Tuesday, March 16, 2021

Loose Associations Between Various Clinical Attributes of Patients Sharing a Diagnosis

Doctors are being pressured more and more to make quicker and quicker clinical decisions using information from studies that show correlations between certain symptoms and certain genes and other patient attributes, both psychological and physical. They are at risk of becoming sucked into a Lake Wobegon effect - the fictional town where it was said everyone’s intelligence is above average. Clinicians should still  assess each individual separately in order to avoid conclusions based on the ecological fallacy (thinking all patients with a particular disorder react exactly the same as the average patient with the disorder). In fact, most people with a certain characteristic fall outside of the average value when looking at another characteristic that seems to correlate with it.

Some people object to personality disorder diagnosis for a similar reason: each person with these disorders presents somewhat differently than the “average” person with the disorder.

However, this whole problem does not mean that we should completely avoid making diagnoses or looking for other correlating characteristics which are likely to be found in the average person in a group. This information can provide a clinician with clues as to where to start an investigation. The diagnosis of borderline personality disorder, for instance, correlates strongly with the spoiler role developed in families in which the parents have conflicts over the role of being a parent. This knowledge is useful in telling a physician or therapist what questions to ask the patient when exploration of the relevant factors begins. 

It does not, however, mean that an initial inquiry is guaranteed to be fruitful, only that in many cases it will be, so starting there may save the clinician a lot of time looking at things that will turn out to be irrelevant. For it to be useful, however, the doctor must treat an initial idea as an hypothesis, and be completely open to the possibility that the hypothesis may turn out to be wrong.

Obviously, the strength of any discovered correlation affects how soon one should start looking for it. Much of the personality disorder literature these days is clinically useless because the correlations “discovered” are either too small, too obvious, or too unimportant to mean much of anything in helping the clinician develop some sort of psychotherapeutic strategy. A study by Wildey et. al. in the  August 2020 issue of the Journal of Personality Disorders found correlations between so-called externalizing disorders like antisocial personality disorder or substance abuse and the negative quality of a person’s relationships using multiple assessment methods. Ya think?

Some studies even look for associations between qualities that are actually intrinsic to the definition of a personality disorder and the disorders themselves. A rose is a rose last time I looked. Like people with BPD being very reactive. I kid you not.

Even dumber is looking at so-called “mediating” variables. For example, one study in the Journal of Personality Disorders by Sato et. al, in the April 2020 issue found that “rejection sensitivity” was mediated by attachment anxiety, the need to belong, and self-criticism. You mean if you think negatively about yourself but still want to fit in, that might lead to anxiety? I know academics face the need to publish or perish, but as I am fond of saying, “No Sh*t, Sherlock.”