Reliability of the ICD-11 personality disorder severity ratings and diagnosis.

Abstract

The present study aimed to investigate the interrater reliability of the dichotomous and dimensional personality disorder (PD) diagnoses based on the overall severity assessment on a rating form consisting of 18 anchored items encompassing diagnostic requirements of the International Classification of Diseases 11th Revision (ICD-11). We also aimed to examine the extent of consistency within the diagnostic requirements grouped under the domains of self- and interpersonal functioning, specific manifestations of personality dysfunction, and distress and impairment in psychosocial functioning. Our data involved a total of 184 inter-ratings of 46 consenting patients by the same set of four clinicians. The chance-corrected agreement levels were estimated at intraclass correlation coefficient (ICC) = 0.89 for the overall severity composite, ICC = 0.83 for the dimensional PD diagnosis and Fleiss' kappa = 0.77 for the dichotomous PD diagnosis. Internal consistency analysis of the overall severity composite and the domain composites revealed Cronbach's alpha coefficients approaching or exceeding 0.90 level. Our findings suggest that the diagnostic requirements listed in the ICD-11 and related documents for the severity determination in PD compose an internally consistent set. With the guidance of a rating form comprised of anchored items covering this set, competency-level clinicians are likely to perform reliable evaluations of the severity of personality disturbance, and dimensional and dichotomous PD diagnoses. The development of semi-structured interviews that would further facilitate the task of inspecting and rating each diagnostic requirement reliably will possibly enhance the implementation of the ICD-11 classification for PD around the world.