Abstract
The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM–5) presents an alternative model for personality disorders in which severity of personality pathology is evaluated by the Level of Personality Functioning Scale (LPFS). The Structured Interview for the DSM–5 Alternative Model for Personality Disorders, Module I (SCID–5–AMPD I) is a new tool for LPFS assessment, but its interrater reliability (IRR) has not yet been tested. Here we examined the reliability of the Norwegian translation of the SCID–5–AMPD I, applying two different designs: IRR assessment based on ratings of 17 video-recorded SCID–5–AMPD I interviews by five raters; and test–retest IRR based on interviews of 33 patients administered by two different raters within a short interval. For the video-based investigation, intraclass correlation coefficient (ICC) values ranged from .77 to .94 for subdomains, .89 to .95 for domains, and .96 for total LPFS. For the test–retest investigation, ICC ranged from .24 to .72 for subdomains, .59 to .90 for domains, and .75 for total LPFS. The test–retest study revealed questionable reliability estimates for some subdomains. However, overall the level of personality functioning was measured with a sufficient degree of IRR when assessed by the SCID–5–AMPD I.
Acknowledgments
We are indebted to the patients who volunteered in this study; the staff members at Akershus University Hospital, Lovisenberg Hospital, Namsos Hospital, Oslo University Hospital and Sorlandet Hospital; and the members of the Research Group at Department of Personality Psychiatry and the National Competence Services for Personality Psychiatry.