Sexual Boundary Violation Index: A Validation Study

This paper explores the psychometric properties of the Boundary Violation Index (BVI), a screening instrument designed to assess the attitudes, thoughts, and behaviors of physicians at risk of sexual misconduct with patients and staff. Items for the BVI were selected and validated using a two phase process of administration of the tool along with the Sexual Addiction Screening Test (SAST) to physicians referred to a CME course for boundary violating behaviors (n = 60 and 272) along with a control group (n = 118). Criterion-related validity in relation to the SAST was strong (r = 0.68, p < 0.001) and construct validity was demonstrated by the difference between intervention and comparison group BVI scores (p < 0.001). A BVI score of ≥ 6 for interpreting substantive risk had a sensitivity of 83% and specificity of 81%, and represented a greater than 20-fold greater risk (Exp B = 20.5, 95% CI 11.8–35.7, model p < 0.001) for membership in the intervention group. The BVI offers promise as a preliminary tool for identification of physicians at risk for boundary violating behaviors and may have utility for medical education and/or monitoring purposes.