In my previous post, I focused on Tyrer and Mulder's critique of the construct of borderline personality disorder (BPD). In this post, I would like to cover Tyrer and Mulder’s critique of the categorical approach to personality disorder classification. They call for its replacement with a dimensional system, such as is found in the Alternative DSM-5 Model for Personality Disorders (AMPD) found in the back of the DSM-5 or in ICD-11.
Categorical classification schemes divide the world up into discrete, non-overlapping categories. An ideal categorical system is comprehensive, so that it covers the entirety of what is to be classified—there is no “other specified” or “unspecified” (DSM-5) or “not otherwise specified” (DSM-IV) category. The categories are mutually exclusive. In the standard 52-card deck (as long as we exclude the jokers), for example, the four suits function like categories; a card either is a heart, spade, diamond, or club, and no card is sort of one suit or belongs to two or more suits. A dimensional approach to classification, in contrast, deals with continua, with spectra. Classification is more like opening up an image editor and selecting a shade of color along one or more dimensions. A categorical approach to personality disorder leads to the view that there are distinct, crisply delimited personality disorders, so that it would make sense, for example, in a clinical case consultation for two practitioners to get into an argument over whether a patient has borderline personality disorder, antisocial personality disorder, or narcissistic personality disorder. From a dimensional perspective, however, the question would be a bit like asking whether that mysterious newcomer who is the subject of every other 1990s high school movie is really a jock, geek, or theater kid.
The problems with categorical approaches to personality disorder are well known and have been consistently pointed out for at least 30 years. Their virtue, of course, is their apparent simplicity, since, supposedly (and I want to stress that this point is by no means self-evident) they are easier to use in clinical practice. In any case, at least in my own experience, one can easily graduate from a master’s-level program in the mental health field without learning the first thing about any of these debates. I was already aware, before reading this book, that, going back to at least the late 1990s, there was a wide consensus that DSM-5 should substitute a dimensional model of personality disorders for the categorical system transmitted more or less unchanged from DSM-III to DSM-IV. The persistence of the categorical approach is more of a collective action problem than the product of scientific defense, akin to the persistence of the practice of resetting the clocks twice a year (everyone hates it, but no one can agree whether Daylight Saving Time or Standard Time should be the default).
What happened in the 2000s was a decade of political maneuvering and professional infighting described in fascinating detail in Zachar, Krueger, and Kendler, “Personality disorder in DSM-5: An oral history” (2016). (This article is required reading for anyone interested in the micropolitics of psychiatric classification; I also think it would make a great movie, with Woody Harrelson playing Allen Frances.) In a nutshell, most people involved in the process seem to have wanted a dimensional system, but different factions competed to promote their system. The confusing result of that process was two different sections on personality disorder in DSM-5, the main section, which presents the old categorical system more or less unchanged, an optional alternative dimensional model included in the back of the book.
What I appreciate about Tyrer and Mulder’s book is how clearly and devastatingly they lay out the case against the categorical approach while defending the clinical practicality of the dimensional approach in ICD-11. You can find much of the following in the standard clinical handbooks for personality disorder, but not nearly in such a clear and accessible form. In a nutshell:
• The old set of personality disorder categories was largely derived, with the exception of the BPD and schizotypal categories, from observations of Hamburg-area sex workers made by German psychiatrist Kurt Schneider in the 1920s. (If any of my philosophy friends are reading this, yes, it’s that same Schneider that Merleau-Ponty is always quoting.)
Psychiatrists in clinical practice have gone in a different direction down the Schneiderian path. Rather than carry out research to test the value and utility of this clinical/prostitution-derived classification, it [sic] has embraced it uncritically, changed a few names here and there, described each category as though it were unique and definitive when in fact it is swamped by comorbidity of other personality disorders, and then wonder why mixed personality disorder or PD-NOS (personality disorder not otherwise specified) is so commonly used in clinical practice (Tyrer and Mulder, p. 7).
• An enormous proportion of people who meet criteria for one personality disorder meet criteria for at least one other, if not several others. For example, depending on the study, up to 95% of patients with a BPD diagnosis also meet criteria for at least one other personality disorder (p. 34). (Indeed, one of the two main professional handbooks on personality disorders, Livesley and Larstone’s 2018 edited volume, notes that, on average, patients who meet criteria for one personality disorder meet criteria for 2.8 to 4.6 personality disorders [see Chapter 4 of that volume]!) These facts make a mockery of the term “comorbidity.” Indeed, on a conceptual level, it is confusing at the very least to see how someone could have more than one personality disorder
• The inherited categories of personality pathology are completely divorced from psychological research on “normal” personality, which has embraced a dimensional approach and has, more or less, settled on a consensus of roughly five underlying personality factors, extraversion, agreeableness, neuroticism, conscientiousness, and openness to experience. The result is that the personality disorder field remains cloistered in Schneider’s time capsule, removed from contact with mainstream personality science. It is as if (this is my analogy) there were two branches of medicine, physiology/anatomy and pathology, but they used entirely different concepts, conceptual structures, and modes of reasoning. (It is worth noting, however, that not everyone working in the field thinks that the sciences of normal personality and pathological personality can be joined up in this way, at least not at present.)
• The categorical approach leaves little room for grasping the wide spectrum of personality disturbance, from mild sub-syndromal “personality difficulty” to severe personality pathology. In the categorical approach, people either have personality disorders or they do not. This binary approach, Tyrer and Mulder argue, contributes to the stigmatization of personality disorder and to the well-established phenomenon that clinicians massively underdiagnose personality disorder, as established by comparisons between epidemiological estimates of personality disorder prevalence and rates of formal diagnosis in clinical settings, which reveal a difference of “10-20 fold” (p. 99).
• The dimensional approach need not be so complicated as to be useless in general clinical settings (a standard objection, and, as Zachar et al discuss, one of the arguments that helped torpedo the DSM-5 revision efforts). Indeed, the ICD-11 approach is quite simple. Patients are evaluated along two lines: 1) three levels of severity (mild, moderate, or severe) and 2) five trait domains (negative affectivity, detachment, dissociality, disinhibition, and anankastia). (Anankastia is somewhat similar to what DSM’s “obsessive-compulsive personality disorder.”) A diagnosis can be made on the basis of severity alone, and any number of trait domains can be added to the diagnosis (pp. 13-25). Frankly, I would guess this is more or less how most clinicians already reason when assessing for personality disorders. A sub-syndromal category of “personality difficulty” can also be registered using the ICD’s Q-code system.
Again, none of this is new. Much of this argument can already be found, indeed in greater detail, in John Livesley’s seminal 1998 “Suggestions for a Framework for an Empirically Based Classification of Personality Disorder.” What makes it practical from a clinical point of view is that the practitioner begins simply by diagnosing personality disorder. This is the ICD-11 definition of personality disorder:
Personality refers to an individual’s characteristic way of behaving, experiencing life, and of perceiving and interpreting themselves, other people, events, and situations. Personality Disorder is a marked disturbance in personality functioning, which is nearly always associated with considerable personal and social disruption. The central manifestations of Personality Disorder are impairments in functioning of aspects of the self (e.g., identity, self-worth, capacity for self-direction) and/or problems in interpersonal functioning (e.g., developing and maintaining close and mutually satisfying relationships, understanding others’ perspectives, managing conflict in relationships). Impairments in self-functioning and/or interpersonal functioning are manifested in maladaptive (e.g., inflexible or poorly regulated) patterns of cognition, emotional experience, emotional expression, and behaviour.
This is actually very similar to Livesley 1998, who had boiled personality disorder down to two general features, “chronic interpersonal difficulties and problems with a sense of self or identity.” Unlike DSM-5, the ICD does not require that the difficulties have been “relatively stable across time, with onsets that can be traced back to at least adolescence or early adulthood” (DSM-5, p. 761); they need only have been present for two years, allowing for late-onset personality disorder or pediatric personality disorder (Tyrer and Mulder discuss this on pp. 116-117).
After an initial impression of personality disorder, the clinician assigns a severity label of mild, moderate, or severe, based on impairment in functioning; a sub-clinical “personality difficulty” code is also available (pp. 11-14). The clinician then may add one or more of the five trait domain specifiers. (It is worth noting that, at the eleventh hour, the ICD-11 did end up including a “borderline pattern” code alongside the five trait domains; Tyrer and Mulder are none too pleased about this.)
There are some serious objections to the views I have outlined in this post and the last. In my next post, I will wade into more critical territory. I plan to focus mainly on the implications of Tyrer and Mulder’s call for expanding assessment and diagnosis of personality disorders. The authors have the merit of including a chapter on stigma, something they note is missing from the main handbooks on personality disorder (pp. 98-111). I do worry that some of their proposals would actually increase stigma. I also worry that the authors leave under-addressed the question of the difference between personality disorders and other psychiatric conditions, which used to be institutionally enshrined—even if it was never really conceptually work out—in the Axis I/Axis II distinction of DSM-III and DSM-IV. Historically speaking, I think the clearest illustration of this problem is the enduring turf war between the DSM personality disorder working groups and the mood disorder working groups over persistent depression—is it depressive personality disorder or persistent depressive disorder (dysthymia)?
Comments
Post a Comment