Health service use in the year before death by suicide
Overall health service use
Among the 522 ex-serving ADF members who died by suicide (cases) over the study period, 85.6% accessed at least one of the five in-scope health services in their last year of life (12 months before death). This proportion was similar among the 2,497 other ex-serving members (controls), with 86.7% having accessed an in-scope health service. This was largely due to four in five cases (81.0%) and controls (81.9%) accessing an MBS service in the last year.
However, there were substantial differences between the proportion of each cohorts who accessed specific health service types at least once, with cases more likely to access:
- ED presentations (41% compared with 15%)
- hospital admitted care (35% compared with 14%)
- PBS (including RPBS) prescriptions dispensed (75% compared with 62%)
- DVA-funded MBS equivalent services (17% compared with 12%)
- MBS allied health attendances (32% compared with 27%).
Across the different health subservices analysed, GP services were accessed by the greatest proportion of ex-serving ADF members in the 12 months before death (or the matched index date): 77% of cases and 76% of controls.
In addition to analysing service use over the last year of life, AIHW examined health service use in the last month and last week before the index date (the date of death for cases or the matched date for date for controls). In the last month before the index date, 61% of cases accessed a health service, compared with 45% of controls. In the last week before the index date, 37% of cases accessed a health service, higher than the 20% of controls (see Figure 1).
Services such as MBS and PBS showed moderate differences between the cases and controls over the last year, but these gaps widened in the last month and even more so in the last week. Hospital services (ED presentations and hospital admissions), which were more frequently accessed by cases across the last year, exhibited the most striking differences in the most recent timeframes, with substantially higher use among cases than controls in both the last week and month before the index date.
Figure 1: Proportion of ex-serving members who accessed a health service at least once by those who died by suicide in the 12 months before death and other ex-serving members over the same period
Bar chart showing the proportion of ex-serving members who accessed health services in the 12 months before death or matched date, comparing those who died by suicide to those who did not.
Mental health services accessed by ex-serving ADF members were identified based on mental health prescriptions dispensed under the PBS and RPBS, Medicare-subsidised mental health services, mental health ED presentations, and mental health hospitalisations. However, it is important to acknowledge that prescriptions for mental health (psychotropic) medications may not always be linked to a formal mental health diagnosis. Some medications may be used for other conditions, and the presence of a prescription alone does not necessarily indicate that a mental health condition was identified or treated.
The proportion of cases who accessed mental health services was consistently higher than among controls across all time periods. In the year before the index date, 65% of cases accessed mental health services compared to 27% of controls. This difference was more pronounced in the last month (44% compared with 13%) and last week (19% compared with 5%) before the index date.
What characteristics are associated with health service use in the last year?
This section examines the likelihood of ex-serving members accessing health services in the year before death by suicide as opposed to not accessing the health service in the year prior. The outputs show which demographic, socioeconomic, geographic and service-related factors are associated with health service access. The likelihood was assessed for both cases and controls independently of each other.
AIHW analysed the likelihood of accessing any health service use, mental health services and specific health service types. This summary focusses on factors that showed notable and consistent associations, across both cases and controls, as well as those observed in cases only.
Higher odds of having accessed at least one health service were observed for (Figure 2):
- DVA clients
- females (for mental health and ED/hospital services)
- ex-serving members aged 65+ (for ED/hospital services)
- involuntary medical separated (for mental health, PBS and MBS/DVA-funded MBS equivalent services).
Lower odds of having accessed at least one health service were observed for (Figure 2):
- ex-serving members in regional/remote areas and in least accessible areas (based on shortest travel time to health services [see Technical notes] (effect observed in cases only; for any health service, and MBS or DVA-funded MBS equivalent services)
- most (SEIFA quintile 1) and least socioeconomically disadvantaged (SEIFA quintile 5) groups (cases only)
- aged 25–44 years (for PBS and non-mental health services).
Figure 2: Factors associated with health service use among ex-serving members who died by suicide (cases) and other ex-serving members (controls) in the 12 months before death (or matched index date)
Chart showing the odds ratios of health service use compared with non-use among ex-serving members who died by suicide and other ex-serving members.
For more information see Supplementary tables S4.1 and S4.2.
Frequency of health services used
The results in the previous section focussed on the proportion of ex-serving ADF members who accessed health services. However, it is also important to understand the frequency (or volume) of health service use across the full cohort. This section therefore examines the average number of services used by cases (ex-serving members who died by suicide) and controls. The average use was calculated based on all ex-serving members, not just those who used health services, to also consider the persons who did not access any services.
The average number of health services per ex-serving member was much higher among cases than controls. Cases used 55 health services in the year before death, 76% more services in the year in comparison to 31 services used by controls.
The largest relative differences in service volume between cases and controls were observed for:
- ED presentations (0.9 services per case compared to 0.2 services per control)
- hospital admitted care (0.9 services per case compared to 0.2 services per control
- specialist attendances (1.9 services per case compared to 0.8 services per control)
- PBS prescriptions dispensed (19.2 services per case compared to 10.3 services per control).
Health service | Controls | Cases | Relative difference |
|---|---|---|---|
MBS services | 13.3 | 21.3 | 1.6 |
– GP services | 4.4 | 7.1 | 1.6 |
– Specialist attendances | 0.8 | 1.9 | 2.4 |
– Allied health attendances | 0.6 | 1.1 | 1.8 |
DVA-funded MBS equivalent services | 6.7 | 12.1 | 1.8 |
PBS prescriptions dispensed | 10.3 | 19.2 | 1.9 |
ED presentations | 0.2 | 0.9 | 4.5 |
Hospital admissions | 0.3 | 0.9 | 3.0 |
Source: AIHW Veterans Health Dataset (VHD), July 2010–June 2020
Leading medication types dispensed and principal diagnoses in hospitalised ex-serving members
The most common reasons for hospital admission and most frequently dispensed medication types among cases and controls were analysed to identify key patterns in health service use. Comparing these factors provides insight into the differing health burdens and treatment needs of each group, highlighting areas for targeted interventions and resource allocation (Figure 3).
The principal diagnosis is generally the main reason someone needed to be admitted to hospital. AIHW analysed the top five most common principal diagnoses (at the International Classification of Diseases (ICD) Chapter level) for cases and controls, based on the number of hospitalisations.
Mental and behavioural disorders were the most common principal diagnosis among cases (47%) followed by injury and poisoning (15%). In controls, musculoskeletal and connective tissue conditions (18%) were most frequent, with mental and behavioural disorders ranking second (14%).
Medicines are organised into Anatomical Therapeutic Chemical (ATC) classification groups according to the body system or organ on which they primarily act. AIHW analysed the top five ATC3 groups (pharmacological subgroup) dispensed to cases and controls, based on the number of prescriptions dispensed.
Anti-depressants (24%) were the most common medication type dispensed to cases, followed by opioids (10%) and anxiolytics and antipsychotics (7% each). In controls, anti-depressants were also common (14%), but less so than in cases, while opioids were the second most common as with cases (9%), but there was a greater use of lipid-modifying agents (9%) and drugs for peptic ulcer and gastroesophageal reflux disease (8%).
Figure 3: Top five principal diagnoses and medication types dispensed to ex-serving members who died by suicide (cases) and other ex-serving members (controls) in the 12 months before death (or matched index date), from 2010-11 to 2019-20
Chart showing the top 5 hospital diagnoses and top 5 medication types dispensed to ex-serving members in the year before for those who died by suicide and other ex-serving members.
Overall, understanding the distinct patterns in each cohort highlight a higher prevalence of mental health conditions and associated treatments among cases, while controls were more likely to have chronic physical conditions.
Trends in health services accessed over the year
AIHW analysed monthly trends of cases and controls accessing health service over the year preceding the index date to identify patterns in service use. Across all the health services analysed, a higher proportion of cases accessed in-scope health services each month compared to controls (Figure 4). For cases, health service access peaked in the months leading up to the index date, with the highest levels observed in the last month before death.
Analysis of monthly trends showed that PBS prescriptions were the most commonly accessed services among cases, with the proportion of users ranging between 42% and 48% across the last year of life, compared to 29% to 31% among controls.
GP services were the second most commonly accessed service (ranging between 30% and 37%) for cases, with a slight increase to 37% in the month before death by suicide, while controls had stable rates at around 22% to 24%.
The proportion of cases who accessed hospital services (ED presentations and hospital admissions increased in the lead-up to death. There was a large spike in the proportion who presented to ED in the month before death (14% compared to 4 to 7% in the earlier months). The proportion who were admitted to hospital rose gradually from 5.7% to 7.3% in the last five months before the index date (Figure 4).
Figure 4: Proportion of ex-serving members who accessed health service by type in the 12 months before death (or matched index date)
Line chart showing the proportion of ex-serving members accessing health services in each month in the year before death or matched date, comparing those who died by suicide to those who did not.
Last health service accessed in the year before death by suicide
AIHW analysed the last health service used among ex-serving members in the case group who had accessed at least one health service in the year before death by suicide. Analysis of their last health service accessed showed that MBS or DVA funded MBS equivalent services were the predominant last accessed health service (52%), followed by PBS dispensed medications (38%). A relatively small proportion (9%) received hospital care as their final health service (ED presentation or hospital admission). These findings are similar to other studies (DelPozo-Banos et al 2024; John et al 2020).
Of those who accessed health services in the case group, a third (33%) had a mental health service use as their final health service accessed. This included those with a mental health diagnosis in ED or hospital admission, mental health-related prescription, or Medicare-subsidised mental health-specific services.
The median duration of time between final health service use and death by suicide was 10 days. However, the duration was smaller for hospital services and specialist care and longer for primary care and medication dispensation.
Overall, females, people aged 55 and over, people with length of ADF service of 10 years or longer, DVA clients and people with 2 or more comorbidities had a lower median time (between 5 to 8 days) between their last health service use and death by suicide.