School-based caries prevention using silver diamine fluoride: A pragmatic randomized trial in low-income rural children

Background: Dental caries is the world's most prevalent noncommunicable disease, disproportionately affecting children from low-income rural areas. This study assessed the effectiveness of using silver diamine fluoride (SDF) for school-based caries prevention. Methods: The CariedAway 3.0 study was a cluster-randomized pragmatic non-inferiority trial comparing SDF to sealants and atraumatic restorations (ART) for the prevention and control of dental caries. All participants also received fluoride varnish. Analysis consisted of multilevel mixed-effects logistic and negative binomial regression for the prevalence and incidence of dental caries, respectively, and a non-inferiority margin of 10% for the difference between groups was used. Dental caries was defined as an ICDAS score of four or greater. Results: A total of 3345 children were enrolled in the trial, however there was a large proportion of children who were noncompliant and received external dental care. In adjusted analyses of compliant participants (n=1083, consisting of 543 in the SDF group and 540 in the sealant and ART group), there was no difference in the weighted risk difference between treatment groups (B=0.003, 95% CI = -0.0001, 0.0008). The odds of caries was elevated in the SDF group in longitudinal analyses (OR = 1.35, 95% CI = 0.86, 2.11) but was not significant and was below the non-inferiority margin. There were no significant differences between groups for caries incidence in adjusted models (IRR = 1.19, 95% CI = 0.81, 1.74). Results for intent to treat analyses were similar to that of per-protocol. Discussion: In this school-based clinical trial, the prevalence of dental caries in children treated with SDF and fluoride varnish was non-inferior compared to those treated with sealants, ART, and fluoride varnish, although the overall risk was slightly higher. Unfortunately, a high rate of dropout and participant noncompliance was observed, likely due to the impacts of COVID-19 on study procedures. As a result, observed effects may be unreliable beyond the short-term.