ABSTRACT

Background

Despite the burden associated with bipolar disorder (BD), research into early diagnosis and treatment of BD lags approximately 20 years behind the field of early intervention for psychosis. This study evolved through a partnership between Orygen (Melbourne, Australia) and the Daymark Foundation (Toronto, Canada). The primary focus was to answer the question: “How might we advance an early intervention approach for people at‐risk of or with BD?”.

Methods

Semi‐structured interviews were conducted with international experts and other stakeholders in early intervention and BD to identify challenges and barriers in early intervention approaches to BD.

Results

Twenty‐eight experts participated. Nine themes emerged as challenges: (i) limited recognition and understanding of BD across stakeholders; (ii) lack of definitions; (iii) poor resourcing and lack of prioritisation in funding models; (iv) absence of validated tools for diagnosis, monitoring treatment response and/or disorder progression; (v) absence of “big data”; (vi) scarcity of evidence‐based treatments and clinical guidelines for the early stages of the disorder; (vii) obscurity around optimal service models; (viii) the need for better support and involvement of families and significant others; and (ix) the need for collaboration (across disciplines, stakeholders, and settings) to progress the field.

Limitations

Of those approached, 54.9% participated in the study. Given the qualitative nature of the study, recruiting more experts to the study would not necessarily change the outcomes as data saturation was achieved.

Conclusions

This work lays the foundations for developing a collaborative research framework to progress early intervention for young people with BD.

1 Introduction

Bipolar disorder (BD) is an enduring and severe mental health disorder that commonly has its onset before the age of 25 years [1, 2]. Over the disorder's course, individuals can experience depressive and subsyndromal affective symptoms for ~50% of their lives [3]. Despite recovery, functional impairments persist in social [4], occupational [5, 6, 7], psychological and environmental domains [8]. Early mortality can be due to suicide (rates as high as 20%) [9] and preventable physical comorbidities (e.g., cardiovascular disease) [10, 11]. BD is associated with marked burden particularly for the 15–24 years age group [12].

The pathophysiology of the BD is largely unknown [13, 14], making it difficult to identify those at risk of developing the disorder [14]. Diagnostic imprecision is common and there can be years between symptom onset and diagnosis [1, 15]. During this period, individuals may receive inappropriate, ineffective, and/or iatrogenic treatments [16].

Treatment innovation is lagging for BD [17]. Lithium, the mainstay treatment for BD, was discovered in 1949 [18] and most other pharmacotherapies were developed for other disorders (e.g., antipsychotics). Because of treatment ineffectiveness polypharmacy is the norm [19] and can be associated with marked side‐effects [20]. Clinical guidelines highlight the benefits of psychotherapy [21]; however, these therapies are not easily accessed, clinicians lack experience in their delivery, and they are costly [22, 23]. Some data suggest if effective and less aggressive treatments are provided in the early stages of the disorder, then the long‐term burden might be lessened [24, 25, 26]. Research is scant, however, on interventions for the early stages of BD [24, 25, 26, 27].

The potential burden of BD underpins the urgent need for prevention, early identification, and treatment access for BD [28]. Despite this need, research on early intervention for BD lags decades behind the field of early intervention for psychosis (EIP) [14, 24]. Although there is agreement that we should support young people at‐risk and during onset or the early stages of BD [1, 24, 25, 29], it is unclear how best to advance this work, and many unanswered questions remain.

This qualitative study was driven by a partnership between Orygen (world leader in early intervention, Australia) and the Daymark Foundation (private family foundation with a focus on BD, Canada). We asked the question: “How might we advance an early intervention approach for people at‐risk of, or with, BD?”. Early intervention was conceptualised as “diagnosis at the earliest possible point, even pre‐symptomatically, followed by proportional or stage‐specific intervention adapted and sustained for as long as is necessary and effective” (p. 545) [30].

2 Materials and Methods

2.1 Design

Semi‐structured interviews with experts and stakeholders in the field of early intervention for BD were conducted. Reporting of the study's findings was in accordance with the Standards for Reporting Qualitative Research (SRQR) [31].

2.2 Researcher Characteristics and Reflexivity

Four senior team members (S.M.C., M.H., C.M., C.S.), each with 20+ years in youth mental health, conducted the interviews. Three were trained in clinical psychology (M.H., C.M., C.S.) and the other in neuropsychology (S.M.C.).

2.3 Sample

The target population was international experts who had worked with young people or adults at risk of, or experiencing, the early stages of BD in clinical and/or research settings. Other included non‐government and philanthropic organisations who support BD research. Experts were identified using a multi‐pronged approach: (i) the research team's own familiarity with the field and established networks; (ii) a scoping review of recent literature to locate authors publishing in early intervention; (iii) consultation with relevant professional bodies; and (iv) snowballing techniques to identify further candidates.

2.4 Interview Schedule

Because “early intervention” can have different meanings, we provided this preamble: “While the term “early intervention in BD is sometimes used to refer only to treatment provided at the first manic or hypomanic episode, for these questions we are using it in the broader sense of also supporting individuals displaying early but significant mood symptoms (which may or may not be diagnosed as BD), as well as preventative interventions for individuals with major risk factors for developing BD”. The scope of interventions was also set: “… the term “intervention” does not necessarily mean pharmacological treatment and can include any other form of care and support to an individual and/or their family (e.g., psychosocial)”.

The semi‐structured interview comprised nine questions covering: (i) research gaps, research challenges, and identifying what works; (ii) challenges with interventions for at‐risk and early stages of BD; (iii) problems with the early detection of BD; and (iv) challenges faced by the field more broadly (see Table 1 for questions).

2.5 Procedure

Ethics approval was obtained from the University of Melbourne (#26552). Experts were sent an introductory email describing the study. For participants, a 1 h Zoom meeting was scheduled where verbal consent was obtained, and the interview conducted. The interview was recorded, and a transcription document was automatically generated. Research assistants checked the quality of the transcription against the audio recordings, and the transcription was amended as required. Interviews occurred between May and June 2023. Participants received a CAD $200 honorarium (Visa gift card).

2.6 Analyses

Transcripts were examined using reflexive thematic analysis [32]. This framework comprises six steps: (i) familiarisation with the data; (ii) developing code labels capturing meaning from data segments; (iii) generating initial themes from the coded data; (iv) reviewing the accuracy of initial themes; (v) refining, defining, and labelling themes; and (vi) report generation [32]. Analyses were led by SC and MH. Each independently reviewed transcripts and labelled meaningful segments without having preconceived themes. They met and discussed initial themes, resolved any disagreements through consensus, and fine‐tuned the final themes.

3 Results

Fifty‐one experts were approached and 28 consented to partake across 27 interviews (two participants were interviewed together). Participants included clinical researchers (n = 19), researchers (n = 3), clinicians, peer support workers (n = 2), philanthropic funders, and representatives from BD advocacy organisations (n = 4). Some participants disclosed a lived experience of BD (personal experience or as a caregiver). Those who were interviewed were from the US (n = 8), Canada (n = 8), Australia (n = 4), UK (n = 4), Europe (n = 3), and one was from South America. Of the 23 who were contacted but did not response or declined to participate most were clinical researchers (n = 17), followed by researchers (n = 4), and support or philanthropic organisations (n = 2). They had similar geographic locations to those who participated; most were from US (n = 8), Canada (n = 6), Australia (n = 5), UK (n = 3), and one was from South America. Because of the small sample sizes, especially across subgroups (i.e., whether or not participated in the interviews, expert types and geographic location), we did not undertake direct examination of differences.

From the interviews, nine themes were identified (see exemplar quotes in Table 2).Theme 1 Awareness. “There is limited recognition and understanding of BD (including early warning signs and at‐risk states) across all stakeholder groups, from the general community through to mental health providers, clinicians, and services”.

In the community, trivialisation and casual dilution of terminology [Participant ID: 7, 14], stigma [3, 12, 15, 18, 24, 26], and denial [3, 12, 15, 18, 24, 26] were viewed as barriers to help‐seeking, diagnosis, acceptance, and engagement in treatment and research. Two participants reported that denial is common in family members and young people, resulting in attribution of symptoms to alternate causes (e.g., adolescent mood fluctuations) [20, 26]. Education to improve recognition, timely diagnosis, and treatment was considered necessary for the general community [3, 9, 13, 14, 19, 20, 24], families [7, 11, 13, 15, 16, 19, 20, 24–26], schools [3, 20, 26] and peers [26].

Participants were clear that general practitioners [11, 15, 18, 19, 21], primary health care practitioners [8, 9, 16, 18, 19], paediatricians [10, 26], and mental health clinicians [3, 8, 10–16, 19, 20, 23–26] need education regarding screening for BD [3, 6, 8, 13, 15, 20] including for family history of BD [3, 7, 9–11, 13, 14, 16, 19, 20]. Two participants reported that clinicians can be reluctant in diagnosing BD [5, 15] and others indicated that diagnoses such as depression, psychosis, and/or borderline personality disorder (BPD) are more likely to be made than BD [3, 5, 8, 9, 11, 13, 14].Theme 2 Definitions. “There is a lack of consensus on key aspects of early intervention in BD including definitions of at‐risk states and what age groups we should be focusing on”.

The field was viewed as lacking consensus on: (i) what quantifies/qualifies as an “at‐risk” state [6, 8, 9, 15, 18, 19, 21, 22]; (ii) how to define “diagnostic delay” [9, 18]; (iii) what characterises the “early stage of BD” [8]; (iv) at what “age” can BD be diagnosed [7, 9, 10, 15, 16, 20]; (v) what is “early intervention” [8]; and (vi) what “measurement tools” should be used in clinical and/or research settings [22]. Because of variability in definitions, some believed that cohorts differ internationally [9, 15, 16].Theme 3 Resourcing. “BD is under‐prioritised and underfunded relative to other mental health disorders especially considering the personal and economic costs associated with the disorder”.

Highlighted as a significant barrier was the limited prioritisation and funding for BD [2, 5, 6, 9, 13–21]. In clinical settings, funding was considered limited for therapy [2, 9, 15, 17–19] and workforce training [15]. In research, BD was described as at the “bottom of the funding pile” [2, 5, 6, 14, 16, 19, 20]. There was the view that traditional funding bodies were uninterested in BD [6, 16, 21]. Early intervention research funding was considered disproportionately directed to psychosis [6, 19, 20], despite the high potential of early intervention for BD [2, 9]. Identifying alternatives to traditional funders was described as necessary [16]; philanthropic investment was viewed as essential for supporting early research, fostering collaborations, and providing data to support future funding applications [6, 13, 16, 21].Theme 4 Measurement and Assessment. “There is a lack of a core battery of validated tools and approaches to help diagnose, monitor, and measure outcomes in at‐risk and early‐stage bipolar groups”.

Measurement was seen as problematic for delineating at‐risk states, diagnosis, and monitoring changes over time [5, 7, 9, 15, 17]. BD was considered the most internationally unreliable diagnosis [9]. Although structured diagnostic interviews were noted to be the best option for diagnosis [15], they were considered too long [15]; clinicians have poor training in their use [15], and they lack diagnostic reliability [5]. Participants also highlighted that diagnosis was complicated by a lack of skill and resources to assist with differential diagnosis [5, 13, 19, 23] and assessment of comorbidities [1, 3, 15, 16, 20].

Because of the heterogeneity of BD, some participants suggested that multiple tools are necessary to assess different phenomena (e.g., (hypo)mania, depression, psychosis) [7, 13]. Existing tools were viewed as inadequate and burdensome [7, 9, 13, 15]. Common measures of depression (e.g., Montgomery Åsberg Depression Rating Scale) were criticised for not detecting atypical depression [2] and for being insensitive to differences between unipolar and bipolar depression [2, 11]. Hypomania and mixed episodes were considered difficult to assess with current tools [8, 9] and are often missed by clinicians [6].

Temporal assessment was viewed as essential for BD [5, 7, 8, 13, 27]. Tools were reported to be inadequate for retrospective mapping of mood episodes (necessary for diagnosis) [7, 8] and for prospective assessments [5, 7, 13]. Some suggested that digital and assistive devices and applications might be useful for temporal monitoring, but this requires further exploration [7, 13, 27].

Several comments were noted about youth assessment. Some participants highlighted that youth rarely present when hypomanic, so BD gets missed [3, 8, 19]. When young people present with depression, participants noted that clinicians fail to consider bipolar depression [3, 20], do not look for signs of atypical depression (e.g., hypersomnia) [3, 8, 20] and/or antidepressant non‐response [6, 8, 13].Theme 5 Data. “We need more meaningful data, harmonisation of existing data, as well as development of bigger prospective datasets”.

A range of problems was identified with current research and associated data. The first issue pertains to data relevance. Data collected tends to have little meaning to individuals with BD and their families [26]. Some noted that cultural, ethnic, and gender diversity are not addressed [13, 24]. Others highlighted the need to focus on functional outcomes [13, 26].

Second, there is a scarcity of data. Data from randomised controlled trials were limited, and few trials have focused on early‐stage BD; as a result, there is a limited understanding of what works and why [5, 6, 9, 14, 16, 18, 22].

Third, participants indicated that there are issues with types of data collected. Several highlighted the inadequacy of cross‐sectional data and the need for longitudinal and real‐time data, especially around sleep and functioning [5–7, 13, 14, 16, 18, 23].

A final issue relates to the pooling and harmonisation of data. Participants suggested that we should pool datasets (including family history, multi‐presentation symptom profile, clinical characteristics, omics, biomarkers, lithium response, genetic, brain imaging, electronic medical records and registry data, Ecological Momentary Assessments (EMAs)/digital technologies, etc.), and use advanced analytical methods such as machine learning approaches to advance our understanding of BD [9, 14, 17, 20, 22]. Pooling of data (whether existing or prospective data) was viewed as limited due to the heterogeneity of definitions [9], no consensus on minimum datasets [22], absence of common templates for assessments [14], lack of systems to facilitate data collection across sites [13], difficulties in recruiting participants [12], and the lack of specialist services [18].Theme 6 Treatments. “The evidence‐base for early intervention treatments in BD is scarce, including which treatments should be offered to which individuals, and when and how treatments should be delivered”.

Participants mentioned that medication, psychosocial and family interventions, peer‐support, and digital interventions may be useful for early intervention. Some noted that there are no clinical guidelines for at‐risk or early stages of BD [8, 17, 23] and no agreement on the adequate dose of early intervention (e.g., should it be 5 years of optimal care?) [2]. These issues stem from the lack of large randomised controlled trials (RCTs) for the early stages of the disorder [5, 6, 9, 14, 16, 18, 22].

For medications, participants highlighted a lack of robust evidence for use, dose, and sequencing of mood stabilisers, antidepressants, antipsychotics, nutraceuticals, and novel therapies in the early stages of the disorder [6, 10, 27]. Some said that lithium, the gold standard treatment, should be used more frequently [6–8, 15, 16, 23], due to its potential as a “disease modifying” agent if used early [3]. Others spoke about challenges associated with selection of first‐line agents for young people [5, 6], the need for more comparative studies [16], and studies evaluating combinations of medication [14, 16], particularly in the presence of various comorbidities [10, 16].

Antidepressant prescription was raised as a concern. Some suggested that more attention needs to be paid to adverse reactions and/or non‐response to antidepressants [5, 6, 8, 13] due to their potential harm in BD (e.g., triggering mania) [2, 3, 8, 10, 15, 24, 27]. Convincing both people with BD and clinicians that mood stabilisers are preferrable to antidepressants was reported to be challenging [2, 3, 5].

Participants highlighted that more information is required on cost‐benefits of medications [3–6, 8, 10, 15, 26, 27]. In young people, participants noted that there was no information on the appropriate duration of medication prescription [23] and whether their treatment response and side‐effect profiles differ from adults [6, 10]. One participant argued that it is hard to make a case to young people to stay on medication when relapse rates are high and side‐effects significant [6]. Physical health monitoring (metabolic, immune, inflammatory) and intervention, some of which can be predictive and/or preventive, was considered poorly integrated into mental health treatment [7, 13, 17, 27].

The overemphasis on medication as the mainstay treatment for BD was noted [7, 20] and despite international guidelines recommending psychosocial interventions, few people receive them [20]. Two participants described that it is not only difficult for people to access therapy, but clinicians do not have the training, skills, and confidence to deliver BD‐specific psychosocial treatments (e.g., cognitive behaviour therapy (CBT) for BD, Family Focused Therapy (FFT)) [10, 20].

Psychosocial interventions can be a benign treatment for at‐risk populations; however, their evidence‐base was considered limited [6, 22]. Some psychosocial interventions were reported to have been trialled on young people or in the early stages of BD (e.g., FFT) while others have been trialled less (e.g., interpersonal and social rhythm therapy (IPSRT)). There was agreement that more evidence on psychosocial interventions was required [6, 9, 11, 22]. Questions were also raised about active ingredients, targets, or mechanisms of effective psychosocial interventions (e.g., exercise, sleep and circadian and social rhythms, diet, substance use) [10, 20]. The lack of developmentally appropriate treatments for young people was also noted [2, 24].

Peer support interventions were viewed as valuable [11, 17, 19, 24, 26]; providing helping individuals and families to accept diagnosis (11). Whether such interventions are best delivered one‐on‐one or in groups remains unclear. It was suggested that lived experience and family peer support groups need to bring together people with common experience (e.g., age, developmental phase, and stage of disorder of the young person with BD) [26, 27].

Digital interventions were viewed as under‐utilised [13, 15, 17]. Promising areas for use of these technologies included: low intensity interventions [20]; interactive websites [15], tools to facilitate engagement [7]; filling gaps in clinician skill [15]; operationalising CBT initiatives or other psychosocial interventions [7]; interactive games (although these are hard to conceive) [15, 27]; engaging with a clinician remotely [15]; immediate accessibility (wherever, whenever people want) [19, 24]; and passive EMAs for symptom monitoring [7, 13–15, 27].

Smart technologies (e.g., wearables or smartphone apps) were considered important for helping people recognise and understand their disorder [7, 13, 27], improving detection of the characteristics of unstable mood (including associated behaviour change and patterns of activity, texting, interaction with social media) [13, 14], overcoming limitations of self‐report and recall, and picking up signs of deterioration [7]. One participant cautioned that for some young people, monitoring can be an unwanted reminder of BD and EMAs are not a therapy substitute [15]. Another participant mentioned that there are over 10,000 mental health apps and although some appear usable and feasible, most are untested [7]. Digital interventions were seen as having dosing and engagement problems [9, 15] and requiring clinician‐supported access, navigation, and engagement [24]. Testing and validation of technologies was considered essential [13, 15, 17].

Two participants viewed the BD field as neglectful of new treatment discovery [5, 7]. A range of therapies was viewed as underused [13, 27], including ketogenic diet [27], transcranial magnetic stimulation (TMS) [13], bright light therapy [13], and pharmacotherapies that act on melatonin [13].Theme 7 Service Capacity and Models. “There is a lack of capacity and skill among clinicians across a broad range of services to diagnose and treat early‐stage BD. It is unclear which service model is best equipped to support the early stages of BD”.

Participants agreed that primary care settings are where young people with BD are likely to be present but are also likely to be unrecognised and not appropriately treated [2, 8, 9, 16]. Similarly, participants reported that BD is not a priority in mental health services and is likely to be overlooked [5, 18, 20]. A contributing factor is that clinicians in these settings see too few patients with the disorder and as such have little experience, knowledge, and confidence in diagnosis and treatment [18]. Further, participants reported that individuals with BD are commonly faced with service barriers including access delays [11, 19], high thresholds for service entry [8, 11, 19, 24], inadequate follow‐up after diagnosis or hospital discharge [19, 20], and lack of continuity of care [5, 20].

Opinions were mixed regarding whether there should be specialist early intervention services for BD, or whether treatment should be delivered in complex mood disorders or transdiagnostic youth mental health services. Specialised clinical expertise in BD was viewed as hard to achieve in the absence of specialised BD services because clinicians do not see enough people with BD [18]. Some participants suggested it may be best to focus early intervention for BD in EIP services [8, 9, 14]. Others suggested that EIP is likely to be unhelpful as BD requires treatment by clinicians with expertise in mood disorders [8, 9]. Another concern was that EIP services are limited to people with psychotic symptoms; many with BD do not present with psychosis [9, 11]. One participant suggested that for EIP services to extend to BD, a culture change was required around the possibility of BD in clients [14].Theme 8 Families. “Families have a key role in assessment and treatment, and should be better supported”.

Families were viewed as an underutilised and valuable resource in detecting early warning signs, diagnosis and treatment [3, 7, 10, 11, 13, 15, 16, 19, 24–27]. Participants expressed concern that collateral information from family members was often not taken seriously, and families must advocate for consideration of family history, possible BD diagnosis, and appropriate treatment [3, 9, 16, 19].

Families were also reported to need better support [3, 7, 10–13, 19, 24, 25]. One participant suggested that families are poorly supported even during acute episodes, but this was the ideal time for family peer support [10]. In families with a history of BD, education regarding early signs of BD could be helpful to aid early detection [3, 25]. Family peer and professional support groups for parents of young people at risk or in the early stages of BD could assist with recognition of BD but are not available [27]. FFT was recognised as a useful treatment for early intervention in BD [4, 9, 10, 14–16, 19–21, 25].Theme 9 Collaboration. “To advance early intervention in BD, we need to look at new opportunities to collaborate that involve all stakeholders, including those with a lived experience of BD”.

Most participants advocated for improved collaborations [1, 6–10, 12–15, 17–19, 21–26]. Lack of collaboration was reported across stakeholder groups, including the interface between researchers, clinicians, service providers, and funding bodies [1, 9, 15, 17, 19]. There was recognition that people with lived experience and families should be more involved in collaborative work with stakeholders, including researchers [11, 13, 19, 26].

4 Discussion

Despite BD being a leading cause of disability, early intervention efforts for the disorder are in their infancy and there is much work to be done. Here we have taken a novel approach by consulting with international experts and using rigorous qualitative methodologies to identify barriers and challenges to advancing early intervention for BD.

4.1 Promoting Awareness

Although community awareness of BD has increased [33], especially driven by media and celebrities reporting a lived experience [34], experts report that awareness is not where it should be; this contributes to diagnostic delay and stigma [35]. Literacy of BD in the community is lower than for schizophrenia, ADHD, and autism [36]. Community awareness campaigns can improve awareness, but they need to be informed by a stronger evidence‐base of early warning signs and prodromal symptoms [35]. Having an international campaign would be helpful; similar approaches have been successfully implemented for psychosis [37].

Medical physicians and other clinicians also have poor literacy regarding at‐risk states, diagnosis, and treatment. Because of time restrictions, depression often being the incipient episode, lack of clarity around prodrome, the presence of comorbidities, and lack of screening and monitoring tools, BD is likely to go unrecognised [38, 39]. A recent study showed 30% of people receiving psychological treatment for depression in the UK's primary care therapy initiative have undiagnosed BD [40]. Another study estimated that three in 20 patients in general practice have undiagnosed bipolar depression [41]. If clinicians could have improved training and resources, then this might aid with earlier detection of the disorder, better coordination of treatments and referrals to secondary and/or tertiary care when needed, and help improve the management of physical health comorbidities (e.g., metabolic syndrome) [41].

4.2 Better Definitions, Measurement, and Data

Definitional problems plague BD, but consensus is essential for accurate identification of at‐risk states, diagnosis, and delineation of the course and outcomes of BD [42]. For research, consensus is important for allowing meaningful comparison across studies, especially around population identification, selection of measurement and outcome tools, data harmonisation, and strengthening the evidence‐base around treatment response and predictors of poor outcome [42]. There have been attempts to drive consensus definitions through the International Society of Bipolar Disorders (ISBD) and specific taskforces focusing on nomenclature of course and outcomes [42], clinical staging [43], precursors and prodromes [44], and cognition [45]; but validation of these definitions is needed [42].

Related to definitional problems are issues with measurement. Research on the psychometrics of tools to assess BD lags behind other disorders [46]. Most tools have been designed for adult populations [46, 47], a problem facing psychiatric research more broadly [48]. There is little focus on young people at‐risk of, or experiencing, the early stages of BD; this is surprising given that the typical age of onset of BD is between 15 and 24 years [45]. Often the focus is on manic symptom severity and diagnostic interviews, but BD is multifaceted and there is a need to consider other symptoms (atypical depression, hypomania) [46], social functioning [46], sleep and circadian rhythms [49, 50], cognition [45, 51], and quality of life [46]. More precise tools are needed to detect early warning signs [52], for BD II (including hypomanic symptoms) [46], as well as for comorbidities [53]. Screening tools are needed for community samples and there needs to be better patient‐reported measures [46]. There is also a growing interest in using technology for assessment purposes [54, 55, 56]. Furthermore, input from those with a lived experience when developing these definitions and measurement tools, is often lacking but essential for ensuring relevance to those affected by BD [42]. Whether using pre‐existing or new measures, there needs to be reasonable consensus about the best measurement tools to be implemented in research and clinical settings. In research settings, harmonisation of tools would allow greater comparability of research data across studies, bigger datasets, and can allow application of newer machine learning techniques to better understand the complexity of BD [57]. There also needs to be allowance for innovation in measurement and new tools to be validated in large cohorts.

4.3 Treatment Support and Innovation

The evidence base for high‐quality pharmacological, psychological, and service‐based interventions for the early stages of BD is scarce [24, 25, 27], an issue compounded by the lack of diagnostic specificity of early warning signs indicating vulnerability to the disorder [29]. Clinical guidelines provide inconsistent recommendations regarding treatments, particularly around the first episode of mania, maintenance treatments, and how to manage comorbidities [25, 27]. However, clinical guidelines need to be informed by high‐quality trials and real‐world data that not only assess the impact of treatments on symptoms but also on other factors such as tolerability and functioning [25].

Although lithium is the first‐line treatment for the disorder (especially for maintenance therapy and prophylaxis), prescription rates have declined by ~50%, which in part relates to clinicians' low levels of confidence and knowledge around the medication including dosage, plasma monitoring, and term side effects [58, 59, 60]. There is some evidence that lithium can have positive effects in young people, including improved symptomatic and functional recovery [61], reduced recurrence of mood episodes [25, 61] and suicide attempts [62], improved cognition [63], and reduced white matter alterations [64]. More evidence is required from larger trials comparing the effectiveness and tolerability of lithium compared to other mood stabilisers and/or antipsychotics, as well as an adjunct to psychosocial interventions [25]. The effectiveness and tolerability of lithium in at‐risk populations also requires consideration of the ethics of treating an individual who may not develop BD [24].

We are still unclear which treatments, or treatment combinations, work best for individuals at different disorder stages. Although clinical trials are the gold standard for testing the effectiveness and tolerability of interventions, generalisability to clinical populations can be limited and implementation into clinical practice slow [65]. We need to consider innovative ways to ensure effective treatments are translated into practice much earlier. This can extend to what service models best support individuals with BD, such as innovative specialised BD‐specific service models [66] or blended care models [67].

4.4 Limitations

This study has several limitations. The participation rate was 54.9%, which might reflect the tight recruitment timelines; a longer period could have yielded a higher rate. Nonetheless, the distribution of expert types and geographic locations between participants and non‐participants was comparable, and additional recruitment would likely not have altered the findings. Data saturation was achieved, with no new themes or insights emerging from the interviews, suggesting the findings were robust [68, 69].

We reported only limited demographic information about the cohort. Given the relatively small field of early intervention in BD, greater detail could have risked compromising participant confidentiality. Although we included a diverse range of perspectives across expert types and geographic regions, subgroup analyses were not possible due to small numbers and the need to protect identities. Finally, perspectives from experts in low‐ and middle‐income countries were not captured, and thus our findings may not fully represent global views on early intervention for BD.

5 Conclusions

In this study, we interviewed experts on advancing early intervention for BD, a field still in its infancy. There is a consensus on the need for clinical, research, and advocacy work. The engagement of participants in this study was encouraging with a shared goal to improve early intervention for BD. Developing collaborative networks that include individuals with a lived experience is vital for advocacy and research funding, potentially transforming early intervention and improving outcomes for those impacted by the disorder. There is much work to be done.

Funding

The Daymark Foundation provided financial support to SC and MH at Orygen for the conduct of this research.

Conflicts of Interest

The authors declare no conflicts of interest.