De Decker S, Gielen IM, Duchateau L, et al.
Objective-To determine intraobserver, interobserver, and intermethod agreement for results of myelography, computed tomography-myelography (CTM), and low-field magnetic resonance imaging (MRI) in dogs with disk-associated wobbler syndrome (DAWS). Design-Prospective cross-sectional study. Animals-22 dogs with DAWS. Procedures-All dogs underwent myelography, CTM, and low-field MRI. Each imaging study was interpreted twice by 4 observers who were blinded to signalment and clinical information of the patients. The following variables were assessed by all 3 techniques: number, site, and direction of spinal cord compressions; narrowed intervertebral disk spaces; vertebral body abnormalities; spondylosis deformans; and abnormal articular facets. Intervertebral foraminal stenosis was assessed on CTM and MRI images. Intraobserver, interobserver, and intermethod agreement were calculated by kappa and weighted kappa statistics. Results-There was very good to good intraobserver agreement for most variables assessed by myelography and only moderate intraobserver agreement for most variables assessed by CTM and low-field MRI. There was moderate to fair interobserver and intermethod agreement for most variables assessed by the 3 diagnostic techniques. There was very good or good intraobserver, interobserver, or intermethod agreement for the site and direction of the worst spinal cord compression as assessed by all the imaging modalities; abnormal articular facets and intervertebral foraminal stenosis were the least reliably assessed variables, with poor interobserver agreement regardless of imaging modality used. Conclusions and Clinical Relevance-There was considerable variation in image interpretation among observers and between use of various imaging modalities; these imaging techniques should be considered complementary in assessment of dogs with DAWS.