Abstract
Background Evidence-based non-surgical interventions to halt the progression of dental caries, the most prevalent noncommunicable disease in the world, include atraumatic restorative treatment (ART) and silver diamine fluoride (SDF). Data are needed on their effectiveness when used in school caries prevention programs.
Methods In this school-based, cluster-randomized pragmatic trial conducted from February 1, 2018 to June 1, 2023, 48 primary schools in New York City were randomly assigned to receive either silver diamine fluoride or atraumatic restorations for untreated caries on any mesial, occlusal, distal, buccal, and lingual surface of permanent molars, premolars, and primary molars. Children were treated by either dental hygienists, pediatric dentists, or medical nurses (SDF group only). Dental caries was diagnosed as any lesion scoring either 5 or 6 on the ICDAS scale. The primary outcome was the caries control rate.
Results A total of 7418 children were enrolled in the trial, of which 2338 (1193 in the SDF group, 1145 in the ART group) presented with treatable dental caries and completed at least one follow-up observation. The total surface-level failure in the SDF group was 32.5%, compared to 32% in the ART group. There were 461 failures observed in SDF participants over 1105 person-years, compared to 431 ART failures over 1039 person-years, yielding an incidence rate ratio of 1.01 (95% CI = 0.88, 1.15). Within the SDF group, children treated by nurses were at lower risk of failure than those treated by other clinicians (IRR = 0.75, 95% CI = 0.59, 0.96). Additionally, children with more affected surfaces had higher chances of failure (OR = 1.11, 95% CI = 1.09, 1.13), though this did not differ across treatment groups.
Conclusions In this four-year pragmatic trial of school-based utilization of minimally invasive interventions for dental caries, similar control rates were observed in children receiving either SDF or ART. These results support the use of secondary preventive therapies for school dental programs.
1 Introduction
Dental caries (tooth decay) is a worldwide public health crisis, afflicting billions of children and adults [1] who often lack access to effective preventive or therapeutic care [2]. It is also highly inequitable, as those from low-income or minority families shoulder a disproportionate burden of disease [3]. Untreated caries may increase the risk of systemic noncommunicable diseases including cancer, diabetes mellitus, cardiovascular diseases, and neurodegenerative conditions [4]. Caries also affects child development, reducing educational performance [5] and oral health-related quality of life [6], and is attributed to over thirty million hours of lost seat time in schools per year [7].
Integrating preventive and therapeutic dental services into schools can increase access to care, reduce the risk of caries, and may improve educational performance [8]. The U.S. Department of Health and Human Services’ Community Preventive Services Task Force recommends school sealant programs to prevent dental caries [9], and the Centers for Disease Control and Prevention funds school-based sealant programs (SBSPs) in multiple states. These programs are both clinically and cost-effective [10; 11]. However, managing existing caries in children who are unlikely to seek out traditional, office-based care remains a critical issue in dental public health.
Silver diamine fluoride (SDF) and glass ionomer atraumatic restorative treatment (atraumatic restorations, or ART) are minimally invasive interventions that can effectively arrest or control caries, and are included on the World Health Organizations’ Model List of Essential Medicines. SDF is a topical solution that inhibits the growth of cariogenic bacteria and contributes to the remineralization of enamel and dentine caries [12], whereas atraumatic restorative treatment removes decalcified tissue with hand instruments before applying adhesive fillings to restore the cavity [13]. Silver diamine fluoride is estimated to arrest anywhere from 47 to 90 percent of caries lesions after one application [14] and is considered to be a practical, affordable approach to community caries prevention, particularly in low-socioeconomic areas [15]. In contrast, the average annual failure rate for single- and multiple-surface ART restorations in primary molars is estimated to be 5 and 17%, respectively [16], though this may be outperformed by conventional restorations [17]. Like SDF, prior studies of ART in community settings conclude that it is acceptable and effective in socioeconomically deprived groups [18].
The CariedAway study was a pragmatic randomized trial that assessed SDF and ART for the control of dental caries when used in a school-based oral health program [19]. In this paper, we estimate the overall and comparative effects of treatment on caries control over four years. A secondary objective was to determine whether posterior application of silver diamine fluoride resulted in subsequent anterior caries control.
2 Methods
CariedAway was a longitudinal, cluster-randomized pragmatic trial conducted from February 1, 2018 to June 1, 2023 in eligible elementary schools in New York, NY, USA [20]. The study received IRB approval from the New York University School of Medicine and is registered at www.clinicaltrials.gov (#NCT03442309). CariedAway results are reported following the Enhancing the Quality and Transparency of Health Research (EQUATOR) guidelines.
Participants
Any school in the New York City geographic area with a student population consisting of at least 50% black and/or Hispanic/Latino and at least 80% receiving free and reduced lunch was eligible to participate. These inclusion criteria were used as they typically have the highest burden of disease in the New York metropolitan area. Amongst participating in schools, any child with parental informed consent and child assent was enrolled in the study. While care was provided to any child meeting these criteria, analysis was restricted to those aged 5-13 years.
Randomization
Schools were block randomized to either the experimental or active control condition using a random number generator performed by RRR and verified by TBG. All children in a school received the same treatment.
Interventions and Procedures
For untreated caries, study participants received either silver diamine fluoride followed by fluoride varnish (to mask the bitter aftertaste of SDF) or glass ionomer cement (GIC) atraumatic restorative treatment followed by fluoride varnish. In the experimental group, a 38% concentration SDF solution (2.24 F-ion mg/dose) was applied to any posterior, asymptomatic, cavitated lesions using a microbrush. In the active control group, placement of GIC atraumatic restorations were applied to all frank, asymptomatic, cavitated lesions. Additionally, dental sealants were applied to sound dentition in the active control, but were not included in analysis. Treatments were provided in a private, dedicated room in a school using portable dental mirrors, explorers, and dental chairs with a frontal light source. All treatments were administered by either dental hygienists (SDF and ART), a supervising pediatric dentist (SDF or ART), or registered nurses (SDF) under patient-specific standing orders signed by a supervising dentist. Examination and treatment were provided biannually, except for disruptions in care due to COVID-19.
Data Collection and Diagnosis Protocol
Demographic data including age, sex, and race/ethnicity were obtained prior to examination from informed consent documents. Caries diagnosis was performed following a visual-tactile dental screening according to the International Caries Detection and Assessment System (ICDAS) adapted criteria in epidemiology and clinical research settings [21]. Any lesion with an ICDAS score of 5 (distinct cavity with visible dentine) or 6 (extensive/more than half the surface distinct cavity with visible dentine) was recorded as caries. At each observational visit, the first instance of dental caries was recorded and treated following the described clinical protocols. SDF was applied at each observational visit regardless of surface caries status, while ART was only reapplied if failed.
Outcome Definition
Caries control was determined if a carious surface (mesial, occlusal, distal, buccal, and lingual) on treatable teeth (permanent molars, premolars, and primary molars) received the assigned treatment and did not present with recurrent caries at any subsequent observation over the course of the study. Any caries recurrence was classified as control failure, and any surfaces on which the first observation of caries occurred at the subjects’ last study visit were removed from analysis. The per-person control rate
was then computed as the number of controlled surfaces divided by the total number of treated surfaces, and the per-person failure rate was computed as
The time elapsed for each controlled or failed surface from initial treatment was then recorded.
Power
As caries control over four years was not a primary endpoint of the CariedAway trial, an a priori power analysis was not performed.
Statistical Analysis
Descriptive statistics for the full study enrollment and for those participants with at least one follow-up observation and at least one carious surface were computed overall and by treatment group (means, standard deviations, and frequencies). We calculated the total number of controlled and failed surfaces observed for each study participant over a maximum of four years of followup, estimating the incidence rate ratio of failure. Controlled and failed surfaces were also determined by sex, race/ethnicity, and whether the participant was treated by either a dental hygienist, pediatric dentist, or registered nurse (SDF group only). To account for potential within-school correlation, we subsequently analyzed failure rates using binomial regression with cluster adjustment. To determine the effect of disease severity on failure rates we included a predictor for the per-person number of carious surfaces and an interaction effect between disease severity and treatment. We then considered the possibility of recurrent surface failure using multi-event survival analysis with predictors for treatment type and selected participant demographic variables. Finally to determine the association between posterior SDF application and anterior surface caries control, we restricted the sample to those participants assigned to the silver diamine fluoride group that presented with both posterior and anterior decay on any surface at baseline. We then calculated the number of decayed anterior surfaces at baseline and number of controlled surfaces at first follow-up and measured the strength of the association using Pearson’s correlation. Statistical analysis was conducted in R v.4.3.3. Statistical significance was determined at p<0.05.
3 Results
A total of 7418 participants were enrolled in the CariedAway trial, consisting of 3739 (50.4%) randomized to the silver diamine fluoride group and 3679 (49.6%) randomized to the atraumatic restoration group (Table 1, Figure 1). The average baseline decay prevalence was 26.7%, and the average age at baseline was 7.6 years (SD=1.9). Over 65% of enrolled participants reported being from either Hispanic/Latino or black race/ethnicities. For the analytic sample (participants with at least one follow-up observation and presented with at least one surface-level carious lesion over the course of the trial), 2338 participants met inclusion criteria consisting of 1193 (51%) randomized to the SDF group and 1145 (49%) randomized to the ART group. Approximately 85% of the analytic sample had initial caries at baseline, while the remaining 15% presented with initial caries at a subsequent observation. The analytic sample was equally distributed with respect to baseline decay, sex, age, and race/ethnicity.
The distribution of surface-level caries in the analytic sample (Figure 2) indicate that the majority of participants had between one and six surfaces with decay (e.g., 314 subjects presented with one carious surface, 494 subjects with two carious surfaces, and 258 subjects with three carious surfaces). Within the SDF group, there were 7557 total carious surfaces that received treatment (Table 2). Of these surfaces, 2456 presented with subsequent decay, for a failure rate of 32.5%. In the ART group, 6634 total carious surfaces were treated, of which 2123 presented with later decay, for a failure rate of 32%. The person-level failure rates were slightly higher: in the SDF group, 461 participants experienced any control failure (38.6%), compared to 431 participants in the ART group (37.6%). When considering the total time of observation (1105 person-years in the SDF group, 1039 person-years in the ART group), the incidence rate was 417.2 and 414.8 for SDF and ART groups, respectively, for an incidence rate ratio of 1.0057 (95% CI = 0.88, 1.15, p = .9321). Similar rates were found across treatments by sex and racial and ethnic groups.
Within the SDF group, the surface level failure rates were 24% and 35% for children with caries treated by registered medical nurses and dental hygienists, respectively. At the individual level, the incident rate ratio for any failure comparing nurses to hygienists was 0.75 (95% CI = 0.59, 0.96).
After adjusting for the potential clustering within school (Table 3), there remained no differences in failure rates between treatment groups (OR = 1.04, 95% CI = 0.76, 1.42). Disease severity was significantly related to control failure, with each additional decayed surface increasing the odds of failure by approximately 11% (OR = 1.11, 95% CI = 1.09, 1.13). However, there was no significant interaction between treatment and disease severity, indicating that severity results in equivalent likelihood of failure whether treated with silver diamine fluoride or atraumatic restorations (OR = 0.97, 95% CI = 0.93, 1.01).
Adjusted odds of failure by treatment and disease severity.
Finally, there were 462 anterior surfaces with decay at baseline across 108 participants that had both posterior and anterior decay and received silver diamine fluoride on posterior carious lesions. In these participants, 920 posterior teeth received SDF. At the first follow-up observation, six anterior surfaces were arrested (0.013%) across two subjects. There was no correlation between anterior surface arrest and the number of posterior teeth treated with SDF (r = -0.001, p = 0.991).
4 Discussion
The CariedAway study was a school-based clinical trial of minimally invasive treatments for caries management. Over 2300 children received either silver diamine fluoride or atraumatic restorations on more than 14000 carious tooth surfaces. The surface- and participant-level failures rates were 32.5 and 38.6% in the SDF group and 32 and 37.6% in the ART group, respectively. Nurses were also significantly more effective than dental hygienists in treating caries with SDF.
While the use of SDF and ART in United States schools is limited, prior school-based application of atraumatic restorations reported reductions in the per-visit risk of caries by approximately 20% in children with untreated decay [22], and a single application of silver diamine fluoride or ART halted caries progression in 56% and 46% of high-risk schoolchildren two years after initial treatment, respectively, [23], which was statistically non-inferior. These previous estimates were computed at the individual-level, while our data are based on more robust surface-level failure rates and include participants with a total time of observation ranging from six months to 3.5 years. Results are also similar to international studies, where SDF has been provided as part of a preschool oral health program in Hong Kong, demonstrating both clinical effectiveness and a high rate of acceptance among parents [15]. Similarly, an ART program in Hong Kong kindergartens reported 30-month restoration survival rates ranging from 79% to 33% [24], depending on restoration classification, and one in Brazilian public schools estimated a six-month success rate of approximately 82% [25]. Outside of school settings, both silver fluoride and ART were found to be effective in a community trial in Australian Aboriginal children [26], and a mobile dental program operating in underserved Mexican communities reported a 66% retention rate for atraumatic restorations after two years [27].
The effectiveness of using SDF and ART for caries management in schools may be modified by the severity of disease, such as size and depth of the carious lesion. The American Academy of Pediatric Dentistry (AAPD) reports a wide range of caries arrest rates for SDF, with effectiveness varying by cavity size and tooth location [14], and notes the arresting capacity of SDF may be greater in less severe lesions. For ART, a systematic review of atraumatic restorative treatment in children estimates a 71% and 67% success rate after one and two years, respectively, but the success of multiple surface restorations was significantly lower than single-surface ones [27]. Accordingly, the AAPD policy reports that single surface applications have higher survival than multi-surface restorations [28]. Caries diagnosis in the present study was based on an ICDAS score of 5 or 6, but we were unable to stratify by severity in analysis. However, differences in arrest attributed to treatment may also be affected by disease severity [29]. In adjusted analyses, disease severity significantly increased the chances of failure but had no appreciable effect across groups, suggesting that more severe cases tend to impact both treatments equally.
Frequency of application is also likely to impact caries control. A review of SDF studies suggests semiannual application [30] to mitigate the reduction in effectiveness for caries arrest that may occur over time, while the AAPD Clinical Practice Guidelines recommends a single application with followup after two to four weeks, reapplying where indicated [14]. Other approaches use more intensive early application, such as weekly application over a three week period, in an attempt to enhance arresting effects [31]. Similar to other school-based dental programs [32], the clinical protocols for CariedAway stipulated semiannual application of both SDF and ART (if not retained, in the case of atraumatic restorations). However, this initial schedule was disrupted due to school closings related to the COVID-19 pandemic. As a result, the elapsed time between treatments had considerable variability for enrolled participants. Regardless, the observed failure rates are lower than prior estimates based on a single application [23]. This supports the general recommendation that silver diamine fluoride is more effective with repeated application, particularly for community-based prevention programs.
Prior research of SDF in schools found that nurses may be similar to hygienists in the prevention of new caries in children with [19] and without caries at baseline [33], and a single application of SDF similar control rates between providers after two years [33]. Our results support the use of nurses in the application of SDF for untreated caries, outperforming dental hygienists in the four year control rate. A prior survey of nurses concluded that nurses believe they should provide oral health services for low-access children and are willing to expand their scope of practice to include SDF, provided that education and training are available [34]. While the present data on nurse-SDF application does not address the potential impact on their professional practice, the number of nurses in schools across the United States—equivalent to approximately three quarters of all dental hygienists nationwide [35]—could be leveraged to further increase access to care.
The direct effects of SDF has received considerable study [12; 36; 37]: the mechanistic action of silver diamine fluoride includes reducing the growth of cariogenic bacteria through the antibacterial effect of silver, and also by promoting the remineralization of enamel and dentine caries [12]. Silver ions in SDF are thought to prevent bacterial aggregation through reaction with bacterial cellular surface proteins [38], harden soft carious lesions through reaction with phosphate or chloride ions resulting in the formation of silver salts (e.g., silver chloride), and increase the alkalinity of the environment through the formation of ammonium compounds hypothesized to have an acid-buffering effect [39]. However, the potential indirect effects of application on adjacent caries activity has not been meaningfully examined. SDF application in CariedAway was restricted to posterior dentition to mitigate any negative impact that SDF staining might have on facial aesthetics and oral health-related quality of life. We subsequently explored whether posterior application controlled anterior lesions and if there was any dose-response relationship. Our data do not indicate any such association and we observed no correlation between the number of posterior surfaces treated and any subsequent anterior arrest.
There are a number of study limitations in addition to those already discussed. Due to the gaps in care due to COVID-19, exfoliation of primary dentition in the intervening period may have occurred. As a result, follow-up for affected tooth surfaces would stop if recorded as missing, and then resume for newly erupted permanent dentition if later diagnosed to be carious. Additionally, it is possible that if a study participant were to visit an outside dental provider, the examining dentist may put a permanent restoration in place of any dentition previously treated with SDF or ART. While we consider this to be unlikely given the target population enrolled, any observation of a filling on a surface previously treated with either SDF or ART would not be recorded as decayed. This may have inflated the success percentage of treatments. Finally, the provider (nurse and hygienist) assignment in the SDF arm was not randomized. In implementing the CariedAway study, participants were seen on a first come, first served basis. There were no selection criteria to determine whether a patient was treated by a hygienist or registered nurse. Therefore, we expect this bias to be low.
5 Conclusions
Both silver diamine fluoride and atraumatic restorative treatment are effective in the longitudinal control of dental caries when used in school-based prevention. Incorporating these and other minimally invasive techniques can help improve access to critical preventive oral medicine and decrease health inequities.
Funding Statement
Research presented in this report was funded by the Patient-Centered Outcomes Research Institute (#PCS-160936724).
Competing interests
The authors have no competing interests to declare.
Footnotes
Author roles: RRR: conceptualization, formal analysis, methodology, funding acquisition, supervision, writing-original draft; AG: formal analysis; QX: formal analysis; TJBG: data curation, investigation.