Reliability and measurement error of sagittal spinal motion parameters in 220 patients with chronic low back pain using a three-dimensional measurement device

Abstract Background context A basic premise for any instrument measuring spinal motion is that reliable outcomes can be obtained on a relevant sample under standardized conditions. Purpose The purpose of this study was to assess the overall reliability and measurement error of regional spinal sagittal plane motion in patients with chronic low back pain (LBP), and then to evaluate the influence of body mass index, examiner, gender, stability of pain, and pain distribution on reliability and measurement error. Study design/setting This study comprises a test-retest design separated by 7 to 14 days. Patient sample The patient cohort consisted of 220 individuals with chronic LBP. Outcome measures Kinematics of the lumbar spine were sampled during standardized spinal extension-flexion testing using a 6- df instrumented spatial linkage system. Methods Test-retest reliability and measurement error were evaluated using interclass correlation coefficients (ICC 1,1 ) and Bland-Altman limits of agreement (LOAs). Results The overall test-retest reliability (ICC 1,1 ) for various motion parameters ranged from 0.51 to 0.70, and relatively wide LOAs were observed for all parameters. Reliability measures in patient subgroups (ICC 1,1 ) ranged between 0.34 and 0.77. In general, greater (ICC 1,1 ) coefficients and smaller LOAs were found in subgroups with patients examined by the same examiner, patients with a stable pain level, patients with a body mass index less than below 30 kg/m 2 , patients who were men, and patients in the Quebec Task Force classifications Group 1. Conclusions This study shows that sagittal plane kinematic data from patients with chronic LBP may be sufficiently reliable in measurements of groups of patients. However, because of the large LOAs, this test procedure appears unusable at the individual patient level. Furthermore, reliability and measurement error varies substantially among subgroups of patients.