Social inequalities in COVID-19 deaths in Canada

Technical notes

Information about the COVID-19 age-standardized mortality estimation, stratification by individual-level measures and stratification by area-level measures.

COVID-19 age-standardized mortality estimation

• Numerators: Deaths were those for which coronavirus disease 2019 (COVID-19) was identified as the underlying cause of death in the provisional death data. Two ICD-10 codes were used to identify when COVID-19 was reported as a cause of death: U071 for COVID-19 specified as confirmed by a positive test result, and U072 for COVID-19 described as "possible," "probable," or "pending a (positive) test result." Deaths are reported for province of residence.
• Denominators: The Canadian population based on 2016 Census.
• Mortality rates were age-standardized to the 2011 Canadian population, using 5-year age groups.
• The COVID-19 data presented are provisional, as they are not based on all deaths that occurred during the reference period because of reporting delays.
• National-level data do not include Yukon due to unavailability of mortality data from Yukon after 2017.

Stratification by individual-level measures

• Stratification measures include:
  • household size
  • household type
  • dwelling type
  • low-income status (according to the after-tax Low Income Measure [LIM])
• Data are disaggregated by the above measures and by sex (both sexes, females, and males).).
• Source: Provisional Vital Statistics-Deaths Database (January 1 2020 to March 31 2021), linked to 2016 short-form Census. A rounded total of 10,845 COVID-19 deaths (4,820 among females, 6,025 among males) were recorded in this dataset.
• During the initial period of reporting (January to July 2020), over 4,400 COVID deaths were reported for individuals living in private dwellings in Canada. The current reporting period therefore contains approximately 6,400 more deaths. Generally, the age-standardized mortality rates in the updated tables are higher due to more deaths being captured within the time period. However, due to the provisional nature of the data and revised numbers from the initial reporting period (that is, an indication of the under-coverage of the previous tables), an additional 400 deaths were captured. Due to this observed under-coverage, comparing the results from the initial and current reporting periods should be done and interpreted with caution.
• These reported deaths occurred among residents of private dwellings. Deaths occurring among residents of collective dwellings, such as long-term care homes, are not included in these data. During this period of the COVID-19 pandemic, half of COVID-19 deaths (50%) occurred in long-term care homes (Canadian Institute for Health Information, 2020, Pandemic experience in the long-term care sector).
• A small portion of the deaths attributable to COVID-19 were not linked to the 2016 short-form Census and were excluded from these analyses. These missing data have a negligible impact on rate estimates.
• Rates based on death counts smaller than 10 are suppressed.
• There is a gap between the short-form Census information collection (data as of May 1, 2016) and when the COVID-related deaths occurred (January 2020 to March 2021). Individuals’ characteristics may have changed between 2016 and 2020–2021, and this represents a potential source of measurement error in these analyses.

Stratification by area-level measures

• Stratification measures include:
  • ethno-cultural composition quintile
  • neighbourhood after-tax per-person equivalent income quintile (local)
  • neighbourhood after-tax per-person equivalent income quintile (national)
  • large urban centre (CMA) residence
• Data are disaggregated by the above measures and by sex (females, males).
• Source:Preliminary Vital Statistics – Deaths Database (January 1 to December 31, 2020) linked with 2016 Census Area Profile data (in the PCCF+ file). A rounded total of 16,120 COVID-19 deaths (8,340 among females, 7,780 among males) were recorded in this dataset.
• During the initial period of reporting (January to August 2020), over 9,250 COVID deaths were reported for individuals living in private dwellings in Canada. The current reporting period therefore contains approximately 6,870 more deaths. Generally, the age-standardized mortality rates in the updated tables are higher due to more deaths being captured within the time period. However, due to the provisional nature of the data and revised numbers from the initial reporting period (that is, an indication of the under-coverage of the previous tables), an additional 140 deaths were captured. Due to this observed under-coverage, comparing the results from the initial and current reporting periods should be done and interpreted with caution.
• Rates based on death counts smaller than 5 are supressed.
• The Preliminary Vital Statistics - Death Database includes deaths occurring in Canada of Canadian residents and non-residents as well as deaths in long-term care homes.
• A small portion of the deaths attributable to COVID-19 were not linked with PCCF+ file due to missing postal code and were excluded from these analyses. These missing data have a negligible impact on rate estimates.
• Area-level data are based on the 2016 Census. The characteristics of the areas in which deaths occurred may have changed between 2016 and the mortality reporting period (January to December, 2020). However, previous research has noted a relative stability of socioeconomic characteristics through time (for example, Blair et al. (2015), The longitudinal effects of neighbourhood social and material deprivation change on psychological distress in urban, community-dwelling Canadian adults).
• Ethno-cultural composition: Statistics Canada's measure of area-level "ethno-cultural composition" captures the relative area-level concentration of individuals who were designated as a visible minority, recently immigrated to Canada or were born outside of Canada, or have no knowledge of either official language of Canada (English and French). For the development and dimensions of Canadian Index of Multiple Deprivation, please refer to Canadian Index of Multiple Deprivation User Guide.
• Neighbourhood income: Neighbourhood income after tax per single person equivalent is a household size-adjusted measure of household income, based on 2016 Census Summary Profile data at the dissemination area (DA) level and using person-equivalents implied by the 2016 low-income cut-offs (LICOs).
  • National-level neighbourhood income quintiles were constructed by ranking all DAs from the lowest income per single-person equivalent to the highest, and by assigning DAs to 5 groups, such that each group contained approximately one-fifth of the total in-scope population. DAs were weighted by population size to ensure an even distribution of population in each quintile.
  • Local-level neighbourhood income quintiles were constructed separately for each census metropolitan area, census agglomeration or residual areas within each province. DAs within each such area were ranked from the lowest average income per single-person equivalent to the highest, and DAs were assigned to 5 groups, such that each group contained approximately one-fifth of the total in-scope population of each area. The quintile data were then pooled across the areas.
• Census metropolitan areas (CMAs): CMAs have a total population of at least 100,000 residents, of which 50,000 or more live within the urban core. See list of the 2016 CMAs in Canada.