Accessibility and responsiveness of Better Access treatment services: Insights from the use of linked administrative data in the evaluation of Better Access

Abstract

Objective:

This study was conducted as part of an evaluation of the Better Access initiative. It examined (1) pathways into Better Access treatment; (2) the proportion of Better Access treatment users who are ‘new’; (3) patterns of use and non-use of Better Access treatment services in relation to need; and (4) socioeconomic differences in Better Access treatment service use.

Methods:

We used linked administrative and survey data available through the Person-Level Integrated Data Asset (PLIDA). More specifically, we used Medicare Benefits Schedule (MBS) data, Pharmaceutical Benefits Scheme (PBS) data, 2016 Census data, and data from the 2017/18 National Health Survey.

Results:

About two-thirds of individuals who have had a mental health treatment plan prepared for them receive Better Access treatment services (albeit often after a considerable wait), but one-third do not. Although Better Access is reaching those with high levels of need, access is not equitable. It is harder for new users to access the programme than it was previously, as the number of continuing users and the number of treatment sessions provided to them have increased. People on low incomes are less likely to receive psychological treatment through Better Access (and more likely to be prescribed antidepressant or anxiolytic medication), and if they do receive Better Access services, they typically wait longer than their high-income counterparts to see a provider.

Conclusion:

Better Access appears to be responsive to need, but there are equity issues regarding its accessibility. These equity issues should be addressed as Better Access continues.

Introduction

The Australian Government introduced the Better Access to Psychiatrists, Psychologists and General Practitioners through the Medicare Benefits Schedule initiative (Better Access) in 2006 to improve access to mental health care for people with common mental disorders. Better Access makes rebates available through a series of items on the Medicare Benefits Schedule (MBS). The core items provide rebates for 10 individual sessions per year of psychological therapy services delivered by clinical psychologists and focussed psychological strategies delivered by psychologists, social workers, occupational therapists, general practitioners (GPs) and prescribed medical practitioners (PMPs); we call these ‘Better Access treatment services’. Certain arrangements govern how the treatment items operate. Consumers can only access treatment from one of the above providers after a mental health treatment plan (MHTP) has been prepared by a GP/PMP, and a review by the GP/PMP is required after the sixth session.

In 2021–22, we were commissioned by the (then) Department of Health to evaluate Better Access. The evaluation occurred during the COVID-19 pandemic, when certain rules around Better Access treatment services were changed to mitigate the pandemic’s potential mental health impacts. In March 2020, the maximum number of available individual sessions per year was increased from 10 to 20. New phone and telehealth items were also introduced, enabling individuals to receive sessions in formats other than face-to-face (prior to this, videoconferencing sessions had only been available to consumers in rural/remote locations). In December 2022, the additional 10 sessions were revoked (Australian Government Department of Health, 2022b), but telehealth and phone options continued (Australian Government Department of Health, 2022a). Since our evaluation, the arrangements for GPs/PMPs have changed; in November 2025, the preparation and review of MHTPs were linked to GPs/PMPs in consumers’ MyMedicare practices or to their usual medical practitioner, and general attendance MBS items replaced Better Access-specific review and consultation items (Australian Government Department of Health Disability and Ageing, 2025).

Our evaluation involved 10 studies which considered the programme’s accessibility, responsiveness, appropriateness, effectiveness and sustainability (Pirkis et al., 2022). The current study examined accessibility and responsiveness and involved linkage of MBS data to other administrative and survey data available through the Person-Level Integrated Data Asset (PLIDA, previously known as the Multi-Agency Data Integration Project or MADIP) (Australian Bureau of Statistics, 2022). The longitudinal, linked nature of PLIDA data (available from 2016 to 2021) enabled us to examine use and non-use of the Better Access treatment items at a population level over time and to explore the extent to which use was based on need and/or driven by socioeconomic factors. More specifically, we considered four themes: (1) pathways into Better Access treatment; (2) the proportion of Better Access treatment users who are ‘new’; (3) patterns of use and non-use of Better Access treatment services in relation to need; and (4) socioeconomic differences in Better Access treatment service use.

The current study is accompanied by eight others in this issue of the Australian and New Zealand Journal of Psychiatry. These involved secondary analysis of routinely collected data (Tapp et al., 2026b; Pirkis et al., 2026a) or data from prior studies (Arya et al., 2026; Harris et al., 2026) and analysis of new data gathered specifically for the evaluation via interviews, surveys and an online virtual forum (Currier et al., 2026; Newton et al., 2026; Pirkis et al., 2026b; Tapp et al., 2026a).

Methods

Data source

We used PLIDA as our data source. PLIDA is a secure data asset combining administrative information on health, education, government payments, income and taxation and employment with information such as population demographic characteristics from the Census (Australian Bureau of Statistics, 2023). PLIDA provides an enduring, regularly updated data linkage of these Australian Government data sources. The Australian Bureau of Statistics (ABS) manages the integration and the process of data access. Deterministic and probabilistic linkage of the different data sources is based on a Person Linkage Spine developed from three sources: the Medicare Consumer Directory, the DOMINO Centrelink Administrative Dataset and the Personal Income Tax dataset. Access to PLIDA data is protected and only authorised researchers with approved projects have access to de-identified data.

Analysis cohorts

Our analyses were based on those who completed the 2016 Census (23.7 million respondents). After excluding visitors to Australia, those who died prior to the analysis period and respondents who were unable to be linked or for whom there were not unique linkages to the Person Linkage Spine and other datasets, the final analysis sample comprised 20,263,132 individuals (in 2018). The analysis sample varied across the study years (2018–2021) based on age inclusion criteria and the exclusion of those identified as having died according to the National Death Index (NDI).

A subset of analyses used data from those in the analysis sample who were aged ⩾18 and completed the Kessler 10 (K-10) in the 2017/18 National Health Survey. The K-10 is a 10-item measure of psychological distress which asks about symptoms of depression and anxiety over the past 4 weeks (Kessler et al., 2002).

Using the 2016 Census and the 2017/18 National Health Survey (and the NDI) as our starting point allowed us to identify two cohorts.

• Cohort 1 included Australians identified in the 2016 Census who were aged ⩾18 in each year of the analysis (2018, 2019, 2020 and 2021) and had a unique linkage to the Person Linkage Spine and specifically to the Medicare Consumer Directory (and therefore to Better Access items). In 2018, this cohort included 16,084,885 individuals.

• Cohort 2 was based on PLIDA linkage to the 2017/18 National Health Survey. The National Health Survey dataset includes a set of population weights generated by the ABS that can be used to generate estimates that better reflect the overall Australian population. Analyses using Cohort 2 were adjusted to correct for the non-linkage of the National Health Survey respondents in PLIDA (based on age and sex). This cohort comprised 14,340 individuals.

Data used for analysis

We used MBS data to identify use (and date of service) of Better Access treatment services, focusing on the occasions of service associated with psychological therapy services delivered by clinical psychologists and focussed psychological strategies delivered by psychologists, social workers, occupational therapists and GPs/PMPs.

In addition, we used the NDI to exclude people who died during each analysis year or earlier, and Pharmaceutical Benefits Scheme (PBS) data to identify people who were prescribed antidepressant or anxiolytic medication (defined using the Anatomical Therapeutic Chemical [ATC] codes N05B and N06A).

We described people in terms of various sociodemographic characteristics (e.g. age, sex, equivalised household income) by using data collected in the 2016 Census and 2017/18 National Health Survey. We also used an indicator of First Nations people developed by the ABS from information collected through various data sources. We used K-10 scores from the 2017/18 National Health Survey as an indicator of need.

Analyses

We present descriptive statistics (e.g. frequency and percentage of the population using given services in a calendar year). We differentiate the results for key sub-populations (e.g. contrasting new Better Access users in a year vs continuing users of Better Access services; considering key groups in the population based on need or household income). Most analyses are person-based (i.e. describing Better Access users), but some are service-based (i.e. describing number of treatment sessions). To provide information on the number or nature of services used by consumers during a year, we use PLIDA data to calculate medians, interquartile ranges and means of services used and contrast different service types (e.g. telehealth vs face-to-face). Although measures of statistical significance are not informative in the analysis of population-level data, we report the results of statistical models (using generalised linear models with a log link to evaluate binary outcome measures and report measures of relative risk) for some analyses with smaller denominators.

Approvals

The Australian National University Human Research Ethics Committee approved the ethical aspects of this project (Protocol 2022/611).

Results

Pathways into Better Access treatment

We explored Better Access service use pathways for those in Cohort 1, comparing Better Access users who received a MHTP between January and March in 2017 with those who did so in the same 3 months of 2020. The analysis was restricted to those who had not used any Better Access services in the previous year and had not died in the following two years. The two groups allowed us to investigate whether there were changes over time in the rates of access to Better Access treatment following the preparation of a MHTP and the wait time between receiving a plan and the first treatment session. We also examined whether the likelihood of receiving treatment following a plan and wait times varied by income and other sociodemographic factors.

Figure 1 shows the steps in study design and sample selection, as well as the number in each group who went on to receive Better Access treatment following the preparation of a MHTP. Overall, two-thirds (66.4%) of adults who received a MHTP in the first 3 months of the relevant year (and who had not received any Better Access services the previous year) had received Better Access treatment services by the end of the following year, but one-third (33.6%) had not. These figures were 68.6% versus 31.4% and 64.4% versus 35.6% for the 2017 and 2020 groups, respectively.

Figure 1.

2017 and 2020 groups (Cohort 1).

We explored the factors that were associated with not receiving Better Access treatment services following a MHTP further in a generalised linear regression model (with log link). We found that the likelihood of not receiving treatment was greatest for those living in a low-income household, living in a more disadvantaged area and with other sociodemographic factors, including being young (aged 18–24), being male, identifying as First Nations, living in an outer regional or remote location and coming from New South Wales, the Northern Territory or the Australian Capital Territory. These factors are summarised in Table 1.

Table 1.

Factors associated with not receiving any Better Access treatment services following a mental health treatment plan (Cohort 1).

Factor	Level	IRR	95% CI
Cohort	2017 (ref)	1.00
	2020	1.15	1.12–1.15
Age	18–24 years (ref)	1.00
	25–44 years	0.95	0.93–0.97
	35–64 years	0.92	0.90–0.93
	65 years +	0.94	0.91–0.96
Prior use of MBS antidepressant or anxiolytic medication	No (ref)	1.00
	Yes	1.04	1.03–1.06
Male gender	No (ref)	1.00
	Yes	1.08	1.07–1.10
Equivalised household income	Lowest income (1)	1.25	1.22–1.28
	Group 2	1.14	1.12–1.17
	Group 3	1.07	1.05–1.09
	Group 4	0.97	0.95–0.99
	Highest income (5; ref)	1.00
	Not available	1.08	1.04–1.12
Residential remoteness	Major city (ref)	1.00
	Inner regional	1.08	1.06–1.09
	Outer regional	1.25	1.22–1.28
	Remote	1.52	1.44–1.62
	Very remote	1.36	1.24–1.49
State	NSW (ref)	1.00
	Vic	0.93	0.91–0.94
	Qld	0.88	0.86–0.89
	SA	0.89	0.87–0.91
	WA	0.92	0.90–0.95
	Tas	0.73	0.70–0.76
	NT	1.16	1.08–1.24
	ACT	1.12	1.07–1.18
Received non-Better Access MBS mental health services	No (ref)	1.00
	Yes	0.90	0.86–0.95
First Nations person	No (ref)	1.00
	Yes	1.27	1.24–1.31
Area-based disadvantage (Index of Relative Socio-economic Advantage and Disavantage)	Most disadvantaged (1)	1.26	1.24–1.29
	Group 2	1.22	1.19–1.24
	Group 3	1.14	1.12–1.16
	Group 4	1.06	1.04–1.09
	Most advantaged (5; ref)	1.00
	Not available	0.61	0.35–1.08

We extended the linear regression model to examine the apparent difference in the proportion of people receiving no Better Access treatment in 2017–2018 versus 2020–2021. More specifically, we included an interaction between group and treatment. There was an absolute difference of −4.2% (95% CI = −4.51 to −3.84) and a relative difference of 0.94 (95% CI = 0.93–0.95). In other words, the likelihood of people accessing treatment following a MHTP just over four percentage points lower in 2020–2021 than it was in 2017–2018, and this represented a 6% decline in the use of treatment services.

We then focussed on those in each group who did receive Better Access treatment and examined wait times between receiving a MHTP and accessing treatment. Figure 1 shows that half of those who received treatment following their plan in 2017 waited 14 days to receive treatment. In 2020, half waited 19 days. We also examined the relationship between income and wait times and the results are presented in Figure 2. This figure shows a clear socioeconomic gradient, with the median wait times being greatest for those on the lowest incomes. The inequities were more pronounced for the 2020 group than the 2017 group.

Figure 2.

Median wait times between receiving a mental health treatment plan and accessing Better Access treatment services by equivalised household income (in quintiles), 2017–2018 and 2020–2021 (Cohort 1).

The proportion of Better Access treatment users who are ‘new’

The longitudinal nature of the PLIDA data allowed us to follow individuals in Cohort 1 to determine the proportion of Better Access treatment service users in any given year who were ‘new’ (i.e. did not access Better Access treatment services in the previous year).

Table 2 shows that although there was an increase over time in the percentage of the cohort who used Better Access treatment services, the proportion of users who were ‘new’ declined. In 2018, 56.0% of those who accessed Better Access treatment service were new users, but in 2021 this figure had dropped to 49.9%. The most marked decline occurred between 2020 and 2021.

Table 2.

New and continuing Better Access treatment service users, 2018 to 2021 (Cohort 1).

		2018	2019	2020	2021
Service	N people	16,084,885	16,199,009	16,310,197	16,499,245
Any Better Access treatment service	N total service users	852,676	897,536	916,378	916,898
	% of cohort using services	5.3%	5.5%	5.6%	5.6%
	N continuing users	375,133	404,371	429,057	459,675
	% of current users who are continuing	43.9%	45.0%	46.8%	50.1%
	N new users	477,543	493,165	487,321	457,223
	% current users who are new	56.0%	55.0%	53.2%	49.9%

We explored whether the decline in new users might relate to the introduction of the additional 10 sessions in 2020. This may have resulted in existing consumers receiving more sessions and reduced providers’ capacity to take on new consumers. Table 3 shows that in 2020, 4.8% of new users accessed the additional 10 sessions, compared with 11.5% of continuing users. This gap increased further in 2021, with 8.1% of new users accessing the additional 10 sessions and 26.8% of continuing users doing so.

Table 3.

Use of additional 10 sessions and face-to-face sessions only, by new and continuing users, 2020–2021 (Cohort 1).

		2020	2021
Use of additional 10 sessions	New	4.8%	8.1%
	Continuing	11.5%	26.8%
Use of face-to-face sessions only	New	63.6%	64.6%
	Continuing	55.5%	52.2%

We also considered whether the widespread availability of telehealth options, also introduced during 2020, may have disproportionately benefitted existing consumers. Providers may have found it more difficult to initiate treatment with new consumers by telehealth than to build on relationships that they may have already established face-to-face with existing consumers. Table 3 indicates that in 2020, new users were more likely to only receive their treatment services face-to-face compared to continuing users (63.6% versus 55.5%). The same was true in 2021 (64.6% versus 52.2%).

Patterns of use and non-use of Better Access treatment services in relation to need

We examined patterns of use and non-use of Better Access in relation to need by using K-10 data from the 2017/18 National Health Survey and linked MBS data for those in Cohort 2, capitalising on the fact that they completed the K-10 as part of the National Health Survey. Others have shown that the levels of psychological distress that individuals report on the K-10 at different time points (i.e. 2 years or more apart) are relatively consistent (Welsh et al., 2020), and that scores on the K-10 correspond closely to diagnoses of mental disorders (Slade et al., 2011). Therefore, it is likely that many of those identified with high or very high levels of distress in the National Health Survey experienced longer-term or chronic distress and poor mental health. For this reason, we examined use of Better Access treatment services during the 12 months before and after each National Health Survey participants’ interview. Data from the National Health Survey were weighted to the general population.

Figure 3 shows that there was a strong association between Better Access treatment service use and psychological distress, with greater likelihood of use being associated with greater levels of psychological distress. In total, 25.0% of Australian adults with very high psychological distress used any Better Access treatment service in the 12 months before or after their National Health Survey interview, and 21.2% of those with high levels of psychological distress did so.

Figure 3.

Use of any Better Access treatment services (2017–2019), by levels of psychological distress (Cohort 2).

Simple generalised linear models (using a log link) confirmed these findings. Compared with those who reported low levels of psychological distress, those who reported high or very high levels of psychological distress were more likely to use any Better Access treatment service (RR = 3.84 [95% CI = 3.28–4.50] and RR = 4.54 [95% CI = 3.72–5.53], respectively).

Socioeconomic differences in Better Access treatment service use

We used linked Census and MBS data to examine the relationship between individual-level indicators of socioeconomic status and use of Better Access services over time. As a point of comparison, we also considered the relationship between individual-level socioeconomic status and access to psychotropic medication through the PBS. These analyses used data from Cohort 1.

Figure 4 shows the relationship between self-reported equivalised household income (taken from the 2016 Census) and use of Better Access treatment services and antidepressants/anxiolytics over time (considering those with any use in a calendar year). Income data are presented in quintiles, and data are presented for those in major cities only, to reduce the potential confounding effect of service availability.

Figure 4.

Use of Better Access treatment services and antidepressants/anxiolytics by equivalised household income (in quintiles), 2018–2021 (Cohort 1, residents of major cities only).

The profile of users of Better Access treatment services shows that between 2018 and 2021 those with comparatively higher levels of income were most likely to access these services. This gap has widened over time as the percentage of those in the higher income groups using these services has increased and the percentage of those in the lower income groups doing so has decreased.

The profile of users of antidepressants/anxiolytics over the same period was quite different. Across all years, people on lower incomes were consistently more likely to have antidepressants/anxiolytics dispensed for them. Rates of use for all groups have increased slightly over time.

Discussion

Although two-thirds of those who have a MHTP prepared for them go on to receive Better Access treatment services, one-third do not. Various reasons may explain this, including that GPs/PMPs provide consumers with mental health treatment themselves (using other MBS items), or that they make referrals to non-Better Access providers and services. Consumers themselves may choose not to take up mental health treatment or may do so through other avenues. Nonetheless, those who do not receive Better Access treatment services following a MHTP are significant and increasing in number.

Those who do move through the service pathway from a MHTP to receipt of Better Access treatment services often have to wait some time to be seen by the relevant mental health professional. These wait times are increasing, presumably because many providers are at or close to capacity. A survey by the Australian Psychological Society showed that 88% of psychologists experienced increased demand for services during the COVID-19 pandemic, resulting in longer wait times for consumers (Australian Psychological Society, 2022). Other studies in our evaluation support this, with a survey of providers showing that wait times and capacity issues were paramount (Tapp et al., 2026a), and surveys and interviews with consumers suggesting that wait times are a major barrier to accessing care (Newton et al., 2026; Pirkis et al., 2026b).

Capacity issues may also have influenced the reduction in new users of Better Access treatment services that we observed. Between 2018 and 2021, the proportion of those who accessed any Better Access treatment services who were new users declined, with the steepest drop occurring between 2020 and 2021. The fact that providers were able to offer consumers the additional 10 sessions during this time may have benefitted continuing users but exacerbated ‘blockages’ for those not previously using treatment services. The telehealth options introduced during the pandemic may also help to explain the findings with respect to new users. Providers rapidly transitioned to offering telehealth services when these became an option (Reay et al., 2021), but we found they were not delivered equally to new and continuing users. In both 2020 and 2021, new users were more likely than continuing users to receive face-to-face treatment only. This suggests that providers may have found it easier to provide telehealth services to consumers with whom they had existing relationships, a finding that has been reported elsewhere (Appleton et al., 2021; Barnett et al., 2021; Irvine et al., 2020). Alternatively, it may reflect a preference of new users in initiating a therapeutic relationship. The potential unintended consequences of policy changes have been considered in other contexts and could be further explored here (Oliver et al., 2019). Further work might be done to determine whether the patterns with respect to new and continuing users have changed now that the additional 10 sessions have been removed and telehealth has become more embedded as a format.

Our study provides population-level evidence that Better Access is reaching those with comparatively high levels of need in the Australian community. We found that 25% of those with very high levels of distress accessed psychological treatment through Better Access, whereas elsewhere in our evaluation we showed that 10% of the general population did so (Tapp et al., 2026b). That said, people with relatively lower levels of need are also accessing Better Access, albeit at lower rates. In absolute terms, given that most people in the population report low levels of psychological distress, this means that significant numbers of people with lower levels of psychological distress are accessing the programme.

Our study highlights a number of equity issues in relation to use of Better Access services and suggest that these may be worsening. Those on the lowest incomes are least likely to access services but most likely to access mental health medication through the PBS. For example, 5.1% of those in the lowest socioeconomic quintile used any Better Access treatment services in 2021 compared with 6.6% in the highest quintile. In the same year, only 56.5% of those in the lowest quintile proceeded to treatment from a MHTP compared with 69.3% of their high-income counterparts. The wait times for treatment for those who did progress from a plan to treatment were also longer for those in the lowest income quintile; their median wait time was 22 days compared with 17 days for those in the highest quintile. It has been suggested that those on the lowest level of income may be preferentially offered pharmacological treatment instead of psychological therapies (Butterworth et al., 2013; Comino et al., 2000), a finding consistent with the current results. All of these indicators have worsened over time. Results from other studies in our evaluation suggest that rising out-of-pocket costs may be driving these socioeconomic gradients in access and uptake, with affordability emerging as a consistent theme. However, there may also be other factors at play, including attitudes and preferences of consumers and providers, and these warrant exploration.

Strengths and limitations

Our use of PLIDA had many benefits. No other data source provides longitudinal data on Better Access service use and individual-level measures of need and socioeconomic circumstances for the Australian population. However, PLIDA has some limitations, including its exclusive focus on Australian Government-funded MBS and PBS services. Our results should be considered in the context of programmes that are the responsibility of other sectors (e.g. services that are commissioned by Primary Health Networks or delivered through community health centres, Aboriginal Community Controlled Health Organisations, private hospitals or public sector inpatient and community services). This caveat is important for interpreting the finding that one-third of those who had a MHTP did not go on to receive Better Access treatment services. As noted, some of these people may have received mental health services through other sources. As a side note, the recent changes to Better Access that involve replacing some GP/PMP items with general attendance items has implications for tracking Better Access service use trajectories through PLIDA in the future.

The fact that we were able to examine Better Access treatment service use in the context of need and to generalise the findings to Australian adults is positive. However, we used a single indicator of need, based on an assessment of psychological distress (per the K-10) made at a single time point. We assumed (based on prior research) that this represented an individual’s level of distress over a longer period (Welsh et al., 2020); however, levels of distress may fluctuate.

Our income measures were based on individuals’ household incomes. This is positive because most other population-based studies of MBS services rely on area-based income indicators, which are a proxy for individual socioeconomic status and do not cater for within-area socioeconomic variability (Moss et al., 2021). However, because income was assessed at a single timepoint (i.e. in the 2016 Census), our measures do not capture change over time. It may be that the incomes of some people with the highest incomes in 2016 declined, and this might partially explain increasing rates of service use within these groups. However, it might also be expected that the economic circumstances of a proportion of those on the lowest incomes would improve. More importantly, the general social patterning of use of Better Access treatment was evident in the first year of observations.

Summary and conclusions

Two-thirds of individuals who have a MHTP prepared for them receive Better Access treatment services (albeit often after a considerable wait), but one-third do not. Although Better Access is reaching those with high levels of need, access is not equitable. It is harder for new users to access the programme than it was previously, as the number of continuing users and the number of treatment sessions provided to them has increased. People on low incomes are less likely to receive psychological treatment through Better Access (but more likely to be prescribed antidepressant or anxiolytic medication), and if they do receive Better Access services, they typically wait longer than their high-income counterparts to see a provider. Better Access appears to be responsive to need but there are equity issues regarding its accessibility which should be addressed as Better Access continues.