Role of Self-efficacy in the Predictive Relationship of Motor Ability to Functional Performance After Task-Related Training in Stroke: A Secondary Analysis of Longitudinal Data

Objective: To determine whether self-efficacy (SE) mediates or moderates the relationship between motor ability at pretest and functional use of the affected arm at posttest in task-related training for stroke. Design: Retrospective, observational cohort study. Setting: Outpatient rehabilitation settings. Participants: Eighty patients with chronic stroke (N=80). Interventions: The training was delivered to the participants for 60-90 min/session, 3-5 sessions/wk for 4-6 weeks. The training involved specific robot-assisted, mirror, or combined therapy, followed by functional task practice for approximately 30 minutes in each session. Main Outcome Measures: The outcome measure was the perceived amount of functional arm use and quality of movement evaluated by the Motor Activity Log (MAL) at posttest. The predictor was scores on the Fugl-Meyer Assessment (FMA)–Upper Extremity subscale at pretest. The tested mediator and moderator were scores on the Stroke Self-Efficacy Questionnaire (SSEQ) at pretest and posttest. Results: The SSEQ scores at pretest and posttest moderated the predictive relationship of pretest FMA to posttest MAL. The interaction between pretest FMA and SSEQ accounted for an additional 3.14%-5.37% of the variance in the posttest MAL. The predictive relationship between FMA and MAL was its greatest when the SSEQ was high, with a less amplified positive relationship at low levels of SSEQ. Conclusions: The results suggest the evaluation of SE at pretest for a better prediction of an individual patient's functional arm use after an intervention and recommend aiming at SE during training to make the most of motor ability transferred to functional use. Future research may compare the effectiveness of task-related training with and without SE building to verify the findings of this study.