Innovation in Acute Adolescent Mental Health Care: Evaluating a Responsive and Sustainable Mental Health Care Model

1. Introduction

Mental health conditions frequently emerge during childhood and adolescence, and are related to a higher risk of adult mental disorders [1]. This is a global health concern, with profound effects on families and communities, as about one in seven adolescents will experience serious mental health problems [2], accounting for an estimated 14% of the total burden of disease [3]. While the majority of adolescents with mental health problems are cared for in the community, inpatient care is indicated particularly when there is a need for intensive interventions that cannot be provided in the community, such as when there is an increased risk of harm to self or others or when the young person presents with a deterioration in physical health as a result of mental health difficulties [4].

The models of care in acute adolescent inpatient units around the world are diverse [5], and the models are shaped by myriad factors such as mental health literacy, care pathways, stigma, extent of family engagement and integration, and availability of follow-up arrangements [6], staffing profiles, resource availability [7], differing thresholds for admission, . In Europe, the acute adolescent inpatient unit is part of a continuum-of-care model linked to outpatient treatment and community-based services [8]. The Child and Adolescent Mental Health Service (CAMHS) is delivered through a tiered model of service organisation with the Tier-4 being the highest tier of severity and complexity, involving highly specialized inpatient care funded by the government [9]. In 2017, a United Kingdom-based study examining the outcomes of inpatient psychiatric admissions for adolescents provided evidence to support the efficacy of a tier-4 service [10]. In the United States, the model of care has gradually evolved from institutional and fragmented service delivery toward more community-based and integrated models, although fragmentation has remained a persistent problem [11]. Although over half a million adolescents receive mental health services through inpatient psychiatric care [12,13], the evidence base evaluating the effectiveness of inpatient care for adolescents with mental health problems is limited.

In Australia, the vast geographic landmass creates differences in population distribution across urban, regional, remote, and very remote areas, requiring innovative service development. With the paucity of emergency, residential, day-patient and in-home services for adolescents and their families, there is growing pressure on inpatient units to admit patients for clinical reasons as well as for non- clinical reasons of respite and primary child protection issues [4]. ^3. Thus, inpatient care consumes the majority of child and adolescent mental health resources, and as societal mental health demands increase, many inpatient services have been encouraged to reduce costs and to justify the spending via evidence on outcomes and effectiveness [12,14] as well as expectations to increase efficiency and cater to more patients within their existing pool of funds [12]. Therefore, there is a need to examine valid and reliable outcome measures, which capture changes in symptoms and functioning, and which determine the effectiveness of inpatient psychiatric admissions for adolescents. These measures provide a means for evaluation of an inpatient service as a whole and highlight areas that warrant improvement [15]. Subsequently, all states and territories have implemented routine outcome measurements (ROM) across public mental health services. Core measures in the adolescent setting include the Health of Nations Outcome Scales for Children and Adolescents (HoNOSCA) [16] and the Children’s Global Assessment Scale (CGAS) [17].

In many high-income countries, healthcare systems use performance indicators or Key Performance Indicators (KPIs) to monitor, measure, and manage the performance of their healthcare systems to ensure effectiveness, efficiency, and equity [18]. The Donabedian conceptual model provides a framework for evaluating healthcare services and quality of care. KPIs are typically organised according to the Donabedian Framework, which includes the structure, process, and outcome domains [19]. Structure describes the context in which healthcare is delivered; process includes all transactions between patients and providers throughout the delivery of care, and outcome refers to the effect of healthcare on the health status of patients and the population [19]. In Australia, the National Mental Health Performance Framework (NMHPF) is a key strategy for facilitating a culture of continuous quality improvement in mental health service delivery [20]. The framework supports Australian state and territory governments’ commitment to improving accountability and transparency at the Mental Health Service Organisation (MHSO) level. The NMHPF follows a tiered, domain-based structure, with Tier 3 dedicated to health system performance, as monitored via KPIs. The KPIs reflect an inpatient unit’s efficacy, effectiveness and safety through metrics of the length of stay (LOS), 28-day readmission rate, rate of seclusion and restraint and 7-day post-discharge follow-up rate [21]. However, few studies have examined these KPIs to assess the effectiveness of MOC.

Taken together, there is little published literature about effective models of care in adolescent mental health inpatient units and to oour knowledge, there are few studies in Australia that compare different models of care using outcome measures and KPIs. The aim of this study was to evaluate the effectiveness of an innovative model of care, the Brief Intervention Model of Care (BIMOC), compared with the Traditional MOC (TMOC) in the acute adolescent inpatient unit. Comparisons were based on clinical outcome measures (HoNOSCA and CGAS) and several KPIs (e.g., length of stay (LOS), 28-day readmission rate, 7-day follow-up rate, and seclusion rate), which were tracked over the subsequent years (2022-2025) to evaluate BIMOC’s effectiveness and sustainability. HoNOSCA and CGAS were chosen because they cover the three critical dimensions of the concept of outcomes: clinical severity and disability through clinician rating. The findings are expected to contribute to efficiently improving mental-health care for young people.

2. Materials and Methods

2.1. Design

A quasi-experimental, non-randomised before–and–after study design was used. This cross-sectional study utilised an Electronic Medical Records (EMR)-based file audit of routinely collected data from consumers, their parents, and carers attending the BIMOC and TMOC Service. This study compares two models of care: TMOC (2016 &2017) and BIMOC (2018). The outcome-driven sustainability of BIMOC was examined using KPIs and clinical data collected from 2022 to 2025.

2.2. Site Descriptions

The inpatient unit Gna Ka Lun (GKL) (indigenous word meaning “healing of mind”) was based within the Campbelltown hospital premises of South Western Sydney. It covers a wide, densely populated geographical region, including both metropolitan and rural communities. GKL is a 10-bed acute adolescent unit established in the early 2000s, with an evidence-based care model that aligned with similar inpatient units. The standard evidence-based MOC referred to as TMOC includes a clinical assessment, evidence-based treatment endorsed by the Royal Australian and New Zealand College of Psychiatrists (RANZCP) guidelines, and a discharge planning phase that involves follow-up arrangements through the Infant, Child and Adolescent Mental Health Service (ICAMHS), the private sector, or the patient’s General Practitioner.

2.3. Overview of New Brief-Intervention Model of Care (BIMOC) and TMOC:

Traditional MOC represents a conventional, non-stratified acute inpatient psychiatric care model delivering standard assessment, treatment, and discharge planning without structured, time-limited or differentiated intervention pathways. The BIMOC was developed after analysing the pre-existing TMOC, the legislative requirement for service delivery, and the examination and inclusion of evidence-based practice. Within 24-48 hours of admission to the in-patient unit, a comprehensive diagnostic and psycho-social assessment of the adolescent and the family is conducted by a Multi-Disciplinary Team (MDT) comprising a Consultant Psychiatrist, a Psychiatry Registrar, a Clinical Psychologist, a Social Worker, and a Nurse. The MDT constructs a bio-psycho-social formulation and streamlines the adolescent as either a “crisis” (adolescents with severe affective dysregulation, presenting with abrupt onset suicidal/self-harm behaviour in the context of trauma or psycho-social stressors) or an “acute” admission (Adolescents with enduring mental health issues illness presenting with relapse or exacerbation, or an unsuccessful trial of community treatment) and a treatment process plan is drafted (Appendix A). This plan identifies the primary treating team, outlines the goals of treatment, the length of stay, the estimated date of discharge, and the schedule for family meetings and individual sessions with the admitted patient. A copy of this document is given to the adolescent, family and the community team.

A brief intervention is offered to those identified as crisis admissions. This is a time–limited, goal-oriented, patient- and family-centred, culturally sensitive intervention. The components in this model (Appendix B) include nominating a ‘therapeutic lead’, agreeing upon targeted family assessments, brief interventions based on trauma-informed care, collaborating with the adolescent to develop self-harm prevention strategies, community CAMHS integration and timely, well-communicated discharge planning [22]. In addition, skills training is provided for the adolescent and the family, while the discharge planning is developed in collaboration with the adolescent, the family and the community team. This approach aims to break down barriers to adolescents engaging with the service, whilst reducing the rate of seclusion and minimising the ‘revolving-door’ phenomenon of repeated hospital presentations. The adolescents categorised as ‘acute’ continued to receive care as espoused by the TMOC.

2.4. Participants:

All adolescents aged 10-18 years admitted to the GKL unit between January 2016 and December 2018 and between January 2022 and December 2025 were included in this study (2016 and 2017 – TMOC; 2018 – BIMOC, which continued through from 2022-2025).

2.5. Data Sources:

Data for this study were collected retrospectively across the study period (2016–2018 and 2022–2025) using EMR systems, primarily the EMR Discern Analytics reporting tool and the NSW Health reporting platform, Power Business Intelligence (BI) (Mental Health Inform). Power BI was used to extract selected service-level indicators. All other variables were sourced from EMR Discern Analytics.

2.6. Measures:

2.6.1. Socio-Demographic and KPI Measures

Socio-demographic variables included gender (male, female, others), age (10-12 years, 13-18 years), and ethnicity (Aboriginal, Caucasian, and Culturally and Linguistically Diverse [CALD]). KPIs included length of stay, 28-day re-admission, 7-day follow-up completed, seclusion, suicide/self-harm incidents and total admissions. The choice of these measures and KPIs was based on the extensive review of existing studies [12,23,24,25,26] and clinical relevance. Service-level KPIs as described above were compared across models.

2.6.2. Outcome Measures

The outcome measures were completed on admission and discharge using the routinely collected measures within CAMHS services and were recorded in EMR systems. Measures were completed by a clinician who had previously received training in administering these instruments.

2.6.2.1. Health of the Nation Outcome Scale-Child and Adolescents (HoNOSCA):

The HoNOSCA is a clinician-rated instrument for people under 18 years of age, comprising of 15 items covering clinical and psychosocial functioning including behaviour, clinical symptoms, social problems, information problems, and impairment [16]. This was used in this study as the primary measure of psychopathology.

2.6.2.2. Clinical Global Assessment Scale (CGAS):

The CGAS is a rating of functioning for children and young people aged 6-17 years [17]. The child is given a single score between 1 and 100, based on a clinician’s assessment of various aspects of the child’s psychological and social functioning. Scores fall into 10 categories ranging from’ extremely impaired’ (1-10) to ‘doing very well’ (91-100). This measure was used as the primary measure of functioning.

2.7. Data Analysis

The data was analysed in three stages. First, descriptive statistics were used to summarise sample demographics and assess comparability between TMOC and BIMOC. Second, the KPIs were examined across these MOC using descriptive statistics. Finally, clinical outcome measures (HoNOSCA and CGAS) were analysed using paired-samples t-tests to evaluate pre–post changes within each model. Effect sizes (Cohen’s d) were calculated to assess the magnitude of change from admission to discharge. Between-model comparisons of effect sizes were conducted using z-tests to examine differences in the magnitude of improvements between TMOC and BIMOC.

2.8. Ethics

This project received ethics approval from the Human Research Ethics Committee (HREC) of the South Western Sydney Local Health District (Reference Number: 2018/ETH00719). The HREC granted a waiver of informed consent from parents/carers and/or assent from young people since this project involved no more than low risk and analysed de-identified routinely collected EMR data obtained during standard clinical care. The current study was conducted in accordance with the Declaration of Helsinki [27].

3. Results

3.1. Demographic Characteristics of Participants

Table 1 presents the demographic profile of admitted young people in the TMOC and BIMOC cohorts over 2016–2018. Females consistently accounted for the majority of admissions (approximately 67%–73%), with a smaller proportion of males (approximately 27%–33%) and minimal representation from other gender categories. Most adolescents were aged 13–18 years (over 92% in all years), indicating that inpatient admissions were predominantly among adolescents. In terms of ethnicity, Caucasian adolescents constituted the largest group across all years, although there was some variation, with a lower proportion observed in 2017 (68.8%) compared to 2016 and 2018 (~80%). The proportion of Aboriginal adolescents remained relatively stable (approximately 7%–8%). Notably, total admissions increased from 114 to 154 under TMOC to 184 under BIMOC, suggesting higher service throughput without substantial demographic shifts.

3.2. Sustained Improvements in Selected KPIs Following BIMOC Implementation

Table 2 shows the comparison of KPIs across the TMOC (2016–2017) and the BIMOC (2018 onwards), demonstrating notable improvements following the implementation of the BIMOC. In 2018, the initial year of BIMOC, there were marked reductions in average length of stay (13.6 days vs 21.0 days in 2016), 28-day readmission rates (8.1% vs 11.9–17.0%), and seclusion events (4.6 vs 13.8–20.1 per 1,000 occupied bed days), alongside a substantial increase in 7-day follow-up completion (76.8% vs ~60%). These improvements occurred alongside an increase in service capacity (admissions), with total admissions rising from 114–154 under TMOC to 184 in 2018. Suicide and self-harm incidents also decreased initially (101 in 2018 compared to 141 in 2017) and continued to decline over time, reaching 12 incidents by 2025, although these figures represent combined inpatient and community events. In subsequent years (2022–2025), improvements were largely sustained, particularly in follow-up rates (68.9%–90.0%) and seclusion events (2–7 per 1,000 bed-days), while total admissions remained consistently higher than TMOC levels despite some fluctuation (123–138 admissions per year). Although variability was observed in length of stay and readmission rates over time, these outcomes were generally comparable to, or at times better than, TMOC benchmarks. Overall, the findings suggest that the BIMOC was associated with optimised service efficiency (increased throughput), reduced restrictive practices, decreasing trends in suicide/self-harm events, and enhanced continuity of care, with several gains maintained longitudinally.

3.3. Changes in Clinical Outcomes and Comparative Effectiveness of Models of Care

Across both the TMOC and BIMOC, statistically significant improvements were observed in clinical outcomes from admission to discharge. Under the TMOC, CGAS scores increased and HoNOSCA scores decreased significantly, indicating improvements in functioning and symptom severity. Similarly, under the BIMOC, substantial improvements were observed across both acute and crisis subgroups, with significant increases in CGAS scores and reductions in HoNOSCA scores. Effect size analysis demonstrated a greater magnitude of improvement in functioning under the BIMOC (d = 1.89) compared with the TMOC (d = 1.22). This difference was statistically significant (z = -3.694, p = 0.0002), indicating superior functional outcomes under BIMOC. In contrast, improvements in HoNOSCA scores were comparable across models (z = -0.769, p = 0.4418), suggesting no significant difference in symptom improvement.

Table 3. Changes in clinical outcomes (CGAS and HoNOSCA) from admission to discharge across TMOC and BIMOC: within- and between-model comparisons.

MOC (Year)	CGAS Admission (SD)	CGAS Discharge (SD)	t-test	Effect size (d, 95% CI)	HoNOSCA Admission (SD)	HoNOSCA Discharge (SD)	t-test	Effect size (d, 95% CI)
TMOC 2016	37.14 (10.49)	57.60 (11.73)			19.00 (7.65)	9.24 (5.34)
TMOC 2017	31.67 (13.15)	52.39 (14.09)			17.99 (5.55)	11.58 (4.04)
TMOC Total	34.16 (12.29)	54.87 (13.23)	t(133) = −14.18, p < .001	1.22 [1.00–1.45]	18.45 (6.58)	10.57 (4.78)	t(146) = 15.94, p < .001	1.31 [1.09–1.54]
BIMOC 2018 (Acute)	23.44 (9.91)	48.67 (9.06)			19.82 (6.66)	10.61 (6.24)
BIMOC 2018 (Crisis)	30.27 (10.55)	51.54 (8.89)			19.11 (6.72)	9.58 (5.18)
BIMOC 2018 Total	27.67 (10.79)	49.76 (9.01)	t(144) = −22.74, p < .001	1.89 [1.62–2.16]	19.30 (6.67)	9.96 (5.62)	t(151) = 17.74, p < .001	1.44 [1.21–1.67]

Notes: Demographic data from 2022-2025 is not available in Power BI. Effect sizes (Cohen’s d) represent magnitude of change from admission to discharge. Between-model comparisons of effect sizes were conducted using z-tests: CGAS (z = −3.694, p = 0.0002); HoNOSCA (z = −0.769, p = 0.4418). SD: Standard Deviation.

Table 3. Changes in clinical outcomes (CGAS and HoNOSCA) from admission to discharge across TMOC and BIMOC: within- and between-model comparisons.

MOC (Year)	CGAS Admission (SD)	CGAS Discharge (SD)	t-test	Effect size (d, 95% CI)	HoNOSCA Admission (SD)	HoNOSCA Discharge (SD)	t-test	Effect size (d, 95% CI)
TMOC 2016	37.14 (10.49)	57.60 (11.73)			19.00 (7.65)	9.24 (5.34)
TMOC 2017	31.67 (13.15)	52.39 (14.09)			17.99 (5.55)	11.58 (4.04)
TMOC Total	34.16 (12.29)	54.87 (13.23)	t(133) = −14.18, p < .001	1.22 [1.00–1.45]	18.45 (6.58)	10.57 (4.78)	t(146) = 15.94, p < .001	1.31 [1.09–1.54]
BIMOC 2018 (Acute)	23.44 (9.91)	48.67 (9.06)			19.82 (6.66)	10.61 (6.24)
BIMOC 2018 (Crisis)	30.27 (10.55)	51.54 (8.89)			19.11 (6.72)	9.58 (5.18)
BIMOC 2018 Total	27.67 (10.79)	49.76 (9.01)	t(144) = −22.74, p < .001	1.89 [1.62–2.16]	19.30 (6.67)	9.96 (5.62)	t(151) = 17.74, p < .001	1.44 [1.21–1.67]

Notes: Demographic data from 2022-2025 is not available in Power BI. Effect sizes (Cohen’s d) represent magnitude of change from admission to discharge. Between-model comparisons of effect sizes were conducted using z-tests: CGAS (z = −3.694, p = 0.0002); HoNOSCA (z = −0.769, p = 0.4418). SD: Standard Deviation.

4. Discussion

This study addresses a key gap in the peer-reviewed literature by comparing different inpatient MOC using both clinical outcome measures and service-level KPIs within a single service setting in South Western Sydney. We found that implementation of the BIMOC maintained clinical effectiveness while demonstrating greater improvements in functional outcomes, alongside improved service efficiency and safety, as reflected in reduced LOS, reduced critical incidents and readmission rates in the early phase, alongside increased service input without compromising clinical outcomes. Our findings also indicated that several of these improvements were sustained over time, particularly higher follow-up rates, reduced seclusion/self-harm incidents, and consistently increased admission capacity, supporting the model’s long-term viability. The BIMOC was also associated with a reduction in restrictive practices (e.g., seclusion events), indicating improved ward safety and reduced reliance on coercive interventions. In addition, a downward trend in suicide and self-harm incidents was observed over time, suggesting potential improvements in safety, risk management and continuity of care. Our findings contribute to the emerging evidence base on the effectiveness and sustainability of inpatient MOC for adolescents globally and within the Australian context.

The current study indicated that the introduction of the BIMOC maintained clinical effectiveness, while demonstrating greater improvement in functional outcomes compared with the TMOC, while symptom improvements were comparable across models, evidenced by initial reductions in LOS and readmission rates, along with increased service input (admissions), all achieved without compromising clinical outcomes. These findings are particularly important within acute adolescent mental health inpatient units in Australia, which play a critical role in managing high-risk presentations [28] and face ongoing pressures related to limited bed capacity and increasing demand [29]. The observed efficiency gains suggest that structured, time-limited MOC may facilitate timely access to inpatient services while maintaining clinical safety and effectiveness in crisis settings [30,31]. Our findings are consistent with previous literature, indicating that similar models, such as the Rapid Stabilisation Pathway [32], have reported shorter lengths of stay without increases in readmission rates, enabling improved access to limited inpatient resources [33]. However, the broader literature also highlights variability in outcomes, with some studies reporting persistent challenges in reducing readmission rates due to clinical complexity and post-discharge factors [34]. The comparatively favourable outcomes observed in the current study may reflect the integrated and responsive design of the BIMOC, including enhanced follow-up and continuity of care, which are known to influence service utilisation and relapse risk. Overall, these findings contribute to the growing literature supporting innovative MOC that can address both clinical and operational demands in acute adolescent mental health services [35].

Our findings also indicated that several of these improvements were sustained over time, particularly higher follow-up rates, reduced seclusion and self-harm incidents, and consistently increased admission capacity, supporting the model’s long-term viability. These findings are broadly consistent with existing literature, which demonstrates that system-level and organisational interventions can produce sustained reductions in restrictive practices, particularly when models prioritise relational, patient-centred care [36]. Moreover, the sustained increase in admission capacity observed in our study aligns with evidence from crisis stabilisation and brief inpatient models, which have shown that efficient, structured care pathways can improve patient flow and optimise use of limited inpatient beds without compromising outcomes [32,36]. However, the existing literature also demonstrates variability in the sustainability of such improvements, with literature reporting challenges in maintaining gains over time due to workforce constraints, service fragmentation, and increasing clinical complexity among admitted adolescents [13]. The sustained benefits observed in the current study may, therefore, reflect the structured design and embedded continuity mechanisms within the BIMOC, including enhanced follow-up processes and consistent service integration, which may mitigate some of these challenges.

The current study’s findings suggest that the BIMOC may have supported a shift toward less coercive practices (e.g., seclusion events), more patient-centred care (less self-harm incidents), consistent with contemporary trauma-informed care principles in adolescent mental health services [36]. These findings align with existing literature demonstrating that system-level and organisational interventions can significantly reduce the use of restrictive practices such as physical restraints and seclusions [37]. Similarly, a systematic review by Kelly et al. [36] found that trauma-informed interventions and service redesign strategies were consistently associated with reductions in seclusion and restraint in youth mental health settings. These studies support the view that changes in care models, particularly those focusing on relational care, staff training, and organisational culture, can effectively reduce reliance on coercive interventions. Overall, the present findings contribute to the growing evidence that innovative models of care can reduce restrictive practices while maintaining patient safety in high-risk adolescent populations.

Additionally, the present study identified the observed downward trend in suicide and self-harm incidents over time, suggesting potential improvements in risk management and continuity of care. This is clinically and operationally significant in acute child and adolescent mental health inpatient settings. Clinically, this may reflect enhanced engagement, more effective safety planning, and enhanced therapeutic alliance [38], while operationally it indicates a more stable ward environment with fewer behavioural escalations requiring intensive intervention [39]. These findings are consistent with emerging evidence demonstrating that structured, brief, and continuity-focused interventions can reduce suicidal behaviours and rehospitalisation risk. For example, Goldstein and colleagues [40] reported that a brief post-discharge safety intervention was associated with lower rates of rehospitalisation and suicidal behaviour among adolescents, highlighting the importance of continuity of care. In contrast, other studies have suggested that self-harm and suicide-related behaviours remain persistent challenges in inpatient settings, particularly in populations with complex comorbidities and limited access to follow-up care [13]. The divergence between these findings and the current study may be explained by differences in MOC structure. Overall, the sustained reduction in suicide and self-harm incidents observed in this study provides important evidence that structured, coordinated models of care can improve both immediate safety outcomes and longer-term clinical stability in high-risk adolescent populations.

4.1. Implications for Policy and Clinical Practice

The current study’s findings highlight the importance of developing effective and sustainable inpatient MOC for young people. The BIMOC demonstrated maintained clinical effectiveness alongside greater improvements in functional outcomes, while achieving sustained improvements in key service-level indicators, including reduced critical incidents, optimised follow-up, and increased service throughput over time. These results highlight the value of models that integrate clinical effectiveness with operational efficiency and sustainability. From a policy perspective, the findings support the need for adolescent-focused, trauma-informed, and developmentally appropriate care frameworks, as well as stronger integration among inpatient, community, child protection, and educational services to ensure continuity of care. Improved follow-up rates and reduced restrictive practices reinforce the importance of coordinated care pathways and transition planning.

At a systems level, increased (safe) service capacity without additional financial investment in a setting under constant pressure suggests that service redesign can improve efficiency while maintaining equity and access, which is critical in high-demand settings. Clinically, structured, goal-directed, and family-inclusive approaches may enhance engagement and safety, with potential to reduce self-harm risk. Future policy should prioritise workforce development, multidisciplinary collaboration, and routine outcome monitoring to sustain improvements and support the delivery of high-quality care for young people requiring inpatient mental health services.

4.2. Strengths, Limitations and Directions for Future Research

A key strength of this study is the use of clinician-rated outcome measures (HoNOSCA and CGAS) alongside service-level KPIs over multiple years, which allowed to evaluate clinical and operational outcomes. To our knowledge, no study has been conducted within adolescent inpatient setting in Australia, comparing the effectiveness of MOC using KPIs and psychometric measures. However, the findings should be interpreted in light of several limitations. First, demographic data for the period 2022–2025 and KPI data for 28-day readmissions were incomplete. In addition, the study only used clinical outcome data (HoNOSCA and CGAS) and these were available only for 2016–2018 due to resource constraints and limitations in data access. The incomplete availability of some long-term data reduces the ability to assess all outcomes consistently across the full study period.

These limitations were primarily driven by a statewide transition in reporting platforms, including changes associated with the relocation and not using the same data system for the GKL inpatient unit in 2022, which restricted the retrieval of historical data. Consequently, complete datasets for the entire study period could not be accessed retrospectively, despite efforts by the Mental Health Information Management team to retrieve them. Additionally, CGAS data were not available through the Discern Analytics system and would have required manual extraction from clinical records, which was not feasible within available resources. Importantly, the quasi-experimental before-and-after design limits causal inference. Future research should aim to incorporate a broader range of psychometric measures and ensure more comprehensive and consistent data capture over extended time periods, particularly by addressing gaps in demographic, clinical, and KPI data. Efforts to improve data accessibility and integration across evolving reporting systems, alongside designs that strengthen causal inference, would enhance the robustness of evaluations of service effectiveness and sustainability.

5. Conclusions

This study demonstrates that implementing a redesigned, structured model of care (BIMOC) in an acute adolescent mental health inpatient unit can maintain clinical effectiveness, with evidence of greater improvement in functional outcomes compared with the TMOC while improving service efficiency and performance outcomes. The BIMOC was associated with initial reductions in length of stay and readmission rates, alongside increased service input, indicating enhanced system efficiency without compromising clinical outcomes. Importantly, several improvements were sustained over time, including higher follow-up rates, reduced use of restrictive practices (particularly seclusion), and consistently increased admission capacity, supporting the model’s long-term viability. The observed downward trend in suicide and self-harm incidents further suggests potential improvements in safety, risk management and continuity of care within this high-risk population. In addition, comparative analysis of clinical outcomes indicated that while both models were effective, the BIMOC demonstrated greater gains in functional outcomes, whereas improvements in symptom severity were comparable between models. These findings highlight the value of structured, goal-directed, and integrated models of care, underpinned by multidisciplinary collaboration and coordinated service delivery, in addressing both clinical and operational demands in adolescent inpatient mental health settings. Overall, this study contributes to the emerging evidence base on the effectiveness and sustainability of innovative inpatient models of care for adolescents, particularly within the Australian context, and supports ongoing efforts to optimise service delivery in high-demand mental health systems.