Summary Some 20 trials of cognitive – behavioural therapy (CBT) for psychosis have re-established psychotherapy as a credible treatment for psychosis. However, it is not without its detractors and problems, including uncertainty about the nature of its active ingredients. We believe that the way forward is to abandon the neuroleptic metaphor of CBT for psychosis and to develop targeted interventions that are informed by the growing understanding of the interface between emotion and psychosis.
How times have changed! It was not long ago that talking to people about their psychotic beliefs was deemed impossible or harmful. Yet cognitive–behavioural therapy (CBT) for psychosis is now recommended by the National Institute for Clinical Excellence (2002) to ‘reduce psychotic symptoms, increase insight and promote medication adherence’. This mainly British innovation has been based on some 20 randomised controlled trials (Tarrier & Wykes, 2004) using predominantly standard psychosis outcomes (Positive and Negative Syndrome Scale and relapse). CBT for psychosis has mirrored the evaluation of neuroleptics to reestablish psychotherapy as a credible treatment for psychosis, and it has succeeded. But was this the right approach and are we moving in the right direction? The development and evaluation of CBT for psychosis have tended to follow the drug metaphor in the way that they have pragmatically applied an intervention that is successful in treating one disorder (in this case depression) to another (psychosis), and applied the same criteria for success (psychosis symptoms and relapse). We have participated in this development, but we believe that it has led to many unintended consequences that cannot be rectified without a decisive change of course. Contemporary CBT for psychosis began in the 1980s with the work of Tarrier and colleagues which aimed to help patients to cope with their symptoms (Tarrier et al, 1993). At the same time, Chadwick & Lowe (1990) showed that it was possible to ‘reality test’ delusional beliefs. Then followed the full armamentarium of CBT, emphasising individual formulation and bringing in the assumptions and techniques that are used for patients with depression, including an emphasis on dysfunctional thinking styles, early trauma, etc. As in so many areas of psychiatry, practice has run ahead of theory.
One of the main consequences of this has been well highlighted by Turkington et al (2003), who express concern that CBT for psychosis now refers to a wide range of CBT treatments which vary in length and emphasis, and call for greater precision in identifying their active elements. They argue for further trials with better control groups and process measures to assist in this enterprise.
WHAT IS COGNITIVE – BEHAVIOURAL THERAPY?
Cognitive–behavioural therapy is a therapy for emotional disorders which has its own well-validated assumptions about what is responsible for those disorders (maladaptive cognitions), in the context of certain adverse life circumstances. At the heart of this is the link between thinking and emotion/behaviour – that emotional and behavioural responses are largely influenced by the cognitive appraisals that are made – and recent evidence suggests that cognition and emotion can mutually influence one another. In retrospect it is curious (and it seems to have gone unnoticed) that a therapy for (and theory of) affective disorder was considered appropriate for a ‘non-affective’ illness. It is even more curious that only a minority of trials of CBT for psychosis have used distress and emotional dysfunction as a secondary outcome, and only one has used it as a primary outcome. Is this still CBT? Or has CBT for psychosis strayed from its conceptual roots and become something else?
THE REMARRIAGE OF EMOTION AND PSYCHOSIS
Emotion and psychosis were divorced from one another in the middle of the 20th century, principally by Jaspers, who argued that we should separate ‘ affective illness from madness proper’ (Jaspers, 1963). Recent years have seen a renewed courtship (Birchwood, 2003; Freeman & Garety, 2003). First, there is well-documented evidence of the sheer scale of affective disorder in psychosis (unhelpfully referred to as ‘ comorbidity’), including depression, social anxiety and post-traumatic stress disorder (Birchwood, 2003). Second, factor-analytical studies of psychosis symptoms have revealed that depression is a distinct dimension of psychosis (Murray et al, 2005). Third, there is now strong evidence that the way in which people make sense of anomalous (primary) experiences such as voices provides the main causal pathway to distress and depression associated with these experiences (Birchwood et al, 2004), and the same is true for the way in which people react to the diagnosis (post-psychotic depression; Iqbal et al, 2000). In other words, distress and behaviour associated with psychotic symptoms may not always be ‘caused’ by the presence of psychotic experience per se, but rather they may result from the appraisal of it (e.g. its potential for threat). The primacy of distress is a core principle of CBT, but CBT for psychosis has not always followed this, given its primary emphasis on psychosis outcomes.
Emotion is now clearly implicated in the ontogeny of psychosis. Epidemiological studies have revealed that ‘neuroticism’ is a major risk factor for psychosis (Krabbendam et al, 2002), and there is evidence for an affective pathway to psychosis (Hanssen et al, 2003). In studies of the transition to psychosis in high-risk populations, or of the transition to relapse, it has been found that depression, anxiety and in particular social anxiety are among the strongest predictors (Owens et al, 2005). This interaction is complex and may involve cortical processes in affect regulation and common psychosocial developmental pathways (Birchwood, 2003). It will be crucial to gain an understanding of the nature of this interaction through laboratory studies and also through the action of CBT for psychosis.
As CBT for psychosis was developed further in the 1990s, the role and magnitude of emotional dysfunction in ‘nonaffective’ psychosis were not recognised. This resulted in a pragmatic application of CBT technique to delusions, and to assumptions being made about dysfunctional thinking and the genesis of psychosis.
COGNITIVE – BEHAVIOURAL THERAPY FOR (EMOTIONAL DYSFUNCTION IN) PSYCHOSIS
We believe that further large-scale pragmatic trials of CBT for psychosis, as currently designed (beyond those in progress), will neither shed further light on the active agents of CBT for psychosis nor initiate a process that will improve the effectiveness or specificity of CBT for psychosis. In fact they run the risk of doing the opposite. The next generation of therapy needs to focus on theory-driven studies of emotional dysfunction and/or behavioural anomaly in psychosis, including treatment studies which are themselves effective in ameliorating distress, but which may also have a secondary effect on the psychotic phenomena. Some possible foci are listed below.
CBT can be used to reduce distress, depression and problem behaviour associated with persecutory delusions and voices. For example, Trower et al (2004) demonstrated a reduction in compliance with command hallucinations and distress without a reduction in voice activity.
CBT can focus on anxiety, depression and interpersonal difficulty in individuals at high risk of developing psychosis. Morrison et al (2004) conducted a randomised controlled trial of CBT to prevent transition to psychosis in a high-risk group, and focused principally on these problems rather than on the attenuated psychosis symptoms that defined the high-risk group.
CBT can focus on the relapse prodrome to prevent relapse in psychosis. Gumley et al (2003) demonstrated a reduction in relapse as a result of working with the earliest (affective) signs of relapse and the way in which patients catastrophised them.
CBT can focus on ‘comorbid’ depression and social anxiety, including the patient’s appraisal of the diagnosis and its stigmatising consequences (Iqbal et al, 2000).
CBT can be used to reduce stress reactivity, thereby increasing resilience to life stress and preventing psychotic relapse (Myin-Germeys et al, 2005).
CBT can be used to increase self-esteem and social confidence in people with psychosis (Hall & Tarrier, 2003).
We believe that the future development and improvement of CBT for psychosis require us to move decisively away from the neuroleptic metaphor. Neuroleptics do what neuroleptics do. The future of CBT for psychosis lies in understanding the (cognitive) interface between emotion and psychosis and in developing interventions either to resolve emotional/behavioural dysfunction alone or to prevent or mitigate psychosis and its positive symptoms. Thus CBT can sit alongside the neuroleptics with a distinctive and complementary emphasis, rather than merely being brought on as a substitute in extra time.
- Received July 6, 2005.
- Revision received September 16, 2005.
- Accepted September 30, 2005.
- © 2006 Royal College of Psychiatrists