Measuring activity limitation outcomes in youth with spinal cord injury.

STUDY DESIGN: A cross-sectional study. OBJECTIVES: The Pediatric Spinal Cord Injury Activity Measure (PEDI-SCI AM), which includes calibrated item banks (child and parent versions) for general mobility, daily routines, wheeled mobility and ambulation, can be administered using computerized adaptive tests (CATs) or short forms (SFs). The study objectives are as follows: (1) to examine the psychometric properties of the PEDI-SCI AM item banks and 10-item CATs; and (2) to develop and evaluate the psychometric properties of PEDI-SCI AM SFs. SETTING: US Shriners Hospitals for Children (California, Illinois and Pennsylvania). METHODS: Calibration data from a convenience sample of 381 children and adolescents with SCI and 322 parents or caregivers were used to examine PEDI-SCI AM item banks, 10-item CATs and SF scores. We calculated group reliability, internal consistency (Cronbach's alpha) and interclass coefficients (ICCs) to assess agreement between 10-item CATs, SFs and item banks. The percent of the sample with highest (ceiling) and lowest (floor) scores was also determined. An expert panel selected items for 14 SFs. RESULTS: PEDI-SCI item banks, 10-item CATs and SFs demonstrate acceptable group reliability (0.73-0.96) and internal consistency (0.77-0.98). ICC values show strong agreement with item banks for 10-item CATs (0.72-0.99) and SFs. Floor effects are minimal (<15%). Ceiling effects are minimal for children with tetraplegia but high in children with paraplegia for general mobility (13.41-26.05%) and daily activities (12.99-32.71%). CONCLUSIONS: The PEDI-SCI AM exhibited strong psychometric properties for children with tetraplegia. Replenishment of the general mobility and daily routine item banks is needed to reduce ceiling effects noted for youth with paraplegia.