Category: Measurement; Clinical Practice (assessment, diagnosis, treatment, knowledge translation/EBP, implementation science, program development)
To examine the psychometric properties of 18 Upper Extremity Functional Index (UEFI) items using the one-parameter partial credit model and two-parameter graded response model.
Design : Cross-sectional survey study.
Setting : Outpatient rehabilitation clinics.
Participants (or Animals, Specimens, Cadavers) :
Clinical data of 982 patients (female = 52.5%, mean [SD] age = 50.1 [14.3]), requested from the Focus on Therapeutic Outcomes, Inc. with orthopedic shoulder impairments seeking outpatient rehabilitation therapy were analyzed.
Interventions : Not applicable.
Main Outcome Measure(s) : UEFI
Results : Internal consistency of items was 0.94. With a separation index of 4.0, the 18 items can separate patient sample into 5.7 statistically distinct strata. Comparing to the mean (SD) of patient ability level of 0.34 (SD = 0.39) logits, the item difficulty estimates [mean (SD) = 0.0 (0.04) logits] matched well with the functional levels of the sample. With only 12 patients had extreme scores, there were no ceiling or floor effects. Item 'lifting a bag of groceries above your head' appeared to be the most difficult item, followed by 'throwing a ball', while tem 'doing up buttons' was the easiest. Items 'preparing food', 'cleaning' and 'laundering clothes' were able to discriminate patients' upper extremity function better than other items, while 'sleeping' and 'throwing a ball' had the lowest discrimination abilities. Meanwhile, both 'sleeping' and 'throwing a ball' showed high infit and outfit statistics, suggesting those items deviate from unidimensionality in the data. We presented the keyform with a clinical example for the application of clinical interpretation of the patient's improvement and goal setting.
Results may improve clinical interpretation of the UEFI measure and assist clinicians using patient-reported outcomes during clinical practice.
Inga Wang– Associate Professor, University of Wisconsin Milwaukee, Milwaukee, Wisconsin
Jay Kapellusch– Associate Professor, Chair, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin
Bhagwant Sindhu– Associate Professor, OT Program Director, Graduate Program Coordinator, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin
Xiaoyan Li– Assistant Professor, University of Texas Health Science Center at Houston, Houston, Texas
Sheng-Che Yen– Assistant Professor, Northeastern University, Boston, Massachusetts
Leigh Lehman– Associate Professor, Brenau University, Norcross, Georgia