Even though radiography is one of the most frequently used imaging techniques for orthopaedic disorders, it has been demonstrated that the interpretation can vary between assessors. As such, the purpose of this study was to examine the intraobserver and interobserver agreement and the influence of level of expertise on the interpretation of radiographs of the stifle in dogs with and without cranial cruciate ligament rupture (CCLR). Sixteen observers, divided in four groups according to their level of experience, evaluated 30 radiographs (15 cases with CCLR and 15 control stifles) twice. Each observer was asked to evaluate joint effusion, presence and location of degenerative joint disease, joint instability and whether CCLR was present or absent. Overall, intraobserver and interobserver agreement ranged from fair to almost perfect with a trend towards increased agreement for more experienced observers. Additionally, it was found that stifles that were classified with high agreement have either overt disease characteristics or no disease characteristics at all, in comparison to the ones that are classified with a low agreement. Overall, the agreement on radiographic interpretation of CCLR was high, which is important, as it is the basis of a correct diagnosis and treatment.

No abstract