Intraoral Radiographic Interpretation Agreement Between Veterinary Students, Veterinary Dental Residents and Veterinary Dental Specialists.

This study evaluated observer variations in the interpretation of radiographic evidence for periodontal disease, tooth resorption, and endodontic disease in dogs. Forty dental radiographs were evaluated for 12 different parameters (periapical destruction of bone, wider than expected root canal, narrower than expected root canal, apical root resorption, loss of alveolar bone, external surface resorption, external replacement resorption, external inflammatory resorption, external cervical root resorption, internal surface resorption, internal replacement resorption, internal inflammatory resorption). Interpretations by 20 veterinary dentists, 10 veterinary dental residents, and 10 veterinary students were analyzed for consistency within groups and between groups by intraclass correlation coefficients (ICCs [95% CI]). Additionally, the agreement rate between groups compared to histopathological diagnosis, when available, and to a consensus group were evaluated. The results showed fair to good interobserver agreement for all participants when looking at all questions simultaneously (0.578 [0.515-0.635]) and with the consensus group (0.483 [0.451-0.517]). However, questions pertaining to various types of tooth resorption scored the lowest ICCs ranging from 0.005 (-0.311 to 0.321) to 0.189 (-0.105 to 0.402) across individual groups. Students had the lowest agreement compared to the consensus group for all questions (0.383 [0.347-0.421]) with fair to good agreement involving groups of residents (0.501 [0.465-0.538]), recently boarded diplomates (0.541 [0.506-0.578]), and more experienced diplomates (0.545 [0.510-0.582]). While dental radiographs are essential for clinical decision making, this study shows that interpretation of radiographs is highly subjective.
Competing Interests: Declaration of Conflicting InterestsThe authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.