Copenhagen Burnout Inventory (CBI)

Structural validity Moderatewell-establishedevidence form: mixed

direct for the UK optometrist Rasch study; otherwise indirect (US nurses, US academic staff, student adaptations)

The proposed three-factor structure (personal, work-related, client-related) has been supported by confirmatory factor analysis in several occupational samples: 928 US hospital nurses (Montgomery A. 2021) and 1,679 US academic-health-centre employees (Thrush C. 2020). A UK Rasch analysis of 1,303 patient-facing optometrists found the three subscales unidimensional with good discrimination, though the client/patient-related scale showed a 22% floor effect (Retallic N. 2026). Student adaptations restructure the client scale (for example a four-factor CBI-SS in 635 Nigerian students, with one item removed for low average variance extracted; Oluwadiya K. 2024). The core three-domain structure is comparatively stable across working-adult samples.

Convergent and discriminant validity Moderatewell-establishedevidence form: canonical

indirect principally US healthcare and Danish human-service samples

CBI scales correlate as expected with the work environment, job satisfaction and intention to leave (moderate-to-high, correct directions) in US nurses (Montgomery A. 2021), and discriminant validity against a measure of meaningful work was supported in US academic-health-centre employees (Thrush C. 2020). In the PUMA development study the scales differentiated well between human-service occupational groups and showed the expected correlations with fatigue and psychological well-being (Kristensen T. 2005).

Criterion validity: reference standard Absent (a finding about the literature)untestedevidence form: canonical

indirect no reference-standard study located, and no diagnostic standard exists to validate against

No diagnostic gold standard for burnout exists, so the CBI has not been validated against a clinical reference standard; the COSMIN-based systematic review of occupational burnout measures found only sparse criterion-validity evidence for the CBI, with the instrument far less studied than the MBI (Shoman Y. 2021). This property is essentially untested for the CBI.

Criterion validity: organisational Moderatewell-establishedevidence form: canonical

indirect prospective Danish human-service workers rather than a general UK working population, and the absence outcome is self-reported rather than register-based

The CBI carries an unusually strong prospective organisational link, though against self-reported rather than register-based absence. In a three-year follow-up of 824 human-service workers from the PUMA study, the work-related burnout scale predicted subsequent self-reported sickness absence: mean absence rose from 5.4 days per year in the lowest quartile to 13.6 days in the highest, and a one-SD increase in work-related burnout predicted a 21% increase in self-reported sickness-absence days after adjustment for gender, age, organisation, socioeconomic status, lifestyle, family status and disease (rate ratio 1.21, 95% CI 1.11 to 1.32) (Borritz 2006). The PUMA baseline paper documents the design and the associations of work burnout with the psychosocial work environment and absence (Borritz 2006). This prospective evidence is a relative strength of the CBI among burnout inventories, but the outcome was self-reported (survey-based, subject to recall and reporting bias) rather than drawn from administrative or registry absence records, so it is weaker than register-based evidence would be.

Internal consistency Highwell-establishedevidence form: mixed

indirect Danish, US and Nigerian samples, no UK-specific alpha in the retrieved primary set beyond the UK Rasch reliability evidence

Internal consistency is consistently high. US hospital nurses gave Cronbach's alpha of 0.91 (personal), 0.89 (work-related) and 0.92 (client-related) (Montgomery A. 2021); the PUMA development study reported very high reliability for all three scales (Kristensen T. 2005); and the Nigerian CBI-SS gave subscale alphas of 0.862 to 0.914 with a total of 0.957 (Oluwadiya K. 2024). Reliability is a strength of the CBI.

Test-retest reliability Absent (a finding about the literature)thin

indirectness not applicable; no coefficient located

No numeric test-retest (temporal stability) coefficient for the CBI was located in the latest review pass. The PUMA study was longitudinal and designed to track change rather than to establish short-interval retest stability, and the systematic review noted reliability evidence beyond internal consistency is sparse for the CBI. The absence of a reported retest coefficient is itself the finding: temporal stability of the CBI is effectively unevidenced in the retrieved literature.

Measurement invariance Lowcontestedevidence form: canonical

indirectness partly direct (UK optometrists) and partly indirect (US healthcare staff)

Evidence is partial. In 1,679 US academic-health-centre employees, configural and metric invariance held across professional role (physicians, nurses/PAs, other staff), gender and age, but scalar invariance was not established, cautioning against latent-mean comparisons across those groups (Thrush C. 2020). A UK Rasch analysis of 1,303 optometrists found no significant differential item functioning by gender or age, supporting cross-group validity at item level (Retallic N. 2026). Higher-order (scalar) invariance across occupations remains the weak point.

Sub-grades (evidence differs by subgroup):

• {"subgroup": "Configural / metric", "grade": "Moderate", "note": "supported across role, gender and age in a large US sample"}
• {"subgroup": "Scalar", "grade": "Very low", "note": "not established; group mean comparisons not yet justified"}

Responsiveness and MIC Very lowthinevidence form: canonical

indirect Danish human-service intervention context; no MIC anchor

The CBI was embedded in the five-year PUMA intervention study and used to track change over time (Borritz M. 2006), but no formal responsiveness statistic or minimal important change (MIC) value for the CBI was located, and the systematic review did not identify robust responsiveness evidence (Shoman Y. 2021).