RESUMO
Background: Arthroscopy of the temporomandibular joint (TMJ) plays a long-established role in the diagnostics and therapy of patients suffering from arthrogenic temporomandibular disorders (TMDs), which do not respond adequately to conservative/non-invasive therapy. However, the interpretation of arthroscopic findings remains challenging. This study investigates the reliability and variability of assessing arthroscopic views of pathologies in patients with TMDs by non-specialists in arthroscopy and whether a standardized assessment tool may improve correctness. Methods: Following a comprehensive one-semester lecture, dental students in the clinical stage of education were asked to rate 25 arthroscopic views (freeze images and corresponding video clips) regarding the severity of synovitis, adhesions, and degenerative changes on a scale of 0-10 (T1). The results were compared to ratings stated by two European-board-qualified academic OMF surgeons. In a second round (T2), the students were asked to repeat the ratings using a 10-point rating scheme. Results: With regard to all three subcategories, congruency with the surgeons' results at T1 was at a low level (p < 0.05 in 19/75 cases) and even decreased at T2 after the implementation of the TMDs-SevS (p < 0.05 in 38/75 cases). For both T1 and T2, therefore, the inter-rater agreement was at a low level, showing only a slight agreement for all three subcategories (Fleiss' Kappa (κ) between 0.014 and 0.099). Conclusions: The judgement of the arthroscopic pathologies of the TMJ remains an area of temporomandibular surgery that requires wide experience and training in TMDs to achieve expertise in TMJ arthroscopic assessments, which cannot be transferred by theoretical instruction alone.