Early intervention in psychosis teams

This article focuses on how to build on recovery from a first episode of psychosis following treatment with antipsychotics. Early intervention in psychosis (EIP) teams provide an intense and comprehensive treatment package to aid recovery and they are recommended in NICE guidelines (National Institute for Health and Clinical Excellence 2013; National Institute for Health and Care Excellence 2014). Designed for young people during the early stages of mental illness, these teams smooth the transition from child and adolescent to adult mental health services. They generally provide treatment for individuals aged 14–35 years, although this can vary. This service model has a strong focus on the needs of young people more generally as they negotiate the path into adulthood (Box 1).

Individuals are referred to EIP teams on the basis of psychotic symptoms and other threshold criteria. There are a number of differential and comorbid diagnoses to consider, in addition to schizophrenia, as the clinical picture emerges over time (Box 2). Although some have argued against allocating resources to EIP teams, there is evidence to support their effectiveness, particularly during their engagement in an individual’s treatment (Reference Marshall and RathboneMarshall 2011). Patient satisfaction with the EIP approach is also high (Reference Lester, Khan and JonesLester 2011).

More research is needed to determine the most effective ingredients of the EIP approach. It seems likely that one of the most important components is active engagement. Following engagement, individuals can make use of a range of psychosocial and medical interventions that are likely to be effective in preventing relapse of psychosis and improving outcomes more generally. However, the art of engagement is subtle. The findings of the UK National Evaluation of the Development and Impact of Early Intervention Services (EDEN) group highlight how overzealous approaches by clinicians can be experienced as intrusive by young patients (Reference Lester, Khan and JonesLester 2011).

EIP teams usually work with adolescents as well as young adults because most (but not all) psychotic disorders have their onset at this stage in life (Reference JonesJones 2013). Despite being most at risk for a range of mental disorders, young people are the least likely to use services (Reference Gulliver, Griffiths and ChristensenGulliver 2010). Hence, the importance of a flexible engagement approach, which takes account of age-appropriate developmental and social factors. Popular communication technologies can provide a youth-friendly format for disseminating information about mental health. Such resources are likely to be most useful if linked to actual local youth and mental health services.

Medication strategies following recovery from first-episode psychosis may be seen as relatively straightforward and easy to access compared with psychological treatments. Further examination, however, reveals the complexity that surrounds this aspect of treatment. A recent survey of staff in an EIP team provides us with a starting point for the discussion of maintenance medication: all the staff surveyed found it ‘difficult’ or ‘very difficult’ to persuade patients to stay on medication once positive symptoms had receded (details available from the author on request).

Maintenance medication

NICE guidelines highlight the high risk of relapse if antipsychotic medication is stopped following recovery from a first episode of psychosis or schizophrenia, and the necessity of considering each patient’s needs and preferences individually (National Institute for Health and Care Excellence 2014). For those under 18, there is a lack of long-term follow-up studies to guide practice. Nevertheless, NICE advises against intermittent symptom-targeted medication (see below) as a routine approach following recovery (National Institute for Health and Clinical Excellence 2013; National Institute for Health and Care Excellence 2014).

There are difficulties comparing studies that have examined the risk of relapse after first-episode psychosis. For example, definitions of relapse and the availability of non-biological treatments vary in different studies. Diagnostic terminology is also a problem. Some of the more recent research has used the broader definition of first-episode psychosis, whereas earlier studies tended to focus on schizophrenia. Use of this broader definition is in keeping with the aforementioned EIP service delivery model adopted in many countries over the past 20 years, including the UK.

The most recent systematic reviews (Reference Alvarez-Jimenez, Parker and HetrickAlvarez-Jimenez 2009; Reference Leucht, Tardy and KomossaLeucht 2012) concluded that maintenance antipsychotic treatment reduces relapse after both first-episode schizophrenia and first-episode psychosis. However, none of the studies reviewed characterise the subgroup of individuals who can successfully discontinue medication. Individual studies will also be of keen interest to patients and clinicians. Longer-term follow-up studies have indicated relatively high rates of relapse even on maintenance medication. A 5-year follow-up of 118 patients with first-episode schizophrenia or schizoaffective disorder found that relapse rates were high overall, at 80%, and increased by a factor of five in those who discontinued antipsychotic treatment (Reference Robinson, Waerner and AlvirRobinson 1999). More recent follow-up studies have generally indicated significantly increased rates of relapse for individuals who discontinue antipsychotics or significantly reduce their dose compared with those who stay on maintenance treatment.

For example, a Hong Kong group have reported a controlled trial in which 178 individuals with first-episode psychosis treated with quetiapine for 1 year were subsequently randomised equally to quetiapine or to placebo (Reference Chen, Christy and MayChen 2010). The estimated risk of relapse at 12-month follow-up was 41% for the quetiapine group and 79% for the placebo group.

In contrast, a Dutch group reporting on a 7-year follow-up of 103 people with first-episode psychosis found no significant difference in the rates of relapse for both groups (Reference Wunderink, Nieboer and WiersmaWunderink 2013). However, those who discontinued medication or significantly reduced the dose had better recovery than the maintenance treatment group (40% v. 17.5%). It should be noted that recovery was restrictively defined as symptomatic and functional remission for only the preceding 6 months at the time of the follow-up. Furthermore, the overall number of participants who had recovered was small (21 who had discontinued or significantly reduced medication v. 9 on maintenance treatment). The authors themselves point to the need for replication of their findings before change of practice can be recommended. In a related commentary, Reference McGorry, Alvarez-Jimenez and KillackeyMcGorry et al (2013) argue for further discussion on agreeing relevant outcome measures rather than focusing on relapse rates alone. Nevertheless, relapse is a significant life disruption for most individuals, especially if admission to hospital is required. Risk of relapse needs to inform discussions about treatment.

Cognitive decline

The issue of cognitive decline associated with first-episode psychosis deserves particular attention. Reference Lewandowski, Cohen and OngurLewandowski et al (2011) provide a useful overview and highlight the difference between schizophrenia and bipolar disorder. In schizophrenia, the biggest changes occur during and immediately after the first psychotic episode. In bipolar disorder, each successive relapse leads to further deterioration in cognitive functioning. For most individuals with first-episode psychosis the diagnosis and pattern of future cognitive problems will be unclear. However, a key question will be the likely effect of antipsychotic medication on mental abilities. There is evidence that effective treatment of positive symptoms using antipsychotic medication can improve cognitive functioning in schizophrenia and first-episode psychosis (Reference Keefe, Seidman and ChristensenKeefe 2004). However, a recent meta-analysis indicates that antipsychotics may cause structural brain change (Reference Fusar-Poli, Smieskova and KemptonFusar-Poli 2013). The relationship between structural brain changes and actual cognitive decline requires further exploration (Reference Leeson, Sharma and HarrisonLeeson 2011). Of note in this regard is a small study involving 42 patients with first-episode psychosis which indicated that antipsychotic discontinuation or dose reduction led to improved neurocognition (Reference Faber, Smid and van GoolFaber 2011).

Problems with maintenance medication

The potentially serious short- and longer-term medical complications and other side-effects associated with antipsychotic treatment are now well recognised (Reference Foley and MoreleyFoley 2011). Young adults and adolescents are more likely to experience extrapyramidal side-effects even when taking second-generation antipsychotics (Reference CorrellCorrell 2008). Metabolic disturbances also seem to be more likely in younger individuals (Reference CorrellCorrell 2008). Clinically significant antipsychotic-induced hyperprolactinaemia can cause hypogonadism. This in turn will adversely affect the important bone mineralisation that occurs during the teens and 20s.

In response to the range of adverse effects there is now widespread advice and guidelines aimed at harm minimisation. Monitoring of metabolic and cardiovascular parameters for individuals before the start of and during antipsychotic treatment are basic requirements. However, repeated audits of clinicians’ adherence to monitoring guidelines for patients on antipsychotics demonstrate inconsistent outcomes (Reference Foley and MoreleyFoley 2011). Logistical challenges in carrying out regular blood tests and physical health checks in community settings may account for some of the difficulties. Furthermore, it is not clear what effect monitoring regimes have on the decisions of patients and their families to continue with maintenance medication. Some individuals will find the monitoring procedures reassuring, whereas for others they will be a deterrent to continuing treatment. The regular blood tests may be experienced as inconvenient and unpleasant, and they can contribute to feelings of stigma and raise understandable anxieties about possible health problems in later life. Relapse of psychosis on discontinuing medication may seem an acceptable risk if it means avoidance of serious medication side-effects.

Thus, we can see how the balancing of probabilities for individual patients is an imprecise task. There is lack of certainty about who is at risk, both of psychotic relapse and medication side-effects, and when. Our evidence base concerning first-episode psychosis is hampered by inconsistent study methods. Comparing different studies is problematic because of a long list of factors. These include not only variability in non-medical treatments, service organisation, definitions and categorisations of patient samples and outcomes, but also confounding factors such as substance misuse (Reference Marshall and RathboneMarshall 2011). Even what we think we know about risks and benefits of maintenance medication relates to groups of patients. Predictions of outcomes for specific individuals are therefore unreliable.

It could be argued that degrees of uncertainty are a feature of all prognostic predictions in psychiatry and medicine as a whole. This may be a source of anxiety for clinicians and patients alike, particularly when possible adverse outcomes are serious. A return to the evidence concerning informed decision-making, psychoeducation and doctors’ communication skills will help to guide clinical practice in the face of ambiguity.

Cognitive development

In considering how young people may reason and think about mental illness and treatment, it is useful to return to the theories of Jean Piaget developed during the 1950s. Piaget and others described how ‘formal operational thinking’ can be expected to develop during puberty (Reference ColemanColeman 2011). Formal operational thought involves hypothesis-testing and abstract thinking, and it can be demonstrated in experimental problem-solving situations.

Piaget’s theories have subsequently been revised and criticised, but their essence remains relevant. We now know that many individuals do not acquire formal operational thinking capabilities in a predictable manner, and the developmental process is variable and may be delayed into adulthood (Reference ColemanColeman 2011). Also, environmental factors, including educational opportunities, are influential.

Our understanding of cognitive development has become more sophisticated with the expansion of neuroscience specialties over recent years. Reference Blakemore and ChoudhuryBlakemore & Choudhury (2006) note that structural brain changes continue well beyond adolescence and into the late 20s. Reference Crone, Bullens and Van Der PlasCrone et al (2008), among others, have proposed neurophysiological models to explain the observed shifts away from emotional, impulsive decision-making towards more rational approaches during adolescence. Brain system theories also highlight how young people’s abilities to incorporate the views of others concerning risk improve with age. These neurocognitive models indicate that tasks such as planning life goals may be affected by a tendency to respond to immediate rewards and emotions in preference to considering future risks.

Clearly, these issues will affect care planning for young people recovering from first-episode psychosis. Their consideration of the debates concerning maintenance medication will be influenced by their incompletely developed reasoning. In addition, a psychotic episode is likely to interrupt social, educational, emotional and cognitive development. This disruption may further delay maturation at exactly the time when affected individuals are being asked to adjust to the onset of major difficulties and make complex choices about treatment. Any cognitive impairment that occurs leading up to, during or after a first episode of psychosis will compound the situation (Reference Lewandowski, Cohen and OngurLewandowski 2011). Psychiatrists working with adults with more established mental illness, including schizophrenia, are aware of how executive functioning, memory and processing speed problems affect decision-making capacity.

Social development and families

The above considerations highlight the need for young people with first-episode psychosis to be supported by their parents and other trusted adults, including mental health professionals, when making decisions about treatment. The importance of involving families in psychoeducation is also recognised (Reference Lincoln, Wilhelm and NestoriucLincoln 2007). Adolescents and young adults are increasingly influenced by peers as they grow older. Nevertheless, research into this area has consistently shown that parents remain influential and an important source of advice (Reference GrahamGraham 2004).

Reference MiklowitzMiklowitz (2010) has researched and described family-focused approaches to treatments for individuals, including adolescents, with bipolar disorder. The author provides a useful stepped approach to addressing medication non-adherence within family interviews. Of particular note is the need to clarify the symbolic significance of non-adherence. For example, non-adherence can be interpreted as an expression of a young person’s quest for autonomy within a family. Much of this detail about family interventions is likely to be applicable to young people with first-episode psychosis. Where relationships with parents are fragmented, other sources of adult support and mentoring will be needed.

Adolescent identity and relationships with professionals

Several other developmental aspects associated with adolescence and young adulthood will affect how individuals respond to the onset of mental illness, recovery and treatment. The emerging sense of an adult self and identity will be disrupted by the arrival of mental illness. This change in identity was explored in the study published by EDEN group mentioned earlier (Reference Lester, Khan and JonesLester 2011). The young participants highlighted their experiences of a changed sense of self and of perceived stigma. Changes in appearance, including weight gain due to medication side-effects, were contributors in both these areas. The anticipated autonomy and separation from parents and important adults is also affected. A young person’s resistance to the implications of first-episode psychosis is likely to be a reflection of these kinds of issue.

As regards the planning of medication treatment, clinicians might expect to encounter questioning and resistance to advice. However, such challenges can be seen as a positive sign of maturation. Debate is healthy and it frees up clinicians to pose questions about reasons for non-adherence. It is also in keeping with the principles regarding patient-centred decision-making outlined in this article. A Socratic dialogue which encourages the patient to think through their decisions is one technique that can be usefully employed (Box 4).

Conclusions

Following a first episode of psychosis there will be many questions as to how well and in what way a young person can engage in decision-making about their treatment. Thanks to the availability and specific characteristics of EIP services, these questions can be explored over time and in an age-appropriate way. Psychiatrists have a central role and particular skills to offer in this process: first, in making sense of the latest evidence concerning medication strategies; second, in applying knowledge of the developmental processes affecting and arising from mental illness; and third, in applying their extensive training in communication and interviewing skills. These are also areas where further research is needed (Box 5).

MCQs

Select the single best option for each question stem

1. 1 As regards medication adherence in young people:

   1. a parental advice can be influential

   2. b it is best to avoid considering identity issues as part of the discussion

   3. c younger individuals are not able to consider the full implications of medication non-adherence

   4. d the reasons for non-adherence are usually obvious

   5. e providing too much detailed information about side-effects will reduce adherence rates.

2. 2 In young people recovering from first-episode psychosis, maintenance antipsychotic medication for 1–2 years:

   1. a reduces risk of relapse for most individuals

   2. b should not be considered unless a firm diagnosis of schizophrenia has been made

   3. c is of limited benefit if psychosocial treatments are available

   4. d is the most important part of the care plan

   5. e may raise concerns about physical health side-effects, but these never outweigh the benefits of relapse prevention.

3. 3 The effects of maintenance antipsychotics on neurocognitive functioning:

   1. a are unrelated to number of relapses

   2. b are not evident immediately

   3. c are easy to predict

   4. d will not need to be assessed unless there is a complaint of drowsiness

   5. e are hard to differentiate from other causes of neurocognitive impairment.

4. 4 As regards first-episode psychosis:

   1. a clinical characteristics reliably predict which individuals can discontinue antipsychotic medication on remission

   2. b monitoring of calorie intake need not occur until antipsychotic treatment is well established

   3. c most antipsychotics are licensed for the treatment of first-episode psychosis

   4. d young people with first-episode psychosis need to identify for themselves what is a positive outcome of treatment

   5. e community treatment orders are likely to be helpful at this stage of illness.

5. 5 Young people with psychosis:

   1. a are likely to seek help at an early stage

   2. b are no more likely to engage with early intervention services than with generic mental health services

   3. c are more sensitive to medication side-effects

   4. d are unlikely to consider the effects of stigma

   5. e can be given a firm diagnosis at first presentation.

MCQ answers

1

a

2

a

3

e

4

d

5

c