Intervention strategies for ultra-high risk for psychosis: Progress in delaying the onset and reducing the impact of first-episode psychosis

Abstract

Over a quarter of a century ago, the formulation of the “at risk mental state” and operational criteria to prospectively identify individuals at “clinical” or “ultra-high risk” (UHR) for psychosis created a global wave of research momentum aimed at predicting and preventing first-episode psychosis. A substantial number of randomized controlled trials (RCTs) were conducted to determine if transition to psychosis could be delayed or even prevented. The efficacy of a range of interventions was examined, with standard meta-analyses clearly indicating that these could at least delay transition for 1–2 years and that outcomes improve. Recently, network meta-analyses have attempted to identify the most effective intervention. These highlighted the fact that no one form of intervention is superior to the rest, a finding interpreted in such a way as to create doubts concerning the value of intervening. These doubts have been reinforced by a subsequent Cochrane review which judged the quality of the evidence as low or very low. Here, we report a narrative review of findings from RCTs and meta-analyses on the efficacy of interventions in UHR. We also critique the network meta-analyses and the Cochrane review, and indicate that many of the trials were of the highest possible quality for such research, and were published in top ranked psychiatry journals, which demand such quality. Although outcomes vary, and the UHR group is clearly heterogeneous, we highlight the clinical benefits of psychosocial treatment. The next generation of clinical trials seek to elucidate the optimal type, duration and sequence of interventions.

Introduction

The creation of the “at risk mental state” and criteria to prospectively identify individuals at “clinical” or “ultra-high risk” (UHR) for psychosis was the catalyst for global research seeking the elusive goal of preventing the onset of psychotic disorder (Cannon et al., 2008; Yung et al., 1996; Yung et al., 2003). A paradigm shift led to a range of candidate interventions being tested to determine their efficacy in preventing transition to psychosis and improving symptomatic and functional outcomes.

Results from initial randomized controlled trials (RCTs) were encouraging and assembled conclusive Cochrane level 1 evidence demonstrating that intervening could reduce the risk of transition from UHR state to a psychotic disorder by around 50% for at least 1–2 years and relieve current symptoms (Preti and Cella, 2010; Stafford et al., 2013; van der Gaag et al., 2013). As with treatments for other mental disorders, notably depression (Amick et al., 2015; Barth et al., 2013), several different interventions were equally effective, with similar effect sizes and number needed to treat (NNT) results. This is likely a reflection of clinical heterogeneity in this population, or that several routes lead to the same result.

Recent meta-analyses confirmed that many interventions are as effective as each other, and further concluded that there was a current lack of evidence to support the use of any specific intervention across a range of outcomes, including transition to full threshold psychosis and attenuated psychotic symptoms (Davies et al., 2018a; Davies et al., 2018b; Devoe et al., 2019a). However, there are problems with the choice of research questions and methodology in these recent network meta-analyses (NMAs) (Stein and Norman, 2019), notably treating active control conditions as the equivalent of placebo. Further doubts about the efficacy of treatments in the UHR group were recently reinforced by, in our view, a confusing review published by the Cochrane Schizophrenia Center (Bosnjak Kuharic et al., 2019). While it may be the case that no specific intervention is more efficacious than any other for the UHR group, the interpretation of recent NMAs have served to obscure the benefits of cognitive behaviorally influenced psychosocial interventions. Cognitive behavioral therapy (CBT) is safe and relatively potent (Hutton and Taylor, 2014). It remains the recommended treatment in the National Institute for Health and Care Excellence guidelines (NICE, 2013; NICE, 2014), the British Association of Psychopharmacology guidelines (Barnes et al., 2020), and the Australian Clinical Guidelines for Early Psychosis (Early Psychosis Guidelines Writing Group and EPPIC National Support Program, 2016).

This critical narrative review revisits evidence from individual RCTs in the UHR population, re-examines recent meta-analytic and Cochrane findings, and proposes future research directions.