Abstract
In Queensland in 1996, 112 people aged between 15 and 25 years committed suicide [1]. In nearly half these cases the chosen method of death was hanging. The frequency of hanging as a cause of death of young people has almost doubled in the last decade [2]. Hanging is a particularly lethal method for suicide attempts, and its increased frequency as a method of choice for suicide may be contributing to the rising rate of suicide observed among young men in Australia [3]. The suicide rate among young people may be significantly reduced by preventing deaths by hanging.
This paper describes the characteristics of young people in Queensland who died from hanging in the years 1995 and 1996. By examining police reports on these cases we were seeking to identify features that were associated with the hanging deaths and which might be modified to prevent future deaths.
Method
Data relating to young people who died from hanging were obtained from the Australian Institute for Suicide Research and Prevention of Griffith University, Queensland, Australia. The database consisted of the coroner's reports and police reports on all deaths by hanging of young people in Queensland in the years 1995 and 1996. For the purpose of this study, we defined young people as those who were under 25 years of age. The case reports provided by the coroner's office and the police were descriptive in nature and we extracted from these descriptions the information contained in this report.
The police reports were variable in the amount of information they contained. All the reports provided a description of the scene of the fatality. Notes, when left by the deceased, were recorded. Most reports contained the results of police inquiries concerning the circumstances of the death and some included interviews with friends or acquaintances, but these were not collected routinely. Some of the reports were in a semistructured form, containing questions about the deceased's behaviour, including alcohol and drug use, prior to the fatal event. In most cases, the police reports contained the results of forensic pathology tests for substances, but autopsies were not generally undertaken. While the reports were thorough, they were made mainly to exclude foul play and were not directed to establishing possible antecedent causes to suicide. Consequently, the lack of systematic data collection limits their use for this purpose. In particular, they lack a general approach to interviewing family members, friends or health professionals that would help establish a consistent background for these suicides. The coronial reports generally accepted the police reports and sometimes added further to them. Again, the data recorded were not consistent enough to provide more than a limited basis for comparison.
This is a descriptive study done in the context of a report on hanging deaths in Australia for the Commonwealth Department of Health and Aged Care [4]. It was made possible by a rare opportunity to read the police and coroner's reports. Unfortunately, we were not able to develop a control group due to constraints on resources, and the results from this study are limited for generalisation by this design failing. We have used Australian census data for comparisons with the reported data where possible.
Results
Age, cause, sex and ethnicity
In Queensland between 1 January 1995 and 31 December 1996, there were 137 young people who were aged less than 25 years and who were recorded as dying by hanging. The cause of death in all cases was recorded as suicide. Hanging was six times more frequent among males than among females (119 males compared to 18 females). The median age of the males in our sample was 19.5 years. The youngest subject was 10 years old. About 7% of the sample (nine cases) were aged less than 16 years. There was a steep rise in the incidence of deaths by hanging after this age. Three cases (2%) were identified as people who were of Asian origin.
Of the sample, 26% (35 cases) were identified as being of Aboriginal origin. All of these people were male. About 10% of these indigenous people (three cases) were aged less than 16 years.
Physical status
No physical abnormalities were recorded as being present in any person in the sample. One person had asthma, and six had visited a general medical practitioner for minor medical complaints such as colds and coughs, just before their deaths.
At the time of death, 38.0% of the cases in the sample (52 cases) were recorded as having blood levels of alcohol. Only 5.8% (eight cases) had blood levels of cannabis and 4.4% (six cases) had other drugs recorded in their blood. Fifteen per cent (21 cases) had combinations of alcohol and other drugs. There were 16 cases where this information was not available.
Personal relationships
Nearly 10% of the sample (12 cases) were recorded as being married or living in de facto relationships. The sexuality of the subjects was not usually recorded. No case was recorded as having same-sex preference.
In 22.6% of the sample (31 cases), separation from a significant other person was recorded as occurring immediately prior to the fatal event. In 12.4% (17 cases) a suicide of a friend or relative had occurred at sometime in the past life of the subject (one maternal suicide, two paternal suicides, three siblings, three friends and eight acquaintances).
Mental health and psychiatric illness
Previous attempts at suicide were recorded in 14.6% of the sample (20 cases). In a further 32.2% (44 cases) threats of suicide were made prior to the fatal event. In 20% (27 cases) a suicide note was left.
Eighteen per cent (25 cases) were recorded as having a history of psychiatric illness. Two were recipients of a disability pension for psychiatric illness. Depression was recorded in 5.8% of cases (six cases) and schizophrenia in 8.1% (11 cases). No one was noted as having a diagnosis of a personality disorder. Twelve per cent of the sample (17 cases) had received previous psychiatric treatment and one person had received counselling at an adolescent mental health unit.
Behaviour that was out of character was noted prior to the fatal event in 40.1% of the sample (56 cases). In 12.4% (17 cases) it was recorded that the person had been depressed prior to their death. In 6.6% (nine cases) it was noted that the subject had been angry, and in 5.1% (7 cases) that they were drinking heavily prior to their death. One case was noted as being ‘paranoid’ and another as being ‘withdrawn’. In the other cases, the type of behaviour was not described.
In three cases the police reports noted the possibility of the subject being influenced by watching a video or reading a book in which hanging featured.
Social circumstances
Twenty per cent of the deaths (28 cases) occurred in regional towns. Thirty-seven per cent (51 cases) occurred in rural shires. Sixteen per cent (22 cases) occurred in remote rural areas of the state. This compares with 24% (33 cases) who died within the Brisbane metropolitan area (place of death could not be identified in three cases).
Twenty per cent of the deaths occurred between 04.00 hours and 08.00 hours and a further 20% between 16.00 hours and 20.00 hours. The deaths were spread evenly throughout the week days, but there were more deaths on Sundays than on other days. This suggests that the evenings and early hours of Sunday morning may be a specially dangerous time. There were also more deaths in December than in other months.
The fatal event occurred in the subject's own home in 57% (78 cases); 33.6% (56 cases) occurred in the open air. Only 3% (four cases) occurred in prison, one case in police cells and one in hospital. None were recorded as occurring in a mental health unit, at work or at school.
Twelve per cent (15 cases) were facing court appearances at the time of the fatal event. One was a victim (of rape) and the others were alleged offenders.
Of the total sample 43.7% (60 cases) were unemployed; 10.3% (14 cases) were students and 4.4% (six cases) were employed in community employment programs. Of the Aboriginal people in the sample, 57.0% were unemployed. Of non-Aboriginal people in the sample, 38.0% were unemployed at the time of their deaths.
Method of suicide
Hanging included all forms of asphyxiation by ligature. Eighty-eight per cent (120 cases) were by suspension. In three cases, death occurred by kneeling. In one case, the subject was cross-dressed at the time of death (this may have been an accidental death).
In 61.6% (94 cases) the subjects had been dead for some time before they were found. In 12% (16 cases), unsuccessful attempts were made to revive the subjects.
Discussion
Coronial and police records show serious limitations in providing an account of factors associated with suicidal behaviour. If these are to be more useful in research into this cause of death they will need to include a more comprehensive account of the biography of the subjects. This might include more extensive interviews with relatives and friends who may be available shortly after the fatal incident. More extensive data would allow comparison with corresponding data for randomly selected control subjects. Uncontrolled studies, such as this one, carry considerable potential to produce misleading conclusions about the risk factors and life processes leading to suicide. One response to this data is to call for more comprehensive police investigations into these deaths, aided by guidelines developed by researchers and assisted where possible by research assistants.
Some conclusions can be made with some confidence from this descriptive study of 2 years consecutive deaths by hanging among young people in Queensland.
Young people who kill themselves by hanging do so privately and at times and in places that provide little possibility of their being found before they die. For this reason, if this cause of death is to be prevented, interventions must occur at an earlier time.
Findings that were specially notable include the following: (i) many more males than females died by hanging; (ii) young men from remote and rural areas were overrepresented; and (iii) young persons of Aboriginal origin were greatly overrepresented.
Nearly half the total population of Queensland is in the metropolitan area of Brisbane, the capital city of the state [5]. However, the majority of deaths by hanging occurred in regional, rural and remote areas of the state, not in the metropolitan area.
Unfortunately, the amount of information contained in the police reports about the social context in which these deaths by hanging occurred was very limited. We were unable to draw any conclusions about which social factors may be producing developmental stress from these data. In particular, it is not possible to identify from these data why males are so at risk from suicide by this method. However, our findings do suggest that there is an urgent need to investigate more thoroughly the social and cultural obstacles to healthy psychological maturation of males in rural Australia. Those who are in difficulty, and the reasons for their problems, should be identified in these studies. Plummer [6] has made a careful analysis of accounts provided by young Australian men of their growing up and school experiences. He demonstrated how when some young people transgress social rules and expectations in schools or communities, wittingly or by chance, they can be persecuted by their peers. He argues that this may lead to suicidal ideation.
Unemployment may be a significant factor in the development of the suicidal behaviour. The overall rate of unemployment for 15–24–year-old people in Queensland was about 16% [7]. Within this overall unemployment rate, Aboriginal people had an unemployment rate that is at least twice that of non-Aboriginal people. Consequently, the unemployment rate of those who died by hanging, both non-Aboriginal and Aboriginal people, was three times higher than would be expected in comparison with the general Queensland population of young people. Our comparisons were made with the overall Queensland unemployment rates and not with the unemployment rates of the districts where the deaths occurred. Most of the deaths by hanging occurred in regional and rural parts of the state. Consequently, the high unemployment rates recorded in this sample may be more characteristic of the districts in which the deaths occurred. In the absence of comparative data and the ability to control for confounding factors in this descriptive study, a causal relationship cannot be derived.
The overall proportion of Aboriginal people in the Queensland population is about 2% [8]. Consequently, the expected number of Aboriginal people in our sample would be three cases; there were 35. Aboriginal people were 10 times overrepresented among the deaths by hanging. This is a cause for great concern.
Of the 99 deaths investigated by the Royal Commission into Aboriginal Deaths in Custody in Australia in 1988/1989, 30 were by hanging. The main characteristics of those who died by hanging in custody were that the subjects were young (most were under 25 years of age), they were alone and there were high levels of alcohol in their blood [9]. Our findings were congruent with these, although only a few people in our sample had died while in a custodial situation.
Cultural factors may influence the choice of suicide method. Different methods of suicide characterise different cultures [10]. The method chosen for suicide may in part be related to the culturally defined way one punishes the self [11]. Hanging was the traditional method of capital punishment in Australia, and as such, may have a special cultural significance. In addition to its lethality, its ready accessibility and its convenience, hanging may be chosen because it is a public signifier of punishment [12]. The meaning of this form of self-inflicted death to the Aboriginal population needs to be explored.
Consultation by Government with Aboriginal people who have knowledge in this area is urgently required. A challenge to parents and the whole community is to find ways to help indigenous youth through stressful phases of their adolescence. Psychiatric services need to be more sensitive to the cultural aspects of the mental health of indigenous people, if they are to help. The education of psychiatrists and other mental health professionals may be deficient. Currently, indigenous aspects of psychiatry occupy only a marginal place in the College of Psychiatrists' training curriculum and take up very little of the curricula of other mental health professionals [13,14]. The Centre for Aboriginal Studies of Curtin University has recently produced a handbook for mental health professionals, giving helpful guides to etiquette and customs in order to facilitate more appropriate interactions between them and indigenous people. The text is available on the AusEinet website [15]. There is also an urgent need to find ways to increase the numbers of Aboriginal people working in the mental health professions. Imaginative approaches through early selection, bridging programs, scholarships and the development and promotion of special opportunities may need to be considered by government in conjunction with Higher Education authorities and Aboriginal advisers.
Research into the causes of suicide indicates that social factors are rarely sufficient in themselves to result in completed suicide. Rather, it is the combination of stressful social circumstances with emerging symptoms of depressive illness that can produce suicidal impulses [16,17]. The contribution of depression to suicide in young people is an important one. Epidemiological studies report that depressive illness commonly develops for the first time during adolescence [18]. The recent survey of the mental health of Australians found that the prevalence of anxiety disorder was highest in people under 25 years of age [19]. Such findings indicate that young people have a risk of suicide as a result of the presence of mental disorders.
About 20% of the subjects had a known psychiatric illness. Studies that have collected data on suicide victims from a wide range of sources consistently report higher levels of psychiatric illness, up to 90% [20]. Those with the highest levels usually include substance abuse and personality disorders. It is likely that the discrepancy between the findings from these studies and our own reflects the serious limitations of coronial and police report data as an account of an individual's psychiatric illness and the unsystematic definitions of psychiatric disorder used. This suggests that more systematic reports are needed from the initial investigations of these cases if they are to provide information that will be useful for developing prevention strategies. Even so, there was little evidence of any ongoing psychiatric care for the subjects for whom psychiatric problems were recorded.
How to get help for distressed people before the fatal event occurs is an important consideration for prevention. They need ready access to mental health services so that their treatment needs can be quickly assessed. In order to develop prevention policies, it would be important to know what services currently exist for young people, how accessible they are and what treatment is actually available. Mental health services are clearly inadequate for indigenous people in rural Australia. Eighty per cent of indigenous people living in rural areas do not have mental health services located within 25 km of their dwelling. Only 35% have access to a permanent doctor [21].
Some basic preventative measures emerge from our findings that could be implemented immediately. Young persons who threaten suicide should be given priority for specialised mental health assessment, where possible by psychiatrists. Threats should be treated as potentially life-endangering situations. Young people who face court hearings need personal support as this appears to be a particularly stressful time. In the absence of adequate family support, they should be offered support through the justice, health or welfare systems. People who have been treated for psychiatric illnesses or attempted suicide need adequate medical follow up, especially if they live in rural areas. Young people who are feeling depressed should know that treatment is available for this condition. It will be important for policy makers to use the information about the pathways to services for young people, and the barriers they encounter when trying to reach them, that will emerge from the recently conducted national survey of the mental health of children and adolescents [21].
Footnotes
Acknowledgements
The authors would like to thank Pierre Baume and Chris Cantor and the staff of the Australian Institute for Suicide Prevention and Research for their help. This project was funded by a grant from the Department of Health and Family Services of the Commonwealth of Australia.