Validation and Clinical Applications of a Measurement Model for Upper-Extremity Function

SIGNIFICANCE: Weight of objects and distance from the floor are strong predictors for the challenging levels of gross upper-extremity (GUE) activities. This finding provides insight into an underlying measurement theory of UE function and may serve as an effective treatment intervention.

INNOVATION: In this study, we investigated the feasibility of model-building methods. The promising results indicate that the Rasch and multiple regression analyses are useful in developing a measurement model for GUE activities.

RESEARCH QUESTION: Is it feasible to build a measurement model accounting for a large variance of GUE function in individuals with musculoskeletal impairments?

BACKGROUND: Stenner and Stone (2010) demonstrated how the Lexile theory can objectively measure the complex construct of reading ability by using regression models. Presently, there have been few, if any, efforts to develop measurement models that underlie the challenge levels of upper-extremity functional activities. On the basis of Stenner and Stone’s methods, we demonstrate the development of a measurement theory underlying the GUE measure of the International Classification of Functioning, Disability and Health (ICF; World Health Organization, 2001). With a measurement theory, researchers may clarify critical components of UE function. Furthermore, measures derived from this theory may effectively guide clinical interventions designed to improve UE function in patients.

METHOD: A total of 203 outpatients with UE disability or low back pain participated in the study. The participants’ average age was 48.3 ± 17.9 yr (range = 18 to 89 yr). Of the participants, 64% were female, and 36% were male. The ICF–GUE consists of 27 items that reflect different levels of UE function. The dependent variable for the multiple regression analyses was the Rasch linear calibration of the 27 ICF–GUE test items. The independent variables were five task-analysis derived measures: object weight (lb), lifting distance from floor (in.), carrying (yes/no), lifting (yes/no), and pushing (yes/no). The multiple regression analysis was run using SAS 9.3, and the independent variable contribution to the model was determined using adjusted variance.

RESULTS: Object weight and lifting distance were the only significant independent variables in the regression model, accounting for 83% of the variance (p < .01). The regression model (Item Challenge = −2.52 + 0.16 × Object Weight + 0.02 × Distance From Floor) indicates that with a 1-lb increase in object weight, item challenge increases by 0.16 logits (p < .00), and with a 1-in. increase in distance lifted from floor, item challenge increases by 0.02 logits (p < .03).

CONCLUSION: The findings suggest that the measurement model for the ICF–GUE can be explained by object weight and distance lifted from the floor. This model may also underlie similar UE measures. Furthermore, the model suggests that treatments can be graded by choosing tasks based on the weight of the objects and the distance lifted from floor. Although the model was effective in explaining ICF–GUE, the findings need to be replicated for different target populations and with similar instruments.

References

Stenner, A. J., & Stone, M. (2010). Generally objective measurement of human temperature and reading ability: Some corollaries. Journal of Applied Measurement, 11, 244–252.

World Health Organization. (2001). International classification of functioning, disability and health. Geneva: Author.