Abstract
This methodological study presents the development and implementation of technological tools for the online training and calibration of dentists participating in the SB Brasil 2023 survey, using the Moodle® platform. The training and in-lux calibration process employed 10 and 25 sets of photographs, respectively. Conditions including dental crown status (dmft and DMFT indexes), caries consequences (pufa/PUFA index), malocclusion (canine relation, overjet, overbite, posterior crossbite, and Dental Aesthetic Index), and dental trauma were evaluated, based on WHO or SB Brasil 2010 criteria. Software was developed to record codes, automated calculations of agreement coefficients (overall percentage, score-specific, and simple and weighted kappa), and generate reports. Examiners were allowed to repeat calibration attempts until they achieved a minimum agreement (Kappa ≥ 0.61). In total, 1,513 examiners used the software, and 728 successfully completed the calibration with substantial or higher agreement across all conditions/indexes. Individual reports described discrepancies and monitored attempts. Occlusal conditions had the lowest percentages of almost perfect agreement and required more attempts to achieve calibration. The technological tools implemented in SB Brasil 2023 enabled online training and calibration, promoting consistency among examiners for fieldwork. These findings demonstrate the feasibility of remote strategies for calibration in epidemiological surveys, particularly in scenarios involving multiple geographically distributed examiners, with potential applications in other contexts and health areas.