Patterns of health service use during the last year of life for those who died by suicide

The final set of analysis conducted by AIHW was to examine how patterns of health service use evolved over time during the year before death among persons who died by suicide. This longitudinal analysis focused exclusively on suicide cases to identify trajectories in the timing and intensity of service use.

By showing different pathways of service use leading up to death, the analyses aimed to identify escalation patterns, periods of intense health service use, or disengagement from health services. Understanding these trends may provide insights into critical time periods for intervention and how health service use may change or intensify before death by suicide.

The analysis was conducted through group-based trajectory modelling and limited to ex-serving ADF members who died by suicide between 1 July 2014 and 30 June 2020.

The results of the trajectory modelling were a series of trends for subgroups of ex-serving members grouped by their patterns of health service use in the year before death. The number of subgroups varied dependent on the health service type. Descriptive labels (such as low stable or high variable) were assigned to these groups based on the observed patterns of health service use within each group. These labels are interpretative summaries to support communication and understanding and were not based on pre-defined thresholds. 

Three main health service types were examined: MBS, PBS/RPBS and hospital (ED and admitted care combined). Each health service type was further categorised into mental and non-mental health services.

Figure 8 presents the trajectories of each group by each health service type, for mental, non-mental and any health service. It shows that there were four different user groups for any MBS services, three for any PBS services and two for any hospital services.

Figure 8: Trajectories of health service use among ex-serving members who died by suicide in the year before death

Line chart showing patterns of health service use over 12 months before death among ex-serving members who died by suicide, highlighting different trajectories of service use over time. 

Line chart showing patterns of health service use over 12 months before death among ex-serving members who died by suicide, highlighting different trajectories of service use over time. 

Patterns of MBS service use

When split by the patterns of MBS service use in the year before death by suicide, there were four groups of ex-serving members who died by suicide. Each of the four groups are described as follows:

  • Low stable: the lowest users of MBS services, with no variation over time. This group accounted for 31% of the cohort.
  • Medium stable: relatively stable and medium level of MBS service use throughout the 12 months. They represented 40% of the cohort.
  • Medium increasing: moderate MBS service use overall, with steady increase over the 12 months. They represented 19% of the cohort.
  • High variable: high MBS service users, peaking around the 10 months before death followed by a gradual decline prior to death. This group comprised 9% of the cohort.

These patterns were broadly similar based on MBS non-mental health services. However, when analysing mental health MBS service use there were only two user groups: a low increasing group (84%) and a smaller high decreasing group (16%).

The distribution of characteristics (demographics, service-related and health-related) across the four MBS (mental and non-mental health combined) service use groups were examined. The results for the two higher use groups shown in Figure 9 indicate which characteristics were more commonly observed in each group, based on their relative proportions.

Figure 9: Characteristics of MBS service user groups

Heatmap showing how each subgroup is distributed across MBS service user groups, with proportions showing the distribution within each subgroup.

Heatmap showing how each subgroup is distributed across MBS service user groups, with proportions showing the distribution within each subgroup.

Supplementary tables S8.1 to 8.4a show the distribution of ex-serving members by characteristic within each MBS service user group.

Patterns of PBS service use

When split by the patterns of PBS service use in the year before death by suicide, there were three groups of ex-serving members who died by suicide. Each of the three groups are described as follows:

  • Low stable: the lowest users of PBS services. This group accounted for 49% of the cohort.
  • Medium stable: ex-serving members with relatively stable and medium level of PBS service use throughout the 12 months. They represented 40% of the cohort. 
  • High variable: this group had high PBS service use with a small decline before use peaked at 4 months prior to death with a further decline until the month prior to death. This group comprised 11% of the cohort.

There were also three user groups with broadly similar patterns based on PBS mental health services. Although the PBS mental health high user group was more stable over the year. There were only two user groups based on non-mental health PBS service use: a low stable group (74%) and a smaller high variable user group (26%).

The distribution of characteristics (demographics, service-related and health-related) across the three PBS service user groups (mental and non-mental health combined) were examined. Figure 10 indicates which characteristics were more commonly observed in the two higher use groups, based on their relative proportions.

Figure 10: Characteristics of PBS service user groups

Heatmap showing how each subgroup is distributed across PBS service user groups, with proportions showing the distribution within each subgroup.

Heatmap showing how each subgroup is distributed across PBS service user groups, with proportions showing the distribution within each subgroup.

Supplementary tables S8.5 to 8.7a show the distribution of ex-serving members by characteristic within each MBS service user group.

Patterns of hospital service use

There were two groups of ex-serving members who died by suicide when split by the patterns of hospital service use (including admitted care and ED care) in the year before death by suicide. 

The low stable group accounted for 82% of the cohort and maintained stable low hospital service use throughout the period, showing only a modest increase in the final months. The smaller high variable group (18%) exhibited a more complex pattern, characterised by moderate to high hospital service use and then a marked escalation in hospital service use in the month preceding death. 

There were also two user groups with broadly similar patterns based on hospital mental health services. However, the hospital mental health high user group was smaller (5%) and had mental health hospital service use peak 9 months before death by suicide instead of the month preceding death. There were also two user groups based on non-mental health hospital service use. The low stable group was steady over the year with no increase in non-mental hospital service use before death.

AIHW analysed the distribution of characteristics (demographics, service-related and health-related) across the two hospital service user groups (mental and non-mental health combined). Figure 11 shows which characteristics were more commonly observed in each group, based on their relative proportions. 

Figure 11: Characteristics of hospital service user groups

Heatmap showing how each subgroup is distributed across hospital service user groups, with proportions showing the distribution within each subgroup.

Heatmap showing how each subgroup is distributed across hospital service user groups, with proportions showing the distribution within each subgroup.

Supplementary tables S8.8 to 8.10a show the distribution of ex-serving members by characteristic within each ED/hospital service user group.

Previous page
Next page