Test-retest reliability and minimal detectable change of repeated sit-to-stand analysis using one body fixed sensor in geriatric patients

Abstract

A majority of geriatric patients experience difficulty in performing sit-to-stand (SiSt) transitions. A detailed assessment of SiSt ability is a prerequisite for successful rehabilitation. Body fixed sensors (BFSs) are increasingly used to assess functional performances. As to date there is no system which analyzes clinically relevant phases of SiSt, the aim of this study was to determine the reliability of an automated approach for quantifying durations and angular velocities of trunk flexion and extension during repeated SiSt transitions using one BFS (DynaPort® Hybrid). Forty multimorbid geriatric patients aged 84.1 ± 6.6 years were included. Each patient participated in two test sessions with a 5 min rest period in between. Intra- and interrater reliability was assessed. Intraclass correlation coefficients (ICCs), absolute and relative standard measurement errors (SEMs, SEMs%) and minimal detectable changes (MDCs95, MDCs95%) were calculated. ICCs were good to excellent for all variables in the total sample (0.80–0.94). The intraobserver group (50%) showed a higher number of excellent ICCs (≥.9) compared to the interobserver subgroup (10%). SEM% was low for all variables (6.9–12.7%). MDC95% ranged 19.2–34.4% and more variables ≤30% were found in the intra- (80%) compared to the inter-observer group (60%). Study results demonstrate that the BFS system provides a reliable analysis of SiSt phases in geriatric patients, and is a substantial improvement over the stopwatch approach used in clinical practice today.