Effects of WIC Participation on Children’s Use of Oral Health Services

Abstract

Objectives. We estimated the effects of the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) on dental services use by Medicaid children in North Carolina.

Methods. We used linked Medicaid claims and enrollment files, WIC files, and the area resource file to compare dental services use for children enrolled in WIC with those not enrolled. We used multivariate models that controlled for child clustering and employed 2-step methodology to control for selection bias.

Results. Children who participated in WIC had an increased probability of having a dental visit, were more likely to use preventive and restorative services, and were less likely to use emergency services.

Conclusions. Children’s WIC participation improved access to dental care services that should lead to improved oral health.

Dental disease and access to dental care are major public health problems for young low-income children living in the United States. This national dilemma has recently come under close scrutiny by policymakers, health care providers, and researchers.1,2 Because of frequent interactions with low-income children and their families, several public health programs can help alleviate dental problems and improve access to dental care. An example of such a program is the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC). WIC is administered by the Food and Nutrition Services of the US Department of Agriculture and serves more than 7.4 million individuals.3 Over one third of infants born in the United States are enrolled in the WIC program. It is often the first contact with the health care system for many poor women and children. The primary goal of WIC is to “improve the health of women and children by providing nutritious foods, nutrition education and good health care during pregnancy, the postpartum period, infancy and early childhood.”4(p391) To achieve its goal, WIC agencies work to improve the linkage between clients and health care providers, including dentists, through referrals and networking.5

WIC and Medical Care

Several investigations have demonstrated the effectiveness of WIC in providing positive health outcomes. Among the benefits are a reduced frequency of low-birthweight deliveries,6–8 reduced Medicaid costs for newborns,9 reduced rates of anemia in children,6 and increased nutrient intake in children.10 The health effects of WIC participation are attributed to the direct nutrition-related benefits of the program rather than to the benefits of coordination of health and social services.11 However, the beneficial effects of the coordination of health services cannot be dismissed readily.12 Referral of children to health and social services is an important component of WIC programs. WIC counselors advise families about the location of clinics and programs, such as Medicaid and the State Children’s Health Insurance Program, that help pay for health care.

Few studies have addressed the effects of WIC on the utilization of health care services.3,6,10 One investigation demonstrated that WIC participants used a children’s clinic more frequently than did nonparticipants.12 In a recently published study, Buescher et al. reported that “Medicaid-enrolled children participating in the WIC program use all types of health care services compared with Medicaid-enrolled children who were not WIC participants.”13(p145) They concluded that “the health care needs of low-income children who participate in WIC may be better met than those of low-income children who were not WIC participants.”13(p145)

WIC and Oral Health Care

Only 2 small, descriptive studies have investigated the relation between participation in WIC programs and referrals for oral health care. McCunniff et al.14 examined dental referral rates by WIC clinics in Missouri, reporting that of the 1850 participants seen during a 2-month period at a clinic site, 27% of children and 17% of infants were referred for services outside the WIC clinic. Using selfreported information from caregivers, McCunniff et al.14 also reported a statistically significant bivariate relation between WIC referral and having a dental visit. Dental referrals constituted 10% of these referrals made for infants and children. Sargent et al.15 used a survey of WIC employees in an inner-city clinic to examine referral patterns. WIC nutritionists at this site offered a variety of referrals to their clients. One fifth of WIC participants were referred to health services, and referrals for dental care constituted the majority of these referrals. Results of these 2 studies suggest that dental referrals do take place within WIC clinics, but the proportion of all referrals due to a dental problem varied considerably. Furthermore, the outcome of these referrals is largely unknown.

Findings from previous WIC and health care utilization investigations are further limited because they do not control for the nonrandomized nature of WIC participation.13,16 We build on these past studies in an investigation of WIC participation and use of dental services by using claims data rather than self-reports and by employing multistage modeling to control for selection bias. Using these more sophisticated analytic techniques in our exploration of the effects of WIC on use of oral health services, we provide an estimate of the effectiveness of WIC programs in linking clients with dental providers. Specifically, we determined whether, compared with Medicaid children who did not participate in WIC, preschool-aged Medicaid children enrolled in WIC had an increased likelihood of having had a dental visit and whether it was of a certain type (preventive, restorative, or emergency.

METHODS

We tested the hypothesis that children enrolled in WIC have better access to dental care than children not enrolled in WIC. For purposes of this investigation, we defined access as use of services that was represented by Medicaid claims data. We measured the effects of children’s WIC participation on the probability of a child’s having had a dental visit and the different types of visits. Available sociodemographic variables were included as control variables in our analyses.

Data Sources and Study Cohort

We used the following linked North Carolina administrative data sets for our investigation: composite birth records, Medicaid eligibility enrollment files, Medicaid dental claims, WIC files, and the area resource file. The linkage process for these files has previously been reported, and a matching rate of 98.5% was etablished.13

All children born in North Carolina in calendar year 1992 who were enrolled in the Medicaid program were eligible for inclusion in the study. Children were excluded if they had more than 1 Medicaid identification in their records (759 children) or if they had recorded periods of Medicaid enrollment indicated before the date of birth (1371 children). A Medicaid enrollment history was created for each child in which enrollment status was indicated for each month of life from birth to the age of 5 years (months 1 through 60).

Analysis Strategy

Using various multiple regression analyses with control variables, we determined the relation of children’s WIC participation and the use of oral health services. Several analytic challenges had to be considered. Because our investigation followed children for 5 years, the data set contained multiple observations per child that can result in correlated error terms and bias results. To control for this problem, we used robust standard errors and controls for clustering around the child in the regression analysis. Panel data techniques using random effects models were used in each analysis.

Another important analysis consideration was the potential for selection bias. Because WIC participation was not randomly determined among Medicaid children, we anticipated that children who participated in WIC would more likely be users of health care services than other Medicaid children. This assumption suggests the potential for correlation between the WIC participation variable and the error term of our main analysis. Failure to account for this correlation could bias our results. To control for this potential bias, we incorporated a 2-stage multilevel model using instrumental variables as outlined by Bollen et al.17 We screened potential instrumental variables for their close association with WIC participation. Three instrumental variables (number of WIC clinics per county, number of full-time WIC workers per county, and WIC hours of operation per county) were correlated with WIC participation but not with dental utilization and were used in our analysis.

The primary measure of oral health utilization was a 3-level variable defined as no visits, 1 visit, or 2 or more visits per year as represented by dental claims. These categories were used because the recommended number of dental visits is 2 per year.18 We did not distinguish among the number of visits greater than 2 per year because those differences are likely to depend on the severity of dental disease rather than basic issues associated with access to care. Because the measure of this dependent variable was coded as 0, 1, and 2, we used an ordered probit analysis with controls for clustering to examine WIC and oral health services utilization for each observation year (1 through 4).

We used separate random effects logit analyses (logistic regression) for each type of oral health services used. Type of services received was classified as (1) diagnostic/preventive services, (2) restorative services, and (3) emergency services. The American Dental Association procedure codes were used to classify service categories.18 Additionally, if a child had a hospital emergency room claim with a primary diagnosis of dental caries19 it was included as an emergency visit. These visit types were coded 0 or 1 to represent each type of visit per year as indicated by Medicaid claims.

Our major explanatory variable was children’s WIC participation measured as the number of months when any WIC voucher was redeemed during each year of life. We also incorporated relevant available sociodemographic control variables into our analysis. These included maternal educational level (years of school completed), maternal age (years), household income (actual dollar amounts), marital status (unmarried vs married), and minority status (non-White vs White). Many of these variables have been reported as important determinants of access to oral health care.20–22 In addition, we used length of Medicaid enrollment (months) and dentists-to-population ratio in our analyses.

RESULTS

Descriptive Statistics

Summary statistics for the study cohort are presented in Table 1 ▶. Of the 81 518 live births in North Carolina in 1992, 53 591 children were enrolled in Medicaid and 49 795 met the study inclusion criteria at birth. Our cohort was reduced to 21 277 at 1 year of age because the eligibility for Medicaid changes from 185% of the federal poverty level during the first year of life to 133% of the federal poverty level thereafter. Approximately 18% of the children born with Medicaid benefits remained continuously enrolled. Approximately 12% stayed continuously enrolled in WIC for the entire 5-year study period. The average number of months per year enrolled in Medicaid was 7.6. More than 50% of the cohort was on WIC at any time during the study period. The average length of children’s WIC participation was 4.4 months per year. The average maternal age was 21 years, with an average educational level of 11th grade. Forty-eight percent of the population was non-White. In each observation year, 7% of the sample had a dental visit of any type, 6% had a preventive visit, 2% had a restorative visit, and 1% had an emergency visit. Overall, 12.1% of the cohort used oral health services before the age of 5 years.

TABLE 1—

Characteristics of the Study Population: North Carolina, 1992–1997

	Mean or Percentage	SD	Minimum	Maximum
WIC variables
Time WIC vouchers were used, mo per year	4.47	3.06	0	12
Infant WIC participation	51%	0.33	0	1
Outcome variables
Any dental visit in a year	7.1%	0.25	0	1
Number of dental visits per year	0.4	2.02	0	6
Had a preventive visit in a year	6.6%	0.25	0	1
Had a restorative visit in a year	2.2%	0.15	0	1
Had an emergency visit in a year	1.5%	0.12	0	1
Control variables
Maternal age, y	21	5.51	13	39
Maternal education	11th grade	4.39	9	18
Medicaid enrollment, mo per year	7.6	5.26	1	12
Household income, $	20 550	4140	12 200	29 130
No. dentists/population	6.820	3.81	0	17.6
Unmarried	54%	0.48	0	1
Non-White	48%	0.50	0	1

Note. WIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

Analytic Results

The analytic file consisted of 4 observations per child (n = 49 795) representing 1 for each year of life until the age of 5 years, for a total of 199 180 child-year observations. Because the first year of life represented the infant year and use of oral health services was extremely low, we omitted this year from our analysis, leaving 4 observation years for ages 1 through 4 years. Table 2 ▶ illustrates the model results of our analysis of children’s WIC participation and oral health services utilization (any visit, 1 visit, 2 or more visits per year). Children’s WIC participation was a significant (P < .01) factor with a positive effect on oral health services utilization. Also significant in the model were Medicaid enrollment (P < .01), household income (P < .05), dentist-to-population ratio (P < .05), and being non-White (P < .05).

TABLE 2—

Ordered Probit Model Results for WIC and Use of Oral Health Services: North Carolina, 1992–1997

Variables	Estimated Coefficient (SE)
WIC variables
Child WIC participation	0.043*** (0.013)
Infant WIC participation	0.032*** (0.011)
Control variables
Maternal age	0.014* (0.00086)
Maternal education	−0.0017 (0.0016)
Medicaid enrollment	0.101*** (0.006)
Household income	0.060** (0.014)
Dentists/population	0.015** (0.007)
Unmarried	−0.0013* (0.00078)
Non-White	−0.32** (0.014)
Constant 1	2.52 (0.030)
Constant 2	2.89 (0.030)

Note. WIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

*P < .10; **P < .05; ***P < .01.

Table 3 ▶ illustrates the likelihood of dental visits by level of WIC participation. These results were calculated using the base-case child (White, maternal age of 21 years, maternal educational level of grade 11, household income of $20 550, and mother married and enrolled in Medicaid for 7.6 months). Children who participated in WIC for a full year were about 1.7 times more likely to have 2 or more dental visits per year than those children who never participated in WIC (odds ratio [OR] = 1.67; 95% confidence interval [CI] = 1.45, 1.78). Children who participated in WIC for a full year were about 1.5 times more likely to have 1 dental visit than children not on WIC (OR = 1.46; 95% CI = 1.32, 1.56).

TABLE 3—

Odds Ratios for Dental Services Use, by Level of WIC Participation: North Carolina, 1992–1997

	No WIC Participation (Reference Group)	6-Month WIC Participation	1-Year WIC Participation
≥ 2 Dental visits per year	1.0	1.51***	1.67***
1 Dental visit per year	1.0	1.26**	1.46***
No dental visits	1.0	0.85*	0.79**
Preventive care visit	1.0	1.59***	1.98***
Restorative care visit	1.0	1.26***	1.69***
Emergency care visit	1.0	0.79*	0.68*

Note. WIC = Special Supplemental Nutrition Program for Women, Infants, and Children; OR = odds ratio.

*P < .10; **P < .05; ***P < .01.

Table 4 ▶ presents results for the 3 logit models for children’s WIC participation and type of dental visit. Children’s WIC participation was significant (P < .01) and had a positive effect on the likelihood of a preventive or restorative visit and a negative and marginally significant effect (P < .10) on emergency visits. Table 3 ▶ presents the likelihood of type of dental visit by level of WIC participation. Children who participated in WIC for 12 months were more likely to have a preventive visit (OR = 1.98; 95% CI = 1.85, 2.02) and a restorative visit (OR = 1.69; 95% CI = 1.56, 1.87) but less likely to have an emergency visit (OR = 0.68; 95% CI = 0.56, 0.89) than children who did not receive WIC services.

TABLE 4—

Random Effects Logit Models for WIC and Type of Visit: North Carolina, 1992–1997

	Preventive Visit (SE)	Restorative Visit (SE)	Emergency Visit (SE)
WIC variables
Child WIC participation	0.065 (0.026)***	0.047 (0.0035)***	−0.036 (0.0021)*
Control variables
Maternal age	0.0048 (0.0011)***	0.037 (0.0014)***	0.00070 (0.0015)
Maternal education	−0.0050 (0.0015)***	−0.0097 (0.0026)***	−0.011 (0.0028)***
Medicaid enrollment	0.140 (0.0018)***	0.12 (0.0028)***	0.11 (0.0029)***
Household income	−0.0030 (0.0015)**	−0.020 (0.0085)**	−0.015 (0.0045)***
No. dentists/population	0.023 (0.0096)**	0.016 (0.0021)**	−0.020 (0.0024)***
Unmarried	−0.0041 (0.0025)*	−0.002 (0.018)	−0.033 (0.019)*
Non-White	−0.038 (0.030)***	−0.014 (0.0072)**	−0.047 (0.018)***
Constant	−2.70 (0.58)***	−3.13 (0.77)***	−3.05 (0.87)***

Note. WIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

*P < .10; **P < .05; ***P < .01.

DISCUSSION

This investigation represents the first detailed examination of children’s WIC participation and oral health care utilization. Previous studies have substantiated that WIC participation has an effect on the use of prenatal care and children’s medical care.3,6,10 Our findings indicate that children who participate in WIC are more likely to have a dental visit, thus increasing their access to oral health care. Because children on Medicaid are a high-risk population who often need more frequent and extensive dental services than other children, the association between WIC and greater use of services suggests that children on both Medicaid and WIC are more likely to receive the care they need.

Our results also showed that young children participating in the WIC program are more likely to use preventive and restorative services and are less likely to use emergency services than WIC nonparticipants. These findings suggest that children participating in WIC may have a better connection to the health care system that can lead to care that is more planned and less urgent. If further studies can confirm a causal link between activities of a WIC nutritionist, dental referral, and use of dental services, it would appear that an important facet of the mission of WIC programs—appropriate referrals to health and social services—is being addressed. Our study contributes to the literature in 2 major ways: by broadening the understanding of the effects of WIC participation to oral health care and by making a methodological improvement on the way WIC effects are examined.

Methodological Contributions

Our study is the first to examine the WIC program and health services utilization with the 2-step statistical modeling approach. A strong criticism of previous WIC children’s health studies is their inability to control for the potential selection bias of enrollment in the WIC program.16 We conducted extensive tests for these sources of bias in the relation between WIC enrollment and use of oral health services and found that selection bias did exist. Random assignment of families to WIC participation would be a stronger design and would help overcome any selection bias. However, the practical problem of implementing this strategy in a community-based setting would be daunting, and such a design is not ethically defensible. To help control for selection bias, we used the 2-step methodology as described by Bollen et al.17 This approach makes a significant contribution to the literature about the WIC program because the majority of studies that examine the WIC program have been unsuccessful in recognizing and correcting for selection bias (endogeneity).16 Our study also demonstrates the feasibility of using 2-step analysis to control for selection bias when examining the effects of WIC on oral health care use.

Sporadic and continuously enrolled Medicaid children differ demographically, socially, and economically,23 so we did not limit the cohort to continuously enrolled children. Doing so would have biased the sample and threatened generalizability. Continuously enrolled children in our cohort had utilization rates for oral health services that were 3 times greater than those not enrolled continuously in Medicaid (30% vs 10%).24 Instead, we controlled for Medicaid enrollment with a variable for duration of months per year enrolled in the program. However this approach can create another bias because low-income children qualifying for Medicaid may receive dental care under private insurance, with fee for service, or at no cost during periods in which they are not enrolled in Medicaid. Although low-income children are likely to use medical care when not enrolled in public insurance programs, it is unlikely that they receive dental care, particularly young children in North Carolina, where excess demand for services exists and dentists’ participation in Medicaid is low.25 Furthermore, dental care is the most prevalent unmet need in uninsured children who are eligible for Medicaid.26 These findings underscore the difficulty young low-income children have in gaining access to oral health care regardless of insurance coverage.

Policy Contributions

In the policy area, we have several noteworthy findings. Our results indicate that children’s WIC participation has a significant and positive effect on oral health services utilization during the first 5 years of life. This finding is important because inadequate access to dental care is commonplace among children of families living in poverty. This situation has been documented in numerous national and state reports including those from the Office of the Inspector General,27 the American Dental Association,19 the General Accounting Office,28 the surgeon general,2 and the North Carolina Institute of Medicine.29

In the North Carolina Institute of Medicine report on access to dental care, it was reported that fewer than 13% of children aged 1 to 5 years received any dental services. Our findings indicated that WIC participation could increase to 23% the use of dental services for children at this age. This is an increase of almost 50% compared with findings of the North Carolina Institute of Medicine. Even with this rather dramatic percentage increase in use, the absolute utilization of 23% is still relatively low; more work needs to be done to address the access to dental care crisis among low-income young children. Medicaid alone is not enough to provide sufficient access to oral health care for young children, but when available in combination with another public health program such as WIC, access to oral care health can be greatly improved.

Limitations

These results should be considered in light of the study’s limitations. First, we have a short study duration of 4 years (ages 1 through 4). For this reason, we chose the random effects model and not the fixed effects as suggested by Judge.30 The random effects model was able to control time-invariant variables as well as time-variant ones. We analyzed the child-level claims records and used panel data techniques that have not been applied in previous studies of WIC participation and use of services. Although we feel that we have made methodological improvements in the evaluation of WIC participation, future work should draw on additional years of panel data and address future questions such as the long-term oral health effects of the WIC program. In addition, because we examined the WIC program in only 1 state, this study needs to be replicated in other states. A single, expanded study with several states would involve considerable time and costs but could provide more generalizable and precise effects of the WIC program on access to oral health services.

We do not have information on children’s oral health status in the study. It is well documented that Medicaid children and children living in poverty have disproportionately more dental disease.31 Thus, any dental visit is likely to be beneficial to this high-risk population. The association of WIC with higher use of services may mean that oral health care needs of the children on Medicaid who participate in WIC are being better met. The findings for number of visits and type of use support this conclusion. Also, the adjusted odds of a child on Medicaid with WIC participation having 1 or more dental claims during the year was 1.3 to 1.5 times greater than that of a child with no WIC participation, across the 4 age groups. Studies have suggested that dental care is a serious unmet need among children in poverty, and our study suggests that use of dental care is enhanced among children participating in WIC services.