Title: Agent-based Model of Fluoride Effects on Caries Development in Children
Brenda Heaton (Presenter)
Sarah Cherng, Boston University
Woosung Sohn, Boston University
Raul Garcia, Boston University
Sandro Galea, Boston University
Objectives: We developed an agent-based model simulating the progression of sound primary teeth to decayed, filled or missing due to decay in order to explore the impact of intervention processes within a hypothesized system of caries development in young children.
Methods: The model simulated a hypothetical cohort of 2,500 children, following caries development and outcomes in their primary teeth. Model feedback mechanisms included the following: increases in the number of teeth with untreated decay resulted in increases in the sound-to-decayed rate; receipt of fluoride decreased the probability of transitions from sound to decayed. To estimate baseline parameters for each child, we fit the model with corresponding 2013-2014 National Health and Nutrition Examination Survey data. We evaluated three scenarios under which fluoride effects could be introduced – 1) at the time of restorative care, and at regular intervals for children with 2) no caries experience or 3) with caries experience, at baseline.
Results: The NHANES data fall within the 95% confidence interval of our model results, indicating good model fit. The proportion of teeth with untreated decay in our data-fitted model increased from 2% at baseline to 11.6% at the end of the model run. Compared to the total population, fluoride administered only at the time of restorative care, at regular intervals for children with sound teeth, or caries experience at baseline resulted in 8.6%, 25% and 26% decreases, respectively. Time trends reveal that while fluoride minimizes the development of caries for children without caries experience, it prevents increases in untreated decay for those with caries experience at baseline.
Conclusions: Our results support findings related to fluoride effects in research literature. Computational models such as ours allow us to account for complex dynamics and feedbacks between the person- and tooth-level that contribute to caries development and outcomes, facilitating future evaluation of multi- level contributors affecting caries.