Health, death and indigenous Australians in the coronial system.

Carpenter, Belinda ; Tait, Gordon

Abstract: This paper details research conducted in Queensland during the first year of operation of the new Coroners Act 2003. Information was gathered from all completed investigations between December 2003 and December 2004 across five categories of death: accidental, suicide, natural, medical and homicide. It was found that 25 percent of the total number of Indigenous deaths recorded in 2004 were reported to, and investigated by, the Coroner, in comparison to 9.4 percent of non-Indigenous deaths. Moreover, Indigenous people were found to be over-represented in each category of death, except in death in a medical setting, where they were absent. This paper discusses these findings in detail, following the insights gained from the work of Tatz (1999, 2001, 2005) and Morrissey (2003). It also discusses a further outcome of this situation--the over-representation of Indigenous people in figures for full internal autopsy.

Introduction

According to the Coroners Act 2003 (Queensland), the purpose of the coronial investigation is to determine a cause of death finding for any death that is considered suspicious, violent or unnatural. The coronial system is also called upon to determine cause of death when the death is neither suspicious, violent, nor unnatural, but where a medical officer cannot, or will not, write a cause of death certificate. More recently, the coronial system has been required to investigate deaths in institutions, like prisons and aged care facilities, as well as those that occur in institutions while under the purview of certain forms of legislation, like mental health or disability.

As a consequence of the parameters of the 2003 Coronial Act in Queensland, it is perhaps not surprising that Indigenous Australians would be over-represented in certain categories of death, and overwhelmingly under-represented in others. Violence in many Indigenous communities is well documented and may lead to their over-representation in homicide statistics within the coronial system (Memmott et al. 2001). Poor access to health care due to location and poverty would go some way to explaining the absence of Indigenous deaths in a medical setting (Ring and Brown 2001), while lower life expectancies and high rates of chronic disease may help explain Indigenous over-representation in deaths by natural causes (Hogg 1992; Paradies and Cunningham 2002). The related problems of alcohol and drug misuse can be connected to over-representations in accidental death, especially motor vehicle accidents (Macaulay et al. 2003), while a history of dispossession, poverty and community sense of injustice may go some way to explaining Indigenous overrepresentation in suicide statistics (Tatz 1999, 2001, 2005).

As well as explanations about why Indigenous people may be over-represented in the Queensland coronial system, this paper also identifies one of the effects of this situation, which is the subsequent over-representation of Indigenous people subjected to a coronial autopsy. Indigenous Australian cultural proscriptions against autopsy are in place to protect the spirit of the deceased, which it is argued would be harmed by a mutilation of the body and thus prevented from entering the Dreaming (Lynch 1999:72). As Vines (2007:17) identifies:

 while the range of beliefs amongst Aboriginal
 peoples is broad, we can say that the relationship
 between Aboriginal people and their
 dead is one of custodianship. The body
 should remain whole so that the spirit will
 have somewhere to go. Autopsy is often seen
 as desecration and destructive of the spirit.

More specifically, Wilson (1998:49), citing an appeal in a Western Australian court against a coronial autopsy, includes affidavit material that maintains that 'if the body of the child is cut up then that will mean that his spirit will not rest according to our belief. It is the Aboriginal belief also that the spirit will be roaming around and will not enter dreamtime.'

The forensic or coronial autopsy has been described as a coercive process, since it takes place by force of law and, unlike hospital or clinical autopsies, irrespective of the wishes of the family of the deceased (Segal 2006:101). While it may be the case that coronial autopsies can still proceed without the consent of the family, all Australian jurisdictions now have a capacity to contemplate objections to coronial autopsy. However, only the State of Queensland and the Australian Capital Territory mention cultural matters in the context of such objections (Vines 2007). In the Coroners Act 1997 (ACT), s28 provides for a capacity to object to a post-mortem based on 'the desirability of minimising the causing of distress or offence to persons who, because of their cultural attitudes or spiritual beliefs, could reasonably be expected to be distressed or offended by the making of that decision'.

In Queensland, s19 of the Coroners Act 2003 stipulates that before ordering an internal examination, the coroner must, whenever practicable, consider that a deceased person's family may be distressed by such a procedure because of spiritual or cultural beliefs. Further, a coroner must consider any concerns raised by a family member, or other person with 'sufficient interest', and if the coroner decides to order an internal autopsy despite such concerns being expressed, a copy of the order must be given to the person who raised the concern. However, the Act does not provide any mechanism for a concerned family member to challenge an order, although the decision would be subject to review in the Supreme Court at the instigation of the relative pursuant to the Judicial Review Act 1991. The effect of such changes in legislation for autopsy decision making in Queensland is considered in more detail later in the paper. At this point, we will review national research, statistics and literature on the overrepresentation of Indigenous Australians in death investigations.

Indigenous Australians, health and death

It is well documented that the 'health status of Indigenous Australians is greatly inferior to that of the non-Indigenous population and that this is most clearly demonstrated by their much shorter life expectancy' (Morrissey 2003:32). More specifically, Indigenous Australians suffer greater ill health, are more likely to experience a disability and reduced quality of life, and, as a consequence, will die 20 years earlier than non-Indigenous Australians (Trewin and Madden 2003:129; Condon et al. 2004). In terms of age-standardised death rates, Indigenous Australians have higher infant mortality and higher accident, poisoning and violence mortality, as well as double the national rate of death from circulatory diseases (Morrissey 2003:32).

It has consistently been noted that the poor health status of Indigenous Australians is related to social and economic factors including diseases triggered by poverty, overcrowded housing and a lack of home ownership, poor sanitation, a lack of access to education, poor nutrition and higher levels of unemployment (Thomson et al. 2007:31). As a consequence, it has been argued that the further health risks of smoking, obesity and alcohol misuse, which impact on the onset, maintenance and prognosis of a variety of chronic diseases, place 35-54-year-old Indigenous Australians with age-specific death rates five times higher than the corresponding total non-Indigenous Australian population rates (Trewin and Madden 2003:201).

Such issues are not in themselves a reason for the over-representation of Indigenous Australians in the coronial system. However, when coupled with their location in remote and rural areas, and their subsequent lack of access to, or resistance to, medical care for accurate diagnosis and treatment, at least a partial explanation is proffered. According to recent figures, '15% of Indigenous people have no access to a doctor, 17% have no access to a nurse and 22% have no access to an Aboriginal health worker within 25km of where they live' (Shepherd et al. 2003:632). This lack of access to health services, their lower life expectancy and chronic ill health may mean that the coronial system is more often notified of natural Indigenous deaths because a medical practitioner cannot, or will not, issue a cause of death certificate.

While 'diseases of the circulatory system' was the leading cause of natural death among Indigenous Australians, accounting for 680 deaths in the period 1999-2001, 'external causes of morbidity and mortality' was the second major contributor, at 512 Indigenous deaths during the same period (Trewin and Madden 2003:192). Given the role of the coronial system in investigating deaths of an unnatural, violent or suspicious nature, the large percentage of Indigenous deaths from external causes of morbidity and mortality goes some way to explaining their over-representation in homicide, suicide and accidental deaths.

It has been argued that violence in Indigenous communities manifests itself in, and is the result of, a range of health, economic and social issues. For example, when Indigenous Australian and non-Indigenous Australian rates of violent death are compared, death by suicide is between two and 2.6 times higher, assault mortality rates are between five and ten times higher, and deaths due to motor vehicle accidents are 3.4 times higher for Indigenous Australians (Stevenson et al. 1998, Mouzos 2001, Anderson 2002). It is also maintained that violence in Indigenous communities shares many of the fundamental antecedents that are part of the poor health and high mortality rates of Indigenous Australians. For example, it is argued that reasons for the over-representation of Indigenous Australians in motor vehicle accidents include a lack of access to advanced emergency medical care when an injury occurs; misuse of alcohol, and its illegality in many Indigenous communities, which means that residents who drink may drive long distances while impaired; the over-representation in the Indigenous population of young people (due to lower than average life expectancy), who are at higher risk of injury due to an increased level of risk-taking behaviour; and the over-representation of Indigenous people in predominantly rural and remote areas, which means that they frequently travel on roads that have higher speed limits and that are often poorly maintained or unsealed (Stevenson et al. 1998).

In their substantial review of literature over the past 20 years, Memmott et al. (2001) identify three explanations for the causes of violence in Indigenous communities: precipitating causes, situational factors and underlying factors. Precipitating causes are defined as a type of social event that triggers an episode of violence and could include making a pass at a de facto; quarrelling between a husband and wife; children fighting at school; accidentally knocking someone over at sport; arguing over a game of cards; and not inviting someone to a wedding or birthday (Memmott et al. 2001:21). Situational factors exacerbate violence in combination with direct causes, and may include family problems, financial problems, bereavement, unemployment, psychological problems, anger and alcohol intoxication (Memmott et al. 2001:18). Underlying factors are generally historical and involve the ways in which violence has been perpetrated upon Indigenous people since European settlement, and the impact of this socially, economically, physically, psychologically and emotionally. The identification of underlying factors is at least partially supported by statistics that demonstrate that the four places with the worst incidence of violent crime in Queensland are Indigenous communities that were once mission centres--places with a long history of European contact, and where large numbers of displaced Indigenous people were brought from other areas. They are thus places where 'maximum dysfunctional cultural changes have occurred' (Memmott et al. 2001:14).

The over-representation of Indigenous people in the coronial system has often, therefore, been explained by the myriad social and economic factors that underpin their low life expectancy, chronic ill health and high level of violent death. The rest of this paper examines this set of explanations through analysis of coronial data gathered in Queensland during 2004.

Methodology

Study design

All completed coronial investigations for the 12-month period between December 2003 and December 2004 were examined over a four-month period by a team of non-medical researchers. Checklists were established for the five categories of death (natural causes, accidental deaths, suicides, medical deaths and homicides), and these were divided into three areas of information. The first recorded basic demographic information about the deceased, including date of birth, date of death, gender, religion if stated, Indigeneity if stated and whether or not a family concern was raised against autopsy. The second section recorded information about the autopsy decision-making process including the geographic region within which the autopsy decision was made, the type of autopsy to be conducted (external only, external and partial internal, or external and full internal), the timing of the decision, the place where the autopsy was carried out, and, if internal, the extent of the autopsy. Using the post-mortem report as the guide, the extent of the internal autopsy was coded, with areas of the body identified as brain, neck, heart, lungs, abdomen, spine, limbs, other. For these to be included as an area of dissection, each had to be specifically named in the report as being dissected. Where five or more areas of the body were dissected, this was noted as a full internal autopsy; when four or less were dissected, this was noted as a partial internal autopsy; and when there was only a viewing of the body, and blood taken for toxicology, this was noted as an external autopsy.

The first two sections of the checklist gathered the same information irrespective of the type of death identified in the police investigation. The third section recorded information about evidence gathered at the scene by the police. This section was specific to each of the categories of death previously identified. In the case of accidental death, six factors were noted: whether or not unexplained injuries were evident; whether or not the accident was witnessed; whether or not the witness may have caused the accident; whether unexplained blood and/or signs of a struggle were evident; whether an autopsy might identify a perpetrator; and whether litigation was likely. For death by suicide, five factors were noted: the existence (or not) of a suicide note; whether triggers to the suicide were identified; whether or not the house was secure; whether a pre-existing mental illness could be identified; and whether previous attempts at suicide had been made within the past six months. For death by natural causes, six factors were noted: the existence (or not) of unexplained injuries; the existence (or not) of unexplained blood or signs of a struggle at the scene; whether or not the death was expected; whether or not the deceased had a life-threatening illness; whether or not the family or treating doctor would form an opinion as to cause of death; and whether or not the house was secure. For death in a medical setting, four factors were noted: whether or not allegations had been made by the family; whether or not a lack of care was evident or asserted; whether or not a lack of skill was evident; and whether or not medical reports were available. For those deaths that were considered a potential homicide by police, three factors were noted: the existence or not of foul play; whether or not police considered the death suspicious; and whether or not homicide could be excluded.

This checklist was first tested against 50 complete investigations from 2005 and further refined before being used for the 2003-2004 data collection process. Information about the death was gained from a variety of sources on file, including the initial police report; the postmortem report; any follow-up police reports; the cause-of-death certificate; any follow-up pathology findings; and any written communication between police, coroners, pathologists, family and counsellors.

Limitations and advantages of this data set

The timing of the research coincided with the enactment of the Coroners Act 2003 (Queensland), as well as the appointment of the State's first Chief Coroner. In 2006 the choice to analyse data from two years previously was based on a need to access the greatest number of complete closed files available, and reflected the time taken to complete accident reports by police, histology and other forensic findings by pathologists, and inquests by magistrates. However, an inherent limitation of this file sample is that it involved only cases from the first year of a new system that varied markedly from the previous one and which was introduced with very limited training for those required to implement it. Most significantly, this lack of training may have impacted on the capacity of police to gather as much information from the scene as was warranted in every case, especially in terms of the Indigenous status of the deceased. Police often relied on their own assessment of the Indigeneity of the deceased, and this may not always have been reliable. Moreover, it has been suggested by Cuneen (2001) that, given the history of relations between Indigenous Australians and police, bereaved families would not necessarily inform police of their Indigenous identity in coronial investigations, implying that the determination of Indigenous descent of deceased persons may be under-represented in this data. Furthermore, given that a death investigation by police concentrates only on the identification (or not) of Indigenous status, the possibility of any detailed identification, as to whether the deceased was Aboriginal or Torres Strait Islander, is negated.

Finally, one debate in the research on death is concerned with the accuracy of death classifications, especially when determining death by suicide (O'Carroll 1989). The current research attempted to avoid such problems by using a range of information from a variety of sources in the coronial investigation to create a distinct picture of each death. The research findings are thus probably more accurate than those drawn solely from statistics from the Australian Bureau of Statistics, for example, which rely, in turn, on cause-of-death certificates, as these only record the cause of death (e.g. asphyxiation) and not the circumstances of death (e.g. suicide) (McKenzie et al. 2004). Nonetheless, the current research may not have always correctly identified Indigenous suicide, as it has relied on more universal markers of suicide, such as apparent mental illness and the presence of suicide notes, both of which have been questioned in terms of their reliability as a determinant of Indigenous suicide (Tatz 1999, 2001, 2005).

Analysis

To enable consistency of coding, all information was entered into a database by one researcher. Statistical comparison between Indigenous and non-Indigenous people for each category of death was then undertaken. In order to assess whether or not the results were statistically significant, a chi-square test of proportions ([chi square]) was completed. By examining the differences between the observed and expected frequencies, this test can determine if there are significant differences between sample populations, and/or whether these differences are simply due to chance.

Results

The general demographic findings from the research are outlined here in terms of statistical differences between Indigenous and non-Indigenous people for each category of death--natural causes, medical setting, suicide, accident and homicide--as well as the differences in autopsy figures. While this section explores the categories of death by suicide, accidental death and death by natural causes in some detail, it does not examine suspected homicide or death in a medical setting. Although death by homicide represents the greatest rate of over-representation for Indigenous people in the coronial system, it was excluded from more detailed analysis because of the low sample size (four suspected homicides were Indigenous). Death in a medical setting was excluded because no Indigenous people were identified in this category of death during the period in question.

Demographics

In 2004 there were 24 500 registered deaths in Queensland, of which 519 were recorded as Indigenous (ABS 2006). Between December 2003 and December 2004, 3000 deaths were identified as reportable, and of these 2446 deaths were investigated by the Queensland coronial system. Of these, 133 persons were identified by police in their initial report as being Indigenous. Thus, 25.62 percent of Indigenous deaths were investigated by the coroner, in contrast to 9.64 percent of non-Indigenous deaths. Statistically, this is a highly significant difference. (1)

When violent death in general is examined (combining suicide, accident and suspected homicide), and compared with overall death rates, there is a significant difference between the 10.78 percent of Indigenous people (56 of 519) and the 4.46 percent of non-Indigenous people (1071 of 23,981) who died violently in the same timeframe. (2) These statistics support the large body of research that demonstrates that Indigenous people are far more likely to die violently than non-Indigenous people (Mouzos 2001, Anderson 2002, Thomson et al. 2007).

For each category of death, we find that Indigenous people are either over-represented or completely invisible. As shown in Table 1, in terms of the conventional rate of expressing deaths for comparative purposes (as rates per 100 000 of the population), with the exception of medical deaths our research demonstrated Indigenous over-representation when compared to the non-Indigenous population.

The fact that Indigenous people fail to come to the attention of the Coroner as the result of a death in a medical setting may be a combination of lack of access to medical attention and low life expectancy. However, it may also be due to social reasons like poverty, and historical relations of distrust between hospitals and Indigenous people due, among other factors, to their role in creating the Stolen Generations during the twentieth century (HREOC 1997). Such claims need further research, which is beyond the scope of this paper. However, in terms of deaths by suicide, accident and natural causes, we are able to offer some tentative findings that provide support for previous research and challenge some of the more dominant assumptions about Indigenous death.

Suicide

The results represented in Table 2 show that Indigenous people who suicide are more likely to use self-asphyxiation via hanging as their method, and that suicide is predominant within the young. According to Elliot-Farrelly (2004:2), suicide was not known in traditional Aboriginal society and, even up to the 1960s, 'suicide was a rarity'. However, by the 1990s suicide in Indigenous society was endemic, with the highest rates being in the 15-24 year age group. The predominance of self-asphyxiation via hanging as a method is considered by Hunter et al. (1999) as both a political and symbolic statement, which, they argue, is related to, and influenced by, recent deaths in custody, as well as historical punishment during colonisation, and thus contains within it connotations of capital punishment, genocide, injustice and oppression.

Further differences between Indigenous and non-Indigenous suicide have also been noted. According to Tatz (1999, 2001, 2005), suicide notes, and what is commonly (but not necessarily accurately) believed to be mental illness, are not useful markers of Indigenous suicide. Our research results in this area are represented in Table 3.

The data lends support to Tatz's (1999, 2001, 2005) claim that the high level of illiteracy within the Indigenous population makes suicide notes a less than useful indicator of Indigenous suicide. It also suggests that diagnosed mental health conditions are not a predominant pre-cursor to Indigenous suicide, despite research that suggests that depression is the precipitating cause of Aboriginal suicidality (Elliot-Farrelly 2004). In Tatz's (1999, 2001, 2005) study, he reported little evidence of clinical depression and little to no correlation between Indigenous suicide and diagnosable mental illness, and our research results offer initial support for these findings.

Accident

Unlike suicide, accidental death does not demonstrate a great difference in terms of mode of death between Indigenous and non-Indigenous people. Thus, for both groups, accidental deaths due to motor vehicle accidents were predominant, constituting 278 of 498 (55.82 percent) accidental deaths for non-Indigenous people and 16 of 27 (59.26 percent) accidental deaths for Indigenous people. However, when taken as a percentage of all deaths, there is a significant difference between these groups, as 3.08 percent of all Indigenous deaths were due to motor vehicle accidents, compared with 1.15 percent of non-Indigenous deaths.3 This data lends support to national findings by the Australian Transport Safety Bureau, which suggest that Indigenous Australians are over-represented in road fatalities by a ratio of 3:1 (Macaulay et al. 2003:8).

After motor vehicle accident, the most common form of accidental death for non-Indigenous people was drowning (10.84 percent), closely followed by accidental overdose (10.64 percent). For Indigenous people, drowning only constituted 3.70 percent of all accidental deaths, while accidental overdose was the second most common form of accidental death, at 7.41 percent. However, as noted in the previous discussion, the use of markers such as suicide notes and diagnosed mental illness to identify suicides in the Indigenous population may not be reliable, and thus may challenge the accuracy of an accidental--as opposed to intentional--overdose determination. This observation requires further research.

Interestingly, while only eight Indigenous women were identified as victims of motor vehicle accidents in the research timeframe, three died as pedestrians (37.5 percent). In comparison, seven of 63 non-Indigenous women died as a result of being a pedestrian (11.1 percent). While these proportions are too small to garner statistical significance, they do suggest an over-representation of Indigenous women in the statistics for pedestrian deaths, and offer tentative support for the more recent Australian Transport Safety Bureau Report Update, which found that Indigenous people were over-represented in this category (Styles and Edmonston 2006).

Death by natural causes

For both Indigenous and non-Indigenous people, circulatory disease was identified as the major cause of reported death by natural causes, constituting 858 of 1179 (72.77 percent) natural deaths for non-Indigenous people, and 54 of 77 (70.13 percent) natural deaths for Indigenous people. Respiratory disease was the next most common cause of reported natural death for both groups, but only comprised 84 deaths (7.12 percent) for non-Indigenous people and nine deaths (11.69 percent) for Indigenous people. While these differences between Indigenous and non-Indigenous people in the coronial system are not significant, when taken as a percentage of all deaths they demonstrate that 14.8 percent of Indigenous people are referred to the coroner for investigation of a natural death, compared with 4.9 percent of non-Indigenous people. This is a highly significant difference, (4) and suggests that a lack of access to medical services is impacting on the ability of doctors to confirm cause of death in the Indigenous population.

In the course of this research, information gathered by investigating police at the scene of death was recorded. For death by natural causes, this included whether or not the death was expected, and whether the deceased had a life-threatening illness prior to death. For non-Indigenous people, the death was unexpected in 28.4 percent of cases, while for Indigenous people the death was unexpected in 35.06 percent of cases. While this is not a significant difference, it does demonstrate a trend towards more unexpected deaths in the Indigenous population. In contrast, there was little to no significant difference between Indigenous and non-Indigenous people in terms of the existence of a life-threatening illness, at 57.14 percent and 56.06 percent respectively. However, when this is combined with data that shows that the highest proportion of Indigenous people reported to the Coroner for death by natural causes were in the age group 45-55, a trend of low life expectancy, coupled with poor chronic health and lack of access to medical care, is established.

Autopsies

In terms of the number of full internal autopsies performed for suicides, death by natural causes and accidental death, there was no statistical difference between the Indigenous (72.9 percent) and non-Indigenous (69.5 percent) data. Thus the identification of Indigenous status by police during the initial stages of a coronial investigation did not influence orders for a full internal autopsy any more than the communication of no Indigenous status. However, 25 percent of Indigenous deaths were investigated by the Coroner in 2004 in comparison with 9 percent of non-Indigenous deaths. There is, therefore, a highly significant difference between the percentage of Indigenous deaths in 2004 that were fully internally autopsied in the coronial system (19.4 percent or 101 of 519) and those for non-Indigenous deaths (6.83 percent or 1584 of 23 981). (5)

These figures indicate that while it is Indigenous people's over-representation in deaths of a violent nature, as well as their high rates of chronic ill health, low life expectancy and poor access to medical attention, that underpins their over-representation in the coronial system, the effect of this is a further over-representation in the statistics for full internal autopsies. As previously noted, there are cultural proscriptions against autopsy, and this over-representation has implications for Indigenous cultural beliefs about the journey of the spirit after death. This paper will now engage with the position implicit in the work of Morrissey (2003:33) and Tatz (1999, 2001, 2005), which demonstrates how 'the structures which support the good health of the majority, are clearly failing this minority', and, also, that Indigenous people are not offered 'incentives for remaining in life' (Tatz 1999:9).

Discussion

According to Wilson et al. (2007:184), 'considerable pessimism surrounds the issue of Indigenous mortality and health in Australia'. Such pessimism is based upon two issues: first, mortality statistics, which demonstrate an increasing and significant disparity in life expectancy between Indigenous and non-Indigenous Australians; and, second, statistics that demonstrate that Australian Indigenous life expectancy is lower than that of Indigenous people in other 'settler societies'. While the growing disparity in mortality between Indigenous and non-Indigenous Australians may be partly explained by the increasing improvement in the health of the non-Indigenous population (Morrissey 2003), this does not explain the increasing disparity between the life expectancy of Indigenous Australians and Indigenous people in North America and New Zealand, who, it is argued, share the social and economic conditions of Indigenous Australians, as well as the historical condition of a minority culture, in an affluent nation, dispossessed of their country and marginalised (Ring and Firman 1998).

Explanations for the poor progress of Indigenous Australians' health status, in comparison with other Indigenous populations like New Zealand and North America, include a lack of specialised Indigenous health programs, especially those employing specialised Indigenous staff (Shepherd et al. 2003); the absence of a land treaty, which is argued to have played a 'significant and useful role in the development of health services, and in social and economic issues for the Indigenous people of New Zealand, The United States and Canada' (Ring and Firman 1998); the funding relationship, and subsequent competition and lack of co-ordination between federal and State governments, for the provision of Indigenous health care (Kunitz 1994, cited in Morrissey 2003); and, finally, the loss of control over their own lives that these issues raise, which work together to create and reinforce a sense of powerlessness in Indigenous Australians' lives, which is less tangible in other Indigenous peoples (Syme 1997).

However, as Morrissey (2003:32) states, the focus in much of the research on Indigenous health as a health problem first and foremost is misleading and tends to support inaccurate methodological and theoretical approaches. This is for two reasons. First is the focus on the poor health in Third World countries of Indigenous populations who have a low life expectancy because of acute communicable diseases. While Indigenous Australians' poor health and low life expectancy have been likened to 'Third World', this is a misleading comparison. Indigenous Australians' low life expectancy is due to chronic illness like diabetes and atherosclerosis, not to acute diseases like AIDS or cholera. Second, most of the theoretical models for understanding chronic illness, and the methodological processes for comparing them, have been developed in 'First World' countries, which again offer little useful comparison with Indigenous Australians. Instead, using 'First World' aetiologies of chronic disease for 'Third World' people has reduced large-scale social, political and historical issues to individual level risk reductions. 'Individuals become inter-changeable across different social structures, across different cultures, different social histories, different positions in structures of race or ethnic inequality' (Morrissey 2003:35).

Morrissey (2003, following Goffman 1974) argues for a different theoretical approach to those causes of low life expectancy in 'Third World' peoples that are acute and those that are chronic, by maintaining that chronic diseases in 'Third World' populations have virtually no natural history at all, and that the focus should be on their social 'nature'. In this way of thinking, chronic diseases in Indigenous Australians do not have 'social causes, rather they are in essence social constructions' (Morrissey 2003: 34).

According to Morrissey (2003:34), such a conceptualisation allows for the movement back and forth between the personal and the public, between the self and the society, in terms of the aetiology of the 'disease', and this enables a more wide ranging and yet also specific engagement with Indigenous health. First, such an approach would encourage researchers to seek qualitatively 'rich descriptions of the causal pathways between social variables and health outcomes' (Morrissey 2003:38). For example, an injury is often interpreted by Indigenous Australians in relation to its effects on their subsequent ability to fulfil social obligations. When this impact is perceived as negative, treatment is likely to be refused or discontinued. Thus the wellbeing of the individual is only perceived in relation to the wellbeing of the relevant social group. Second, such an approach would also eliminate reductive, depoliticised accounts of social concepts like 'community', which is often not a genuine site of sociality given their historical creation, or 'poverty', which, when treated as an isolated risk factor, does not explain a great deal (Morrissey 2003:40). Finally, such an approach would enable Indigenous Australians to move to the centre of control of the agenda, conduct and uses of research in Indigenous health (Morrissey 2003:40). As Morrissey (2003:37) notes, between 1990 and 2002, 433 articles dealt with the social aspects of Indigenous health but only 28 made any effort to explain the poor health outcomes of Indigenous Australians in terms of the historical, social or political structures within which their poor health can be best understood and explained.

Similarly, Tatz (1999, 2001, 2005) argues that Aboriginal issues of health and mortality have their own unique social, historical, political and cultural contexts, and must be seen as distinct phenomena. While focusing on Indigenous youth suicide, his argument has resonance for all forms of violent death of Australian Indigenous people, because he differentiates between those factors that impact on Indigenous society generally and those specific to the underlying causes of violence. In terms of the former, Tatz (1999:13) identifies the confusion, ambiguity and contradiction in State and federal Indigenous policy, which, he argues, is the basis for uncertainty, unease, ambivalence and a confusing diffusion of responsibility: such ambiguity 'bewilders' not only Indigenous Australians but also 'those responsible for bringing government policies to fruition'. In order to illustrate this point, Tatz (1999:13-20) cites seven areas that impinge upon the daily lives of most Indigenous communities, and that have within them high levels of ambiguity, contradiction and confusion. These include the confusion over land rights and the serious turning back of hard-won decisions since 1973; Aboriginal participation in decisions affecting them, and the retreat from Aboriginal control over their own affairs since it was stated as a universal goal in the 1960s; removed children and the need for a national apology as opposed to 'burying the mistakes of the past'; and the 'One Australia' philosophy, 'which brooks no special treatment for any one group but which ignores the effects of history and those who are already powerful'.

Importantly for this discussion, such large-scale ambiguity has serious consequences and, according to Tatz (1999:20), there is a causal link between it and the patterns of violent behaviour in many Aboriginal societies. The stages are 1) a feeling of frustration; followed by 2) a sense of alienation from society, of not belonging, of foreignness; then 3) withdrawal from society, no longer caring about membership, loyalty, law-abidingness; and then 4) the threat of, or actual, violence. Moreover, as Tatz (1999:20) explains, the violence in Aboriginal life is focused on 'appealing violence--a cry for help when one is at the end of one's road ... confined to self, to kin and at times to those who work with and for communities'.

On the more specific issue of suicide, Tatz (1999:84) identifies eight specific 'Aboriginal factors' that increase the likelihood of suicide, 'These are specific, smaller in scope and more tangible' than the issues in relation to violence that have been discussed above, yet they still have implications for violent death of Indigenous people more generally. Some of these include a lack of a sense of purpose in life, of an inability to exert one's will, to believe that life could be better; high levels of molestations and abuse of children, despite the fact that child molestation and sexual abuse have always been absolute taboos in all Aboriginal societies; drinking and drug taking, which, according to Tatz (1999:90), is motivated most likely by a user's need to obliterate his or her present existence, and probably to mask a suppressed wish for permanent obliteration; continuous cycles of grief in which there is no time to complete the grieving before another death ensues, and where there is little grief counselling; and high levels of illiteracy in a wider society (and health service) predicated on literacy, which means a lack of communication, whether in writing, reading and/or verbally.

There is no shortage of explanations for violent deaths in Indigenous society and many of the issues raised by Tatz (1999, 2001, 2005) have been identified in previous research. For example, the Commonwealth report Violence in Indigenous Communities, which was based on a thorough review of the literature on Indigenous violence up to that point, focused its attention on social explanations like poverty, unemployment, substance abuse and welfare dependency; psychological problems like powerlessness, lack of self-esteem, hopelessness and depression; and cultural issues like kinship and familial relations (Memmott et al. 2001). However, the report failed to give any detailed consideration to the historical or political context of violence by Indigenous people. The national report Australian Indigenous Road Safety was based on a review of literature on Indigenous road fatalities up to 2003, and explained Indigenous over-representation as being due to behavioural risk factors like alcohol intoxication, overcrowding and illegal seating positions in vehicles; environmental risk factors like increased exposure through greater distances travelled, higher speed limits and poorer road quality; and vehicle risk factors like older vehicles, lower vehicle ownership and lack of public transport (Macaulay et al. 2003). A more recent report, Australian Indigenous Road Safety: Update 2005, also focused on these known risk factors (Styles and Edmonston 2006). As a consequence, neither report, nor the policy recommendations they supported, engaged with economic or social factors that might underpin issues such as low vehicle ownership and subsequent overcrowding, or dealt with the historical and political factors that Tatz (1999) identified as linked to excessive alcohol consumption. Finally, The Overview of Australian Indigenous Health Status 2007 cited a range of factors impinging upon the high level of injury among the Indigenous population, including many of those that have already been discussed (cultural, environmental and lifestyle variables, socio-economic disadvantage, and geographical isolation), but offered no detailed discussion of the political or historical context of many of these issues, and thus no analysis of the relationship between history, politics, culture, the psyche and the social (Thomson et al. 2007).

Importantly, Tatz's (1999:126) conclusions about the method and theory of research into Indigenous suicide are very similar to Morrissey's (2003) conclusions about Indigenous health: explanations must have a greater focus on historical, political and social factors, and current research needs to liberate itself from a narrowly focused 'bio-medical model' previously discussed. Moreover, Tatz's (1999, 2001, 2005) research offers a detailed template on how such work might proceed, being both methodologically qualitative in its approach and theoretically diverse in its explanations. It also offers a more central role for the coronial investigation not only to more accurately identify Indigenous suicide and other violent deaths, but also to gather more appropriate social data at the scene of death, which would enable more appropriate statistics to be gathered and thus offer a more detailed and informed approach to Indigenous death.

Conclusion

The research outlined in this paper has demonstrated that the over-representation of Indigenous people in the coronial system has led to their subsequent over-representation in coronial autopsies, despite their cultural proscription against dissection of the body. More complicated, however, is how such over-representation may be addressed. This paper has examined the multitude of explanations given in the literature for the high rate of injury and low life expectancy of Indigenous Australians, with the work of Tatz (1999, 2001, 2005) and Morrissey (2003) offering the more complex explanations and, as a consequence, suggesting the more detailed pathways out of this situation.

In terms of the coronial investigation, the recent change in legislation in Queensland, which enables Indigenous status to be communicated to the Coroner, along with family concerns against autopsy, is an important first step towards decreasing the number of coronial autopsies performed for Indigenous deaths. However, this research demonstrated that Indigenous status had no effect on decisions to conduct autopsies, and Indigenous people were not likely to raise concerns about the autopsy with the investigating police officer. More specifically, while 144 families raised concerns about the autopsy procedure during the research timeframe, only six were motivated by an Indigenous cultural concern. This number may increase if further training is given to police officers and coroners about cultural proscriptions against autopsy. However, unless they are also trained in Indigenous communication, made conversant with explanations for Indigenous suicide (for example), and understand the social, political and historical factors surrounding the low life expectancy of Indigenous people, knowledge of proscriptions against autopsy may not induce change. Tatz (1999, 2001, 2005) has suggested that the American model of utilising forensic anthropologists in death investigations may be a useful addition to the coronial system in Australia, as would the increase in status and numbers of Aboriginal community liaison officers, who operate in many rural communities and 'are in the forefront of practically every facet of Aboriginal life' (Tatz 1999:145).

Finally, it may be that the coronial system can also help alleviate the number of Indigenous people coming under its purview, at least in terms of violent death, by instituting a more rigorous identification of Indigenous people in their investigations, and by identifying the social factors surrounding each death, rather than simply the individual and, in the case of suicide, the mental health issues. The one constant in research into the health and mortality of Indigenous people is the lack of available statistical data. Gathering more detailed knowledge of the social circumstances of all Indigenous deaths in the coronial system would increase understanding of this phenomena, enable more detailed explanations and, following Tatz (1999:124), offer successful alleviation, or at least mitigation, of these 'trends toward mortality'.

ACKNOWLEDGMENTS

This research has been funded by the Australian Research Council and the Queensland Department of Health. Access to coronial files was made possible by the Office of the State Coroner and the Director General of the Office of Justice and Attorney General.

REFERENCES

ABS (Australian Bureau of Statistics) 2006, <abs.gov. au>.

Anderson, Ian 2002 'Understanding Indigenous violence', Australian and New Zealand Journal of Public Health 26(5):408-9.

Condon, John, Tony Barnes, Joan Cunningham and Len Smith 2004 'Improvements in Indigenous mortality in the Northern Territory over four decades', Australian and New Zealand Journal of Public Health 28(5):445-51.

Coroners Act 1997 (ACT)

Coroners Act 2003 (Queensland)

Cuneen, Chris 2001 Conflicts, Police and Crime: Aboriginal communities and the police, Allen & Unwin, Sydney.

Elliot-Farrelly, Terri 2004 'Australian Aboriginal suicide: The need for an Aboriginal suicidology?', Australian e-journal for the Advancement of Mental Health 3(3):1-8.

Goffman, Erving 1974 Asylums: Essays on the social nature of mental patients and other inmates, Pelican, New York.

Hogg, Russell 1992 'Indigenous mortality: Placing Australian Aboriginal mortality within a broader context', Social Science and Medicine 35(3):335-46.

HREOC (Human Rights and Equal Opportunity Commission) 1997 Bringing them Home: Report of the National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children from their Families, Commonwealth Government, Canberra

Hunter, Ernest, Joseph Reser, Mercy Baird and Paul Reser 1999 An Analysis of Suicide in Indigenous Communities of North Queensland: The historical, cultural and symbolic landscape, report for the Mental Health/Health Services Development Branch, Commonwealth of Australia, Canberra.

Kunitz, Stephen J. 1994 Disease and Social Diversity: The European impact on the health of non-Europeans, Oxford University Press, New York.

Lynch, Mathew 1999 'Forensic pathology: Redefining medico-legal death investigation', Journal of Law and Medicine 7(August):67-74.

Macaulay, Jemima, Rachel Thomas, Nick Mabbott, Tanya Styles, Colin Edmonston, Mary Sheehan and Cynthia Schonfeld 2003 Contract Report: Australian Indigenous road safety, Australian Transport Safety Bureau, Canberra (ATSB Report, RC2321).

McKenzie, Kirsten, Sue Walker and Ron Casey 2004 'Some issues to consider when assessing concordance of death certificates with registry reports', Australian and New Zealand Journal of Public Health 28(1):87.

Memmott, Paul, Rachel Stacy, Catherine Chambers and Catherine Keys 2001 Violence in Indigenous Communities, report to the Crime Prevention Branch of the Commonwealth Attorney General's Department, Canberra.

Morrissey, Michael 2003 'The social determinants of Indigenous health: A research agenda', Health Sociology Review 12(1):31-44.

Mouzos, Jenny 2001 Indigenous and Non-Indigenous Homicides in Australia: A comparative analysis, Australian Institute of Criminology, Canberra (Trends and Issues in Crime and Criminal Justice series 210).

O'Carroll, Patrick 1989 'A consideration of the validity and reliability of suicide mortality data', Suicide and Life Threatening Behaviour 19(1):1-16.

Paradies, Yin and Joan Cunningham 2002 'Placing Aboriginal and Torres Strait Islander mortality in an international context', Australian and New Zealand Journal of Public Health 26(1):11-16.

Ring, Ian and Ngaire Brown 2001 'Indigenous health: Chronically inadequate responses to damning statistics', Medical Journal of Australia 177(11): 629-31.

--and David Firman 1998 'Reducing Indigenous mortality In Australia: Lessons from other countries', Medical Journal of Australia 169:528-33.

Segal, George 2006 'Law and practice in relation to coronial postmortems--A social perspective', Medicine and Law 25:101-13.

Shepherd, Frank, Kristine Battye and Elizabeth Chalmers 2003 'Improving access to cardiac rehabilitation for remote Indigenous clients', Australian and New Zealand Journal of Public Health 27(6):632-6.

Stevenson, Mark, LJ Wallace, James Harrison, Jerry Moiler and Richard Smith 1998 'At risk in two worlds: Injury mortality among Indigenous people in the US and Australia, 1990-92', Australian and New Zealand Journal of Public Health 22(6):641-4.

Styles, Tanya and Colin Edmonston 2006 Australian Indigenous Road Safety: 2005 update, Australian Transport Safety Bureau, Canberra (ATSB Research and Analysis Report, CR225).

Syme, S Leonard 1997 'Individual vs community interventions in public health practice: Some thoughts about a new approach', Health Promotion Matters (VicHealth) July(2): 2-12.

Tatz, Colin 1999 Aboriginal Suicide is Different, a report to the Criminology Research Council on CRC Project 25/96-7, Criminology Research Council, Canberra.

--2001 Aboriginal Suicide is Different: A portrait of life and self destruction, Aboriginal Studies Press, Canberra.

--2005 Aboriginal Suicide is Different: A portrait of life and self destruction (2nd edn), Aboriginal Studies Press, Canberra.

Thomson, Nell, Jane Burns, Amy Hardy, Ineke Amy and Sasha Stumpers 2007 Overview of Australian Indigenous Health Status, Australian Indigenous Health InfoNet, Perth.

Trewin, Dennis and Richard Madden 2003 The Health and Welfare of Australia's Aboriginal and Torres Strait Islander Peoples, Australian Bureau of Statistics and Australian Institute of Health and Welfare, Commonwealth of Australia, Canberra.

Vines, Pru 2007 'The sacred and the profane: The role of property concepts in disputes about post-mortem examinations', Sydney Law Review (University of New South Wales Faculty of Law Research Paper 13).

Wilson, Tony 1998 'Body rites--Aspects of possession and control of the dead human body', Australian Property Law Bulletin 13(5):48-51.

Wilson, Tom, John Condon and Tony Barnes 2007 'Northern Territory Indigenous life expectancy improvements, 1967-2004', Australian and New Zealand Journal of Public Health 31(2):184-8.

Belinda Carpenter

School of Justice, Faculty of Law, Queensland University of Technology

Gordon Tait

School of Cultural and Language Studies in Education, Faculty of Education, Queensland

NOTES

(1.) Chi-square test: [[chi square].sub.0.[infinity]] = 104>10.83

(2.) Chi-square test: [[chi square].sub.0.[infinity]] = 39.9>10.83

(3.) Chi-squaretest: [[chi square].sub.0.[infinity]] = 15.2>10.83

(4.) Chi-square test: [[chi square].sub.0.[infinity]] = 85.1>10.83

(5.) Chi-square test: [[chi square].sub.0.[infinity]] = 102>10.83

Dr Belinda Carpenter is an Associate Professor in the School of Justice, Faculty of Law, at Queensland University of Technology. She has been researching and teaching in the area of death investigation since 2004. Her other research interests include violent offending women, and sex, crime and morality.

<b.carpenter@qut.edu.au>

Dr Gordon Tait is a Senior Lecturer in the School of Cultural and Language Studies at Queensland University of Technology. His research interests include the philosophy of education, youth and governance, the sociology of death, and the rise of behaviour disorders and schooling.

<g.tait@qut.edu.au>

Table 1: Indigenous over-representation in the
coronial system, per 100 000 of the population

Category Indigenous Non-Indigenous
of death people people

Accidental 22.4 13.2
Suicide 20.8 14.4
Homicide 3.3 0.82
Medical 0 1.67
Natural 64.16 31.3

Table 2: Indigenous and non-Indigenous suicide in
the coronial system

 Significance
 level
 chi-square
 test:
 Non- [[chi square].
 Indigenous Indigenous sub.0.0]

Suicide by 23 (92%) 217% (40%) 8.07>6.64
hanging
All age groups

All suicides 18 (72%) 186 (34%) 5.55>3.84
35 and under

Suicide by 16 (88.8%) 92 (49.46%) 10.2>6.64
hanging
35 and under

Table 3: Suicide notes and 'mental illness': a
comparison of Indigenous and non-Indigenous
coronial deaths

 Significance
 level
 chi-square
 test
 Indigenous Non- [[chi square].
 Indigenous sub.0.0]

Suicide 8 (32%) 311 (57.3%) 6.26>3.84
notes

Mental 8 (32%) 284 (52.3%) 3.98>3.84
illness