Participant Perception of Recovery as Criterion to Establish Importance of Improvement for Constraint-Induced Movement Therapy Outcome Measures: A Preliminary Study

• Published in: Physical therapy 2007-02, Vol.87 (2), p.170-178
• Main Authors: Fritz, Stacy L, George, Steven Z, Wolf, Steven L, Light, Kathye E
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Care and treatment; Changes; Data analysis; Data collection; Disability; Disability Evaluation; Exercise Therapy; Female; Follow-Up Studies; Health aspects; Humans; Male; Middle Aged; Movement; Movement therapy; Multivariate Analysis; Neuropsychological Tests; Physical therapists; Prognosis; Recovery of Function; Regression Analysis; ROC Curve; Stroke; Stroke (Disease); Stroke Rehabilitation; Studies; Therapy; Training; Treatment Outcome
• ISSN: 0031-9023; 1538-6724
• DOI: 10.2522/ptj.20060101

Changes in function following constraint-induced movement therapy (CIMT) are characterized primarily by improvements in performance; however, the importance of these outcome measures to the participant may be unclear. The primary purpose of this study was to determine whether either change scores or raw follow-up scores for the Motor Activity Log amount scale (MALa) and the Wolf Motor Function Test (WMFT) predicted participants' self-reports of recovery of upper-extremity function at 4 to 6 months after starting CIMT. This study was a secondary analysis of a cohort of subjects (N=46) who participated in CIMT trials. completed measures at baseline and 4 to 6 months later. Hierarchical regression models determined whether change scores or raw follow-up scores of CIMT outcome measures were predictive of perceived recovery. Receiver operating characteristic (ROC) curves determined cutoff scores for measures that significantly contributed to participants' reports of perceived recovery. The regression models indicated that raw follow-up MALa scores (beta=0.80, P=.024) and WMFT scores (beta=-0.37, P=.03) contributed to perceived recovery. Proposed cutoff scores for the MALa scores were less than 1.15 (negative likelihood ratio [LR]=0.17) for predicting less than 50% recovery and greater than 2.50 (positive LR=2.75) for predicting 50% or greater recovery. Proposed cutoff scores for follow-up WMFT scores were greater than 34.0 seconds (negative LR=0.24) for predicting less than 50% recovery and less than 11.0 seconds (positive LR=5.96) for predicting 50% or greater recovery. Raw follow-up scores for the MALa and WMFT were better predictors of self-report of recovery in comparison with change scores. These data also serve as a starting point for developing cutoff scores that accurately predict self-report of recovery.