Reliability of quality measures of movement in lumbar spine flexion-extension radiography

Background data: For center of rotation (COR) and centroid length variables in lumbar flexion-extension radiography, enough reliability information is not available. Objective: The objective of this study was to determine the intra-rater reliability of COR and centroid length of lumbar intervertebral movement. Methods: A methodological research for reliability was designed. Fifteen males with chronic non specific low back pain aged between 22-43 years participated in this study. Study variables included COR of full, flexion and extension arc and centroid length. The proposed method from Putto was used for flexion-extension radiography. Five x-rays were taken in neutral, full flexion and extension, mid-flexion and mid-extension positions. The variables were calculated using Computer Aided Radiographic Analysis of Spine (CARA) software after scanning. The intraclass correlation coefficient (ICC) and standard error of measurement (SEM) were utilized for relative and absolute reliability respectively. Results: The results demonstrated high ICC values on x axis (range: 0.67-0.88) and y axis (range: 0.55-0.90) for full arc COR, high ICC values of flexion arc COR on x axis (range: 0.54-0.97) and y axis (range: 0.48-0.99). ICC values results of extension arc ranged between 0.19-0.60 for x axis and 0.16-0.83 for y axis. ICC values for centroid length ranged between 0.22-0.88. Conclusion: Results support previous studies which show very high reliability of full arc COR. In addition, flexion arc COR has desirable reliability and low measurement errors. But, extension arc COR and centroid length have moderate reliability and large measurement errors.