I just spent a while reading all the papers in the American
Journal of Psychiatry about the future DSM (1-5). As you might expect many people have many
things to say and that is as true inside as outside the field. We are on the cusp of another epoch of DSM
articles in the popular press that will predictably vary from inadequate to
horrific. Those articles will claim that the DSM is published as a source of
revenue for the American Psychiatric Association (APA), as a way for
pharmaceutical companies to make money, and as a “Bible” for
psychiatrists. There will be
philosophical musings tangentially related to the field but extremely critical. There will be the usual antipsychiatry
screeds about how it is unscientific, how there are better systems out there,
and how the diagnoses are mere labels that mean nothing. Most of those opinions
will be written by people who have never practiced psychiatry or been treated
for a mental illness. It seems that just
about anybody believes that they are an expert in psychiatry.
For those of us familiar with the field - our backgrounds
are more uniform. A significant number of people are like me – undergrad
science majors who are always interested in biological science and medicine. We
practiced in acute care settings and saw people with significant medical
comorbidity. We made plenty of medical
and neurological diagnoses that nobody else made and were a resource for that
kind of referral. We knew early on that
many of the diagnoses listed in the DSM were questionable and we never used
them. It turns out we are the last people the DSM is designed for and after
reviewing recent opinion pieces I will tell you why and how it can be
corrected.
The lead paper by Oquendo, et al (1) briefly reviews common
cited problems with the DSM and possible remedies. The first criticism is that it is
atheoretical. That is less of a problem than described. Any reader of the DSM sees immediately that
despite the stated atheoretical stance there are clear stated etiologies for
DSM listed diagnoses. To keep it simple, I refer any reader to the table Diagnoses
associated with substance class (p. 482).
That table contains 127 diagnoses associated with specific substances. There are similarly many diagnoses that
identify a psychosocial factor as being involved in the etiology. Categories versus spectrums are listed as
problem 2, despite the fact there probably are no spectrums (from the genetic
side) and all polygenic medical conditions (hypertension and diabetes mellitus
for example) have the same limitations. There are 4 additional uninteresting
points and proposed solutions. One of
the subcommittees is focused on the Dahlgren-Whitehead framework for social
determinants of health. At the same time
another committee is looking at instruments to assure a more comprehensive sociocultural
assessment. It made me wonder whether
anyone on the committee had ever read a current comprehensive psychiatric
assessment. Every psychiatrist should
have concerns about more checklists.
The second paper by Cuthbert, et al (2) was about biomarkers
and biological factors. The discussion was long on biomarkers and short on
biology. To neuropsychiatrists this
section of the DSM has always been a disappointment. As an example, the section
with the most biology – neurocognitive disorders has surprisingly little
discussion of associated medical features (like a gross characterization of EEG
in delirium) or a discussion of neuropathology without any additional
discussion of what that looks like clinically.
The Oquendo committee (1) has proposed changing the name of the DSM to
the Diagnostic and Scientific Manual because it is no longer used to
collect statistics. If that occurs, they
need to put a lot more science into it, and this is the area for it.
I have proposed a separate DSM for psychiatrists in the past
but a separate volume on the current science of psychiatry would be as useful.
I am talking about more than just a review of unproven research, but how the science-based
psychiatrist translates what we know so far into clinical practice. I would start with a rewrite of the section
on Neurocognitive Disorders and all the important variations before worrying
about plasma biomarkers and whether they are FDA approved. There are volumes
written on this subject that have been lost on the DSM. To cite a few examples – should a
psychiatrist be able to recognize presentations of encephalitis, meningitis,
and the various presentations of vascular dementia from their own assessment
and available imaging and lab studies? Should
a psychiatrist be able to diagnose various forms of aphasia and do the
indicated evaluation? Of course, they should – and it is all part of the rule
out criteria for psychiatric disorders. It is not enough to leave the medicine
and neurology of psychiatry to somebody else.
But very little is mentioned in the DSM except the rule out conditions:
“the disturbance is (or is not) attributable to the physiological effects of a
substance or another medical condition (or mental disorder).” That is too vague for psychiatrists.
The Structure and Dimensions Committee (3) is charged with
coming up with the most clinically useful structure of the future DSM. That involves incorporating recent
research. They have produced a lengthy
table summarizing the total categories, named categories and prevailing
frameworks and theories used for all the DSM starting with the first one. That
number goes from 4 to 22 categories in the DSM 5-TR. There is usually criticism about diagnostic
proliferation – but not much about category proliferation. When I encounter these numbers – I remind
myself that we started with a unitary psychosis model in the 19th
century. By 1918 (6) the situation not
much better with the major diagnostic categories being psychosis or not
psychosis. It could be argued that early diagnostic and classification
efforts failed to recognize or include mental disorders that had been observed
since ancient times rather than lower numbers being more ideal.
The fourth paper (4) is focused on Quality of Life (QOL) as
an essential part of psychiatric diagnoses. They establish premises based on
the often-quoted literature on disability associated with psychiatric
diagnoses. They describe a bidirectional
relationship: “… symptoms of a mental
illness can impair the individual’s functioning in daily life, and poor
functioning can in turn lead to or exacerbate the symptoms of a mental illness.” The paper has two definitions of QOL. The author’s definition is “a person’s
subjective perception of their emotional, psychological, and social well-being.” The paper also contains the World Health
Organization (WHO) definition of QOL “incorporates how an individual feels
about their emotional, social, and physical well-being, which can affect and be
affected by their mental health condition(s).” WHO further defines QOL as “one’s perceptions
of their position in life, contextualized by the culture and value systems in
which they live, in relation to their expectations, goals, and standards.” There is a related discussion on the Global
Assessment of Functioning (GAF) from previous DSMs. QOL metrics were decided to be subjective
rather than clinical ranking like the GAF.
The GAF was also thought to conflate symptoms of mental illness with
functioning even though there is a clear relationship.
The authors discuss the World Health Organization Disability
Assessment Schedule 2.0) (WHODAS-2.0) and it’s use to rate psychiatric
disability. It is a 36-item, 100-point self-administered, 6-dimension rating
scale. Administration and scoring in
full clinical schedules was considered a limiting factor, but clinically the
question is what happens with more identified problems? Does the treatment plan expand
proportionally? Will psychiatrists be
expected to either treat directly or develop referral sources for all the disabilities
identified as communication, mobility, self-care, interpersonal, life and
societal activities. Additional briefer QOL instruments are discussed as well
as brief interventions.
A critical concept that was not mentioned was the patient’s
baseline function. With every patient I saw, I had a subjective (and often
other informant) description of their baseline level occupational, academic,
and interpersonal functioning. In some
case it involved activities of daily living (ADLs) and instrumental activities
of daily living (IADLS). On inpatient
units those ADLs were often documented by occupational therapists. In my outpatient Alzheimer’s Disease and
Memory Disorder clinic – every new patient had their ADLs documented by the RN
staffing the clinic. It required hours
of work per day that were not reimbursed.
My clinic was eventually shut down because of that unreimbursed work and
my refusal to do the work myself for free. The additional cost and time for
these assessments is a reality factor in the modern rationed health care
system.
The fifth paper is entitled: “The Future of DSM: A Strategic
Vision for Incorporating Socioeconomic, Cultural, and Environmental
Determinants and Intersectionality.” The
definition of intersectionality is “a framework for understanding how various
social and political identities—such as race, gender, class, sexuality, and
ability—overlap and intersect to create unique combinations of privilege and
systemic discrimination.” I have a
problem with the use of a vague term that is used rhetorically being implemented
in a DSM. The DSM is a target of
rhetoric and putting rhetoric in the manual is likely to amplify its role as a
target. I have also reviewed ample
evidence that the major journal of the APA – was unable to separate rhetoric
from reality in the case of clear
historical evidence about racial discrimination. This highlights the need for clear definitions
and avoiding political rhetoric in any rethinking of this manual. It also highlights the need for clear
evidence rather than rhetoric and that commentaries – even in the flagship
journal of the American Psychiatric Association cannot be depended upon for
that evidence.
Intersectionality is unnecessary to get at what the authors
hope to accomplish. Cross cultural
psychiatric evaluations are the case in point. They involve an assessment of
cultural differences and how the culture affects disease definitions and
presentations, the sick role in that culture, and how demographic factors affect
how a person is advantaged or disadvantaged in their original or adopted
culture. The authors suggest it is necessary to promote various public health
prevention strategies and promote health care equity. As far as I can tell, health care equity in
the US is strictly in the purview of politics and in one year a massive amount
has been destroyed by the Trump administration.
Political features should be avoided as much as rhetorical features in a
DSM, especially given the abysmal track record of physician medical
organizations in politics.
The authors define socioeconomic, cultural, and
environmental determinants of health (SCE-DoH) as the key focus (along with
intersectionality). These determinants are all well known to any psychiatrist
who has recorded a social history for a detailed assessment and that should
include all of us. They conceptualize
them as modifiable or non-modifiable risk factors and how they may be relevant
for prevention strategies. Much of the
prevention is outside the scope of psychiatric practice and advocacy by
professional organizations has questionable impacts. They also use the Dahlgren-Whitehead model of
main health determinants and cover suggestions of screening patient populations
for these variables. They conclude that
the next DSM should include recommendations to use multiple “vetted
instruments” to make these SCE-DoH assessments.
They give an example of how this assessment can be built into routine
clinical care. Interestingly, the
psychiatric assessment is not included in the “routine diagnostic workflow”
(see figure 2). Looking at the strategy
2 where the SCE-DoH is used to determine “management as usual” versus “enhanced
case management” – I made that determination myself for 40 years. For the last
25 years that “enhanced case management” was not available for most people
needing it. That tells me that the suggested assessment is already being done
by some people and the necessary resources are not there. I found myself documenting that fact in too
many cases.
The Committee realizes that they cannot create an additional
burden on clinicians who already have unrealistic demands and provide far too
much work for free in rationed environments.
That translates to less time to do comprehensive assessments – not more.
Even though these are very preliminary statements about the
future DSM – I am not very hopeful at this point. The commentaries so far seem directed at
criticisms from outside of the field rather than what psychiatrists need. Apart from the criticism I have offered so
far what is noticeable:
1: The lack of
commentary on medical and neurological diagnoses – in any psychiatric
classification it is either explicit (or implicit) that what are considered the
current psychiatric diagnoses are not caused by a substance or another medical
diagnosis. The non-DSM diagnostic
systems are generally just focused on the listed symptoms of these disorders
and there is no provision for other medical conditions. It is also not explicit enough in medical
training. At some level this is explained away and needing to utilize whatever
resources are available. That is not
enough. The DSM should have a section of
diseases by system that need to be diagnosed if they are present and at least a
reference to how that should be done.
There is not nearly enough information on what medical diagnoses
psychiatrists make. This is also an
important feature for resident education since it would suggest how much clinical
medicine and neurology residents need to be exposed to and whether they are
seeing relevant cases.
2: Philosophical
criticisms while minimizing biology and history – in several of the papers the authors talk
about “natural kinds” and “carving nature at the joints”. This is philosophy speak that has been used
to obfuscate the field. The first time I encountered these arguments they
struck me as obvious nonsense. That was
first suggested by Thomas Sydenham when he made this statement in about 1640:
“In writing the history of a disease, every philosophical
hypothesis whatsoever, that has previously occupied the mind of the author,
should lie in abeyance. This being done, the clear and natural phenomena of the
disease should be noted — these, and these only…” (7)
DeGowin and DeGowin (8) summed up the process over the next
three centuries:
"For several thousand years physicians have recorded
observations and studies about their patients.
In the accumulating facts they have recognized patterns of disordered
bodily functions and structures as well as forms of mental aberration. When such categories were sufficiently
distinctive, they were termed diseases and given specific names.”
It seems that the conceptual clarity here requires no
reference to naturalism or essentialism.
It only requires empiricism and a determination of sufficiently
distinctive. In my long and
intensive career – the only place I have encountered these philosophical
arguments was in a literature that was generally critical of psychiatry. In the process it also requires psychiatrists
to suspend the idea that empirical adequacy is not all that is required, but
also all that we were taught.
Conceptual expansionism or semantic drift has been used to
criticize the DSM and psychiatry and that needs to be called out wherever it
happens. By that I mean a concept that
is developed within one academic silo that is suddenly applied without
precedent or a clear basis to psychiatry.
On this blog I have criticized several of these applications including
epistemic and hermeneutical injustice.
Although none of the Am J Psychiatry papers used the term, I did
encounter folk psychology now being applied to criticize the DSM (9) in
a mailing. That is a concept I was
familiar with from Andy Clark’s work (10). If you are not familiar with the concept a generally
accepted definition would be: “The
everyday ability to predict and explain the behavior of ourselves and others by
attributing mental states—such as beliefs, desires, intentions, and fears.” In other words – you see somebody doing
something and come up with a theory of why they are doing it. I have written about it on this blog as a
reason why many people seem confident in their knowledge of psychiatry and
psychology even though they have never been trained in either. There are
several theories of how a folk psychology theory can apply, but the original debate
centered on how the ascribed beliefs, desires, intentions, etc. had no neural
equivalent and therefore that at some point these mental states would be
replaced by more scientific terms. In other word suggesting that the DSM is folk
psychology is basically saying the signs and symptoms used as descriptors
have no brain equivalent and therefore it is an invalid classification. This
argument is essentially the same argument that there is an explanatory gap
between what most people consider consciousness to be and the neural substrates
that causes it. Consciousness is approximately
represented in neural substrate and the same thing can be said for mental
disorder symptoms.
3: The continued lack
of focus on what might be useful to psychiatrists -
When I think about a DSM that might be useful to
psychiatrists or at least the kind of psychiatrists I am used to working with –
there needs to be more than the usual slicing and dicing of diagnostic
criteria. Adding more work with more
rating scales is also a disappointment.
A manual breaking down the current work with examples and a suggestion
of the potential exhaustive data points might be. For example, pointing out
that the typical phenomenology of a disorder should be adequately represented
in the history of the present illness.
That obviously includes any precipitating factors irrespective of what
they might be – biological or sociocultural. The next section should include a
discussion of the past psychiatric and medical histories as well as comorbid
conditions. Psychiatrists should be
expected to know relevant medical diagnoses, how medical comorbidity affects
psychiatric treatment, and medical causes of psychiatric presentations. The usual disclaimer about medical conditions
is as inadequate as a disclaimer about sociocultural aspects of care. The new DSM should not be a mere collection
of psychosocial determinants completely devoid of medicine.
A more formal formulation section should be there. In the DSM-5 for example it is referred to as
a “concise summery of the social, psychological, and biological factors that
may have contributed to developing a given mental disorder.” (p. 19).
There are multiple ways to write a formulation (behavioral,
psychodynamic, neuropsychiatric, and others) and they should all be discussed
in the DSM.
4: A theory section
on the biology of psychiatric diagnoses – why they are complex and how that
complexity should be approached. There
are experts in the field who can comment on how polygenes produce quantitative
diagnoses that can blend imperceptibility into the normative states. Some of those same experts can discuss the
statistical methods used to try to improve classifications and how that works
clinically. There should be a comparison
with other commonly described quantitative disorders like hypertension and
diabetes mellitus Type 2. The
classification system of rheumatology could be discussed as a direct comparison
to the DSM.
I have written about the problem with the term transdiagnostic.
I do not think it adds any specificity
to interventions. In psychiatry what is
considered a transdiagnostic symptom can also conceal a potential primary
problem. One of the most common scenarios I encountered in practice was longstanding
insomnia prior to the onset of depression. In the transdiagnostic scenario,
insomnia could be considered just that or a symptom of another disorder rather
than a primary sleep disorder. All these issues including categorical versus
dimensional diagnoses should be covered in this theory section written by our experts. There are plenty of reasons not to blindly
accept the transdiagnostic jargon as being that relevant.
Psychometrics can be discussed in the theory section. We have all heard and read about reliability
of diagnoses for decades and a lack of validity. Reliability statistics are
available for a range of DSM categories and that could be included as a single
graphic with a brief discussion. The
discussion of validity needs to be more extensive and nuanced rather than just
dismissed. Study groups from DSM-5 were
working on 11
validity indicators. It is time to
see them on graphics like what can be constructed for reliability. The data
should be included where it exists.
5: A genetics section: Genetics and the associated molecular biology
is the future of medicine and psychiatry. A summary of that data should be
available in the DSM as well as the clear importance of this information. At the biological level, the discussion
should be clearly focused on changes in brain systems associated with disorders
and the problem of many genes affecting these systems.
6:
Definition/Threshold of a disorder:
There is always criticism about the dysfunction threshold
for making a diagnostic assessment.
There is never much discussion about why it is necessary or why there
are consensus diagnoses. Even a superficial
look at other specialties that treat polygenic heterogeneous entities invites
comparison. Rheumatology is a case
in point:
“Rheumatologists face unique challenges in discriminating
between rheumatologic and non-rheumatologic disorders with similar
manifestations, and in discriminating among rheumatologic disorders with shared
features. The majority of rheumatic
diseases are multisystem disorders with poorly understood etiology; they tend
to be heterogeneous in their presentation, course, and outcome, and do not have
a single clinical, laboratory, pathological, or radiological feature that could
serve as a “gold standard” in support of diagnosis and/or classification.”
A recent review of polymyalgia rheumatica (PMR) in the NEJM
(11) looked at diagnostic algorithms for both acute PMR and treatment. The introduction involved the statement: “The diagnosis of polymyalgia rheumatica is
made on the basis of clinical grounds by combining characteristic signs and
symptoms with laboratory findings and ruling out common mimickers such as
late-onset gout and pseudogout and others.” (p. 1099).
I counted 23 conditions in the differential diagnosis. One of the criteria for the diagnosis is “functional
impairment”. The implication is that it
is due to morning stiffness or possible pain but that is not specific. There are limited reviews of how to establish
diagnostic criteria for diseases and disorders that lack objective tests
(12). I think the degree of dysfunction
is obviously relevant when assessing disorders that are based on purely
subjective signs and symptoms. It
factors into routine clinical care of both known and unknown diagnoses. On this
blog I have documented examples from numerous medical and surgical specialties.
That is my criticism after reading 5 current papers on the
direction of the DSM. I really do not
want the next volume to look like what has been described so far. When I think
about my final 1500-2500 word assessments that contain just about everything
the authors of these papers discuss and much more – I do not want to see all
that good work sacrificed because somebody wants to include more checklists or
dimensions of questionable value. I have had people tell me years and in some
cases decades later, that they found those assessments to be valuable and
useful for future evaluation and treatment of that same person.
If I had to capture three elements that the
future DSM planning seems to miss it is that phenomenological assessments can
easily contain as much or more data than checklists, that psychiatry is a
medical specialty, and that like all medical specialties the field has
boundaries. The current suggestions from these papers stretch those boundaries into
activism, politics, and importing criticism from other academic silos rather
than a restatement of what is relevant for psychiatric assessment and
classification.
That should be the priority…
George Dawson, MD, DFAPA
References:
1: Oquendo MA,
Abi-Dargham A, Alpert JE, Benton TD, Clarke DE, Compton WM, Drexler K, Fung KP,
Kas MJH, Malaspina D, O'Keefe VM, Öngür D, Wainberg ML, Yonkers KA, Yousif L,
Gogtay N. Initial Strategy for the Future of DSM. Am J Psychiatry. 2026 Jan
28:appiajp20250878. doi: 10.1176/appi.ajp.20250878. Epub ahead of print. PMID:
41593833
2: Cuthbert B,
Ajilore O, Alpert JE, Clarke DE, Compton WM, Drexler K, Fung KP, Gogtay N, Kas
MJH, Kumar A, Malaspina D, O'Keefe VM, Öngür D, Tamminga C, Wainberg ML,
Yonkers KA, Yousif L, Abi-Dargham A, Oquendo MA. The Future of DSM: Role of
Candidate Biomarkers and Biological Factors. Am J Psychiatry. 2026 Jan
28:appiajp20250877. doi: 10.1176/appi.ajp.20250877. Epub ahead of print. PMID:
41593830.
3: Öngür D,
Abi-Dargham A, Clarke DE, Compton WM, Cuthbert B, Fung KP, Gogtay N, Kas MJH,
Kumar A, Malaspina D, O'Keefe VM, Oquendo MA, Wainberg ML, Yonkers KA, Yousif
L, Alpert JE. The Future of DSM: A Report From the Structure and Dimensions
Subcommittee. Am J Psychiatry. 2026 Jan 28:appiajp20250876. doi:
10.1176/appi.ajp.20250876. Epub ahead of print. PMID: 41593835.
4: Drexler K, Alpert
JE, Benton TD, Fung KP, Gogtay N, Malaspina D, O'Keefe VM, Oquendo MA, Wainberg
ML, Yonkers KA, Yousif L, Clarke DE. The Future of DSM: Are Functioning and
Quality of Life Essential Elements of a Complete Psychiatric Diagnosis? Am J
Psychiatry. 2026 Jan 28:appiajp20250874. doi: 10.1176/appi.ajp.20250874. Epub
ahead of print. PMID: 41593851.
5: Wainberg ML,
Alpert JE, Benton TD, Clarke DE, Drexler K, Fung KP, Gogtay N, Malaspina D,
O'Keefe VM, Oquendo MA, Yonkers KA, Yousif L. The Future of DSM: A Strategic
Vision for Incorporating Socioeconomic, Cultural, and Environmental
Determinants and Intersectionality. Am J Psychiatry. 2026 Jan
28:appiajp20250875. doi: 10.1176/appi.ajp.20250875. Epub ahead of print. PMID:
41593836.
6: American Medico-Psychological Association. Statistical
Manual for the Use of Institutions for the Insane. 1918: https://dn790008.ca.archive.org/0/items/statisticalmanu00assogoog/statisticalmanu00assogoog.pdf
7: Sydenham, Thomas,
1624-1689; Greenhill, William Alexander, 1814-1894; Latham, R. G. (Robert
Gordon), 1812-1888. The works of Thomas
Sydenham, M.D. Volume 1, London.
Sydenham Society. 1848-1850. P. 14
https://archive.org/details/worksofthomassyd01sydeiala/page/lv/mode/1up?q=abeyance
Translation of Medical Observations by Thomas Sydenham,
London, 1669. The Preface. Original was
in Latin.
8: DeGowin EL, DeGowin
RL. Bedside Diagnostic Examination, 3rd
ed. New York. Macmillan Publishing Company, Inc. 1976. P. 1.
9: Aftab A. The
Future DSM: Bold redesign, lingering blind spots. Psychiatric Times. March 2026: 12-16.
10: Clark A. Microcognition: Philosophy, cognitive
science, and parallel distributed processing.
Cambridge, MA. The MIT press.
1989.
11: Dejaco C,
Matteson EL. Polymyalgia Rheumatica. N Engl J Med. 2026 Mar
12;394(11):1097-1109. doi: 10.1056/NEJMcp2506817. PMID: 41812194.
12: White SJ, Barker
TH, Merlin T, Holland G, Sanders S, O'Mahony A, Pathirana T, Theiss R, Pollock
D, Reid N, Munn Z. Methods for developing diagnostic criteria for conditions
without objective tests, biomarkers, or reference standards: a scoping review.
J Clin Epidemiol. 2026 Feb;190:112052. doi: 10.1016/j.jclinepi.2025.112052.
Epub 2025 Nov 18. PMID: 41265667.
