1. Introduction
The National Institute for Health and Care Excellence (NICE) defines psychosis as a major psychiatric disorder in which a person’s perception, thoughts, mood and behaviour are significantly altered. Symptoms are divided into positive symptoms, negative symptoms, and cognitive symptoms (NICE Guidelines, 2014). This review will explore how modern psychiatry is now aimed at preventing the onset of psychosis, owing to its severe and disabling outcomes, both when left untreated and when treated within standard care (SC). We will review the current model of care compared to standard care, as well as reviewing its evidence base in preventing a first episode (FE) of psychosis, or in providing care for those with a FE.
2. A scientific revolution for modern psychiatry
Prevention, in conjunction with evidence-based treatment, has played a large part in reducing the instance of physical health diagnoses. There are few reasons why this preventative approach cannot be explored with psychopathology. Indeed, modern psychiatry has now shifted toward early intervention and prevention within an evidence-based staging model dependent on the progression of the illness.
Early Intervention (EI) services are now recommended by NICE in their most recent guidelines for the treatment of psychosis (NICE, 2014) and the National Institute of Mental Health (NIMH) as part of their initiative for recovery after a first episode of psychosis (Goldstein & Azrin, 2014). Indeed, research has shown that the outbreak of psychosis is preceded in around 70-100% of cases by an initial prodrome which lasts on average 5 years. This is considered a “critical period” (Hafner et al., 1992; Birchwood & Macmillan, 1993; Birchwood & Fiorillo, 2000) when a patient’s prognosis is poor unless treatment is provided (Crumlish et al., 2009; Lutgens et al., 2015). The notion of a prodrome is nothing new and has been discussed as early as the Kraepelin era (Kraepelin, 1921). However, establishing a population of individuals deemed to be at “Ultra-High Risk” (UHR) of developing psychosis has only been the focus of research (Yung et al., 1996; Fusar-Poli et al., 2012) and indeed clinical practice (McGorry, 2015) in the past few decades. The UHR sample can be defined as those that present with symptoms of psychosis that are subthreshold and of reduced severity to warrant a diagnosis of a psychotic episode, or schizophrenia, for example (Fusar-Poli et al., 2013). In developing this concept, we are aiming to prevent or delay the onset of a FE of psychosis.
EI began with McGorry and colleagues in Melbourne, Australia in a 10-bed clinical research unit (McGorry, 1985; Copolov et al., 1989) and has now been widely implemented to varying degrees throughout England (Broome et al., 2005), Canada (Iyer et al., 2015) and European countries (Maric et al., 2018) to point out a few. EI services are community-based multidisciplinary teams that seek to reduce duration of untreated psychosis and improve outcomes (Craig et al., 2004; Correll et al., 2018). This can include low-dose antipsychotic medication with close case management encouraging adherence, cognitive behavioural therapy (CBT), family treatment and psychoeducation, as well as reintegration into employment and/or education (Lutgens et al., 2015; Kane et al., 2016; Correll et al., 2018).
3. The paradigm shift
The clinical staging model as a premise for EI services is an appropriate view for the treatment of psychosis as different levels and intensity of treatment are likely necessary at different stages throughout the duration of the illness, as opposed to a generic all-round intervention posited in SC. SC in the United Kingdom can be defined as the regular care with a Community Mental Health Team (CMHT) of whom have often received no additional training in treating psychosis (Craig et al., 2004; Lutgens et al., 2015). Staging models may also allow us to establish reliable predictors of transition into FE and in turn prevent this.
When comparing SC and EI services, “diagnosis” of being in the UHR sample differs to a diagnosis of FE or schizophrenia, owing to the nature of a prodromal stage of illness. Diagnosis of schizophrenia, for example, is conducted via diagnostic interview using criteria specified either in the DSM-IV (American Psychiatric Association, 2013) or ICD-10 (World Health Organisation, 1992). On the other hand, semi-structured interviews conducted by Jackson and McGorry (Jackson et al., 1994) pathed the way for implementation of the Comprehensive Assessment of At-Risk Mental States (CAARS; Yung et al., 2005) which is now widely implemented in EI services. Of interest when comparing EI and SC, the CAARMS requires functional decline over 12 months, whereas DSM-IV criteria requires active symptoms/episode present for at least 1 day and less than 1 month, suggesting a protracted UHR state of gradual global decline compared to a definitive time-sensitive episode of psychosis i.e. the 5 year “critical period” (Birchwood & Fiorillo, 2000).
4.1 What does normal science say?
Alvarez-Jiminez et al (2009) conducted a meta-analysis into comparison of EI services versus SC. By comparing the two services, the most efficacious method of treatment can be ascertained. They found that EI services in 3 RCTs were more effective in preventing relapse. The reverse has also been seen to be true in that, duration of untreated psychosis (DUP) defined as time spent from first episode of psychotic symptoms before receiving psychological care (Watson et al., 2018), is a predictor of poor prognosis in psychosis in both the short-term (Marshall et al., 2005; Singh, 2007) and the long-term (Penttila et al., 2014). Even further so, EI services generally produced significantly reduced conversion to a FE psychosis at 6-48 month follow-up compared to SC (Schmidt et al., 2015) with a meta-analysis predicting the estimation transition risk for those engaged in EI services to be 18% after 6 month follow-up period, 22% after 1 year, 29% after 2 years and 36% after 3 years (Fusar-Poli et al., 2012). Therefore, not only is the duration and impact of the illness less severe and protracted with treatment in EI services, but it can also be prevented in a large percentage of the UHR population, when compared to SC. Indeed, in a meta-analysis including over 1,000 patients, global functioning showed significant improvement in EI services compared to SC (Correll et al., 2018). Wider outcomes spanning 6-24 months of EI services found that coordinated and integrated treatment was associated with superior outcomes compared with SC.
Correll and colleagues (2018) conducted a meta-analysis, finding that the risk of at least 1 psychiatric hospitalisation among 2,105 patients was significantly lower with EI services than SC (32.2% vs 42.4%), with hospitalisation being associated with higher costs. Indeed, annual National Health Service (NHS) costs were significantly lower for patients using EI services compared to non-users, with an annual saving of £4,031 per patient (Tsiachistas et al., 2016). Furthermore, the economical and occupational consequences of late intervention are vast when we consider that the onset of psychosis is, on average, late adolescence and early adulthood (Kessler et al., 2007) when such people are paving the way for their adult lives (Insel & Fenton 2005; McGorry 2015) and may present with reduced occupational functioning and discontinue their education. Indeed, if all people with FE were treated with EI services, savings are estimated at £44 million per year, providing good support for the validation and implementation of such services compared to SC with an estimated 35% of patients in employment under EI services, compared to 12% for SC (NICE Guidelines, 2014).
5.1 Future puzzle-solving
Up to 70-80% of the UHR sample have at least once Axis I comorbid diagnosis (Fusar-Poli et al., 2012; Salokangas et al., 2012) with prevalence rates of 40.7% for depressive disorders and 15.3% for anxiety disorders in UHR individuals (Fusar-Poli et al., 2014). This may mean that the premorbid state of both the UHR population and an FE patient, may affect their presentation or clinical outcome. For example, abnormal cortisol & function of the hypothalamic-pituitary-adrenal (HPA) axis in the presence of a stressor has been implicated in psychosis and indeed discussed as a predictor of FE. But, this dysfunction has also been noted in other psychiatric disorders such as depression and PTSD (Chida & Steptoe, 2009; Ulrike et al., 2013; Pruessner et al., 2017). Some researchers have proposed that this should lead to an at-risk state for general psychopathology (van Os & Guloksuz, 2017; Albert et al., 2018), however this is outside of the scope of this review.
Gender may also be a potential moderator for transition in FE. Poorer baseline social functioning and greater positive symptoms predicted higher conversion rates in male but not female patients (Walder et al., 2013). Furthermore, hippocampal volume (HV) has been correlated with positive and negative symptoms of psychosis (Brambilla et al., 2013; Mathew et al., 2014), but smaller bilateral HV was significantly related to positive psychotic symptoms and lower levels of functioning only in male but not in female patients (Pruessner et al., 2015; Pruessner et al., 2017). Therefore, a source of heterogeneity in research may indicate patient subgroups based on gender. This could also implicate a necessary adaptation of treatment within EI services based on the manifestation of the disease for each gender.