Canadian Criminal Justice Association 32nd Annual Congress on Criminal Justice Deaths in Custody Plenary October 31st, 2009 Halifax, Nova Scotia Deaths in Correctional Custody: A Review of the Literature


Outline of Presentation

  1. Definitions, Typologies, and Limitations 
  2. Natural and Unnatural Causes of Death 
  3. Contributing Factors 
  4. Gaps in the Literature 
  5. Lessons Learned / Best Practices 

A Problem with Definitions

  • There is no commonly accepted definition for 'in-custody deaths'. 
    • Refers to a variety of custodial situations in a singular definition (police lock-ups, youth detention centres, prisons, pre-trial and remand custody, immigration detainment centres, etc.)
    • Others include pre-arrest, apprehension, and post-release in the definition of 'custody'.
    • Suggest that Canadian Roundtable establishes its own working definition

Typologies: What Should We Use?

  • The natural / unnatural typology is a preferred typology as it is broad enough to encompass a variety of custodial situations. 
    • Other typologies refer to causes of death in custody in terms of levels of violence, degree of predictability, and internal/external causality.

Other Limitations

  • Data Limitations 
    • Missing data
    • Changes in legislation / reporting practices over time
    • Focus on suicide
    • Underlying assumptions of offender behaviour
  • Statistical Limitations 
    • Small sample sizes, increased likelihood of finding 'statistically significant' results

Natural Causes of Death

  • Most Common Medical Causes of Death in Custody 
    • Cardiovascular illnesses
    • Cancers
    • Respiratory illness/ diseases
    • Infectious diseases
  • Comparisons to the Population as a Whole 
    • In Canada, natural death rates appear to be higher in federal corrections than in the population as whole
    • The Canadian experience appears contrary to that of international experiences
  • Age 
    • There is some evidence to suggest that offenders are dying younger than would be expected in the population as a whole
  • Ethnic Background & Gender 
    • Do not appear to impact rates of natural deaths
  • Sentence Length 
    • Does not specifically impact, however, longer sentences would increase the likelihood of an individual staying in prison until an older age and many diseases are more prominent in older populations
  • Other Considerations 
    • Underlying assumptions that offenders are "harder on their bodies"
    • Literature accepts that natural deaths cannot be prevented and are an inevitable part of the prison experience

Unnatural Causes of Death

  • Most Common 
    • Suicides
    • Overdoses
    • Homicides
    • Accidents
  • Comparisons to the Population as a whole 
    • Rates of unnatural death are higher in custody than in the population as a whole, particularly in the case of suicide (3 to 9 times the national average)
  • Background 
    • Tend to be young (under 40 years of age), male, Caucasian, and in a medium-security facility
    • In Canada, trends in ethnicity and suicide is consistent with the ethnicity's representation in the prison population
  • Gender 
    • Women offenders are much more likely to engage in self-harming behaviours that do not result in suicide than men
    • Suicide rates amongst women offenders are lower than would have been expected in the Canadian context
  • Sentence Length 
    • Longer sentences are associated with a stronger likelihood of committing suicide in Canada
    • Suicides are more likely to occur in earlier stages of incarceration, particularly following stressful events such as: institutional transfers, negative appeal or parole decisions, and/or the cancellation of visits

Contributing Factors


  • Personal History 
    • Criminal history, family / friends, socio-economic background, education, personal relationships, history of abuse, etc.
  • Mental Health Issues 
  • Substance Abuse / Addictions 
    • Drugs (incl. prescriptions) and alcohol
  • Previous Self-Harming/ Self-Injurious Behaviour 
    • There is a relationship between self-harm and suicide, though not necessarily one of 'cause and effect'

Conditions of Confinement

  • Prison Environment 
    • Lack of control of surroundings
    • Segregation / isolation
    • Physical infrastructure and deficiencies in design
  • Health Care / Mental Health Care Availability 
    • Often particularly problematic during after hours coverage (e.g. midnight shift)
    • Mental health care is a concern across institutions
  • Staff Response 
    • Monitoring and response to medical emergencies is critical
  • Preventative Programming 
    • Availability and limited use of programs can contribute to deaths in custody

Gaps in the Literature

  • Natural deaths in custody 
    • Are inmates harder on their bodies and therefore more likely to die younger?
    • Assessments of health care quality and monitoring in Canadian prisons
    • Comparisons of natural death rates in custody across countries
    • Effect of 'prisonization' on physical and mental health
  • Canadian vs. international context 
    • Further evidence to support / refute higher deaths in federal custody than in provincial custody
    • Comparisons of mean morbidity rates of custodial populations versus the general population
    • How Canada compares to other countries?
  • Differences in deaths in custody across custodial groups 
    • Different security levels
    • Different sentence lengths
    • Remanded, pre-trial, and sentenced individuals

Lessons Learned / Best Practices

  • Staff Activities 
    • Regular and active interaction with offenders (dynamic security)
    • Verification of "Live" body counts
  • Health Care 
    • Improvements in the provision of care (incl. capacity, staffing, and response to emergencies)
    • Increased emphasis on the role of physical and mental health care in the prevention of deaths in custody
    • In the case of suicides, avoiding involuntary transfers
  • Preventative Programming 
    • Increase access and availability
  • Communication 
    • Improved communication between all staff (incl. health staff and correctional staff, as well as with other institutions)
  • Training 
    • Improved staff training in the recognition of and response to at-risk individuals and/or emergency situations
  • Supports 
    • Increased offender access to supports (including volunteer organizations) to build relationships outside of the institution
  • Surveillance 
    • Increased and improved surveillance by staff
    • Alterations of the physical environment (i.e. reduce blind spots and remove/reduce potential suicide ligature points)

Date modified 


Banner Image
Office of the Correctional Investigator - Report