Objective: To develop and validate a self-reported questionnaire assessing the barriers to physical activity (PA) among stroke survivors. Design: Psychometric study. Setting: Ambulatory stroke care.
Participants: A total of one hundred and forty-six (N=146) individuals were included in this study. In stage 1, community-living stroke survivors (n=37; 13 women) with low-moderate disability (modified Rankin Score 0-3, stroke >3mo) were included. In stage 2, participants (n=109; 40 women) with same characteristics were included. Nine professionals experienced in PA for poststroke patients formed an expert panel.
Interventions: In stage 1, semistructured interviews identified perceived barriers to PA, which were then selected by the expert panel and grouped on a Barriers to Physical Activity After Stroke (BAPAS) scale. In stage 2, stroke participants completed a personal information questionnaire and the BAPAS scale.
Main Outcome Measures: An item selection process with factor analysis was carried out. The suitability of the data set was analyzed using the Kaiser-Meyer-Olkin coefficient, internal consistency was evaluated by Cronbach α, and concurrent validity was assessed with Spearman correlation coefficients between the BAPAS scale and the modified Rankin Scale. Test-retest repeatability was estimated using 2-way random effects intraclass correlation coefficient model 2,1 at 4-6 day follow-up (n=21).
Results: Factor analysis supported a 14-item BAPAS that explained 62% of total variance (Kaiser-Meyer-Olkin=0.82) and total score calculated higher than 70 (higher scores for higher barriers). Cronbach α was 0.86, Spearman correlation with the modified Rankin Scale was r=0.65 (P<.001), and test-retest intraclass correlation coefficient was 0.91 (95% CI, 0.79-0.97). The BAPAS scores were higher in patients with greater disabilities and in those with a longer time since the stroke event (P<.01).
Conclusion: We developed and validated the BAPAS scale to assess barriers to PA in stroke survivors with low-moderate disability with promising psychometric properties.