Affiliations: Community Health Science, University of Northern
British Columbia, Prince George, BC, Canada
Note: [] Corresponding author: Lois E. Lochhead, University of Northern
British Columbia, Women North Network, Northern FIRE, 10-4040, 3333
University Way, Prince George, BC, V2N 4Z9, Canada. Tel.: +1 250 960 5789; Fax:
+1 250 960 5744; E-mail: [email protected]
Abstract: BACKGROUND: The Oswestry disability index (ODI) is the most widely
used measure of perceived disability for low back conditions. It has been
adopted without adaptation in functional capacity evaluation (FCE). Rigorous
testing of the ODI with modern psychometric methods, in this setting, is
warranted. OBJECTIVE: To determine the psychometric properties of the ODI in
FCE: unidimensionality; differential item functioning; item coverage and to
identify poorly functioning items, allowing for improvement of these items and
recalibration of the scale. METHODS: Rasch analysis, specifically Masters' partial credit model,
was conducted on data. PARTICIPANTS: 133 work-disabled individuals presenting for FCE in
northern British Columbia, Canada. RESULTS: All items had one poorly functioning option. Items were
rescaled from six categories to five, improving the psychometric properties of
the ODI as a unidimensional (disability due to back pain) scale. Item
difficulty range is sufficient for a population with mild to severe
disability. CONCLUSION: Although two of the ten ODI items functioned marginally
unsatisfactorily in the unrevised state, the 5-option revised ODI appears
superior. Use in clinical settings across a broad spectrum of disability levels
could help establish its psychometric properties. Health professionals should
be aware that the ODI may perform differently depending on client
population.
