Abstract
The 5-Year-Olds' Index is considered to be a relatively simple and useful evaluation method; however, much remains unknown about its accuracy and reproducibility. We conducted a study on the accuracy and reproducibility of evaluations with this index.
In this study, evaluations were made by 21 dentists using standard models in a total of 10 cases, two each ((1), (2)) for each classification (Groups 1-5) in the 5-Year-Olds' Index distributed at a workshop held by the Japan Cleft Committee. Based on the results, weighted Kappa values were obtained for agreement with the correct responses and intra-rater agreement. In addition, we compared years of experience (A assessors: ≥ 10 years, B assessors: 2-9 years, C assessors: < 2 years) with the relationship of agreement with the correct responses. Agreement with the correct responses and tendencies for errors in evaluation were also examined for each standard model. Agreement with the correct responses (Kappa value) was “Very good” for A (0.83) and B (0.83) assessors, and “Moderate” for C assessors (0.53). Agreement with the correct responses was also compared by assessors' years of experience, and the results showed a significant difference between A and B assessors and C assessors (p < 0.05). The correct response rate for each of the standard models was the highest for Group 5-(1)), at 95.2%. The most common evaluation error was judging Group 3-(2) to be Group 4 (40.5%).
The standard models used in this study were appropriate as standard models for the 5-Year-Olds' Index. The results suggest that in order to raise the evaluation accuracy and reproducibility of the 5-Year-Olds' Index, it is necessary to provide thorough training on evaluation using the standard models and to increase opportunities for involvement in the treatment of children with unilateral cleft lip and palate. Repeated training for evaluation using the standard models is also considered to be important for correct assessment of the judgment criteria.
