Wednesday, September 10, 2014

Combinatorics and the Folly of Diagnostic Labels for Personality Disorders

by Randi Dellosa
This post has been edited since it's first incarnation. I view DSM Personality Disorder classification and labelling as conceptually grotesque -- not only inelegant but ugly, preposterous and unscientific. Here's why.

The DSM provides a classification system based on checking off items from a list. If the number checked is greater than X then the person has a specific Personality Disorder (PD) -- with significant ramifications for their life.

For instance, according to the National Institute of Mental Health, by "the DSM-IV)[1], to be diagnosed with Borderline Personality Disorder (BPD), a person must show an enduring pattern of behaviour that includes at least five of the following [nine] symptoms...


Combinatorics and BPD

So just by counting the number of combinations of having 5 or more distinct symptoms out of 9, there are 9!/(5!4!) + 9!/(6!3!) + 9!/(7!2!) + 9!/8! + 1 = 256 different phenotypes of BPD, with markedly different symptoms -- as required by the diagnostic criteria of the DSM itself. For those criteria to be meaningful, this implies that not only life experiences are different amongst the 256 phenotypes of BPD, but also ways of being, and of viewing one's self and others. Indeed this diversity is reflected in personal experiences of people with BPD I have run across in online forums. It is often a source of frustration and possibly confusion for people with this diagnosis.

Nor does the counting of combinations of symptoms take into account differences in their severity, as long it is over a qualitative threshold. Neither does it include co-morbidities (more than one PD diagnosis), which I get to in the next post. But all individuals fitting into this mould get the same label.


Other Personality Disorders

Both DSM-5 and -IV list ten different specific PDs, with the third most commonly used diagnosis being another one "Not Otherwise Specified". Some disorders overlap in symptoms with others, which leads to high levels of co-morbidity, and inconsistent diagnoses. 

I will show the absurd folly driven by combinatorial explosion in the number of phenotypes the DSM defines for Personality Disorders as a class, if one takes its categorical system seriously to determine whether or not an individual has one or more PDs in the next post too. It is a catastrophe involving astronomical numbers of symptom combinations or phenotypes within the class of Personality Disorders. It is the death of any categorical system based on factorials with numbers up to 10 or so.

The standard, prevailing methods for diagnosing other PDs also use a checklist algorithm. Hence people with such diagnoses also have not only vastly different symptoms from others with the same diagnosis, but also vastly different experiences, ways of being, interacting as social beings, and attitudes to life.


Harms Obscuring Beauty

The oversimplification of not taking into account the variance or variability of people with the same diagnosis, leads to widespread cultural misunderstanding and type casting. It is hard to see how it does any good. Due to the astronomical numbers involved (as I show in the next post), either psychiatry's categorical diagnostic system groups people together based on the wrong criteria, or they cannot really be grouped at all.

In the context of psychiatry's labelling and our cultural use of these labels, individual experience gets swallowed by a hungry, munching monster, turning everything except the most undigestible parts into mush. I find it more than sad; indeed I see beauty in these varieties of human experience, beyond and within each label -- the variety itself is intriguing, and fascinating. It is part of what makes us human. 

To give one example of the harm of type casting, I quote at the end of this post a recent exchange on sociopathworld.  This type of situation is faced by people with BPD on a regular basis.


Online Research Ethics

Regarding my journey into anthropology of online voices, I have decided for now to stick to wholly anonymous posts -- those that lack even a pseudonym. The question of privacy of online texts on the public internet is not the same as individual privacy. It takes into account the connection, if any, between a text and an individual person.  

For instance Michael Thelwall writes "A simple but strong argument for researching published information on the public web without consent is that the object investigated is the publication and not the person."

To be conservative, online comments posted anonymously have no connection to an individual, by any reasonable measure, and can be treated as any other publicly available text: newspapers, blogs, research papers, wikipedia etc. This is a prevailing norm in the still unsettled area of online research ethics.


How labels harm -- an online excerpt


Anonymous writing to another poster: ... could you answer a few questions about borderline's for me? I recommended the book about Hitler called "The Psychopathic God." Are Borderlines:

  • A) Purposefully self sabotaging. Do they LIKE to set themselves up for failure?
  • B) They are extremely contradictory. It's ALL or NOTHING.
  • C) They can display "many faces," in over the course of a few minutes. Almost as though demon possessed. Look at the hundreds of "faces" "good" mother Casey Anthony expressed in the 1,000's of photographs taken of her.
  • D) They play sex games involving poop. I don't know WHY, but that has been a noted trait.

I don't know whether this is true in your case, and you've said you worked in a mental hospital setting. The hospital TA's are sadistic and often w orse than the patients, but you said you had a "responsible" interactive position as an "intake nurse." Is it so? You've held it together pretty good for a Borderline. Any explanations?


DoctorSciFi: Anon, regarding your post:

  • All your questions are ill posed. They are not even wrong. To be diagnosed with BPD one needs at least 5 distinct symptoms out of 9 possible ones to be checked off. 
  • By combinatorics this means there are 256 different types of BPD [even disregarding the actual severity of each symptom over a threshold]. Each person will have some symptoms and NOT have others, as demanded by the classification system itself.
  • Just because a person has a high degree of one symptom does not demand they have any other particular symptom at all. That is a requirement of the classification system itself.
  • So any question like Are Borderlines:"A) Purposefully self sabotaging. Do they LIKE to set themselves up for failure?" misses the point entirely.
  • Assuming for the sake of argument that A) fits into one of the 9 possible symptoms, it is not true that "borderlines are A)". Some are and some aren't.

-------------------
[1] The current edition, the DSM-5, retains the nine main symptoms of BPD with slightly different wording.

No comments:

Post a Comment