LPFS-BF 2.0 in chronic pain: a 12-item screener you can run in five minutes before the first session
- **12 items, two domains (Self / Interpersonal), 4-point Likert.** Validated in n = 892 adults consecutively referred to the Oslo University Hospital outpatient pain clinic (mean age 49, 64% female). Administration time ~3-5 minutes; scoring is a simple sum.
- **Internal consistency:** general-factor ω = 0.89, α = 0.79. Bifactor model produced the best fit (CFI = 0.995, TLI = 0.993, RMSEA = 0.023, SRMR = 0.030). Explained common variance for the general factor was 68% — i.e., the LPFS-BF behaves as essentially **unidimensional** in chronic pain populations.
- **Concurrent validity:** moderate-to-strong associations with HADS anxiety, HADS depression, EQ-5D health-related quality of life and past-week pain intensity — magnitudes comparable to community and psychiatric samples reported by Hutsebaut and colleagues, supporting construct stability across diagnostic frames.
- **Domain caveat with clinical teeth:** the Self subscale alone showed poor reliability (ω = 0.25) and the Interpersonal subscale was questionable (ω = 0.68). In chronic pain, **interpret only the total score** — do not split into Self vs Interpersonal as some AMPD adopters do in psychiatric outpatient work.
If you work with chronic pain, ME/CFS, fibromyalgia, post-COVID, or any "medically unexplained" referral stream, you already know the bottleneck: somewhere between 25% and 50% of these patients carry significant personality pathology that sabotages pain-management protocols, and you have neither time nor permission to administer the SCID-5-AMPD. The LPFS-BF 2.0 is the pragmatic answer published in 2018 by Weekers, Hutsebaut and Kamphuis as a 12-item screener for the DSM-5 Alternative Model Criterion A (level of personality functioning). What this 2026 paper from Eikeseth and the Oslo Pain Registry adds is the missing piece: psychometric evidence in 892 chronic-pain patients, not in psychology students or community volunteers.
How it works
You give the patient 12 statements before the first session — items like "I often don't know who I really am" (Self) and "I find it difficult to feel close to others" (Interpersonal), each rated 0-3. Sum gives a total 0-36. The original validation studies in psychiatric samples suggested cut-offs around ≥17-19 for clinically significant impairment, with Hutsebaut's Dutch work proposing graded thresholds (mild / moderate / severe) at roughly 12 / 17 / 24. The Oslo paper does not publish a chronic-pain-specific cut-off — it instead establishes that the structure and reliability hold up well enough that you can keep using the existing thresholds with appropriate caution.
The bifactor structure deserves attention. The authors tested four competing models (one-factor, two-factor, second-order, bifactor) and bifactor won decisively. Translation: there is a strong general "personality dysfunction" factor running through all 12 items, and the Self/Interpersonal split that looks intuitive on paper does not carry independent variance worth interpreting in this population. The chronic-pain context likely conflates the two — pain reorganises identity and relationships simultaneously. So when you score the LPFS-BF 2.0 in your pain patient, trust the total, ignore the subscales.
In practice
Use case one: triage. A patient referred for CBT-CP or pain-rehabilitation programme scores 22 on the LPFS-BF total. That is not a contraindication — it is information. You know going in that adherence will be choppy, the therapeutic alliance will be tested, and standard pain-coping protocols may need pacing modifications. You would not have spotted that from the HADS or the Pain Catastrophizing Scale alone.
Use case two: outcome tracking. The LPFS-BF 2.0 is sensitive to change in personality-disorder treatment trials (mentalisation-based therapy, schema therapy, TFP). In chronic-pain settings where pain intensity may not move much but functional and interpersonal recovery does, an LPFS-BF delta of -4 to -6 over six months is meaningful evidence that the work is landing somewhere.
Use case three: differential. A patient scoring 29 on LPFS-BF + high pain catastrophising + 4+ ED visits in 12 months is probably not a candidate for short-term protocolised pain CBT. Refer for a longer formulation-based intervention or a DBT-informed pain group. The 12-item screener gave you that decision in five minutes.
In chronic pain, don't fight to interpret Self vs Interpersonal — the LPFS-BF 2.0 in this population is essentially one number, and that number is enough to change your treatment plan.
Cross-sectional, single-site (Oslo University Hospital), Norwegian-language version, predominantly white European sample. The poor Self subscale reliability is a real ceiling — readers wanting separate Criterion A domain scores need a longer instrument (LPFS-SR, SIPP-118). No chronic-pain-specific cut-off was derived; clinicians must lean on cross-sample thresholds.