Date Presented 4/8/2016
This systematic review elucidates trends in research using item response theory (IRT) models to develop and validate patient-reported outcome measures. Implications of findings regarding use of IRT models, method(s) of item fit evaluation, and statistical software are discussed.
Primary Author and Speaker: Brittany N. Hand
Contributing Author: Paul De Boeck
PURPOSE: To quantify the real-world activity and affected upper-limb (UL) use of children with cerebral palsy (CP) who participated in constraint-induced movement therapy (CIMT), we (1) determine the feasibility of using accelerometers to quantify UL movements during CIMT, (2) compare activity level of children with hemiplegic CP with that of typical peers, and (3) determine the relationship between change in activity/duration of affected UL movements (as measured by accelerometer) and change in performance on standardized assessments.
BACKGROUND: CIMT is a high-dosage rehabilitation approach used for children with hemiplegic CP. Questions remain regarding how much movement is required during CIMT programs to change the real-world affected UL use of children with CP. Understanding optimal therapy dosage is important for maximizing outcomes and offering family-friendly, achievable interventions. Accelerometer data could further elucidate the relationship between high-intensity CIMT and changes in patterns of affected arm use.
DESIGN: A pretest–posttest design was used with 12 children with hemiplegic CP (mean = 4.9 yr) who completed a 30-hr camp-based CIMT program.
ASSESSMENTS: Three developmental assessments, the Melbourne Unilateral Upper Limb Assessment–2, the Children’s Hand-Use Experience Questionnaire, and the Pediatric Evaluation of Disability Inventory, were administered pre–post CIMT program. Accelerometers using the ActiGraph GT9X Link were successfully worn before, during, and directly after the program to collect UL and activity data.
ANALYSIS: Repeated-measures analysis of variance was used to examine changes in accelerometer data of the affected UL’s active duration and mean activity counts before, during, and after the CIMT program. Tukey’s procedure was used for pairwise post hoc comparisons between days. In addition, paired t tests were used to assess differences between pre- and posttest assessment scores. A p < .05 was considered statistically significant.
RESULTS: All 12 children successfully wore the accelerometers 6 hr/day before, during, and after the CIMT program to collect the UL and activity data. Children demonstrated lower levels of moderate to vigorous activity compared with typical peers (p < .05). Significant improvements were seen on all three developmental assessments (p < .05), and children demonstrated a significant increase in the duration of affected UL use during each camp day in comparison with pretest data (p < .01 for the main effect of day and p < .05 for Tukey post hoc). Children also demonstrated a significant increase in the mean activity count of affected UL use during 3 of 5 days in comparison to pretest data (p < .01 for the main effect of day and p < .05 for Tukey post hoc). However, Tukey post hoc comparisons showed no significant changes in UL active duration/mean activity count pre- versus post-CIMT (p > .05).
CONCLUSION: Although improvements were seen in assessment scores measuring capacity of affected UL use, accelerometer data suggest that some children may not have incorporated new movements into daily habits after a short-dosage CIMT program. These preliminary findings on the use of accelerometers as an outcome measure for pediatric CIMT are promising but will need confirmation in a larger study with a program that offers a higher dosage of intensive task-specific practice (e.g., 60 hr).
IMPACT STATEMENT: Wearable accelerometers could increase our understanding of repetitions required to change motor movements and lead to new ways to provide the active ingredients of intensive therapies in different practice settings (i.e., school based, early intervention, outpatient).
