The reliability and minimal detectable change in gait outcomes among individuals with unilateral transfemoral amputation: A generalizability theory approach.

Abstract

Reliable gait assessments are critical for identifying gait deviations in individuals with unilateral transfemoral amputation (TFA), allowing targeted interventions to improve mobility and long-term outcomes. This study aimed to 1) evaluate the intra- and inter-visit reliability and minimum detectable change (MDC) of gait biomechanics outcomes in individuals with TFA and 2) determine the minimum number of steps per limb required for reliable gait assessments in this population. Ten participants (6 males, age = 36.9 ± 10.4 years, K level ≥ 3) with unilateral TFA completed a treadmill walking protocol twice, with a one-month interval between visits. Spatiotemporal, kinetic, and kinematic outcomes were analyzed using 14 steps per limb at a self-selected treadmill walking speed. Generalizability theory (GT) was applied to assess reliability and MDCs, while decision studies (D-studies) examined the effects of step counts (1 to 30) and visit numbers (2 to 5) on reliability and MDC outcomes. Reliability greater than 0.8 was considered desirable. Most gait outcomes demonstrated desirable intra- and inter-visit reliability. D-studies revealed that a minimum of 10 steps per limb is sufficient for reliable gait assessment, with additional steps or visits offering only marginal improvements in reliability and MDCs outcomes. In contrast to the minimum clinically important difference, which reflects patient-perceived meaningful change, the MDC values established in this study provide a benchmark for identifying gait alterations that exceed measurement error. These MDC values can assist researchers and clinicians with more precise longitudinal gait assessments for individuals with TFA. Collectively, our findings support reliable gait evaluation and effective long-term monitoring in this clinical population.