Can we reliably assess spine movement quality in clinics? A comparison of systems to evaluate movement reliability in a healthy population

Abstract

Inertial measurement units (IMUs) have the potential to facilitate a large influx of spine movement and motor control data to help stratify low back pain (LBP) diagnosis and care; however, uncertainties related to validity and equipment/movement reliability are preventing widespread use and acceptance. This study evaluated the concurrent validity of Xsens DOT IMUs relative to gold-standard optical motion capture equipment, and compared within- and between-day reliability of both systems to track spine range of motion (ROM) and movement quality (MQ) by evaluating intraclass correlation coefficients (ICC), standard error of measurement (SEM), coefficient of variation (CV), and minimum detectable difference (MDD). ROM was evaluated during planar ROM movements, and local dynamic stability (LDS; λ max), mean absolute relative phase (MARP) and deviation phase (DP) were estimated from repetitive trunk flexion at 3 speeds, in 15 healthy controls to assess MQ. Results showed no statistically significant differences between systems for all metrics, and ICCs ≥ 0.86; therefore, validity was confirmed for tracking primary axis ROM and MQ. IMU data revealed that absolute (C7, T12, and S1) and relative (thoracic, lumbar, and total) ROM was the most reliable metric, followed by λ max, DP, and MARP. Reliability was similar between systems, suggesting that the poorer between-day reliability (higher SEM and CV, lower ICC) observed is attributable to movement variability and sensor placement rather than equipment error. The MDDs can provide thresholds to researchers and clinicians for identifying changes in MQ. Further standardization of evaluated movements/metrics, and patient subgrouping are suggested to improve reliability assessments and refine MDDs in future work.