Lost in the system: lived experiences of eating disorder service models in Australia Authors: Rachel Knight, Karen Stagnitti, Danielle Hitch, Genevieve Pepin Published: 2026 ABSTRACT BackgroundIn eating disorders, service models provide a structure for the delivery of assessment and treatment to people across a range of settings. High-quality service models can improve a person’s satisfaction with the care provided, sustain them in treatment and promote better outcomes.ObjectiveTo capture and explore the perspectives of former service users, carers and eating disorder clinicians on their experiences of eating disorder service models.MethodsThis phenomenological study used semi-structured interviews to explore perspectives of former Australian service users (n = 5), carers (n = 15) and eating disorder clinicians (n = 29) on eating disorder service models.ResultsThematic analysis identified four themes: (1) Navigating the service system, (2) Weight as a measure of progress, (3) Competent clinicians: a needle in a haystack and (4) Who is steering the ship? The first three themes were shared across all participant groups, while the fourth was specific to clinicians. Service models were described by participants as overstretched, contributing to shorter treatment duration, with skilled clinician shortages further constraining care. Participants reported over-reliance on weight to determine service model access (and thus, access to treatment) and as a trigger for discharge, despite consensus that recovery cannot be measured by weight alone. To enhance eating disorder service models, participants proposed strengthening the workforce, facilitating timely access, using more holistic outcome measures, and tailoring care to individual needs within coordinated, stepped service pathways.ConclusionThese findings highlight opportunities for system-level reform and the need for clearer leadership to guide consistent standards and accountability.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40337-026-01680-0. Introduction Service models are multidimensional frameworks that define how healthcare is organised and should be delivered to address the heath related needs of people [1, 2]. Service models specify the structures, processes and resources required to ensure coordinated and consistent care for a defined group of people [2]. When clearly articulated, a service model provides a consistent approach to delivering care that aligns with the health-related needs of people who access it. Furthermore, effective service models will maintain or improve the health of a specific group of people by ensuring care is accessible, timely and high quality [2]. Service models exist across service settings. Service settings are environments where a service model is based, examples of service settings include hospital or the community [3]. Australians typically access eating disorder care through public specialist and general mental health service models. A recent rapid review of eating disorder service models found that specialist service models improved identification of eating disorders, increased treatment completion and reduced mortality rates [3]. It was further identified that integration across service models has been rarely examined, and a stepped-care system integrating prevention to intensive hospital-based service models was recommended as it could provide more comprehensive, personalised and cost-effective support [3]. Because eating disorders can be challenging to diagnose, there are often delays in accessing service models that can provide assessment, support and treatment for eating disorders [4]. Primary care service models are typically where people first present for help [3]. Clinicians within these service models (such as general practitioners) have a crucial role in identifying eating disorders and referring to appropriate service models for ongoing care [4, 5]. However, unskilled primary care clinicians have been well documented as a barrier to detecting eating disorders [3, 6, 7]. Consequently, it is estimated that only 17%−31% of people with eating disorders commence treatment within recommended timeframes [8, 9]. An Australian study of people aged between 12 and 60 years old reported an average gap of 5.28 years between symptom onset and treatment, a delay reported across age groups and diagnoses [10]. These delays can entrench illness and complicate engagement with service models; conversely, earlier detection and treatment are consistently associated with better recovery [10, 11]. Once a person accesses services, their experiences of service models strongly influence engagement with treatment. A meta-synthesis of 22 studies, mostly from the United Kingdom (UK) (n = 13), Australia (n = 5) and the United States of America (USA) (n = 3), explored treatment experiences, with an emphasis on treatment provided in hospital-based service models, from the perspective of people with eating disorders [12]. The importance of collaborative, person-centred care, delivered by knowledgeable and supportive clinicians enhanced hope, perceived treatment effectiveness, and recovery. Carer studies similarly emphasised person-centred care and reported difficulty accessing expertise, strict eligibility criteria to access service models, and under-resourcing of service models [13, 14]. Further research has examined the experiences of people with eating disorders receiving inpatient care. Systematic reviews and meta-syntheses of studies conducted with people who have received treatment in hospital-based service models consistently highlight the importance of clinician expertise and the value of person-centred approaches to treatment [15, 16]. At the same time, these reviews identify challenges associated with service model processes and protocols (such as restrictions on movement or visitors) which may be perceived as unhelpful or counterproductive [16, 17]. In addition, characteristics of the inpatient environment itself, such as non-negotiable boundaries and predictable routines, have been shown to shape a person’s recovery experience [17]. These findings underscore a critical gap between the design of hospital-based service models and how care is experienced in practice. Most studies of eating disorder service models focus on a single stakeholder group, like people with lived experience or carers. However, a systematic review and thematic synthesis of 63 qualitative (n = 44) and quantitative (n = 24) studies (some using mixed methodologies) from ten countries, found broad agreement among people with eating disorders, carers and clinicians about the barriers to quality care [18]. These included delayed diagnosis and treatment, disrupted continuity of care, and long waiting lists. The authors emphasised the need for greater collaboration and workforce capacity to improve continuity across settings [18]. Current evidence describing perspectives and experiences of eating disorder service models largely arises from studies that focus on treatment experiences, leaving scope to build on findings specific to service models. Much of this literature focuses on people in inpatient settings [15–17, 19–21], and comparatively few primary studies include more than one stakeholder group. To date, most studies are predominantly from the UK or the USA. This study addresses that gap by exploring the perspectives and lived experiences of former service users, carers and clinicians regarding current eating disorder service models in Victoria, Australia. Methods Phenomenology was selected to examine how participants make sense of eating disorder service models and their impacts [22, 23]. This study adopted a descriptive phenomenological approach, which captures detailed, subjective experiences, to better understand the phenomenon and its meaning for participants [22, 24, 25]. Ethics approval was granted by Deakin University Human Research Ethics Committee – Health, and reporting followed the COREQ (COnsolidated criteria for REporting Qualitative research) Checklist (Additional file 1) [26]. Researcher Positionality/Stance In phenomenological research, investigators remain open and curious while recognising that their beliefs and experiences inevitably shape the process [27, 28]. Acknowledging this subjectivity, supervision, peer debriefing and field notes were used to address preconceptions, minimise bias, and document authentic accounts of participants’ experiences. All members of the research team are female occupational therapists. RK, DH and GP bring extensive clinical experience in mental health and eating disorders, alongside local, national and international knowledge of mental health and eating disorder service models and systems. Both RK and DH have provided clinical services within the local service context and the wider system, in which this study is situated, providing them with in-depth contextual insight. In her clinical work, RK observed inconsistencies across eating disorder service models, which prompted her interest in examining this under-researched area. GP is an experienced qualitative researcher whose work has focused on understanding carer experiences in eating disorders. Through her clinical experiences providing eating disorder treatment in Canada, she has also witnessed the impact of limited skilled workforce and difficulty accessing eating disorder treatment across service models. DH, working primarily within a health service organisation, brings a translational research focus with expertise in complex care. KS is an academic with expertise in play therapy and community-based interventions. Although she does not specialise in eating disorders, she contributes a broader perspective on health and service systems, supporting critical reflection beyond the immediate eating disorders context. Sample The study included three participant groups: former service users, carers, and clinicians within Victorian public mental health services. Former service users who self-identified as recovered and had received assessment or treatment for an eating disorder were included, along with carers who currently or previously supported someone who had received these services. Clinicians working in specialist or generalist service models providing eating disorder care were also included. All participants were aged ≥ 18 years; those not meeting the criteria were excluded. There are no predetermined rules for sample size in qualitative research, which prioritises rich information and relevance to the research question [29]. A systematic review of descriptive phenomenology research reported sample sizes from 1 to 31 participants, and no association between sample size and study quality [30]. Information power is a concept used to determine the adequacy of sample size in qualitative research [31]. Information power considers factors like the aim, scope, sample specificity and analysis strategy. The higher the information power, the smaller sample required [31]. In this study, a specific target group, in-depth analysis and interviews by clinically experienced researchers (RK, GP) enhanced the information power of the sample recruited [31]. Therefore, a sample of 5–10 former service users or carers was determined. A larger number of clinicians (n = 30) was targeted. This decision was made to ensure representation from child, youth and adult public mental health services in Victoria, including those with specialist eating disorder models, to capture a broad range of experiences across different service models. Recruitment Participants were recruited via purposive sampling and snowball sampling, to capture participants with relevant knowledge and experience [32–34]. Statewide and national eating disorder organisations (e.g., the Victorian Centre of Excellence in Eating Disorders) disseminated study information through professional networks, and a guest blog post by a recovered advocate for eating disorders included research team contact details. GP screened enquiries for eligibility, provided a plain-language statement and consent form, and obtained written consent before data collection commenced. Data collection Semi-structured interviews captured participants’ subjective experiences [22]. Individual interviews were conducted face-to-face or by phone (~ 60 min) by RK, using bespoke guides developed specifically for this study (Appendices 1 and 2) as no existing interview guide had been identified in existing literature. Questions were developed to be neutral, ensuring that the strengths and limitations of current service models could be explored. The interview guide was pre-tested with three clinicians, and three carers and no changes were recommended [35]. Data was not collected in the pre-testing of the interview guide. Rather, pre-testing focused on ensuring the questions were neutral, suitable for the study and prompted responses relevant to the study. The interview guide had two sections. The first section recorded age, gender, diagnosis, co-occurring conditions, age at treatment onset, and hospital admissions for former service users and carers (regarding their loved ones). Professional background, highest qualification, years of experience in mental health and eating disorders, and current workplace setting/location were recorded for clinicians. The second section explored service model perceptions and experiences, including accessibility, treatment options, care teams, and perceived strengths/limitations. Interviews were audio-recorded and transcribed verbatim by RK and a transcription service. Participants reviewed these transcripts for accuracy [36], and a reflexive field journal supplemented interview data. Pseudonyms replaced names and identifying information was removed to maintain confidentiality. Data analysis Data were analysed using Braun and Clarke’s [37] six-step thematic analysis in NVivo (Version 12) [38]. RK and GP immersed themselves in the data and independently generated inductive codes (steps 1–2), collated codes into themes (step 3), reviewed and refined themes through face-to-face discussions to consensus (steps 4–5) and completed final refinement of the analysis during the write-up (step 6). Trustworthiness Trustworthiness was addressed using established strategies targeting credibility, transferability, dependability and confirmability [39, 40]. Credibility was supported by study design, reflexive field journaling and participant transcript review, and investigator triangulation occurred through multiple researchers contributing to design and analysis [41]. Dependability was achieved with a detailed methodological and analytical description, while transferability was strengthened through the inclusion of multiple perspectives from three participant groups [40]. Confirmability was supported by acknowledging research subjectivity throughout the analysis. Results Participants Forty-nine participants were interviewed: former service users (n = 5), carers (n = 15) and clinicians (n = 29). The overall sample size aligns with descriptive phenomenology norms [23]. Findings were interpreted cautiously, given the small service user subsample and differences between groups. All former service users were female (18–42 years) with a history of anorexia nervosa. Nearly two-thirds (60%) had co-occurring anxiety/depression. Most (80%) were diagnosed and began treatment at 10–19 years, with treatment received in the last 2–5 years (60%) or > 5 years (40%). Their care was primarily through metropolitan child and youth specialist services (inpatient and outpatient) with some participants accessing more than one type of service model. Because of the small sample size and to maintain confidentiality, further demographic information is not provided in the Participant Demographics and Characteristics table (Table 1). Carers were mostly female (93%), aged 20–59 years, with just over half (53%) aged over 50 years. They supported loved ones with diagnosed anorexia nervosa (73%), who were diagnosed and began treatment aged 10–19 years (87%). Most reported co-occurring conditions (67%), and most of their loved ones remained in treatment (80%). Their experiences spanned multiple service models, but predominantly metropolitan Child and Youth Mental Health Services. No carers were related to participating former service users. Clinicians were predominantly female (76%) and aged 30–39 (48%). They included clinicians from multiple disciplines, including social work (37%), psychology (21%), occupational therapy (14%), nursing (7%) and dietetics (7%). Just over half (55%) held a masters degree, and most (59%) worked in metropolitan Child and Youth Mental Health Services. They typically had less than 10 years’ experience working in eating disorders (62%). See Table 1 for full participant demographics. Table 1Participant demographics and characteristicsParticipant groupNumber (n)Percentage (%)Carers GenderMale17Female1493 Age18–291730–390040–4964050 or above853 Age of loved one (at time of eating disorder diagnosis)10–19138720–29213 Age when loved one commenced treatment10–19138720–29213 Eating Disorder diagnosisAnorexia1173Atypical Anorexia213Bulimia17EDNOS17 Co-occurring conditions (of loved one)Yes1067No533 Co-occurring condition diagnosis (of loved one)cAnxiety533Depression427Autism320Bipolar Affective Disorder17OCD17 Hospital admissions (of loved one)Psychiatric5Medical9 Recency of eating disorder treatment (for loved one)Currently ongoing1280Previous 5 years2135 + years ago17 Public service settings accessed for eating disorder assessment and/or treatment (for loved one)dAdult ED inpatient213Adult ED day program213Adult MH outpatient17C&Y ED inpatient427C&Y Medical inpatient213C&Y ED day program427C&Y ED outpatient320C&Y MH outpatient960 Service setting locationeMetropolitan1280Regional/rural427Eating disorder clinicians GenderMale724Female2276 Age18–292730–39144840–4951750 or above828 DisciplineSocial work1137Psychology621Occupational Therapy414Psychiatry414Nursing27Dietetics27 Highest level of qualificationBachelor517Post Graduate Diploma414Masters1655Fellowship414 No. years working in eating disordersLess than 10186210–1982820–292730–3913 Service settingMH Child & Youth1759MH Adult27ED Child & Youth517ED Adult517 Service setting locationMetropolitan1965Regional621Rural414 ED eating disorder (specialist), MH mental health (generalist), C&Y child and youth, EDNOS Eating disorder not otherwise specified, PTSD Post traumatic stress disorder a Some participants reported multiple co-occurring conditions b Some participants accessed more than one type of service c Some participants reported multiple co-occurring conditions d Some participants accessed more than one type of service e The loved one of one carer accessed service models in both metropolitan and regional/rural areas Table 1Participant demographics and characteristicsParticipant groupNumber (n)Percentage (%)Carers GenderMale17Female1493 Age18–291730–390040–4964050 or above853 Age of loved one (at time of eating disorder diagnosis)10–19138720–29213 Age when loved one commenced treatment10–19138720–29213 Eating Disorder diagnosisAnorexia1173Atypical Anorexia213Bulimia17EDNOS17 Co-occurring conditions (of loved one)Yes1067No533 Co-occurring condition diagnosis (of loved one)cAnxiety533Depression427Autism320Bipolar Affective Disorder17OCD17 Hospital admissions (of loved one)Psychiatric5Medical9 Recency of eating disorder treatment (for loved one)Currently ongoing1280Previous 5 years2135 + years ago17 Public service settings accessed for eating disorder assessment and/or treatment (for loved one)dAdult ED inpatient213Adult ED day program213Adult MH outpatient17C&Y ED inpatient427C&Y Medical inpatient213C&Y ED day program427C&Y ED outpatient320C&Y MH outpatient960 Service setting locationeMetropolitan1280Regional/rural427Eating disorder clinicians GenderMale724Female2276 Age18–292730–39144840–4951750 or above828 DisciplineSocial work1137Psychology621Occupational Therapy414Psychiatry414Nursing27Dietetics27 Highest level of qualificationBachelor517Post Graduate Diploma414Masters1655Fellowship414 No. years working in eating disordersLess than 10186210–1982820–292730–3913 Service settingMH Child & Youth1759MH Adult27ED Child & Youth517ED Adult517 Service setting locationMetropolitan1965Regional621Rural414 ED eating disorder (specialist), MH mental health (generalist), C&Y child and youth, EDNOS Eating disorder not otherwise specified, PTSD Post traumatic stress disorder a Some participants reported multiple co-occurring conditions b Some participants accessed more than one type of service c Some participants reported multiple co-occurring conditions d Some participants accessed more than one type of service e The loved one of one carer accessed service models in both metropolitan and regional/rural areas Participant demographics and characteristics ED eating disorder (specialist), MH mental health (generalist), C&Y child and youth, EDNOS Eating disorder not otherwise specified, PTSD Post traumatic stress disorder a Some participants reported multiple co-occurring conditions b Some participants accessed more than one type of service c Some participants reported multiple co-occurring conditions d Some participants accessed more than one type of service e The loved one of one carer accessed service models in both metropolitan and regional/rural areas Findings Data analysis generated four themes: (1) Navigating the service system (2) Weight as a measure of progress, (3) Competent clinicians: a needle in a haystack and (4) Who is steering the ship? Themes 1, 2, and 3 were shared by all participant groups, but theme 4 was identified by clinicians only, as seen in Fig. 1. Fig. 1Themes per participant group Themes per participant group Theme 1: Navigating the service system This theme discussed how people move through eating disorder service models, from initial access to treatment to discharge, and how navigation was experienced as complex and confusing. Eligibility criteria were a major access barrier, especially for people with multiple needs (co-occurring conditions, medical/psychiatric risk, or low treatment motivation). Clinicians and carers expressed their frustration that specialist eating disorder service models were not transparent ‘about who they will or who they won’t see’, and that community-based eating disorder specialist service models often did not accept people deemed ‘complex’ (such as those with co-occurring conditions, medical or psychiatric risks and/or low motivation) leading many to receive care in generalist community-based mental health service models where clinicians had less eating disorder expertise. As one clinician noted, ‘you would think the specialist eating disorder services would pick them up, but it doesn’t work that way’ (Alice, Clinician). Geography also constrained access. Services typically operated within catchment areas, creating inequities where eating disorder service models were absent. One carer explained, ‘So, we have no choices about where to go, despite the fact I would happily throw money at it’ (Gerry, Carer). Participants highlighted the absence of ‘a consistent model of care for people with an eating disorder’ meaning that ‘the type of care or the quality of care or the intensity of care is determined by where you live’, which compounded pre-existing inequities. Carers and service users from regional and rural areas spoke about the measures they had to take to access care, including significant travel and for some, relocating where they lived. Clinicians highlighted the underrepresentation of some groups accessing treatment in specialist eating disorder service models (both hospital and community -based), such as people with binge eating disorder and Aboriginal and Torres Strait Islander peoples, ‘I don’t know that we do diversity too well. We’re versed in it, we’ve trained in it, but … if we look at what our demographics are …it doesn’t represent the demographic in the geographical catchment area’ (Laura, Clinician). After service entry, participants reported difficulties navigating treatment and recovery. They wanted greater choice and flexibility, and while person-centred care was viewed as essential, it was rarely experienced. One former service user explained that while receiving inpatient eating disorder treatment, ‘they would just automatically, just turn to medication, in both the public and private, rather than try other things’ (Audrey, Former Service User). Clinicians linked an inability to provide person-centred care to limited resources, funding constraints, and pressure for shorter treatment episodes. Many service models prioritised short-term outcomes, with service users expressing ‘longer would have been better’, when it came to the length of time they received outpatient treatment for their eating disorder. People were often discharged if they did not progress within expected timeframes with little follow-up, shifting the burden to carers. As one carer recounted:The community-based team said to us that they couldn’t work with her any longer, that they weren’t making any headway with her psychological state. So, they discharged her from their service and referred her to the acute team. And the acute team, we were led to believe, were going to call us that afternoon, but they didn’t phone. I phoned them on the Monday, and they saw us on the Monday afternoon, and they discharged her from their service because she didn’t have a suicide plan (Brianna, Carer). The community-based team said to us that they couldn’t work with her any longer, that they weren’t making any headway with her psychological state. So, they discharged her from their service and referred her to the acute team. And the acute team, we were led to believe, were going to call us that afternoon, but they didn’t phone. I phoned them on the Monday, and they saw us on the Monday afternoon, and they discharged her from their service because she didn’t have a suicide plan (Brianna, Carer). Participants from all groups expressed a desire for service models to be more flexible. Specifically, carer and service user participants had sought out intensive outreach eating disorder care (that helped keep people out of hospital), and carers living in rural areas of Victoria wanted inpatient units that could accommodate the entire family but noted this service model did not exist or were unavailable. Theme 2: Weight as a measure of progress This theme, exploring strategies to monitor recovery, revealed over-reliance on weight, with one participant noting, ‘once your weight is restored, it didn’t really matter anymore’ (Cassandra, Previous Service User). Participants agreed that weight offers a narrow view of recovery. Bella, a carer participant expressed, ‘we were the stars of family based treatment, because she put on weight very quickly, but mentally, she wasn’t right’. Carers described ongoing symptoms and struggles with the eating disorder despite their loved one reaching their ‘expected’ weight, often viewing weight targets set by outpatient clinical teams as too low. One carer expressed that ‘everybody’s got to get rid of this fat phobia’, and others emphasised that specialist eating disorder outpatient clinical teams did not always acknowledge the carer perspective:When they [clinical team] gave me her target weight I was like “I know my kid, that’s too low …she’s never going to be well” …if I’m telling them, no that’s not enough weight, they need to take that on board (Willow, Carer). When they [clinical team] gave me her target weight I was like “I know my kid, that’s too low …she’s never going to be well” …if I’m telling them, no that’s not enough weight, they need to take that on board (Willow, Carer). Perspectives on weight and expected weight outcomes differed between participant groups. Many carers preferred less focus on a specific number, ‘I mean, it helps us monitor progress, but they don’t need to know what the numbers are’; while some former service users wanted a transparent discussion of weight targets, with Emma explaining, ‘talking about a health weight range is fine until you go over a BMI of 24, which is what I did’. Others rejected weight-focused treatment as misaligned with the core problem, as one clinician working in a community-based service model shared,‘by and large it’s about health restoration, and when I say health restoration, I don’t just mean weight. I mean a correction of relationships that have gone off track or being able to do what someone of their age should be able to do in terms of school or study or friendships or developmental tasks’ (Brooke, Clinician). ‘by and large it’s about health restoration, and when I say health restoration, I don’t just mean weight. I mean a correction of relationships that have gone off track or being able to do what someone of their age should be able to do in terms of school or study or friendships or developmental tasks’ (Brooke, Clinician). All stakeholders observed that discharge decisions were chiefly weight based rather than holistic, with people sometimes discharged upon nearing target weight despite incomplete recovery. Theme 3: Competent clinicians: a needle in a haystack Participants highlighted the value of competent eating disorder clinicians and the difficulty in finding them. Delayed diagnosis was common at first contact (often within primary care) and participants attributed this to clinicians receiving limited training in eating disorders and their early signs and therefore lacking confidence. One clinician identified that ‘there is a lot of fear and a lot of people not feeling confident working in eating disorders, therefore they are kind of shying away from it’ (Rebecca, Clinician). Some participants reported that primary care clinicians dismissed their concerns and/or suspicion that their loved one had an eating disorder, as one carer recalled, ‘…and I said, well she has lost quite a bit of weight, she said ‘oh that just how, you know, they grow and they lose weight and that is normal’, and she said there was nothing wrong’ (Mila, Carer). Some participants had also experienced stigma and discrimination from clinicians who lacked expertise in eating disorders. Reflecting on her admission to a generalist mental health inpatient unit because of her eating disorder, Nelly shared; ‘it has taken me longer to get over the comments that nurses and that have said to me, than the actual eating disorder’ (Nelly, Previous Service User). A carer recalled that while her loved one was receiving treatment in an eating disorder inpatient unit; ‘I had tears on the phone [from her] after inappropriate comments from an agency nurse, who wasn’t trained or experienced in eating disorders’. On the other hand, former service users highlighted the value of ‘empathetic and compassionate’ clinicians who understood their experience of an eating disorder, as Rose explained; ‘the people who helped me the most have understood what it is like but then also understand how my brain works. Like, they understand me and also the illness’. Specialist eating disorder clinicians were few, in high demand and overworked, meaning that accessing clinicians in service models could be ‘…like trying to get an audience with the Pope. You just cannot get in to see them’ (Willow, Carer). One previous service user blamed themselves for the difficulties they experienced in finding a knowledgeable and helpful clinician during recovery, stating ‘I think it’s me, after seeing like six different people or something. I think I am the issue’. Workforce shortages were also perceived as driving high staff turnover, which also disrupted continuity of care within all service models who support people with eating disorders;People are away or on holidays, or sick or leaving and it gets handed over to someone else, and we had that happen a lot and things were just really poorly managed… I thought the people we dealt with were lovely, wonderful human beings, but they are clearly, whether it’s a time issue, whether they’ve got too much workload, I don’t know what it is, but the whole thing is just a bit of a dog’s breakfast basically (Aggie, Carer). People are away or on holidays, or sick or leaving and it gets handed over to someone else, and we had that happen a lot and things were just really poorly managed… I thought the people we dealt with were lovely, wonderful human beings, but they are clearly, whether it’s a time issue, whether they’ve got too much workload, I don’t know what it is, but the whole thing is just a bit of a dog’s breakfast basically (Aggie, Carer). Some participants identified service model leaders (within both generalist and specialist service models) had taken steps to build and improve competence, capacity and continuity of care by mandating eating disorder training for all clinicians or allocating two or more clinicians per person ‘to get some consistency’ so a familiar clinician was usually available. Furthermore, Carl, a clinician working in a community based-service model identified that their ‘biggest strength’ was that they had ‘physical and mental health care in the same place’ which had resulted in improved integration and meant that the eating disorder service model had ‘staff with skills in multiple areas, to deal with the complexity of the presentation’. Theme 4: Who is steering the ship? This theme was specific to clinicians and explored the interface between different service models and who provides leadership for eating disorders. Integration across service models was seen as critical for treatment quality and safety, yet clinicians noted the absence of ‘minimum standards’ and limited accountability; with one clinician noting ‘it is extraordinary that there is no uniformity of criteria for admission for medical rescue for paeds or adults across the state’ (Alice, Clinician). Regional clinicians prioritised building ‘good relationships with other service providers. So, there’s very little opportunity to fall in between services or have the experience of patients being handballed on’ (Oliver, Clinician) however, this was led by clinicians rather than leaders. Effective collaboration was supported by clear communication channels, joint care-team meetings and teams being co-located. Eating disorders leadership was viewed as essential to the delivery of high-quality care. ‘Less and less coordination and leadership which means that things have become fragmented, if service providers know it is fragmented, the consumers will be feeling it…’ (Debbie, Clinician). Clinicians called for leadership at multiple levels: within teams (clinical support), from state-wide peak bodies (consultation, training and education), and nationally (consistent standards, meaningful outcomes and targets, and more effective resource distribution). Discussion This study addresses experiences of eating disorders models from the perspective of multiple stakeholder groups, providing more nuanced findings than single group research. All participant groups experienced eating disorder service models as complex and inconsistent, often leading to inequitable support and treatment. A unique finding was that clinicians identified a need for stronger sector leadership. The findings have highlighted several areas where service models could be improved to enhance the quality of care. Many participants, especially in regional or underserved metropolitan areas, reported exclusion from treatment due to service model eligibility criteria and geographical zoning. Consistent with Johns et al. [18], people with complex presentations often did not meet the criteria, leaving those with the greatest need without specialist care. Participants also noted limited person-centred care and service models that are designed to focus on short, time-limited episodes to boost throughput and reduce wait lists, which conflicts with empirically supported treatment models, recommending up to 40 sessions of psychological therapy [42]. For example, a Danish study found 57% of families required 12 months to complete Family Based Treatment [43], and optimal duration varies by diagnosis, age, co-occurring conditions and clinical complexity, supporting the argument for flexibility put forward by participants in this study [42, 44]. Participants reported that service models often focused on weight as a sole recovery indicator, which they felt was insufficient in determining recovery. This likely reflects models that are designed primarily for anorexia nervosa, where weight restoration is necessary to support health and recovery [45]. A rapid review of peer-reviewed literature on treatment outcomes identified only one weight-related predictor - higher baseline BMI - and this only for anorexia nervosa [46], while a Dutch study found that lower depression scores and shorter illness duration were strongest overall predictors [47]. Participants in this study agreed recovery is complex and not determined by weight alone, underscoring the need to recognise non-weight predictors within person-centred treatment and support provided in service models. Aligning with previous research, this study underscores the pivotal role of primary care (e.g., general practitioners) and service models in identifying eating disorders. An unskilled workforce remains a key barrier to timely detection, particularly in primary care, where people typically first seek help [3]. Beyond primary care, gaps in specialist knowledge persist across the broader mental health workforce in service models. To address capability, Australian workforce competencies for eating disorders delineate roles and required skills across settings [48]. Nonetheless, shortages of clinicians skilled in eating disorders continue [4, 10], reflecting wider mental health workforce challenges which have been attributed to stigma, rising demand and unclear scopes of practice [49]. Potential solutions to an unskilled workforce include hub-and-spoke knowledge transfer models, as exemplified by the remote consultation and education provided by Project ECHO in the United States [50], and the statewide model in Queensland, Australia, which offers eating disorders advice and education to less experienced clinicians from a central hub [51]. Broader adoption of these strategies could address workforce shortages more effectively and improve clinician competence. A distinctive finding of this study was clinicians’ perception of weak sector leadership and calls for stronger, more visible structures to guide development. Evidence from healthcare reform underscores that clear leadership and governance are critical for system integration and coordination [52, 53]. In eating disorders, leadership has enabled the reform and development of large-scale service models [51, 54]; for example, the New South Wales Service Plan was achieved through collaboration between government, health executives and specialist institutes [55]. Together, these findings suggest strong leadership structures and development should be part of service and model design going forward. Findings from this study reinforce that current service models who support people with eating disorders are highly varied, challenging to access, and are often unable to deliver authentic, recovery-oriented, person-centred care due to factors at the system level. Key contributing factors include strict eligibility criteria for service models, the absence of established minimum standards for care and outcomes, workforce shortages and skill gaps, and a lack of clear, visible leadership. Several effective and innovative service models already exist internationally and could guide service model reform in Australia. With a 21% increase in the number of Australians living with an eating disorder since 2012 [56], there is an urgent imperative to act, change, and improve eating disorder service models. Study strengths and limitations. The qualitative design of this study captured rich descriptions of participants’ experiences and perspectives [33, 57]. Including former service users, carers and clinicians provided a multidimensional view of strengths, limitations and opportunities for improvement of eating disorder service models, and yielded novel insights such as the need for stronger leadership. The limitations of this study include its conduct in a single jurisdiction (Victoria, Australia) and challenges in recruiting former service users, which limits transferability. Despite strategies to enhance recruitment, lived experience participation can be constrained by ongoing symptoms, stigma or reluctance to revisit past experiences [58]. The findings should therefore be interpreted cautiously given differences between groups and the smaller service user subsample. Most service users and carers reported a diagnosis of, or caring for, a female with anorexia nervosa, reflecting the focus of many current service models but limiting perspectives from other genders and diagnoses. Conclusion The perspectives and lived experience of 49 former service users, carers and clinicians were captured across four themes. These themes identified that eating disorder service models can be difficult to navigate and are often perceived as overly focused on weight during recovery. Clinician expertise in eating disorders appears to be insufficient to meet demand and the sector would benefit from greater leadership to guide reform. This study highlights that better coordination, integration and streamlined access is required for eating disorder service models. Care provided in these models should be person-centred and tailored to individual needs, with outcomes (extending beyond weight) co-developed with service users and carers. Strong, inclusive leadership is also essential to improve equity and consistency of care, and build a sustainable, skilled workforce. As this study focused on service models rather than treatments, future research should examine how service models and workforce capability intersect with treatment type and duration to shape recovery and experiences. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.