# Source provenance and clinical audit **Last verified:** 2026-06-12 **Scope:** PCL-5 content, scoring, interpretation, translations and crisis link. ## 1. Primary instrument source National Center for PTSD. **PTSD Checklist for DSM-5 (PCL-5)**: - Measure page: - Standard form, version date 29 August 2023: - Clinical guide: The measure page carried `dcterms.dateAccepted=20260610` when checked on 2026-06-12. It states that PCL-5 was developed by National Center for PTSD staff, is public domain and is not copyrighted, and is intended for qualified health professionals and researchers. ### Mapping to repository | Repository field | Primary source | |---|---| | `pcl5-data.js items[*].en` | Official standard form, items 1–20 | | `anchors.en` | Official 0–4 anchors | | `instructions.month.en` | Official standard instructions | | `pcl5-engine.js CLUSTERS` | VA scoring guidance | | `DX_RULE`, `ENDORSE_MIN=2` | VA provisional DSM-5 symptom rule | | `CUTOFF_MIN=31`, `CUTOFF_MAX=33` | VA probable-PTSD range | | `delta().response10` | VA suggested response indicator | ## 2. Interpretation decisions ### Cut-point VA guidance says 31–33 is reasonable, but selection depends on population and assessment purpose. Lower values favor sensitivity; higher values favor specificity. The UI therefore requires an explicit choice from 31, 32 or 33 and defaults to 33. ### No severity categories The VA clinical guide explicitly says there are no empirically derived PCL-5 severity ranges. This repository intentionally does not label totals as mild, moderate or severe and does not create a “gray zone.” ### Past-week adaptation VA states that PCL-5 is intended for the past month and that other timeframes have not been validated. The optional week mode is visibly marked as unvalidated and suppresses cut-point interpretation. ### B–E rule An item is endorsed at `≥2`. The symptom pattern is met at: `≥1 B + ≥1 C + ≥2 D + ≥2 E`. The UI calls this a symptom-pattern/provisional evaluation, not a diagnosis. Criteria A, F, G and H still require clinical assessment. ## 3. Change scores Marx BP, Lee DJ, Norman SB, et al. (2022). *Reliable and clinically significant change in the Clinician-Administered PTSD Scale for DSM-5 and PTSD Checklist for DSM-5 among male veterans.* Psychological Assessment, 34(2), 197–203. - DOI: - PubMed: - PMC: Findings used: - PCL-5 reliable change: 15 points in Sample 1 and 18 in Sample 2. - Final PCL-5 `≤28`: clinically significant change margin in both samples. Limits preserved in the UI and documentation: - treatment-seeking US military veterans; - male-only analytic samples; - small test–retest samples; - point estimates that may differ in other populations; - authors caution against equating these values with recovery. VA separately suggests about 10 points as an indicator of response while stating that evidence is limited. The previous draft's `≥5` rule was removed. ## 4. Ukrainian text Reference adaptation: Bezsheiko V. (2016). **Адаптація Шкали для клінічної діагностики ПТСР та опитувальника "Перелік симптомів ПТСР" для української популяції.** Психосоматична медицина та загальна практика, 1(1), e010108. - Article: - Published PCL-5 PDF: The 2016 article reported preliminary reliability/validity data from about 50 participants and planned a final sample of 100. It did not provide final psychometric estimates in the article. Current linguistic audit: Figueiredo S, Hefter D. (2025). **PTSD assessment in Ukrainian refugees using the PCL-5: linguistic and cultural challenges.** Frontiers in Psychology, 16:1659301. - DOI: - PMC: The authors found no fully validated Ukrainian PCL-5, reported mediocre semantic/structural overlap in available translations and highlighted major issues in items 3, 7, 16–18 and domain E. `items[*].uk` is a linguistically revised clinical adaptation aligned to the official English meaning. Notable corrections include: - item 3: restores the “back there reliving it” explanation; - item 7: restores the full external-reminder examples; - item 9: preserves “completely dangerous”; - item 12: keeps all enjoyable activities, not only work; - items 16–18: distinguishes risky behavior, hypervigilance and exaggerated startle; - item 20: includes both falling and staying asleep. This revision has not undergone forward/back translation, cognitive interviewing, measurement invariance testing, ROC validation or test–retest validation. It must not be represented as psychometrically validated. ## 5. Russian text Figueiredo & Hefter evaluated a Russian PCL-5 published by Los Angeles County Department of Mental Health in 2019 and likewise reported no psychometric validation plus item-level semantic discrepancies. `items[*].ru` is a close clinical translation from the official English text, written to preserve the same distinctions listed above. It has not undergone formal psychometric validation and must not be represented as validated. ## 6. Criterion A and related tools The optional event block in this repository provides context only. It is not a validated substitute for the official extended Criterion A assessment. Official LEC-5: CAPS-5 remains the recommended structured diagnostic interview: ## 7. ICD note The report reminds clinicians that PCL-5 follows DSM-5 symptom criteria: - ICD-10: F43.1, post-traumatic stress disorder. - ICD-11: 6B40, post traumatic stress disorder. - ICD-11: 6B41, complex post traumatic stress disorder. PCL-5 does not measure the ICD-11 disturbances in self-organization required for complex PTSD. Consider the International Trauma Questionnaire when that construct is relevant. Official ICD browsers: - ICD-11 MMS: - ICD-10 2019: ## 8. Crisis information LifeLine Ukraine's official site was checked on 2026-06-12 and states that 7333 provides suicide-prevention and mental-health support 24/7: The UI also states that PCL-5 does not assess suicide risk and directs people in immediate danger to clinicians or emergency services. ## 9. Software behavior - All scoring functions are pure and covered by `docs/test-engine.js`. - No network request is made by the application. - Draft PHI is stored only in `sessionStorage` and expires after one hour. - Language preference alone is stored in `localStorage`. - Generated reports escape user-entered text before producing HTML.