Multielement psychosocial interventions for first-episode psychosis are feasible and show promise in generic mental health services

What is already known on this topic

Specialist First Episode Psychosis (FEP) services are considered the gold standard of early intervention for psychosis. They adopt a youth-friendly, recovery-focused model and controlled studies have demonstrated their effectiveness in improving a number of key outcomes including psychotic symptoms, relapse rates, functional recovery, engagement with services and patient satisfaction.1 ,2 However, when the deployment of specialist services is not possible, an alternative may be implementing the principles of early intervention for psychosis within existing mental health services. This study is aimed at evaluating the effectiveness of multielement psychosocial interventions delivered within generic mental health services in patients with FEP.

Methods of the study

The study was a cluster-randomized controlled trial (c-RCT), with 96 community mental health services being allocated to either receiving expert training and support in psychosocial interventions for FEP patients (N=48) or treatment as usual (TAU) (N=48). The multi-element psychosocial intervention included cognitive behavioral therapy for psychosis (CBT-p), targeted family interventions and case-management. The study population comprised 444 clinically remitted participants (272 receiving the multi-faceted psychosocial intervention and 172 receiving TAU) meeting a broad definition of FEP. Primary outcome measures included psychotic symptoms and days of hospital admissions at 9 months, with secondary outcome variables comprising appraisal of psychotic symptoms, global functioning, depression and treatment disengagement.

What does this paper add

• It is feasible to successfully train in a short period of time (ie, 6 months) clinicians from generic mental health settings in delivering specialised psychosocial interventions for FEP. Approximately 95% of those trained in CBT and family interventions achieved the required competence score, and fidelity to the principles of CBT was judged to be medium-to-high by expert supervisors.

• While multielement interventions for FEP delivered in generic mental health settings may produce small effects in terms of positive and negative psychotic symptoms and hospital admissions when compared with TAU, they showed promise in improving domains related to general functioning (effect size 0.35, 95% CI 0.06 to 0.64), depression (−0.25, 95% CI −0.48 to −0.03) and subjective appraisal of delusions (−0.82, 95% CI −1.29 to −0.35).

Limitations

• A major limitation of the study is that study personnel were not blind to treatment allocation. This may introduce biases that go above and beyond the assessment of outcomes. For example, the leakage study carried out by the researchers to ensure accuracy of recruitment showed that potential cases were more likely to be missed in the TAU group compared with the experimental group (45 vs 18 missed cases, respectively).

• The intake criteria employed by the researchers relied on a broad definition of FEP. Specifically, participants could be recruited if they experienced only hallucinations or delusions, or presented with loss of interest along with social withdrawal. This inclusive definition could have resulted in a subset of participants meeting traditional ultrahigh risk (UHR) of psychosis criteria being recruited into the study,3 which has implications for the external validity of the study findings. This further raises questions as to the appropriateness of the CBTp intervention evaluated in the study (which has been previously modified to be applied in UHR samples).4

• The pre-established threshold for optimal dosage (20 sessions for patients with FEP and 10 for families) was received by 66.3% of patients and 71.4% of relatives. This level of exposure to specialised psychosocial interventions is considerably smaller than that provided by specialised FEP services1 and recommended by guidelines.5

• The cluster RCT design may have reduced statistical power to detect intervention effects and introduced biases related to the characteristics of different mental health services (clusters) being different across treatment conditions (this was not formally tested). However, the cluster randomization was a good choice for this study and the statistical analysis was appropriate.

What next in research

• Further studies should assess the effects of applying the principles of early intervention for psychosis within generic mental health services on social and functional recovery. In addition, future research should investigate the cost-effectiveness, long-term effects and family outcomes of this approach to early intervention for psychosis.

Do these results change your practices and why?

This study does not address the issue of the effectiveness or advantages of this approach to early intervention in psychosis when compared with specialist FEP services. That said, the results from the study add to the body of research demonstrating that specific interventions improve outcomes in young people with psychosis when compared with TAU. Moreover, the study showed that it is possible to instil principles consistent with evidence-based interventions for FEP within mainstream mental health services.