The PRISMA 2020 flow diagram, describing the search strategy and selection schema, is displayed in Fig. 1. The search returned 76 articles for potential inclusion in the review (PubMed [n = 17], Embase [n = 19], and Scopus [n = 40]). Of these, 47 (62%) unique articles remained following removal of duplicates (38% [n = 29]). After title and abstract screening, 30 (64%) articles were excluded for being literature reviews (4% [n = 2]), not investigating ED transitions (57% [n = 27]), or not being peer reviewed (2% [n = 1]). Accordingly, 17 (36%) articles were assessed for eligibility, with 4 (24%) deemed non-eligible for inclusion. One additional article was identified via reference checking. There were no inaccessible full texts. Between independent reviewers who screened articles, there was almost perfect agreement (Cohen’s k = 0.94).

The final review comprised 14 articles, with a total of 747 (53.36 [mean] ± 92.58 [std dev.]) participants. The first study was published in 2008 (n = 1), the largest number was published between 2020 and 2021 (n = 6), and the most recent was published in 2023 (n = 3). Regarding location, most studies were conducted in England (36% [n = 5]) and Canada (36% [n = 5]), followed by Norway (21% [n = 3]) and France (7% [n = 1]). At the time of data abstraction, articles were grouped into ‘qualitative’, ‘cross-sectional’, and ‘longitudinal cohort’ studies by research design. Ten studies conducted focus groups and/or qualitative interviews (n = 153 [15.30 ± 8.12]), two studies distributed prospective surveys at one time point (n = 66 [33.00 ± 7.07]), and two studies analyzed data retrospectively (n = 528 [264.00 ± 82.02]). Concerning participants, five studies recruited patients (n = 559 [111.80 ± 144.90]), one study recruited caregivers (n = 12 [12.00 ± 0.00]), four studies recruited providers (n = 87 [21.75 ± 12.50]), and four studies recruited more than one type (n = 89 [22.25 ± 10.28]).

See Table 1 for a summary of the qualitative studies included in this review, which are further described below. Figure 2 displays the quality of evidence, specifically the extent to which each study addressed the possibility of bias in its design, conduct, and analysis. Overall, qualitative studies were of moderate-to-high quality, with almost perfect agreement between independent reviewers who performed critical appraisals (Cohen’s k = 0.92).

Quality assessment of qualitative studies

Dimitropoulos et al. [17] were the first to conduct qualitative research on the perspectives of providers, covering the transition of young people from CAMHS to AMHS for AN. Two in-depth, semi-structured focus groups (2 h) were conducted, one at a pediatric ED program (n = 8) and one at an adult ED program (n = 10). Providers had direct experience with transitioning patients in the last 12 months. In addition, five in-depth, semi-structured qualitative interviews (1 h) were conducted in tertiary and community care settings with providers treating patients who had exited pediatric ED programs, yet who had not accessed adult ED services. Using grounded theory methodology, an approach for generating theory based on systematic data [37], three inter-related themes were identified: (1) AN-related factors, namely denial of illness and ambivalence towards recovery; (2) developmental interruptions in adolescence, such as failure to cultivate autonomy and social connections; and (3) reduced parental involvement in adult ED services, with change in authority and legal power. There was congruency within and across focus groups and qualitative interviews.

In a follow-up study [38] that recruited providers from the same ED programs, investigators sought to identify factors that impact effective transitions for AN. Similarly, two in-depth, semi-structured focus groups (2 h) were conducted, one at a pediatric ED program (n = 8) and one at an adult ED program (n = 10), along with five in-depth, semi-structured qualitative interviews (1 h). The latter involved professionals outside of these programs (n = 5). Their analysis revealed two major themes. First, transitions to AMHS should be determined by patient and family readiness, not by age, as emphasized in the pediatric ED focus group and qualitative interviews. However, this was not identified as problematic in the adult ED focus group. Second, transitions can be improved by family interventions and psychoeducation, patient interventions and self-management skills, and coordinated medical care before or immediately after transitioning from CAMHS. These suggestions were shared in both focus groups and qualitative interviews. Further, all providers agreed that interventions should foster greater patient autonomy and independence.

Findings from both studies motivated a third [39], regarding the experiences of young adults (aged 17–21) with EDs who transitioned from CAMHS to AMHS. Patients were eligible if they met diagnostic criteria for AN or BN and received treatment in a tertiary pediatric ED program within the past two years. In-depth, structured qualitative interviews (1 h) were conducted across two sites (n = 15), which were geared towards patients’ retrospective experiences. Three key themes emerged: (1) difficulty navigating care during the transition period, largely due to inconsistent procedures and minimal provider discussions; (2) challenge achieving and maintaining recovery post-transition as a result of systematic barriers, such as accessing care and trained providers, and the intensity of available programs interfering with higher education; and (3) transitions can be improved by increased coordination, communication, and collaboration between services and providers.

As part of a larger, three-study qualitative project, Lockertsen et al. [40] explored how providers experience the transition process for AN. Using snowball sampling, providers (n = 8) were recruited from in- and out-patient facilities who had experience treating and/or coordinating treatment for AN patients, plus transitioning them across care. One dialectic, multi-step focus group (1.5 h) and two in-depth, semi-structured qualitative interviews (1.5 h) were conducted. Leveraging systematic text condensation, a method for analyzing cross-case qualitative data [41], their analysis revealed four primary barriers to transitioning. This included (1) differences in treatment cultures between services, with respect to family involvement and patient responsibility; (2) a lack of mutual understanding between services, regarding administrative systems and treatment ideologies and approaches; (3) feelings of ineptitude and insecurity and poor self-confidence among providers, negatively impacting patients and their transition experience; and (4) a lack of trust between services—across systems, providers, and patients—and focus on building transfer alliance. Additionally, providers expressed that parents and patients were not adequately prepared for the transition process.

These results extended to another study [42] that examined patients’ lived experiences transitioning from CAMHS to AMHS. Ten patients with AN, previously treated at in-patient units (n = 6) or specialized ED programs (n = 3), were recruited through providers and ED support groups. They partook in one dialectic, multi-step focus group (1–1.5 h) and five in-depth, semi-structured qualitative interviews (1–1.5 h). Four inter-related themes were identified: (1) a lack of preparedness for transitioning and related loneliness and stress; (2) time needed to establish relationships with providers, deemed essential for developing trust, mutual understanding, and motivation for treatment; (3) poor acknowledgement of patient individuality and self-sufficiency, with transitions based on age versus developmental stage and readiness; and (4) dependency on systems and provider impressions, causing feelings of powerlessness and hopelessness.

As the final study in this project, Lockertsen et al. [43] assessed how parents experience the transition process for adolescents with AN. Through snowball sampling, they recruited mothers (n = 9) and fathers (n = 3) via therapist referrals and study advertisements. All parents had a child diagnosed with AN who transitioned to AMHS at 18 years of age. In-depth, semi-structured qualitative interviews were conducted (1–1.5 h) in settings chosen by parents. Using systematic text condensation, their analysis revealed six themes: (1) sudden discharge from CAMHS that was determined by age, rather than by process; (2) lack of continuity between services, followed by mental health deterioration during waiting periods; (3) poor involvement and access to information, resulting in psychological distress; (4) overwhelming and complex responsibility, making the timing of transitions challenging; (5) provider incompetency and ignorance of both patients’ and parents’ ED knowledge and understanding, contributing to many unnecessary transitions; and (6) a lack of professional support from mental health services.

Taking a prospective approach, one study [44] investigated impending transitions for adolescents (n = 5) and their caregivers (n = 5). Patients were eligible if they met diagnostic criteria for an ED, as determined by a psychologist or physician; were actively being treated in CAMHS, specifically in a tertiary ED program; and were aged 17–18, waiting to undergo the transition to AMHS. In-depth, semi-structured qualitative interviews (0.5–1 h) were conducted. Using inductive reasoning, analyses showed that stakeholders largely had a limited understanding of the transition process. This applied more so to caregivers than adolescents. Several barriers to transitioning were described, including late timing of discussions, a lack of professional support during waiting periods, and re-explaining information to adult providers. Further, adolescents and caregivers expressed that successful transitions could be achieved through greater parental and provider involvement, as well as by implementing a ‘transition coordinator’. This would ideally be accompanied by a ‘transition passport’, giving adolescents instant access to their medical information, like their prescriptions, that could be shared with AMHS providers.

Wales et al. [45] explored the experience of transitioning, aiming to identify barriers and facilitators to this process. Providers were recruited via email from CAMHS and AMHS community ED teams, and asked to reflect on their past experiences. Four in-depth, semi-structured focus groups (1 h) were conducted, two in CAMHS (n = 10) and two in AMHS (n = 12). Additionally, ED patients (n = 29) who were eligible for transitions in the past two years, whether they transitioned or not, along with caregivers (n = 28), were recruited via letters from local CAMHS and AMHS. These individuals partook in in-depth, semi-structured qualitative interviews (50 min), either in person or over telephone. Thematic analysis revealed three core themes that served as barriers or facilitators to transitioning: (1) communication between CAMHS and AMHS internally and with patients and caregivers externally; (2) operational differences between CAMHS and AMHS, as well as uncertainty about transitions to AMHS; and (3) timing of transitions that are determined by rigid age boundaries and coincide with important life events.

Taking a resource perspective, one study [46] examined the perceived value of educational supports in facilitating effective transitions across ED care. Patients diagnosed with AN (n = 6), who transitioned to AMHS for at least one year, were recruited by administrative staff from a tertiary adolescent medicine program, supported by a dedicated ED team. In-depth, semi-structured qualitative interviews (30 min) were conducted via telephone. Thematic analysis identified three primary themes, namely unique challenges for ED patients (i.e., late age of ED onset and sudden discharge, fluctuating symptoms, and transition versus discharge), issues in adult care (i.e., greater autonomy and new psychiatric comorbidities), and the value and content of educational resources (i.e., as symptom benchmarks and provider connection tools). Specifically, patients thought educational supports would aid successful transitions into adult care, which should be distinguished from discharges, describe administrative changes, and set expectations for new roles and responsibilities.

Finally, Scanferla et al. [47] captured transition experiences shared by ED patients and their immediate family members. Recruited from an inpatient ED unit, 12 patients with AN and six related caregivers were included in the study. In-depth, semi-structured qualitative interviews were conducted in person for patients and over telephone for caregivers. Leveraging interpretative phenomenological analysis, an inductive approach that details how individuals make sense of their personal and social world [48], investigators revealed four main themes: (1) the transition experience in and of itself, comprising new caregiver roles, treatment modalities, and administrative cultures; (2) associated emotions, such as fear and abandonment (patients) and destabilization and helplessness (caregivers); (3) challenges of transitioning, like delayed access to services and lack of provider support; and (4) facilitators to improve the transition process, including smooth transfers, support of personal life goals, involvement of caregivers, and being accompanied and welcomed into adult care.

See Table 2 for a summary of the cross-sectional studies included in this review, which are further described below. Figure 3 displays the quality of evidence, specifically the extent to which each study addressed the possibility of bias in its design, conduct, and analysis. Overall, cross-sectional studies were of poor-to-moderate quality, with perfect agreement between independent reviewers who performed critical appraisals (Cohen’s k = 1.00).

Quality assessment of cross-sectional studies

As an extension of their qualitative findings, Wales et al. [49] sought to determine which of the factors that influenced ED transitions were most important to medical stakeholders. Patients (n = 12), caregivers (n = 8), and providers (n = 8) were recruited from ED charities, support groups, and professional bodies. Eligible patients had been offered transitions from CAMHS to AMHS for an ED, whether they transitioned or not, and eligible providers had experience transitioning patients to AMHS, whether for an ED or an unrelated condition. Leveraging Q-methodology, a ‘qualiquantological’ approach that focuses on individual viewpoints and subsequently identifies shared ones [50], participants completed a Q-sort rank task, evaluating 40 qualitative statements based on their agreement. Principal component factor analysis identified four items, explaining 52% of the variance that highlighted: parent and caregiver inclusion, patient support during the transition, timely and patient-centered care, and effective transitions between services.

Most recently, a study [51] conducted in England assessed provider compliance with the National Health Services’ (NHS) recommendations for managing ED transitions for children and young people. During a series of national training days hosted by the Royal College of Psychiatrists, a survey was distributed to 70 teams that provided ED treatment. Of the 38 that participated, 97.4% (n = 37) had fixed transition boundaries, commonly set at 18 years of age. Moreover, 73.7% (n = 28) of teams reported that some young people were admitted to specialized or community-based programs for EDs or mental health issues, respectively, whereas 15.8% (n = 6) always transitioned to tertiary ED care. Further, 53.3% (n = 21) stated that a subset of ED patients did not meet clinical thresholds for AMHS, thereby failing to transition. Most teams complied with NHS recommendations, with 52.6% (n = 20) using ED-specific transition protocols, 78.9% (n = 30) creating individualized transition plans, 81.6% (n = 31) allowing flexible transition times, 89.5% (n = 34) jointly working with AMHS, and 73.7% (n = 28) providing transition support to caregivers and families. Regarding treatment, 97.4% (n = 37) of teams reported asymmetry between their therapeutic model and the services they transitioned care to, with more family-based approaches in CAMHS. Relatedly, 71.1% (n = 27) reported that providers discussed treatment differences with patients during the transition process, though 15.8% (n = 6) did not respond to this question.

See Table 3 for a summary of the longitudinal cohort studies included in this review, which are further described below. Figure 4 displays the quality of evidence, specifically the extent to which each study addressed the possibility of bias in its design, conduct, and analysis. Overall, longitudinal cohort studies were of moderate-to-high quality, with almost perfect agreement between independent reviewers who performed critical appraisals (Cohen’s k = 0.88).

Quality assessment of longitudinal cohort studies

Arcelus et al. [52] were the first to compare new ED patients in AMHS to those with prior contact in CAMHS, examining differences in illness severity and complexity. Referrals to AMHS were made from local primary care and secondary psychiatric services. Through a medical chart review, 206 young people (aged 16–25) were identified over a four-year period (2002–2005). Results showed that 27.7% (n = 57)  were previously treated in CAMHS for an ED, either as in-patients (57.9% [n = 33]) or out-patients (42.1% [n = 24]). Surprisingly, 43.9% (n = 25) were referred to AMHS by general practitioners versus CAMHS. Further, patients with prior CAMHS contact, particularly as in-patients, had lower self-esteem and higher maturity fears than those without; the latter referring to a desire to return to childhood, or a fear of adulthood and its many demands [53].

To further characterize service utilization, another study [54] analyzed a consecutive cohort of patients (aged 13–17) treated in CAMHS for an ED over a five-year period (2009–2014). Data regarding their use of AMHS for ED care as young adults (aged 18–25), within a follow-up period (2013–2017), were extracted from local and national hospital records. A total of 322 patients were identified and eligible for inclusion. Of these patients, 67.0% (n = 216) made no contact with AMHS, 13.0% (n = 42) received brief ED treatment, 10.0% (n = 32) received extended ED treatment, and 10.0% (n = 32) received non-ED treatment. Moreover, 10.8% of patients were directly referred to AMHS, whereas 7.6% were re-referred by CAMHS back to their general practitioner following discharge. Older age at presentation to CAMHS predicted increased ED care in young adulthood, as did the amount of CAMHS service use (i.e., the number and length of out-patient, day-patient, and/or in-patient episodes). No/brief use of AMHS was associated with the longest duration of untreated EDs.