ICDAS Coding by Dental Faculty Using Images and Extracted Teeth

Objectives: To compare the inter- and intra-examiner reproducibility and the accuracy of ICDAS coding by restorative and non-restorative dental faculty using either photographic images or extracted teeth as tools for training.

 Materials and Methods: Thirty five extracted permanent human teeth with a range of ICDAS codes were carefully cleaned and a suitable surface on each selected for the study. Camsight images were taken of those surfaces. The ICDAS codes for the surfaces of these reference sets of teeth and images were validated by an international expert. All full-time dental faculty were trained using an E-learning program, extracted teeth exercises and photograph discussions and assigned into one of two groups, restorative or non-restorative specialty. Thereafter, on one occasion they coded the selected surfaces of the teeth twice and on another the images, also twice. Kappa (weighted: Wkappa) statistics and Yates corrected Chi Square tests (SPSS 17.0) were used to analyze the data.

Results: For extracted teeth, inter-examiner Wkappa was good at 0.69-0.89 for restorative dental faculty and at 0.57-0.82 for non-restorative dental faculty, the comparable intra-examiner values also being good at 0.76-0.88 and 0.48-0.83.

There was no significant difference between the ICDAS coding performances (p>0.05) of restorative and non-restorative faculty.

However, as a training tool the set of extracted teeth was statistically significantly better than the set of photographic images with restorative dental faculty being more accurate in detecting code 0, 2 and 3 lesions (p<0.05).

Conclusion: The ICDAS can be learned and undertaken by dental faculty of different specializations with a set of extracted teeth being a better training tool than a set of photographic images.