The MHC and WA Health do not use the existing data effectively to inform how they deliver services, or to target changes in the mix of services. The MHC use data on activity, mainly instances of care, to monitor service delivery. However, activity data does not provide information on how many people used mental health services. It also does not provide insight on whether they are efficient and effective or if people are using them as intended. Focusing on people, and how they use services over time, could help the MHC and WA Health prioritise investment and service reform.
MHC and WA Health did not know how many people accessed mental health services
While activity data tells the MHC and WA Health how many times a service has been delivered, it does not tell them how many people accessed services, or the pathways they took through those services. This is because the data records instances of care as single events, rather than linking them together to show people’s episodes of care, which show how they use services.
The MHC and WA Health’s activity data tells them that 4.8 million instances of care were delivered across the various hospital and community treatment services from 2013 to 2017. These events included a range of care types, such as short phone calls to stays in a mental health hospital bed. It also included 253,000 EDs visits. What the activity data does not show the MHC and WA Health is that more than 212,000 people accessed mental health care from 2013 to 2017.
To identify that more than 212,000 people used mental health services between 2013 and 2017, we looked at people’s pathways in accessing care in WA Health’s hospitals, EDs and community treatment services. This allowed us to identify that people use combinations of services in a series of episodes often over many years. It also allowed us to identify particular groups of people who use services in a way that indicates that the service is not working as intended. This is not information that the MHC can generate from activity data.
For instance, by looking at people’s episodes of care we were able to see that one person was in hospital for 173 consecutive days across 3 hospitals (Figure 7). The MHC and WA Health would have captured this as 6 separate instances of care; 3 ED visits and three hospital stays. For reporting purposes, it would not have been clear to the entities if this was 1 person or up to 6 people.
This pattern of access suggests that the person was not able to get the right level of care in the first regional centre. After an extended stay in the larger regional hospital they required further care and were eventually transferred to a major metropolitan hospital. Measuring instances of care individually, such as this person’s hospital stays, would not provide an accurate reflection of their overall care. Reviewing episodes of care allows entities to better understand people’s pathways through mental health services, improving their ability to deliver people-centred services most efficiently and effectively.
Existing activity data does not enable MHC to know if there is sufficient service capacity to reach all the people who need it
As the basis for the Plan, the MHC estimated that around 3% of the WA population would require State-managed care for severe mental health issues each year. The model was based on the National Mental Health Service Planning Framework adapted to take into account WA’s population and geographical distribution. This allowed the MHC to build estimates of the service capacity required to meet demand. The Framework estimates the volume of services required, taking into account that people may present to services multiple times. However, the MHC does not regularly assess the actual number of people accessing care against the estimated number, and so is not identifying if there is an overall service gap.
Our analysis indicates that a lower than forecast number of people are accessing community treatment services and hospital mental health care. This could indicate a few things. If the estimate of need is reliable, it may suggest that the volume of available State-managed care is not sufficient or people could be accessing services elsewhere. It could also mean that there are too few options to meet people’s needs, although the service gap was smaller in 2017 than 2013. For example, in 2013, 54,800 people accessed State-managed mental health care, 24,300 less than the MHC estimated. In 2017, the figure was 63,700, which was 16,500 less than the estimate (Figure 8). Monitoring this information would give the MHC better information on whether they are reaching more people by introducing new services or expanding existing services. It would also inform the MHC on whether their ongoing estimates of need are proving to be accurate, and enable them to adjust the Plan and its roll-out.
Analysing pathways shows that 10% of people used 90% of in-patient care and 50% of emergency and community care
By using WA Health’s hospital, emergency and community treatment service data from 2013 to 2017 and looking at pathways we were able to identify clear patterns in how services were being used by groups of people. By comparing them to how the MHC and WA Health expect services to work, we were able to identify patterns of use that indicate that services are not working as intended for some groups. This information can inform more efficient and effective management of services and help to prioritise investment in service reform.
Of the 212,000 people who accessed State-managed mental health care, 30% of these people only accessed care once. This suggests that for this group either their mental health issues were resolved, or that they were accessing care through alternative providers which can include primary care (general practitioners), private services or non-government organisations care. Although we did not have access to data from primary healthcare, private mental health services or non-government organisations services to verify this, this group are not the most numerous or resource-intensive consumers of State-managed mental health services.
The main consumers of State-managed services are the 10% or 21,000 people who, over the 5 years we reviewed, consumed 90% of the hospital care provided and around half of the community treatment services and ED mental health care (Figure 9). This is one of the key groups of people who the MHC and WA Health should focus on understanding so they can provide pathways that enable these people to spend as much time as possible in the community and then move through more intensive services as they need to. This approach would reduce the amount of time these people spend in very expensive hospital beds and EDs.
Older adults are over-represented in the 10% of people who accessed the most in-patient mental health care
Within the top 10% of people who accessed the most hospital care, older adults (people aged 65 and over), were over-represented. Older adults represented 15% of all the people who had accessed mental health services between 2013 and 2017. However, they represented 25% of the 10% of people who had used the most in-patient hospital care (Figure 10). Their episodes of care tend to show long-stays in an in-patient setting, indicating a lack of community-based care options.
Reviewing the most common diagnoses of older adults in this group showed that they often required care for Alzheimer’s, delirium and dementia. Clinicians reported that the consumption of hospital care by this group of older adults was in part due to a shortage of aged care services equipped to manage people with mental health issues, and limited clinically appropriate residential care settings in the community. The MHC reported that the State and Commonwealth governments are also working together to introduce a specialist dementia program for this cohort.
Understanding people’s pathways in accessing care would allow the MHC to target, prioritise and demonstrate the expected benefits of investment under the Plan to change the mix of mental health services. Being able to find alternative care settings for this group would allow them to access care in a more appropriate setting, and would effectively increase the available capacity of hospital beds, improving access for those who need them.
Current KPIs do not meaningfully inform MHC on the outcomes of care
The MHC reports on the rate of readmission to hospital within 28 days of discharge from a specialised mental health ward as an indicator of the effectiveness of care. The MHC also reports the percentage of occasions that someone is contacted by community treatment services within 7 days of discharge, to assess the effectiveness of follow-up and service integration. These are standard nationally-reported metrics however, neither measure provides a robust view of actual performance or outcomes for people in the system. Rather, they provide an indication of when the system may be under pressure, prompting further review by the MHC:
- The 7-day follow-up measure counts whether someone leaving hospital is contacted by community mental health services within 7 days. However, counting contacts has a number of limitations in measuring whether there is effective integration between services, and whether people access the care they need. The measure does not capture whether the person gets an appointment with community services in a reasonable timeframe and whether they keep that appointment, both of which impact whether that person accesses care.
- Second, counting the contacts does not provide a view of the quality of the contact. Face-to-face contact with the community treatment service is preferred, as this is likely to be more effective in helping someone to access services. In practice, they more frequently use phone calls, which often come from the hospital, not the community treatment service. Across the 5 years of data we analysed, follow-up phone calls increased from 34% to 60%, while face-to-face contacts fell from 60% to 30%. The KPI does not identify this.
The MHC also monitors the readmissions to hospital within 28 days of discharge from specialised mental health units. This KPI currently counts how many readmissions occurred within 28 days of discharge from a hospital bed. Some patients are readmitted as part of their care plans, but the measure and data do not distinguish between planned and unplanned readmissions. The national reporting framework acknowledges that there are data limitations that prevent the ability to differentiate between planned and unplanned readmissions.
Analysing people’s pathways, rather than counting these events, would allow the MHC to measure useful indicators. These could include how many people accessed community services within a given time following a hospital stay, if contact was by phone or in person, and whether contacts resulted in subsequent access to care. This would provide a better measure of service integration. It would also inform the MHC whether people’s access to services follow the models of care, and enable a more sophisticated view of readmissions within 28 days.