Psychometric properties of the Flemish version of the Life Balance Inventory, a patient-reported outcome measure in chronic stage after stroke

Purpose: To investigate test-retest reliability, measurement error, and convergent validity of the Flemish version of the Life Balance Inventory (Fl-LBI) in chronic community-dwelling stroke survivors.Methods: Stroke survivors living at home (>6 months post-stroke) were recruited. The Fl-LBI was administered twice (1st home-visit, 2nd postal), with a 1-week time interval.Results: Thirty-two stroke-survivors participated: median age 61 years (IQR: 54-67), median time since stroke 571 days (IQR: 433-734). Median total Fl-LBI score was 2.47(IQR: 2.18-2.80) out of 3 on the first occasion (n = 32) and 2.55(IQR: 2.35-2.77) on the second occasion (n = 29), 91% retention rate. The intraclass correlation coefficient [95% confidence interval] for total Fl-LBI score was 0.88 [0.77-0.94] and for subscale scores >0.74 [0.48-0.94] indicating moderate to good test-retest reliability. Standard error of measurement was 0.14 out of 3(7%) and minimal detectable difference 0.39(20%). Agreement for activity category scores ranged from moderate to excellent. Construct validity was supported by moderate associations between Fl-LBI and participation (rs = 0.46); mobility (rs = 0.36) and emotion (rs = 0.37) subscales of the Stroke Impact Scale.Conclusion: In a pilot sample, the FL-LBI has satisfactory test-retest reliability, acceptable measurement error and minimal detectable difference and acceptable convergent validity. It can be used in clinical practice and research to assess life balance in chronic community-dwelling stroke survivors.Implications for rehabilitationLife balance can be a useful concept in stroke rehabilitation; it corresponds to individuals' perception of having a satisfying activity pattern tailored to their life circumstances.The Flemish version of the Life Balance Inventory can be used to evaluate life balance in community-dwelling chronic stroke.Satisfactory test-retest reliability and acceptable measurement error and minimal detectable difference were demonstrated in a pilot sample for the Flemish version of the Life Balance Inventory.status: Published onlin