PEDIATRICS Vol. 105 No. 4 April 2000, pp. 780-788

An Elementary School-Based Health Clinic: Can It Reduce Medicaid Costs?

E. Kathleen Adams, PhD^* and Veda Johnson, MD^{, §}

From ^* Rollins School of Public Health, Emory University;  Whitefoord Elementary School-Based Health Clinic; and ^§ Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia.

	ABSTRACT

Top Abstract Methods Results Discussion References

Objective.  School-based health services have evolved from primarily controlling communicable disease to comprehensive programs with direct services, education, and improvement of the school environment. School-based health clinics (SBHCs), currently 1157 in number, are used to reach children for preventive and other routine care. Although several studies have examined the costs and effects of such programs, few, if any, have examined their potential to save Medicaid program outlays. The objective of this study was to assess the effect of the Whitefoord Elementary School-Based Health Clinic (WESBHC), located in Atlanta, Georgia, on health care costs paid by Georgia Medicaid over the 1994-1996 period. This clinic has been in operation since late 1994.

Design.  The analysis uses 1994-1996 claims data for Medicaid-enrolled children 4 through 12 years old served by the WESBHC and similar children in a comparison school district without such a clinic. Descriptive and multivariate analyses are used to discern the differences in the changes in Medicaid expenses per child-year enrolled for these 2 groups of children. Both those who only used the WESBHC sporadically and those for whom it was their medical home were identified for analysis.

Results.  The descriptive analysis shows that although there were no significant differences in the Medicaid expenses for the WESBHC and comparison children in 1994, before the operation of the WESBHC, by 1995, the WESBHC children had significantly lower emergency department expenses. In addition, they had higher Early Periodic Screening Diagnosis and Treatment preventive care expenses. By 1996, the WESBHC children had significantly lower inpatient, nonemergency department transportation, drug, and emergency department Medi- caid expenses. Multivariate analysis confirmed the effect of the WESBHC on lowering emergency department expenses.

Conclusions.  The results strongly suggest that the operation of a SBHC can have effects on the child's use of services and health care expenses. Given that these clinics serve all those who come for care and many of these are low-income children, these savings are likely to accrue to the Medicaid program of the state. As states continue to implement Medicaid-managed care for their child populations, they will need to consider the ability of SBHCs to participate in and receive Medicaid revenues through health maintenance organization networks.  Key words:  school clinics, medicaid, costs.

School health services, most often provided by nurses, have existed in this country for >100 years. Although these activities initially consisted primarily of communicable disease control, a broader comprehensive view of school health programs began to emerge in the 1970s. The concept of comprehensive school health programming in which direct services, education, and improvement of school environment are considered complementary efforts is now accepted in the public health and educational communities. A recent national survey found that community-based health care providers often play a key role in school health services, while school nurses continue to operate in a traditional fashion.¹

The Whitefoord Elementary School-Based Health Clinic (WESBHC) has been in operation in metro Atlanta since November of 1994. Its purpose, like other school-based health clinics (SBHCs), is to overcome barriers to care for the wide range of medical and social problems that confront children residing and attending school in a relatively poor community. The purpose of this analysis is assess the effect of the WESBHC on the children's use of expensive health care services, such as emergency department and inpatient hospital, while potentially increasing the use of preventive and other primary care services funded by the Medicaid program.

As noted, the concept of school-based services has grown in popularity during the past 2 decades. In a recent survey, the services provided by school nurses included primarily direct service/intervention, screening, special education, and prevention.¹ One national survey documented that school nurse practitioners diagnosed 87% of the presenting health problems, achieved 96% resolution of those problems, and avoided unnecessary duplication of ser- vices.² There are also school-based clinics, which encompass the regular supervision and visitation by physician providers of service, such as the WESBHC.

School-based and school-linked health centers now serve as comprehensive centers for medical and mental health screening and treatment for children on or near the school grounds. They are designed to address barriers to care, such as transportation, inconvenient appointment times, burdensome out-of-pocket costs, and other personal barriers to seeking care. School-based clinics are being widely used to reach children for preventive and other routine care; there are currently over 1100 operating across the country. Forty-six percent serve high school students, 16% serve middle school students, 28% serve elementary school children, and 10% serve some combination.³ In addition to their broad design, these clinics are often serving a large number of uninsured children. One report showed that nearly 40% of all school-based and school-linked health center users are completely uninsured; at a California SBHC, 93% of clinic enrollees reported no other source of medical care.⁴ Children without insurance are less likely to receive any ambulatory care, even when medically indicated, than their privately insured counterparts.⁵

WESBHC in Atlanta, Georgia is an expanded model of an SBHC serving preschool and elementary school-aged children. It is a joint effort of the public schools, Emory University School of Medicine, and several other community, state, and private organizations. Its overall goal is to integrate relevant programs to serve the children of Whitefoord Elementary School and their preschool siblings in such a manner that their overall health and well being are sustained and/or improved. Ultimately, the program is aimed at improving school attendance and classroom performance and the longer-term prospects for these children as they mature.

This SBHC was organized by the Department of Pediatrics, Emory University School of Medicine and is located within the Whitefoord Elementary school. The clinic was initially opened in November of 1994; the clinic services were expanded to include preschool children (0-4-year-olds) in July of 1996. By its immediate proximity to the child during the day, the WESBHC removes major access barriers to primary care (eg, transportation time and costs, lost work time, etc) that other providers cannot. The child can immediately enter the health care system because the WESBHC is located within the school walls. It is open all year and does have after hour (24 hours) phone triage by a nurse with physician backup. Further, the WESBHC strives to promote wellness through an interdisciplinary approach and to provide education to help parents take responsibility for their child's health and health care. The key services provided by the clinic include: 1) diagnostic and treatment for acute illnesses and injuries, 2) management of chronic illnesses, 3) preventive care and screening, 4) mental health screenings and management, 5) dental care, 6) health promotion/education, 7) social services, and 8) referrals to medical subspecialties and community agencies.

While there have been earlier evaluations of the effects of SBHCS, these are relatively few in number, lack comprehensiveness, and have sometimes been based on simulated data rather than the actual experience of those served by a SBHC. For example, earlier studies have concluded that school nurse practitioner services and school-based high school health services can result in savings for the schools and families.⁶ This study found lower costs for parent and child of a visit to a SBHC compared with a physician's office. Further, service costs are believed to be lower for some services at a SBHC because of the greater substitution of nursing for physician time. There is also some evidence that their outcomes are as good or better than those achieved by physician's primary care.⁷ One study on the cost-effectiveness of school nurse practitioner services reported that multiple case histories involving medical problems could be cited in which school nurse practitioners had averted unnecessary adaptive education placements while freeing placement slots for students in real need of those programs.⁸

Another earlier study focused on outcomes and found that comprehensive elementary school health education programs increased the health knowledge of younger groups but not older ones.⁹ Still other studies have asked more comprehensive questions about the costs and benefits of such programs, while not carrying out a specific evaluation.¹⁰ Most studies have not included all costs affected by school clinics or chosen appropriate comparison groups for evaluating effects. None, to our knowledge, have focused on Medicaid program savings.

	METHODS

Top Abstract Methods Results Discussion References

The overall goal of this analysis is to evaluate and compare the Medicaid health care costs per child for those children whose primary caregiver is the WESBHC to children in an area without a SBHC. In completing this analysis, we are taking the perspective of Medicaid, a publicly financed program designed to care for the poor and recently expanded to cover more children at or near poverty. To achieve this overall goal, we ask:

• Have the types of services used (eg, preventive care, ER, etc) for those whose primary care provider is the WESBHC changed over time? How does this change compare to the change in utilization over time for a similar group of children whose primary care provider is not a SBHC?
• Are Medicaid health care expenses lower for those whose primary source of care is the WESBHC before and after the start of the clinic and/or compared with those whose primary care is not a school-based clinic?

Our major underlying hypothesis is that the increased access to primary care provided by the WESBHC will reduce children's emergency department use as well as maintain the child's health so that hospitalizations are avoided and Medicaid expenses are lowered.

Our overall design involves using Medicaid claims data to measure the health care use and expenses of children served by the WESBHC in 1995, compared with those used by them in 1994 and to those used by children who do not have access to a SBHC. The Medicaid program is designed to fund, rather than directly deliver care and hence patterns of delivery will vary across children served in different areas. We use data for periods before and after (1994-1996) the implementation of the clinic's service. These data include information on all outcomes of interest (eg, use of emergency room, hospitalizations, etc) for our analysis.

The 1994 period is largely before the implementation of the WESBHC and provides a baseline or control period for the Whitefoord and comparison children. We assembled data from the Georgia Department of Medical Assistance (DMA) or Georgia Medicaid program for children served by the Whitefoord clinic and for a comparison group of children over the 1994-1996 period. We are only able to study the use/expense of either group while they are enrolled in Medicaid. Because these children are in high poverty areas, they are likely to qualify for Medicaid coverage. Indeed, over 90% of the Whitefoord children were enrolled in Medicaid sometime during this period.

The Medicaid identifications of the children enrolled (having parental approval for care) and/or using the Whitefoord SBHC were obtained from the clinic, while Medicaid identifications for the comparison group were drawn from the DMA files. Using these identifications, we matched the WESBHC children to the Medicaid enrollment files; of the 696 listed, between 598 and 646 were enrolled in Medicaid at some time during the 1994-1996 period. Enrollment and claims histories were pulled for the Whitefoord and comparison children for the 1994, 1995, and 1996 periods. The claims history provides all outpatient and inpatient utilization by date of service and includes details on diagnoses (International Classification of Diseases, 9th Revision [ICD-9] coding), procedures (Current Procedural Terminology-4 and state-specific), amounts billed/paid, provider type, place of service, category of service, etc. The Medicaid enrollment files provide data on age, sex, race, residential zip, and Medicaid eligibility group. We constructed monthly enrollment and utilization records for each child.

We then drew a sample of children believed to be similar in their sociodemographic background to those with access to the WESBHC, enrolled in Medicaid, and without access to a SBHC. To do this, we drew from a geographic area with similar household characteristics (eg, income and family education) that generally affect a child's health status and utilization by using a set of 5-digit zip codes. These zip codes comprise the school district for a nearby school, East Lake Elementary, which has no SBHC program. The 1990 census data indicated that 63% of the children in the East Lake Elementary census tracts were in poverty, whereas 68% of those served by Whitefoord Elementary were in poverty; from 92% to 93% of each group were in the school lunch program.

Children found on the Medicaid enrollment files anytime during 1994, 1995, or 1996 and residing in the zip codes served by East Lake Elementary comprised our comparison group. The total number of children enrolled in Medicaid residing in the East Lake Elementary zip codes during both 1994 and 1995 or 19941996 and in the same age group (4-12 years old in 1994) totaled a little <500 in some years. This sample size more than satisfied our criteria for detecting 30% to 40% differences in emergency department use based on an  of .05 and power of 80.

As noted, we used all those children whose participation in the clinic was approved by their parents as the initial cases for our analysis. Most of these children are actual users of the clinic during the year but some are not. We were, therefore, interested in identifying the subset of Whitefoord children for whom the Whitefoord clinic is their usual provider of primary care. To define usual providers, we first arrayed the claims data associated with primary care. We used physician visit codes, all Early and Periodic Screening, Diagnostic and Treatment (EPSDT) claims, all immunization claims, and all those claims with category of service equal to nurse practitioner as primary care. In addition, all claims with a V diagnostic code were included to capture all well-child visits in this definition.

For each child, we identified the (unique) provider identification on most claims of the above types. A tie between 2 providers was broken by using the provider with the most dollars paid. We expect to find stronger evidence of our hypothesized effects, such as reduced emergency department use, for WESBHC children who use the clinic as their medical home. Although the comparison group children generally had physicians as their usual provider, followed by public health departments, these entities cannot reduce access barriers in the same manner as the WESBHC.

In the above process, we found that a large number of the Whitefoord clinic claims for EPSDT and other primary care services had not been billed properly during late 1996 and the volume of the Whitefoord clinic was understated. The 1996 data for these categories of service overstate our case, because the Whitefoord children look less expensive because of the under-billing. Therefore, we focus on the identification of the usual provider during the 1995 period and although we present 1996 data for all categories of service, we focus on those not affected by the billing problem, such as emergency room and inpatient services. Finally, we were also interested in the effects of the Whitefoord clinic for children with a chronic and costly condition, such as asthma. Additional analyses are shown for children with this diagnosis. We identified asthmatic children as those with an ICD-9 diagnosis code of 493 anytime during the year.

Our primary statistical methods include testing for differences in means for the Whitefoord and comparison sample before and after the clinic opening. We also use multivariate analysis. In the multivariate analysis, we derive a measure of the differences in differences for the 2 groups and test its significance. That is, we derive a measure of the difference between 1994 and 1995 for Whitefoord and comparison children's groups and then compare this difference across the 2 groups. The coefficients on the interaction of a dummy variable denoting membership in the Whitefoord group and the time variable (1994-1995) are used to derive this measure.

Specifically, the regression coefficients are as follows:

[($_{t = post,D = W}  $_{t = pre,D = W})  ($_{t post,D = C} $_{t = pre,D = C})]/X,

where

$_{t =  post,D = W} is the 1995 or post period expense per Whitefoord child;

$_{t = pre,D = W} is the 1994 or pre period expense per Whitefoord child;

$_{t = post,D = C} is the 1995 or post period expense per comparison child;

$_{t = pret,D = C} is the 1995 or pre period expense per comparison child; and

X =  a vector of control variables in the form of characteristics of the child and a time dummy variable.

Throughout the descriptive and multivariate analyses, the Medicaid expense data are adjusted for the relative length of Medicaid enrollment of each child in the WESBHC and comparison groups.

	RESULTS

Top Abstract Methods Results Discussion References

Overall, the results indicate that the introduction of the WESBHC did alter the utilization and expense patterns of children enrolled in Medicaid and residing in the area served by the WESBHC and program.

Descriptive Data

In Table 1 the average Medicaid amounts paid in 1994 and 1995 for children who were enrolled in Medicaid in 1994 and 1995 and in the comparison zip code or the WESBHC during 1995 are shown. In the third column, we provide the ratio of the Medicaid expenses for the comparison group to those for the WESBHC children and in the last column we provide the value of the t test on the difference in these mean Medicaid amounts. We note that these data are for children who were 4 to 12 years old in 1994, enrolled in Medicaid, and users of some services in the year. The children identified as Whitefoord in Table 1 are those who used the clinic at least once during 1995. Although the comparison group does not include children enrolled in Medicaid and never using services during the year, this is the appropriate comparison group because the Whitefoord children are users by definition. The Medicaid expenses of both groups are adjusted for the length of time they were enrolled during the year.

As these data show, the average amount paid by Georgia Medicaid for health care for the WESBHC children in 1994, before the implementation of the clinic, was lower than for the comparison children. Medicaid expenses for the WESBHC children equaled $1742, whereas for the comparison children, the average was $1772. However, this difference was not statistically significant, and there were no statistical differences in the mean expenses of the 2 groups in any single service category in this pre-period.

The second bank of data in Table 1 shows that although Medicaid expenses declined for both groups of children, they declined more for the WESBHC children. The decline from 1994 to 1995 was $536 for the Whitefoord children versus $277 for the comparison children. Much of the decrease in Medicaid expenses per child in both groups was in nonemergency transportation, a category of service in which the Georgia DMA sought cost savings during these years. The substantially greater decline in non-emergency expenses for the Whitefoord versus comparison children is likely attributable to better access to primary care services provided by the clinic after 1994.

These results also show 2 findings important to our basic hypothesis. First, the WESBHC children have higher expenses for EPSDT preventive care services in 1995 than they did in 1994 and higher than the comparison children in 1995; this difference is statistically significant (P = .01). Second, their emergency department expenses are lower than those of the comparison group by 1995, and this difference is also statistically significant (P = .10).

We also show in Table 1 the 1996 inpatient, nonemergency transport, drug, and emergency department expenses per child for the Whitefoord (user of Whitefoord in 1995) and comparison groups. We note that the comparison group of children is markedly lower in number in 1996 attributable to: 1) disenrollment of the child from the Medicaid program; 2) movement out of the zip code; or 3) nonuse of any Medicaid service in 1996. The Medicaid amounts presented in Table 1 are generally lower in 1996 and the differences are statistically significant at either the .01 or .05 level. These data indicate that the WESBHC users cost the Georgia Medicaid program only $197 per child-year enrolled in 1996 inpatient care expenses, whereas the comparison group cost the program $749. The difference in emergency department expenses is also significant. Recall that the 1996 EPSDT and physician primary care data are understated because of missing claims for the Whitefoord clinic.

Similar types of data are shown in Table 2. The children listed as Whitefoord children in this table, however, are those for whom the clinic is the usual provider of primary care during 1995. We expect stronger effects for this subset of the Whitefoord children examined in Table 1. Although this classification is perhaps a better measure of the impact of the clinic on the child's health care utilization, we note that the children so identified are smaller in number and that there may be some overstatement of the effect of the clinic on emergency department use. Our identification of the WESBHC as the usual provider means it provides the most primary care services to the child during 1995. By definition, then, their use of a non-WESBHC physician in the emergency room for primary care has to be smaller than their use of the WESBHC. Still, the child's use of the emergency room for any services in 1995 could be higher or lower than that use in 1994 and it is this change that is the focus of the analysis.

These data also show a statistically significant difference in the total expenses of the 2 groups over the 1994-1995 period and significant changes in their utilization patterns. Medicaid expenses per child drops from $1797 in 1994 for the WESBHC usual provider group to only $901 in 1995; total Medicaid expenses are significantly lower (P = .01) for the WESBHC than for the comparison children whose expenses are almost $1494 in 1995. Underlying this overall change for the Whitefoord children is a drop in Medicaid expenses for all categories of service except mental health and EPSDT. Moreover, although the only statistically significant difference between the 2 groups in 1994 was for other services, by 1995, the WESBHC usual provider group has significantly lower nonemergency transportation, physician, and emergency room expenses paid by Medicaid. The lower emergency room and inpatient expense also hold in 1996.

The equality of expenses between the 2 groups of children in 1994 reflects, in part, a comparability of their underlying characteristics as we had desired. The mean age in 1994 was just over 6 years old for both groups and both groups were predominantly black (97%-98%). Children in both groups were enrolled in Medicaid ~11.5 months of the year. The children were also comparable in terms of their distribution across eligibility groups within the Medicaid program in 1994. We do note, however, that the comparison group of children was comprised of relatively more disabled children by 1995; 3.5% of the comparison group fell into this eligibility category as opposed to 1.2% of the Whitefoord (usual provider) children in 1995. Given that disabled children have more chronic and high-cost medical conditions, this would make the comparison group of users more expensive. Yet, we would not expect them to have greater use of the emergency room unless their care is not well managed. This points to the need to examine the impact of the WESBHC in a multivariate context, holding constant the percentage of disabled children.

Regression Analysis

In Table 3, we show the results of multivariate regression analysis of the total, emergency department, and inpatient expenses per child-year enrolled for the pooled 1994 and 1995 observations for the WESBHC (usual provider) and comparison children. By using multivariate analysis, we are able to hold constant certain characteristics of the child that may affect their use of Medicaid services and expenses. We include the child's age, sex, race, and Medicaid eligibility category as covariates; the means and standard deviations of these variables for the pooled 1994 and 1995 samples are shown in Table 4.

The results in Table 3 are largely consistent with the descriptive analyses, although there is not generally statistical significance on the Whitefoord effect. Although the coefficient on the time variable (TIME) indicates an overall downward trend in total Medicaid expenses in the first equation, it is not significant. Being in the WESBHC usual provider group (WHITEPROV) shows no significant effect for total Medicaid nor emergency department expenses. The coefficient of most interest is that on the time and Whitefoord group interaction (TIME * WHITEPROV); this coefficient, as noted in our "Methods" section, measures the differences in differences, or the difference between the Whitefoord and comparison children in the change in Medicaid expenses over the pre and post period. This coefficient is negative and significant (P = .10) in the emergency department expense equation.

Other results shown in Table 3 are consistent with expectations but not generally significant. Expenses are generally higher for younger (4-7 years old in 1994 [AGE 1]) children relative to older (8-12 years old in 1994) children. Younger children have greater periodicity of preventive care and also tend to have more acute care illnesses than older children. Although there is a negative coefficient on the race variable (NONWHITE), there is so little variation in this variable this result is hard to interpret. The findings with regard to Medicaid eligibility group are also not surprising. They indicate that the nondisabled children's groups, those related to the Aid to Families With Dependent Children (AFDC) or in the newer poverty-related expansion groups (POVREL) are generally less expensive than children in the (omitted) disabled child category, as expected.

Asthmatic Children

Asthma is a chronic condition that affects a large and growing number of minority children across the country. It is a particular problem for inner-city youth and numerous efforts are being made to better manage the clinical manifestations and costs of this condition. We chose to look further at this condition by subsetting the children based on evidence that they suffered from asthma. To do so, we flagged children with an ICD-9 code of 493 on any inpatient or outpatient claim during the study. As the data in Table 5 show, there were relatively few of these children in the WESBHC and comparison groups in the 1994-1996 period and the percentage of children affected was similar.

The descriptive data on asthmatic children (see Table 5) again indicate there are not significant differences for the Whitefoord and comparison children in the period before 1994. Total Medicaid expenses per asthmatic child-year enrolled are higher, as expected, than the average (Table 1), ranging from $2373 to $2414 for the WESBHC and comparison groups in 1994. By 1995, the expenses of those asthmatic children served by the WESBHC were lower, equal to $1758, whereas expenses per asthmatic child for those in the comparison group had increased to $2541. Although neither the 1994 nor 1995 total Medicaid expense were significantly different across the 2 groups, there was a statistically significant difference in inpatient hospital expenses by 1995. In 1995, the WESBHC asthmatic cost the Medicaid program an average of $352 in inpatient expenses, the comparison group cost over 3 times that or $1259. Although we cannot control for possible differences in asthma severity, the results do indicate the pattern of care changed from the 1994 to the 1995 period to lower costs for the Whitefoord asthmatics relative to the comparison group asthmatics. The 1996 data also indicate inpatient and drug expenses for the Whitefoord asthmatic children were significantly lower.

	DISCUSSION

Top Abstract Methods Results Discussion References

The forgoing data and analysis indicate that the placement and operation of the WESBHC in the Whitefoord Elementary community had an impact on Medicaid expenses for the state of Georgia. A major effect of the clinic was the reduction in the probability that children use the emergency department, and hence, emergency room expenses. The data also indicate the clinic enhanced the downward trend in total Medicaid expenses taking place for children in these inner-city Atlanta communities over the 1994-1995 period. The overall downward trend likely reflects the continued implementation of primary care case management within Georgia's Medicaid program and the implementation of controls on nonemergency transportation expenses over this period. The WESBHC enhanced this overall trend by lowering emergency department expenses.

Another important effect of the clinic was an increase in the EPSDT expenses per child. It seems that the WESBHC, by increasing access, had the expected effect on increased primary care. This is consistent with the belief of the WESBHC that they were able to shape the parent's perceptions of the importance of primary care and having a medical home for their child. Staff members at the WESBHC report spending significant time educating and reenforcing these concepts for the parents in what are often dysfunctional homes in these poor neighborhoods. If the parent does come to see the clinic as the child's medical home, this allows the WESBHC to shape the parents' response to an illness and better control the child's overall care.

These are important and encouraging findings for the potential of SBHCs to save money while improving the use of preventive care. We note, however, that we have only focused here on the costs of services and clinic revenues for those who are Medicaid enrolled. Hence, we have not looked at the total costs of running a SBHC such as the WESBHC. This clinic, as well as others, must cover their total costs through Medicaid revenues plus subsidies from the taxpayers of the federal and/or state governments.

Although our findings indicate that public or private investment in a SBHC has the potential to produce savings for public programs such as Medicaid, we are not able to analyze total costs for either group or compare total costs when a clinic is the usual provider to that of an office-based physician. We do note, however, that there is significant evidence of a lack of physician practices in inner-urban areas,^12,13 and hence, the alternative source of care for these children is likely to be another subsidized clinic rather than a physician's office. From a societal perspective, the private subsidization of the Whitefoord clinic might reflect a cost-shift from the public to private sector if these children would have otherwise been served in a public clinic.

Just as we have not examined all costs, we also have not examined full benefits of the WESBHC. For example, if students and their families are assured better access to health care services, this can reduce travel time, parent's lost time at work, and increase the child's school attendance. If conditions such as asthma are better managed/controlled, the student may actually be able to achieve more in the time they are attending school. Beyond this, there are other less readily measured benefits to the child, family, and community that likely accrue with the presence of a program such as the Whitefoord Community program and WESBHC.

The findings here also have relevance to broader health policy issues. In 1989, Congress sought to achieve the improvement of children's participation in the Medicaid EPSDT program through the Omnibus Budget Reconciliation Act of 1989 in a number of ways,¹⁴ including increased provider participation. An analysis of EPSDT before and after the implementation of the Omnibus Budget Reconciliation Act of 1989 found increases in the Medicaid child case-loads of clinic providers inclusive of school-based clinics.¹⁴ Because school attendance is mandatory and SBHCs do not typically turn anyone away, they serve as a ready source of access for children's primary and preventive care.

Although states may, therefore, see SBHCs as an avenue through which to increase access and participation in the EPSDT program, they will need to carefully maintain them as a source of care if they are expanding enrollment in capitated managed care. As noted, the WESBHC and others like it are at least partially dependent on Medicaid as a source of revenue. However, these clinics will not be reimbursed for services provided to Medicaid children unless they become part of the provider network of health maintenance organizations (HMOs). They are often not set up as billing entities, and hence, may need to become affiliated with others to enroll in networks. Unless states carve out the types of services provided by SBHCs or require HMOs to include these providers in their network, SBHCs may see diminished Medicaid revenues without diminished children's needs, causing increased financial stress. A more constructive policy might integrate the SBHC into the HMO network of providers to pay for its services and to allow for the exchange of clinical information for the better management of the child's health and well being. We recognize, however, that HMOs will not likely be able to pay the full costs of these clinics and that subsidizations from taxpayers or other sources will likely be necessary to cover their full costs.

	FOOTNOTES

Received for publication Mar 22, 1999; accepted Oct 5, 1999.

Reprint requests to (E.K.A.) Rollins School of Public Health, 1518 Clifton Rd NE, Atlanta, GA 30322. E-mail: eadam01{at}sph.emory.edu

	ABBREVIATIONS

WESBHC, Whitefoord Elementary School-Based Health Clinic; SBHC, school-based health clinic; EPSDT, Early and Periodic Screening, Diagnostic and Treatment; DMA, Department of Medical Assistance; ICD-9, International Classification of Diseases, 9th Revision; TIME, coefficient on the time variable; WHITEPROV, WESBHC usual provider group; TIME * WHITEPROV, coefficient of the time and Whitefoord group interaction; AGE 1, 4 to 7 years old in 1994; NONWHITE, coefficient on the race variable; AFDC, Aid to Families With Dependent Children; POVREL, newer poverty-related expansion groups; HMO, health maintenance organization.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Hacker K, Fried LE, Bablouzian L, Roeber J A nationwide survey of school health services delivery in urban schools. J Sch Health 1994; 64:279-283 [Medline]
2. Meeker RJ, DeAngelis C, Berman B, Freeman HE, Oda D A comprehensive school health initiative. Image J Nurs Sch 1986; 18:86-91 [Medline]
3. School Health Resources Services. School-Based Health Centers: The Facts: 1995 Resource Packet Series. Denver, CO: University of Colorado Health Sciences Center; 1995
4. Health Resources and Services Administration, Bureau of Primary Health Care. School-based Clinics That Work. Washington DC: Division of Programs for Special Populations; Health Resources and Services Administration 93-248P
5. Weinick RM, Weigers ME, JW Cohen Children's health insurance, access to care and health status: new findings. Health Aff 1998; 17:127-136 [Abstract]
6. Siegel LP, Krieble TA Evaluation of school-based, high school health services. J Sch Health 1987; 57:323-325 [Medline]
7. Igoe JB, Giordano BP. Expanding School Health Services to Serve Families in the 21st Century. Denver, CO: School Health Resources Services, University of Colorado Health Sciences Center; 1995
8. Sobolewski SD Cost-effective school nurse practitioner services. J Sch Health 1981; 51:585-588 [Medline]
9. DuShaw ML A comparative study of three model comprehensive elementary school health education programs. J Sch Health 1984; 54:397-400 [Medline]
10. Newman IM, Newman E, Martin GL School health services: what costs? What benefits? J Sch Health 1981; 51:423-426 [Medline]
11. Hurley RE, Freund DA, Taylor DE Emergency room use and primary care case management: evidence from four Medicaid demonstration programs. Am J Public Health 1989; 79:843-847 [Abstract/Free Full Text]
12. Fossett, JW, Perloff JD, Peterson JA, Kletke PR Medicaid in the inner city: the case of maternity care in Chicago. Milbank Q 1990; 68:111-139
13. Goldstein A. Many doctors in few places: where doctors work, health care in Washington. The Washington Post. July 31, 1994. Healthcare in Washington section
14. Gavin NI, Adams EK, Herz EJ, The use of EPSDT and other health care services by children enrolled in Medicaid: the impact of OBRA89. Milbank Q 1998; 76:207-250 [CrossRef][Medline]