Classification systems: categories, dimensions and their limits

Transcript

SoheeTwo patients in a depression trial. Both diagnosed with major depressive disorder by independent interviewers who agreed. Neither has any other diagnosis.

RyanThe first one can't sleep, has lost weight, and feels restless. The second sleeps too much, has gained weight, and moves slowly.

SoheeSame diagnosis. Almost opposite pictures.

RyanWelcome to EPPP Prep, the podcast. I'm Ryan.

SoheeI'm Sohee. Today we're on classification systems: why we sort psychological problems into categories at all, the two big systems, and what gets lost when we draw yes-or-no lines.

RyanStart with why. OpenStax says a proper diagnosis lets professionals use a common language with others in the field, helps communication with patients, colleagues and the public, and is an essential element in guiding proper treatment. That's why classification systems are necessary.

SoheeIn the United States, the system most mental health professionals use is the Diagnostic and Statistical Manual of Mental Disorders, published by the American Psychiatric Association. Its first edition came out in 1952.

RyanAnd it has changed a lot. OpenStax notes the first two editions listed homosexuality as a disorder, and in 1973 the American Psychiatric Association voted to remove it. Classification systems reflect their times.

SoheeIt's also grown. The first edition included 106 diagnoses in 130 pages. The third edition had 265 diagnoses in 886 pages. And critics worry that new diagnoses might overpathologize the human condition, turning common human problems into mental illnesses.

RyanThe second big system is the International Classification of Diseases, the ICD, published by the World Health Organization. OpenStax's glossary describes it as an authoritative index of mental and physical diseases, including infectious diseases, and the criteria for their diagnosis.

SoheeAnd besides clinical use, it's used to examine the general health of populations and to monitor the prevalence of diseases internationally.

RyanHere's the comparison the exam likes. A study comparing the two found that worldwide, the ICD is more frequently used for clinical diagnosis, while the American manual is more valued for research. And in the United States, the American manual is the system of choice.

SoheeLet's pause and think. A global health agency wants to compare rates of depression, tuberculosis and heart disease across 40 countries, using one shared coding system. Which system?

RyanThe ICD. Mental and physical diseases in one system, built for international health monitoring. The American manual covers mental disorders only.

SoheeNow, both systems mostly use categories: you meet criteria, or you don't. The National Institute of Mental Health lays out the tradeoffs in its description of a research framework called RDoC, the Research Domain Criteria.

RyanAnd it's fair about the benefits first. Diagnosing by the number and type of symptoms provides benefits like reliability and ease of diagnosis across many settings. But it has come at the cost of numerous tradeoffs.

SoheeTradeoff one: heterogeneity, because of the varied ways people can qualify for a symptom-based diagnosis. Two people can be diagnosed with the same disorder despite having few, or sometimes no, symptoms in common.

RyanThat's our opening pair. And SAMHSA's depression symptom list shows how it happens: insomnia or sleeping too much, appetite changes leading to weight loss or weight gain. The list runs in both directions.

SoheeAnd it matters beyond the clinic. NIMH notes that the psychological and neurobiological mechanisms may differ greatly among patients who share little to no specific symptoms. One of our practice items has a neuroimaging lab finding very different patterns of brain activity in patients who all carry the same agreed-upon diagnosis.

SoheeTradeoff two: comorbidity. People who meet criteria for one mental disorder often meet criteria for others too. Remember the distinction: heterogeneity is variety within one diagnosis. Comorbidity is one person with more than one diagnosis.

RyanAnd it's common. OpenStax reports that more than half of people with a primary diagnosis of a depressive disorder are estimated to have some sort of anxiety disorder, and the reverse holds too.

SoheeNIMH adds a research wrinkle. To reduce heterogeneity, researchers often exclude people with comorbid disorders and study only those with a single, pure diagnosis. That raises the worry that dimensions shared across disorders aren't being properly studied.

RyanTradeoff three: arbitrary cutoffs. NIMH says clinical criteria, though created through expert consensus, are somewhat arbitrary, and research shows important similarities between people who meet the criteria and those who narrowly miss the cutoff.

SoheeSo, in NIMH's words, to understand the full spectrum of mental health and illness, it's important to adopt dimensional conceptualizations. Sorting people into well and ill by symptom counts can obscure how problems gradually emerge across development, how risk factors work, and how to measure outcomes in prevention and treatment trials.

RyanAnd one more subtle point. Symptom-defined categories, by definition, reflect well-advanced problems that are harder to treat. NIMH adds that problems typically shift gradually toward illness, so understanding that path has the potential to head off disorders with the right interventions.

SoheeSecond pause and think. In a depression study, participants with five or more symptoms are coded ill and those with four or fewer are coded well. The researcher notices that people right below the cutoff closely resemble those right above it, wherever the cutoff is set, and wants to keep every participant's severity. Best change?

RyanRate symptoms along a continuum instead of present or absent. Moving the cutoff up or down still splits people into two groups. Only the line moves.

SoheeA related item: a youth prevention program wants to spot teenagers whose anxiety is gradually increasing, long before any would meet criteria for a disorder, and track each teen over time. Same answer. Continuous ratings show gradual change in every teen. A yes-or-no diagnosis would show nothing until it's well advanced.

SoheeSo what is RDoC exactly? It began in 2009, and NIMH is clear that it's not meant to serve as a diagnostic guide, or to replace current diagnostic systems. It's a set of research principles.

RyanIt looks at mental health in terms of varying degrees of dysfunction in basic psychological and biological systems. Six major functional domains, each studied along the full range from normal to abnormal. And it encourages researchers to integrate several units of analysis: behavior, genetics, physiology and self-report.

SoheeNIMH describes it as reversing the usual perspective. Instead of starting from symptoms and then hunting for causes, it starts from deviations in normal functions that could account for particular symptoms, often across several diagnoses at once.

SoheeThe other dimensional system to know is HiTOP, the Hierarchical Taxonomy of Psychopathology. A review in Dialogues in Clinical Neuroscience describes it as a hierarchical, dimensional system for organizing psychological symptoms, built from factor-analytic studies.

RyanBroad spectra sit at the top, with narrower dimensions below. For example, the Internalizing spectrum includes subfactors called Fear and Distress, which in turn include more specific problems like social anxiety or insomnia.

SoheeTwo strengths the review highlights. Continuous symptom dimensions show higher reliabilities than categorical diagnoses. And a dimensional approach can accommodate heterogeneity and the way conditions co-occur. Generalized anxiety and major depression, which often occur together, share features captured by the Distress subfactor.

RyanWith an honest caveat: HiTOP is a work in progress and will be updated on the basis of new data.

SoheeSo in a question asking which system most clinicians worldwide use for diagnosis, RDoC and HiTOP are distractors. One is a research framework, and the other is still being built.

RyanBefore we recap, one value that matters more than any label. OpenStax reminds us that a psychological disorder is not what a person is. It's something a person has, through no fault of their own. Which is why it avoids words that turn people into their diagnoses.

SoheeLet's recap. We classify for a common language and to guide treatment. The American manual is the system of choice in the United States, and worldwide it's more valued for research. The ICD is used more for clinical diagnosis worldwide, covers mental and physical diseases, and monitors population health.

RyanCategories bring reliability and ease. Their costs are heterogeneity, where people share a diagnosis but few symptoms, comorbidity, where one person meets several, and arbitrary cutoffs.

SoheeDimensional approaches rate severity along a continuum. RDoC is a research framework, not a diagnostic guide. HiTOP is a hierarchical, dimensional model that's still a work in progress.

RyanIf classification has felt dry, remember those two patients from the start. These distinctions are about seeing the person accurately.

SoheeAnd the same goes for you. One category on a score report is not the full picture of what you know. Keep building. See you next time on EPPP Prep.

Loading the read-along timings…