Background For patients with axial spondyloarthritis (axSpA) it is recommended to monitor spinal mobility regularly. Cervical rotation (CR) is usually measured with goniometer or myrinometer, but might give imprecise results. Technological innovation provides opportunities to develop and improve measurement methods for use in clinical practice.
Objectives To evaluate validity and reliability of a sensor developed to measure rotation around the vertical axis in a laboratory setting.
Methods A sensor was developed by engineers in close collaboration with health professionals and patients with axSpA. It was tested in a laboratory setting and the data collected was compared with Cybex 6000 simulation data as gold standard. The sensor was mounted to the cybex with a bracket in six different positions, 1) straight up 2) straight down 3) up, tilted 30°around the frontal axis 4) down, tilted 30°around the frontal axis 5) up, tilted 30°around the frontal axis, 20° around the sagittal axis 6) down, tilted 30°around the frontal axis, 20° around the sagittal axis. Reference angles were set at approximately 10, 30, 60, 90 and 120 degrees. From a centre position the sensor was moved from left to right (one cycle) by manual force at the speed allowed (60°/second) until stop brackets were reached.
Criterion validity and reliability for single measures and for the mean of three trials (left, right or cycle) was evaluated with two ways mixed interclass correlation coefficientagreement (ICC). Limits of agreement (LoA) (Bland and Altman method), and smallest detectable changeagreement (SDC95%) with 95%CI were calculated to evaluate the measurement error of the sensor.
Results A total number of 261 cycles were performed. The ICC values were excellent for all measurement (ICC: 1.0; 95% CI 1.0, 1.0). One cycle of movement showed lower measure error (LoA) compared to single measures (left or right) (table 1). The measurement error was further improved by using the mean of 3 cycles (figure 1). The SDC95 was 0.87 for one cycle and 0.80 for the mean of three cycles.
Conclusions The sensor showed excellent criterion validity and reliability for rotation around the vertical axis in the range of motion from 10 to 120 degrees, especially for cycles and mean of 3 cycles. When assessing a single cycle or the average of three cycles, a change of 1 degree is needed for real change beyond measurement error. Our findings justify proceeding with further evaluations of the sensor in patients with axSpA.
Disclosure of Interest None declared