Diagnostic Accuracy of PID‐5‐BF for Screening Personality Disorders: A Cross‐Sectional Study in Clinical and Nonclinical Samples

Abstract

ABSTRACT The DSM‐5 Alternative Model for Personality Disorders introduced a dimensional approach to personality pathology, with maladaptive traits assessed by the Personality Inventory for DSM‐5 Brief Form (PID‐5‐BF). Despite extensive validation, evidence on its diagnostic accuracy and empirically derived cutoff scores remains limited. This study aims to compare PID‐5‐BF trait levels and profiles across individuals with personality disorders (PDs), clinical participants without PDs, and nonclinical participants and to evaluate its diagnostic accuracy and screening cutoffs. This cross‐sectional study included 1168 Hungarian participants: 867 nonclinical, 237 clinical without PD, and 64 with PD. Group differences were examined using one‐way ANOVAs, whereas domain profiles were evaluated with mixed‐design repeated‐measures ANOVA. Diagnostic accuracy was assessed using ROC analyses adjusted for age, sex, and education, and optimal cutoffs were determined using Youden's index. Participants with PD showed the highest pathological trait scores, followed by clinical participants without PD and nonclinical participants. The largest group effects emerged for disinhibition, total score, and detachment. For distinguishing PD from nonclinical participants, the adjusted AUC for the PID‐5‐BF total score was 0.849, with an optimal cutoff of ≥ 1.32. Domain‐specific cutoffs ranged from 0.60 to 1.60 (median = 1.20). Negative predictive values were high (mean NPV = 0.934), whereas positive predictive values were modest (mean PPV = 0.327). The PID‐5‐BF may be useful as a brief screening measure for identifying individuals requiring further personality assessment. However, its limited positive predictive value and weaker discrimination between PD and other clinical conditions preclude its use as a stand‐alone diagnostic instrument.