Fine, so it’s not valid. It’s still useful!

What good is a psychiatric label?

As I’ve been reporting, biopsychiatry’s proponents have in recent days publicly admitted that their diagnoses lack validity. It’s not a new realization on their part. As one blogger describes:

The reaction to the 1970s crisis of American psychiatry was to use claims about the ‘reliability’ of diagnosis to strengthen the profession’s ‘scientificity’ in appearance but not reality.

What they’ve done, in effect, is play the game of using the alleged reliability of their diagnoses to imply or suggest validity,* and they’ve been remarkably successful at it for several decades. But this pretense has faced growing challenges, leading to the public admissions of the past couple of weeks.

The champions of biopsychiatry certainly recognize the significance of any public recognition of the validity problem - if they didn’t, they wouldn’t have spent decades trying to hide it. Now, however, they’re attempting to argue that a lack of validity isn’t really a fatal flaw for a psychiatric diagnosis. A recent New York Times article – “Psychiatry’s Guide Is Out of Touch With Science, Experts Say” – quotes NIMH head Thomas Insel stating that, while his agency will abandon these categories as invalid,** “his motivation was not to disparage the D.S.M. as a clinical tool.” In the interview, he calls the DSM “the best tool now available for clinicians treating patients” and says it “should not be tossed out.”

While they faithfully hold to their assumption that their bold “new” biopsychiatric research program will produce useful diagnoses, these people know that they have none to offer at present. So they’re left to assert that what exists must be useful. Michael First, also quoted in the article, claims that while NIMH’s RDoC “is clearly the way of the future,” it “can’t do what the D.S.M. does. The D.S.M. is what clinicians use. Patients will always come into offices with symptoms.” This comports well with the APA’s spin - as David Kupfer put it (quoted in my earlier post): “DSM, at its core, is a guidebook to help clinicians describe and diagnose the behaviors and symptoms of their patients. It provides clinicians with a common language to deliver the best patient care possible.”

So these are the arguments used to support the continued reliance on the DSM: people continue to experience psychological problems, it’s what physicians use in understanding these problems, it’s reliable, and there’s no available alternative. Added to these is a very real concern about the denial of legitimacy and care to people who aren’t officially diagnosed using these categories. Despite its fundamental lack of validity, then, it’s claimed that the DSM remains a useful clinical tool.

The first of these arguments is correct: it is overwhelmingly used in psychiatric diagnosis. The claims about reliability and nonexistent alternatives are not accurate, and rest on a series of problematic assumptions. For the sake of argument, though, let’s assume that all of these claims are sound, and evaluate the basic case for continuing to recognize the DSM “diagnoses.”

Obviously, the fact that physicians use a tool doesn’t make it clinically useful. Nor does the fact that it serves the function of matching people to drugs or other interventions. The distinction between usefulness to pharmaceutical corporations, psychiatrists or psychologists, the criminal justice system, and so on and usefulness to people experiencing problems has to be maintained. The system is not the client.

The third argument is the most astonishing. What they seem to be saying is that since people continue to experience psychological distress or exhibit behaviors deemed undesirable (I’m not playing along with their “patient” and “symptom” language), and the DSM provides a reliable means of classifying these experiences or behaviors, that makes it a useful clinical tool. How would his work, precisely?

The purpose of a clinical diagnosis, as I understand it, is to identify a real condition so as to facilitate effective treatment. But a diagnosis that isn’t valid, that doesn’t identify a real condition, is just a gratuitous label. How is that useful as a clinical tool? It’s interesting how often the DSM is described as psychiatry’s “Bible,” because a similar set of diagnostic classifications could be derived from the literal Bible. A reliable diagnosis of Bipolar, for example, could be replaced by a reliable diagnosis of Demonic Possession. I can’t imagine that anyone reasonable would contend that this would be a clinically useful diagnosis for the person so labeled. The same could be said about reliable diagnoses based on different chakra imbalances or what have you.

It seems plainly that the “clinical” function of the diagnostic labels is to prop up belief in the biological model and promote the drugs. What we’re talking about in psychiatry is a simulacrum of medical diagnosis. We can see this if we imagine what prescribing the drugs would look like if everyone knew and understood that the diagnoses were invalid. Many other interventions can happen without labels, but psychopharmacology pretty much requires them.***

The last argument - that we should be concerned about people not receiving needed help or respect if these diagnoses are scrapped - is an important one, if irrelevant to the question of their usefulness as clinical tools. The system is set up around invalid diagnoses. But the response to that fact isn’t to retain those labels in order to preserve the system but to abandon them - and the larger obsession with medical models - and create a system in which people can receive help and respect without having to adopt a pseudoclinical label. (Not least because the diagnoses don’t contribute to effective interventions and their actual effect is to stigmatize and delegitimize people’s experiences and their sociopolitical concerns.) That system would look very different.

* I like Marcia Angell’s simple description of the difference between reliability and validity:

The DSM-III was almost certainly more ‘reliable’ than the earlier versions, but reliability is not the same thing as validity. Reliability…is used to mean consistency; validity refers to correctness or soundness. If nearly all physicians agreed that freckles were a sign of cancer, the diagnosis would be ‘reliable’, but not valid.

** One Boring Old Man sums up Insel’s likely motive:

Dr. Insel hears a great sucking sound over at the APA offices and he’s trying to get out of its way before it sucks him down with it. His NIMH was a major partner with the APA in the DSM-5 conferences and planning. His RDoC was born in that failure as a way to keep the dream alive when the APA failed.

The Times piece notes another consideration: “[Insel] added that he hoped researchers would also participate in projects funded through the Obama administration’s new brain initiative.”

*** This isn’t to imply a one-to-one matching of diagnosis to drug. It works much more profitably as organized now: around the notion that one drug can treat multiple disparate disorders and a single disorder might respond to several different drugs.