Derivation and testing of an interval-level score for measuring locomotor disability in epidemiological studies of middle and old age.

PURPOSE

Locomotor disability is prevalent in middle and old age and although interval-level measurement instruments have been developed that relate to locomotor disability, there is currently no brief, interval-level score developed specifically to measure the severity of locomotor disability in the general population. We aimed to create such a score from five items of the SF-36 Physical Functioning subscale (PF-10): three walking- and two stair-climbing-related, and assess the scoring mechanism's measurement properties.

METHODS

Data were collected from postal surveys of adults aged 50 years and over. The walking and stair-climbing items were formed into two super-items to remove response dependency. The Rasch measurement model was used to form an interval-level score and fit to this model was assessed. The scoring mechanism was applied to three external data sets and fit to the Rasch model was assessed. Repeatability, construct validity, and responsiveness of the new interval scale were examined.

RESULTS

Overall, the fit of the super-items to the Rasch model was good, with little evidence of misfit. Our analyses broadly support the generalisability, repeatability, construct validity, and responsiveness of this new interval-level score.

CONCLUSIONS

An interval-level score for locomotor disability was created from five items of the PF-10 completed in a general population of adults aged 50 years and over. Further work is needed to assess the generalisability of this Rasch score to other populations and to determine a clinically meaningful change score.