Background: The International Classification of Diseases, 11th Edition (ICD-11) includes a new personality disorder (PD) diagnosis, with five optional trait qualifier domains (Negative Affectivity, Detachment, Dissociality, Disinhibition, Anankastia) that can be assigned. Previous research has validated self-report measures of ICD-11 PD traits (see Oltmanns, 2021), but little research exists on other measurement modalities and how they converge. This study examined the convergence of clinician-ratings, self-reports, and lay informant ratings on the five trait domains using a multi-trait multi-method approach.
Method: We used a community mental health sample from New Zealand (n=336). Participants were rated on traits after a series of structured clinical interviews, completed the Personality Inventory for ICD-11 (Oltmanns & Widiger, 2018); and nominated an informant who completed the PiCD-Informant Version. We also examined associations between the various trait domain measures with other measures of traditional DSM-5 PDs, and the five-factor model of personality, to evaluate the comparability of validity of each of these approaches.
Results: We estimated a hierarchical CFA to evaluate the five-factor structure of the ICD-11 PD trait model, using the clinician-rating, self-report, and informant-report measures. We also estimated three method factors that were orthogonal to the trait factors. The overall model was associated with good fit to observed data (CFI=.965, TLI=.939, RMSEA=.053, SRMR= .045). All measures loaded meaningfully (.30+) on their respective latent trait factors, with most loading on their relevant method factors as well. The latent trait correlations ranged from .01 to .85 (median: .35). Furthermore, the trait measures correlated in a comparable manner with a range of conceptually relevant personality trait scales and traditional PD symptoms scores (ICCs = .67-.96).
Conclusion: The ICD-11 trait domain qualifier operationalizations show promising convergence, including when controlling for method variance, and have good support for validity. Clinicians are thus afforded a number of options for their assessments.