Charlotte Lanhers, Bruno Pereira, Eliot Picard, Emmanuel Coudeyre, Christian Dualé
Of 107 eligible studies, 23 (22%) applied a Rasch analysis. The ODI (k = 11) and the RMDQ (k = 10) were more frequently analyzed than the QBPDS (k = 2). Most studies (87%) had Rasch analysis as a primary objective. Reporting rates varied widely. The software used (91%), unidimensionality testing (83%), and fit statistics (78%) were well documented, whereas response dependency (30%) and person fit tests (17%) were rarely reported.
BACKGROUND: The questionnaires currently used to assess physical functioning in low back pain (LBP) are the Oswestry Disability Index (ODI), the Roland-Morris Disability Questionnaire (RMDQ), and the Quebec Back Pain Disability Scale (QBPDS). Whereas those tools had traditionally been validated according to the classical test theory, Rasch analysis - based on the item response theory - offers a more refined item-level assessment. However, despite recommendations, reporting standards for Rasch analyses remain inconsistent. This review aimed to describe the use of Rasch analysis to validate these tools and assess their reporting quality.
METHODS: A systematic search of PubMed, Embase, Google Scholar, and Cochrane was conducted without time restrictions, including studies in English or French analyzing Rasch models for the ODI, RMDQ, or QBPDS. Keywords included terms related to "low back pain," "validation," and "Rasch analysis."
RESULTS: Of 107 eligible studies, 23 (22%) applied a Rasch analysis. The ODI (k = 11) and the RMDQ (k = 10) were more frequently analyzed than the QBPDS (k = 2). Most studies (87%) had Rasch analysis as a primary objective. Reporting rates varied widely. The software used (91%), unidimensionality testing (83%), and fit statistics (78%) were well documented, whereas response dependency (30%) and person fit tests (17%) were rarely reported.
DISCUSSION: Although recommended to validate questionnaires of physical functioning in LBP, Rasch analysis is underused and inconsistently reported. Variable terminology and software output complicate comparisons. Insufficient critical appraisal tools hinder the adoption of methods. Standardized reporting guidelines and enhanced expertise are needed to improve the integration of Rasch into psychometric assessments.
PROSPERO REGISTRATION: This systematic review and meta-analysis was registered in PROSPERO (registration number: 1157921, date of registration: 29 September 2025).