Dental age in Japanese children using a modified Demirjian method

In a retrospective cross-sectional study, dental development of 1,620 healthy Japanese children, aged 3-15.9 years, were evaluated by the Demirjian method. A subset of 55 pantomograms were randomly chosen and re-examined. The intra-class correlation coefficient on maturity scores was 0.99. The Cohen's Kappa coefficient to agreement stage by teeth and overall agreement by stage resulted in average 0.85 and 0.82 respectively, both interpreted as "substantially reliable". The children were classified by sex and age. In each age group, dental age and chronological age was compared using paired t-test. Most of age groups were overestimated and had a significant difference. New standards for Japanese children were built using a logistic curve with the equation y = 100*{beta/(1 + e(-alpha(x-x0)))} as a basis. A residual analysis was made to verify if the curve was appropriate. The explained variance for the regression curve resulted in 94.25% for males and 95.07% for females. For both sexes, a linear regression line between chronological and estimated dental age showed 94% of association. The Demirjian standards were not suitable for the presentday Japanese children, so it was necessary to establish specific standard.