Evidence for preventive treatments in young patients at clinical high risk of psychosis

citation for published version (APA) Nelson, B., Amminger, G. P., Bechdolf, A., French, P., Malla, A., Morrison, A. P., Schmidt, S. J., Shah, J. L., Thompson, A., Van der Gaag, M., Wood, S. J., Woods, S. W., Yung, A. R., & McGorry, P. D. (2020). Evidence for preventive treatments in young patients at clinical high risk of psychosis: the need for context. The Lancet. Psychiatry, 7(5), 378-380. https://doi.org/10.1016/S2215-0366(19)30513-9


IM Varney/Science Photo Library
Evidence for preventive treatments in young patients at clinical high risk of psychosis: the need for context Cochrane reviews, as rigorous evaluations of evidence in health care, have a substantial effect on clinical and policy decision-making; however, their findings and methods need to be contextualised. These reviews are done by groups of academics who might or might not have adequate expertise or clinical experience in the field they examine, and we feel the methods can be indeterminate and conservative. The recent Cochrane review 1 of intervention trials for patients at clinical high risk of psychosis concluded that, despite the considerable research effort in this area, the evidence base was weak and firm conclusions could not yet be drawn. The authors noted that the "strongest weak evidence" supported the ability of omega-3 fatty acids to prevent the onset of psychosis in the clinical high risk population, but that the quality of evidence overall was low to very low.
We have several methodological concerns about the Cochrane review. 1 First, a major contributor to the lowto-very-low quality rating of studies was their risk of bias (eg, randomisation and allocation concealment methods not being described, the risk of unblinding, and high attrition). However, many studies included in the review used rigorous methods of randomisation and allocation concealment without detailing these in print. 2 Moreover, most mentioned studies were psychosocial or psychotherapy trials, in which it is impossible to implement masking of therapists and notoriously difficult to maintain patient masking. High attrition is also common in all trials involving youth with mental disorders.
Second, derived from studies of medications for acutely unwell patients with psychosis, the criterion of a 50% reduction in symptoms used to judge clinical improvement might be inappropriate for the clinical high risk group and represents an unrealistic goal for a group of patients who, by definition, have symptoms of moderate intensity. 3 Even in clinical trials of pharmacological and psychological interventions for acutely ill patients with first-episode psychosis and schizophrenia, response is usually set between 20% and 50% symptom reduction. 4 Finally, the Cochrane review compared different categories of interventions across randomised controlled trials (RCTs) with control conditions. 1 Although this method is arguably a better approach than network meta-analyses, it meant that the critical issue of whether, when all pooled together, specific targeted interventions were superior to standard treatment was left unaddressed. When this issue has been addressed, the onset of psychosis in the clinical high risk population could at least be delayed through specific targeted treatments, with a 50% risk reduction over 12 months. 5,6 The Cochrane review did show the benefits of cognitive behavioural therapy (CBT) over supportive therapy, with a number needed to treat (NNT) of 13 over 1 year and a relative risk of 0·45 (about 8% vs 16% transition rate), however, the conclusion and summary sections downplayed this important finding. 1 An NNT of 13 is certainly clinically meaningful and compares favourably with antipsychotic medication preventing psychosis relapse, as well as treatments in other areas of medicine. 7,8 The decline in the effect of CBT over time is not unique, with treatments needing to be sustained in many other health conditions (eg, antipsychotic medication for psychosis relapse prevention and insulin for diabetes).
The Cochrane review ignored biological analysis from the RCTs that supported the protective function of omega-3 fatty acids. 1 These studies indicated that omega-3 concentrations at baseline and their increase in clinical high risk trial participants predict clinical improvement, highlighting the potential value of omega-3 fatty acids as a treatment option. 9 The authors correctly pointed out that treatment studies have consistently been underpowered. Although prioritising multisite studies to increase sample size is certainly one solution, the Cochrane review 1 ignored the importance of enriching samples in order to evaluate preventive treatments. Stratification of risk within clinical high risk cohorts is a highly active area of research.
Standard treatment is not a fixed entity. Background service-level contextual factors are likely to have improved over time and standard treatment to have been refined. Therefore, the control and comparison conditions have probably become more effective in recent trials, which, coupled with the observed rise in placebo response, means that trial interventions need to improve on already efficacious treatments.
The authors recommended a two-stage approach for future treatment studies, which firstly compares low-dose, antipsychotic medication with psychosocial treatment as usual before progressing to a second step that compares different components of the psychosocial treatment as usual. The recommendation to use antipsychotic medications first is incongruent with the evidence base, as outlined in the Cochrane review itself, with an unfavourable risk-benefit ratio. 1 The evidence base needs to be regularly reviewed; however, the details of the methodological approach taken and the broader context of the treatment trials are both essential. Although we agree that more highquality research trials are needed to enhance outcomes and determine the most effective type and sequence of interventions, the evidence base shows that helpseeking individuals at clinical high risk benefit from the available treatments, including standard treatment, without iatrogenic harm associated with antipsychotic medications. 10 This potential for improvement is a key message for patients, families, and practitioners. We are concerned that this recent Cochrane review instead negatively framed the evidence base and did not convey any treatment benefits. 1 If heeded, the review's message would result in many help-seeking young people being denied much needed psychosocial care and being exposed to the risks of worsening symptoms and functioning.
We declare no competing interests.

Ethnicity and mental health: a new beginning
The UK is a high-income country with a publicly funded mental health-care system, which is free at the point of use. The UK's National Health Service (NHS), like other public bodies, is legally obliged to ensure fair and equal services. However, ethnic inequalities remain established in most aspects of mental health care in the UK. 1 Black and minority ethnic (BME) people do not do as well as the white majority in any aspect of mental health care and, generally, they fare much worse.
The nature and extent of racial discrimination in mental health care has been known for over half a century. Over the years, however, there has been no change in the experiences of people from BME communities who use mental health services. Despite the continuing rhetoric on race and mental health, 2,3 and more promises of change, 4 there is no parity between BME communities and the white majority in access, experience, or outcomes of mental health care. 5 There are several reasons for this absence of progress in reducing ethnic inequalities in mental health care. First, most academic and professional focus in this area has been on the probable reasons for the ethnically differentiated nature of psychiatric care, rather than on the shortcomings within current services. Second, changing established practices creates a sense of helplessness and pessimism. Third, despite various policy initiatives on race and mental health, there is still no national plan or strategy to reduce race inequalities in treatment and outcomes. Fourth, political and professional leadership has largely been absent in both the government and the NHS in tackling ethnic inequalities. The lack of commitment is evident from the continuing failure to implement the recommendations from various reviews and national inquiries, and omission of any investment in this area.
Further inquiries or reviews will add little to what is already known. The problems are already well understood, and despite the complexity of underlying issues, it is clear what changes are required. For example, a wealth of evidence exists that is based on the experience of service users and the black communities, and many examples of what works for the benefit of patients and their families (panel). Most crucially, the BME communities and agencies are engaged and willing to work with statutory providers to bring about change.
The Ethnicity and Mental Health Improvement Project (EMHIP) in Wandsworth, southwest London, UK, is an attempt to bridge the gap between policy rhetoric and practice. EMHIP is a collaborative project involving the local mental health service, South West London and St George's Mental Health NHS Trust (SWLSTG), and a BME community mental health organisation, Wandsworth Community Empowerment Network (WCEN). WCEN has been at the forefront of challenging the unjust patters of mental health care in southwest London as well as mobilising resources and creating networks in the local community. 6 Although SWLSTG and WCEN have worked together for over a decade, they have been unable to change the ethnically differentiated pattern of mental health care locally. Over the years, it has become clear that fundamental reconfiguration Panel: Evidence for changing ethnic disparities in mental health care • Evidence based on research, national inquiries, and reports • Evidence based on experience: service users and black and minority ethnic communities • Examples of good practice • Local evidence from services and communities