Data sources

The report used a range of linked data assets.

Table B. 1: Description of datasets

Data source

Description and Scope

ADF population


Personnel Management Key Solution


The PMKeyS data defines the scope of the ADF population for the current analysis. It is a Defence staff and payroll management system that contains demographic and ADF service information for those with at least 1 day of ADF service since 1 January 2001. Name-based linkage was conducted between the PMKeyS, DVA client data, MEF and NDI.

Excludes those who separated prior to 1 January 2001, and those hired after 31 December 2018.

Suicides in scope for the Review are those that occurred between 1 January 2001 and 31 December 2018.


ADF personnel with at least 1 day of ADF service since 1 January 2001.

For current serving and Reserve members, data from the PMKeyS is reported as at the extract date (September 2020). As the PMKeyS data is stored destructively (new information overrides old information), it is not possible to determine the service characteristics (e.g. rank) as at 31 December 2018 for current serving and Reserve members. This results in a difference between the numerator and the denominator of approximately 2 years for disaggregation by service characteristics. This means that for disaggregation by some service characteristics, the scope of the denominator exceeds the numerator and in turn, rates of suicide are underestimated. As such, rates of suicide disaggregated by these characteristics will only be reported for ex-serving members.

For ex-serving members, service characteristics are reported as at separation, and do not change thereafter.

For the ex-serving population time at risk while holding given service characteristics is equal to time in the ex-serving population


Defence health and transition data

The Defence health and transition data is a collection of data sets held across several areas of Defence.

The data contained in these databases are by-products of administrative processes including those related to Defence health service provision, post-deployment mental health screening, WHS reporting and transition support services.



Health services contract data

The data set does not include records prior to 2012. It does not capture health service use before or after ADF service.

Additionally, it does not capture health services used outside of ADF arrangements.

Pharmacy dispensing record

The data set does not capture pharmacy dispensing before or after ADF service.

Data is limited to dispensing by Defence pharmaceutical chain on base, excludes off base dispensing.

It does not include records prior to 1997.

Does not capture pharmacy dispensing used outside of ADF arrangements.Work health and safety (WHS) data

As the data is collected from the supervisor of the involved ADF member, it is self-reported data.

Analysis of trends over time is difficult due to changes in reporting. System improvements and confidence in reporting have contributed to an overall increase in WHS events. Changes in military training, operations, event support, and WHS legislation also contribute to clusters at certain time points.

Each incident is classified into 6 severity types. Three of these (fatality, dangerous incidents and serious injury/illness) are notifiable to Comcare and Defence under the WHS Act. Exposures and minor injuries are more frequent but not required to be reported to Comcare. As a result, data quality has a tendency to be better in the notifiable incidents and there is higher confidence that they are captured.


Defence Suicide Database

The DSD contains information on confirmed and suspected suicides that occurred during active ADF service, since 2000.

Confirmed and suspected suicides of active ADF personnel since 2000.

The database does not include deaths among Reserves or ex-serving ADF members.

Department of Veterans’ Affairs client data

The DVA client data will be used to determine whether making a DVA claim, or receiving a particular claim outcome, is associated with increased or decreased risk of suicide, with a particular focus on clams for mental health conditions.

DVA clients

There have been significant DVA policy changes over the study period. In particular, there has been increasing access to non-liability mental health care. Initially this was only for selected conditions, but since July 2016 all current and former ADF personnel have been entitled to non-liability health care for all mental health conditions.

Since July 2018 all personnel discharged from the ADF have been automatically issued with a DVA health card (formerly a white card) that entitles them to non-liability health care for any mental health condition for the rest of their lives. ADF members are considered DVA clients from the date they first use this card.

Due to these policy changes, it is expected that the number of DVA funded mental health treatments will have significantly increased over the study period. This does not necessarily indicate an increase in the prevalence of mental health conditions in the ADF population.

The proportion of ADF members who are DVA clients (and white cardholders) is also expected to have increased since 2016.

Data range: 2001 to 2018

DVA National Treatment Account

The DVA NTA is an administrative data set containing records of health services provided to DVA card holders and funded by DVA. The relationship between the MBS and National Treatment Account is similar to that between the PBS and Repatriation Pharmaceutical Benefits Scheme (RPBS): similar services are available under both schedules, but services claimed by DVA card holders are funded through DVA’s NTA.

DVA clients/card holders

As the MBS claims data does not include services funded by DVA, the level and patterns of all government-subsidised medical service use among DVA clients cannot be determined using the currently approved MBS variables alone.

DVA NTA data has not previously been combined with MBS data. As such, it is unclear how well the data items compare across the two data sets. Data range: 2000 to 2018

The DVA hospitals data includes diagnoses using ICD-10.4, which is not the most recent version of this classification scheme. Some diagnoses recorded using a more recent version of ICD-10 are not included. Data range: 2000 to 2018 for admitted patient care and 2015−16 to 2018−19 for emergency department data.

Medicare Enrolment File

Contains identifying information for all those registered in the Medicare system since 1984.

To be used as an intermediary data set to supplement linkage information on PMKeyS and DVA client data for subsequent linkage to the NDI.

Using a series of map files, it will allow merging to:

  • MBS
  • PBS

Demographic information is also requested from the MEF for analysis.

All ADF members enrolled in Medicare.

Only includes persons enrolled in Medicare.

Medicare Benefits Schedule (MBS)

Contains records of all health services funded under MBS and corresponding claims data.

It will be used in conjunction with DVA NTA data to determine whether receiving particular kinds of medical treatment is associated with an increase or decreased risk of suicide.

ADF members who used MBS funded health services.

Claims data does not include:

  • Services where no benefit was processed (even if it is eligible for a rebate)
  • Services provided to public patients in hospitals
  • Services subsidised by the DVA
  • Services delivered in public outpatient departments/public accident and emergency departments
  • Services for an injury/illness for which the patient’s insurer has accepted liability
  • Non-hospital services subsidised by private health insurance
  • Services provided through other publicly funded programs, including jurisdictional salaried GP services provided in remote outreach clinics
  • Health screening services

Claims data identifies health service use but does not provide corresponding diagnostic information.

Use of Medicare-subsidised services targeting specific health conditions do not reflect quality of care or prevalence of health conditions.

If a patient used other forms of health care (eg. Standard GP consultations) to manage their health conditions, it will not be reflected in the database.

Pharmaceutical Benefits Scheme (PBS)/Repatriation Pharmaceutical Benefits Scheme (RPBS)

These are two main Australian Government subsidy schemes for medicines, and their dataset provides a record of prescriptions that are dispensed under them.

ADF members who received prescriptions under these schemes.

Data does not cover:

  • Pharmaceuticals that are prescribed but not dispensed
  • Pharmaceuticals not listed on the PBS
  • Over-the-counter purchases (non-prescription)
  • Private prescriptions
  • Medicines supplied to admitted patients in public hospitals

In data prior to April 2012, dispensing data is only available for prescriptions that attracted a government subsidy (i.e. Cost of prescription exceeded co-payment threshold).

Safety net threshold of these schemes ($390 for concession and $1550 for general beneficiaries as of 2019) when met, allows provision of free pharmaceuticals to the consumer, which means that before July 2012, all medication dispensed to general beneficiaries will be captured only once this safety net is met.

National Death Index

Contains fact of death information from 1980 onwards, cause of death information from 1985 onwards and associated causes of death from 1997 onwards.

It will be used to determine vital status of those in the study and control population, and identify those whose cause of death was suicide or sequelae of intentional self-harm (ICD-9 codes: E950-E959; ICD-10 codes: X60-X84, Y87.0).

ADF members who died by suicide from January 2001 to December 2018.

Cause of death data may change where a death is being investigated by a Coroner and more up to date information becomes available as a result of the ABS revisions process.

  • Lag in cause of death information for the most recent year of data, where a death is registered in the following year. Analysis in this study is based on year of occurrence of death. The NDI is the source of information on fact of death in this study.
  • The NDI contains fact of death information linked with cause of death information from the National Mortality Database (NMD). Analysis of the NMD for all Australian deaths shows that between 4% and 7% of deaths are not registered until the next year. These deaths are not captured in cause of death information, until data for the next year become available.

ABS Psychosocial coding data

Psychosocial coding for ADF members in study cohort who died by suicide on the basis of their NCIS files.

Used to investigate psychosocial factors associated with suicide among ADF members (i.e. Factors relating to social process/structure that could influence individual thought or behaviour that impact health outcomes).

ADF members who died by suicide.

Coding of psychosocial causes of death will only be undertaken for ADF members who died by suicide, and Australians who died by suicide between 2017 and 2019, making it impossible to compare to wider ADF members or Australian populations.

Some psychosocial risk factors do not align with the ICD-10 codes or are part of a non-specific code. These cannot be easily analysed once tabulated.

Differences in collection of psychosocial factor information across coronial jurisdictions produces inconsistencies in coverage across Australian deaths by suicide.

Some cases of deaths by suicide remain open in the coronial system, for which psychosocial coding will be incomplete. The NCIS is continually updated as cases are closed and quality assured. As new information comes to light, historical cases may be re-opened by a coroner. Occasionally this results in changes to data and published mortality figures over time.

National Coronial Information System

National repository containing data on deaths reported to a coroner in Australia and New Zealand. Sourced from coronial briefs provided by each jurisdiction, created as part of the investigation conducted by a coroner into the death of an individual.

Investigations determine the identity of the deceased and the cause of death. International Cause of Death (ICD-10) coding is provided by the Australian Bureau of Statistics and the New Zealand Ministry of Health.

The AIHW will receive coronial files specific to suicides in the ADF population, and suicides in the general Australian population that occurred between 2001 and 2018. AIHW will analyse coded data from the NCIS online system. These variables include those relating to the demographics of the deceased, details of the incident preceding death, the mechanism of injury, and cause(s) of death.

Data will be used to explore how the circumstances of the individual preceding death relate to suicide risk, and analysis of the details of the death may inform environment-specific prevention strategies.

ADF members who died by suicide between 2001 – 2018.

Data is collected from primary source material such as the police notification of death report, autopsy, and toxicology reports and coronial findings from each jurisdiction. The level of detail contained in these documents varies between and within each jurisdiction. Institutional practices and legislative differences impact the information collected and reported by each jurisdiction. Therefore, there will be differences in the qualitative and quantitative comprehensiveness of the data.

The NCIS is continually updated as cases are closed and quality assured. As new information comes to light, historical cases may be re-opened by a coroner. Occasionally this results in changes to data so there may be some changes to published mortality figures over time.

Cases contained on the NCIS can be restricted at the discretion of the State or Chief Coroner. Coroners may restrict access to cases that are particularly sensitive in the community.


National Health Survey

The 2017–18 National Health Survey (NHS) was conducted by the Australian Bureau of Statistics (ABS) from July 2017 to June 2018.The 2017–18 NHS is the most recent in a series of Australia-wide health surveys conducted by the ABS. It was designed to collect a range of information about the health of Australians, including:

  • prevalence of long-term health conditions
  • health risk factors such as smoking, overweight and obesity, alcohol consumption and physical activity
  • use of health services such as consultations with health practitioners and actions people have recently taken for their health
  • demographic and socioeconomic characteristics.

Respondents were also asked:

  • whether they had ever served in the Australian Defence Force
  • if they were a client of DVA
  • if they had ever received a benefit or support from DVA.

This information was used to identify the estimates of ADF members and DVA clients in Australia as well as selected demographics and socioeconomic status in this report.

Alive ADF members 2017 - 2018

Alive ADF members are those people who responded to the National Health Survey, 2017-18, and identified as having served in the ADF. They will include ADF members who served prior to 2001.

The question relating to DVA benefits was asked regardless of whether the respondent was a current or former serving member of the Australian Defence Force.