COVID-19 SYMPTOM STUDY
April 15, 2020, Toronto
fifth of Ontario households self-report any of seven listed symptoms of
key symptoms of COVID-19: fever, cough and shortness of breath, are
characteristic of just less than a tenth of households (8%).
Incidence of the three key symptoms (a new fever, cough, or shortness of breath
- 8%) is higher among the largest households (five or more - 11%) and the least
wealthy (less than $20K - 11%).
in fifty households has a member who's been tested for COVID-19 (2%).
these tests, one fifth returned positive (18%). This is less than half of one
percent of the population (0.4%).
More than one third of Ontarians report an underlying condition which might
aggravate a COVID-19 infection (37%). Of these, the most common is asthma (24%
of those reporting a condition).
About one fifth report being extremely or very emotionally stressed (20%).
About one fifth report being somewhat or very pessimistic about the future
order to avoid "social desirability bias", or the impulse to give a
socially acceptable answer, researchers often ask how "people in the
neighbourhood" behave, as a proxy for respondent behaviour.
- The wide majority of household members
say those in their neighbourhood are observing social distance (86%) and
staying at home (91%), and this draws almost universal approval.
ONE-FIFTH OF ONTARIO HOUSEHOLDS HAVE ONE OR
MORE SYMPTOMS OF COVID-19
In a new poll conducted
jointly by Forum Research and Mainstreet Research among 5045 randomly selected
Ontario households on April 11 and 12, one-fifth reported at least one of seven
listed symptoms of COVID-19 (19%). The symptoms listed were a new fever (3%),
cough (4%), headaches (10%), sore throat (5%), diarrhea (5%), loss of taste or
smell (2%), or shortness of breath (3%).
Incidence of any symptom
is higher among the largest households (five or more - 24%), in the lowest
income bracket (less than $20K - 22%), areas serviced by the Mississauga Halton
LHIN (22%), the Toronto Central LHIN (22%), and the North East LHIN (23%). Two
thirds of those households with any symptoms had 3 or more occupants (61%).
Among those households
interviewed in a Brampton pre-test (n = 4243), just less than one tenth reported
one or more of four listed symptoms, all those shown here with the exception of
diarrhea, shortness of breath and headaches. In the Fort MacMurray pre-test
(Regional Municipality of Wood Buffalo, n = 3643), incidence of any
symptom of the four was 1-in-7 (14%).
(FEVER, COUGH, SHORTNESS OF BREATH) AFFECT ONE IN TEN HOUSEHOLDS
Three of the seven listed
symptoms are seen as key indicators; fever, cough, and shortness of breath.
Together, incidence of having one or more of these three symptoms in the
household is just less than a tenth (8%), and this is especially common to the
largest households (five or more - 11%), households in the lowest income
bracket ($20K or less - 11%) and in the area serviced by the South East LHIN
HAVE BEEN TESTED FOR COVID-19...
One in fifty households
have had a member tested for COVID-19 (2%). This number matches the reported
actual penetration of testing in Ontario as of April 12 (2%). Among those
reporting any symptoms, 1-in-20 have been tested (5%), while among those with
the three key symptoms, one tenth have a-had a test (10%). Of those tested, one
seventh (15%) are still awaiting results.
ONE FIFTH HAVE TESTED POSITIVE
Among the very small group
of households that report having been tested for COVID-19, just less than a
fifth report a positive test (18%), and this may be more common in single
person households, in those with less formal education and among the lowest
income households (caution: small sample sizes). Among the total population,
less than one half of one percent have been tested (0.4%). Among those with any
symptoms, one third have received a positive result (32%), while among those
with the three key symptoms, slightly more have a positive result (39%).
MORE THAN ONE
THIRD HAVE A HEALTH CONDITION WHICH MIGHT AGGRAVATE COVID-19
More than one third of
Ontarians report an underlying health condition which might aggravate a
COVID-19 infection (37%), and this is most likely to be asthma (24% among those
reporting a condition) followed by diabetes (18%) a lung condition (17%) or a
heart condition (13%). Incidence of having this type of condition is most
commonin the area serviced by the Central East LHIN (43%), and the North East
In Brampton, incidence of a comorbidity is
about 3-in-10 (29%), most commonly diabetes (29%), followed by a heart
condition (24%), a lung condition (15%) or asthma (10%). In Fort MacMurray,
3-in-10 also have an underlying condition (31%), most commonly a lung condition
(24%), followed by asthma (22%), diabetes (20%), or a heart condition (13%). Among
those with any symptoms, one half have an underlying condition (48%).
REPORT BEING "EXTREMELY/VERY STRESSED" EMOTIONALLY
One fifth of Ontario
households say their level of emotional stress is "extremely/very
stressed" (20%), and this is especially common in larger households (4
people - 23%, five or more - 24%), in the lowest income group ($20K or less -
23%) and in the area serviced by the Central East LHIN (27%).
In Brampton, one fifth are
also "extremely/very stressed" (18%). In Fort MacMurray, a similar
figure applies (23%).
REPORT BEING "SOMEWHAT/VERY PESSIMISTIC" ABOUT THE FUTURE
One fifth of Ontarians say
they are "somewhat/very pessimistic" about the future (22%), and this
is characteristic of larger households (4 people - 25%), those in the very
highest income group, notably, ($100K+ - 28%), and in the areas serviced by the
Toronto Central LHIN (26%) and the Hamilton, Niagara, Haldimand, Brant LHIN
(25%). In Brampton, this level of pessimism is common to one tenth (12%),
whereas in Fort MacMurray it is almost twice as high (21%).
"SOCIAL DESIRABILITY BIAS"
In order to avoid
"social desirability bias", or the impulse to give a socially
acceptable answer, researchers often ask how "people in the
neighbourhood" behave, as a proxy for respondent behaviour.
OBSERVING SOCIAL DISTANCING AND STAYING HOME
The very wide majority of Ontario
households are observing social distancing (86% reported of neighbours) and are
staying home (91%). Among the minority not doing so (14%), are those in the
lowest income groups ($20K or less - 18%, $20K to $40K - 17%), in the area
serviced by the Erie St Clair LHIN (16%), the Central West LHIN (18%), and the
North Simcoe Muskoka LHIN (19%). Somewhat disturbingly, those who report any
symptoms of COVID-19 are much more likely to say their neighbourhood doesn't
practice social distancing (20%) or staying at home (15%).
APPROVE OF SOCIAL DISTANCING AND STAYING HOME
Ontarians approve of social distancing (97%) and staying home (95%). Among
those very few who don't approve of social distancing, the largest households
are overrepresented (6%), as are respondents in the area serviced by the North
East LHIN (6%). Once again, households with potential COVID-19 symptoms are,
paradoxically, less likely to approve of social distancing. The same general
patterns apply to approval of staying home.
"Just 5% of households reporting
new symptoms consistent with COVID-19 have had testing, and almost two thirds
of households with someone experiencing new symptoms had three or more
occupants. There is considerable room for expansion of COVID-19 testing among
community-dwelling people with symptoms in Ontario. Syndromic surveillance through phone polling
can supplement existing COVID-19 data sources by providing much needed
information on community members who are experiencing possible COVID-19
symptoms but have not yet been tested," said Dr. Lauren Lapointe-Shaw of the
findings of this survey. Dr. Lapointe-Shaw is supporting this
project by sharing her clinical and research expertise to ensure results are
directly relevant to public health administrators and other policymakers.
This poll represents an unprecedented collaboration between two of
Canada's leading market research and public opinion polling firms, Forum
Research and Mainstreet Research. These two firms have come together to
demonstrate proof-of-concept of a nationwide symptoms screening poll, conducted
by telephone, using Interactive Voice Response (IVR) technology, as an
effective method of measuring prevalence and location of people with symptoms
of COVID-19 infection, and identifying clusters, or potential "hot
spots" before they become too advanced to manage.
"Symptom surveillance will provide a better
understanding of the scale and evolution of COVID-19 in Canada. This is because
symptoms start 1-2 weeks before test results, and at present many people with
symptoms cannot access testing," said Dr Lauren Lapointe-Shaw, when asked about the
efficacy of this methodology.
There are key advantages to this methodology:
- Based on a
random probability sample
to the general population
robust sample sizes
in real time
tracking, ongoing reporting, dashboard
- SMS text
messaging included in sampling scheme
functionality allows targeted messaging
social desirability bias
Our methodology captures an
extremely robust probability sample, from which statistically valid projections
can be made, while the freshness of the data, which can be gathered nightly and
reported on a dashboard the next day, ensures key decisions can be made based
on real time data. SMS text messaging is included in the data collection
protocol. Other benefits include interactive functionality missing in web-based
applications, and the elimination of any self-selection bias
This methodology has already been pilot-tested in two
municipalities in Canada, the City of Brampton and the Regional Municipality of
Wood Buffalo (Fort MacMurray), on very large samples, on March 30. Results from
these two pretests have been cited here where indicated.
PARTICIPANTS IN THE STUDY
With offices across Canada
and around the world, 100% Canadian-owned Forum Research is one of the
country’s leading survey research firms, having been the only poll to correctly
call a Liberal majority in 2015. Forum Research is a member of the American Association
for Public Opinion Research (AAPOR) and the Canadian Research and Insights
Differentiated by its
large sample sizes, Mainstreet Research provides accurate snapshots of public
opinion, predicting the majority NDP government in Alberta, the 2018 Ontario
election and a CAQ majority win in the 2018 Quebec election. Mainstreet
Research is a member of the World Association for Public Opinion Research
(WAPOR) and the Market Research Intelligence Association (MRIA) and meets
international and Canadian publication standards.
Lauren Lapointe-Shaw, MD, PhD, is an
Assistant Professor and Clinician Scientist in the Department of Medicine at
the University of Toronto, and a Staff General Internal Medicine Physician at
the University Health Network. She was previously awarded a CIHR
Fellowship Award and CIHR-IHSPR’s Rising Star Award for her doctoral thesis in
Clinical Epidemiology and Healthcare Research. Her research is broadly focused
on measuring the quality of healthcare and the effects of health policy on
Polling across Ontario was
conducted jointly by Forum Research and Mainstreet Research, on April 11 and
12, among a randomly selected sample of 5045 Ontario residents. Respondents were interviewed by landline,
cell phone and SMS text messaging. The poll was conducted by Interactive Voice
Response (Smart IVR) and has a margin of error of plus or minus 1.5%, nineteen
out of twenty times asking. Sample is segmented by 14 Local Health Integrated
Networks (LHINs) at approximately 300 interviews each. Margin of error on
individual LHIN samples is plus or minus 5.7%, nineteen out of twenty times
Where appropriate, data
have been weighted to ensure that data matches the actual population profile
according to the latest Statistics Canada data. Totals may not add up to 100%
due to rounding.
The Brampton poll was
conducted by Mainstreet Research on March 30, among a randomly selected sample
of 4243 residents of the city. Respondents were interviewed by landline and
cell phone. Polling was conducted by Interactive Voice Response (Smart IVR) and
has a margin of error of plus or minus 1.5%, nineteen out of twenty times
asking. Totals may not add up to 100% due to rounding.
(REGIONAL MUNICIPALITY OF WOOD BUFFALO) POLL
The Fort MacMurray
(Regional Municipality of Wood Buffalo) poll was conducted by Mainstreet
Research on March 30, among a randomly selected sample of 3643 residents of the
regional municipality. Respondents were interviewed by landline and cell phone.
Polling was conducted by Interactive Voice Response (Smart IVR) and has a
margin of error of plus or minus 1.7%, nineteen out of twenty times asking. Totals
may not add up to 100% due to rounding.