Reading focus
Syllabus: When Do Personality Patterns Become Disabling?
Name the kind of stability behind each estimate before interpreting it.
Why you’re reading it. Replace a simple stable/changeable verdict with comparisons of diagnoses, criterion levels and relative ordering.
A question to consider. How could a sample show stable rank ordering, declining average difficulty and substantial diagnostic remission at the same time?
Why this reading
Claims that personality disorders are enduring can refer to different outcomes. This review separates retention of a diagnosis, change in average criterion levels and consistency in people’s relative positions. It makes the definition of stability part of the question rather than an assumption.
Before you read
Diagnostic stability depends on a threshold. Mean-level stability concerns an average, while rank-order stability concerns relative ordering. These can differ without contradiction. A pooled estimate combines selected studies; it is not an individual prognosis or a measure of treatment benefit.
Read for
- Compare eligibility, follow-up intervals and assessment methods. How were multiple reports from the same dataset handled?
- Inspect which diagnoses contribute enough evidence to each analysis and how much results vary between studies.
- Distinguish dimensional counts of diagnostic criteria from the functioning and trait dimensions in newer classification models.
Consider afterwards
- Construct a small hypothetical dataset in which everyone improves but rank ordering remains unchanged. What happens if some people cross a diagnostic threshold?
- What information about a person’s course is lost when only the first and last assessments are compared?
Editorial context
The search ended in December 2022. Forty studies entered the review, with high heterogeneity and limited evidence for some comparisons. Using first and last observations cannot describe the intervening course, and reliability affects apparent stability. The review excludes studies designed to test treatment effectiveness; it does not demonstrate untreated change or identify what caused improvement. Its criterion-count dimensions should not be treated as a direct evaluation of ICD-11 severity.