when we calibrated our model on 20 Sep 2021 (the model calibration date), we were not yet able to robustly estimate the degree to which reopening affected transmission: few transmission events that occurred after 7 Sep 2021 would have led to reported cases by 20 Sep 2021.

To explore possible effects of school reopening, we consider three scenarios for a change in the transmission rate on 7 Sep 2021 (the forecast scenario start date). Transmission rate changes are expressed as a percentage of the most-recently fitted transmission rate (calibrated over the month of September):

  1. as calibrated, i.e. a status quo forecast;
  2. a 15% increase in transmission;
  3. a 15% decrease in transmission.

Due to the delay in infection reports, differences between the forecast scenarios in the plot above do not become apparent until about two weeks after the forecast scenario start date (i.e. the projections begin to diverge around mid-to-late September).

Variants

We model the Delta variant as taking over Alpha through the fit period, under the assumption that Delta is 50% more transmissible than Alpha, based on variant surveillence conducted by Public Health Ontario. We continue to model any further increases to Delta frequency through the forecast period, though these changes are minimal as virtually all infections in the province have been caused by Delta since early August (see Figure 5 in the Public Health Ontario Daily Epidemiologic Report).

Vaccination

For vaccination, we use the reported number of first and second doses administered each day in the fit period. In the forecast period, we assume that first doses saturate to 90% of the eligible population (aged 12+ in 2021) and second doses saturate to 85%. (As of 21 Sep 2021, 86% of the eligible population had received at least one dose, while 80% had received two doses, according to COVID-19 Tracker Canada.)

The per-dose infection-blocking vaccine effectivenesses differ between Alpha and Delta, so we adjust them in the model depending on the daily estimated frequency of the Delta variant. The baseline effectivenesses used are:

First dose Second dose
Alpha 60% 90%
Delta 30% 80%


The vaccination model also includes the following assumptions:

  1. Vaccine-derived immunity takes an average of 14 days to develop after a dose is administered.
  2. If a vaccinated individual is infected, their probability of having a less-transmissible, asymptomatic infection increases with each dose (20% for the unvaccinated, 60% for those protected by one dose, and 90% for those protected by two doses).
  3. Vaccines are administered at random in the population (we do not explicitly account for age or other characteristics of individuals).

Related post: Ontario COVID-19 forecasts, 5 June 2021

Back to the MacTheobio COVID Modelling Group page

  1. This report has been written independently of the Ontario Modelling Consensus Table, the Ontario Science Advisory Table, and the Health Coordination Table. The views expressed in this report are solely the authors’.↩︎