The “Overdiagnosis” Confusion
“Overdiagnosis” is a blunt verdict offering blunt solutions
“Madness never could be among us—not among us—because it was behind asylum walls. Whatever we may have cannot be the old madness, not something isolated, different, pure deficit, meaningless, with nothing to say—but rather it would have to be something new, more familiar, something we come across in friends and family, on TV and in ourselves, something more understandable, having a voice—something better called something like mental health problems, fading into ordinary or extraordinary problems of living. Madness transforms into something else when it is in the community...”
Derek Bolton, What is Mental Disorder? (OUP, 2008), p. 242
There are few things as muddled and irksome as the public debate on “overdiagnosis” in mental healthcare. Public discussions frequently hinge on vague complaints such as “too many people these days have ADHD, anxiety, autism…” without clear agreement on what exactly constitutes “too many,” or even a consensus on what overdiagnosis actually means.
Here are different sorts of things people have in mind when they use the term “overdiagnosis”:
Diagnosing conditions unlikely to cause significant distress or harm if left undiagnosed and untreated.
Diagnosing transient psychological distress that would resolve spontaneously.
Labeling minimally impairing cognitive and behavioral differences as disorders.
Mislabeling diagnostic errors as overdiagnosis.
Diagnosing subthreshold symptoms not meeting official diagnostic criteria.
Equating rising diagnostic rates directly with overdiagnosis without considering improved recognition and epidemiological estimates based on official criteria.
Expressing cultural anxiety about mental illness’s increased visibility.
Criticizing broadening diagnostic criteria in diagnostic manuals (DSM, ICD).
Assuming patients seek diagnoses because they are trendy or fashionable.
Viewing people as desiring the sick role and lacking genuine impairment.
Confusing overdiagnosis with overtreatment.
Expressing discomfort with viewing emotional distress through a medical lens.
Clarifying the fundamental role of diagnosis
The fundamental concern when it comes to clinical diagnosis is recognizing distressing, impairing, harmful, or at-risk states for which clinical interventions (psychotherapy, medications, brain stimulation, etc.) can be helpful. There is suffering and harm, or the prospect thereof, and healthcare professionals have something to offer. This something is usually access to treatment but also includes things such as assessment, work-up, clinical perspective, official diagnosis, accommodations, etc. Recognizing this fundamental role helps clarify and resolve common misunderstandings surrounding diagnosis.
Seen in this manner, the fundamental concern of diagnosis is not identification of disease entities or “disorder” judgments based on biological dysfunction, nor is it an issue of medicine-vs-psychology/social work turf wars. But it does bring up important questions of its own, questions that I take seriously. In our efforts to recognize distressing, impairing, harmful, or at-risk states for which clinical interventions can be helpful, what do we get wrong about psychiatric diagnosis and the medical model? How do these misconceptions, such as essentialist thinking, derail our sense of identity and make us vulnerable to iatrogenic harm? How can we best negotiate the practical trade-offs surrounding clinical diagnoses?
Even professionals opposed to diagnosis cannot avoid acknowledging distress and impairment warranting professional care. A staunchly anti-diagnosis psychotherapist may bemoan that too many are being diagnosed with “mental disorders” (and being prescribed medications), and yet the anti-diagnosis therapist also offers open-ended, weekly psychotherapy to people who are experiencing, by their own description, “normal, understandable reactions to life circumstances.” Why allegedly “normal, understandable reactions” require professional treatment offered by clinicians is a contradiction that is never satisfactorily addressed. What such a therapist is offering is the superficial comfort of normalization while also behaving in a manner that implicitly recognizes that the person’s suffering is such that it can benefit from clinical care, and hence their services.
“Overdiagnosis” in the strict sense (#1, 2, 3)
Overdiagnosis, strictly speaking, describes a scenario in which conditions unlikely to cause significant harm or impairment if left untreated are identified as targets for clinical treatment, usually through proactive screening efforts. Cancer screening exemplifies this clearly: routine tests can reveal slow-growing tumors that pose minimal threat to one’s life. Translated to mental health, overdiagnosis could refer to screening in healthcare settings that detects transient distress likely to resolve spontaneously, or labeling individuals with mild attention difficulties or social awkwardness that cause no or minimal impairment.
While this certainly happens, it’s not clear that the magnitude with which this happens merits the degree of concern and alarm that is directed towards it. The rates of clinical diagnoses of various psychiatric disorders generally do not exceed epidemiological estimates by a wide margin in the Western world, and if anything, they are usually below what is estimated.
To put this into perspective, in the UK Longitudinal Household Study, there were 12 times as many people with “undiagnosed distress” (symptoms severe enough to meet clinical diagnostic criteria) as there were people with diagnoses in the absence of clinically significant symptoms.
Clinically speaking, it is also rare for me to encounter a patient who I believe had been overdiagnosed in the strict sense; I may disagree with the exact diagnosis they have been given in the past, but the degree of distress and impairment is almost always to such a degree that some clinical diagnosis is warranted. We don’t have compelling evidence to suggest an ongoing epidemic of asymptomatic or unimpaired individuals receiving psychiatric diagnoses.
Self-diagnosis is a somewhat different story. Some self-diagnosis does happen in the absence of substantial impairment warranting professional help. However, such self-diagnosis is often conceptualized in non-disordered, non-clinical ways, such as neurodivergence. Hence, however people may feel about seemingly functional, nerdy, socially awkward individuals using the term “autism” to refer to themselves, it is not an instance of clinical overdiagnosis. It is an instance of people reclaiming a clinical concept and extending it beyond the clinical sphere because of its epistemic value, because it allows them to make better sense of who they are.
Misdiagnosis and subthreshold diagnosis (#4, 5)
Much public debate around overdiagnosis conflates the strict definition above with related phenomena, such as diagnostic error (misdiagnosis), diagnosis of subthreshold symptoms, diagnostic lumping of mild and severe impairment in the same category, and excessive treatment (overtreatment). If a person is incorrectly labeled as having a condition that they do not actually have (i.e. they don’t meet accepted clinical criteria for that condition), that’s not overdiagnosis; that’s misdiagnosis.
When some take issue with specific diagnoses such as ADHD, autism, or bipolar disorder being overdiagnosed, they are usually thinking of misdiagnosis. I personally do think that misdiagnosis is fairly common in this sense, but this also has to be understood within the context of poor inter-rater reliabilities of psychiatric diagnoses. I frequently re-diagnose patients I see, but I cannot pretend to be the one, true arbiter of correct diagnostic judgments.
What annoys me, however, is that this point is used by various professionals to push for their pet diagnoses. It is not uncommon for bipolar experts to say, “Most cases of ADHD diagnoses are actually bipolar disorder,” and for ADHD experts to say, “Most cases of bipolar disorder diagnoses are actually ADHD.” PTSD is another one of those disorders such that in the opinion of some professionals almost everything is PTSD or almost nothing is. I don’t have much to say about this phenomenon other than: Get over yourselves! There is far more to psychopathology than whatever corner of it you specialize in.
Some people endorse a version of mental illness denialism except that it is circumscribed in its application to select conditions. Some diagnoses are real for such folks, but trendy diagnoses like ADHD, PTSD, and autism are kinda suspect. People clamoring for ADHD, PTSD, and autism diagnoses are just trying to make a big deal out of ordinary life problems. Life is hard; people want it easy. They just need to suck it up; they need to learn to deal with it. Bring back stoicism and the stiff upper lip. Etc. Such sentiments, in my opinion, are products of a failure to recognize the distress and impairment that people experience and the value of clinical treatment in such instances
“Too many people are being diagnosed!” and cultural anxiety (#6, 7)
For many commentators, the judgment of overdiagnosis is entirely based on vibes and nostalgic comparisons to the 1980s. Increasing rates of mental health diagnoses do not provide actual evidence of overdiagnosis since this does not take into account whether higher rates reflect better awareness and recognition, and whether rates are catching up with epidemiological estimates based on current diagnostic criteria.
The cry of overdiagnosis reflects a particular kind of cultural anxiety over the public visibility of mental illness. Some people long for a past when mental disorders were invisible, confined within asylums or hidden by stigma. Some people want mental illness to be rare enough that containment can work as a strategy, and they can pretend that mental disorders don’t exist except when they are talking about them in a hushed and scandalous manner. (Some reject the concept of “mental disorder” altogether or do not view psychological suffering as a legitimate healthcare concern. For such folks, any diagnosis of a “mental disorder” is overdiagnosis.)
All this reflects cultural anxiety over our collective relationship with illness, sickness, disability. Seen in this manner, “overdiagnosis” functions as a shorthand for an unease with contemporary diagnostic culture, not a literal description of clinical excess. It’s easier to lament “overdiagnosis” than it is to recognize and process this anxiety in a productive manner. This is particularly so because these questions about the nature of illness are complex, and trying to answer them drives us to exhaustion.
Diagnostic manuals and overdiagnosis (#8)
Many blame overdiagnosis on the broadening of diagnostic criteria in consecutive revisions of the DSM and ICD. This is the Allen Frances style “Saving Normal” argument, still echoing in contemporary debates. Normality is under threat from out-of-control, ever-expanding classifications. I used to be sympathetic to this line of thinking, but the more I practice, the more I am of the view that the DSM criteria are rather conservative. DSM falls short of a comprehensive description of psychopathological presentations, and the categories are not broad enough to capture all clinically relevant forms of psychological suffering and disabilities for which diagnosis is needed. The criteria for specified diagnoses exclude a significant number of patients who experience considerable distress and impairment in their daily lives.
People fundamentally misunderstand the impetus behind increasing rates of psychiatric diagnosis—by and large, clinicians are not imposing labels or disorder judgments on reluctant patients who would rather not have a name for what they are experiencing (although it does happen). There is a tremendous demand and appetite for diagnostic labels as tools of understanding. The amount of suffering and disability is vast; our traditional folk-psychological categories have been inadequate for this task. The DSM legitimized psychological distress and disability as healthcare problems deserving of recognition and treatment. The profession was unprepared for the magnitude of clinically significant distress out there, and the diagnostic manuals have struggled to keep up.
Much clinical practice involves clinicians responding to real suffering inadequately captured by official criteria. This leads to stretching existing diagnoses or utilizing vague “unspecified” categories, not from diagnostic carelessness but due to genuine gaps in the diagnostic framework. Similarly, frontline clinicians, burdened by time pressures and limited therapeutic tools, often prescribe treatments such as SSRIs broadly regardless of the precise diagnostic criteria. PCPs do not particularly care whether a patient has “adjustment disorder with mixed anxiety and depression,” vs “major depressive disorder, with anxious distress,” vs “dysthymia” vs “generalized anxiety disorder” because all they are going to do is put the patient on Prozac.
Patient-led diagnoses (#9, 10)
“So which is it? Are people victims of professionals labelling them, or are they taking too much control over their medical labels?”
, The Age of Overdiagnosis Panic
Much discourse around overdiagnosis betrays a distrust of individuals’ abilities to accurately recognize their own impairments. Patients generally seek diagnoses willingly, finding validation and clarity in diagnostic labels. Critics of diagnostic culture sometimes imply patients are either passively labeled by professionals or aggressively self-diagnosing driven by social media and disability activism.
A lot of public criticism of “overdiagnosis” does not strike me as very patient-centered. ADHD may be trendy, for good or bad reasons, but that should not stop us from acknowledging a patient’s lifelong struggles with inattention and focus, with resulting negative emotional consequences. DSM may or may not pathologize normal grief, but we should not clinically disregard a patient whose life has been paralyzed by grief for years.
A curious feature of this debate is that some clearly acknowledge that an important motivation is economic and pertains to scarcity of public resources. The UK government seems to think, for instance, that it cannot afford increasing rates of clinical treatment and disability benefits. Rather than transparently stating, “We wish we had sufficient resources to care for everyone, but we must prioritize severe cases,” many stakeholders instead gaslight those with milder impairments, implying that these individuals exaggerate or misunderstand their conditions when there is little evidence that it is happening at a mass scale or that this accounts for the rising demand for disability benefits.
Overtreatment (#11)
Sometimes people talk about overdiagnosis when they actually mean “overtreatment.” Is there overtreatment in the sense that there are many people who take treatment who are not helped by it? That is inevitable. No treatment is perfect. Unless the Number Needed to Treat (NNT) is 1, there will always be “overtreatment” by this definition. The NNT of aspirin for preventing cardiovascular events in diabetics in 153. We must treat 153 diabetic individuals with aspirin to prevent a single cardiovascular event in one person. People deserve to make informed choices about treatment, and people deserve access to care. And people deserve to know about the limitations of treatments and potential harms. Medical treatment is not simply about preference. It is guided and informed by clinical research, considerations of risks vs benefits, public safety considerations, cost effectiveness, public resources available for healthcare and health insurance, etc. But within these larger constraints, there is still a lot of flexibility and room for personal preference. In any case, issues around overtreatment bring up distinct considerations and cannot be subsumed under overdiagnosis.
Emotional suffering viewed through a medical lens (#12)
There is one aspect of what gets covered under “overdiagnosis” that I’m somewhat sympathetic to. This is the old “medicalization” criticism, used in sociological and philosophical critiques to argue that human suffering is now being frequently and inappropriately interpreted through a medical lens. Clinical conceptualizations and treatments are not universally helpful. For some they will be useless, and for some harmful. Some people are better off trying to understand and manage their suffering through non-clinical approaches: self-help, spiritual, lifestyle, coaching, social safety net, political, financial, existential, peer support, etc. The problem is that we don’t have a good way of knowing in advance who will benefit from the clinical approach and who will not. Epidemiological estimates of mental disorders provide a rough upper limit on who could stand to benefit from clinical care, but only a subset will actually benefit.
If clinical diagnosis and treatment are understood as one way of conceptualizing and helping a condition and not necessarily the only way, and not automatically the best way of helping to achieve all or indeed any desired outcomes,1 then this necessitates a robust pluralism of conceptualization and interventions. Diagnosis is a partial perspective on a person’s challenges.
The pluralistic approach offers a productive way forward. Clinical diagnosis is just one perspective among several—medical, psychological, existential, social, spiritual—that can effectively address individual distress. Different contexts demand different approaches. Recognizing this plurality does not negate the value of diagnostic categories, which serve essential pragmatic functions. Instead, it encourages clinicians and society at large to remain flexible, responsive, and open to multiple understandings of human suffering. To reduce this plurality of perspectives to a single dominant narrative, whatever that may be, is to impoverish our existence and to deprive us of the tools we need to make sense of ourselves in relation to our worlds.
Clinical concepts are appealing because they offer a useful hermeneutic lens through which to look at our problems. If we are over-relying on this lens, the answer is not to artificially restrict its use and deny people access to clinical care, but to develop other frameworks that can offer people similar or better utility and facilitate more informed and mindful use of the medical lens.
The answers I ultimately advocate for are thoughtfulness, conceptual clarity, person-centered care, pluralism, and realistic expectations around clinical treatments. “Overdiagnosis” is a blunt verdict offering blunt solutions. We deserve and can achieve better.
See also:
Ahmed Samei Huda, The Medical Model in Mental Health: An Explanation and Evaluation (p. 14):
“The medical model can thus be viewed as follows: ◆One way of conceptualizing and helping a condition and not necessarily the only way. ◆Not automatically the best way of helping to achieve all or indeed any desired outcomes. ◆A model that is willing work in conjunction with other models, in a multi-model, multidisciplinary way (i.e. with others such as carers, other professionals), each contributing a part to achieve desired outcomes. ◆A model that at times may be unhelpful or even harmful for certain conditions.”
Thank you, Awais, for your careful anatomizing of the nebulous term, "overdiagnosis." You rightly point out the many senses in which this dubious charge is leveled against psychiatry and psychiatric diagnosis. I would like to comment specifically on your category #8 (expanding or loosening of diagnostic criteria), since this question has been studied systematically by Fabiano & Haslam [https://doi.org/10.1016/j.cpr.2020.101889] Their meta-analysis concluded that:
* Criteria for diagnosing mental disorders did not loosen from DSM-III to DSM-5.
* No post-DSM-III revision produced significant diagnostic inflation or deflation.
* Certain mental disorders have substantially inflated or deflated.
* Some critiques of diagnostic expansion in the DSM have been over-stated.
Specifically, they found "no overall change in diagnostic stringency from DSM-III to DSM-5" and concluded that "Although serious concerns have been raised about diagnostic inflation or “concept creep” in the DSM, these concerns may have been overstated."
I would also like to make a broader point, more in the nature of philosophy than statistics. In your article, I found four instances of the phrase, “distress and impairment.” I count at least 7 additional instances in which you use the term, “impairment.” As you know, I have been banging on about “suffering and incapacity” for at least the past 45 years, as the most clinically important feature of disease—not constituting an “essential definition” of disease, but the aspects of disease of greatest concern to both patients and clinicians.
If I am correct in this assertion, it seems that the most clinically relevant question to ask of our diagnostic categories is this: how well do they identify clinical conditions characterized by marked “distress and impairment”; or, as I would phrase it, marked suffering and incapacity. The DSMs usually invoke the phrase “distress or impairment.” I would prefer the tighter phrase, “distress and impairment.” Either way, I do not believe critics of psychiatric diagnosis have shown that psychiatrists using DSM-5 criteria are “overdiagnosing” substantial distress and/or impairment in the emotional, cognitive and behavioral realm. If anything, in my view, many people in our society are struggling with high degrees of suffering and incapacity and, alas, are not receiving professional evaluation or treatment.
Ronald W. Pies, MD
Robert Chapman made really good points in their book Empire of Normality. Changing demands in society, in regular school and in commonly available jobs, result in a higher number of people actually experiencing significant difficulties. It's pointless to look for some objectively valid line between "healthy" and "disability" or "disorder" that stays the same regardless of society's demands, because there's no such line drawn in nature: Criteria WILL shift as society shifts.
We could argue that there's something wrong with a society in which a huge number of people fits psychiatric diagnostic criteria, and that politicians and other powerful people should set long-term goals to reduce the number of people who struggle like this (obvs that's never gonna be zero, but it can still be higher or lower). But that's actually quite different (though I'm not sure that everyone understands that it's quite different) from saying there's an over-diagnosis.
Like, we could have a long-term goal to reduce the number of people with cardiovascular problems in society, because we know that environmental factors play a big part. We could have a long-term goal to reduce certain types of cancer, etc. But this is different from saying there's an over-diagnosis.
I blogged about the book here https://jeppssonphilosopherauthor.blogspot.com/2023/12/review-robert-chapmans-empire-of.html