Correlation Between ICDAS II Clinical Scores and Digital Images Scores

Objective:The aim of the study is to determine how inter- and intra-examiner agreement using ICDAS II on clinical digital images compare to the inter- and intra-examiner agreement in the clinical setting. Methods:Currently 569 children ages 5-14 have been consented and enrolled in an NIH-funded study in Puerto Rico using ICDAS II criteria (5R01DE017890-02). Selected digital images (N=160) from tooth surfaces of children enrolled in that study were matched to their corresponding ICDAS II clinical data, which was obtained by an experienced ICDAS examiner at the time the image was taken. Twenty images were selected for each code 1-6, and 40 for code 0. Two PowerPoint presentations were made of all the selected images and presented, under standardized conditions, in two different days to three experienced examiners who had all been previously clinically calibrated against each other on that same population. Results: Repeatability of scores from the digital photographs was high for examiners 1 and 2 and acceptable for examiner 3: examiner 1, (k= 0.81, wk= 0.91); examiner 2, (k= 0.83, wk= 0.93); examiner 3, (k= 0.68, wk= 0.86). Agreement between the clinical examination and the three examiners using the digital photographs was high for examiners 1 and 2 and acceptable for examiner 3: examiner 1, (k= 0.72, wk= 0.87); examiner 2, (k= 0.73, wk= 0.88); examiner 3, (k= 0.50, wk= 0.72). Agreement between the digital examination scores of the secondary examiners and the primary clinical examiner was high for examiners 1 and acceptable for examiner 3: examiner 1, (k= 0.75, wk= 0.89); examiner 3, (k= 0.48, wk= 0.73). Conclusions: Overall intra- and inter-examiner agreement for scoring of the images and scoring of the images versus clinical scores was high or acceptable. These results indicate that using digital images as an introductory training tool for the ICDAS II criteria is an acceptable method.