Introduction

About 4500 people die by suicide in Canada annually. Researchers assert this number is understated; various studies suggesting the real number is 10-30% higher. After stigma, the main reason for underreporting suicide is lack of standards in death determination practices and data collection. A national suicide prevention strategy could set national standards for both these areas. Read more about the accuracy and reliability of suicide statistics in Canada in this Centre for Suicide Prevention editorial.

Compared with the USA, Canada collects a paltry amount of information. Read more about our overall data deficit in the Globe and Mail’s 2019 series here. Complete, robust, timely data are key for designing targeted suicide prevention efforts.

Current practice

  1. Provincial-Territorial (P-T) coroners or medical examiners determine cause of death
    1. Some coroner and medical examiner offices release this data, labeled ‘preliminary’
  2. Death data is collated by P-T Vital Statistics Registries (typically part of P-T Ministry of Justice)
  3. P-T Vital Statistics send datasets to Statistics Canada
    1. P-T Vital Statistics sometimes post this data, labeled ‘preliminary’
    2. Some provinces analyze and interpret the data on an ad hoc basis; Québec does this regularly: contact epidemiologist Pascale Lévesque, MSc, conseillère scientifique à l’Institut national de santé publique du Québec (INSPQ)
  4. Statistics Canada cleans the data, supresses categories lower than 5 to anonymize the deaths, and then publishes the data
    1. Statistics Canada holds the Canadian Coroner and Medical Examiner Database (CCMED)
    2. Statistics Canada will produce descriptive reports upon request

Gaps

System-level lacks:

  • Structure to support collaboration and knowledge exchange between CC/CME
  • Common approaches to death investigation
  • Resources at P-T level (paper processes, no research/ data management units)
  • Ability to consistently identify, evaluate and monitor mortality trends at a national level reducing the opportunity for intervention or prevention.

CCMED limitations:

  • Not nationally comparable (no common approaches to death investigation)
  • Not comprehensive (some P-Ts don’t submit)
  • Not timely (delayed submission and availability)
  • Not easily analyzed (cause and circumstance of deaths is mostly text; lack of core and minimum data sets)

Opportunities for improved death determination and data collection practices

Here is a selected, non-exhaustive list of opportunities for improvement in our surveillance and data collection.

  • Build a Collaborative of Chief Coroners and Medical Examiners, PHAC and Statistics Canada to develop the Common Approach Framework (This has been underway for about a year.) Goals include:
    • Drive knowledge exchange
    • Address system gaps and challenges
    • Co-develop common approaches to data collection to improve comparability
    • Improve CCMED
  • Implement national standards for suicide death investigation practices to promote consistency of reporting (somewhat underway, see Collaborative, above)
  • Determine what data needs to be collected; set national standards; consider a pan-Canadian death certificate to collect consistent demography
    • National standards need to include a breakdown by sex (beyond male and female), gender, age, Indigeneity, ethnicity, race, and occupation for all thirteen jurisdictions to permit intersectional analysis
  • Develop and implement national surveillance: consistently identify, evaluate, and monitor mortality trends at a national level overall (see Québec example, above) and for priority populations over-represented in suicide including First Nations, Métis, Inuit, and some occupations such as first responders
  • There are no standard tables produced at Statistics Canada of mortality rates by occupation, as occupation data is not available on the Vital Stats This type of information needs to come from integrated data products like the Canadian Census Health and Environment Cohorts (CanCHECs) which have occupation data on one data source linked with the death data. The most recent data is the 2016 census linked to deaths up to 2020
  • Mandate and set standard reporting schedule to ensure submission and improve timeliness of it from P-Ts to achieve something closer to real-time access to data
  • Establish core and minimum data sets to facilitate analysis of cause of For example, on death certificates, instead of occupation reporting being optional and inputted textually, occupation field can be mandatory and inputted by selecting an option from a standardized list