Thursday 9 August 2018

Psychiatric diagnosis and informed consent: challenges and opportunities

It seems that there is an emerging middle ground in the debate about psychiatric diagnosis. Two recent pieces struck a note between polar positions: Jay Watts reminds us that diagnosis can be a lifesaver as well as a destructive force. Akiko Hart argues that diagnosis should be regarded as something over which people have a choice, rather than something to be entirely embraced or rejected.

One response to this uneasy stalemate has been to invoke the ethical requirement of informed consent. This would get around the variation in preferences by handing more choice to the recipients of diagnosis. It seems appropriate because informed consent is already the framework within which other healthcare interventions are offered. I have seen increasing numbers of people describe themselves as being "pro-choice" about diagnosis. This presumably involves taking something like an informed consent approach to the issue.

However, although it obviously moves us in a positive direction by protecting individual choice and autonomy, the idea of informed consent does not apply so easily to diagnosis as it first appears. There are at least three challenges for an informed consent framework in relation to the use of psychiatric diagnoses:

1. Scope 

The first challenge has to do with the scope of the process to which individuals would be consenting. Would informed consent  include questions about whether diagnostic categories would be used in an individual's treatment planning or in their clinical notes? Some clinicians already elect to exclude diagnostic labels in their reports - others do not. We can imagine the exclusion of diagnostic terminology from notes, but what if, given a clinician's working understanding of a person's problem, this exclusion were only token, or cosmetic?

Drilling down further, would the informed consent process also be supposed to apply to the process of clinical decision making? Here's a common situation: a psychiatrist elects not to prescribe SSRIs for an individual's depression on the ground that they believe the individual is at high risk of a manic episode. The SSRI is avoided because these medications are liable to set off mania in people with a bipolar diathesis. Here the reasoning is grounded in diagnosis (suspected risk of a bipolar disorder): the clinician does something in the best interests of an individual on the basis of a diagnostic considerations. Does informed consent to diagnosis preclude this sort of reasoning? If I have declined to consent to psychiatric diagnosis, do I also intend to prevent my psychiatrist from making this kind of judgment? What is the ethical clinician to do here? They seem to be acting without the informed consent of the individual (if they have deployed diagnostic reasoning to direct that person's care), but it seems frankly harmful (and therefore clearly unethical) to go ahead and prescribe the SSRIs given what they know.

This is a concern that could be answered if a more detailed account could be given of what "diagnosis" is supposed to include. It seems pretty clear the individual refusing to consent to diagnosis would be declining to understand themselves as "having" an illness called depression, or OCD or what have you. This seems reasonable. Such self-construals are arguably* beyond the purview of clinicians anyway. If a person makes sense of their experience in spiritual or religious or magical terms that are distinct from the language of diagnosis - clinicians may offer an alternative perspective, but they have no mandate to enforce a change in worldview. Clinicians have no right to tell people how to think.

When we propose an informed consent model for psychiatric diagnosis, we need clarity over what people are being asked to consent to.

2. Possibility  

The second challenge has to do with whether informed consent to a diagnosis is practically possible. First there are issues that pertain to the witholding and disclosure of relevant information. You see your psychiatrist and they say, "I have a diagnostic formulation, would you like to know it?" If you have the willpower to say "no" then you have to live with the knowledge that they have formed an opinion that may be relevant to you. If you decide you want to hear what the diagnosis is, you have to undertake some mental gymnastics in order to simultaneously know what you have heard while also discounting it as irrelevant to your self-understanding.

Then, given the transformative nature of diagnosis as a type of information, is informed consent to it even possible? Diagnosis changes your conception of yourself. Once you start to identify in terms of a diagnosis, the way you perceive your self and your life will shift in unanticipated ways. Perhaps you don't want to receive a diagnosis now, but once you know what it is you might change your mind. Perhaps you feel hostile to your diagnosis at first, but as time goes on you come to accept it. Perhaps your initial enthusiasm about a diagnosis turns to regret as you realise how it has limited you. Diagnosis - if it is a transformative experience in the sense intended by LA Paul - may not be the sort of thing to which one can consent.

To return to our opening proposition - that there are some people who find it helpful and some who do not - how can anyone anticipate in advance what sort of a person they will be? We need to have a better understanding of how clinicians would be expected to handle an informed consent process.

3. Coherence

The final challenge has to do with whether it is even coherent to think about providing informed consent to a diagnosis. On the one hand, we are talking about something like a procedure here, so it seems appropriate that people need to consent to it. On the other, a diagnosis is meant to be algorithmic, so (provided you have already consented to some form of assessment) consent seems misplaced. Either you meet criteria for a DSM disorder or you don't. Consider a situation in which a psychologist assesses someone and concludes that they meet criteria for disorder [x]. That is information, whichever way you interpret it. What is changed by withholding such information from someone? Is such withholding even ethical?

The coherence challenge also applies to the ways that diagnosis functions socially. Consider this situation: two people consult the same psychiatrist for mental health treatment. They meet in the waiting room and get talking. Person A says "Dr. Z has been very helpful to me. Since she diagnosed me with depression I have had a whole new way of understanding my bleak mood and my life has become far more liveable." Person B says "that's interesting, Dr. Z also felt I had depression, but I don't like diagnostic labels. She and I discussed the fact that depression is a construct and we agreed that I had a choice about whether to accept the diagnosis. I prefer to regard myself as not having an illness." One outcome here is that the two agree to disagree and take satisfaction in knowing that their provider can accommodate different preferences. But another outcome is that both feel unease. Person A's understanding of their situation may be changed now - if there is choice about this issue, do they perhaps not have an illness after all? As for Person B, they must wonder whether the doctor is simply humoring them. Yes they have agreed to behave "as if" Person B is not suffering from an illness, but a doubt arises: perhaps Dr. Z does actually believe in an illness process and is secretly regarding Person B with this cognitive set in mind.

This highly simplistic scenario gets at the problem that a diagnosis is never just about one person. Every actually-existing instance of a class tells you something about the class. If some sub-set of people within a diagnostic grouping is being told by their clinicians that their diagnosis doesn't track anything real, that has implications for the others in the same grouping. Imagine the situation from such a person's perspective. Their clinician says they have a disorder, but they are also suggesting to other people (ostensibly with the same disorder) that they can opt out of it. This looks thoroughly inconsistent. Whether you think this inconsistency is particularly troubling will depend on your views of diagnosis.

So there is a significant worry here. I suggest we can start to address it in two ways:

The first response is to acknowledge that psychiatric diagnosis is often actually applied with some degree of lassitude. A clinician doesn't have to be (in fact shouldn't be) entirely guided the algorithm. They can conclude that the symptoms aren't qualitatively the same as those described in the DSM and thus avoid diagnosing someone who otherwise resembles the criterion. Alternatively they can conclude that the person resembles the prototype for a disorder even if they fall short of the criteria. Additionally, many diagnoses have a criterion that includes this sort of language: "clinically significant distress or impairment in social, occupational, or other important areas of functioning." This introduces further subjectivity and may account for some of the variable reliability of diagnosis.

Given this lassitude, we can start to see a way for clinicians to be coherent in their inconsistency. Two people with extremely similar experiences in regard to their mood and behavior may nonetheless interpret these things in very different ways. One feels themselves to be overcome by an inane hopelessness that cannot be understood; the other believes they are in the midst of a profound spiritual or time of life crisis. Might not their functioning be impacted differently, and might it not therefore be appropriate to diagnose one and not the other as a result?

A second response is that diagnosis is open with regard to how far it applies to people's self identity. Here the get-out provided has to do with the clincian's work in helping an individual to achieve some degree of distance from any diagnosis that is used for clinical purposes. Imagine a clinician says something like "I conceive of your difficulty as an instance of DSM-disorder [x], but look, it is up to you whether and how you internalize that information - the DSM is at best a very rough guide to help us make sense of people's experiences. If you feel disinclined to think of yourself as suffering from a disorder then I would encourage you to develop other ways of making sense of it."

So depending on what is meant, it is not necessarily incoherent to seek informed consent to diagnosis, but more work is needed on how clinicians could start to implement it as a deliberate approach.

There are at least three areas of potential fudge to be worked through if we are to promote a pro-choice attitude to diagnosis. Over what activities would this framework scope? How would we manage the withholding of classificatory information? and how can we introduce an informed consent model without unduly negatively impacting the desirable social functions of diagnosis?

__________________________________________________________

*I say arguably because this is a bit of an ethical grey area. On the one hand ethics codes presuppose a respect for peoples' diverse worldviews. On the other hand, the professional impetus of psychiatry and psychology includes the idea that some worldviews are appropriate targets for change. If a person believes they are being persecuted by the FBI; that is not  a belief that is necessarily respected in the same manner as the belief in a benevolent creator. Equally, the idea of "psychoeducation" implies that there is information (such as about the nature of panic attacks) it behooves clinicians to share.