Development and Psychometric Validation of Capacity Assessment of Prosthetic Performance for the Upper Limb (CAPPFUL).

ABSTRACT

OBJECTIVES: (1) To develop a performance-based measure for adult upper limb (UL) prosthetic functioning through broad (ie, overall performance) and functional domain-specific (eg, control skills) assessment of commonplace activities; (2) to conduct initial psychometric evaluation of the Capacity Assessment of Prosthetic Performance for the Upper Limb (CAPPFUL). DESIGN: Internal consistency of CAPPFUL and interrater reliability for task, functional domain, and full-scale (sub)scores among 3 independent raters were estimated. Known-group validity was examined comparing scores by amputation level. Convergent validity was assessed between CAPPFUL and 2 hand dexterity or function tests; discriminant validity was assessed against self-reported disability. SETTING: Six prosthetic rehabilitation centers across the United States. PARTICIPANTS: Subjects (N=60) with UL amputation using a prosthesis. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Not applicable. RESULTS: Interrater reliability was excellent for scoring on the task, domain, and full-scale scores (intraclass correlation coefficients=.88-.99). Internal consistency was good (α=.79-.82). Generally, subjects with higher UL amputation levels scored lower (worse) than subjects with lower UL amputation levels. CAPPFUL demonstrated strong correlations with measures of hand dexterity or functioning (rs=-.58 to .72) and moderate correlation with self-reported disability (r=-.35). CONCLUSIONS: CAPPFUL was designed as a versatile, low-burden measure of prosthesis performance for any UL functional prosthetic device type and any UL amputation level. CAPPFUL assesses overall performance and 5 functional performance domains during completion of 11 tasks that require movement in all planes while manipulating everyday objects requiring multiple grasp patterns. Psychometric evaluation indicates good interrater reliability, internal consistency, known-group validity, and convergent and discriminant validity.