Skip to main content
The MSI-BPD is a ten-question yes/no screen for probable borderline personality disorder. It takes about a minute, and it exists to answer one question: does this presentation warrant a proper structured interview?

​
What it measures

The nine DSM-IV borderline criteria, mapped onto ten items — the self-harm and suicidal-behaviour criterion gets its own question, which is why there are ten items for nine criteria. The items cover unstable relationships, deliberate self-harm and suicide attempts, impulsivity, mood instability, anger, distrust, dissociation and derealization, chronic emptiness, identity disturbance, and efforts to avoid abandonment. The frame is “generally … over the past 2 years (or as long as you can remember)” — this is a trait-level screen, not a state measure.

​
When to use it

  • When the presentation doesn’t fit the referral. A depression or anxiety referral that keeps producing relationship crises, self-harm, or abrupt treatment ruptures.
  • Before committing to a treatment model. The answer changes whether DBT, mentalization-based work, or schema-focused work belongs on the table.
  • Alongside the BSL-23. The MSI-BPD screens for whether the picture is present; the BSL-23 tracks severity over treatment. Screen once, track with the other.
  • In the assessment phase, not the crisis. A client in acute distress will over-endorse trait items; the screen reads more cleanly when things are stable.

​
How clients fill it

About a minute. The stem reads: “Please answer Yes or No to each of the following questions about how you have generally been over the past 2 years (or as long as you can remember).” All ten items carry the same two options:
  • No
  • Yes

​
How Rivet scores it

Yes scores 1, No scores 0. The total is the count of Yes answers across all ten items. Range: 0–10. No reverse-scored items. Item 2 is also tracked on its own — see below.

​
Score bands

Seven is the cutoff from Zanarini 2003, which reported sensitivity of 81% and specificity of 85% in a sample of psychiatric outpatients. A positive screen means the structured interview is indicated — not that the diagnosis is made. Ten yes/no questions cannot carry a personality disorder diagnosis, and the consequences of an incorrectly applied borderline label are large enough that this distinction matters more here than on most measures.

​
Item 2 — self-harm and suicide attempts

Item 2 asks directly about deliberate self-harm and suicide attempts. A Yes on item 2 warrants a risk conversation regardless of the total. An MSI-BPD of 4 with item 2 endorsed matters more than an MSI-BPD of 9 with item 2 at No.Rivet flags item 2 separately in the scoring pill when it’s endorsed. What happens next — history-taking, a C-SSRS, a safety plan, a consult — is your clinical call.
Item 2 reads: “Have you deliberately hurt yourself physically (e.g., punched yourself, cut yourself, burned yourself)? Or made a suicide attempt?” It’s a compound question covering two different behaviours with different risk implications, and the client answers it with one word. The item tells you to have the conversation; it doesn’t tell you which of the two they meant, how recent it was, or what the intent was. That’s the follow-up.

​
Clinical change

The MSI-BPD isn’t a change measure. The frame is the past two years, so re-administering it during treatment measures very little. Screen once; track severity with the BSL-23.

​
When NOT to use it

  • As a diagnostic instrument. It’s a screen. The diagnosis comes from a structured clinical interview.
  • During acute crisis. Trait items get answered as state items when the client is in the middle of something.
  • With adolescents, as if it were a diagnosis. The instrument was validated in adult psychiatric outpatients, and applying a personality disorder label to a developing adolescent on the strength of a screen is not defensible.
  • As a way to explain a difficult therapeutic relationship. If the reason you’re reaching for it is that the work is hard, that’s a supervision conversation, not a screening one.

​
MSI-BPD vs BSL-23

The MSI-BPD asks whether the criteria are met; the BSL-23 asks how bad this week was. Ten yes/no items mapped to the nine DSM criteria, scored against a cutoff of 7 — that’s a screen, and a positive one points to a structured clinical interview, not to a diagnosis. The BSL-23 has no diagnostic cutoff at all and isn’t trying to have one. It tracks borderline symptom severity over the past week, which is what you want once the formulation is settled and you’re doing DBT or similar work. The order most practices use: MSI-BPD once, BSL-23 repeatedly.

​
Is it free to use?

Yes. Free clinical use, and widely cited.

​
Citation

Zanarini, Vujanovic, Parachini, Boulanger, Frankenburg & Hennen (2003). Journal of Personality Disorders, 17: 568–573. Free for clinical use. Rivet preserves the canonical item wording verbatim.

BSL-23

Borderline Symptom List — the severity tracker to use once the screen is positive.

C-SSRS

The suicide-risk assessment to run when item 2 is endorsed.

Safety plan

The six-step collaborative plan that follows an actionable risk finding.

Screening overview

When to screen and the risk-flag rule across measures.