A new approach for interexaminer reliability data analysis on dental caries calibration.

Objectives: a) to evaluate the interexaminer reliability in caries detection considering different diagnostic thresholds and b) to indicate, by using Kappa statistics, the best way of measuring interexaminer agreement during the calibration process in dental caries surveys.

Methods: Eleven dentists participated in the initial training, which was divided into theoretical discussions and practical activities, and calibration exercises, performed at baseline, 3 and 6 months after the initial training. For the examinations of 6-7-year-old schoolchildren, the World Health Organization (WHO) recommendations were followed and different diagnostic thresholds were used: WHO (decayed/missing/filled teeth - DMFT index) and WHO + IL (initial lesion) diagnostic thresholds. The interexaminer reliability was calculated by Kappa statistics, according to WHO and WHO+IL thresholds considering: a) the entire dentition; b) upper/lower jaws; c) sextants; d) each tooth individually.

Results: Interexaminer reliability was high for both diagnostic thresholds; nevertheless, it decreased in all calibration sections when considering teeth individually.

Conclusion: The interexaminer reliability was possible during the period of 6 months, under both caries diagnosis thresholds. However, great disagreement was observed for posterior teeth, especially using the WHO+IL criteria. Analysis considering dental elements individually was the best way of detecting interexaminer disagreement during the calibration sections.