Validating a fear-of-falling–related activity avoidance scale in lower limb prosthesis users

Abstract

Background: Limited available data suggest that fear of falling (FoF) may be common among lower limb prosthesis users (LLPUs) and associated with negative rehabilitation outcomes. The impact of FoF on outcomes may be exacerbated when fear leads to self-imposed declines in activity. There is a need to identify the validity of fear-of-falling–related activity avoidance scales in LLPUs. Objective: To evaluate the convergent, discriminant, and known-group construct validity of the modified Survey of Activities and FoF in the Elderly (mSAFE) for LLPUs. Secondarily, we sought to determine if the mSAFE scale could be reduced without losing information regarding construct validity. Design: Online survey. Setting: Not applicable. Participants: Fifty-nine persons with unilateral or bilateral amputation at the transtibial or transfemoral level, with ≥6 months experience using a definitive prosthesis for activities other than transfers. Intervention: Participants completed an online survey that consisted of the mSAFE and questions to assess FoF (yes/no), fear-of-falling–related activity avoidance (yes/no), as well as previously validated scales capturing related and unrelated constructs. T-tests were used to compare mSAFE score between yes/no respondents for known-group construct validity. Correlations between mSAFE and previously validated surveys assessed convergent and discriminant validity. Analyses were repeated on a reduced number of mSAFE items following a redundancy analysis. Main Outcome Measures: Not applicable. Results: Significant correlation coefficients of |0.440–0.825| were found for convergent validity with evidence of known-group construct validity (p