Waiting for psychiatry’s periodic table
In this first post I consider the DSM-5 in light of the problems inherent in diagnostic classification. In part two I reflect on how changes to the DSM-5 have affected my clinical practice.
Publication of the DSM-5 in 2013 caused controversy, with much of the commentary casting psychiatry in an unflattering light. The gist of it was that “the bible of psychiatry” was making arbitrary decisions about what was and wasn’t a mental illness. People who had Asperger’s disorder lost the diagnosis; and people who were grieving now found themselves susceptible to one. Mostly the articles were critical, but their exploration of the issues was shallow: they rarely passed comment on the difficulties inherent in the whole classificatory enterprise.
Since DSM-5’s publication I have read only two articles that examined the task of diagnostic classification with any degree of nuance and intelligence — one by Ian Hacking in the London Review of Books, and the other by Nick Haslam in The Monthly.[1. A scholarly volume that explores the DSM-5 has just been released. I have yet to read it, but a review is here.] Both argue that the DSM has assumed a cultural reach well beyond its original bureaucratic purpose. As originally conceived, its function was to determine whether or not someone had a diagnosis for the purposes of collecting data. But as Haslam writes, the DSM:
… has produced tectonic changes in [the] landscape and shaped how we think about abnormality. It dictates how treatment is practised and funded, professionals are trained, research is conducted and legal responsibility is assessed.
It has come to frame the way that clinicians think about their patients, and how patients think about themselves.
Both Hacking and Haslam point to the failure of the DSM as inherent in its very structure; a structure that Hacking suggests is akin to a botanical taxonomy and Haslam to a birdwatcher’s field guide. These comparisons are instructive, because the DSM assumes that mental illnesses can be classified in the same way that Linnaeus classified animal and plant life. Each disorder exists within a chapter, much as a species lies within a genus; with the chapters together encompassing all of mental ill health.
There is a major difference though between the structure of mental illness and the structure of animal and plant life. As Haslam writes, a bird can only be a member of one species, but a person’s clinical presentation can place it in multiple diagnostic categories. When a person presents with pervasive low mood, fatigue, and excessive worry they might receive diagnoses of major depressive disorder and generalised anxiety disorder. This is not like comorbidity in the rest of medicine. A person with diabetes and pneumonia has two distinct illnesses with distinct biological causes — one is caused by insulin dysfunction and the other by a pathogen.[1. Even if diabetes has made them more vulnerable to pneumonia.] Our distressed patient can only sensibly be said to have one illness, despite what the DSM says.
The truth is that mental illnesses are not structured like the animal and plant kingdoms, and a Linnaean taxonomy for them is wrong. Hacking notes that Linnaeus produced a lesser known taxonomy for minerals. We don’t know about it because it so clearly failed to capture the structure of the material world. The Linnaean system worked for living things because of evolutionary descent, unknown to Linnaeus at the time. For the chemical elements though, his taxonomy “did not represent nature”[1. Hacking, LRB] and was eventually displaced by the periodic table, which captured the chemical world’s much deeper structure. Hacking suggests that we are awaiting something similar to describe the structure of mental illnesses, and I agree. Whether the Research Domain Criteria takes us in that direction remains to be seen.
