Compared with mainstream medicine and complementary and alternative therapies, the practice of psychotherapy has enjoyed a relative pass when it comes to ethical evaluation. Therefore, contributions to the, although slowly growing, body of literature on psychotherapy ethics are to be welcomed. In his paper ‘Psychotherapy, placebos, and informed consent’, Garson Leder takes issue with what he calls the ‘go open’ project in psychotherapy ethics—the idea that the so-called ‘common factors’ in therapy should be disclosed to prospective patients. Although Leder does not give a detailed list, the common factors include therapist characteristics (empathy, positive regard, positive expectations that therapy will succeed), patient characteristics (expectations about therapy including its plausibility, confidence in the therapist), and the working alliance (how well both therapist and patient work well together during sessions). He argues that the project advocating disclosure of these factors is flawed on two grounds: (1) that information about common factors is not necessary for informed consent; and (2) clarity about specific mechanisms of change in therapy is consistent with ‘many theory-specific forms of psychotherapy’. There are multiple serious problems with Leder’s critique of the recent literature, including how he represents the contours of the debate, which I list, and address in this response.
- applied and professional ethics
- informed consent
- clinical ethics
Statistics from Altmetric.com
The ‘Go Open Project’
Leder characterises the ethical debate surrounding the disclosure of common factors as a single project.1 By drawing on evidence for the importance of the common factors in psychotherapy, and contrasting this with the ‘specific factors’ in treatment, he notes that proponents of this project claim that—as currently practised—psychotherapy works as a ‘placebo’. This is a reasonable representation—as far as it goes—and Leder cites three publications to support this perspective in his paper. However, most of the recent papers advocating common factors in informed consent processes do not depend on any interpretation of psychotherapy as a placebo (disclosure: I should know because I wrote them)2–5 or, alternatively, advocate caution about any such interpretation (again: one of my own).6 Regrettably, Leder does not delve into this literature even though it is not terribly extensive. I say this not on the grounds of self-importance but because, when critiquing a project or field of thought, it is better to examine the literature sufficiently well to avoid replicating lines of inquiry or strawman-ing, and thereby missing opportunities to finesse and strengthen one’s analysis. To illustrate, as my colleagues and I noted in 2016,
[T]he term ‘placebo’, when applied to psychotherapy, may invite more questions than it can easily resolve. Nonetheless… the core moral debate about clinical placebos raises important themes that are transferable to a psychotherapy context: namely, are therapists providing adequate information to patients about psychotherapy works…?6
Mechanisms versus mediators of change
So, let us focus, then, on the themes, including the central concern about what adequate informed consent to psychotherapy might encompass. One consideration is that the common factors should be disclosed to patients. Leder counters that, “The ‘go open’ argument errs in assuming that the common factors provide an explanation of the efficacy of therapy. The argument confuses the possible mediators of change in psychotherapy with mechanisms of change”. He urges, “the common factors may be mediators of change, but they are not necessarily mechanisms”.
The point is not particularly contentious. In fact, it has been articulated in multiple previous papers aligned with the ‘common factors project’.2–5 7 To pluck one example, in 2018 I argued, “when it comes to process research, correlation does not necessarily mean causation, and no evidence so far demonstrates that specific factors (or, indeed, the common factors) are best described as the causal determinants of change in therapy.”3
It is important to pause here, to reflect on what we can infer from randomised placebo-controlled trials (RCTs) in psychotherapy, and to give a brief flavour of the challenges facing psychotherapy trialists, including the methodological problems that do not arise in biomedical clinical research.2 3 8 9 Unlike in biomedical clinical trials, double blinding in RCTs in psychotherapy is impossible since treatments are dependent on interpersonal interactions. This means it is enormously difficult to isolate and to test the efficacy of the so-called specific factors of any psychological modality. In addition, controlling for the so-called common factors presents a major challenge to the robustness of psychotherapy clinical trials. Complicating matters further is wide variation in how investigators interpret the concept of ‘placebo’ in psychotherapy research.8 Notwithstanding, comparison trials of different psychotherapies (the so-called ‘horse race’ studies) suggest that the various specific techniques (across psychodynamic, humanistic, cognitive behavioural and other modalities) may be less important for patient outcomes than the common factors.9 It is this research that has given rise to the hypothesis that the most important mediators of change may be, in fact, be the common factors.
The nub of the issue, then, remains: what should be communicated to patients about how psychotherapy works? Leder proffers an answer: “An explanation of the ‘real engine of treatment’ will need to provide the patient with an account of the mechanisms of change that are responsible for the efficacy of the treatment. And crucially, a number of theory-specific therapies attempt to do this” (italics added). Certainly, these theory-specific therapies may attempt to do this, but that does not make them true. This is the crux of the debate about disclosure of common factors. The nettle that Leder fails to grasp is that the epistemological limitations of RCTs in psychotherapy research do not permit inferences about the explanatory truth of specific techniques. Focusing on cognitive–behavioural therapy (CBT) (“one of the most widely applied and studied therapies in the Anglophone world”) Leder insists: “The point is that not all forms of psychotherapy are equally as plausible (they are not)” (italics added). Again, on the contrary, this is the point. It is precisely why researchers hypothesise that the common factors may be the most important mediators of change, and why some ethicists have argued patients should know about them.
To recap: comparative treatment studies and placebo RCTs are valuable for determining that a treatment works. In ideal circumstances, placebo RCTs can also be helpful in identifying the mediators of efficacy, without thereby supplying a clear mechanistic explanation about how a treatment is effective—for that we need basic research to take up the baton.
Disappointingly, Leder fails to engage with these scientific and research problems, instead taking psychotherapy theories (specifically, CBT) at face value. This is to miss the key rationale behind advocating for common factors disclosure where a central problem is that therapists have, for too long, exhibited a facile approach to evidence-based practice.3 4 7
‘Going open’ may ‘seriously harm patients’
Leder claims that, “Common factors are consistent with theory-specific explanations of healing”. He argues there should be no extra onus on therapists to disclose common factors and that, “Practitioners should discuss mediators…. to whatever level of specificity that they choose (including no discussion)…” Going further, he urges, “The ‘go open’ argument misinforms patients”, putting them in a “worse epistemic position” and may even “seriously harm patients” (italics added).
These are strong claims. Taking the first point, I have emphasised—now on several occasions2–4 10—that, “there is compelling agreement across diverse psychotherapy traditions that such factors [the common factors] play a significant role in treating clients”.7 This invites the question about why it would be problematic—epistemically worse—to disclose information about these factors, in an understandable way, to prospective patients. As to the second point, in a recent web-based experiment on CBT disclosure practices, Kelley and I found that the specificity of disclosures is likely to be relevant to patient understanding.5 We found that among psychotherapy naïve-participants, a standard CBT disclosure (focusing on the specific factors) was not significantly different in influencing lay opinions about CBT than no disclosure. However, an enhanced disclosure—one that emphasised all of the potentially relevant mediating factors, both CBT and the common factors—did lead respondents to increase their ratings of the value of therapist-related factors. Relatedly, and to the third point: is it harmful for patients to know that therapist factors might be relevant to outcome? On the contrary, the attrition rate in therapy is high, posing a potential real harm to patients; as I have emphasised,
[W]hen patients ignore common factors (and place a premium on specific factors) they may undervalue the importance of a trusting relationship with their therapist. It is also conceivable that, if patients decide to discontinue therapy, they may come away with the false impression that the treatment (eg, CBT) is “not for me” rather than reflect on the role of other factors such as their relationship with the therapist, or their expectations about the treatment, that may have influenced their outcome.3
Leder owes us a rebuttal about (1) why disclosing these factors misrepresents the evidence-base; (2) is unnecessary; and (3) might even be harmful.
Common factors disclosures ‘sets too high a standard’
Finally, Leder argues that disclosing common factors sets too high a bar for psychotherapy: ‘given the pervasiveness of psychosocial factors on healing’, he urges, “it is far too onerous a requirement that all medical treatments include an exhaustive discussion of all potential mediators of change”. The point is exaggerated. As we have seen, Leder suggests that therapists need disclose only the specific techniques as mediators of therapy. In previous papers, I have articulated how therapists might reasonably extend this to the common factors, without encumbering patients with information.3–5
Perhaps, it is the mere change of practice that is considered onerous. In response, just because widespread practices are embedded does not provide justification for failing to disrupt them, especially if those practices are found wanting. Put another way, the question about whether it is considered too onerous is irrelevant if the practice is unethical. Indeed, there is growing attention in biomedical contexts to the ethical importance of disclosing contextual factors in care as important mediators of patient outcomes for certain conditions and symptoms (see the literature on ‘open label placebo’ studies).
There are many important papers on why ethical disclosures in psychotherapy should be commensurate with, and informed by, the evolving evidence base. While a welcome contribution, Leder’s paper fails mount a convincing case for why ‘going open’ should be resisted or why doing so may have ‘serious detrimental effects on patients’.
Collaborators Not applicable.
Contributors CB is sole author of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.