• Statistics
  • The Institut
  • Surveys
    Go back to document

    Weekly number of deaths in Québec: data sources and methodology

    RED deaths database

    The data in the new series of tables are taken from the deaths database of the Registre des événements démographiques du Québec (RED), maintained by the Institut de la statistique du Québec (ISQ) as mandated by the Ministère de la Santé et des Services sociaux (MSSS). The MSSS uses this database to monitor the health status of the population, while the ISQ uses it to produce demographic data on the Québec population. It should be noted that the RED deaths database is not intended to provide a real-time estimate of the number of deaths.

    The RED deaths database is created from death bulletins (SP-3), which are sent to the ISQ when a death occurs in Québec. These bulletins come mainly from healthcare facilities (hospitals, CHSLDs, etc.) and palliative-care homes. They may also be sent by the Bureau du coroner in the case of deaths involving an investigation. The database also includes deaths of Québec residents that occurred in other Canadian provinces, whose records are provided by Statistics Canada through its own collection process. Deaths occurring in other countries are included when the information is transmitted to the ISQ.

    The statistics disseminated from the RED deaths database cover all deaths of individuals residing in Québec.

    In order to ensure the best possible completeness and quality, data should be considered final approximately 24 months after the end of a year.

    Currently, data on deaths are final up to 2018.

    Production of preliminary data

    The number of weekly deaths disseminated for 2019 to 2021, which are still preliminary, have been adjusted to take into account the coverage of the database, which is still incomplete. The results for these years are based on the number of records present in the database at the time of production and on an estimate of the number of cases still missing (late registrations, deaths submitted to a coroner, deaths outside Québec). The estimated number of missing cases is based primarily on a historical analysis of the rate at which death bulletins are received and processed.

    Under normal circumstances, it is estimated that approximately 80% of records are entered into the database within 4 weeks following the death. The database completeness rate reaches 90% after 8 weeks and 95% after 26 weeks. The majority of cases that are received late are cases referred to a coroner for which the bulletins are completed at the end of the investigation, as well as deaths that have occurred in other Canadian provinces.

    The adjustment made to the 2018 and 2019 data is minor given the database completeness for those years, but it is higher for 2020, especially for the most recent weeks. Differential correction factors are applied based on the age group, sex, and region grouping.

    The additional delay in releasing the data is about 4 weeks for Québec-wide data, and about seven weeks for regional data. These delays are necessary to ensure the quality of the first estimates produced for a given week.

    Cautionary note: Data for the most recent weeks should be interpreted with caution. Due to the current pandemic, the rate of receipt of bulletins may be different from that in the reference period used in the adjustment model. Changes in the rate of receipt are monitored to adjust the correction factors as required.

    Methodological notes

    Universe:

    Deaths of individuals residing in Québec, all causes of death.

    Recurrence:

    Data for the weeks after October 3, 2020 are updated every two weeks, if the database completeness permits.

    CDC week:

    Weeks as defined in several epidemiological studies, including those of the U.S. Centers for Disease Control and Prevention (CDC).

    CDC weeks start on a Sunday and end on a Saturday. They are numbered sequentially from 1 to 52 or 53 depending on the year.

    Data status by year:

    2010 to 2018: Final data (d).

    2019 to 2021: Preliminary data (p) adjusted to take into account the database completeness. Estimated completeness ranges from 99% for the first few weeks of 2019 to approximately 80% for the most recent week of 2021, before the adjustment. Data for the weeks after October 3, 2020 are revised from one release to the next.

    The data presented in dotted lines are for weeks with lower database completeness and, therefore, higher weights. These data are more likely to be revised during updates and must be interpreted with greater caution.

    Rounding:

    Preliminary data are rounded to a multiple of 5; subtotals may not add up to totals due to rounding.