Reliability and minimum detectable change of the gait profile score for post-stroke patients

The objectives of this work were (i) to determine Gait Profile Score (GPS) for hemiparetic stroke patients, (ii) to evaluate its reliability within and between sessions, and (iii) to establish its minimal detectable change (MDC). Seventeen hemiparetic patients (mean age 54.9±10.5years; 9 men and 8 women; 6 hemiparetic on the left side and 11 on the right side; mean time after stroke 6.1±3.5months) participated in 2 gait assessment sessions within an interval of 2-7 days. Intra-session reliability was obtained from the intraclass correlation coefficient (ICC) between the three strides of each session. Inter-session reliability was estimated by the ICC from the averages of that three strides. GPS value of non paretic lower limb (NPLL) (13.9±2.4°) was greater than that of paretic lower limb (PLL) (12.0±2.8°) and overall GPS (GPS_O) was 13.7±2.5°. The Gait Variable Scores (GVS), GPS and GPS_O exhibited intra-session ICC values between 0.70 and 0.99, suggesting high intra-day stability. Most of GVS exhibited excellent inter-session reliability (ICC between 0.81 and 0.93). Only hip rotation, hip abduction of PLL exhibited moderate reliability with ICC/MDC values of 0.57/10.0° and 0.71/3.1°, respectively. ICC/MDC values of GPS were 0.92/2.3° and 0.93/1.9° for PLL and NPLL, respectively. GPS_O exhibited excellent test-retest reliability (ICC=0.95) and MDC of 1.7°. Given its reliability, the GPS has proven to be a suitable tool for therapeutic assessment of hemiparetic patients after stroke.