General dentists typically do not follow the profession’s recommendations^10,11 of a first dental visit for all children on eruption of their first tooth but no later than 12 months of age.^2,4 The willingness of dentists in private practice to treat young children often is influenced by whether their care is paid for by Medicaid. In Ohio, the State Children’s Health Insurance Program (SCHIP) is part of Medicaid. Forty-five percent of children aged 0 to 4 years in the state were enrolled in Medicaid in state fiscal year 2003.¹² In 2002, only 26 percent of Ohio dentists submitted one or more Medicaid claims, and only 15 percent saw more than 50 patients enrolled in Medicaid during the previous year.¹³

Many low-income families turn to safety-net (SN) dental clinics for their children’s care. These clinics generally treat patients regardless of their ability to pay and have a substantial share of the patient mix that represents uninsured, Medicaid and other vulnerable populations.¹⁴ SN dental clinics are found in a variety of settings, including hospitals, dental schools, local health departments, community health centers and clinics operated by other nonprofit organizations. There may be considerable variation among SN dental clinics with regard to the scope of services, hours of operation, staff expertise and mission.

During much of the 1990s, Head Start programs and parents of children in Head Start programs nationwide reported access to dental care as their number one health issue.⁷ Head Start programs primarily target low-income families with children aged 0 to 5 years. Ten percent of enrollment slots (regardless of income) are to be filled by children with disabilities.¹⁵

We conducted a study to determine the extent to which Ohio primary dental care providers (general and pediatric dentists and SN dental clinics) treat very young children (defined as 0–2 years of age for the purposes of this study) and young children (defined as 3–5 years of age), those who are enrolled in Medicaid, those enrolled in Head Start and those with disabilities. We also sought to identify factors that influence dentists’ willingness to care for these children.

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Using an Ohio State Dental Board database, we selected a random sample of 13 percent of the state’s 4,984 general practitioner (GP) dentists and 72 percent of the 139 pediatric dentists (PDs) to receive a pretested 22-question survey that was mailed in late 2002. We sent a second mailing to nonresponders. After adjusting for undeliverable surveys and retired dentists, the response rate was 63.2 percent (63.6 percent for GPs and 60.6 percent for PDs).

We sent a separate mailing to all Ohio SN dental clinics, with the exception of those known to limit care to nonpediatric populations. A second mailing to nonresponders and telephone follow-up ensured responses from all 72 clinics surveyed.

In the data analysis, we stratified responses for GPs in private practice according to years since graduation and geographic character of their practices (for example, urbanized, rural). For descriptive analysis, we established three years–since-graduation categories on the basis of information from the ADA, which defines new practitioners as those who graduated from dental school less than 10 years ago,¹⁶ as well as from reports indicating that independent dentists’ incomes peak with the 25- to 29-year postgraduation cohort, which suggests that dentists begin to cut back on their practices beginning at approximately 30 years after graduation.¹⁷ For correlation analysis and analysis of variance (ANOVA), we maintained years since graduation as a continuous variable. We established the geographic character of respondents in private practice with geocoding using geographic information system software (ArcView GIS, Version 3.3, Environmental Systems Research Institute, Redlands, Calif.). Using census definitions,¹⁸ we categorized dentists into mutually exclusive geographic classifications (Box).

We created several scaled variables to group-related survey items according to underlying issues (Table 1). The groupings reflected constructs that are supported in the literature on dental care access.^19–21 We created each scaled variable by running a reliability analysis (Chronbach’s alpha)—a measure of internal consistency of a set of items that make up a scale. If an item lowered the reliability and was not thought to be central to the construct, we dropped it. We calculated the grand mean for remaining items.

Not all respondents answered each question. In the analysis, we did not attribute a value to a nonresponse to a given question. However, we considered negative responses to the question regarding whether at least one child aged 0 through 5 years was treated in the previous year to be negative responses to the same questions for 0-through-2-year-olds and for 3-through-5-year-olds. In this report, the expressions "0 through 2" and "3 through 5" refer to ages birth through 2 years and 3 through 5 years, respectively.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Dentists’ treatment of young children. PDs generally provide care to children of all ages, while GPs are more likely to see 3-through-5-year-olds than 0-through-2-year-olds (Figure 1⁸). When asked about their ability to manage the behavior of 0-through-2-year-olds, only 31 percent of GPs strongly or somewhat agreed that they had a high comfort level, while 71 percent of GPs strongly or somewhat agreed that they had a high comfort level. Forty-seven percent of SN-General and 67 percent of SN-GPR/AEGD clinics strongly or somewhat agreed that they had a high comfort level in treating 0-through-2-year-olds, while 89 and 100 percent, respectively, strongly or somewhat agreed that they had a high comfort level in treating 3-through-5-year-olds.

View larger version (33K): [in this window] [in a new window]   Figure 1. Ohio primary dental care providers’ treatment of young children during preceding 12 months. Reprinted with permission of the American Public Health Association from Siegal and colleagues.8 GPR/AEGD: General Practice Residency/Advanced Education in General Dentistry.

Dentists’ treatment of young Medicaid recipients. PDs were more than three times as likely as GPs to treat 0-through-5-year-olds covered by Medicaid (Figure 1). Fifty-seven percent of PDs and 69 percent of GPs who saw children enrolled in Medicaid had limitations on their involvement; only 7 percent of all GPs and 29 percent of all PDs reported accepting patients enrolled in Medicaid without limitation compared with 80 percent of SN dental clinics.

While most dentists who treated patients enrolled in Medicaid indicated having more than one limitation for accepting these patients into their practices, the most common limitation for GPs (40 percent) was to accept patients of record but not new patients, and the most common limitation for PDs (35 percent) was to accept only referred patients. On the basis of our survey’s findings, we estimate that 13 percent of Ohio GPs and 59 percent of Ohio PDs accept new 0-through-5-year-old patients enrolled in Medicaid. Other common bases for accepting patients enrolled in Medicaid were to limit them to emergencies (GP = 21 percent, PD = 25 percent), children enrolled in a Head Start program (GP = 18 percent, PD = 15 percent) or a patient-by-patient determination (GP = 25 percent, PD = 20 percent).

For three dimensions that might limit participation in Medicaid—administrative issues, reimbursement and client behavioral issues—we found differences between PDs and GPs on those dependent variables (F_1,321 = 5.686, P < .01, partial ² = .051). Follow-up ANOVA showed that PDs were not as likely to perceive Medicaid administrative issues (F_1,321 = 17.090, P < .001, partial ² = .051) or Medicaid reimbursement (F_1,321 = 8.237, P < .01, partial ² = .025) as barriers to treating Medicaid-enrolled children compared with GPs. We found that the Medicaid dimensions were not significantly correlated with years of practice.

Using the same dimensions as those of dependent variables, MANOVA results revealed differences between GPs who were not participating in Medicaid and those who were (F_1,247 = 16.722, P < .001, partial ² = .171). Follow-up ANOVA tests showed that GPs who were not participating in the Medicaid program were more likely to perceive Medicaid administrative issues (F_1,247 = 48.209, P < .001, partial ² = .164) and Medicaid reimbursement (F_1,247 = 29.263, P < .001, partial ² = .106) as barriers compared with GPs who were participating in the Medicaid program. The majority of SN dental clinics disagreed with all of the factors, except "missed/late for appointments"—the factor most frequently reported as a barrier for all groups of dentists.

Dentists’ treatment of children in Head Start programs. Figure 1 shows that PDs were approximately twice as likely as GPs to have treated a child enrolled in a Head Start program in the past 12 months, (² [1, n = 373] = 28.31, P < .001). Essentially, all of the SN dental clinics saw children enrolled in Head Start programs. PDs, with few exceptions, reported providing both examinations and restorative care to the majority of children in Head Start programs, as did essentially all SN-Pediatric and SN-GPR/AEGD clinics. More than one-third of GPs (39 percent) and SN-General clinics (36 percent) limited care for children in Head Start programs to examinations.

A young child’s being enrolled in Head Start appears to make dentists in private practice, but not SN dental clinics, somewhat more willing to see the child whose care is paid for by Medicaid. Eighteen percent of GPs and 13 percent of PDs in private practice reported that they would be more willing to treat Medicaid children enrolled in Head Start than Medicaid children who were not enrolled in Head Start.

Dentists’ treatment of children with disabilities. Figure 2 shows that GPs’ treatment of 0-through-5-year-old children with disabilities is situational. PDs were more likely than GPs to treat children with disabilities, (² [3, n = 351] = 153.10, P < .001). A one-way ANOVA showed that PDs were influenced less by type and severity of disability than were GPs (F_{1, 362} = 73.11, P < .001, partial ² = .168).

View larger version (30K): [in this window] [in a new window]   Figure 2. Ohio dental care providers’ treatment of young children with disabilities. GPR/AEGD: General Practice Residency/Advanced Education in General Dentistry.

We used MANOVA and follow-up ANOVA tests to detect significant differences in GPs’ geographic settings related to 10 variables, using years of practice as a covariate. The only significant difference we found was that dentists in central city areas were more likely to consider access to public transportation in selecting practice location than were their counterparts in rural and small city locations (F_3,198 = 7.58, P < .001, partial ² = .105). GPs in urban clusters were the most likely to see children with Medicaid coverage (46.4 percent), and those in urban areas were the least likely (11.9 percent) (² [3, n = 316] = 31.064, P < .001). GPs in urban clusters and rural areas were more likely than their urban counterparts to treat children in Head Start (² [3, n = 316] = 34.376, P < .001).

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
GPs must be the core of any strategy to care for the majority of young children, including those who are most vulnerable owing to familial socioeconomic status. PDs and SN dental clinics may be more likely to treat young children and those with Medicaid coverage than are GPs, but there are approximately 25 times as many GPs in Ohio as PDs and SN dental clinics combined.

Our study confirms that new strategies for ensuring dental care access for young children from low-income families are necessary. Such strategies may include interpeer advocacy, education, practice incentives, or GP and PD teams coordinating their respective skills to serve vulnerable young children better.

Our study found that being younger than 3 years and enrolled in Medicaid were the greatest barriers to young children being treated by GPs. Being enrolled in Head Start increased dentists’ willingness to treat children. Our study also found that dentists’ attitudes and behaviors toward treating very young children (0–2 years of age) did not reflect the profession’s recommendation for the first dental visit to be by 12 months of age. This finding was consistent with that of a national study,² as was our finding that treatment of young children was not associated primarily with dentists’ perceptions of inadequate training or that they did not enjoy treating children. Ohio GPs indicated that their offices were "not well-equipped to provide care to children aged 0 through 2 years," and those in the national study indicated that their offices "were not geared to children." To the extent that those responses related to relatively inexpensive items such as papoose boards and mouth props, rather than staff to administer sedation, it is difficult for us to understand the lack of equipment as being an insurmountable barrier. The next most common response, child’s behavior (for example, crying or screaming) disrupting the office, provides insight into the reason for GPs’ reluctance to see young children. Dentists likely will require encouragement and incentives to change their practices.

Although newer dentists were more willing to provide more difficult patient care, they were not significantly more willing to accept young patients enrolled in Medicaid. Further study using a more controlled design could determine better whether newer dentists’ willingness to provide more difficult care for younger children relates to their being less selective in accepting patients into developing practices or to a cohort effect in which newer dentists are better prepared to treat young children.

The issues with the greatest implication for access to dental care for high-risk young children are dentists’ willingness to accept Medicaid and the limitations placed by those who do. Our study’s finding of dentists’ self-reported participation in Medicaid (22 percent) is consistent with the 26 percent of Ohio dentists who had Medicaid claims in 2002, regardless of patient age.¹³ Our study provided insight into the distinction between being a Medicaid provider of dental care and being an access point for patients enrolled in Medicaid. Most GPs who see patients enrolled in Medicaid place limitations on which ones they will see; being enrolled in Head Start or having a need for emergency care, however, improve a child’s chances.

Typical reasons that dentists offer for not treating Medicaid recipients can be grouped as administrative issues (for example, fees, payment speed, getting questions answered) and client behavioral issues (for example, keeping appointments/timeliness, unruly children in office).^3,22 Contrary to another study that found that administrative issues were more serious problems than were those related to client behaviors,⁴ our study found that broken appointments was the biggest complaint, and fees were ranked as a lesser issue.

Across all categories of primary care dentists, a higher percentage reported treating children enrolled in Head Start than young children enrolled in Medicaid (who account for approximately three-fourths of all children enrolled in Head Start) in the past 12 months. This finding has several possible explanations, including

– efforts that Head Start may make to ensure that patients keep appointments;

– some dentists’ perceptions that they are limiting their Medicaid exposure by serving only children in Head Start;

– the role of Head Start as a payer (of last resort) for dental care in expanding access to children not enrolled in Medicaid.

Our finding that 39 percent of GPs were more likely to examine young children’s mouths than to restore their carious teeth contributes to Ohio Head Start programs’ reporting difficulty in finding dentists to treat their children.⁸

Our study demonstrated that Ohio SN dental clinics mirror their private practice counterparts. When it comes to dental care for young children, the institutionally based programs (particularly those that train PDs) are most like PDs, and the more common SN-General dental clinics are more like GPs, the major exception being the willingness of clinics to accept Medicaid.

GPs’ attitudes about the age for a first dental visit is one significant barrier. A multifaceted effort, including dentist training, parent education and enhanced Medicaid fees, however, showed some success in increasing the percentage of children up to 3 years of age who had a dental visit.²³

The primary limitations of our study, as with all such studies, are the extent to which the responses reflect office practices of respondents and the extent to which the 37 percent of dentists who did not respond are similar to those who did with regard to the issues of interest. In addition, our study could not determine whether more recent dental-school graduates were more likely to treat very young children because they have better training or because they cannot afford to be as selective when building their practices as established dentists can.

	CONCLUSIONS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Because the numbers of Ohio PDs and SN dental clinics are so inadequate compared with the treatment needs of young children, particularly those enrolled in Medicaid, GPs must be the primary providers for this population. Few GPs subscribe to the ADA/American Academy of Pediatric Dentistry recommendation for a child’s first dental visit to be by 1 year of age. Most GPs do not see children younger than 3 years, and when they do, the children are most likely to receive an examination or emergency treatment.

The likelihood of an Ohio dentist treating a young child is reduced dramatically if Medicaid pays for the child’s care; however, it is increased marginally if the child is enrolled in Head Start. Although Ohio PDs and GPs in private practice share the same concerns about Medicaid, PDs are three times more likely to treat patients enrolled in Medicaid. Most GPs who see patients enrolled in Medicaid limit whom they will treat, most often to patients of record. We estimate that 13 percent of Ohio GPs and 59 percent of Ohio PDs accept into their practices new 0-to-5-year-old Medicaid recipients. Differences between dentists’ attitudes and practices for the populations of young children in question were minimal with regard to years since graduation and geographic character of the practice.