Model of care in Early Intervention in Psychosis: how do we make the square peg rounder?

Michael Jewell

Ryan Lord

M. Aamer Sarfraz


Early Intervention in Psychosis Service (EIS) is a multidisciplinary community mental health service that provides treatment and support, typically for 3-5 years, to people experiencing or at high risk of developing psychosis [1]. This approach is based on the early detection and treatment of emerging symptoms of psychosis as it is thought to prevent relapses and reduce the long-term impact of the condition. Built on influential research on first episode psychosis from the 1980s, frontline early psychosis clinical services were established, first in Melbourne and soon afterwards in many major locations in the UK, Europe, North America and Asia [2]. They were developed as a stand-alone model of care, but with ample multiprofessional resources, and a specific focus on treating people with a first-episode of psychosis. Since this initiative was a part of the National Service Framework (1999), most mental health trusts in the UK adopted it from 2001 onwards, following the publication of the Mental Health Policy Implementation Guide (MHPIG) by the Department of Health (2001) [3].

EIS started with much fanfare but came under criticism soon after its establishment because it had the same in-built challenges, which hampered progress in mental health services more widely, including stigma, cynicism, the silence that surrounds the mentally ill, and doubts about future investment [4]. It also struggled to address conceptual and operational challenges regarding its evidence base, expansion at the cost of other services, rationale for the treatments offered, choice of cases, and inherent chronicity of psychotic illnesses [4,5]. Despite robust defence of this remarkable innovation on behalf of the early adopters [2,6], most of the early criticism has not gone away since, including the long-term outcome of treatments offered, and the increasing burden of questions regarding its cost effectiveness [7, 8].

The perceived effectiveness of EIS has encouraged a view that it favours a shift away from generic towards specialist community care, and a suggestion that the creation of separate specialist teams, has led to the fragmentation and discontinuity of care [9]. The NHS Long Term Plan (2019) has pledged a new framework of “transformation and modernisation” of existing community care that requires a certain threshold of severity of mental health to qualify for support. The proposed new “core” community mental health service will incorporate existing CMHTs with primary care mental health services (for people who need more than Improving Access to Psychological Therapies, IAPT, services can offer but currently fall below CMHT thresholds) and “residential care” services such as supported housing and care homes. This “whole person, whole population” service will offer a wide range of interventions, including care coordination, advocacy, psychological therapies, employment, housing and benefits support, and physical health care.

This article aims to orientate a traditional EIS to the NHS Long Term Plan’s vision for joined-up community mental health care. Our approach is threefold: to better understand the warrants of EIS’ siloed model of care; to describe a case study of a London EIS; and to synthesise and draw conclusions from the first two with the aim of presenting a view on how to situate the future EIS within the NHS Long Term Plan.

Evidence base for EIS model of care

EIS is a multi-component specialist community mental health service which aims to prevent, detect early and treat first-episode psychosis. The central assumption is that earlier intervention results in better outcomes. In the case of psychosis, it is hypothesised that a critical period, when the psychosis first develops, is implicated as a major risk factor for a poorer prognosis [10]. Thus, therapeutically intervening in this early critical period is believed to improve prognosis – this motivated the clinical concept of minimising the duration of untreated psychosis (DUP). The DUP is the time between onset of psychosis and the initiation of treatment. Since the implementation of the UK’s first EIS at the turn of the century, a maturing international evidence base has sought to test these claims. Various quantitative systematic reviews offer insights into the following four major warrants of EIS’ model of care: the link between the DUP and prognosis (2.1); EIS effects on the DUP (2.2); the clinical effectiveness of EIS (2.3); and the cost-effectiveness of EIS (2.4). In addition, qualitative studies have identified that stakeholder perspectives, particularly those of service-users, highly value EIS [11,12,13].

2.1 The link between the DUP and prognosis

The theoretical basis of EIS relates to the conception of chronic psychotic conditions, such as paranoid schizophrenia, as varieties of neurodevelopmental disorders. The concept of the DUP aims to clinically represent a particular neurodevelopmental period when an early adult’s biology and social functioning is most vulnerable to the deleterious effects of psychosis – the ‘critical period’ hypothesis [13]. The prognostic significance of this early untreated period is supported by observational studies that prospectively tracked the natural history of psychotic conditions following first onset [10]. Although it has been contended that the DUP is definitionally imprecise (for example, does initial treatment with psychosocial or pharmacological intervention mark the endpoint of the DUP?), two meta-analyses in 2005 showed that the DUP is related to poorer 6-12 months’ multi-dimensional outcomes [15,16]. In particular, the DUP was an independent predictor of, with small to moderate effects on, symptomatic and functional recovery. Moreover, a meta-analysis (2014) found that long DUP significantly correlated with poorer long-term outcomes, such as severer symptoms and poorer social functioning but was not associated with quality of life or hospital treatment [17]. Thus, it seems that there is a plausible link between the DUP and prognosis.

2.2 How does EIS affect the DUP?

The DUP is a relatively independent measure of prognosis in psychotic conditions. Therefore, it’s important to show that EIS can actually intervene on the DUP. A Norwegian study (TIPS), in 2005, compared the effects of two regions, with and without EIS, on the DUP [18]. The regions with EIS had reductions in the DUP with averages between five and sixteen weeks [19]. This study was not randomised but showed that EIS can reduce the DUP. However, a 2016 US study (RAISE) used a cluster randomisation design to examine the interaction between EIS and the DUP. At baseline, there was no significant difference in the DUP between EIS and the control group. Following two years of EIS intervention, they found that having a shorter DUP (less than the study median of 74 weeks) resulted in better quality of life and symptom outcomes [20]. The authors concluded that a timely EIS intervention in reducing the DUP can optimise outcomes [20]. These studies suggest a complicated interaction between the DUP and EIS; EIS can reduce the DUP, but the DUP can also moderate the effectiveness of EIS. Despite the RAISE study’s findings of EIS and the DUP interacting, a meta-analysis (2017) found that the DUP did not significantly moderate the clinical effectiveness of EIS [21]. One answer is that the meta-analysed median DUPs from the individual studies varied significantly (from 8 – 74 weeks) creating too much noise [21]. Secondly, although EIS has at least two distinct aims of early detection and treatment of First Episode Psychosis, the trials included in the meta-analysis vitiated the first aim by controlling for the DUP at baseline between EIS and standard care arms. Thus, the true efficacy of EIS, in both intervening on the DUP and improving clinical outcomes, was masked. This is in part due to the confusion over what the term ‘early intervention’ denotes: treatment of pre- or post-onset psychosis, or both [22]. Thus, if implemented correctly, the suggestion is that EIS can effectively reduce the DUP.

2.3 The clinical effectiveness of EIS

Studies have tested two independent objectives of EIS: pre- and post-onset psychosis intervention. The quality of pre-onset intervention depends on correctly identifying a clinically high-risk group of developing psychosis. A meta-analysis (2013) found that early detection and intervention on a high-risk group can prevent or delay psychosis, particularly in the first year of treatment [23]. However, prevention interventions are contested in light of the high false positive rate indicated by the small proportion (34.9%) of individuals in the high-risk group progressing to psychosis [24]. Thus, the benefits of pre-onset interventions ought to be weighed against the iatrogenic harms, such as diagnostic mislabelling and unnecessary treatment, of misidentifying false for true cases of psychosis [25]. In response to minimizing iatrogenic harm, and to improve identification of future morbidity, clinical staging models are evolving towards trans-diagnostic models [26]. There is less emphasis on sub-threshold categories like clinically high-risk groups and a greater focus on the dimensions of morbidity. An advantage of this broader approach may provide better predictive tools compared to those based on conventional categorical diagnoses. This is an important area for future research.

In terms of post-onset psychosis intervention, a meta-analysis (2018) of EIS versus standard care for FEP found that EIS was superior in all thirteen outcomes, relating to symptoms, relapses, hospitalisations, functional status and quality of life [21]. Thus, establishing Cochrane level 1 evidence for the effectiveness of EIS. Despite the quality of evidence, there are some lingering concerns. First, the meta-analysed studies only reported short-term outcomes (6 – 24 months) [21]. Long-term follow-up has shown that the beneficial gains of EIS are actually lost [27,28]. Second, although EIS was superior in all thirteen outcomes, only relapse prevention, remission and psychiatric hospitalisation were clinically significant (defined by the authors as a number needed to treat of less than or equal to 10). Third, EIS is a better-funded service with more highly motivated staff compared to standard care. The clinical efficacy of EIS has not been established to be independent from this funding disparity [29]. This issue is made more pressing by the lack of evidence supporting individual elements of EIS, such as psychosocial or pharmacological intervention, that account for its superiority to standard care [30]. It seems that system-level factors such as smaller caseloads, more funding and staff with specialist training could be more important than distinctly specialist modes of intervention.

In summary, evidence supports the case that pre- and post-onset psychosis intervention by EIS is clinically effective. However, interpretations of the effectiveness of EIS are complicated by the risk of iatrogenic harm for the former and, for the latter, the loss of clinical benefit over the long-term and the blunt effectiveness of EIS which could be more closely tied to systemic-level factors.

2.4 The cost effectiveness of EIS

Most studies which support the cost-effectiveness of EIS are service-evaluation studies [31]. In the literature, reduction in hospitalisations is a common index of cost saving in favour of EIS [32]. When a clustered randomisation trial (2016) measured cost-effectiveness, they found that EIS had a high probability (0.94) of being more cost-effective than standard care [33]. EIS benefits also exceeded costs in terms of monetised Quality Adjusted Life Years. A 2019 meta-analysis had also found that the total evidence was consistent with the cost-effectiveness of EIS compared to standard care across various countries [8].

The cost-effectiveness of EIS is connected to the argument that EIS model of care is discriminatory because there are not any relevantly specialist services that are funded for other mental disorders, such as affective or personality disorders [28]. A response could be that the costs saved from EIS could be re-invested into standard care. Furthermore, there is not a relevantly similar critical period hypothesis that justifies early intervention in affective or personality disorders. However, the specialism of EIS has been viewed as a token of a broader approach to adult community mental health service that emphasises specialist teams that aim to address the needs of specifically diagnosed populations. This is a shift away from generic to specialist community teams, which bypasses the aforementioned discrimination. This is the model of care adopted in some London NHS Trusts and elsewhere, but may not be the typical model of care in the rest of the UK.

In summary, the strengths of the EIS model of care can be seen to be represented by the clinical relevance of the DUP, the ability of EIS to reduce the DUP and the superior clinical and cost-related outcomes of EIS over standard care. It is also a service which is highly valued by the staff, carers and, most importantly, patients. The weaknesses include complications in interpreting the clinical effectiveness of EIS, such as iatrogenic harm, diminished benefits over the long-term and the salience of system-level factors, such as better funding.

Case study: a description of a regional EIS

3.1 Clinical descriptors

Figure 1. Total EIS caseload in a London NHS Trust

We examined three EIS teams who each cover a London Borough. All three services were commissioned in 2006. Regarding referrals, males were more commonly referred than females. In terms of ethnicity, most patients were white (approx. 70%), followed by black (approx. 10%); the latter were overrepresented relative to the UK general population but not the local population as described by the 2021 UK Census [35]. Since being commissioned, the total caseload had risen to around 360 (See Figure 1). It rose sharply to 260 patients to begin with, but in late 2011 plateaued and has risen slowly since then apart from an interesting spike in the first year of the COVID pandemic. The average length of an EIS episode of care ranged from 407 to 531 days across the three teams.

In terms of Emergency Department (ED) attendances, the majority of the referrals to EIS (62%) did not present to ED. The average number of ED attendances for mental health reasons per discharged EIS service-user, during their EIS care episode, was between 0.08 and 0.18. In other words, between 1 in 10 and 1 in 20 EIS service-users presented to ED with mental health needs during their EIS care episode. The average rate of psychiatric hospitalisations per EIS referral was similar ranging from 0.12 to 0.16. Following EIS treatment, most service-users (18%) were discharged to the care of their GP.

Regarding serious untoward incidents (SUIs), 47 were reported in total since 2012 excluding self-harm. Most (n=21) were related to abuse and aggression. There was a total of 32 incidents of self-harm over the same period; most were related to actual harm rather than threats.

3.2 Budget descriptors

An examination of the financial statement of EIS in this London NHS Trust indicated that staff salaries contributed most to budget expenditure. For the financial year, up to and including October 2022 (YTD), total salary expenditure was £1,123,621 (See Figure 2). More money was spent than what was budgeted for. The largest contributors to overspend included agency staff (£70,686) and unanticipated hiring of Trust staff (average of £85,003). However, these costs, which were not accounted for in the original budget, were offset by the absence of other staff roles in the multi-disciplinary team (MDT). Therefore, the total overspend on staff salaries up until and including October for the 2022 financial year was only £4,532.

In terms of MDT sizes between the three EISs in the trust, whole-time equivalent (WTE) personnel ranged from 9.36 to 14.73. The gross WTE was less than what was budgeted for averaged across the three EIS by 0.25 for the YTD (October 2022). It is interesting to note that the salary budget was exceeded despite employing a gross WTE that was less than what was budgeted. This perhaps highlights the significant cost of agency work.

Figure 2. EIS total salary expenditure (October 2022 – financial year to date (YTD))

The future of EIS implementation in the NHS

The NHS Long Term Plan outlines a vision for the future of healthcare in the UK, including a commitment to improving community mental health services. The Community Mental Health Framework for Adults and Older Adults within the plan aims to enhance community-based care for individuals with mental health conditions [9]. Integrating EIS and other siloed models of care, according to this framework, can help achieve these objectives. This entails seamless collaboration between EIS, primary care, and other community mental health services to maximise a continuum of care. Integration offers several advantages, such as the potential for improved care coordination, reduced service fragmentation, and enhanced support for individuals with psychosis beyond the critical period.

A theme that emerged from our critical summary of the EIS literature was that EIS is clinically effective, but there are some caveats. First, there is some uncertainty about long-term prognoses following transition from EIS to other community MHS. Second, the kind of clinical effectiveness evidenced in the literature demonstrates its bluntness rather than the incisive quality of its particular therapeutic components. We also identified some local EIS budgetary issues which amounted to excessive agency fees and insufficient recruitment. Joined-up care in community MHS has the advantage of potentially finding solutions to the matters that we have highlighted in the EIS literature and the case study.

4.1 Collaborative care models

A shift from siloed, such as EIS, to joined-up models of care is the topical goal for community mental health service in the NHS Long-term Plan. While EIS excel in the early stages, there is a need to provide continued care and support for individuals with psychosis as they transition into the community; this will help to mitigate the clinical losses stemming from moving between separate models of care that we identified in the EIS literature.

A collaborative approach embodies this shift towards joined-up care and aims to maximise continuity of care. In the case of EIS, this could be achieved by embedding EIS within community MHS which could both extend duration of EIS involvement and promote a gradual and seamless transition between services [13]. An embedded approach fosters integration between siloed care models. The integration can involve shared caseloads, joint decision-making, and regular communication between EIS and other community MHS practitioners.

As well as having the possible advantage of maximising the long-term prognoses of EIS patients, a joined-up care approach could also improve efficiency of resources. Our EIS case-study signalled an over-reliance on agency staff that was costly. Blurring the boundaries between separate community mental health teams promotes shared care and responsibility. A positive upshot is that there is potential for greater flexibility with regard to staff allocation across community services. If applied judiciously, staff shortages could be accommodated internally without relying on expensive agency staff, thus maximising efficiency of resources.


EIS have demonstrated their effectiveness in reducing DUP and improving FEP outcomes. To successfully integrate EIS into the NHS Long Term Plan, a shift from a siloed model of care to a collaborative, community-based approach is essential. This transition should involve expanding the scope of EIS, promoting collaborative care models, enhancing training and education, engaging with communities, and implementing robust monitoring and evaluation processes. As EIS evolve to integrate within the NHS Long Term Plan’s Community Mental Health Framework, it is essential to address the challenges and harness the benefits of a more integrated and whole-person approach to mental healthcare. By doing so, the UK can improve community mental health services and provide comprehensive support for individuals with psychosis throughout their recovery journey.


[1] National Institute for Health and Care Excellence. Psychosis and schizophrenia: treatment and management. (Clinical guideline 178.) 2014.].
[2] McGorry PD, Killackey E, Yung A. Early intervention in psychosis: concepts, evidence and future directions. World Psychiatry. 2008 Oct;7(3):148-56.
[3] Corsico, P., Griffin-Doyle, M. and Singh, I. (2018), What constitutes ‘good practice’ in early intervention for psychosis? Analysis of clinical guidelines. Child Adolesc Ment Health, 23: 185-193.
[4] Horton R. Launching a new movement for mental health. Lancet 2007;370:806.
[5] Castle DJ. The truth, and nothing but the truth, about early intervention in psychosis. Australian & New Zealand Journal of Psychiatry. 2012;46(1):10-13. doi:10.1177/0004867411432553
[6] McGorry PD. Pre-Emptive Intervention in Psychosis: Agnostic Rather than Diagnostic. Australian & New Zealand Journal of Psychiatry. 2011;45(7):515-519. doi:10.3109/00048674.2011.581648
[7] Puntis S, Minichino A, De Crescenzo F, Harrison R, Cipriani A, Lennox B. Specialised early intervention teams for recent‐onset psychosis. Cochrane Database of Systematic Reviews 2020, Issue 11. Art. No.: CD013288. DOI: 10.1002/14651858.CD013288.pub2. Accessed 24 October 2023.
[8] Aceituno, David; Vera, Norha; Prina, A. Matthew; McCrone, Paul (July 2019). “Cost-effectiveness of early intervention in psychosis: systematic review”. The British Journal of Psychiatry. 215 (1): 388–394. doi:10.1192/bjp.2018.298. ISSN 0007-1250.
[9] England N. NHS long-term plan community-mental-health-framework-for-adults-and-older-adults. 2019.
[10] Birchwood M, C Jackson TP. Early intervention in psychosis: the critical period hypothesis. British Journal of Psychiatry. 1998;172:53-9.
[11] Watkins S, Sanderson C, Richards V. Service User Perspectives of an Early Intervention in Psychosis Service. A Service Evaluation. Mental Health Review Journal. 2018;23(3):156-64.
[12] Darker CD, Nicolson G, Reddon H, O’Connor K, Jennings R, O’Connell N. Barriers and facilitators to the implementation of an early intervention in psychosis service in three demonstration sites in Ireland. BMC Health Serv Res. 2023;23(1):653.
[13] Lavis A, Lester H, Everard L, Freemantle N, Amos T, Fowler D, et al. Layers of listening: qualitative analysis of the impact of early intervention services for first-episode psychosis on carers’ experiences. Br J Psychiatry. 2015;207(2):135-42.
[14] Murphy BP, Brewer WJ. Early intervention in psychosis: strengths and limitations of services. Advances in Psychiatric Treatment. 2011;17(6):401-7.
[15] Perkins DO, Gu H, Boteva K, Lieberman JA. Relationship between duration of untreated psychosis and outcome in first-episode schizophrenia: a critical review and meta-analysis. American Journal of Psychiatry. 2005;162:1785-804.
[16] Marshall M, Lewis S, Lockwood A, Drake R, Jones P, Croudace T. Association between duration of untreated psychosis and outcome in cohorts of first-episode patients: a systematic review. Archives of General Psychiatry. 2005;62:975-83.
[17] Penttila M, Jaaskelainen E, Hirvonen N, Isohanni M, Miettunen J. Duration of untreated psychosis as predictor of long-term outcome in schizophrenia: systematic review and meta-analysis. Br J Psychiatry. 2014;205(2):88-94.
[18] Johanessen JO, Larsen TK, Inge J, Melle I, Friis S, O’pjordsmoen S, et al. Pathways to care for first-episode psychosis in an early detection healthcare sector: part of the Scandinavian TIPS study. British Journal of Psychiatry. 2005;48:24-8.
[19] Melle I, Larsen TK, Haahr U, Friis S, Johanessen JO, Opjordsemoen S, et al. Reducing the Duration of Untreated First-Episode Psychosis. Archives of General Psychiatry. 2004;61.
[20] Kane JM, Robinson DG, Schooler NR, Mueser KT, Penn DL, Rosenheck RA, et al. Comprehensive Versus Usual Community Care for First-Episode Psychosis: 2-Year Outcomes From the NIMH RAISE Early Treatment Program. Am J Psychiatry. 2016;173(4):362-72.
[21] Correll CU, Galling B, Pawar A, Krivko A, Bonetto C, Ruggeri M, et al. Comparison of Early Intervention Services vs Treatment as Usual for Early-Phase Psychosis: A Systematic Review, Meta-analysis, and Meta-regression. JAMA Psychiatry. 2018;75(6):555-65.
[22] Singh SP. Early intervention in psychosis. Br J Psychiatry. 2010;196(5):343-5.
[23] Van der Gaag M, Smit F, Bechdolf A, French P, Linszen DH, Yung AR, et al. Preventing a first episode of psychosis: meta-analysis of randomized controlled prevention trials of 12 month and longer-term follow-ups. Schizophr Res. 2013;149(1-3):56-62.
[24] Nelson B, al. E. Long-term follow-up of a group at ultra high risk (“prodromal”) for psychosis: the PACE 400 study. JAMA Psychiatry. 2013;70(8):793-802.
[25] Killackey E, Yung AR. Effectiveness of early intervention. Current Opinion in Psychiatry. 2007;20:121-5.
[26] Shah JL, Scott J, McGorry PD, Cross SPM, Keshavan MS, Nelson B, et al. Transdiagnostic clinical staging in youth mental health: a first international consensus statement. World Psychiatry. 2020;19(2):233-42.
[27] Albert N, Melau M, Jensen H, Emborg C, Jepsen JR, Fagerlund B, et al. Five years of specialised early intervention versus two years of specialised early intervention followed by three years of standard treatment for patients with a first episode psychosis: randomised, superiority, parallel group trial in Denmark (OPUS II). BMJ. 2017;356:i6681.
[28] Neale A, Kinnair D. Early intervention in psychosis services. Br J Gen Pract. 2017;67(661):370-1.
[29] Craig TK, Garety P, Power P, Rahaman N, Colbert S, Fornells-Ambrojo M, et al. The Lambeth Early Onset (LEO) Team: randomised controlled trial of the effectiveness of specialised care for early psychosis. BMJ. 2004;329(7474):1067.
[30] Thorup A, Petersen L, Jeppesen P, Ohlenschlaeger J, Christensen T, Krarup G, et al. Integrated treatment ameliorates negative symptoms in first episode psychosis–results from the Danish OPUS trial. Schizophr Res. 2005;79(1):95-105.
[31] Catts SV, O’Toole BI, Carr VJ, Lewin T, Neil A, Harris MG, et al. Appraising evidence for intervention effectiveness in early psychosis: conceptual framework and review of evaluation approaches. Australian and New Zealand Journal of Psychiatry 2010(44):195-219.
[32] Cullberg J, Mattsson M, Levander S, Holmqvist R, Tomsmark L, Elingfors C, et al. Treatment costs and clinical outcome for first episode schizophrenia patients: a 3-year follow-up of the Swedish “Parachute Project” and two comparison groups. Acta Psychiatr Scand. 2006;114(4):274-81.
[33] Rosenheck R, Leslie D, Sint K, Lin H, Robinson DG, Schooler NR, et al. Cost-Effectiveness of Comprehensive, Integrated Care for First Episode Psychosis in the NIMH RAISE Early Treatment Program. Schizophr Bull. 2016;42(4):896-906.
[34] Statistics OfN. UK Census 2021. 2021.

About the authors

Dr Michael Jewell is a Higher Trainee (ST6) in psychiatry working at Oxleas NHS Trust.

Mr Ryan Lord is a Senior Quality Data Analyst working at Oxleas NHS Trust.

Prof. M Aamer Sarfraz is a Consultant Psychiatrist, and visiting professor at the Canterbury Christ Church University.