Consumers’ experiences with and outcomes from Better Access: Results from a national survey

Abstract

Objective:

We sought consumers’ views about Better Access, which funds sessions of care with eligible providers via the Medicare Benefits Schedule (MBS).

Methods:

We surveyed a stratified random sample of consumers who saw a clinical psychologist, psychologist, social worker or occupational therapist (OT) via Better Access during 2021. The survey focussed on consumers’ experiences with receiving treatment through Better Access, and the outcomes of this treatment. Survey data were linked to MBS claims data for consenting participants.

Results:

In total, 2013 individuals completed the survey; linked MBS data were available for 1317 (65.4%). The majority (85.2%) were satisfied with their care, although they raised some issues, particularly around affordability. When asked to rate their mental health before and after treatment, 91.9% indicated it had significantly improved. Overall, 77.5% attributed this improvement to treatment by the mental health professional. For the full sample, baseline self-rated mental health was predictive of improvement, as was the number of sessions. For the sub-sample with linked data, these factors also predicted improvement, as did whether they paid a co-payment.

Conclusion:

In general, consumers who use Better Access appear to appreciate the programme and benefit from the care it provides. However, affordability remains an issue.

Introduction

In Australia, the commonest way for someone to visit a clinical psychologist, psychologist, social worker or occupational therapist (OT) for mental health problems is through the Better Access to Psychiatrists, Psychologists and General Practitioners through the Medicare Benefits Schedule (MBS) initiative (Better Access). Medicare partially or wholly funds treatment sessions with these providers via MBS item numbers; where the costs are not fully covered consumers pay a co-payment. Consumers with a diagnosed mental disorder can access a certain number of individual and group sessions per year following referral by a general practitioner (GP), psychiatrist or paediatrician (Australian Government Department of Health, 2021; Australian Government Department of Health and Aged Care, 2022).

Better Access began in 2006 and we evaluated it in 2009–2011. One evaluation component involved assessing outcomes for 883 consumers (recruited by clinical psychologists, psychologists and GPs). Those 883 consumers and an additional 530 (recruited by social workers and OTs) were asked about their experiences of care, via interviews/surveys. Participants were positive about the programme, and many reported reductions in symptoms/distress and improvements in wellbeing (Ftanou et al., 2014; Pirkis et al., 2010a, 2010b, 2011a, 2011b).

In 2021–2022, we conducted a second Better Access evaluation (Pirkis et al., 2022). The programme had expanded and some of its requirements had been modified. Notably, certain arrangements were implemented during the COVID-19 pandemic; 10 additional sessions were offered (raising the limit to 20), and telehealth/phone sessions became more readily available. The additional 10 sessions have since been revoked (Australian Government Department of Health, 2022b), but telehealth/phone options are now permanent (Australian Government Department of Health, 2022a).

Our 2021–2022 evaluation included 10 studies (Pirkis et al., 2022); this study and eight others are reported in this issue of the Australian and New Zealand Journal of Psychiatry (Arya et al., 2026; Chilver et al., 2026; Currier et al., 2026; Harris et al., 2026; Newton et al., 2026; Pirkis et al., 2026; Tapp et al., 2026a, 2026b). This study involved a survey of consumers who had received treatment from clinical psychologists, psychologists, social workers or OTs via Better Access during 2021. Services Australia (which administers MBS payments) assisted us with recruitment. We chose this approach to avoid potential biases associated with our previous method of recruiting through providers (Allen and Jackson, 2011; Hickie et al., 2011). The survey focussed on consumers’ experiences with and outcomes of receiving treatment through Better Access. Survey data were linked to MBS claims data for consenting participants.

Methods

Sampling and recruitment

Services Australia identified a stratified random sample of consumers aged ⩾ 18 who had received care through the Better Access treatment item numbers during 2021. More specifically, Services Australia classified eligible consumers on the basis of their location of residence and the services they received into 16 mutually exclusive strata (2 × location of residence, 2 × receipt of the additional 10 sessions and 4 × treating provider type) and randomly selected up to 2500 consumers within each stratum. In strata with fewer than 2500 consumers, all were included. In total, 27,167 consumers were identified (Supplementary Table 1).

Services Australia sent each consumer an invitation to complete the survey on our behalf. Each invitation directed consumers to the survey via stratum-specific URL/QR codes. Invitations were sent on 31 January 2022 and the survey closed on 4 March 2022.

Procedure

Interested consumers accessed the survey online via their URL/QR code. They were presented with a plain language statement describing the survey requirements. Those who consented clicked on a box which took them to the survey.

After completing the survey, participants were presented with another plain language statement describing the MBS data linkage component. Those who consented to linkage provided relevant details, so that Services Australia could locate their information.

Survey data were automatically captured in a database held by our independent subcontractor, Logicly, which delivered the final data set to us by secure means. MBS data were provided to us by Services Australia and linked using a linkage key.

The survey instrument

The survey sought demographic details from participants (e.g. age and sex) and asked them for their postcode which was later mapped to the Index of Relative Socioeconomic Disadvantage (IRSD) of the Socioeconomic Indexes for Areas (SEIFAs) (Australian Bureau of Statistics, 2018).

The survey asked participants about the mental health professional they saw in 2021 (or the main professional if they had seen more than one), the sessions of care, how the care was paid for, the process of seeing the mental health professional, their overall satisfaction with care, and the outcomes of care. Additional detail is provided below about the latter three:

• Consumers were asked to rate their agreement/disagreement with 11 statements about the process of seeing the mental health professional (e.g. ‘I found the referral process straightforward’) on a scale of 1 (strongly disagree) to 5 (strongly agree), modified from statements used in the UK National Audit of Psychological Therapies (Royal College of Psychiatrists, 2013).

• Satisfaction with care was assessed by the question ‘How satisfied were you with your care?’ with five response options from 1 (very dissatisfied) to 5 (very satisfied), which was similar to questions used in other Australian health care settings (Pearse, 2005).

• Regarding outcomes, participants were asked about their mental health before and after care: On a scale of 1 to 10, where 1 is the worst possible mental health and 10 is the best possible mental health, how would you rate your mental health [before/after] your [first/most recent] session with the mental health professional? These questions were based on those used in surveys like the Australian Health Survey (Australian Bureau of Statistics, 2013). Participants were then asked, ‘To what extent would you attribute any change in your mental health to the treatment you received from the mental health professional?’, with three response options: ‘entirely responsible’, ‘partially responsible’ or ‘not at all responsible’.

Data analysis

We conducted descriptive analyses, reporting frequencies, percentages and means (and standard deviations) for all variables.

We measured outcomes of care as the difference between participants’ self-rated mental health after versus before receipt of care from the mental health professional (scores ranged from −9 to + 9; negative and positive scores indicated deterioration and improvement, respectively). We adopted a standardised difference approach to classifying outcomes on this scale, using a small-to-medium effect size of 0.3 of a standard deviation of the self-rated mental health before receipt of care as the indicator of change (Cohen, 1988). This yielded three outcome groups: ‘Significantly deteriorated’; ‘No significant change’; and ‘Significantly improved’. We chose an effect size of 0.3 to correspond with other studies in the evaluation that used standardised measures like the Patient Health Questionnaire-9 (PHQ-9) (Kroenke et al., 2001) for which 0.3 equates to a minimally important difference for consumers (Arya et al., 2026; Harris et al., 2026; Pirkis et al., 2026).

We used multivariable logistic regression to estimate the strength of association between selected covariates and improvement (‘Significantly improved’ versus ‘Significantly deteriorated’/‘No significant change’). The covariates were age, sex, sexual identity, country of birth, First Nations status, area of residence, socioeconomic status (SEIFA IRSD), baseline self-rated mental health, provider type and self-reported number of sessions.

Our primary analyses were conducted using data from all survey participants. We conducted supplementary analyses with data from the sub-sample with linked data. For these analyses, we used MBS data instead of self-report data for all session-related variables, and included out-of-pocket costs as a covariate in the logistic regression.

All analyses were conducted in STATA v15.1 (StataCorp, 2017).

Approvals

The University of Melbourne Human Research Ethics Committee approved the study (HREC 2022-22999-30805-6). Services Australia External Requests Evaluation Committee approved the mailout (EREC RMS2089) and MBS data linkage (EREC RMS2092).

Results

In total, 2013 individuals (7.4%) completed the survey and provided usable data. Linked MBS data were available for 1317 of these participants (65.4%). The socio-demographic characteristics of the whole sample and the sub-sample with linked data were similar (Table 1). In both cases, the majority of participants were relatively young (with over half being aged < 40), but all age groups were represented. Three quarters of each group were female. Three quarters identified as straight or heterosexual, and one fifth as lesbian, gay, homosexual or bisexual. Three quarters were born in Australia. All states/territories were represented. There was good representation from people in outer regional, remote and very remote areas due to our sampling strategy; one third of the sample and sub-sample resided in these areas. There was also good representation across areas with differing levels of socioeconomic disadvantage, with around one fifth of the sample and sub-sample in most of the five quintiles of the SEIFA IRSD (Australian Bureau of Statistics, 2018). First Nations people were under-represented relative to the general population (Australian Bureau of Statistics, 2021), however, accounting for only 2%.

Table 1.

Socio-demographic profile of the total survey sample and the sub-sample for whom linked MBS claims data were available.

		All participants (N = 2013)		Participants with linked MBS claims data (N = 1317)
		Frequency	%	Frequency	%
Age a	18–19	82	4.1	56	4.3
	20–29	454	22.9	302	23.0
	30–39	531	26.8	344	26.2
	40–49	382	19.3	252	19.2
	50–59	256	12.9	170	12.9
	60–69	197	9.9	138	10.5
	70–79	78	4.0	53	4.1
Sex a	Female	1336	74.6	888	73.9
	Male	399	22.3	282	23.5
	Non-binary sex	39	2.2	25	2.1
	Prefer not to say	16	0.9	7	0.6
Sexual identity a	Lesbian, gay or homosexual	125	7.0	90	7.5
	Straight or heterosexual	1349	75.1	904	75.0
	Bisexual	213	11.9	146	12.1
	Something else	44	2.4	28	2.3
	Don’t know/prefer not to say	65	3.6	37	3.1
Country of birth a	Australia	1449	75.3	981	78.0
	Overseas	476	24.7	276	22.0
First Nations status a	First Nations	36	2.1	26	2.2
	Not First Nations	1744	98.0	1179	97.8
State of residence a	NSW	390	20.8	254	20.2
	VIC	664	35.5	444	35.3
	QLD	417	22.3	285	22.7
	SA	90	4.8	59	4.7
	WA	164	8.8	111	8.8
	TAS	88	4.7	67	5.3
	ACT	25	1.3	17	1.4
	NT	33	1.8	20	1.6
Area of residence a	Major city	1399	69.5	889	67.5
	Regional, rural and remote	614	30.5	428	32.5
Area-level socioeconomic status (SEIFA IRSD quintiles)a,b	Q1 (most disadvantaged)	253	13.6	180	14.4
	Q2	363	19.5	243	19.4
	Q3	391	21.0	267	21.3
	Q4	374	20.1	254	20.3
	Q5 (least disadvantaged)	484	26.0	308	24.6

^aMissing data excluded.

^bSocioeconomic status was ascribed to participants on the basis of their postcode, using quintiles.

^cDerived from the Index of Relative Socioeconomic Disadvantage (IRSD) of the Socioeconomic Indexes for Areas (SEIFA). More specifically, the SEIFA concordance file was used to assign the IRSD. The IRSD file reports deciles that were then converted into quintiles.

Primary analyses: total sample (n = 2013)

Mental health professionals seen by participants via Better Access in 2021

According to their MBS claims data, 34.5% of the 2013 participants saw a clinical psychologist in 2021, 30.2% saw a psychologist, 25.1% saw a social worker and 10.2% saw an OT.

Diagnosis at the time of receiving care

In total, 52.8% of participants said they were given a diagnosis when they sought Better Access care, and 11.0% said they were unsure. The remaining 36.2% said they were not given a diagnosis. Those who were given a diagnosis were typically told they had an anxiety disorder (71.8%) and/or depression (70.2%).

Sessions of care

Table 2 profiles participants’ sessions of care. The majority (99.4%) received individual sessions. Most (79.2%) received face-to-face sessions but 51.0% received telehealth sessions and 20.9% received phone sessions. Many participants received sessions in a combination of formats.

Table 2.

Sessions of Better Access care received by all participants.

			Frequency	%
Session format (n = 2000) a	Individual	Yes	1988	99.4
		No	12	0.6
	Group	Yes	38	1.9
		No	1962	98.1
Session modality (n = 2000) a	Face-to-face	Yes	1583	79.2
		No	417	20.9
	Via telehealth	Yes	1020	51.0
		No	980	49.0
	Via phone	Yes	414	20.7
		No	1586	79.3
Still receiving sessions of care (n = 1999) a	Yes		1365	68.3
	No		508	25.4
	Unsure		126	6.3
Number of sessions of care (n = 1246) a	1–2		83	6.7
	3–4		111	8.9
	5–6		144	11.6
	7–10		244	19.6
	11 +		664	53.3
Adequacy of number of sessions (n = 1978) a	Too many		10	0.5
	Too few		747	37.8
	Just right		989	50.0
	Unsure		232	11.7

^aMissing data excluded.

In total, 68.3% of participants were still receiving sessions of care at the time of the survey and 25.4% were no longer receiving care. Participants were asked about the number of sessions they had attended or expected to attend (in the case of those who were still receiving care). The mean number of sessions was 12.45 (SD = 6.64). Half (53.3%) reported receiving (or expecting to receive) 11 or more sessions. The number of sessions is likely to be skewed because of the way we sampled participants. One of the characteristics used to create the 16 strata was receipt (or non-receipt) of the additional 10 sessions that became available during the pandemic, meaning we over-sampled consumers who had received additional sessions in 2021.

Participants were also asked about the adequacy of the number of sessions they had received or expected to receive. Half (50.0%) indicated that the number of sessions was just right, but a further 37.8% regarded it as too few.

Additional consideration was given to those who said they were no longer seeing the mental health professional. Of these, 51.4% indicated that they had continued seeing the mental health professional for as long as possible, but 44.1% had stopped seeing the mental health professional before their session limit was reached. When this group was asked about their reasons for ceasing treatment, 38.4% indicated that they did not find the sessions helpful, 31.7% said the out-of-pocket costs were too expensive and 30.4% said they did not like the mental health professional’s manner or approach. Around one quarter (27.7%) indicated that they stopped seeing the mental health professional because they felt better.

Payment for Better Access care

For the majority of participants (70.0%), Medicare covered only some of the cost of care and they paid the residual (Table 3). For 26.1% of participants, however, Medicare covered the total cost. Participants were evenly split between considering that the fee was affordable (35.6%) and too expensive (34.5%).

Table 3.

Payment for Better Access care by all participants.

		Frequency	%
Payment source (n = 1987) a	Medicare covered all of the costs	519	26.1
	Medicare covered some of the costs, but I paid at least some of the costs out of my own pocket	1391	70.0
	Some other payment arrangement	36	1.8
	Unsure	41	2.1
Perceptions of affordability (n = 1945) a	I didn’t pay anything; Medicare covered all of the cost	510	26.2
	I paid a fee that was affordable	693	35.6
	I paid a fee that was too expensive	670	34.5
	Unsure	72	3.7

^aMissing data excluded.

Experiences of seeing the mental health professional

The majority of participants reported positive experiences with Better Access: 77.0% found the referral process straightforward, 83.8% were offered sessions at a time that suited them, 91.6% found the mental health professional empathic, 94.2% thought that the mental health professional listened to them, 93.9% indicated that the mental health professional respected their right to make decisions, 86.5% believed that the mental health professional equipped them with strategies to address the issues they were facing, 82.0% considered that the support or care provided by the mental health professional met their needs, and 90.2% had a good relationship with the mental health professional. Only 8.2% said they had to travel too far to see the mental health professional.

There were two notable exceptions. Nearly, one third of participants (30.1%) believed they had to wait too long for an appointment with the mental health professional, and only a little over one third (38.1%) indicated that they were offered the opportunity for their family/friends to be involved in their support or care if they wanted this.

The majority of participants were satisfied with their care (40.7% satisfied, 44.5% very satisfied).

Outcomes of care

Participants were extremely positive about the outcomes of their care. Using the 1–10 scale (1 = worst possible mental health and 10 = best possible mental health), participants’ mean self-rated mental health scores before and after receiving care were 3.48 (SD = 1.73) and 7.04 (SD = 1.74), respectively. Figure 1 provides more detail. Panel 1 shows that prior to receiving care, nearly 60% of participants rated their mental health at ⩽ 3. After treatment, the picture reversed, with close to 50% of participants rating their mental health at ⩾ 8. Panel 2 translates participants’ before-and-after ratings to raw outcome scores by subtracting their self-rated mental health score prior to using Better Access from their self-rated mental health score after doing so. The skew to the right indicates that most participants experienced positive change. Panel 3 shows that when the raw outcome scores were translated into outcome groups, 91.9% of participants were ‘significantly improved’.

Figure 1.

Outcomes of care.

Overall, 77.5% of those whose mental health improved attributed this, in part or entirely, to treatment by the mental health professional.

Predictors of improvement

Only three covariates had a statistically significant association with significant improvement (highlighted in grey): sex, baseline mental health and number of sessions (Table 4). Being female was associated with higher odds of improvement (odds ratio [OR] = 2.69; 95% confidence interval [CI] = [1.42, 5.08]). The better participants’ self-rated mental health was at the beginning of their episode of care, the lower their odds of showing improvement (OR = 0.53; 95% CI = [0.45, 0.63]). Compared with those who had one to two sessions, those who had (or expected to have) more sessions had greater odds of improving. There was some suggestion of a dose response effect, with increasingly greater odds associated with increasingly higher numbers of sessions, although the 95% CIs overlapped: three to four sessions (OR = 5.18; 95% CI = [1.69, 15.87]), five to six sessions (OR = 6.28; 95% CI = [2.18, 18.03]), seven to ten sessions (OR = 7.45; 95% CI = [2.74, 20.25]) and 11+ sessions (OR = 8.86; 95% CI = [3.60, 21.79]).

Table 4.

Predictors of improvement in self-rated mental health for all participants. ^a

		Odds ratio	95% CI low	95% CI high	p-value
Age	⩽ 29	1.00
	30–39	1.54	0.68	3.48	0.301
	40–49	1.50	0.62	3.60	0.367
	50–59	1.45	0.54	3.92	0.460
	60–69	1.38	0.48	3.97	0.556
	⩾ 70	7.43	0.68	81.25	0.100
Sex	Male	1.00
	Female	2.69	1.42	5.08	0.002
	Non-binary sex	0.77	0.17	3.48	0.735
	Prefer not to say	1.73	0.14	21.86	0.674
Sexual identity	Straight or heterosexual	1.00
	Lesbian, gay or homosexual	1.76	0.43	7.25	0.432
	Bisexual	0.47	0.19	1.12	0.088
	Something else	0.47	0.12	1.85	0.278
	Don’t know/prefer not to say	0.62	0.12	3.29	0.573
Country of birth	Australia	1.00
	Overseas	1.23	0.57	2.70	0.597
First Nations status	Not First Nations	1.00
	First Nations	0.52	0.08	3.29	0.486
Area of residence	Major city	1.00
	Regional, rural and remote	0.90	0.44	1.85	0.779
SEIFA IRSD quintile b	Q1 (most disadvantaged)	1.00
	Q2	1.20	0.43	3.34	0.725
	Q3	1.15	0.43	3.07	0.781
	Q4	1.90	0.63	5.75	0.256
	Q5 (least disadvantaged)	1.48	0.53	4.13	0.459
Self-rated mental health	Before care	0.53	0.45	0.63	0.000
Provider type c	Clinical psychologist	1.00
	Psychologist	0.75	0.36	1.58	0.453
	Social worker	0.64	0.29	1.40	0.260
	Occupational therapist	1.23	0.37	4.10	0.738
No. of sessions	1–2	1.00
	3–4	5.18	1.69	15.87	0.004
	5–6	6.28	2.18	18.03	0.001
	7–10	7.45	2.74	20.25	0.000
	11+	8.86	3.60	21.79	0.000
Face-to-face sessions	No	1.00
	Yes	1.00	0.47	2.10	0.995
Telehealth sessions	No	1.00
	Yes	0.76	0.36	1.57	0.452
Phone sessions	No	1.00
	Yes	0.95	0.45	2.02	0.902

^aMissing data excluded.

^bSocioeconomic status was ascribed to participants on the basis of their postcode, using quintiles derived from the Index of Relative Socioeconomic Disadvantage (IRSD) of the of the Socioeconomic Indexes for Areas (SEIFA). More specifically, the SEIFA concordance file was used to assign the IRSD. The IRSD file reports deciles that were then converted into quintiles.

^cProvider type according to Services Australia.

Supplementary analyses: sub-sample of participants with linked MBS data (n = 1317)

According to their MBS data, 35.0% of participants for whom linked MBS data were available saw a clinical psychologist, 31.0% saw a psychologist, 24.3% saw a social worker and 9.7% saw an OT. In total, 16.5% of the sub-sample received one to two sessions of care, 14.8% received three to four sessions, 13.6% received five to six sessions, 23.0% received seven to ten sessions and 32.1% received 11 or more sessions. Among the sub-sample, 78.4% received face-to-face sessions, 43.5% received sessions via telehealth and 15.2% received sessions via phone (noting that an individual participant could receive sessions via more than one modality).

In total, 67.7% of the sub-sample paid a co-payment for at least one session. Table 5 shows that for the sub-set of sessions where a co-payment was made, the median out-of-pocket cost per session was AUD$ 71.60 (inter-quartile range [IQR] = AUD$37.40–$91.75). Table 5 also shows the median and IQR for those who indicated via the survey that the fee they paid was affordable (54.9%) and those who thought the fee they paid was too expensive (45.1%). Unsurprisingly, the median was lower for those who considered the fee to be affordable than it was for those who thought it was too expensive (AUD$61.75 [IQR = AUD$31.60–82.20] vs AUD$77.55 [IQR = AUD$51.75–100.45]).

Table 5.

Co-payments paid by participants with linked MBS claims data.

		Median (AUD$)	IQR (AUD$)
Co-payments by provider type	Any provider	71.60	37.40–91.75
	Clinical psychologist	70.45	37.40–91.60
	Psychologist	72.55	42.55–92.55
	Social worker	72.20	42.20–82.90
	Occupational therapist	72.20	22.90–102.20
Co-payments by perceived affordability	I paid a fee that was affordable	61.75	31.60–82.20
	I paid a fee that was too expensive	77.55	51.75–100.45

IQR: inter-quartile range.

Like the overall group of survey participants, the vast majority of the sub-sample who provided pre- and post-responses on their self-rated mental health indicated that their mental health got better over the course of their care: 91.9% significantly improved, 5.4% experienced no significant change and 2.7% significantly deteriorated.

We repeated the logistic regression analysis for the sub-sample, using MBS data for all session-related variables and including out-of-pocket costs as a covariate. The results largely mirrored the findings from the primary analyses (Table 6). Participants’ self-rated mental health at the beginning of the episode of care was significantly associated with improvement; the better their initial self-rated mental health, the lower their odds of showing improvement (OR = 0.47; 95% CI = [0.40, 0.55]). The number of sessions was also related to improvement; compared with those who had one to two sessions, those who had more sessions had greater odds of improving: three to four sessions (OR = 2.66; 95% CI = [1.03, 6.82]), five to six sessions (OR = 3.18; 95% CI = [1.15, 8.78]); 11+ sessions (OR = 4.18; 95% CI = [1.67, 10.48]). Having made a co-payment for at least one session was also associated with greater odds of showing improvement (OR = 4.08; 95% CI = [2.17, 7.69]).

Table 6.

Predictors of improvement in self-rated mental health for participants with linked MBS claims data. ^a

		Odds ratio	95% CI low	95% CI high	p-value
Age	⩽ 29	1.00
	30–39	1.20	0.53	2.72	0.661
	40–49	1.31	0.54	3.19	0.557
	50–59	0.72	0.28	1.85	0.498
	60–69	1.14	0.41	3.13	0.803
	⩾ 70	4.60	0.77	27.36	0.094
Sex	Male	1.00
	Female	1.60	0.84	3.06	0.153
	Non-binary sex/prefer not to say	0.66	0.14	3.05	0.594
Sexual identity	Straight or heterosexual	1.00
	Lesbian, gay or homosexual	0.98	0.30	3.20	0.979
	Bisexual	0.54	0.23	1.30	0.170
	Something else	0.48	0.10	2.39	0.374
	Don’t know/prefer not to say	1.05	0.12	8.88	0.966
Country of birth	Australia	1.00
	Overseas	0.74	0.36	1.55	0.426
First Nations status	Not First Nations	1.00
	First Nations	0.60	0.08	4.21	0.605
Area of residence	Major city	1.00
	Regional, rural and remote	1.46	0.73	2.94	0.287
SEIFA IRSD quintile b	Q1 (most disadvantaged)	1.00
	Q2	1.30	0.50	3.38	0.590
	Q3	1.49	0.59	3.77	0.404
	Q4	1.30	0.47	3.57	0.614
	Q5 (least disadvantaged)	1.54	0.57	4.17	0.392
Self-rated mental health	Before care	0.47	0.40	0.55	0.000
Provider type c	Clinical psychologist	1.00
	Psychologist	0.87	0.44	1.70	0.676
	Social worker	1.22	0.56	2.63	0.619
	Occupational therapist	3.61	0.87	15.09	0.078
No. of sessions d	1–2	1.00
	3–4	2.66	1.03	6.82	0.042
	5–6	3.18	1.15	8.78	0.026
	7–10	2.19	0.95	5.09	0.067
	11 +	4.18	1.67	10.48	0.002
Face-to-face sessions d	No	1.00
	Yes	0.80	0.38	1.68	0.564
Telehealth sessions d	No	1.00
	Yes	0.57	0.27	1.20	0.140
Phone sessions d	No	1.00
	Yes	0.83	0.39	1.81	0.646
Out-of-pocket costs d	No	1.00
	Yes	4.08	2.17	7.69	0.000

^aMissing data excluded.

^bSocioeconomic status was ascribed to participants on the basis of their postcode, using quintiles derived from the Index of Relative Socioeconomic Disadvantage (IRSD) of the of the Socioeconomic Indexes for Areas (SEIFA). More specifically, the SEIFA concordance file was used to assign the IRSD. The IRSD file reports deciles that were then converted into quintiles.

^cProvider type according to Services Australia.

^dSession-related information according to MBS claims data from Services Australia.

Discussion

This study presents a positive picture of Better Access from the perspective of a large, diverse sample of consumers who saw a range of providers and received varying amounts of care delivered via different modalities. Most valued their relationship with their mental health professional and believed that the strategies that the mental health professional gave them met their needs. The vast majority were satisfied with their care. A smaller but still substantial proportion noted aspects of their care that could be improved. Some ceased their sessions early because they did not find them helpful or did not like the mental health professional’s manner or approach. Some considered that they had to wait too long for an appointment, and some found the out-of-pocket costs too high. These findings are broadly consistent with those from our 2009–2011 evaluation (Ftanou et al., 2014; Pirkis et al., 2010a, 2010b, 2011a, 2011b). It is difficult to make comparisons with consumer experiences from overseas because of differences in providers/programmes and evaluation methods, but a 2013 survey of 15,078 Improving Access to Psychological Therapies (IAPTs) service users in England/Wales yielded similar positive results, with 81% indicating they were satisfied with their experience of treatment (Royal College of Psychiatrists, 2013).

The issue of out-of-pocket costs is especially salient; other studies in our evaluation showed that the proportion of Better Access sessions attracting a co-payment and the median co-payment amount have increased over time (Tapp et al., 2026b) and that consumers, providers and other stakeholders rank affordability as a top reform priority (Currier et al., 2026). Other studies have also found that out-of-pocket costs are influencing decisions about mental health care. For example, the Australian Bureau of Statistics (2023) Patient Experiences Survey found that 24.4% of those who needed to see a psychologist delayed doing so or did not do so at all at least once because of cost. The equivalent figure for those who needed to see a social worker, OT, mental health nurse and/or counsellor was 18.6%.

Overwhelmingly, participants experienced good outcomes from their Better Access care. Again, this is in line with the findings from our previous evaluation (Ftanou et al., 2014; Pirkis et al., 2010a, 2010b, 2011a, 2011b). In the current study, the self-rated mental health of 91% of all participants improved. A majority attributed this improvement – at least in part – to the treatment they received from the mental health professional. This latter point is critical because although other studies in our evaluation also showed that Better Access treatment was associated with positive outcomes on various standardised measures, all relied on pre- and post-data collected for different purposes, and it wasn’t possible to definitively attribute improvements to Better Access (Arya et al., 2026; Harris et al., 2026; Pirkis et al., 2026).

Those who entered their Better Access episode of care with worse self-rated mental health were significantly more likely to show improvement. This makes sense; those who begin treatment with severe mental health symptoms have greater opportunity for improvement. It is also consistent with the broader literature on the outcomes of psychotherapy (Amati et al., 2018).

Those who received more than one to two sessions also had greater odds of improvement than those who received one to two sessions. There was some evidence of a dose–response effect, with additional levels of improvement associated with increasingly higher numbers of sessions, but the 95% CIs overlapped, so we were not able to determine the optimal number of sessions. A meta-regression that examined the relationship between the amount, frequency and intensity of therapy and outcomes for people with depression also only found a small association between increasingly greater numbers of sessions and improvement (Cuijpers et al., 2013). As noted, the permissible number of sessions available annually through Better Access reverted to the pre-pandemic 10 after we completed the evaluation (Australian Government Department of Health, 2022b).

The findings from the sub-sample of participants with linked MBS data reinforce those from the full sample; they had similar patterns of care, showed similar improvements in self-rated mental health, and the strongest predictors of improvement for them were baseline severity and number of sessions. The findings from the sub-sample also provide additional detail about issues relating to costs and affordability. On one hand, they show that the more people pay out-of-pocket, the less affordable they find their care; the median per-session co-payment was higher for participants who thought their care was too expensive than for participants who thought it was affordable (AUD$77.55 versus AUD$61.75). On the other hand, those who paid a co-payment had a greater likelihood of reporting improvement in self-rated mental health, perhaps because contributing to the cost of care can lead to greater commitment to treatment (Aubry et al., 2000).

A final important finding relates to diagnosis. Over half of our participants were given a diagnosis at the time of seeking Better Access care (usually anxiety and/or depression). However, one third said that they were not given a diagnosis. This warrants further exploration, given that having a diagnosed mental disorder is an eligibility requirement for Better Access. Some of these consumers may have been given a diagnosis but did not recall this or were not made aware of the specific diagnosis. There may also be other issues at play, including inappropriate referrals, inadequate communication between providers and referrers, and stigma.

Strengths and limitations

We recruited a substantial, diverse sample (n = 2013) of known Better Access users. However, our response rate was 7.4% and some groups were under-represented (First Nations people) or not represented (children). We know from elsewhere in the evaluation that in 2021, 15% of First Nations people and 3% of children used Better Access (Chilvers et al., 2026b). Our over-sampling of particular groups (e.g. those in non-metropolitan areas) ensured their participation but may also have implications for generalisability.

Because we recruited participants in early-2022 and our selection criteria relied on people having received Better Access services in 2021, two thirds of our participants were still receiving care when they completed the survey. This meant that many participants were reflecting on their experiences when their care was ongoing. It also meant that we could not capture outcomes using standardised measures administered at baseline and follow-up, although, as noted, this was achieved in other studies in our evaluation which also showed positive results (Arya et al., 2026; Harris et al., 2026; Pirkis et al., 2026).

The survey’s reliance on self-report may have introduced recall and social desirability biases (e.g. participants may have had difficulty remembering what their mental health was like before and after their episode of mental health care, and/or may have been inclined to indicate that it was better afterwards).

Conclusion

In general, consumers who use Better Access appreciate the programme and benefit from the care it provides. Those who enter care with particularly poor self-rated mental health and receive more than one to two sessions benefit the most. Despite these generally positive findings, affordability remains an issue.