Have the benefits of cognitive behavioural therapy for psychosis been oversold – and the potential harm of it relatively ignored?
By Professor Keith R Laws, Department of Psychology, University of Hertfordshire
In 1952, the year that chlorpromazine arrived as the first antipsychotic medication for schizophrenia, Aaron Beck published a paper outlining a form of cognitive-based talk therapy to treat delusions. While chlorpromazine initiated the era of drug treatments in psychiatry, Beck’s psychological approach was a slow-burner. Eventually cognitive behavioural therapy (CBT), as we now know it, became part of mainstream treatment for psychosis in the UK in 2002 when the National Institute of Health and Care Excellence (NICE) endorsed it; and again in 2009 and 2014, when it recommended CBT be offered “to all people with psychosis or schizophrenia.”
But the past 12 months has seen CBT for psychosis (CBTp) become the focus of a heated debate about whether it has been oversold. In that time we have seen new clinical trials, critiques, counter-critiques, meta-analyses, a new NICE guideline and a large public discussion at the 50th Maudsley Debate.
A taste of the diversity of opinion can be garnered from the following quotes. Authors of the largest ever meta-analytic review of the literature recently concluded: “…the UK government’s continued vigorous advocacy of this form of treatment might be considered puzzling. [And] claims that CBT is effective against these symptoms of the disorder are no longer tenable.” (Jauhar et al, 2014)
Or from the recent review by the independent Cochrane organisation “…at present, it is not possible to assert any substantial benefit for cognitive behavioural therapy over other psychological therapies [in regard to schizophrenia].” (Jones et al, 2012)
By contrast, the recent British Psychological Society (BPS) document Understanding Psychosis and Schizophrenia (2014) argued that “[t]here is a consistent evidence base suggesting that many people find CBTp helpful.” And perhaps controversially, that “...there is general consensus that on average, people gain around as much benefit from CBT as they do from taking psychiatric medication.”
In the words of American theoretical physicist Richard Feynman: “Extraordinary claims require extraordinary evidence.” However, can we regard the evidence concerning CBTp for psychosis as extraordinary?
The shrinking impact of CBT
The ‘gold standard’ for evaluating any intervention comes from clinical trials where volunteers are randomly assigned to the intervention (e.g. CBT) or a control condition (e.g. treatment as usual) before the intervention begins. Following the intervention, the groups are compared to see if those who received the intervention had a better outcome. Using meta-analysis, researchers can then combine the data from numerous trials to quantify the average treatment effect.
Surprising to some, perhaps, is that CBT researchers have chosen to use it as a kind of pseudo-neuroleptic, targeting symptom reduction as their key outcome. These trials have witnessed an apparent shrinking of the therapeutic power of CBTp over the past 15 years. This reflects how the efficacy of CBT was initially exaggerated and subsequently reigned-in as better-controlled studies emerged. In practical terms, this shrinking effect means that 15 years ago, CBT treatment might have predicted a favourable outcome for one in three patients, but today this figure may be as few as one in 20.
Of course, one in 20 could be worthwhile, especially for those individuals whose symptoms have not responded well to antipsychotics. Nevertheless, the economic cost of such an intervention is necessarily high. From this viewpoint, other more readily available and cost-effective forms of psychological support might prove as effective as CBT.
Indeed, clinical trials show that CBT produces no better outcomes than supportive counselling and befriending (for meta analyses, see Lynch et al, 2010; Jones et al, 2012; Jauhar et al, 2014; Turner et al, 2014). Any change following CBT is therefore not specific to CBT but shared with other psychological approaches – a placebo response due to the attention received. Such findings would seem to mitigate in favour of exploring other forms of psychological support, especially if they can be delivered more widely, quickly and cheaply.
What are the benefits of CBT compared to psychiatric medication?
The efficacy of CBT has – almost exclusively – been examined in people who are currently taking antipsychotic medication. Nonetheless, as previously mentioned, the BPS document claimed that: “… people gain around as much benefit from CBT as they do from taking psychiatric medication.” Crucially, despite the audacity of this claim, it totally lacks support. No studies have yet compared CBT directly to antipsychotic drugs. Nevertheless, we do know that all of the most commonly used antipsychotics produce much greater symptom reduction than CBT.
Might CBT still be helpful to those who decide not to take antipsychotics? Evidence on this topic is restricted to just one trial, by Morrison et al (2014). At the end of the trial, symptoms were no better in those who had received CBT than in controls. Moreover, almost one in four who received CBT showed symptom worsening during the intervention, slightly higher than for the control group (17%) who, having declined medication, were often simply ‘discharged’.
The issue of possible worsening is explored further below. The ‘tolerability’ of CBT in such individuals is also questionable, as 40% dropped out during the nine-month intervention, rising to more than half by follow-up. Data from the recent National Audit of Schizophrenia covering CBT provision in all 64 NHS trusts in England and Wales shows that about 40% of people with schizophrenia are offered CBT and more than half decline it. We should not simply assume that a course of CBT is ‘easy’ or necessarily a more tolerable alternative to medication.
But surely if the evidence is that weak, why does NICE continue to recommend CBTp? Remarkably, three quarters of all published trials find that CBT does not reduce the symptoms of psychosis. Ironically, NICE’s own meta-analytic evidence has found no support for a reduction of either positive or negative symptoms – even when trials lasted the advocated 16-week period.
Given such surprising and overwhelming negative evidence, we might reasonably ask ‘how does CBT for psychosis survive as a recommended intervention?’ One possibility concerns the ‘wriggle room’ between data and how authors portray it. In relation to this issue, Marcus Muanfo and colleagues at the University of Bristol recently reported that CBTp trials consistently claim more favourable outcomes in their abstracts than their results would warrant.
In this context, it is worth noting that in 2014 NICE produced its latest guide, CG178 ‘Psychosis and schizophrenia in adults: treatment and management’, but neglected to update its earlier meta-analysis. This means current advice regarding CBT is based on trials conducted only up until 2008. NICE chose not to include any of more than 15 newer high-quality trials, the overwhelming majority of which have reported no significant impact of CBT.
In an unprecedented critique of the latest NICE guide, the chair of the SIGN committee, the Scottish equivalent of NICE, Professor Mark Taylor, said: “In our view NICE CG178 promotes some psychosocial interventions, especially CBT, beyond the evidence.
NICE CG178 also makes strong non evidence-based recommendations [and that] ...it is unfortunate that the guideline appears at times to reflect the interests of its authors rather than impartial up-to-date evidence.” This damning indictment has a wider resonance in terms of who should sit on such committees and the extent to which their evaluations provide reliable assessments.
Returning to the – neglected – possibility of negative consequences, while the evidence suggests little or no benefit for reducing positive symptoms per se, or more specifically for hallucinations (see Jauhar et al, 2014), the evidence on negative symptoms is potentially more disconcerting. Psychotherapy is traditionally seen as benign, but, in contrast to medications, it is far less frequently assessed for potential harms.
The absence of evidence for adverse effects in psychotherapy is not evidence of absence, as it has been tested far less rigorously. Adequate reporting of the benefits and risks of all interventions is crucial to facilitate clinical decisions, inform policymakers and, crucially, enable individual patients to make informed choices.
Currently, the monitoring of ‘harms’ in psychotherapy research falls behind that of psychopharmacology research. In their review of clinical trials published in high impact journals, Vaughan et al (2014) reported that harms were considered in only 60% of psychotherapy interventions compared to 100% of medication interventions. Similarly, in their review, Jonsson et al (2014) found that only 21% of psychotherapy trials published in 2010 mentioned harms, concluding: “the risk of reporting bias is nontrivial in conclusions about the risk-benefit ratio of psychological treatments.”
Returning to CBTp, a recent review by Velthorst and colleagues found that the negative symptoms of psychosis may worsen following CBT. Not one of more than 20 trials in the past 10 years has reported a significant reduction of negative symptoms following CBT – though most reported greater and indeed significant worsening of negative symptoms.
Additionally, six of eight studies have found greater relapse rates in people who have received CBT compared to controls (Lynch et al, 2010). It is crucial therefore to view CBT not just in terms of the potential benefits claimed, but also potential harms.
While CBTp remains the one-stop-shop government-endorsed psychological intervention for psychosis, we will be in denial of the current evidence base. CBT advocates have rightly been proud of their efforts to provide a sound scientific evidence base for clinical practice, but this area seems to lack some of that judgement. While CBT is prioritised, people will not be offered equally effective, cheaper and more available alternatives such as supportive counselling or befriending and the development of potentially more efficacious psychological alternatives will be hindered.
Beck AT (1952) Successful outpatient psychotherapy of a chronic schizophrenic with a delusion based on borrowed guilt. Psychiatry 15(3) 305-312.
Cooke A, ed (2014) British Psychological Society Division of Clinical Psychology. Understanding Psychosis and Schizophrenia. London: British Psychological Society.
Jauhar S, McKenna PJ, Radua J, Fung E, Salvador R & Laws KR (2014) Cognitive-behavioural therapy for the symptoms of schizophrenia: systematic review and meta-analysis with examination of potential bias. The British Journal of Psychiatry 204(1) 20-29.
Jones C, Hacker D, Cormac I, Meaden A & Irving CB (2012) Cognitive behavioural therapy versus other psychosocial treatments for schizophrenia. The Cochrane Library. 4 CD008712
Jonsson U, Alaie I, Parling T & Arnberg FK (2014) Reporting of harms in randomized controlled trials of psychological interventions for mental and behavioral disorders: A review of current practice. Contemporary Clinical Trials 38(1) 1-8.
Leucht S, Cipriani A, Spineli L et al (2013) Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple treatments meta-analysis. The Lancet 382(9896) 951–62.
Lynch D, Laws KR & McKenna PJ (2010) Cognitive behavioural therapy for major psychiatric disorder: does it really work? A meta-analytical review of well-controlled trials. Psychological Medicine 40(01) 9-24.
Morrison AP, Turkington D, Pyle M, Spencer H, Brabban A, Dunn G et al (2014) Cognitive therapy for people with schizophrenia spectrum disorders not taking antipsychotic drugs: a single-blind randomised controlled trial. The Lancet 383(9926) 1395-1403.
Munafo MR, Hunt DF, Rees NEJ & Laws KR (2014) Citation analysis reveals disproportionate emphasis on positive claims in study abstracts. British Journal of Psychiatry (eLetter). http://bjp.rcpsych.org/content/204/1/20.full#responses
National Institute for Clinical Excellence (2002) Schizophrenia. Core interventions in the treatment and management of schizophrenia in primary and secondary care. London: NICE.
National Institute for Health and Care Excellence (2009) Schizophrenia: core interventions in the treatment and management of schizophrenia in adults in primary and secondary care. (Clinical guideline 82)
National Institute for Health and Care Excellence (2014) Psychosis and schizophrenia: treatment and management. (Clinical guideline 178)
Perera U & Taylor M (2014) NICE CG178 on psychosis‐an evidence based guideline? Prescriber 25(9) 6-8.
Royal College of Psychiatrists (2014) Report of the National Audit of Schizophrenia (NAS) 2014. Healthcare Quality Improvement Partnership.
Turner DT, van der Gaag M, Karyotaki E & Cuijpers P (2014) Psychological interventions for psychosis: a meta-analysis of comparative outcome studies. American Journal of Psychiatry 171(5) 523–38.
Vaughan B, Goldstein MH, Alikakos M, Cohen LJ & Serby MJ (2014) Frequency of reporting of adverse events in randomized controlled trials of psychotherapy vs. psychopharmacotherapy. Comprehensive Psychiatry 55(4) 849-855.
Velthorst E, Koeter M, van der Gaag M, Nieman DH, Fett AK, Smit F, Staring BP, Meijer C & de Haan L (2015) Adapted cognitive– behavioural therapy required for targeting negative symptoms in schizophrenia: meta-analysis and meta-regression. Psychological Medicine 45(03) 453-465.