Evaluation of the Patient-Reported Outcomes Measurement Information System Upper Extremity Computer Adaptive Test

Abstract

The Patient-Reported Outcomes Measurement Information System Upper Extremity Computer Adaptive Test (UE CAT) has recently been made available by the National Institutes of Health to measure physical function outcomes in the upper extremity. We hypothesized that the UE CAT would psychometrically outperform the Disabilities of the Arm, Shoulder, and Hand (DASH) and the Patient-Reported Outcomes Measurement Information System Physical Function Computer Adaptive Test (PF CAT) in a hand patient population. The UE CAT, PF CAT, and DASH were each electronically administered to all adult patients who presented to a tertiary hand and upper extremity (nonshoulder) orthopedic clinic. Patient responses were retrospectively studied to determine the validity, reliability, and floor/ceiling effects of all 3 instruments using the Rasch Partial Credit Model. Responder burden and Pearson correlations were calculated for each instrument. A total of 379 patients completed the UE CAT, PF CAT, and the DASH. On average, 6 UE CAT, 9 PF CAT, and 30 DASH questions were administered to each patient. All 3 instruments were each highly correlated with each other. Floor effects were low and similar between all instruments; however, ceiling effects were higher in the UE CAT (10.82%) than in the PF CAT (1.32%) or DASH (5.28%). High person reliability (PR) and item reliability (IR) were found for all 3 metrics: UE CAT (α= 0.99; PR= 0.91; IR= 0.94); PF CAT (α= 0.95; PR= 0.89; IR= 0.96); and DASH (α= 0.97; PR= 0.95; IR= 0.99). The UE CAT questions had the best item-fit: only 1 of 15 UE CAT items had poor fit in contrast to 4 of 30 DASH items and 7 of 33 PF CAT items. The psychometric properties of the UE CAT compare favorably with the PF CAT and the DASH in nonshoulder upper extremity patients. The relatively large ceiling effect found in the UE CAT could be improved with item bank expansion to include items at the upper end of function. The UE CAT is a useful patient-reported outcome measure that merits further investigation.