The Influence of Examiner Experience on Inter- and Intra-Rater Agreement in Dental Restorations Identification on Three-Dimensional Digital Models.

Background and Objectives: In recent decades the prevalence of dental caries has continued to increase despite widespread access to modern dental care. This study aimed to evaluate the diagnostic relevance of differentiating healthy, carious, and restored dental structures using intraoral scans-a non-irradiating imaging technique. Materials and Methods: A cross-sectional reliability (diagnostic agreement) study was carried out. All 36 examiners underwent pre-calibration on the ICDAS scores recordings. They filled in ICDAS files for each randomly assigned patient through three different methods: clinical examinations and three-dimensional digital models obtained with two different intraoral scanners. Cohen's weighted kappa test and Prevalence and Biased Adjusted Kappa (PABAK) were utilized to evaluate agreements. A corresponding p-value for agreement was computed for each agreement coefficient. Results: When ICDAS values recorded using intraoral scanners were compared for examiners with different experience levels, there was an inter-rater substantial agreement for all teeth (PABAK = 0.688; CI 95% = 0.344-0.894), but also for anterior teeth only, as well as for smooth surfaces only. Regardless of the examiner's clinical experience, PABAK inter- and intra-rater agreement on fillings identification on three-dimensional digital models obtained by intraoral scanning were at least moderate in most cases. Conclusions: Three-dimensional digital models offer reliable diagnostic information, especially for experienced clinicians, supporting their use as a standardized tool in routine practice to obtain an accurate, dynamic view of patients' caries status, although limitations such as controlled study conditions and variability in scanning algorithms across systems must also be acknowledged in daily dental diagnostics.