An Observational study of changes in cervical inter-vertebral motion and the relationship with patient-reported outcomes in patients undergoing spinal manipulative therapy for neck pain.

Abstract

Spinal manipulation is a commonly sought therapy for neck pain. The present work examined sagittal plane cervical inter-vertebral kinematics in patients and healthy volunteers to explore whether motion is different in patients with neck pain, if manipulation is associated with changing cervical kinematics, and if changes are related to patient-reported outcomes. A standardised quantitative fluoroscopy (QF) image acquisition protocol for the cervical spine was developed and tested. A model of a cervical segment with a rigidly fitted digital inclinometer was rotated in the sagittal plane whilst being fluoroscopically imaged and QF results were compared for accuracy to that of the inclinometer. QF imaging sequences from ten subjects were analysed twice, six weeks apart, to assess repeatability. Finally, 30 patients and 30 age/gender-matched healthy volunteers had two cervical spine QF assessments four weeks apart. Only patients received spinal manipulation and completed patient-reported outcome measures (PROMs). Kinematic variables of interest included IV-RoM, segmental hypo-mobility, paradoxical motion, instantaneous axis of rotation (IAR) location, and laxity/attainment rate. The acquisition protocol allowed for imaging sequences to be achieved in a manner acceptable to participants. QF was found to be accurate to 0.5° for rotational range of motion. Intra- and inter-observer repeatability studies revealed substantial agreement and reliability for the QF measurement of C1 to C6 rotational motion (largest standard error of measurement (SEM) = 1.14°, lowest intra-class correlation coefficient (ICC) = 0.895) but not for IAR location (largest SEM = 7.66mm, lowest ICC = -0.080). Agreement and reliability were moderate-substantial for laxity/attainment rate (largest SEM = 0.04, lowest ICC = 0.70). There were no significant differences at baseline between patients and healthy volunteers in IV-RoM, or in the number of hypo-mobile, paradoxical or lax motion segments. Spinal manipulation was weakly associated with IV-RoM increases above the minimum detectable change calculated from healthy volunteers, in a dose response manner (Rho = 0.39 (95% CI: 0.014 to 0.663) p = 0.04). While the majority (87%) of patients reported clinically significant reductions in pain and disability,