CBT: the received view

The ‘received view’ of CBT amounts to the claim that mental disorders such as depression result from faulty or dysfunctional cognition and behaviour. I take the core principles of CBT to be located in the work of Aaron Beck14 and published in manualised form by Judith Beck.15 For proponents of CBT, mental disorders are ‘thinking disorders’. On this view, faulty thoughts cause what Beck dubs maladaptive behaviours. The goal of CBT is to help the patient identify problematic automatic thoughts (what are termed ‘cognitive distortions’), and to help the individual challenge the validity of these thoughts. According to the received view depression arises because of faulty learning: cognitive therapy is an attempt to redress this purported maladaptive learning by helping patients to become ‘realistic’ in their thinking. It is contended that therapists engage patients in a collaborative and experimental process, the outcome of which is purportedly the promotion of healthy and realistic cognitive restructuring. The role of empathy is considered a means to an end in this process: without positive feedback and support, it is contended, the patient will not be sufficiently motivated to become engaged in the therapeutic process which is dubbed ‘collaborative empiricism’.9

This view of CBT is propagated in patient literature by leading health authorities in the UK and USA. For example, the NHS's ‘Health A–Z’ website on CBT explains that, “You and your therapist will look at your thoughts, feelings and behaviours to work out if they are unrealistic or unhelpful and to determine the effect they have on each other and on you”.16 Similarly, the NIMH explains that “The therapist helps the patient learn how to identify distorted or unhelpful thinking patterns, recognise and change inaccurate beliefs, relate to others in more positive ways, and change behaviours accordingly”.5 In the UK NICE Guidelines also advise that health professionals: “include an explanation of the CBT model, encourage tasks between sessions, and use thought-challenging and active monitoring of behaviour4”.

Common factors research

There are major problems with the scientific core of the received view of CBT. This should come as no surprise: its tenets were first formulated nearly 40 years ago.^v While this is not the place to review scientific evidence against the theoretical hard core of CBT we can survey solid scientific research aimed at explaining why psychotherapy is effective.

Research directly undermines the importance of the ‘cognitive’ components of CBT: it is estimated that most improvement takes place prior to the introduction of cognitive restructuring techniques.17 More than 300 studies have found that less than 1% of the variability in patient outcome is due to the specific techniques of distinctive treatments.18 ,19 ,20 An estimated 40% of improvement in psychotherapy is due to external or extratherapeutic factors (eg, changes in the patient's social environment).21 So what are the common factors responsible for the success of psychotherapy (including CBT)? Surprisingly, research consistently shows that therapists’ allegiance to a particular theory is more important than their adherence to treatment protocols.22 ,23 One estimate puts allegiance effects at around 10% of treatment outcome. Studies even show detrimental effects of adherence: Henry et al found that therapists who strictly adhered to treatment manuals were more likely to display negative attitudes such as less approval, less optimism and more authoritarian behaviour towards patients.23

The most significant factors in treatment outcome appear to be therapist-related. ‘The essence of therapy,’ it is claimed, ‘is embodied in the therapist’.9 Comparative studies of therapists who use the same treatment show that 30% of the variability in outcome is attributable to the therapist.24 In other words some therapists are highly effective while others are ineffective, even among experienced and well-trained practitioners.9 ,25 ,26 The ‘therapist effect’ undermines the received view of CBT where the role of therapist qualities are considered ancillary (‘necessary but not sufficient’)9 to the therapeutic process.27 But what exactly should we understand by the term ‘therapist effect’? Research partitions this into: (1) therapist factors such as degree of empathy, encouragement, trust and unconditional positive regard for the patient; and (2) agreement between therapist and patient about the goals of therapy (often dubbed ‘the therapeutic alliance’ or ‘the alliance’). In respect of the latter, alliance is said to encompass the degree of collaboration and commitment to those goals by patient and therapist. It is proposed that agreement about goals helps to harness patient compliance. Indeed, numerous studies show that early assessments of therapeutic alliance between patient and therapist predict outcome with up to 65% of symptomatic relief occurring within one to seven visits.28 ,29 Moreover, contra conventional wisdom, patients are better judges of therapeutic alliance than therapists.30,^vi

Finally, the patient's expectation that the treatment will be effective is also considered to be a key component in outcome. The placebo effect has been defined as an expectancy response whereby the expectation that a treatment will be effective elicits a self-fulfilling therapeutic benefit in the patient.31 The placebo effect has been found to be a significant therapeutic effect for a range of conditions (including depression, anxiety and chronic pain).32 In short, the placebo effect is no less relevant in psychotherapy than in pharmacological interventions.33 Ascertaining the size of the placebo effect in psychotherapy is notoriously problematic not least because double-blinded placebo psychotherapy is not experimentally feasible.33 Nonetheless expectancy can be assessed without problem and studies show that treatments with cogent rationales are more effective than psychotherapy interventions that have no explanation.

Informed consent guidelines

Informed consent procedures for professional organisations such as the British Association for Counselling and Psychotherapy (BACP), the American Psychological Association, are commensurate with General Medical Council (GMC), American Medical Association (AMA) and American Psychiatric Association guidelines. Just like doctors, psychotherapists who are members of organisations such as the BACP and American Psychological Association are duty bound to respect patient autonomy and to provide adequate informed consent to patients. The BACP stipulates, “Clients should be adequately informed about the nature of the services being offered. Practitioners should obtain adequately informed consent from their clients and respect a patient's right to choose…”34 Similarly, the American Psychological Association specifies that psychologists should, “[O]btain the informed consent of the individual”35 and, “inform clients/patients as early as is feasible in the therapeutic relationship about the nature and anticipated course of the therapy”.36 It also stipulates: “Psychologists should seek to promote accuracy, honesty, and truthfulness in the science, teaching, and practice of psychology”. Doctors are required to tell the truth to patients and to obtain informed consent before undertaking any medical intervention: in the UK the GMC asserts that physicians must obtain “consent or other valid authority”;37 physicians must “discuss with patients what their diagnosis, prognosis, treatment, and care involve”; in the USA, the AMA states that, “withholding medical information from patients without their knowledge or consent is ethically unacceptable”;38 and “The physician has an ethical obligation to help the patient make choices from among the therapeutic alternatives consistent with good medical practice”.39 Finally, the American Psychiatric Association stipulates that, “A psychiatrist shall not withhold information that the patient needs or reasonably could use to make informed treatment decisions”.40

How should we understand these ethical dicta? Beauchamp and Childress usefully dissect informed consent into three phases: (1) preconditions for informed consent (competence and voluntariness of the patient): (2) information (the disclosure material, recommendations, and the patient's understanding of material and recommendations); and (3) consent elements (the patient's decision and its authorisation).41 This paper focuses on the second cluster of elements since this is most relevant to information disclosure in CBT.

Consider the disclosure of material information. Decisions about what to disclose stems from one fundamental issue: what information is relevant in order for a patient to decide whether to consent to or refuse a treatment? Notice that relevant disclosure need not provide an in-depth account of every material fact about CBT. Disclosure clearly cannot be evocative of Lewis Carroll's cartographer whose maps had ‘the scale of a mile to a mile’. Instead, and as a heuristic, practitioners must provide germane information according to what Beauchamp and Childress refer to as a ‘reasonable person standard’: the facts relevant for a reasonable person to consent to a treatment and to make well-informed choices about that treatment.41 Furthermore, as Beauchamp and Childress point out, failure to provide representative information through omission, or ignoring patients’ false beliefs about treatment, can be said to undermine the disclosure process.

The failure to uphold ethics guidelines

CBT therapists (and primary care doctors at the referral stage) are not meeting their own ethical standards of material disclosure: NHS, NICE and NIMH guidelines on CBT propagate the received view much as it appears in Judith Beck's manualised account: the goal of the initial session we are told is ‘Educating the patient about her disorder, about the cognitive model, and about the process of therapy’.15 We do not know if CBT therapists are disclosing anything more nuanced and in line with advanced scientific research—the indication that they update their disclosures is not promising given these guidelines.

However, it might be argued that the disclosure accurately describes the procedure of CBT—that the information presented to patients is accurate, even if the outcomes are not due to these specific procedures.^vii In response, it not clear that this information does accurately describe the procedure—it may be a description of the intentions of therapy but: (1) in practice therapists may routinely stray from rigidly adhering to these techniques;22 and (2) (more importantly) we can contest whether it is even possible to implement these techniques—the phenomenon of depressive realism shows that mild–moderate depressive thoughts are realistic even if they are unhelpful (see footnote^v).42 In short, the debate about informed consent in psychotherapy proceeds with the assumption that the various and distinct models of psychotherapy are correct on their own terms. For example, a report on informed consent for the National Register of Health Service Psychologists states that,[D]epending on their treatment modality psychologists should provide clients with information about the overall approach they will use to treat the presenting problem, and likely techniques that the approach may entail e.g., exposure therapy, dream analysis, detailed developmental history, conjoint family sessions, behaviour contracts, or any other information relevant to making an informed decision to engage in treatment.43

This involves the promulgation of equivocation: it allows patients to assume that the received view of CBT can be taken at face value. Indeed, the report notes—without self-consciousness—“In addition, some therapists choose to inform clients of the empirical evidence guiding their treatment choice”. This ambiguity—the notion that some therapists choose to provide scientific support for their therapy but others can (by intimation) choose not to justify their treatment choice inverts the very rational for disclosure. Informed consent procedures are not about tailoring information to the therapist but to the patient. Therapists are therefore failing to inform patients about common research factors: (1) the importance of the patient ‘getting on board’ with the goals and aims of the treatment; (2) the importance of therapist attributes (such as empathy and support); and (3) that the patient is the best judge of the quality of the bond between the therapist and him or herself.^viii

Do these failures to meet informed guidelines constitute an ethical failing in the clinical practice of psychotherapy, and specifically, CBT?

Beneficence versus autonomy: a false dilemma in CBT?

The challenge in informed consent is to disclose to the patient information that is pertinent to treatment choice and autonomous decision-making. It might be argued, however, that therapists do indeed provide adequate information to their patients—furthermore, regardless of the conventional (yet pseudoscientific) explanation for how CBT works there appears to be scientific evidence that it does work. Isn't this science enough for maintaining the status quo when it comes to informed consent? Thus, if a patient asks, “How does CBT work?” and the therapist responds, “In these sessions we use various techniques that will help you to overcome your depression” it might be claimed that such a response is ethically justified on the grounds that sufficient information has been disclosed to warrant the designation ‘informed consent’. On this line of reasoning, the various techniques in CBT form the ‘vehicles’ for treatment and it is the collaborative agreement over these techniques that carry the causal therapeutic effects: thus, it might be argued, the therapist has met the requirement of respecting patient autonomy.

Take the first claim: the claim that CBT is effective is not a robust defence for disclosing false theories to patients. Indeed, it might be contended that patients with CBT would fare even better (regardless of the ethical issue of respecting individuals’ self-determination) if they were equipped with adequate information. Moreover, the fact that a treatment works does not license doctors or therapists to disclose whatever they choose in respect of that treatment. Second, in response to the claim that therapists merely describe the vehicles of treatment which (it is argued) transmit the causally effective factors in CBT, we might retort that (at best) this is something of a sleight of hand. Certainly, the therapist has thereby described the intended practice of therapy but they have not answered the relevant question of the causal explanation for the effectiveness of therapy. We can concede that the plausibility of the form of therapy to the patient (including the techniques involved) is relevant to outcome but it is this factor that needs to be conveyed to patients. It is not sufficient to say, ‘CBT works because of cognitive restructuring techniques aimed at changing unhelpful or faulty cognition’; rather, truthful disclosure would involve: ‘CBT uses techniques based on a theory about cognitive restructuring. But the evidence shows that it works best for those patients who can readily commit to the theory and techniques behind the treatment.’ Indeed, as noted, this is one factor among others that patients should be informed about since research shows that the therapeutic alliance is the most significant factor in outcome and patients appear to be the best judges of this.

It might be countered, however, that current disclosure policies maximise the therapeutic effect of treatment. On this view, the duty to be open and honest about treatment clashes with beneficence. Some might consider the evidence for the significance of patient expectations about particular treatments and the cogency of the rationale of the therapy for patients as supporting the idea that truthful disclosure would significantly undermine the ‘magic’ of treatment.

But this stance ignores other findings that may support the opposite conclusion. For example, although open placebo studies are still in their infancy they^ix have begun to yield promising results which suggest that practitioner deception may not be necessary to elicit placebo effects: similarly, it may be that the effectiveness of CBT will not be undermined by improved disclosure. Furthermore, we already know that some therapists put a misplaced premium on the techniques of CBT in their adherence to treatment (of which more, later). Thus, it is certainly possible that the impediment of informed consent may even produce negative health consequences for some patients. So far empirical studies have found: “[C]lients approach therapy with expectations regarding the nature of therapeutic interview and the roles they and their counsellors/therapists will assume...[C]lients’ expectations may either facilitate or hinder the communication process and the process and the effectiveness of therapy. For instance, expectancies influence the choice of help giver...Expectancies may be related to persistence in therapy, and a client's decision to discontinue therapy”.44 It may be, for example, that if there is lack of progress patients erroneously blame themselves for the failure of CBT to work. If (as a consequence) they drop out of therapy the outcome may be clinically harmful and it may negatively affect patients’ future trust in therapy, therapists and even referring doctors. Patients may come away with the false impression that ‘psychotherapy is not for me’ rather than the clearer perception that (for example) CBT may not have been the best treatment for them, or even that their therapist may not have been suitable for their particular needs.^x Thus, the claim that CBT is effective and that this is sufficient defence for upholding current practices can be challenged: again, CBT may have the potential to be even more effective for more people, when aligned with adequate informed consent policies.

But is informed consent too tall an order for therapists? Does it place impossible demands on clinicians in requiring them to relay scientific research on psychotherapy to patients? It might even be argued that professional psychotherapists should not be expected to grasp the science behind psychotherapy. This defence of the status quo is unconvincing. Relevant information about the nature of the treatment can certainly be representative of current scientific knowledge without being exhaustive. This need not be an onerous demand: indeed, it is a standard expectation for other treatments in orthodox medicine. Furthermore, when it comes to training psychotherapists we might add that the science of psychotherapy is not more challenging than CBT theory. Indeed, we might rhetorically turn the criticism on its head and respond: should we expect psychotherapists to ignore the evidence base for their professional activities, because: (1) therapists might not understand the evidence; and/or (2) it is too much effort to present the science in a digestible, understandable manner to trainees? Deceiving therapists about current scientific understanding about their treatments may even jeopardise the future status of the profession by risking distrust and scepticism among therapists, and ultimately patients. As we've seen therapists’ allegiance to a form of therapy is more effective than their adherence to the protocols of that therapy (and some evidence shows that adherence is detrimental to patients): therefore, the justification for inculcating misinformation to therapists as a means to a beneficent end for patients is risky and (on the grounds of evidence) unpersuasive. Indeed, on consequentialist grounds, we might reach the opposite conclusion: previous controversies such as the landmark informed consent case Osheroff versus Chestnut Lodge (1980) arguably contributed to the demise and diminished reputation of psychoanalysis as a preferred treatment option.^xi ,45 Lessons need to be heeded lest CBT (and other effective versions of psychotherapy) are challenged in the future on the basis of research into the science of psychotherapy and the ways in which this challenges informed consent.