Characteristics of ex-serving Australian Defence Force members hospitalised for suicidality and intentional self-harm

Background

In 2017 the Department of Veterans’ Affairs (DVA) and Australian Institute of Health and Welfare (AIHW) established a partnership to build a comprehensive profile of the health and welfare of Australia’s veteran population.

AIHW has partnered with DVA and Defence to establish a new data asset to provide greater insight into the health service needs of ex-serving ADF members. This data asset allows the service use patterns for individuals (de-identified) to be tracked over time (for more detail see Technical notes).

This project describes the use of certain healthcare services by ex-serving ADF members based on this data asset. This report specifically examines the use of public hospital services for conditions associated with suicidal behaviour (mental health, alcohol and drug use and intentional self-harm) by ex-serving ADF members, including those who died by suicide. Some private hospital data is also included in this report where it was available.

Given 20% of deaths by suicide are linked to a history of self-harm (ABS 2021), understanding hospital-based care for conditions associated with suicidal behaviour may identify intervention opportunities and inform suicide prevention strategies.

AIHW acknowledges that the data presented in this report represent human lives and we acknowledge all of those serving and ex-serving ADF members who have died by suicide. We also acknowledge all of those who have been affected by suicide or intentional self-harm. We are committed to ensuring our work continues to inform improvements in mental health, and suicide awareness and prevention.

What is included in this report

In this first report using the new data asset, AIHW has focussed on exploring the public hospital care use of ex-serving ADF members for conditions related to suicide and suicidality, including those admitted to hospital as well as those presenting to emergency departments. More specifically, AIHW has used the data asset to establish what proportion of ex-serving members who received public hospital care in 2019-20 and received that hospital care primarily for conditions related to suicidal behaviour.

For example, the proportion of ex-serving ADF members admitted to public hospitals for mental health-related care was calculated by taking the number of ex-serving members who were admitted to public hospital at least once during the period primarily for mental health-related care and then dividing this by the total number of ex-serving members who were hospitalised during the period.

Importantly, because this analysis only includes people that received care from a hospital, the data cannot be used to provide an indication of the incidence or prevalence of conditions associated with suicidal behaviour within the populations studied or the relative risk of suicide or suicidality. It also doesn’t provide information on the unmet care needs of those who could have benefited from hospital care but did not receive it.

Incidence (or hospitalisation) rates are recognised as a useful statistical measure of hospital care utilisation, however are not presented here, due to data limitations with gaps in information from some jurisdictions (outlined further below).

This report compares the characteristics of those ex-serving ADF members who accessed hospital services (hospital admissions and ED presentations) and those who died by suicide with the general Australian population who accessed the same sorts of services. While comparisons with other specific population groups, (such as those who have faced similar trauma including first responders) or those who have similar socio-economic characteristics (including housing circumstances, employment status, family structure) could offer valuable insights, data constraints limit such analysis at present.

Further information on population scope, analysis period and methodology can be found in Box 1 and the Technical notes. 

In this report, the term ‘mental health-related care’ is used to refer to a range of psychological disorders consistent with AIHW reporting of mental health admitted care for the Australian population, while ‘mental and behavioural disorders’ is used specifically when discussing diagnosis in line with international classification systems such as the ICD (International Classification of Diseases).

Who is included in this report?

The scope of the analysis for this report is all people who had:

• accessed admitted patient care services at a participating public hospital in Australia (excluding non-DVA clients admitted to public hospitals in Western Australia and Northern territory) from 1 July 2010 to 30 June 2020
• been admitted to a participating public or private hospital in Queensland during the study period from 1 July 2010 to 30 June 2020
• presented to a participating hospital Emergency Department (ED) from 1 July 2013 to 30 June 2020.

These data include around 269,000 ex-serving ADF members (who had served at least one day since 1 January 1985), of whom 226,800 were males and 42,100 were females. Between July 2010 and June 2020, 36% of ex-serving males (81,300) and 40% of females (16,900) received admitted patient care for any condition at a participating public hospital. 

Between July 2010 and June 2020, 600 ex-serving males and 68 ex-serving females died by suicide. Of ex-serving members who died by suicide, 275 males (46%) and 40 females (59%) had received admitted patient care at a participating public hospital prior to their death during the 10-year period.

The small number of observed deaths by suicide among females limits analysis and interpretation of data for females who died by suicide.

Between July 2013 and June 2020, there were around 117,300 ex-serving ADF members who attended a public hospital ED for any reason, this included 98,900 males and 18,400 females.

Further information on the data sources and population scope can be found in Box 1 and the Technical notes. Definitions of the terms used in this section are available in the Glossary.

Notes on measuring the use of admitted patient and emergency department care

There are some factors that affect the analysis of admitted patient care and ED care in the ex-serving population and the comparable Australian population. These include:

• differences in the age structure of the population groups
• the Australian comparator population includes the ex-serving population
• counts of patients were computed corresponding to the disaggregation of year (i.e. analysis by year counted patients per year while counts across the whole analysis period would only count each patient once for the entirety of that period)
• lack of data on private hospital admitted patient and emergency department care
• to a lesser extent, lack of data on public hospital admitted patient and emergency department care (explained below)
• emergency department care and public hospital admitted patient care were analysed separately, and overlapping admissions were not excluded. This should be considered when interpreting the results. 

Data on admitted patient care was available for 2010-2020 while ED care was for 2013 to 2020.

Age structure of the ex-serving member population

The ages of ex-serving members are different to the ages of persons in the whole of Australia population. The ex-serving population has a smaller proportion of older persons and a smaller proportion of younger persons, with a higher proportion of persons aged between 35-54 years and also males aged 55 to 64 years (see Figure 3 in Technical notes). This is a result of the ex-serving population only including persons who have served since 1985, a small share of persons having already separated from the ADF at younger ages and historical differences in female recruitment.

The differences in the age structure between the ex-serving ADF member population and the total Australian population should be considered when interpreting health-related findings. Many health conditions and diseases are directly associated with aging. In contrast, conditions related to suicidal behaviour such as mental and behavioural disorders, alcohol and other drug use and intentional self-harm are more prevalent among younger cohorts (see Intentional self-harm hospitalisations by age groups).

Due to the differences in age structures between the ex-serving population and Australian population and the links between age and health conditions, AIHW has made comparisons with the Australian population based on age groups to control for these differences. Where most of the age group comparisons show the same pattern, greater confidence can be had in using all age comparisons.

Data limitations

The analysis for this report was conducted using multiple linked administrative datasets held in a bespoke data asset built by the AIHW. Comprehensive information about the data sources used for this analysis is available in the Technical notes.

There are some challenges and limitations that affect any type of analysis using linked administrative data including:

• bias from linkage errors where records cannot be linked
• data sources of varying quality
• limitations in the time periods to which data are available.

Specific challenges relating to this report were the lack of private hospital data across most jurisdictions and public hospital data for non-DVA clients for two jurisdictions (Western Australia and Northern Territory).

Lack of hospital data

Overall, half of all episodes of admitted patient care for mental health-related care is provided by private hospitals (AIHW Admitted Patients 2021-22). This proportion varies by condition and population with national data showing that more admitted patient care is provided by private hospitals for depressive episodes and mental and behavioural disorders due to alcohol and other drug use (see Admitted patients mental health-related care).

Private hospital services are not universally available and most are funded by the patient’s private health insurance and/or self-funding - meaning that people from higher socioeconomic groups have higher rates of private hospital admissions (see Understanding the wellbeing characteristics of ex-serving ADF members and Australia’s hospitals at a glance). Notably for ex-serving members, private hospital care is also available through workers compensation arrangements, the Department of Defence for serving ADF members and DVA for eligible ex-serving members.

At the time of this report only Queensland had both public and private hospital data that could be used to compare ex-serving and Australian population use of hospital care. In comparison, all jurisdictions except for Western Australia and Northern Territory had sufficient public hospital data to compare ex-serving and Australian population use of public hospital admitted patient care.

ED data has similar limitations to admitted patient care data. However, most care in ED takes place in public hospitals so the lack of private hospital data is likely to have had a smaller impact on the analysis of ED care.

As a result of the limited private hospital data, the focus of the report is on public hospital admissions. However, to complement the main analysis AIHW also conducted analysis based on data from Queensland to provide a more complete picture of public and private hospital admissions for conditions relating to suicidality. The Queensland comparison is used as a form of case study to better understand the downward bias that was present in the public hospital only analysis and to understand if trends changed given the inclusion of private hospital data based on Queensland. The Queensland results are presented in the Technical notes.

Partial data on Western Australia and Northern Territory

AIHW also used data from DVA for admitted patient and ED care for the analysis. This included care delivered in hospitals across all states and territories including those for which there was no public hospital data through the National Health Data Hub (NHDH). This has enabled the addition of some Western Australian and Northern Territory hospital data (for eligible DVA clients only) but would not be considered representative of hospital data not captured in NHDH.

See the Technical notes for more information on the NHDH and DVA data.

Measures used in the report

Information about the use of admitted patient and emergency department care by the ex-serving population and comparisons with the broader age-matched Australian population is presented using counts, proportion differences and relative differences. Proportion difference is the absolute difference in percentage points between the two populations. Relative difference is the ratio of proportions and measures the scale or extent of the difference.

When calculating counts and proportions, ex-serving ADF members who were admitted multiple times in a period were only counted once each period. In terms of the proportions, if an ex-serving ADF member was admitted multiple times in the same period and at least one admission was primarily for mental health-related care, they were counted as a mental health-related care admission in the numerator (and denominator) whereas if an ex-serving ADF member was admitted multiple times in the same period and no admissions were for mental health-related care, they were included in the denominator.

The same process was taken for all people receiving public hospital care during the period and the proportions compared to provide additional context to the ex-serving member experience.

Confidence intervals (CI) of 95% are provided for the proportion and relative differences to indicate the level of certainty around these estimates due to random fluctuations associated with small numbers. A narrow confidence interval indicates more certainty in the result. Statistically significant differences between ex-serving ADF members and all Australians are not necessarily explained by prior ADF service and may be explained by other factors (such as socio-demographic factors) not accounted for in this report. 

Suicide is a relatively rare event compared to other causes of death. Admission to hospital for intentional self-harm, while more frequent than suicide, is also not common. Small numbers can raise privacy and confidentiality concerns and affect statistical methods and reporting capability. Values based on small numbers have therefore been suppressed to maintain confidentiality and to avoid publishing statistics of low reliability. Statistics based on small numbers of suicides should be interpreted with caution.

Hospitalisation rates of the ex-serving population could not be determined at the time of reporting due to a lack of data on the size of the ex-serving population in each jurisdiction. AIHW is currently exploring other ex-serving population data sources to support these analyses in the future.

Statistics in this report are presented for two distinct periods. Analysis of admitted patient care in the year before suicide used data across the entire study period of 2010-2020. In contrast, analyses of admitted patient care for conditions associated with suicidal behaviour focussed on 2019-20, as prior years revealed similar patterns (See Supplementary Tables for years 2010-11 to 2019-20). The ED analysis focuses exclusively on the entire period for which data was available, being 2013-20.

More information is available in the Technical notes.