Update on COVID-19 in

Canada: Epidemiology
and Modelling
April 23rd, 2021

Canada.ca/coronavirus
 National disease and severity indicators have increased considerably
over the past month
Number of cases, total in On average over the past 7 days: Number of deaths
hospital and ICU 8,444 cases
10,000 4,167 hospitalizations 900
9,000 1,268 ICU admissions 800
46 deaths
8,000 700
7,000
600
6,000
500
5,000
400
4,000
300
3,000
2,000 200
1,000 100
0 0
Mar 20 May 20 Jul 20 Sep 20 Nov 20 Jan 21 Mar 21

Cases Hospitalizations ICU Deaths

Data in figure as of April 20, 2021

Note: Trend lines reflect 7-day moving averages. Total hospitalizations and ICU admissions include all people in hospital and in ICU on that 1
day. The average length of stay in hospital is approximately two weeks.
 For the first time in many weeks, national Rt has fallen below 1

1.5 When Rt is consistently >1,

the epidemic is growing

Canada’s Rt is now <1

0.5 When Rt is consistently <1, the
suggesting early effect of
epidemic is being brought under
public health measures
control

0
Apr 20 May 20 Jun 20 Jul 20 Aug 20 Sep 20 Oct 20 Nov 20 Dec 20 Jan 21 Feb 21 Mar 21 Apr 21
Date of case report
Data as of April 19, 2021
Note: 7-day moving average. 2
 Variants of concern make up more than half of recently reported
COVID-19 cases in Canada
Daily number of VOC Incomplete data, % of cases that
cases due to reporting lag are VOC*
3,500 100
140
120 90
3,000
100 80
2,500 80
70
60
40 60
2,000
20 50
1,500 0
20 Dec 03 Jan 17 Jan 31 Jan 14 Feb 28 Feb 14 Mar 28 Mar 11 Apr 40

1,000 30
20
500
10
0 0
18 Dec 01 Jan 15 Jan 29 Jan 12 Feb 26 Feb 12 Mar 26 Mar 09 Apr
P.1 B.1.351 Lineage not assigned B.1.1.7 % of all COVID-19 cases with a mutation indicative of a VOC
Data as of April 20, 2021
Note: By laboratory specimen collection date. *Includes cases with confirmed VOC lineage and cases with a mutation indicative of a VOC. VOC cases with no
assigned lineage are reported only by ON and PEI, resulting in underestimation of national number of cases with no assigned lineage. Data from SK are excluded
3
from this analysis.
 Increase in COVID-19 incidence and severe illness in heavily impacted
provinces
Number of cases and
hospitalizations per
100,000 population
40 BC 40 AB 40 SK
30 30 30
20 20 20
10 10 10
0 0 0
Apr 20 Aug 20 Dec 20 Apr 21 Apr 20 Aug 20 Dec 20 Apr 21 Apr 20 Aug 20 Dec 20 Apr 21

40 40 40
MB 30
ON QC
30 30
20 20 20
10 10 10
0 0 0
Apr 20 Aug 20 Dec 20 Apr 21 Apr 20 Aug 20 Dec 20 Apr 21 Apr 20 Aug 20 Dec 20 Apr 21
Cases Hospitalizations
Data as of April 20, 2021
4
 Steepest rise in hospitalizations observed among those aged 40-59 years
Number of reported Incomplete data,
hospitalizations due to reporting lag
160

140

120

100
40 to 59
80

60

40

20

0
Jun 20 Jul 20 Aug 20 Sep 20 Oct 20 Nov 20 Dec 20 Jan 21 Feb 21 Mar 21 Apr 21
Date of illness onset*
0 to 19 20 to 39 40 to 59 60 to 79 80+
Data as of April 20, 2021
Note: Trend lines reflect 7-day moving averages. *The earliest of the following dates: Onset date, specimen collection date, laboratory 5
testing date, date reported to province or territory, or date reported to PHAC.
 Vaccination coverage is increasing across Canada, with benefits being
seen in prioritized high-risk populations

• 29.9% of Canadian adults

have received at least one
dose
72.2%
73.7% • 68.8% of adults in the
57.3% Territories have received at
least one dose
27.6%

30.1% 25.1%
27.0% 33.4%
34.1% 25.7%
29.1%

21.4%
28.2%

Data as of April 17, 2021

6
 Dramatic decline in incidence rates among adults aged 80 years and older
Number of reported cases Incomplete data,
per 100 000 population due to reporting lag
35
80+ years
30

25

20

15

10

0
Jun 20 Jul 20 Aug 20 Sep 20 Oct 20 Nov 20 Dec 20 Jan 21 Feb 21 Mar 21 Apr 21
Date of illness onset*
0 to 19 20 to 39 40 to 59 60 to 79 80+

Data as of April 20, 2021

Note: Trend lines reflect 7-day moving averages. *The earliest of the following dates: Onset date, specimen collection date, laboratory 7
testing date, date reported to province or territory, or date reported to PHAC.
 Sustaining public health measures as vaccination coverage expands is
keeping Indigenous communities safer
Doses administered in
Active cases in First
First Nations
Nations communities
communities
6,000 250,000
225,000
5,000
200,000
175,000
4,000
150,000
3,000 125,000
100,000
2,000
75,000
50,000
1,000
25,000
0 0
Mar 20 Apr 20 May 20 Jun 20 Jul 20 Aug 20 Oct 20 Nov 20 Dec 20 Jan 21 Feb 21 Mar 21
Cumulative number of COVID-19 vaccine doses administered (includes 1st and 2nd doses) Active cases

Data as of April 20, 2021

Note: Data from Indigenous Services Canada. Current week data is incomplete. 8
 Short-term forecast predicts slower increase in cases, with a slight
acceleration in deaths
Cumulative cases predicted to May 2, 2021: Cumulative deaths predicted to May 2, 2021:
1,209,780 to 1,281,040 24,000 to 24,570
1,300,000 24,750

Cumulative number of deaths

Cumulative number of cases

24,500
1,250,000
24,250
1,200,000 24,000

1,150,000 23,750

23,500
1,100,000
23,250
1,050,000
23,000
1,000,000 22,750
07 Apr 12 Apr 17 Apr 22 Apr 27 Apr 02 May 07 Apr 12 Apr 17 Apr 22 Apr 27 Apr 02 May
Cumulatively reported cases in Canada by April 17 Cases added since April 17 when the prediction was made
Prediction to May 2 Lower 95% prediction limit Upper 95% prediction limit
Data as of April 20, 2021
Note: Extrapolation based on recent trends using a forecasting model (with ranges of uncertainty). 9
 The previous longer-range modelling forecast from March 26th continues
to play out in the data we are seeing now

*Grey line - with spread

of VOCs and we
maintain the current
number of people we
contact each day

*Purple line - If VOCs

Reported
are controlled by
cases
reducing the current
number of people we
contact each day by 20-
30%
Red points - surveillance
data after the forecast
from March 25th to April
20th

2020-2021
*Model data as of March 24, 2021. Surveillance data as of April 20, 2021. It assumes VoC introduced in mid-Dec (~1 week prior to first detected case in Canada) at very low prevalence;
VoC is 50% more transmissible than wildtype; growth rates AND replacement rate are negatively correlated
Note: Ensemble of output from PHAC-McMaster and Simon Fraser with the strength of public health measures. Proportion of VoC is indirectly fitted when calibrating to data. 10
University. Recent changes in testing rates are not taken into account in this forecast. SFU methods are at
https://www.sfu.ca/magpie/blog/variant-simple-proactive.html
 Longer-range forecast shows strong measures are required to counter
more transmissible variants as vaccines continue to roll out
Without recent changes to
public health measures.
Many jurisdictions have
recently implemented
measures to prevent this
trajectory
Implementation of recent public
Reported health measures and
cases adherence to them reduces
contact between people by
30%

Implementation of recent
public health measures and
adherence to them reduces
contact between people by
40%

Data as of April 19, 2021

Note: Model developed by PHAC and McMaster University. Model considers impact of vaccination and increased transmissibility of VOCs, 11
refer to annex for detailed assumptions on modelling.
 International experience shows strong and sustained measures are
needed to suppress rapid growth of more contagious variants

Rate per 1,000,000 population

Data as of April 20, 2021

Note: 7-day moving average. 12
 High uptake of 1st vaccine dose will influence whether it will be safe to lift
restrictive public health measures this summer
PLAUSIBLE SAFE LIFTING SCENARIO PLAUSIBLE ALTERNATIVE SCENARIO

1st dose vaccine uptake is at least 75% among adults 1st dose vaccine uptake is at least 55% among adults

 OUTCOME: hospital capacity not exceeded  OUTCOME: hospital capacity widely exceeded in fall 2021
Hospital capacity Measures lifted Median hospitalized cases Vaccination period

Physical distancing, mask wearing, and other personal protective measures alongside current levels of testing and
tracing would be maintained even when restrictive public health measures are lifted.

Note: In both scenarios, measures are lifted when 20% of the Canadians have received their second dose. 13
 Many factors influence the epidemiological situation and timing for
adjusting public health measures…

Factors which could allow lifting of Factors which could result in lifting of
restrictive measures SOONER: restrictive measures LATER:

• Vaccines highly effective • Vaccines less effective

• Faster vaccine roll out • Slower vaccine roll out
• High vaccine uptake • Low vaccine uptake
• Transmission is controlled • Transmission not controlled
• Public health is able to test and trace • Public health unable to test and trace
• Hospitals and ICUs have enough • Hospitals and ICUs at risk or unable to
capacity to meet community needs meet community needs

When more people are vaccinated, together with continued adherence to public health measures, we
will get to lower levels of COVID-19, so we will have fewer restrictions to enjoy more activities

14
Summer holds promise if we can all keep doing our part

Chief Medical Officers of Health

Vaccines can help lead us out of this pandemic and we can all
rolling up their sleeves
contribute to their success by: NS

• Get vaccinated as soon as it’s your turn

• Support others to get vaccinated when it is their turn QC

• Keep following public health advice CPHO

• Keep up with Wash, Mask, Space
AB

Deputy Chief Public Health

Officer of Canada

BC

YK
NT

15
ANNEX

16
 Longer-range forecast shows strong measures are required to counter
more transmissible variants while vaccines continue to roll out
Without recent changes to
public health measures.
Many jurisdictions have
recently implemented
measures to prevent this
trajectory
Implementation and
adherence to recent public
Reported health measures reduces
cases contact between people by
30%
Implementation and
adherence to recent public
health measures reduces
contact between people by
40%

Data as of April 19, 2021

Note: Model developed by PHAC and McMaster University. Model considers impact of vaccination and increased transmissibility of VOCs, 17
refer to annex for detailed assumptions on modelling.
 Longer-range forecasting model assumptions
• The forecast for April 23rd uses a compartmental model reflecting the biology of COVID-19 and public health response developed by PHAC in
collaboration with McMaster University. It projects the near future given recent incidence of COVID-19 and scenarios for public health measures,
variants of concern and vaccination.
• The model assumes that VOCs are 50% more transmissible compared to previous strains. This value is used to estimate the rate at which VOCs
replace existing strains.
• VOCs are considered to have been introduced in mid-December (~1 week prior to first detected case in Canada) at very low prevalence. The
proportion of cases due to VOCs is indirectly fitted when calibrating to data.
• Changes to public health measures impact the speed with VOCs replace previous strains; stronger public health measures result in slower growth
and replacement rates.
• Since many jurisdictions have recently strengthened public health measures, and it is too soon to determine the full impacts of these measures, so
their potential impacts are not yet reflected in the forecast. Recent changes in testing rates are not taken into account either.
• The forecast includes a graph showing the expected increase in cases if public health measures had not recently been strengthened (grey line), one
that assumes recent changes to public health measures will decrease transmission by 30% (purple line), and one that assumes recent changes to
public health measures will decrease transmission by 40% (blue line).
• The forecast includes simplified assumptions on vaccine roll-out, including an assumption that vaccinations are 60% effective against infection after
one dose, a simplified roll-out that does not prioritise by age, and a static vaccination rate. Because only simplified assumptions on vaccine roll-out
are included, the forecast is limited to 30 days.
• The forecast from March 26th uses a slightly different approach, with the following differences:
• It combines the PHAC-McMaster model with a model developed at Simon Fraser University to create an ensemble forecast
• The ensemble forecast assumes VOCs are 40-50% more transmissible and includes a graph showing expected increases in cases with VOC spread (grey line), and a
graph which assumes that transmission (including VOCs) is controlled (purple line) by public health measures that are equivalent to a 20-30% reduction in rates of
contact between people in Canada
• Forecasts are limited to mid-April because assumptions related to vaccination roll-out are not incorporated

18
Lifting of restrictive measures modelling assumptions

• Complex models, including a deterministic, age-structured compartment model and an agent-based model, are used to explore
scenarios for the lifting of restrictive public health measures. These models were initially developed to model measures needed to
control COVID-19 and have recently been adapted to model the effects of vaccination and transmission of variants of concern.
• Assumptions in the model are either obtained based on current data and knowledge of the transmission of COVID-19 and updated as
new data and evidence emerge or obtained by fitting the models to surveillance data. Key assumptions informing the scenarios
presented here include:
• VOC were introduced in December 2020 and are 50% more transmissible and 40% more virulent than previous strains, but do not have immune
breakthrough from vaccines;
• The vaccine is 60% effective at preventing infection and 80% effective at preventing hospitalization after one dose, and 92% effective at preventing
infection and 96% effective at preventing hospitalization after two doses;
• Hospital bed capacity in Canada is estimated at 30 per 100,000;
• Vaccination roll-out proceed in order of priority groups as recommended by NACI with a 4-month interval between doses starting from March 4, 2021.
The 4-month delay progressively decreases to the regular 28-day delay by August. The entire vaccination rollout period is Dec 14, 2020 to Sep 5, 2021;
• In the scenarios shown, restrictive measures are lifted and not re-imposed once 20% of the eligible population have received their 2nd dose of vaccine
(which is at the same time point in both scenarios) to show what would happen based on different levels of vaccine acceptance in the population. Until
then, the epidemic is controlled by a combination of restrictive closures, case detection, isolation, contact tracing and quarantine, and personal
distancing;
• The model assumes 4 weeks after restrictive measures are lifting, the border reopens and the number of imported cases increases from 2 per 100,000
per week to 12 per 100,000 based on current reduction in travel volume due to border restrictions, imported cases are estimated from the PHAC
importation model;
• Restrictive measures refers to closures of settings such as workplaces/businesses, leisure facilities, and other public settings. Even when these
measures are lifted, personal distancing by the public and current levels of testing and contact tracing are maintained.
• Vaccine acceptance by age group is estimated using data from two Canadian surveys (2020 Canadian Community Health Survey – September 2020)
and EKOS probability based research panel (January 6-11, 2021). The 75% vaccine acceptance model is the average overall acceptance ranging from
72% (18-44) to 84.7% (65+). The 55% vaccine acceptance model is the average overall acceptance ranging from 50.4% (18-44) to 59.3% (65+).

19