Psychometric properties of the 4-week version of the Borderline Personality Disorder Severity Index-5

This study evaluated the psychometric properties of a new version of the Borderline Personality Disorder Severity Index-5 that assesses BPD symptom severity over the course of 4 weeks instead of the standard 3 months: the BPDSI-5-4wk. Reliability and validity were evaluated in a mixed sample of patients with BPD (n = 92), patients with avoidant personality disorder (APD) as clinical control group (n = 16), and a non-patient control group (n = 20). The study demonstrated very high interrater agreement and test-retest reliability and acceptable to excellent internal consistencies of the BPDSI-5-4wk. Confirmatory factor analysis supported its assumed nine-factor structure. The BPDSI-5-4wk also showed very good construct (i.e. known-group) validity, as well as criterion-related (i.e. concurrent) validity when correlating the BPDSI-5-4wk with the BPDSI-5 (3-month version), the structured clinical interview for DSM personality disorders (SCID-5-P), and several other BPD- and other mental health-related self-report questionnaires. We additionally derived cut-off scores with high sensitivity and specificity for distinguishing BPD from clinical controls (21.02, 85 % sensitivity, 94 % specificity), from non-patient controls (10.50, 98 % sensitivity, 100 % specificity), and from both control groups combined (17.26, 93 % sensitivity, 92 % specificity). A reliable change criterion of 5.88 was established. Preliminary trial data additionally showed the BPDSI-5-4wk's sensitivity to change. The strong reliability and validity of the BPDSI-5-4wk support its value for detailed, dimensional assessment of BPD symptom severity over shorter time frames. This will facilitate frequent treatment response evaluations, rapid indication of follow-up treatment, and better alignment with shorter intervention and follow-up periods in clinical trials.