Revised on 29 May 2020, 9:40 a.m.
Any substantive changes in this HillNote that have been made since the preceding issue are indicated in bold print.
(Disponible en français : Comprendre la transmission et la visualisation des données fédérales durant la pandémie de COVID-19)
One result of the COVID-19 pandemic is an unprecedented effort to collect case data to analyze and respond to its spread. Some results are being disseminated to the public using charts and maps. This HillNote provides explanations and important considerations about the collection, visualization and interpretation of these data, particularly in federal public health information systems.
Canada has had a pandemic plan since 1988, and it has been updated several times based on global events and medical, scientific and technological advances. The plan is called the Canadian Pandemic Influenza Preparedness; Guidance for the Health Sector, or CPIP, and is now developed and maintained by the Public Health Agency of Canada (PHAC).
Surveillance is one of the key components of preparedness and response described in the CPIP. Pandemic surveillance contributes to early detection, informs analysis on the spread of the virus through regions and populations, and supports assessing the clinical severity of a disease by enumerating hospitalizations, intensive care unit (ICU) admissions, and deaths. Although the novel coronavirus is not an influenza virus, the guidance provided in the CPIP, including data collection, still applies.
What Data Are Collected Per Case?
The COVID-19 case report form provided by PHAC guides data collection in public health units in the provinces and territories. The form includes a wide range of questions. Provinces are not obliged to use the form but must, at a minimum, provide information about:
- the reporting province or territory;
- whether the case is confirmed or probable;
- case details (age, gender, occupation);
- symptoms, including onset date;
- pre-existing conditions and risk factors;
- clinical course of illness and outcomes;
- laboratory results, and
- exposures and contact summary.
How Are Case Data Reported and Disseminated?
Rapid and effective dissemination of case data requires collaboration across all levels of government. In 2016, provincial, territorial and federal ministers of health signed the Multi-Lateral Information Sharing Agreement (MLISA) to help streamline data collection and dissemination, and to improve surveillance and pan-Canadian management of public health outbreaks.
The MLISA lists several obligations for each party to the agreement, unless compliance is prevented by a law of Canada, a province or territory. The MLISA further states that the remainder of the agreement provides principles that the parties should make their best effort to follow.
PHAC’s strategy on health data surveillance outlines guiding principles on how pandemic data are collected, modelled and, when required, shared with external partners such as the World Health Organization, while safeguarding privacy.
In response to the COVID-19 pandemic, the federal government issued interim national surveillance guidelines, which recommend that most of the case data listed above be collected and recorded locally with the exception of coronavirus test results. Following the guidelines, patient specimens are sent to authorized public health laboratories. All these data are then forwarded to the provincial or territorial public health authority, which should pass along the information to PHAC within 24 hours for analysis. The process is divided into a data collection stream and a laboratory testing stream.
Reporting process for national notification of human infection with COVID-19
Note: PUI refers to person under investigation
Source: Government of Canada, Interim national surveillance guidelines for human infection with Coronavirus disease (COVID-19), Figure 1, 20 February 2020.
What Can the Data Tell Us?
Case numbers provide some insight into the distribution of the disease, but there are additional considerations including the number of tests performed, the proportion and sector of the population being tested, and the total population of a region.
Rates and proportions can better reflect the prevalence and impact of the disease by identifying, for example:
- percent of cases by demographic factors (e.g., age, sex or gender and ethnicity);
- percent of cases by symptom;
- hospitalization rates, and
- case fatality rates.
Reporting can vary across provinces and territories for several reasons:
- number of tests available and resources available to conduct them;
- intermittent testing and data reporting backlogs;
- differences in testing criteria and/or divergence from the national case definitions guidelines;
- revisions to testing criteria over time, resulting in spikes in reported cases, and
- incomplete case fatality information or reporting.
Charts and Maps
(Examples for illustrative purposes only, please use data sources for current numbers.)
The COVID-19 pandemic data collection is dynamic. For example, PHAC includes shading in its graphs to indicate a time period within which cases may have occurred but not yet reported to PHAC.
More Than Case Numbers: People
Hospitalization and ICU rates can reveal the disease impacts on different population groups. Overall the data reveal that disease severity increases with age.
Age distribution of COVID-19 cases hospitalized, admitted to ICU and deceased in Canada as of 27 May
Source: Public Health Agency of Canada, Daily Epidemiology Update, Coronavirus Disease 2019 (COVID-19), Figure 4. Accessed 28 May 2020.
While females account for more reported cases of COVID-19 to date, males are more likely to experience severe symptoms.
Gender representation in COVID-19
Source: Figure prepared by authors using data obtained from Public Health Agency of Canada, Daily Epidemiology Update, Tables 2 and 4, Coronavirus Disease 2019 (COVID-19). Accessed 28 May 2020.
Charts can be compared most easily when the y-axis (vertical axis) is identical in multiple charts. This can be seen in a comparison of the epidemiological curve of cases over time for all provinces using the relative scale where all the y-axes are the same
Comparing Data That Span a Large Range
When the data span a wide range, data may be transformed with a log scale to make them more easily comparable. The y-axis is calculated so that the distance of 1 to 10 is the same as the distance from 10 to 100 or 100 to 1,000. Using the 500th case reported as a baseline and examining the number of cases accumulated each day since helps illustrate the rate of change of the disease.
Cumulative cases of COVID-19 in Canada compared to other countries by date of report
Source: Public Health Agency of Canada, Daily Epidemiology Update, Coronavirus Disease 2019 (COVID-19), Figure 7. Accessed 28 May 2020
Note: Comparison between countries remains a challenge because of differences in geography, population density, testing regimes and data collection.
Maps are mainly used for explanatory purposes to show a snapshot of number of cases or incidence rates.
The following points should be considered when interpreting maps of pandemic data:
- Proportional circles are best used for counts.
- Shaded area maps are best used for rates.
- Positive test counts do not imply the risk of disease.
- Generally, in locations with high population density where overcrowding occurs, person to person transmission of infectious disease is facilitated.
- Comparison between countries should be made cautiously when there are varying scales in the map, such as counts at a national level next to counts at subnational levels.
- Differences in case counts in neighbouring countries may reflect variation in testing capacity, reporting practices and criteria.
The Canada COVID-19 Situational Awareness Dashboard illustrates total cases using proportional circles by province as well as rates per 100,000 population using a shaded background for an effective comparison between provinces and territories. The Johns Hopkins University’s World COVID-19 Dashboard map is most frequently cited for world data and illustrates incidence rates per 100,000 population, case fatality ratios, hospitalization rates and other measures.
Centers for Disease Control and Prevention, “Principles of Epidemiology in Public Health Practice: An Introduction to Applied Epidemiology and Biostatistics”, 2006.
European Centre for Disease Prevention and Control, Guidelines for presentation of surveillance data, Stockholm: ECDC, 2018.
Henry, B. “Canada’s Pandemic Influenza Preparedness: Surveillance strategy”, Implementation Science, 2018.
Public Health Agency of Canada, Public Health Surveillance.
RSS Statistical Ambassadors, “A statistician`s guide to coronavirus numbers”, Significance, 2020.
Tiedemann, M., “Bill C-5: Public Health Agency of Canada Act”, Library of Parliament, 2006
Young, Ed., “Why the Coronavirus Is So Confusing”, The Atlantic, 2020
Authors: Sonya Norris and Mélanie Zahab, Library of Parliament