Looks like you’re on the US site. Choose another location to see content specific to your location

Home Dental NYU Tests Social-Determinant AI for Caries Risk Prediction
NEWS & BLOGS IMAGES (93)

NYU Tests Social-Determinant AI for Caries Risk Prediction

20th November 2026

Researchers at NYU College of Dentistry have tested statistical and machine learning models that predict caries risk from socio-economic and behavioural data alone, pointing to a new pathway for public health dentistry screening. Published online on 10 October 2026 in Caries Research, the study analysed 23,726 participants from Brazil’s national SB Brasil 2023 survey to assess whether non-clinical predictors could triage who needs examination first.

The models used household wealth, household density, private dental insurance, piped water access, time since last dental visit and self-rated oral health as predictors across three age cohorts: children aged 12, adolescents aged 15 to 19 and adults aged 35 to 44. Predictive performance improved with age, reaching moderate accuracy in the adult cohort, and models forecast untreated caries more reliably than cumulative DMFT experience. Notably, simpler statistical models matched or outperformed the more complex machine learning algorithms for DMFT prediction.

Dr Fernando Neves Hugo, chair of NYU’s Department of Epidemiology and Health Promotion, framed the commercial implication: “These are powerful tools that could be helpful for targeting resources.” The statistical case supports using the models to prioritise examinations over universal or no screening, with household wealth and density emerging as the most influential predictors. The team plans a parallel analysis using US NHANES data. For dental practice management vendors, insurance providers and public health commissioners, social-determinant risk scoring opens a route to triaging preventive care without requiring clinical examination first.

The commercial signal is a widening role for social-determinant data in dental risk stratification, where lightweight statistical models may rival ML complexity in resource-constrained public health deployments. US NHANES replication and integration into practice management or insurance workflows remain critical near-term milestones. Expect preventive care targeting algorithms to move into public dental health commissioning conversations through 2027, particularly across middle-income markets.

For the latest updates and in-depth insights into the world of Dentistry, including breakthrough treatments, industry trends, and regulatory news, contact Kiera Allen today! 

 

Stay informed

Receive the latest industry news, Tips and straight to your inbox.

wpChatIcon
wpChatIcon