Changes in Caries Experience, Untreated Caries, Sealant Prevalence, and Preventive Behavior Among Third-Graders in New York State, 2002–2004 and 2009–2012

Abstract

Objective

This study assessed changes in caries experience, untreated caries, sealant prevalence, and preventive behavior among third-grade children in New York State to monitor progress toward state health objectives.

Methods

We analyzed children's data from the 2002–2004 (n=10,865) and 2009–2012 (n=6,758) New York State Oral Health Survey. We calculated differences in weighted percentages and 95% confidence intervals for caries experience, untreated caries, sealant prevalence, and preventive behavior. We used logistic regression procedures to assess the independent effects and interaction terms on dental caries experience.

Results

The percentage of children with dental caries and untreated caries decreased from 54.1% and 33.0% in 2002–2004 to 45.2% and 23.6% in 2009–2012, respectively. While this decrease was not uniform across income subgroups, the prevalence of sealants, a key measure of the use of preventive services, increased significantly from 16.7% to 36.0% among lower-income children.

Conclusions

Measurable improvement in reducing dental caries prevalence among third-grade children has been made in New York State, but this improvement was not uniform across subgroups. Specifically, disease prevalence among lower-income children remained high, underscoring the need to strengthen existing programs and identify additional policy and programmatic interventions.

Researchers generally agree that the prevalence and severity of dental caries among U.S. and New York State (NYS) school-age children declined steadily from the 1970s to the 1990s. Although this trend has continued for older children in more recent years, this trend is uncertain among younger children aged 2–8 years.^1–3 Findings from analyses of 1988–1994 and 1999–2004 national surveys show that declines in dental caries observed in earlier decades among younger children may have plateaued or dental caries may even be increasing among subgroups of younger children.⁴ Because of the persistent higher disease rate, especially in low-income groups, prevention of tooth decay among children has become the focus of many prevention efforts.^5,6 Since 2001, the Centers for Disease Control and Prevention (CDC) and the Health Resources and Services Administration have provided grants and technical assistance to NYS to strengthen the infrastructure and capacity to promote fluoridation and improve its quality, as well as to strengthen school-based preventive and early treatment programs.

Changes have also been made to increase insurance coverage for dental services and improve annual dental visits.^7,8 Child Health Plus, the state Children's Health Insurance Program (CHIP), was implemented in 1997 to provide public health insurance for near-poor children from families previously not eligible for Medicaid. According to a U.S. Government Accounting Office report, nationally, Medicaid and CHIP beneficiaries, children in particular, showed increases in the use of dental services (from 28% in 1996 to 37% in 2010), but still visited the dentist less frequently than privately insured children (58% in 2010).⁹ In NYS, the Medicaid program enhanced the fee structure for dental procedures in 2000. In addition, professional organizations, advocacy groups, and foundations have made a concerted effort to promote prevention and access to care. The professional recommendation to initiate first dental visit shifted from age 3 to age 1 around 2003.¹⁰ To assess the collective effect of these and other efforts in NYS, we examined data on caries experience, untreated caries, sealant prevalence, and preventive behavior among third-grade children from the 2002–2004 and 2009–2012 NYS Oral Health Survey.

METHODS

We analyzed data from two independent samples derived from oral health surveys of third-grade children conducted in 2002–2004 (10,865 children from 359 schools) and 2009–2012 (6,758 children from 250 schools). The state-based third-grade oral health survey is a component of the National Oral Health Surveillance System developed as a result of the collaboration between CDC and the Association of State and Territorial Dental Directors (ASTDD).¹¹ The oral health indicators were approved by the Council of State and Territorial Epidemiologists for public health surveillance. ASTDD's Basic Screening Survey (BSS) tool provides states with a low-cost, simple, and consistent model for monitoring oral diseases that also supports comparisons within and among states. In addition, the guidance documents include methods for sampling, implementation packets, and methods for analyzing data.¹²

These oral health surveys are based on ASTDD BSS guidance documents and employed the concept of a stratified random sample of clusters to select the sample. The NYS Education Department Profiles of Schools enrollment records were used to construct the sampling frame. Stratification was based on county's and school's two levels of economic status, as determined by the percentage of children in the free and reduced-cost school-lunch program. For each survey, all third-grade children who consented to participate from randomly selected schools were screened for dental caries, untreated caries, and sealants if children were present on the day of the survey. Data reported by a parent or guardian regarding dental insurance coverage, last dental visit, and fluoride tablet use were gathered via a questionnaire.

The questionnaires in both surveys included identically worded questions on the child's school lunch participation. Participation of the reported child in the free or reduced-cost school lunch program was used to categorize the child as lower income; all other children were considered higher income. In NYS, children whose income is below 160% of the federal poverty level (FPL) are eligible for free coverage under Medicaid or CHIP. Eligibility for the free or reduced-cost school lunch program is 185% FPL or an annual income of #$44,123 for a family of four. Trained dental hygienists gathered the data on caries experience, untreated caries, and the presence of dental sealants in accordance with the ASTDD survey protocol. The weight assigned to each child was based on the inverse of the probability of school selection within a stratum and the response rate. Further details regarding the survey method are reported elsewhere.¹³

The two surveys had two methodological differences. First, the number of schools and children selected in the 2009–2012 survey was smaller. Second, the 2002–2004 survey included anthropometric measurements along with a detailed food frequency questionnaire. The two samples did not, however, differ with respect to participation in the free and reduced-cost school lunch program, an important determinant of the outcomes examined in this analysis. The percentages of children who reported participating in the free and reduced-cost school lunch program in the 2002–2004 and 2009–2012 surveys were 44% and 47%, respectively. The response rates for the 2002–2004 and 2009–2012 surveys were 37.7% and 35.5%, respectively.

We calculated differences in weighted percentages and 95% confidence intervals (CIs) for caries experience, untreated caries, sealant on at least one molar tooth, dental insurance coverage, ever using fluoride tablets, and last dental visit within one year. We conducted all descriptive analyses using SAS^® SURVEYFREQ version 9.3.¹⁴ To assess the independent effect of year (2009–2012 vs. 2002–2004), region (New York City [NYC] vs. rest of NYS), and economic status of the child (higher income vs. lower income), we used the logistic regression procedure with weighting in the first model. To determine if the effects observed in the main effect model were dependent on other variables, we included interaction terms in a second model; significance was set using alpha = 0.1 for the interaction effect. Fluoride tablet use and last dental visit variables were not included in the model because of their strong association with region (odds ratio [OR] 5 11.1) and dental insurance coverage (OR52.4).

RESULTS

Among all children, the prevalence of dental caries decreased from 54.1% in 2002–2004 to 45.2% in 2009–2012 (28.9%, 95% CI 215.7, 22.1). We also observed a decrease in the prevalence of untreated caries among all children (29.5%, 95% CI 214.2, 24.8). The percentage of children with dental sealants increased from 26.8% in 2002–2004 to 39.5% in 2009–2012 (12.7%, 95% CI 8.8, 16.5). Insurance coverage, annual dental visit rates, and fluoride tablet use also increased (Table 1).

Table 1.

Changes in the prevalence of dental caries, untreated caries and dental sealants, and dental insurance coverage, by last dental visit within one year and fluoride tablet use, among third-grade children in New York State: 2002–2004 and 2009–2012 New York State Oral Health Survey

^aLower income refers to those children enrolled in the free and reduced-cost school lunch program. The number of lower-income children in the 2002–2004 survey was 4,820 (44%) and in the 2009–2012 survey was 3,147 (47%).

CI = confidence interval

In both surveys, the prevalence of dental caries and untreated caries was highest among lower-income children. Except for dental insurance coverage, lower-income children had fewer dental visits and a lower use of dental sealants and fluoride tablets than higher-income children (Table 1).

Similar changes in dental caries, untreated caries, and sealant prevalence were seen among higher- and lower-income children in NYC and the rest of NYS, but these changes were not uniform across subgroups. Lower-income children in NYC had lower caries experience compared with children in the rest of NYS (44.4% vs. 61.9%). An increase in the use of sealants was particularly evident among lower-income groups in NYC and the rest of NYS (14.1% and 11.8%, respectively), but not in higher-income children in the rest of NYS (data not shown).

The logistic regression model confirmed the effect of year on dental caries experience, which was the primary outcome of interest. The year, income, and region variables were significant in the main effect model. The effect of year on caries experience was not uniform across subgroups, however, as evidenced by the interaction of effect in the second model, such that the lower-income group in the rest of NYS did not gain as much as the other groups (Table 2).

Table 2.

Logistic regression analysis of the effect of selected variables on caries experience among third-grade children, New York State Oral Health Survey, 2002–2004 and 2009–2012^a,b

^aModel 1–2 log L = 614067.75; Wald chi-square = 290.405; DF = 5; p<0.001

^bModel 2–2 log L = 611275.08; Wald chi-square = 251.731; DF = 7; p<0.001

SE = standard error

DF = degree of freedom

DISCUSSION

Our findings showed a decline in the prevalence of dental caries and untreated caries among third-grade children in NYS. While more recent national data on changes in dental caries prevalence are not yet available, the improvements observed are largely consistent with national trends,¹⁵ with notable exceptions. For example, in the national surveys, improvement with respect to a key preventive measure (i.e., dental sealant prevalence among poor 6- to 9-year-old children in 2009–2010) was 25.5% when compared with that of the 1999–2004 National Health and Nutrition Examination Survey estimate of 21.0% (95% CI 10.6, 31.4) among 6- to 8-year-old poor children.^16,17 In NYS, a similar comparison with respect to dental sealant prevalence between the two surveys we examined showed a greater improvement among lower-income children (16.7%, 95% CI 15.0, 18.5 in 2002–2004 vs. 36.0%, 95% CI 29.6, 42.3 in 2009–2012) compared with the previous survey. Because the prevalence of caries declined primarily in the higher-income group, the disparity appears greater now. This widening in disparity with respect to caries prevalence argues for the need to understand the causes of the disparity and find effective solutions.

Through a cooperative agreement with CDC, the results of the initial survey were provided to local health departments, rural health networks, and other stakeholders for local action beginning in 2004. These data were used to develop the State Oral Health Plan, which was widely disseminated through multiple channels. The State Oral Health Plan promoted evidence-based interventions and identified multiple strategies for promoting best practices. In NYS, the fluoridation program reaches 100% of children in NYC and only about 48% of children in the rest of NYS. School-based dental programs are present in about 1,000 targeted schools in underserved areas.¹⁸ These efforts are f-urther buttressed by the strategies to increase insurance coverage and annual dental visits through Medicaid and Child Health Plus insurance programs. The result of these initiatives may have played a large role in the observed improvements. In particular, the increase in the prevalence of dental sealants among lower-income children should lead to a reduction in the prevalence of caries when they become adolescents. This increase in the prevalence of dental sealants, when combined with the decrease in untreated caries, suggests that policies and programs have been effective. For example, the decline in the prevalence of caries was more evident in lower-income children in fluoridated NYC. The prevalence of sealants was also higher among children in schools with a dental sealant program (data not shown). Insurance coverage, annual dental visits, and the use of fluoride tablets all showed improvements over time.

Dye at al. postulated two traditional risk factors, unhealthy eating habits and inadequate dental insurance coverage, that might have played a role in the lack of progress among young children during the 1988–1994 and 1999–2004 periods.⁴ An increase in the consumption of sugar-sweetened beverages (e.g., juice drinks, fruit juices, flavored juices, and sodas) by children and its link to dental caries has received considerable attention.¹⁹ Results from the 2009–2010 NYS Behavioral Risk Factor Surveillance System showed that 25% of children aged 2–5 years and 31% of children aged 6–11 years consumed at least one regular soda or other sugar-sweetened beverage per day.²⁰ During the periods of this survey in NYC, several policies including improved nutrition, provision of extensive nutrition education training to on-site nutrition education workers, and substantial improvements in school cafeteria food were implemented. There was a decrease from the 2006–2007 to the 2010–2011 school year in the prevalence of obesity among public schoolchildren in kindergarten through eighth grades.²¹ Therefore, it is interesting to speculate if the lower dental caries experienced by lower-income children in NYC when compared with that of children in the rest of NYS could be explained in part by their exposure to fluoridated water, lower consumption of sugar-sweetened beverages, or effect modification of fluoridation. Armfield et al. found that fluoridation modified the effect of sugar-sweetened beverages on dental caries.²² Because fluoridated water is also used to process beverages, children are indirectly exposed to fluoride even when they are consuming beverages. Therefore, expanding fluoridation in the rest of NYS may reduce the geographic variation in caries prevalence.

In this survey, both insurance coverage (except in lower-income children) and annual dental visit rates improved. The annual dental visit rate among 1- to 18-year-old children enrolled in the NYS Medicaid program increased from 27.3% in 2000 to 40.9% in 2012.^23,24 Nationally, Wall and colleagues reported a steady increase in the use of dental services from 1997 to 2010 among all children.²⁵ This increase in the use of dental services in children appears to have coincided with a drop in the percentage of children who were uninsured, from 15.1% to 10.2%, and an increase in the percentage of children enrolled in Medicaid and CHIP, from 16.1% to 32.2%. One possible reason for the lack of improvement in insurance coverage among lower-income children in NYS is that the coverage for dental services was already available even before implementation of the federal CHIP.

The findings regarding untreated caries in this oral health survey of third-grade children mostly reflect the missed opportunities for preventing caries in primary teeth. It is important to focus on this population because caries in primary teeth is a good predictor of future caries risk. In NYS, a major policy intervention to allow reimbursement for primary care providers to provide screening, counseling, and fluoride varnish for children younger than 7 years of age was implemented in 2009; however, its effect is too early to detect in this survey. Similarly, the effect of the school dental certificate law that launched in 2009 is also not known.

Despite this progress, the prevalence of both caries and untreated caries among lower-income children remains high, as evidenced by the lack of progress in achieving the Healthy People 2010 oral health objectives for 6- to 8-year-old children (target of #42.0% for caries experience and #21% for untreated decay).²⁶ Furthermore, this higher prevalence of caries and untreated caries underscores the need to strengthen existing programs and explore additional policy and programmatic interventions. To achieve further improvements, the NYS Department of Health has prioritized the prevention of tooth decay in its state health improvement plan, Prevention Agenda 2013. The Prevention Agenda is a call to action to promote evidence-based interventions, with a particular focus on social determinants, environmental and policy changes, and long-lasting clinical interventions.²⁷

Limitations

This study was subject to several limitations. First, measuring the precise magnitude of the changes using cross-sectional surveys conducted over time required some caution for several reasons, such as methodological differences, demographic changes, and shifts in diagnostic criteria. Second, these data were collected for surveillance purposes, and the nature of the approach did not allow for detailed collection of sociodemographic information that may lend itself to hypothesis testing. Third, these findings were based on caries prevalence but not the more important caries severity, which is needed to determine the effectiveness of interventions. Previous surveys, however, have shown that a decline in caries prevalence is accompanied by a concomitant decline in the severity of caries.³ Fourth, the CIs for the 2009–2012 survey were slightly higher than the 2002–2004 survey, and the sample sizes for several subgroups were smaller; as such, the results should be interpreted with caution, especially for subgroups such as NYC children.

Fifth, parental reports of annual dental visits, fluoride tablet use, and dental insurance coverage may have been subject to inaccurate reporting. For example, while the 2012 Centers for Medicare & Medicaid Services 416 Early and Periodic Screening, Diagnostic and Treatment Participation Report shows that only 51.2% of 6- to 9-year-old children had at least one dental visit, the self-reported annual dental visit in our survey for lower-income children was much higher (74.3%).²⁴ Macek et al. have expounded the reasons for the differences among national dental visit estimates and explained why differences in utilization estimates might have occurred in three large national surveys. According to the authors, while such self-reported estimates may not provide true utilization, an assessment of trends should still be valid.²⁸ Lastly, generalizing the results of this survey to all third-grade children requires the assumption that the respondents who provided data are a representative sample of the population. Although the response rate was <38% in both surveys, it was similar to other state surveys such as the Behavioral Risk Factor Surveillance System. Furthermore, the participation of children by school free and reduced-cost lunch status shows that it was largely consistent with that proportion in the population. While such low response rates may lead to biased estimates, our assessment of the trend should still be valid.

CONCLUSIONS

During the last decade, NYS used federal grants to expand the infrastructure and capacity, develop a blueprint for action, and mobilize partners to promote best practices to improve oral health. The initiatives included improvements in fluoridation and school-based sealant programs, increases in dental insurance coverage due to expansion of Medicaid income eligibility, development and dissemination of guidelines to promote oral health during pregnancy and early childhood, expansion of the primary care loan repayment program to include dental professionals, and, more recently, collaborative dental hygiene practice. Nationally, Mandal et al. reported significant improvements in only four of 15 oral health policies from April 2002 to September 2008.²⁹ While it is difficult to determine the contributions of these policies and programs toward reducing dental caries prevalence among third-grade children in NYS, this study showed significant improvements in disease prevalence and use of preventive services. This improvement, however, was not uniform across all subgroups. Specifically, disease prevalence among lower-income children remained high, underscoring the need to identify policy and programmatic interventions that are likely to eliminate disparities.