Diagnosis is one of the most important and controversial areas in mental health science and care. It’s become a commonplace to say that psychiatry lacks the biomarkers—blood tests, imaging, and so forth—that are used by other medical specialties. Instead, clinicians have to take a careful history from the patient, and interpret this in the light of their own observations, the individual’s history, and collateral information from friends and family. In truth, this isn’t a process unique to psychiatry; but psychiatry is unique in that this is still the mainstay of the diagnostic method.
These diagnoses, such as “depression”, “schizophrenia”, and “PTSD” can then be used to support discussion with the individual about their condition, communication with other professionals, treatment planning, and administrative work. Diagnostic categories are also used in research, meaning that therapies can be developed and targeted towards specific conditions.
The major diagnostic manuals used by psychiatrists for the past several decades have been the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM), and the World Health Organisation’s International Classification of Diseases (ICD).
The DSM in particular has come under heavy fire since the fifth edition (DSM-5) was launched with great fanfare and even greater controversy in 2013.There are two main strands of criticism.
The first, which comes from largely outside the psychiatric profession, builds on long-term concerns about the practice of psychiatry in general. This argument goes that diagnostic manuals are houses built on sand, constructions founded on the opinions of largely white, western men, which medicalise social and life problems. The diagnostic approach is misguided at best, actively harmful at worst. I am, of course, oversimplifying, but this critique is not without merits. And you will hear a version of it from quarters that could not by any stretch of the imagination be described as “anti-psychiatric”. Robust criticism of the DSM-5-TR’s treatment of grief is a case in point (Cacciatore, 2022).
The second strand of criticism comes from recent developments in mental health science. Several types of evidence, including genetic studies, suggest that mental health conditions as described in the manuals do not map onto clear and discrete biological categories (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2013). Psychiatrists are, in effect, like infectious disease specialists before microbiology. Some things that look the same might have different underlying causes. Some things that look different might have the same causes. And some symptoms occur in more than one diagnostic category.
‘Psychiatric Diagnosis: A Clinical Guide to Navigating Diagnostic Pluralism’, published in The Journal of Nervous and Mental Disease, is an attempt by leading authors in the mental health field to figure out where we are, and what might, and should, happen next (Aftab et al, 2024).
A way out of the maze?
Methods
This is billed as a review paper, but there is no indication as to how the literature reviewed here was searched for, retrieved, and analysed. This is not necessarily a problem, as this is not the kind of review (such as systematic review and meta-analysis) in which the reader needs to be given a clearly reproducible method. The purpose of this paper is descriptive and reflective. The reader should bear in mind that this is best considered an opinion-based paper based on analysis of the evidence. It is, of course, possible that such papers are selective in their citations, avoiding mention of work that is inconsistent with their argument and conclusions. But this should not detract from the importance of this paper as a reflection of how thought-leaders are considering the challenge of psychiatric diagnosis.
Results
This paper is a good read and sets out the issues around diagnosis clearly, with minimal jargon (the phrase “orthogonal and oblique rotation” and a brief baseball metaphor aside).
The authors review the evidence for how current diagnostic systems are used in practice, and assess the strengths and limitations of the newer systems of classification that have emerged in the past decade, including the Research Domain Criteria (RDoC), the Hierarchical Taxonomy of Psychopathology (HiTOP), and others such as the British Psychological Society’s Power Threat Meaning Framework (PTMF).
One of the most interesting parts of this paper is the authors’ assessment of how clinicians actually use diagnostic manuals such as DSM in practice; they assert that clinicians “rely on prototypical descriptions” to decide on a differential diagnosis, which is then revised with time. Consultation of the chapter-and-verse of the manual, if it happens, comes later in the process. And this is only part of a broader formulation of a narrative incorporating the circumstances in which the patient’s condition has developed, and the factors that have contributed. (It is worth pointing out, though, it is unclear what kind of data, if any, backs up the authors’ statements; the main reference here is to a book.)
As for how well our current systems reflect the reality of mental health problems, the authors characterise the current scientific consensus as “nominalist”. That is to say, there is an underlying reality of mental health conditions, and while current diagnostic systems describe it imperfectly, they can be a useful tool to “capture important differences that are relevant to etiology and treatment” albeit “in a historically contingent, pragmatic, and fallible manner.”
The authors also discuss psychiatry’s problem of defining the boundaries of “normality” versus “illness” based on function, acknowledging that this is difficult not only in mental health care, but across many medical fields. And even if individual psychiatric symptoms exist on a spectrum, there are other cases in medicine where setting an imperfect cutoff on that spectrum is useful for making treatment decisions (e.g. blood pressure).
So it’s not really accurate to call the DSM the “Bible of psychiatry”. It’s more a tool that clinicians use as part of process of care which, by its nature, is complex, holistic, and characterised by fuzzy boundaries. Ironically, the diagnostic process described by the authors is best described in a quote from the actual Bible: “we see through a glass, darkly”.
What of the new contenders?
RDoC is still affected by confusing messaging about the project in 2013, the same year that DSM-5 was launched in 2013. As the authors put it, “one is forced to distinguish between the research framework itself and the provocative manner in which RDoC was presented to scientific community as an alternative to DSM diagnoses”. RDoC provides a potentially useful framework for dividing up neuropsychological functioning into different domains, and for defining the multiple levels at which these domains operate—from “genes” to “paradigms”. It’s a long road from here to clear characterisations of mental health problems, let alone the clinic, although the authors argue that “RDoC domains such as negative valence systems and positive valence systems can enrich a diagnostic case formulation”.
HiTOP, meanwhile, is an attempt to organise symptoms and diagnostic entities on the basis of evidence rather than opinion. A revolutionary idea. The authors set out its two key concepts. First, “dimensionality”, that is, that “psychopathology can be understood in terms of a set of distinct dimensions” along which clinical phenomena fall. Second “hierarchical structure”, meaning that at the bottom of the tree sit the narrow elements of signs and symptoms; in the middle, syndromes, then subfactors (small clusters of strongly correlated syndromes) and spectra (broad groups of subfactors); and at the very top the “general psychopathology factor” p. As the authors point out, the apparently data-driven nature of the HiTOP project can mask the fact that the analysis is shaped by human choices—and, I might add, the demographic limitations of available data sources. As for the clinic, HiTOP is a much more realistic clinical contender than RDoC, but with a couple of caveats: first, it’s an open question as to whether its use would actually yield better results. And second, it’s possible that HiTOP constructs will be used in practice just as loosely as DSM diagnoses are now.
The authors discuss other approaches briefly; the main point of interest for me is how, despite much fanfare, the PTMF, a formulation-based “psychological alternative to psychiatric diagnosis” has “has largely been ignored by clinicians” and remains “virtually unknown in the United States”. This is surprising to me, given the strong critical psychiatry movement in the US. I wonder if this is because its British origins have limited the PTMF’s uptake across the Atlantic, or if insurance-based US health systems are simply not set up for formulation-based treatment. Another reason might be that if—as the authors of the present paper claim, psychiatrists tend to take a narrative, formulation-informed approach anyway—the PTMF has not been resisted as such. It might be the case that its less radical aspects at least were pushing at an open door, being already part of mainstream psychiatric and psychological practice.
Through a glass, darkly.
Conclusions
The authors conclude with an acknowledgement of the limitations of DSM and ICD, and the need for new systems. But nothing, they think, is ready for prime time yet, and certainly nothing is going to make current systems redundant anytime soon. Psychiatry will hang on to ICD and DSM in the spirit of Hilaire Belloc’s Jim: “And always keep a-hold of Nurse / For fear of finding something worse.”
That being said, they argue that new frameworks such as HiTOP, and the kind of thinking they engender, might be used productively alongside current diagnostic systems.
What’s the future? Not a Lord of the Rings-style “one ring to rule them all”, but a plurality of classification systems for different settings and uses.
Not coming soon: one ring to rule them all.
Strengths and limitations
The strengths and limitations of this paper are to an extent one and the same. It’s a subjective piece of work, written by individuals with deep knowledge of the field. It’s simply not possible to compare the clinical benefits of the various systems discussed here as the data isn’t there, so this discursive approach is the best we can get for the time being.
However, this paper is limited by its authors’ perspective as four men based at institutions in high-income countries. It would be interesting to look at issues of psychiatric classification with a more global balance. At present, culturally specific syndromes are covered in a glossary in the appendix of DSM, while the World Health Organisation promotes the use of mhGAP, essentially a stripped-down version of ICD, for use in low-resource settings. I feel there is much more to be said and done about classification systems globally, but the topic barely gets a look-in here. Similarly, feminist critiques of specific psychiatric diagnoses such as borderline personality disorder are not explored in this paper.
Finally, I think that the authors could have outlined in more detail some of the challenges that multiple classification systems will present. It’s fine to have a “horses for courses” approach of different systems for different settings. But my worry is that the research and clinical fields will gradually diverge in terms of focus and coherence. How will understanding and even new treatments from RDoC and HiTOP translate into an ICD- or DSM-based clinical approach? And what will this mean for other important players in the mental health space, such as funders and regulators?
A wider angle needed on diagnostic systems.
Implications for practice
There are two main implications for practice.
The first is that I would advise clinicians to read this paper, in order to get up to speed with how others are thinking about the limitations of current diagnostic symptoms, and what is being proposed instead. It’s important to know what’s over the horizon—and it would not surprise me if HiTOP or something like it eventually makes it to the clinic—and to see what aspects of the new systems might be fruitfully integrated into practice now.
Second, this paper sets out a major challenge for the clinical and research fields. As I said above, the authors think “one ring to rule them all” is not possible. If the foreseeable future is one of multiple classification systems, we will need to figure out a way for such systems to understand and align with one another. In other words, and jumping genre analogies from fantasy to sci-fi, we need the equivalent of the Babel fish.
The Babel fish, you will recall from Douglas Adams’s The Hitch Hiker’s Guide to the Galaxy, is the tiny fish that when inserted into the ear enables the individual to understand anything said to them, in any language. Unfortunately, writes Adams, by “effectively removing all barriers to communication”, the Babel fish “has caused more and bloodier wars than anything else in the history of creation.”
A future of multiple diagnostic systems is probably inevitable, but it will not always be easy.
Needed: a Babel fish.
Statement of interests
This blog reflects my personal opinions, and not those of my employer, Wellcome. Wellcome has provided a grant to the Mental Elf, but has no influence over its content or editorial decisions.
Links
Primary paper
Aftab A, Banicki K, Ruffalo ML, Frances A (2024) Psychiatric Diagnosis: A Clinical Guide to Navigating Diagnostic Pluralism. The Journal of Nervous and Mental Disease 2024 212(8) 445-454 https://pubmed.ncbi.nlm.nih.gov/39079000/ [PubMed abstract]
Other references
Cacciatore J, Frances A (2022) DSM-5-TR turns normal grief into a mental disorder. Lancet Psychiatry 2022 9 e7 https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(22)00150-X
Cross-Disorder Group of the Psychiatric Genomics Consortium (2013) Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide analysis. Lancet 2013 381 1371–1379 https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(12)62129-1
Jones M, Smith P, Anderson T. et al (2014) Antidepressants are safe and effective for women with depression who are pregnant (PDF). BMJ 2014 33(6) 172-183.
