Understanding Psychosis and Schizophrenia: A response to Jauhar and Morrison
Anne Cooke and Peter Kinderman
This is a response to a critique by Sameer Jauhar and Paul Morrison of a public information document published by the British Psychological Society, Understanding Psychosis and Schizophrenia: why some people hear voices, believe things others find strange or appear out of touch with reality, and what can help (Cooke, 2017) (Jauhar & Morrison, 2017). The report is a revised version of one originally published in 2014 (Cooke, 2014): for a selection of other reviews see www.understandingpsychosis.net. We welcome Sameer and Paul’s reflections: one of our aims in writing the report was to open up discussion and debate about this important topic.
Sameer and Paul’s blog focuses on ‘the revised sections, pertaining to definitions, aetiology and treatment’. It is not clear on what basis they decided that these were the revised sections. In fact, as stated both inside the front cover of the report and on the download web page, the only changes compared to the original version relate to the effects of racism and inequality. Following the launch of the first version in 2014 it was pointed out that the report had engaged insufficiently with scholarship regarding the experiences of people from black and minority ethnic communities in relation to psychosis. The amendments in the revised version address these issues and also the relationship between psychosis and social inequality more generally (see Cooke, 2016). In other respects, (including most of the material that Sameer and Paul discuss) the report is unchanged. As also stated, critiques pertaining to other issues will be taken into account in preparing future editions.
Having clarified that, we will now quote and respond to Sameer and Paul’s points relating to each of the three areas that they highlight, and also to their comments on the report’s overall message and implications for practice. It is worth noting that the passages to which they refer comprise a very small proportion of the whole report. We are, however grateful to them for opening discussion and for creating an opportunity for us to respond to some misunderstandings about the report which may not be unique to them.
Definition and nature of psychosis
- Over-inclusive definition of psychosis
Sameer and Paul state: In defining psychosis widely, it becomes difficult to comment on current practice. The authors state the prevalence of hearing voices is 10%, a recent WHO study finding this nearer 5% the authors acknowledge that what is picked up on screening questionnaires for psychotic experiences do not readily overlap with what a diagnostic interview for schizophrenia picks up (Kendler et al., 1996; Ochoa et al., 2008). In other words, they are defining a population of people with “psychosis” who do not meet conventional criteria, and appear to have over-estimated the prevalence…. Symptoms given weight in the document, auditory hallucinations and suspiciousness are seen in 11.6% and 25.2% of people with diagnosed schizophrenia respectively, two years after acute exacerbation (and about 50% in acute illness), which would hardly make them pathognomonic of diagnosed schizophrenia (WHO, 1979). Murray and van Os have made the astute observation that: “the boundaries between normal mentation, common mental disorder and schizophrenia become blurred, if positive psychotic symptoms are used as a distinguisher” (van Os and Murray, 2013)…. The authors are apparently describing people not seen by mental health services; a point they acknowledge….It is therefore unclear how they can comment on care provided to people within current services; a different population to what they define.
Here the argument seems to be that we are not talking about ‘real schizophrenics’. Sameer and Paul are right – one of the messages of the document is that there is no underlying ‘thing’ or disease called schizophrenia that causes these experiences; these are just labels we give to particular experiences which some of us have more of, or experience more frequently or intensely than others. So it is not a question of separating out ‘real schizophrenia’. This is a concept which many people, immersed in the traditional idea of mental illness, find it hard to understand. For example, workers sometimes say to us: ‘Come to the ward where I work and I’ll show you schizophrenia exists’. We have each worked on both acute and long-stay mental health wards, and are only too aware that some people have multiple complex and severe problems and need ongoing help and support. The difference is that for reasons outlined in the report, we do not think that it is scientifically or ethically justified to insist that those problems are ‘positive and negative symptoms’ of a mental illness called schizophrenia. It is, of course, true that people in receipt of different forms of care (or none) are likely to have different problems, but this is not the same as assuming that conventional criteria define an underlying illness entity distinguishable from normality. One of the central theses of the report is that these conventional lines of demarcation do not necessarily reflect valid ‘natural forms’. It is a thesis with which Sameer and Paul, by merely repeating these definitions, do not really engage.
That said, we disagree with their argument that the report refers to a different population to those seen within current services. Indeed, there has been an overwhelming response from current and former service users who feel that the report speaks to precisely the problems they have experienced. Some of the report’s authors had themselves been in receipt of mental health services for long periods and had been diagnosed with schizophrenia.
- Ignoring negative symptoms and disabling psychosis
The blog states: Crucially, symptoms omitted are seen in people with the highest level of impairment, who would fulfil criteria for treatment-resistant psychotic illness; the people for whom secondary services were originally set up. One of the paradoxes in modern mental health services, is that there is under-allocation of resource to those suffering from the most disabling psychiatric syndromes, such as negative syndrome schizophrenia, and a shift of resource to the detection and probable over-treatment of relatively mild psychological issues (de Leon, 2014).
This is the argument often made by American psychiatrist Allen Frances: Anne outlined her reasons for disagreeing with it in a co-authored blog (Frances & Cooke, 2014) with him. We acknowledge that some people (for example an anonymous commenter on Sameer and Paul’s blog) feel that the report de-emphasises the severe and ongoing nature of some people’s problems, and their need for support. It is true that the report focuses on the possibility of recovery: one of our aims was to counter the negative and hopeless messages that people often receive both from services and from available public information about ‘schizophrenia’. However we were also at pains to point out that many people have ongoing, severe and complex problems (including those traditionally thought of as ‘thought disorder’ or ‘negative symptoms’) and need significant support. This is a difficult balance to get right, and we apologise to anyone who feels that we got it wrong. That said, it is also important to reiterate that disputing one particular interpretation of problems (e.g. that they are ‘positive and negative symptoms of schizophrenia’) is absolutely not the same as denying they exist. The report argues that each of us develops our unique set of characteristics and problems for a unique and complex set of interacting reasons. Some of us, for a variety of reasons, experience more, or more severe, problems than most and may need long term support. Ironically, for many people the reasons for ongoing disability can include, amongst other things, iatrogenic effects of drugs and institutionalisation, and the hopelessness induced by being told you have a progressive illness called schizophrenia. This is a very different idea to Sameer and Paul’s hypothesis that the explanation for a heterogeneous range of difficulties lies in an underlying illness with positive and negative symptoms. As we acknowledge in the report, that idea is useful for some and has some advantages: within our current social system, drawing on an illness model can help those affected to access support, for example. However, we think it is important to acknowledge both that it is only one of a number of possible frameworks of understanding, and that as well as some advantages it has significant drawbacks (see e.g. Cooke, 2008; Cooke & Kinderman, in press, Cooke & McGowan, 2013) and has been subject to significant challenge scientifically, as outlined in Section 3 of the report.
- Ignoring ‘bipolar psychosis’
The blog states: There is no mention of bipolar psychosis, which has differences in aetiology, natural history and treatment response to non-affective psychoses. It also has a high degree of reliability (Tolin et al., 2016). By ignoring this, and lumping “psychosis” together, they ignore the increased relapse rate in this, as well as greater number of hospitalisations (Chang et al., 2016).
The blog states: They (the Understanding Psychosis authors) state half of people who experience psychosis will experience problems only once, and recover completely, quoting a review which cites 9 studies….The review stating 50% of people have recovery at follow-up is not a systematic review. A more recent systematic review suggested that only 13% of people with a diagnosis of schizophrenia met the criteria for recovery (Jääskeläinen et al., 2013). Long-term follow-up of people presenting with first episode psychosis show significant heterogeneity in symptomatic and functional outcome: a multi-centre UK first-episode study showing, in geographically defined populations, that 50% were in symptomatic remission at 10 years (a number of whom were taking antipsychotics), though only 15% fulfilled criteria for functional recovery (Revier et al., 2015).
Any report must be updated as new research becomes available. As outlined above, the current revision related only to the effects of racism and inequality: future editions will update other aspects. The research to which Sameer and Paul refer (which has itself been subject to critique: see Slade & Longden, 2015) is part of a wider and growing literature: other studies are more optimistic (e.g. Lally et al, 2017). The different figures are likely to be a reflection of the overall variability and also of the different ways it is possible to operationalise and measure outcome. Sameer and Paul’s overall coverall conclusion, that there is “significant heterogeneity in symptomatic and functional outcome” is consistent with ours. Whilst it is important to be mindful that some difficulties can be lifelong and life-changing, it is equally important not to misrepresent the available literature and risk iatrogenic harm by telling people that they have a brain disease with a poor prognosis.
Aetiology (Reasons that people might experience psychosis)
The blog states: Causes are limited to social deprivation (inequality) and life events (racism, discrimination and childhood trauma). The authors ignore aetiological factors, including head injury, cannabis, and psychoses caused by general medical conditions, e.g. epilepsy, and autoimmune conditions. Missing an organic cause of psychosis constitutes a serious clinical error. The evidence for discrimination is based on a preliminary report, and findings of delusional-like ideas in 7 people, with perceived discrimination, using the CIDI, which has poor psychometric properties compared to diagnostic interview (Jauhar and Lawrie, 2011). The authors state: Experiencing multiple childhood traumas appears to give approximately the same risk of developing psychosis as smoking does for developing lung cancer. They cite a study where items on a screening questionnaire correlated with recall of adverse events, with an odds ratio of 11 (Bentall et al., 2012).It is mathematically impossible to explain how the factors they give could contribute to the prevalence they mention; the odds ratio for childhood trauma is between 2-4 in most meta-analyses (Morgan and Gayer‐Anderson, 2016) (with methodological limitations). The odds ratio for smoking and lung cancer varies depending on smoking status and cell type, anywhere up to around 100 (Pesch et al., 2012) and evidence they put for a similar association is based on items from a screening questionnaire that measures phenomena that are not related to what people conventionally call psychosis. Even then, the odds ratio of 11 does not sit well with odds given for current smokers and lung cancer. Put simply, not everyone who experiences psychosis (however defined) has these risk factors, and not everyone who has these risk factors develops psychosis.
The statement that causes are limited to social deprivation and life events is inaccurate. Whilst one section (Section 6) addresses social factors, there is a parallel section (Section 5) on the role of biology including genetics, neurochemistry, and brain structure and brain function. It is acknowledged that these can play a significant role. Particular mention is made of their clear importance ‘when someone has a psychotic experience when they are physically ill or have taken particular drugs’ (page 37). Section 7 addresses psychological factors.
For a discussion of the report’s comparison of the risk of psychosis in the context of childhood trauma with the risk of lung cancer in the context of smoking, we refer the reader to the comments that our co-author Richard Bentall has already made.
With respect to the role of discrimination and particularly of racism, the report covers a number of sources of evidence, as a search of it using those terms will reveal. The importance of racism both as a potential contributory factor, and in the poor deal that young black men in particular often receive from services, is important to acknowledge.
We agree with Sameer and Paul’s last sentence pointing out that there is no 1-1 correlation between particular adverse events and psychosis. Indeed, the complex and interacting nature of the reasons that people might have particular psychotic experiences is a major and important message in our report, and one which contrasts with the traditional approach of looking for ‘the cause of schizophrenia’.
- Comparison of medication and psychological therapies
The blog states: In the section on psychological therapies they state: On average, people gain around as much benefit from CBT as they do from taking psychiatric medication. The authors cite various meta-analyses, but the effect sizes given are not comparable to those for psychotropic medication, especially when considering effects of blinding, even taking into account criticisms regarding included studies (Jauhar et al., 2014), see below….No comment is made on relapse, the best-conducted study showing no benefit (Garety et al., 2008), which contrasts to evidence for antipsychotics (see below)….The section on medication has factual errors….They state meta-analyses: “suggested that many people experience only slight benefits and only about 20 percent experience a significant improvement or prevention of reoccurrence”. The citation was not available, though the paper we assume they are referring to has caveats to this proclamation, and the article also reports reasonable effect sizes for functioning and quality of life (Leucht et al., 2017a). This refers primarily to people with longer-term illness with acute exacerbation, as opposed to first-episode illness. In reference to the statement above, the effect size is still more than twice that for CBT for overall symptoms in blinded trials, and five times that for positive symptoms (Jauhar et al., 2014)….Regarding the 20% figure regarding relapse, meta-analysis by the same authors indicating 27% versus 64% of people had a relapse at one year after medication discontinuation, a relative risk of 0.4 (Leucht et al., 2017b).
Peter has published a detailed explanation of the comparison between the effectiveness of CBT and antipsychotic medication elsewhere (Kinderman, McKenna & Laws, 2015). Briefly here we note that Sameer and Paul cite only one meta-analysis of CBT for psychosis, namely one authored by Sameer. Other meta-analyses have drawn different conclusions (see Wykes, 2014).
Moreover, CBT is not the only form of psychological therapy. The report discusses several other forms of psychological intervention, for example family work. Again, NICE recommends that everyone diagnosed with psychosis should have access to family interventions, although unfortunately this is far from being the case.
It is also wrong to see the benefits of psychological therapies as being the main message of our report. Its central message relates to the need for a change in the fundamental framework of understanding which guides every conversation within mental health services, not just those formally designated as ‘therapy’. The approach outlined in the report has profound implications for the way that we design services and also for our efforts to prevent psychological distress in the first place.
- Discontinuation of medication
The blog states: The authors are selective in quoting two observational studies which suggest improved outcome where medication was either stopped naturalistically (Harrow and Jobe, 2013) or decreased (Wunderink et al., 2013), whilst ignoring in this context an observational study showing poorer outcome in never treated people with schizophrenia followed up after 14 years (Ran et al., 2015), and a national register study showing longer life expectancy for people with schizophrenia taking antipsychotics (Tiihonen et al., 2009). They do not mention in citing the Wunderink study that at 18 months relapse rates were doubled in the discontinuation group, and only 20% of people actually discontinued medication at 18 months (Wunderink et al., 2007). Furthermore, in the study they cite, people in the discontinuation group were still taking an efficacious, albeit lower dose of antipsychotic than the other group (Wunderink et al., 2013).
We stand by our interpretation of the literature base. There is not space here to outline all the relevant debates and (rapidly accumulating) research to which we would wish to draw readers’ attention: for a fuller critical examination see e.g. the work of Robert Whitaker (e.g. Whitaker, 2010; 2011) and Joanna Moncrieff (e.g. Moncrieff, 2013). It is also worth reiterating that the report is not ‘anti-medication’: we acknowledge that many people find medication helpful. What we are advocating is openness on the part of clinicians that (to quote the report) “Neuroleptics can only be used pragmatically – trying a particular medication and seeing what happens. We need to be honest and take a pragmatic (‘suck it and see’), collaborative approach, talking through options and enabling the person to try different things to see what helps” (page 98).
We would agree that withdrawal from antipsychotic medication is often a difficult process and it can be difficult to distinguish discontinuation effects from a re-occurrence of the original difficulties (‘relapse’). It will be interesting to see the results of the current RADAR trial (Moncrieff, 2016) which directly compares ‘maintenance medication’ with a gradual and supported programme of antipsychotic reduction.
The blog states: The main conclusions drawn are that conventional concepts of psychosis are flawed, with inadequate provision of psychological therapies for people with what is termed “psychosis”, services needing to change.
This is broadly correct, and we stand by these conclusions. We believe that the idea currently popular within Western mental health services that psychotic experiences are symptoms of an underlying mental illness requiring medication, is open to question. We also state that ‘services need to change radically’: Section 13 outlines the changes we would like to see happen. Whilst we also we also suggest increasing provision of psychological therapies, as we have stressed above this is not the report’s main message. The main change we are advocating is much more fundamental than providing more of x or less of y, and it would be wrong to read the report merely as clinical psychology ‘touting its wares’ (see Cooke, 2016 for a first-person account of what motivated the authors to volunteer our time to this project). The report argues for a fundamental change in the guiding ideas that drive both society’s and services’ response to the experiences we sometimes call psychosis:
“Mental health is a contested area. The experiences that are sometimes called mental illness, schizophrenia or psychosis are very real. They can cause extreme distress and offering help and support is a vital public service. We know something about the kinds of things that can contribute to these experiences or cause them to be distressing. However, the causes of a particular individual’s difficulties are always complex. Our knowledge of what might have contributed, and what might help, is always tentative. Professionals need to respect and work with people’s own ideas about what has contributed to their problems. They need to take into account the person’s background and culture, and consider whether their particular life experiences and circumstances (for example poverty, trauma, racism, or homophobia) may have contributed to their difficulties. Some people find it helpful to think of their problems as an illness but others do not. Professionals should not promote any one view, or suggest that any one form of help such as medication or psychological therapy is useful for everyone. Instead we need to support people in whatever way they personally find most helpful, to ensure that the help offered is culturally sensitive and accessible to everyone, and to acknowledge that some people will receive support partly or wholly from outside the mental health system” (Cooke, 2017, p.103)
Implications for practice
The blog states: This document gives theoretical views on what constitutes psychosis and what a group of influential senior psychologists think would be helpful for people who experience the phenomena they define. However, given the definitions used, selective citing of evidence regarding aetiology and treatment, it is unclear how (or why) this guidance should be implemented within current services. It ignores a significant number of people currently seen in secondary services, and by selectively citing literature that corroborates their views, and making errors throughout, it is difficult to know how the conclusions drawn are justified, on the basis of evidence.
It is important to address directly the suggestion that this was a report written by “a group of influential senior psychologists”, directing advice at a separate group of people “who experience the phenomena they define”. One quarter of the 25-strong authorship team had personal experience of psychosis. One of them (Jacqui Dillon) has described the writing of the report as ‘a wonderful example of moving beyond them and us, in process and practice ’. We welcome Paul and Sameer’s recent acknowledgement that they were mistaken in this regard. Quite apart from the ethical consideration of ‘nothing about us, without us’, we believe that in the arena of human experience, it is very odd for one group of people to pronounce on the nature and meaning of others’ experiences without involving them in that conversation. For a fuller discussion of this issue, readers are referred to the resources compiled by the Mad Studies Network.
The report did not claim to be a systematic review of the literature (see Cooke, 2016). Nevertheless, we stand by our scholarship. Like popular science books, public information reports such as ours draw on the available literature in order to offer accurate and useful information to the general reader. Scholars and clinicians both have to make judgements as to how to select, interpret and apply relevant research from the vast quantity available. A major conclusion from our report is that the illness model of mental health is only one way to look at the available evidence. As is inevitable, not everyone agrees with all our interpretations or conclusions. That does not necessarily mean that we are mistaken. It means that there is a very real and important debate to be had.
Cooke, A. (2008). Problems associated with the use of the concept of mental illness. In T. Stickley & T. Basset (Eds.), Learning about mental health practice (pp.329–346). Chichester: John Wiley & Sons Ltd.
Cooke, A. (2016) Changing society’s whole approach to psychosis. Journal of Mental Health, 25 (4). pp. 287-290. ISSN 0963-8237.
Cooke. A. (ed) (2014). Understanding Psychosis and Schizophrenia: why people sometimes hear voices, believe things other people find strange or appear out of touch with reality, and what can help. A report by the British Psychological Society Division of Clinical Psychology. Leicester: British Psychological Society.
Cooke. A. (ed) (2017). Understanding Psychosis and Schizophrenia: why people sometimes hear voices, believe things other people find strange or appear out of touch with reality, and what can help. A report by the British Psychological Society Division of Clinical Psychology. Revised Version. Leicester: British Psychological Society.
Cooke, A. & Kinderman, P. (in press) “But what about real mental illnesses?” Alternatives to the disease model approach to ‘schizophrenia’. Journal of Humanistic Psychology.
Cooke, A. & McGowan, J. (2013) Is Life a Disease? (podcast) Discursive of Tunbridge Wells. https://blogs.canterbury.ac.uk/discursive/is-life-a-disease/
Frances, A. & Cooke (2014). A. Pro and Con: The British Psychological Society Report on Psychosis. Huffington Post, 15th December.
Jauhar, S. & Morrison, P. (2017) #Understandingpsychosis? Mental Elf, 30 October.
Jauhar S, McKenna PJ, Radua J, et al. (2014). Cognitive-behavioural therapy for the symptoms of schizophrenia: systematic review and meta-analysis with examination of potential bias. Br J Psychiatry 204(1):20–9.
Jones, S., Lobban, F. & Cooke, A. (eds.) (2010). Understanding Bipolar Disorder: Why some people experience extreme mood states and what can help. A report by the British Psychological Society Division of Clinical Psychology. Leicester: British Psychological Society
Kinderman, P., McKenna, P. and Laws, K. R. (2015), Are psychological therapies effective in treating schizophrenia and psychosis?. Prog. Neurol. Psychiatry, 19: 17–20.
Kingdon D. (2014) Has cognitive behavioural therapy for psychosis been oversold? BMJ 348: 2295.
Lally, J., Ajnakina, 0., Stubbs, B., Cullinane, M., Murphy, K., Gaughran, F. & Robin Murray, R. Remission and recovery from first-episode psychosis in adults: systematic review and meta-analysis of long-term outcome studies. The British Journal of Psychiatry Oct 2017, bjp.bp.117.201475; DOI: 10.1192/bjp.bp.117.201475
Moncrieff, J. (2013) The Bitterest Pills: the troubling story of antipsychotic drugs. Palgrave Macmillan. http://www.palgrave.com/products/title.aspx?pid=638387
Moncrieff, J. (2016). New research into antipsychotic discontinuation and reduction: the RADAR programme. Blog post, 2nd March. Retrieved from https://joannamoncrieff.com/2016/03/02/new-research-into-antipsychotic-discontinuation-and-reduction-the-radar-programme/ .
Morrison AP, Turkington D, Pyle M, et al. (2014). Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. Lancet 383(9926):1395–403.
National Institute for Health and Care Excellence (2014). Psychosis and schizophrenia in adults: treatment and management. NICE clinical guideline CG178. London: National Institute for Health and Care Excellence, 2014. 9.
Schizophrenia Commission (2012). The abandoned illness: A report from the Schizophrenia Commission. London: Rethink Mental Illness, 2012. www.rethink.org/mediareport_14_nov.pdf
Slade, M. & Longden, E. (2015) Empirical evidence about recovery and mental health. BMC Psychiatry, 15:285.
Whitaker, Robert (2011). Anatomy of an Epidemic. Crown Publishing Group. ISBN 978-0-307-45241-2.)
Whitaker, Robert (2010). Mad in America: Bad Science, Bad Medicine, and the Enduring Mistreatment of the Mentally Ill. New York: Basic Books. ISBN 0-465-02014-3.
Wykes, T. (2014) Cognitive-behaviour therapy and Schizophrenia. Evidence Based Mental Health 17: 67-68. Doi : 10.1136/eb-2014-101887.