Inter-rater reliability and preparation features for diverse grading scales analyzing computer-based grading of 1465 preparations Running title: Inter-rater reliability and preparation features analyzing 1465 preparations

Abstract

Abstract Background: Preparation exercises represent an essential part in the study of dentistry. Digital grading software can be used for evaluation of prepared artificial teeth, which is mainly based on limited number of teeth and raters. In order to improve a standardized evaluation of preparations, the aim of this study was to assess the inter-rater reliability of a high number of teeth by different raters and, to evaluate specific preparation features. Methods: Therefore, a total of 1465 preparations of an artificial lower molar were evaluated with grades from 1 to 5 by four dentists out of three dental schools, and inter-rater reliability was calculated. A specific analysis program was used for objective evaluation of the preparation. Data were analyzed using the Kolmogorov-Smirnov test, Pearson chi-square test, and reliability analysis with a level of significance of p= 0.05. Results: The median value of all grades was 2.75 (IQR: 1.35). Lowest grade variance of 0.00 within the raters’ evaluations was observed in grade group 1. Grade variance of evaluations with grades 1 and 2 differed significantly from all the other grade groups (p≤0.003). The inter-rater reliability factor of all grade groups together was 0.958. Conclusion: Inter-rater reliability was very high, which indicates very good correlation of the given grades. The objective evaluation of preparation features has worsened with ascending grade groups. Significant differences within the grade groups indicate higher agreement when evaluating good preparations. The objective feature convergence and preparation angle, marginal width, and stump height have a crucial impact on evaluation.