[Comparison of reproducibility measurements for calibration of dental caries epidemiological surveys].

This study compares three measurements (Kappa, general agreement percentage, or GAP, and dice index) used to determine the reproducibility of caries diagnosis in epidemiological surveys under different clinical diagnostic thresholds. Eleven examiners with previous experience in epidemiological surveys were submitted to a theoretical and clinical calibration process. Data analysis used two caries detection thresholds: World Health Organization (WHO) and WHO with the inclusion of initial enamel lesions (WHO + IL). Twenty-three children 6-7 years of age were examined, with and without caries. Mean values for Kappa index, GAP, and Dice were considered high (> 0.90), except for the dice index for the WHO + IL threshold (0.69). Since Kappa is an adjusted agreement index, it can be considered the instrument of choice for calibration of examiners. However, when it is impossible to use, the GAP is recommended together with the dice index in order to orient and improve examiners when examining caries lesions.