Medicaid Payment and Fluoride Varnish Application during Pediatric Medical Visits

Abstract

All Medicaid programs pay for fluoride varnish applications during medical visits for infants and toddlers, but receipt of care varies considerably across states. Using 2006-2014 Medicaid data from 22 states, this study examined the association between Medicaid payment and receipt of fluoride varnish during pediatric medical visits. Among 3,393,638 medical visits, fewer than one-in-ten visits included fluoride varnish. Higher Medicaid payment was positively associated with receipt of fluoride varnish during pediatric medical visits. As policymakers consider strategies for increasing young children’s access to preventive oral health services, as well as consider strategies for balancing budgets, attention should be paid to the effects of provider payment on access to pediatric oral health services.

Introduction

Access to dental care is a major public health issue for children from low-income families. Fewer than half of publicly insured children receive an annual dental visit (Medicaid and CHIP Payment and Accesses Commission (MACPAC), 2016), and dental caries, commonly called tooth decay, is concentrated among children living in poverty (Centers for Disease Control and Prevention, 2019). During 2011-2016, two-times as many children aged 2–5 years living in poverty had dental caries than children living in households at or above 200% of federal poverty guidelines (33.9% vs. 15.7%) (Centers for Disease Control and Prevention, 2019). Untreated caries is associated with substantial consequences, including pain, decreased quality of life (Abanto et al., 2011), missed days of school and poor school performance (Jackson, Vann, Kotch, Pahel, & Lee, 2011), and, in extreme cases, death (Casamassimo, Thikkurissy, Edelstein, & Maiorini, 2009).

Application of fluoride varnish is one intervention that can reduce dental caries among young children. Systematic reviews indicate fluoride varnish applications reduce caries in primary teeth, with evidence supporting two-to-four applications annually (Chou et al., 2021; Chou, Cantor, Zakher, Mitchell, & Pappas, 2014; Marinho, Higgins, Logan, & Sheiham, 2002; Marinho, Worthington, Walsh, & Clarkson, 2013). The US Preventive Services Task Force recommends non-dental providers apply fluoride varnish to the teeth of children all children younger than six years of age during medical visits (Moyer, 2014; US Preventive Services Task Force, 2021). The rationale for this service to be delivered in medical settings is that young children are more likely to visit medical offices than dental offices. The 2019 National Survey of Children’s Health reported that 91% of children <5 years of age received medical care in the past year, compared to fewer than two-thirds of children who received dental care (Child and Adolescent Health Measurement Initiative, 2019). Furthermore, young children are recommended to have frequent well-child medical visits, including at 9, 12, 15, 18, 23 months and 3, 4, 5 years of age, and guidelines from the American Academy of Pediatrics and Bright Futures recommend fluoride varnish applications every 3-6 months coinciding with these periodic well-child medical visits (Bright Futures & American Academy of Pediatrics, 2021). Promoting delivery of fluoride varnish in pediatric medical settings is important because studies suggest that young children’s receipt of multiple applications of fluoride varnish during medical visits is associated with fewer dental caries (Kranz, Preisser, & Rozier, 2015) and a lower likelihood of hospitalization for dental treatment (Stearns, Rozier, Kranz, Pahel, & Quinonez, 2012).

Despite the documented benefits associated with of receipt of fluoride varnish during medical visits, few children receive it. Multi-state studies suggest that fewer than 5% of well-child visits paid by private insurers include fluoride varnish applications (Geissler, Dick, Goff, Whaley, & Kranz, 2021), and fewer than 11% of Medicaid-enrolled children typically receive fluoride varnish during medical visits (Arthur & Rozier, 2016; Geiger et al., 2019; Goldstein, Dick, Ross, Stein, & Kranz, 2021). Although all state Medicaid programs have paid for fluoride varnish applications in medical settings since 2017 and earlier, states vary considerably in the number of children receiving it. A 50-state study using data from 2010-2013 found that fewer than two percent of eligible Medicaid-enrollees received fluoride varnish or another type of oral health service in a non-dental setting in 16 states, while only three states reported 20% or more of eligible Medicaid-enrollees receiving it (Arthur & Rozier, 2016). Two of the three state Medicaid programs with the highest rates of children receiving fluoride varnish during medical visits (North Carolina and Washington) pay more than the national average for this service, suggesting that higher Medicaid payments to medical providers for applying fluoride varnish may encourage delivery.

New Contribution

Using Medicaid data from 22 states during 2006-2014, we exploited variation in the amounts that state Medicaid agencies pay for fluoride varnish during medical visits to test the hypothesis that higher payment was associated with increased likelihood of young children receiving fluoride varnish during well-child medical visits. Because states enacted policies allowing for the reimbursement of fluoride varnish application in medical settings at different times, we examined the association between payment and receipt of fluoride varnish conditional on length of policy enactment. Controlling for policy duration is important, as prior studies have shown that these policies take time to impact children’s receipt of care (Goldstein et al., 2021; Kranz et al., 2020; Rozier, Stearns, Pahel, Quinonez, & Park, 2010). This study makes several novel contributions. First, we examine a large and geographically diverse set of states. This is novel as most studies examining children’s receipt of fluoride varnish during medical visits have focused on one or two states (Herndon, Tomar, Catalanotto, Vogel, & Shenkman, 2015; Okunseri, Szabo, Jackson, Pajewski, & Garcia, 2009; Rozier et al., 2010), meaning these study findings may not be generalizable to other states. Second, while a few recent studies have examined more states (Geiger et al., 2019; Goldstein et al., 2021; Kranz et al., 2020), these studies have not explored Medicaid payment nor examined if the different amounts paid by state Medicaid agencies contribute to the different state-level rates of children receiving fluoride varnish during medical visits. One study did examine the association between Medicaid payment and receipt of preventive oral health services in medical settings, finding non-significantly higher rates among states paying more during 2010-2013 (Arthur & Rozier, 2016). This study, however, used aggregate state-level data reported to the Centers for Medicare and Medicaid Services (CMS), which mask important individual-level variation and underreport utilization (Schneider, Rossetti & Crall, 2007). Our use of individual-level data, collected for billing purposes, allows us to conduct a more rigorous analysis to examine the impact of payment on care received. Finally, our study adds to the literature examining the impact of Medicaid payment on receipt of pediatric medical care, which has so far mostly focused on adults (Saulsberry, Seo, & Fung, 2019).

Conceptual Framework

Medicaid payments to clinicians may incentivize the provision of care. Studies consistently show a positive relationship between Medicaid payment and children’s receipt of care from dentists (Beazoglou, Douglass, Myne-Joslin, Baker, & Bailit, 2015; Chalmers & Compton, 2017; Decker, 2011; Fisher-Owens, Amendola, Featherstone, Inge, & Flaherman, 2017; Nasseh & Vujicic, 2015). However, less is known about the relationship between Medicaid payment and children’s receipt of medical care. A 2019 systematic review examining changes in Medicaid payment reported mixed results on utilization, although most studies focused on adults (Saulsberry et al., 2019). A study examining the effect of increased Medicaid payments in two states reported increased well-child visits in Idaho, but not Kentucky (Kenney, Marton, Klein, Pelletier, & Talbert, 2011). Additionally, a study of eight states reported that higher Medicaid payments for vaccines were associated with higher rates of vaccination among young children (Tsai, 2018).

Higher Medicaid payments may help to offset costs in medical settings associated with fluoride varnish applications, such as education, supplies, and clinician time (Quinonez et al., 2014; Silk, Sachs Leicher, Alvarado, Cote, & Cote, 2018). However, it is unclear to what extent higher payments are sufficient to change behavior, given pediatricians report numerous non-financial barriers to delivery, including lack of awareness, insufficient training, and logistical challenges (Isong et al., 2011; O'Callaghan & Douglass, 2013; Quinonez et al., 2014). While Arthur and Rozier found non-significantly higher rates of fluoride varnish and additional preventive oral health services during medical visits in states with higher Medicaid payments (Arthur & Rozier, 2016), the limitations of this study’s data and analysis suggest this relationship is still unclear. Therefore, it is unknown if Medicaid payment amounts may increase, decrease, or not impact the rate of fluoride varnish applications in pediatric medical settings.

Methods

Data and Sample.

Our analysis used repeated cross-sections of annual Medicaid claims data to estimate the association between payment and receipt of fluoride varnish during medical visits. We obtained Medicaid Analytic eXtract files for 45 states based on availability of data files (Research Data Assistance Center, 2019) and high-quality managed care encounter data (Byrd & Dodd, 2012). The dataset included all children <6 years who received fluoride varnish or other preventive oral health services during medical visits and a stratified random sample of Medicaid enrolled children <6 years who never received preventive oral health services during medical visits. We weighted our sample to be representative of all Medicaid enrolled children <6 years in these states and years. We dropped seven states due to our inability to confidently identify fluoride varnish or other preventive oral health services during medical visits, 13 states that did not pay for fluoride varnish during medical visits during our entire study period, and three states for which payment levels could not be confirmed, resulting in 22 states during 2006-2014. Not all states provided data for all years, with each state providing data for an average of five years. Data used for this study were obtained under a data use agreement from the Research Data Assistance Center (ResDAC), a contractor for CMS, and are not publicly available.

The unit of analysis for this study was well-child medical visits for children aged six months to five years of age (weighted N=28,593,936). We examined well-child medical visits during these ages because guidelines from Bright Futures and the American Academy of Pediatrics explicitly recommend fluoride varnish applications during visits at these ages (Bright Futures & American Academy of Pediatrics, 2021). Well-child visits were identified using current procedural terminology (CPT) codes for a new or established visit (99381, 99382, 99383, 99391, 99392, 99393). We excluded managed care encounters because payment could not be determined in our dataset (weighted N=15,908,810, 55% of total visits). A few visits were dropped for missing information on sex (dropped <0.5% of visits), county (dropped <0.5% of visits), and race (dropped 10.6% of visits), and if they occurred the year a policy was enacted but before the enactment date (dropped 2.5% of visits), resulting in a final analytic sample of 3,393,638 unweighted well-child medical visits (weighted N=10,602,564).

Variables.

The primary dependent variable was a dichotomous indicator that a well-child medical visit included a fluoride varnish application. State-specific CPT and current dental terminology (CDT) procedure codes used to identify fluoride varnish applications during medical visits were obtained from the American Academy of Pediatrics Section on Oral Health (2020). Because some states pay for additional preventive oral health services during medical visits, namely risk assessments, oral evaluations, and parental counseling (Arthur & Rozier, 2016), we also constructed a second dependent variable identifying well-child medical visits with any type of preventive oral health service. These CPT and CDT codes were also obtained from the American Academy of Pediatrics.

State Medicaid programs enacted policies at different times allowing for the payment of fluoride varnish during medical visits. We obtained information about the dates of policy enactment from the American Academy of Pediatrics, supplemented when necessary, with information from the literature. Recognizing that policies often take time to influence utilization (Arthur & Rozier, 2016; Geiger CK et al., 2019), we categorized time since policy enactment (<1 year, 1 year, 2 years, 3 years, 4 years, and 5 or more years).

The key explanatory variable, Medicaid payment for fluoride varnish applications during medical visits, was obtained from the fee-for-service (FFS) claims data for applications occurring on the same day as the well-child medical visit. Because some states pay for additional services intended to be provided as a package with fluoride varnish (e.g., anticipatory guidance and risk assessment), we also examined the association between payment for fluoride varnish plus additional preventive oral health services. We constructed a measure of Medicaid payment for all preventive oral health services during medical visits by summing the payment amounts for all of these types of services occurring on the same day as the well-child medical visit.

We constructed child-level control variables: sex, age, race/ethnicity, and number of months enrolled in Medicaid in a year. County-level characteristics were obtained from the Area Health Resource Files and merged to claims by county and year. These characteristics included continuous measures of unemployment and child poverty rates and indicators of whether the county was a full or partial primary care or dental health professional shortage area (HPSA). County rurality was defined according to the 2013 Rural-Urban Continuum Codes, and categorized as metropolitan, non-metropolitan adjacent to metropolitan areas, and nonmetropolitan not adjacent to metropolitan areas; consistent with prior research (Geiger et al., 2019).

Analytic approach.

The analysis exploited variation in Medicaid payment levels for fluoride varnish across states to estimate the association between payment amount and the likelihood that a child received fluoride varnish during a well-child medical visit. We estimated a multivariable logistic regression model in which each observation corresponded to a well-child medical visit and the likelihood a child received fluoride varnish depended on child-, county-, and state-level characteristics and year fixed effects. Building on the approach described by Borusyak, Jaravel, and Spiess (2021), we estimated a model of the form:

$$Y_{ist}=f({\alpha }_t+\beta X_{ist}+\sum _{j=1}^5{\gamma }_j\ast Pos{t}_{j_{st}}+\tau Rat{e}_{st)}$$

where f is the logit function, Post_jst equals one if the policy to reimburse fluoride varnish during medical visits was in place for j years in state s in calendar year t. Rate_st is the Medicaid payment amount for fluoride varnish in state s in year t, α_t are calendar year fixed effects, and X_ist are the state-level fixed effects and the child- and county-level characteristics that were described in the text above.

This approach compares receipt of fluoride varnish in states in each year after the policy was enacted with receipt of fluoride varnish in states with policies enacted for less time and different reimbursement rates, allowing the size of the effect of the policy to vary with the number of years since the policy was enacted and the reimbursement rate. The yearly effects are measured relative to receipt of fluoride varnish in the 11 months of policy enactment. This strategy allows us to use visits in states with newly enacted policies as a control group for visits in states with longer enacted policies, an approach with controls for factors that may be related to the timing of the policy enacted. We do not include a “pre” period in our models because the probability of a visit including fluoride varnish is always zero before a state Medicaid program enacts a policy allowing reimbursement for FV applications during medical visits. Thus, the “pre” period is not informative in this model, nor are states that do not pay non-dental providers to apply fluoride varnish.

We accounted for the sampling scheme by weighting each observation using sampling weights and estimated the standard error of the estimates using empirical Huber-White sandwich estimator to cluster at the state level (Huber, 1967; White, 1980). This clustering allowed for correlation in the outcomes of children who live in the same state, both within-years and across-years, and also accounted for the fact that a child may be observed multiple times in the sample (i.e., children may have multiple visits within a year and over time). We used model estimates to calculate adjusted predicted probabilities of the outcomes by computing the mean of predicted probabilities for the entire sample after setting the payment rate first to the 25^th percentile, then to the 50^th percentile, and then to the 75^th percentile, while keeping all other covariates at their observed values, and varying the years since policy enactment (<1 year, 1 to <2 years, and so on). Payment percentiles were calculated at the state-level.

We conducted three additional analyses to assess the robustness and validity of our results. Building on the above model, we estimated an additional regression model that adjusted for the total payment amount for fluoride varnish plus any additional preventive oral health services paid by the state and examined the odds of receipt of preventive oral health services during well-child visits. Next, to address concerns that the different rates of enrollment in managed care across state Medicaid programs may impact our findings, we estimated a model adjusting for the percentage of our sample enrolled in Medicaid managed care plans in each state and year. Finally, to allay concerns that our findings may not be relevant to managed care enrollees, we estimated a model including all visits, paid via FFS and managed care, and assigned the managed care amount paid for preventive oral health services to be the mean FFS payment for preventive oral health services in the state and year the visit occurred. All analyses were conducted using STATA MP version 15.1 (College Station, TX) and used a significance level of 0.05. The corresponding author’s Institutional Review Board approved this study.

Results

Among an unweighted sample of 3,393,638 well-child medical visits, 9.4% of visits included fluoride varnish (Table 1). Among visits with fluoride varnish, 60.8% occurred in states that had a policy enacted for five or more years and 51.6% were visits for one-year old children. Hispanic children received 31.1% of all well-child visits and received a higher share of visits with fluoride varnish (36.1%). White children received 40.5% of all well-child visits and received a lower share of visits with fluoride varnish (34.3%). Of visits without fluoride varnish, 86.8% occurred in counties designed as primary care HPSAs.

Table 1.

Characteristics of well-child medical visits for Medicaid-enrolled children <6 years in 22 states (2006-2014)

	All visits	Visits with fluoride varnish	Visits without fluoride varnish
Utilization Characteristics
Visits with fluoride varnish, %	9.4	100	0
State Policy Characteristics
Visits, by time since FV policy enactment, %
<1 year	15.7	3.5	17.0
1 year	17.5	9.4	18.3
2 years	16.5	11.5	17.0
3 years	12.9	10.2	13.2
4 years	5.8	4.6	5.9
5 or more years	31.6	60.8	28.6
Child Characteristics
Race/ethnicity, %
White, non-Hispanic	40.5	34.3	41.1
Black, non-Hispanic	25.0	26.4	24.8
Hispanic	31.1	36.1	30.6
Other race/ethnicity groups	3.5	3.2	3.5
Female, %	48.8	48.6	48.9
Months enrolled in Medicaid during calendar year, mean (standard deviation)	11.0 (2.1)	11.3 (1.7)	11.0 (2.1)
Age in years, %
<1 year	22.2	20.4	22.4
1 year	31.2	51.6	29
2 years	13.5	22.2	12.6
3 years	11.2	4.9	11.8
4 years	12.4	0.7	13.6
5 years	9.6	0.3	10.5
County Characteristics
Rurality, %
Metropolitan (metro)	72.6	77.9	72
Non-metro adjacent to metro	17.9	17.2	17.9
Non-metro not adjacent to metro	9.6	4.9	10.1
Poverty rate, mean (standard deviation)	26.1 (9.4)	27.7 (9.7)	25.9 (9.3)
Unemployment rate, mean (standard deviation)	8.5 (2.7)	9.4 (2.8)	8.5 (2.7)
Lived in full/partial primary care HPSA, %	85.8	76.3	86.8
Lived in full/partial dental HPSA, %	84.9	84.0	85.0
Count of total visits (unweighted)	3,393,638	942,207	2,451,431

Note. Table describes characteristics of well-child medical visits paid for fee-for-service and received by children aged 6 months to 5 years in 22 states during 2006-2014. Abbreviation: health professional shortage area, HPSA.

During this study period, Medicaid payments for fluoride varnish during medical visits ranged from $8 to $35, with a median reimbursement rate of $17 (25^th percentile=$14, 75^th percentile=$23). There was little change in payment within states, as payment between the first and last observed year varied by 0 to $2 for nearly all states—only Georgia experienced a larger (-$5) change. States that paid for additional preventive oral health services had a higher median reimbursement rate of $26 (25^th percentile=$15, 75^th percentile=$38).

Following regression adjustment (Table 2), we found that higher Medicaid payment for fluoride varnish was positively associated with children’s receipt of fluoride varnish during medical visits. In our main analysis, we found that doubling the payment for fluoride varnish was associated with 8.6 times greater odds of a medical well-child visit including fluoride varnish (95% confidence interval (CI)=3.2 to 23.2; P<0.001). A 10% increase in the payment for fluoride varnish was associated with 0.9 times greater odds of a medical well-child visit including fluoride varnish. Figure 1 illustrates the relationship between payment-level for fluoride varnish, time since policy enactment, and adjusted predicted probabilities of receipt of fluoride varnish during well-child medical visits. Payment at the 75^th percentile was significantly associated with higher predicted probabilities of receiving fluoride varnish during visits compared to payment at the 25^th and 50^th percentiles after 1 or more years of policy enactment (P<0.001). For example, increasing payment for fluoride varnish from the 25^th percentile to the 75^th percentile (from $14 to $23) in the fourth year after policy enactment, was associated with a 6.7 percentage point increase in the predicted probability that a child received fluoride varnish during a medical visit (P<0.001); there were only an expected four fluoride varnish applications per 100 well-child visits when fluoride varnish is paid at the 25^th percentile ($14), which increased to 11 fluoride varnish applications per 100 well-child visits when fluoride varnish is paid at the 75^th percentile ($23). Similarly, payment at the 50^th percentile was significantly associated with higher predicted probabilities of receiving fluoride varnish during visits compared to payment at the 25^th percentile after 1 or more years of policy enactment (P<0.001).

Table 2.

Odds of receipt of fluoride varnish during well-child medical visits for Medicaid-enrolled children <6 years in 22 states (2006-2014)

	Main Analysis: Receipt of fluoride varnish (FV) during visit	Sensitivity 1: Receipt of any preventive oral health service during visit	Sensitivity 2: Receipt of FV, with adjustment for managed care penetration in state/year	Sensitivity 3: Receipt of FV, with inclusion of visits paid via managed care
	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]
Log payment for fluoride varnish	8.61*** [3.20, 23.18]	5.25*** [2.24, 12.29]†	9.07*** [3.54, 23.22]	19.49*** [6.99, 54.37]
Percentage of sample enrolled in managed care in state/year			0.50* [0.26, 0.95]
Visit paid via fee-for-service (Reference group: visit paid via managed care)				1.32 [0.86, 2.03]
Indicator that state pays for only fluoride varnish (Reference group: state pays for additional oral health services)	0.48** [0.26, 0.76]	2.02*[1.12, 3.65]	0.60 [0.30, 1.19]	0.38*** [0.21, 0.67]
Years since policy enactment (Reference group: <1 year)
1 year	2.38*** [1.76, 3.22]	2.12*** [1.50, 3.01]	2.40*** [1.76, 3.26]	2.79*** [2.20, 3.55]
2 years	3.05*** [2.11, 4.40]	2.59*** [1.79, 3.75]	3.12*** [2.17, 4.51]	3.98*** [2.83, 5.60]
3 years	4.19*** [2.54, 6.91]	3.49*** [2.24, 5.42]	4.40*** [2.70, 7.18]	5.38*** [3.45, 8.39]
4 years	6.43*** [3.88, 10.66]	5.26*** [3.49, 7.93]	6.51*** [4.03, 10.53]	7.62*** [4.19, 13.86]
5 or more years	22.44*** [14.97, 33.65]	9.13*** [6.24, 13.35]	22.57*** [14.82, 34.38]	29.08*** [16.35, 51.73]
Count of total visits (unweighted)	3,360,339	3,393,638	3,360,339	6,729,269

Note. Logistic regression models controlled for year fixed effects, sex, age, race/ethnicity, number of months of Medicaid enrollment, and the following county-level characteristics: unemployment and child poverty rates, indicators of if the county was a full or partial primary care or dental health professional shortage area, and county rurality (metropolitan, non-metropolitan adjacent to metropolitan areas, and nonmetropolitan not adjacent to metropolitan areas). Models were estimated with sampling weights and standard errors were clustered at the state level. The main analysis includes visits paid via FFS and the FFS payment amount for fluoride varnish. Sensitivity 1 includes visits paid via FFS and adjusts for the payment amount for fluoride varnish and other preventive oral health services. Sensitivity 2 includes visits paid via FFS and adjusts for the payment amount for fluoride varnish and the percentage of the sample enrolled in Medicaid managed care plans in each state and year. Sensitivity 3 includes all visits paid via FFS or managed care and adjusts for the FFS payment amount for fluoride varnish. Abbreviations: Odds Ratio, OR; Confidence Interval, CI. ^†OR for log payment for fluoride varnish and additional preventive oral health services.

^*P<0.05

^**P<0.01

^***P<0.001.

Figure 1. Adjusted predicted probabilities of receiving fluoride varnish during well-child medical visits for Medicaid-enrolled children <6 years, payment percentile and years since policy enactment.

Note. Figure presents predicted probabilities of receiving fluoride varnish during a well-child medical visit, which were generated from an adjusted logistic regression model, described in the main text and Table 2 (main analysis). Sample included well-child medical visits paid via fee-for-service and received by children aged 6 months to 5 years in 22 states during 2006-2014.

Results for the sensitivity analyses, shown in columns 3-5 in Table 2, show the main results are robust, as we found no difference in magnitude, direction, or significance for the payment effect, as well as no statistically significance changes in any of the other coefficients. Some states pay for a package of services intended to be provided with fluoride varnish and we found that higher payment for fluoride varnish plus additional preventive oral health services was significantly associated with a higher likelihood of a medical well-child visit including any preventive oral health services (odds ratio (OR)=5.3; 95% CI=2.2, 12.3; P<0.001). Next, we adjusted for the percentage of our sample enrolled in Medicaid managed care plans in each state and year, finding a statistically significant association between higher payment for fluoride varnish and receipt of fluoride varnish during a medical visit (OR=9.1; 95% CI=3.5, 23.2; P<0.001). Finally, we added visits paid via managed care to our main model. We found no statistically significant difference in the odds of a visit included fluoride varnish if paid FFS vs. managed care (OR=1.3; 95% CI=0.9, 2.0; P=0.2) and found statistically significant association between higher payment for fluoride varnish and receipt of fluoride varnish during a medical visit (OR=19.5; 95% CI=7.0, 54.4; P<0.001).

Discussion

This study found that higher levels of Medicaid payment were associated with higher rates of well-child medical visits with fluoride varnish applications for young children. This association was robust in states with longer enacted policies, when examining cumulative payment for fluoride varnish plus additional preventive oral health services delivered during medical visits, and when exploring the impact among managed care enrollees. These findings extend the limited literature on Medicaid payment and children’s receipt of care in medical settings, which is important because this literature has primarily focused on adults (Saulsberry et al., 2019).

Our findings also extend the existing evidence base about the positive impact of state policies allowing for fluoride varnish applications during medical visits (Arthur & Rozier, 2016; Herndon et al., 2015; Kranz et al., 2020; Okunseri et al., 2009; Rozier et al., 2010). Identifying policies and interventions that promote dental care utilization for young children from low-income households is important because dental caries is common (Centers for Disease Control and Prevention, 2019), and its consequences can be severe and costly. In addition to causing pain and decreased quality of life (Abanto et al., 2011; Filstrup et al., 2003; Low, Tan, & Schwartz, 1999), untreated caries may harm school performance. A study of North Carolina children found that children with dental pain were three times more likely to miss school, and that absences due to dental pain were associated with worse academic performance (Jackson et al., 2011). Furthermore, poor oral health is costly. Because treatment of severe dental caries may require hospitalization and general anesthesia for young children, dental treatment costs can range from $1,500 to over $3,000 per episode (Bruen, Steinmetz, Bysshe, Glassman, & Ku, 2016; Meyer, Lee, & Casey, 2017). In a ranking of the 20 most expensive health conditions for children and adolescents, based on personal spending, oral disorders (e.g., oral surgery and dental caries, including fillings, crowns, and extractions) was ranked 5^th at $8.7 billion in 2013 (Bui et al., 2017). Our findings provide evidence about an effective intervention to promote young children’s oral health, an intervention supported by both medical and dental stakeholders and the US Preventive Services Task Force (American Academy of Pediatric Dentistry, 2016; Clark & Slayton, 2014; Moyer, 2014), which is important given the significant and widespread consequences of poor oral health.

Despite evidence that applying fluoride varnish during medical visits improves oral health (Kranz et al., 2015; Kranz et al., 2020), we found that fewer than one-in-ten well-child visits included fluoride varnish. This is troubling because rates of dental visits for this population are also low, particularly among Black children (MACPAC, 2016). This finding highlights the need for more efforts to increase young children’s receipt of preventive oral health services. In 2020, CMS launched a learning collaborative to support states in their efforts to improve the oral health of publicly-insured children, including through greater adoption of the delivery of fluoride varnish in medical settings (Centers for Medicare & Medicaid Services, 2020). This new collaborative will include educational webinars and technical assistance for states designing and implementing quality improvement projects to promote the oral health of publicly-insured children. Efforts from CMS, as well as the 2015 requirement that private health insurers deliver fluoride varnish without cost sharing (Lander, 2015), provide additional support for the delivery of fluoride varnish in medical settings.

Despite these promising findings that higher Medicaid payments increased children’s receipt of fluoride varnish during medical visits, state budget concerns may make it challenging for states to increase payments. Findings from a survey conducted in October 2019 indicate that, even before the COVID-19 pandemic, states were projecting increased Medicaid enrollment and state spending on Medicaid for fiscal year 2020 (Rudowitz, Hinton, Diaz, Guth, & Tian, 2019). Most states surveyed by the Kaiser Family Foundation anticipate exceeding their enrollment projections in fiscal year 2021 (Hinton, Stolyar, & Rudowitz, 2021). Irrespective of budget concerns, states are mandated to have FFS Medicaid payment rates sufficient to ensure access to care equivalent to that of the general population (Rosenbaum, 2019). Although the former Trump administration considered repealing monitoring of this “equal access” rule, there was pushback to this from stakeholders, including the nonpartisan Medicaid and CHIP Payment and Access Commission (MACPAC, 2019), and, despite budget concerns, states are still required to have reasonable payment rates.

This study should be interpreted in the context of its limitations. First, in our main analysis, we limited the sample to medical visits paid on a FFS basis because we could not identify payment amounts for managed care encounters in our dataset, meaning our findings may not be generalizable to managed care enrollees. The lack of usable payment data for Medicaid managed care enrollees is a known issue that has been highlighted by the Department of Health and Human Services Office of Inspector General (2021). Our findings, however, are relevant to the more than 24 million Medicaid-enrollees who did not receive care from comprehensive managed care organization in 2019 (MACPAC, 2020). Moreover, our results were robust in sensitivity analyses, which included examining the FFS payment among visits paid managed care and adjusted for managed care penetration by state and year. Further, as with any nonexperimental study, there are threats to the validity of the causal interpretations. There was little variation in payment rates within states, with an average change in FFS payment for fluoride varnish of $0 (standard deviation=$2). This means that our payment effect estimates are identified largely by cross-sectional variation in payment and and causal interpretation should be made in this context and with caution. We sought to address this by including other contextual charcteristics, such as rurality, rates of poverty and unemployment, and HPSA indicators and by controlling for the timing of policy enactment across states. Additionally, fluoride varnish was not delivered during medical visits prior to policy enactment, meaning we were unable to exmaine pre-period trends. Finally, the states and years included in this study were limited based on data quality and availability, and we do not know if our findings would generalize to states or years we were unable to examine.

This study of 22 states during 2006-2014 found that higher Medicaid payment was associated with higher rates of well-child medical visits with fluoride varnish applications for young children. As states consider strategies for increasing young children’s access to preventive oral health services, as well as strategies for balancing budgets, attention to the effects of provider payment policies on access to pediatric care, including oral health services, is warranted. As rates of fluoride varnish applications remain low in high-risk populations, additional efforts are needed to encourage medical providers to deliver fluoride varnish to young children.