On case conceptualization and diagnostic hypotheses

Many times, in clinical decision-making one must take on binary choices that affect a patient’s treatment course. An obvious example of this is whether the clinician must assign a diagnosis or not. In evidence-based case conceptualization however, the clinician is forced to reduce the patient into distinct and exclusionary variables, behaviors and environmental factors that map out mechanistic explanations for symptom maintenance and disfunction. In this case, the binary decision is not focused on deciding which variables sustain or contradict a certain hypothesis, but rather focused on deciding whether a variable is relevant or not for the case formation.

For example, suicidal ideation may be thought of as a continuum of severity, ranging from passive ideation (“I don’t want to wake up tomorrow”) to active ideation and finally to having an active plan to commit. For clinical practice, suicide “severity” is a descriptive variable that does not hold value; suicide severity is a proxy for the probability that the patient commits suicide within the next year. The best predicting factor of death by suicide is whether a patient has had a suicide attempt in their life history. Would it be relevant to include “suicidal ideation” in the case conceptualization if the patient has active ideation without plan to commit with no history of suicidal behaviors thought their lifetime? Would it be relevant to include “suicidal ideation” in the case conceptualization if the patient has had a numerous history of suicide attempts but is currently on SSRIs and has had no ideation for the past months?

Evidence-based case conceptualization and assessment must function as a predictive tool that, along with patient-therapist discussion, would benefit from Bayesian probability methods for dealing with those kinds of binary decisions. Following the suicide example, a type 2 error is worse than a type 1 error.