Validity and reliability of a sensor-based electronic spinal mobility index for axial spondyloarthritis

Abstract Objective To evaluate the validity and reliability of inertial measurement unit (IMU) sensors in the assessment of spinal mobility in axial spondyloarthritis (axSpA). Methods A repeated measures study design involving 40 participants with axSpA was used. Pairs of IMU sensors were used to measure the maximum range of movement at the cervical (Cx) and lumbar (Lu) spine. A composite IMU score was defined by combining the IMU measures. Conventional metrology and physical function assessment were performed. Validation was assessed considering the agreement of IMU measures with conventional metrology and correlation with physical function. Reliability was assessed using intra-class correlation coefficients (ICCs). Results The composite IMU score correlated closely (r = 0.88) with the BASMI. Conventional Cx rotation and lateral flexion tests correlated closely with IMU equivalents (r = 0.85, 0.84). All IMU movement tests correlated strongly with BASFI, while this was true for only some of the BASMI tests. The reliability of both conventional and IMU tests (except for chest expansion) ranged from good to excellent. Test–retest ICCs for individual conventional tests varied between 0.57 and 0.91, in comparison to a range from 0.74 to 0.98 for each of the IMU tests. Each of the composite regional IMU scores had excellent test–retest reliability (ICCs=0.94–0.97), comparable to the reliability of the BASMI (ICC=0.96). Conclusion Cx and Lu spinal mobility measured using wearable IMU sensors is a valid and reliable assessment in multiple planes (including rotation), in patients with a wide range of axSpA severity.