Asynchronous Teledentistry for Caries Detection: A Diagnostic Accuracy Study

Asynchronous teledentistry has the potential to address oral health disparities by enabling remote caries detection in underserved regions. This study aims to evaluate the diagnostic accuracy of asynchronous, image-based caries detection performed by general dentists in comparison with remote assessments by restorative dentists (reference standard) and clinical examinations by general dentists.

A cross-sectional diagnostic accuracy study was conducted in Mithi, Pakistan. Intraoral images (n = 3350) were captured using a custom-built mobile application, both with and without cheek retractors. After three months, two general dentists remotely assessed the images using a novel annotation tool (Cario-Label). Their findings were compared against a reference standard established by restorative dentists on remote images. The remote image-based assessments of the general dentists were also compared with the clinical examinations performed by them. Diagnostic performance metrics (accuracy, sensitivity, specificity, positive predictive values, negative predictive values) and Cohen's kappa were calculated.

Remote assessments by general dentists achieved an overall accuracy of 97% compared with the reference standard. Sensitivity ranged from 98% to 99% and specificity from 69% to 74%, with consistently high PPV (98%-99%). The presence or absence of cheek retractors did not significantly influence diagnostic performance. Agreement between remote assessments and clinical examinations was substantial, with Cohen's kappa values of 0.72 and 0.80.

Asynchronous teledentistry demonstrated high diagnostic accuracy for dental caries detection, comparable with expert remote assessments, and showed substantial agreement with clinical examination. These findings validate the reliability of image-based remote screening in resource-constrained environments.

This study highlights the feasibility of asynchronous teledentistry as a scalable, cost-effective strategy for early caries detection in underserved populations. By leveraging simple mobile tools for image capture and annotation, task-shifting to general dentists can be achieved without compromising diagnostic quality, ultimately supporting global efforts to improve oral health equity.