The Borderline Symptom List-Interview: development and psychometric evaluation of an observer-based instrument for assessing symptom severity in borderline personality disorder

Abstract

Background: Borderline Personality Disorder (BPD) is a complex mental health condition characterized by pervasive instability in mood, interpersonal relationships, self-concepts, and behavior. A reliable assessment of BPD symptom severity is essential for effective treatment planning and evaluation. This study introduces and evaluates the Borderline Symptom List Interview (BSL-I), a semi-structured interview designed to assess the severity of BPD symptoms comprehensively.

Method: The BSL-I is a freely accessible 31-item interview designed to assess BPD symptom severity. It evaluates (a) the frequency and subjective distress associated with BPD-specific and typical psychopathological symptoms, (b) the behavioral consequences of these symptoms, (c) functional impairment, and (d) facets of positive mental health. The items were developed through an iterative process, incorporating feedback from international experts and individuals with lived experience of BPD. Psychometric properties of the BSL-I were examined cross-sectionally in different samples of clients meeting DSM-5 criteria for BPD (n = 171), clinical controls (n = 89), and healthy controls (n = 43).

Results: The BSL-I demonstrates good internal consistency within the BPD sample (Cronbach's α = 0.82) and good interrater reliability (ICC = 0.768). It significantly discriminates between BPD clients and clinical controls (Cohen's d = 2.02) and healthy controls (Cohen's d = 3.88). High correlations were observed with other established BPD symptom measures, including the number of IPDE criteria (r = 0.70, p < 0.001) and the BSL-23 (r = 0.83, p < 0.001).

Discussion: Our findings indicate that the BSL-I is a reliable and valid multidimensional instrument for assessing the severity of BPD. Both clinical experts and clients found the application of the BSL-I acceptable and feasible. Future research might explore its sensitivity to change resulting from psychosocial treatments and assess its utility for treatment planning and outcome measurement.

Conclusion: The BSL-I is a practical and psychometrically sound instrument for assessing the severity of BPD symptoms in clinical and research contexts.

Keywords: Borderline personality disorder (BPD); Psychometric evaluation; Semi-Structured interview; Severity assessment.

Fig. 1

Boxplots illustrating the distribution of each diagnostic group across the four dimensions of the Borderline Symptom List – Interview (BSL-I). Dimension labels: D1 – Symptom Frequency, D2 – Subjective Distress, D3 – Behavioral Consequences, D4 – Impairment in Daily Life. Diagnostic groups include BPD (Borderline Personality Disorder), CC (Clinical Controls), and HC (Healthy Controls). The dotted line represents the scale score for each subgroup

Fig. 2

ROC curves illustrating the accuracy of the BSL-I in classifying diagnostic status (BPD vs. non-BPD) in two subsamples. Left: Borderline Personality Disorder (BPD) vs. Healthy Controls (HC); Right: Borderline Personality Disorder (BPD) vs. Clinical Controls (CC)

Fig. 3

Bland-Altman plot illustrating the agreement between two raters across clients (N = 80) as part of the inter-rater reliability (IRR) assessment. The red dashed line represents the mean difference between the raters. The blue dotted lines indicate the 95% limits of agreement (mean ± 1.96 × SD)

Fig. 4

Scatterplot illustrating the correlation between IPDE-BPD criteria and BSL-I scores, assessing construct validity in a single-method comparison. BPD: Borderline Personality Disorder, CC: Clinical Controls

Fig. 5

Scatterplot illustrating the correlation between BSL-23 scores (model concept) and BSL-I scores, assessing convergent validity in a multi-method comparison. BPD: Borderline Personality Disorder, CC: Clinical Controls, and HC: Healthy Controls

Fig. 6

Pearson correlations between the BSL-I score and its dimensions with established measures for convergent and discriminant validity in the BPD sample. Abbreviations: BSL-23 = Borderline Symptom List-23; SCL-27 Physical = Physical Symptoms Subscale; SCL-27 Emotional (2 weeks) = Emotional Symptoms Subscale (past 2 weeks); SCL-27 Emotional (lifetime) = Emotional Symptoms Subscale (lifetime); Q-LES-Q-SF = Quality of Life; MWT = premorbid IQ; BSL-I = Borderline Symptom List – Interview (Total Score); BSL-I Dimensions = D1) Frequency, D2) Distress, D3) Behavioral Consequences, D4) Impairment in Daily Life

Fig. 7

Pearson correlations between the BSL-I score and its dimensions with established convergent and discriminant validity measures in the total sample. Abbreviations: BSL-23 = Borderline Symptom List-23; SCL-27 Physical = Physical Symptoms Subscale; SCL-27 Emotional (2 weeks) = Emotional Symptoms Subscale (past 2 weeks); SCL-27 Emotional (lifetime) = Emotional Symptoms Subscale (lifetime); Q-LES-Q-SF = Quality of Life; MWT = premorbid IQ; BSL-I = Borderline Symptom List – Interview (Total Score); BSL-I Dimensions = D1) Frequency, D2) Distress, D3) Behavioral Consequences, D4) Impairment in Daily Life

Fig. 8

Density plot of BSL-I scores across the groups and the six severity degrees. Groups include Borderline Personality Disorder (BPD), Clinical Controls (CC), and Healthy Controls (HC)