Determining meaningful changes in gait speed after hip fracture.

BACKGROUND AND PURPOSE Older subjects after hip fracture walk more slowly than age-matched peers. The extent to which they walk more slowly is difficult to define because the standard error of the measure (SEM), sensitivity to change, and clinically important change have not been reported for gait speed. The purposes of this study were to quantify the SEM for habitual and fast gait speeds among older subjects after hip fracture, to define the minimal detectable change (MDC), and to estimate the minimal clinically important difference (MCID) for habitual gait speed. SUBJECTS A sample of 92 subjects after hip fracture was drawn from 3 studies that collected gait speed data. METHODS An estimate of the MDC was determined by use of the SEM. The MCID was determined from expert opinion and from a receiver operating characteristic (ROC) curve. RESULTS The SEM and the MDC were 0.08 m/s and 0.10 m/s for habitual speed and fast speed, respectively. Both methods of MCID estimation identified 0.10 m/s as a meaningful change in habitual gait speed. DISCUSSION AND CONCLUSION The estimated MCID for gait speed of 0.10 m/s was supported by clinical expert opinion and the cutoff point of the ROC curve.