Elsevier

The Lancet

Volume 358, Issue 9299, 22–29 December 2001, Pages 2110-2112
The Lancet

Articles
Death by homicide, suicide, and other unnatural causes in people with mental illness: a population-based study

https://doi.org/10.1016/S0140-6736(01)07216-6Get rights and content

Summary

Background

People with mental illness are at great risk of suicide, but little is known about their risk of death from other unnatural causes. No study has commented on their risk of being victims of homicide; public concern is pre-occupied with their role as perpetrators. We aimed to calculate standardised mortality ratios (SMRs) and directly standardised rate ratios for death by homicide, suicide, and accident in people admitted to hospital because of mental illness.

Method

We did a population-based study in which we linked the data for 72 208 individuals listed in the Danish Psychiatric Case Register between 1973 and 1993, and who died before Dec 31, 1993, with data in the Danish National Register of Causes of Death.

Findings

17 892 (25%) patients died from unnatural causes. Our results show raised SMRs for homicide, suicide, and accident for most psychiatric diagnoses irrespective of sex. The all-diagnosis SMRs for women and men, respectively, were: 632 (95% CI 517–773) and 609 (493–753) for homicide, 1356 (1322–1391) and 1212 (1184–1241) for suicide, and 318 (305–332) and 466 (448–484) for accident. We recorded an increased risk of dying by homicide in men with schizophrenia and in individuals with affective psychosis. The highest risks of death by homicide and accident were in alcoholism and drug use, whereas the highest risks of suicide were in drug use.

Interpretation

People with mental disorders, including severe mental illness, are at increased risk of death by homicide. Strategies to reduce mortality in the mentally ill are correct to emphasise the high risk of suicide, but they should also focus on other unnatural causes of death.

Introduction

People with mental disorders have high rates of suicide, although estimates of risk vary according to duration of follow-up and in some diagnoses—ie, learning disability and dementia—high risk has not been seen.1 There is less evidence on other kinds of unnatural mortality, and specific causes of death such as accidents and homicides are frequently not distinguished.2 Reports on homicide in association with mental illness focus on the mentally ill as perpetrators rather than victims,3, 4 and discrimination against the mentally ill is thought to arise in part from the perception that they are dangerous.5 However, people with mental illness are also frequently victims of violent crime.6 Our aim was to ascertain the risk of death by homicide, suicide, and accident in psychiatric patients.

Section snippets

Methods

We did a study based on information collected about the Danish population over 21 years and entered in two databases: the Danish Psychiatric Case Register and the Danish National Register of Causes of Death. The Danish Psychiatric Case Register is a record of all people aged 15 years and older who have been admitted to hospital in Denmark with a psychiatric disorder since 1973.7 We used the unique identification number that every Danish resident is assigned to link the data in these two

Results

From 1973 to 1993, 257 720 patients were added to the case register: 119 543 women and 138 177 men. During this period, 29 702 women and 42 506 men died, 17 892 (25%) from unnatural causes (8098 women and 9794 men). Of these unnatural deaths, 181 (1%) were by homicide, 12 977 (73%) by suicide, and 4734 (26%) by accident. The table shows the SMRs and crude rates by diagnosis and sex for homicide, suicide, and accident. In most diagnostic groups, there is a high rate of homicide. The highest

Discussion

Our results indicate high rates of death from homicide, suicide, and accident in people who have been psychiatric inpatients. Most psychiatric diagnoses were associated with increased mortality from all three causes, the highest risks of homicide and accident being in drug use. However, perhaps our most important finding is the high risk of death by homicide among people with mental disorders, including severe mental illnesses such as schizophrenia and affective psychoses. These findings on

References (9)

  • EC Harris et al.

    Suicide as an outcome for mental disorders

    Br J Psychiatry

    (1997)
  • EC Harris et al.

    Excess mortality of mental disorder

    Br J Psychiatry

    (1998)
  • PJ Taylor et al.

    Homicides by people with mental illness: myth and reality

    Br J Psychiatry

    (1999)
  • L Appleby et al.

    Safer services: report of the national confidential inquiry into suicide and homicide by people with mental illness

    (1999)
There are more references available in the full text version of this article.

Cited by (0)

View full text