Precision of spinal radiographs as a screening test for intervertebral disc calcification in Dachshunds

Abstract

Among dog breeds, the Dachshund has the highest lifetime incidence of intervertebral disc disease (IVDD). Intervertebral disc (IVD) calcification is an indicator of severe degeneration that predisposes to disc herniation. IVDD is heritable in Dachshunds, and in some countries, breeding candidates are screened to reduce IVDD occurrence by selecting dogs according to their score of radiographically detectable intervertebral disc calcification (RDIDC) and excluding dogs with ≥5 RDIDCs from breeding. This study evaluated the precision of scoring spinal radiographs for IVD calcification and subsequent classification of Dachshund dogs for breeding based on their RDIDC score. Digital radiographs of the spine were obtained in 19 clinically healthy, young adult Dachshunds, and scored for RDIDC independently by five scorers with varying levels of experience, three times each. Within scorer (repeatability) and between scorer (reproducibility) variability was estimated both at the individual IVD level and at the whole dog level for breeding classification purposes.At the IVD level, some degree of scorer effect was supported by the pairwise repeatability (92.3%; 95% CI: 88.8–94.7%) being marginally higher than the reproducibility (89.2%; 95% CI: 85.7–91.8%). Scorer-specific patterns confirmed the presence of scorer subjectivity. Repeatability significantly increased with scorer experience but the reproducibility did not. RDIDC scoring repeatability and reproducibility substantially decreased at the cervicothoracic spine region, likely due to anatomical superimpositions. At the dog level, a breeding classification could be repeated by the same scorer for 83.6% (95% CI: 73.8–90.2%) of the dogs, and was reproduced between two scorers for 80.2% (95% CI: 66.6–89.1%) of the dogs. The repeatability of breeding classification also seemed to increase with scorer experience but not the reproducibility. Overall, RDIDC scoring revealed some degree of inconsistency explained by scorer subjectivity and inexperience, and anatomical superimpositions. Scorer training and experience is strongly recommended to improve test precision and ensure appropriate classification of Dachshunds for breeding.