Development and validation of a screening instrument for borderline personality disorder (SI-Bord) for use among university students

Background: The screening instrument for borderline personality disorder (SI-Bord) consists of a 5-item self-reported questionnaire on the key features of BPD from the DSM-5 using a 5-point Likert scale. This study investigated its validity and reliability in screening for BPD in university students.

Methods: A cross-sectional study was conducted on a sample of university students in Thailand between November and December 2019. An online assessment gathered demographic data and results from the SI-Bord, the Perceived Stress Scale-10 (PSS-10) and the Patient Health Questionnaire-9 (PHQ-9). Participants whose SI-Bord scores were ≥ 1 were randomly selected to be interviewed and assessed for a BPD diagnosis by four psychiatrists using the Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II) as a reference point. An intraclass correlation coefficient (ICC) of 0.925 (95% CI, 0.805-0.979) ensured inter-rater reliability between the four psychiatrists. The diagnostic sensitivity and specificity of the SI-Bord, as compared to that of the SCID-II, were determined to indicate the cut-off score. The Receiver Operating Characteristics (ROC) was analyzed to evaluate its diagnostic accuracy.

Results: The study included 342 students aged 18-25 years (the mean age was 20.25 ± 1.4 years), 80.4% of whom were female. Among the 68 participants selected for an online interview, 16 were diagnosed with BPD. The cut-off score of the SI-Bord was > 9, as suggested by the Youden index, yielding a sensitivity of 56.3% and a specificity of 92.3%. It had a positive predictive value of 69.2% and negative predictive value of 87.3%. The SI-Bord had adequate discriminative power between cases and non-cases of BPD, with the area under the ROC curve being 0.83. Cronbach's alpha for the SI-Bord was 0.76, indicating acceptable internal consistency. The SI-Bord score was positively correlated to PHQ-9 and PSS-10 scores (r = 0.67 and r = 0.69, p < 0.001, respectively) and negatively correlated to MSPSS (r = - 0.50, p < 0.001). The prevalence of BPD in the sample was 6.4%, according to the cut-off score > 9.

Conclusion: The SI-Bord demonstrated good reliability and validity for screening BPD in university students. However, a study in non-Thai and other population groups should be warranted.