User Guide and Data Dictionary for Preliminary COVID-19 Data

1. Background

COVID‑19 is the disease caused by SARS-CoV2, a novel coronavirus that had not been identified before the first cases were reported in Wuhan, China, in December 2019. The first confirmed cases in Canada appeared in January 2020.

In Canada, the 10 provinces and 3 territories are providing the Public Health Agency of Canada (PHAC) information on COVID‑19 cases on a routine basis. In collaboration with PHAC, Statistics Canada (StatCan) contributes to make these preliminary data available to the research community and to all Canadians.

2. Objectives of the Preliminary dataset on confirmed cases of COVID‑19

PHAC and StatCan have been working closely together to be able to provide preliminary data received from the provinces and territories (PTs) to researchers.

The Preliminary dataset on confirmed cases of COVID‑19 provides easy access to as much data as possible, by provincial regions, while respecting confidentiality of the individuals for which information on COVID‑19 history is reported.

Given the COVID‑19 pandemic is still progressing, the content of this dataset will be updated regularly, making it a unique and relevant product. Each iteration of the dataset will provide up-to-date case information reported by PTs.

This information was originally released in the Detailed preliminary information on confirmed cases of COVID‑19 (Revised) table but due to the increasing number of cases, this dataset could no longer be supported in this format. This table was deleted on Thursday, December 10th, 2020. The information from this table is now available in a downloadable dataset: "Preliminary dataset on confirmed cases of COVID‑19, Public Health Agency of Canada” (13-26-0003).

3. Coverage of the Preliminary dataset on confirmed cases of COVID‑19

The data published by StatCan contains cases for which detailed case information was submitted by provincial or territorial public health authorities to PHAC. The governments of Canada and the provinces and territories agreed on a common Case Report Form (CRF)^Note 1 to be used to report cases to PHAC.

These data may not match the total cases reporting done at the provincial and territorial levels, which are updated routinely by each jurisdiction. Discrepancies are due to factors such as delays in reporting, or variability in reporting cut-offs. Due to these discrepancies, the data are a subset of the total reported cases in Canada.

Routine updates on health outcome status are not made uniformly across Canada, and therefore the data may underestimate the number of hospitalizations, admissions to intensive care units and deaths.

Throughout the pandemic, the CRF has been updated, which has impacted the data published by StatCan. For example information about symptoms was removed from the dataset as of March 2021 since the information was no longer collected on the CRF and the historical information was incomplete.

Variables related to the resolution status of cases were removed from the data file in June 2022. Many provinces and territories were unable to determine the resolution status of a case in alignment with the national definition of a resolved COVID‑19 infection.^Note 2

The variables transmission, asymptomatic, occupation, onset year of symptoms and onset month of symptoms were removed from the data file in January 2023. Most provinces or territories were no longer reporting these variables meaning they are incomplete and no longer representative, creating challenges for interpretation as a result the quality of these variables are no longer deemed sufficient for release.

The data on this dataset is preliminary and subject to change as updated information is received from the provinces and territories.

4. Content of the Preliminary dataset on confirmed cases of COVID‑19^Note 1

This dataset is a subset of the information that provinces and territories collect using the Coronavirus Disease (COVID‑19) Case Report Form^Note 1. The variables selected were those that were considered to be the most important while meeting a certain quality threshold. Also, some “derived variables” were computed by PHAC based on the information contained in the case report forms.

To minimize the risk of disclosure:

1. a few categories from the original questions collected on the form have been grouped together:
   • The provinces and territories have been grouped into the following regions:
     • British Columbia & Yukon
     • Alberta, Saskatchewan, Manitoba & the Northwest Territories
     • Ontario & Nunavut
     • Quebec
     • New Brunswick, Nova Scotia, Prince Edward Island & Newfoundland and Labrador
   • The age in years of individuals has been grouped into age groups:
     • 0-19, 20-29, 30-39, 40-49, 50-59, 60-69, 70-79, 80+
2. a few derived variables have been created:
   • Dates (Earliest Date)
     • Earliest Date is used to derive Episode week, Episode week group and Episode year.
     • All dates were converted to weeks, i.e., Episode week, Episode week group.
     • All cases with an earliest date prior to February 23^rd, 2020 (the first day of the 8^th week of 2020) were grouped with the cases in week 8.
     • If there are insufficient cases by week for a region to be able to release the episode week without compromising the confidentiality, the cases for a given period will be grouped in the week that had the most cases (i.e., not enough cases for the Atlantic region for weeks 23, 24 and 25, most cases were in week 23, all cases will be grouped in week 23). The Episode week group variable will indicate if a grouping was done and appendix IX will indicate which weeks have been grouped together.

Refer to the data dictionary for detailed information about each variable.

The information on this dataset is considered preliminary.

5. Limitations

This dataset includes cases who are confirmed according to the Canadian interim national case definition for COVID‑19^Note 2, that is:

• The detection of at least 1 specific gene target by a validated laboratory-based nucleic acid amplification test (NAAT) assay performed at a community, hospital, or reference laboratory, or
• the detection of at least 1 specific gene target by a validated point-of-care NAAT that has been deemed acceptable to provide a final result, or
• seroconversion or diagnostic rise (at least 4-fold or greater from baseline) in viral specific antibody titre in serum or plasma using a validated laboratory-based serological assay for SARS-CoV-2.

COVID‑19 testing was initially performed for diagnostic purposes only (i.e., to confirm the diagnosis of suspected cases of COVID‑19), and then, increasingly, for screening based on public health priorities (high-risk groups or contact tracing). None of these tests were conducted for research purposes, and the screening was not designed to be conducted in a probabilistic sample representative of the Canadian population.

The expansion of laboratory testing evolved over time following the epidemiology of the disease, i.e., the spread of the disease from China to other countries and the establishment of community transmission in Canada.

With increasing laboratory capacity, some provinces were able to screen people from targeted groups, e.g., residents and staff of long-term care facilities where cases have occurred, or contacts of cases identified in epidemiologic investigations. These expansions of testing did not occur simultaneously across provinces and territories. Additionally, testing capacity and prioritization continue to differ between provinces and territories, thus skewing any inter-jurisdictional comparison.

Due to changes in COVID‑19 testing policies in many jurisdictions, which were in response to a surge in demand for laboratory testing starting in December 2021, cases may be skewed towards populations deemed high priority for laboratory testing, and case counts will underestimate the total burden of COVID‑19 in the population.

Furthermore, following the surge in demand for laboratory testing, COVID‑19 rapid antigen tests (RATs) became increasingly available for public use. While positive laboratory tests have a reporting mechanism to provincial public health authorities, RATs do not, and are therefore not captured within this data set. The availability of RATs for public use limited demand for laboratory testing, resulting in case counts further underestimating of the total burden of COVID‑19 in the population.

The factors listed above must be taken in consideration when interpreting data analysis. Examples of possible bias include:

• Following outbreaks in long-term care facilities, some jurisdictions undertook mass screening in residents of these facilities, which may impact the age distribution of cases. Mass screening in specific segments of the population may lead to their over-representation in the confirmed case data, as general population mass screening has not occurred on a large scale.
• Any comparisons between provinces and territories using demographics or health outcomes may be biased by differences in testing criteria.
• Starting in December 2021, cases may be skewed towards populations deemed high priority for laboratory testing, and case counts will underestimate the total burden of COVID‑19 in the population.

6. Data quality concerns

Routine updates on health outcome status are not made uniformly across Canada, therefore the data may underestimate the number of hospitalizations, admissions to intensive care units and deaths.

There is a high proportion of missing values and some sections of the case report form were provided inconsistently.

Please note that variables may be recoded or changed based on several factors, including but not limited to: new information being reported for historical cases, updates to the case report form, revised reporting by provinces and territories, etc.

StatCan and PHAC are working closely together to improve the quality of the file with the help of all provinces and territories. This will be reflected in each iteration of this dataset.