Brief Symptom Inventory

The BSI (Brief Symptom Inventory) is a 53-item self-report measure of psychological symptom distress across 9 dimensions (Somatization, Obsessive-Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, Psychoticism) plus 3 global indices ^[1].

What does the BSI measure?

The Brief Symptom Inventory (BSI) is a 53-item self-report symptom inventory developed by Leonard Derogatis and Norman Melisaratos as a shortened version of the SCL-90-R ^[1]. It is used in clinical trials, outcome monitoring and psychological research, and it produces scores on 9 symptom dimensions and 3 global indices.

Where only depression or anxiety is in question, a shorter targeted measure such as the PHQ-9, GAD-7 or DASS-21 may be enough; the K10 is a brief single-score distress screen.

BSI nine symptom dimensions

Each dimension produces a raw score (sum of item ratings ÷ number of items in that dimension) and a T-score relative to the normative sample. A dimension T-score ≥63 on two or more subscales, or a GSI T-score ≥63, meets the clinical caseness criterion ^[3].

The nine dimensions and their clinical focus are:

Note: the 53 BSI items are distributed across these nine dimensions, and some items contribute to the global indices without belonging to a dimension subscale. Per-dimension item counts are specified in the publisher's scoring manual.

How do you give the BSI?

Respondents rate each symptom on a 5-point distress scale (0 = not at all to 4 = extremely) for the past seven days.

The BSI-18 is a brief 18-item version covering only Somatization, Depression, and Anxiety with a GSI ^[3], designed for medical settings and population screening. Both BSI and BSI-18 are published by Pearson Assessments and require purchase for clinical and research use.

How do you score and interpret the BSI?

The Global Severity Index (GSI), the mean of all 53 items, is the main summary of current distress level and psychiatric caseness. Scores are interpreted using normative T-scores (mean=50, SD=10); a T-score ≥63 on the GSI (or ≥2 dimensions at T≥63) indicates clinical caseness ^[3].

Interpreting BSI T-scores

BSI dimension and global index scores are expressed as T-scores normed against a published adult reference sample (mean = 50, SD = 10). Because the BSI is a multidimensional profile measure rather than a unidimensional screener, there are no literature-supported mild/moderate/severe total-score bands. The validated threshold is a single caseness criterion.

Clinical caseness is met when either of the following conditions is satisfied ^[3]:

• GSI T-score ≥ 63, or
• T-score ≥ 63 on two or more of the nine symptom dimensions.

Caseness signals that the level of psychological distress warrants further clinical evaluation. Scores below this threshold should still be interpreted within clinical context: the caseness rule identifies a clinically meaningful threshold, not a clean diagnostic boundary.

Choosing the appropriate norm group matters: the same raw score can yield different T-scores when referenced against outpatient versus community norms, and national norms differ ^[3].

Three global indices

The BSI produces three global indices that summarize different aspects of the overall symptom burden. All three are reported alongside the nine dimension T-scores ^[3].

Interpreting the three indices together gives a richer clinical picture than any one index alone:

• High GSI with high PST. Broad, diffuse distress across many symptom domains.
• High GSI with lower PST but high PSDI. Fewer symptoms, each experienced with considerable intensity: a pattern that may reflect a more focal but severe clinical picture.
• High PST with relatively low PSDI. Many symptoms acknowledged but none particularly distressing, which can occur with broad subclinical worry without frank psychological distress.

BSI vs BSI-18 vs SCL-90-R

Three scales from the same family: choose based on clinical setting, administration time, and the dimensional coverage needed.

• SCL-90-R. Preferred when comprehensive dimensional profiling is required and administration time is not a constraint.
• BSI. An acceptable short alternative to the SCL-90-R, with high correlations with the comparable SCL-90-R dimensions ^[1].
• BSI-18. Gives up dimensional breadth for brevity, so it suits rapid distress screening where the full nine-dimension profile is unnecessary ^[3].

Psychometric properties

• Introductory report. Test-retest and internal consistency reliabilities were very good for the primary symptom dimensions, and correlations with the comparable SCL-90-R dimensions were high. High convergence with like dimensions of the MMPI supported convergent validity, and factor analytic studies supported construct validity ^[1].
• Bifactor model. In 688 outpatients, a bifactor model fit the BSI's internal structure better than unidimensional or oblique models and improved diagnostic accuracy (73% versus 61% and 65%) ^[2].
• Swiss norms. In a representative Swiss general-population sample of 1,238 adults, the BSI had good psychometric properties; 18.1% of the general-population participants and 75.2% of psychotherapy outpatients met the caseness criterion ^[3].

Limitations

• Not diagnostic. Caseness signals distress that warrants further evaluation; it is not a diagnosis, and a high dimension score (for example Psychoticism) does not confirm a disorder.
• Profile, not severity bands. There are no literature-supported mild/moderate/severe total-score bands; only the T ≥ 63 caseness rule is validated.
• Norm-group dependence. The same raw score yields different T-scores against outpatient, community or adolescent norms, and national norms differ.
• Self-report. Ratings reflect the past seven days and depend on the respondent's willingness and ability to report symptoms.
• Licensed. The BSI is proprietary (Pearson); per-dimension item counts and normative tables are in the scoring manual.