Modeling Caries Risk in Formerly Sealed Teeth

Objective: We are not aware of any published studies to date that have directly compared caries risk in never sealed teeth (NS) to that in formerly sealed teeth that have fully or partially lost their sealant (FS). This exploratory analysis examined the risk of caries in FS relative to the risk in NS.

Methods: We obtained data on sealant retention and probability of caries in sealed and never sealed teeth from the five split-mouth studies of resin-based sealant included in the Cochrane Review of the effectiveness of sealants in preventing caries in the permanent teeth of children. Assuming that retained sealants are 100% effective, the probability of caries in FS would equal (% sealed teeth with caries at follow up)/(1 - complete retention rate). For each study, we calculated the risk of caries in FS relative to NS. We assumed teeth were independent when calculating the standard error. We used a weighted bivariate model to estimate the summary relative risk ratio (RR), (Probability caries in FS)/(Probability caries in NS) and its 95% confidence interval (CI). We tested for heterogeneity using the quantity I2.

Results: The number of NS and FS teeth were 1405 and 303, respectively. Retention ranged from 79% to 92% (1 year after placement) and from 39% to 74% (3 years after placement). The RR was 0.95 (CI: 0.69-1.29; 3 studies) 1 year after placement, 1.13 (CI: 0.96-1.34; 3 studies) 2 years after placement, and 0.96 (CI: 0.85-1.10; 3 studies) 3 years after placement. Heterogeneity was significant in all but the first year after placement.

Conclusions: Findings suggest that for this group of studies, teeth with fully or partially lost sealant were not at higher risk for caries than were never sealed teeth.