Abstract

We examined the association between state Medicaid spending for children with disabilities and the financial burden reported by families of children with autism. Child and family data were from the 2005–2006 National Survey of Children with Special Health Care Needs (n  =  2,011 insured children with autism). State characteristics were from public sources. The 4 outcomes included any out-of-pocket health care expenditures during the past year, expenditure amount, expenditures as a proportion of family income, and whether additional income was needed to care for a child. We modeled the association between state per capita Medicaid spending for children with disabilities and families' financial burden, controlling for child, family, and state characteristics. Overall, 78% of families raising children with autism had health care expenditures for their child for the prior 12 months; 42% reported expenditures over $500, with 34% spending over 3% of their income. Families living in states with higher per capita Medicaid spending for children with disabilities were significantly less likely to report financial burden. There is a robust relationship between state Medicaid spending for children with disabilities and the financial burdens incurred by families raising children with autism.

Most children with autism need a range of often expensive medical and therapeutic services, including early intervention and intensive treatments (Dawson & Osterling, 1997; Hurth, Shaw, Izeman, Whaley, & Rogers, 1999; Thomas, Ellis, McLaurin, Daniels, & Morrissey, 2007; Thomas, Morrissey, & McLaurin, 2007). Children with autism typically receive services from both school and medical settings, which can obscure coverage responsibility (Autism Speaks, 2007). Moreover, typical private behavioral health coverage for children often excludes autism altogether (Newacheck & Kim, 2005; Peele, Lave, & Kelleher, 2002). Medicaid, therefore, provides important coverage for families in this situation (Chiri & Warfield, 2012; Fox, McManus, & Limb, 2000; Limb, McManus, & Fox, 2001; Ruble, Heflinger, Renfrew, & Saunders, 2005; Young, Ruble, & McGrew, 2009). Nationally, approximately half of children with autism had Medicaid coverage, underscoring the fundamental importance of the program for these children (Chiri & Warfield, 2012). However, the extent to which families' costs of care are associated with Medicaid spending have not been examined.

The societal costs of autism are exceptionally high, estimated at $3.2 million per lifetime (Ganz, 2007). Although this estimate aggregates direct and indirect costs to society, little research has examined the out-of-pocket costs families incur to raise their children with autism. One study found that families of children with autism incurred annual costs that averaged 14% of total household income (Montes & Haterman, 2008). There is no research that has examined whether the direct costs families incur for their children with autism are related with state Medicaid spending.

Few studies have examined the financial burden incurred by families raising children with autism (Young et al., 2009), but this issue has been examined for the larger population of children with special health care needs, which includes the subpopulation of children with autism. For children with special health care needs, greater financial burden is associated with parental work loss and other negative outcomes, particularly among low-income families (Honberg, Kogan, Allen, Strickland, & Newacheck, 2009; Oakumura, VanCleave, Gnanasekaran, & Houtrow, 2009). Families of children with special health care needs living in states with higher Medicaid income eligibility thresholds had lower financial burden, but the relationship between Medicaid and family costs for their children with autism has not been examined (Parish, Shattuck, & Rose, 2009).

Advocates and policymakers are working to expand health care access for children with autism. Much of this activity has occurred at the state level, including the adoption of Medicaid waivers specifically targeted to children with autism. Researchers face a daunting task in trying to understand the impact of state policy on children with autism and their families, for the simple reason that there is little data. Currently, the only existing survey that contains an adequate representation of children with autism in each state (and thus facilitates the modeling of state-level estimates) is the 2005–2006 National Survey of Children with Special Health Care Needs. Other national data sets (e.g., the National Survey of Child Health) lack adequate samples of children in each state, let alone of children with autism, to understand the effects of state policy and program characteristics.

Given the absence of a national autism survey, current insights into autism service use have been developed from local or, at best, multistate data. Recent analyses of the National Survey of Children with Special Health Care Needs indicates that families of children with autism are more likely to have problems accessing care, unmet need for care, and greater financial burden than families with children with other special health care needs (Kogan et al., 2008). Our study combined family- and child-level data reported by parents of children with autism from the National Survey of Children with Special Health Care Needs with state Medicaid data to capitalize on the unique resources of the National Survey of Children with Special Health Care Needs. This innovative combination of data supports multilevel analysis of the association between state Medicaid policy and the financial burden for families of children with autism.

Theoretical Model and Hypotheses

The theoretical model that informs this study is the Behavioral Model for Vulnerable Populations (Gelberg, Andersen, & Leake, 2000), which is a revision of a leading framework in public health research, the behavioral model (Andersen, 1995). This expanded model addresses the particular needs of subgroups at elevated risk for adverse outcomes, which we include to mean individuals with autism. This model states that the interplay of cross-system elements, such as individual and family factors, community factors, and policy factors determine health care access. In this model, states' Medicaid spending is posited to be an enabling resource that facilitates obtaining health care services (Gelberg et al.).

We examined how the financial burden experienced by families raising children with autism was associated with the level of state Medicaid spending for children with disabilities. We hypothesized that families of children with autism living in states with greater levels of Medicaid spending for children with disabilities would report less financial burden, after controlling for family and child characteristics and state wealth. We hypothesize that financial burden, both absolute and relative to family income, will be lowest for children with autism who live in states with the greatest levels of Medicaid spending for children with disabilities.

Method

Data Sources

Child and family data were drawn from the 2005–2006 wave of the National Survey of Children with Special Health Care Needs, a telephone survey described elsewhere (Blumberg et al., 2008; van Dyck, Kogan, McPherson, Weissman, & Newacheck, 2004). The random-digit-dialed sampling strategy used a five-question sequence to determine if a child with special health care needs lived in the household. At least 750 telephone interviews with families of children with special health care needs in every state and the District of Columbia provided representative state population estimates. Survey respondents were the parent or guardian who was most knowledgeable about the child's health care. The final sample included 2,123 special needs interviews with parents of children with autism (U.S. Department of Health and Human Services, Health Resources and Services Administration, & Maternal and Child Health Bureau, 2007). Children in the National Survey of Children with Special Health Care Needs were identified in the analytic sample as children with autism if their parent or guardian responded affirmatively to the question, “To the best of your knowledge, does [your child] currently have Autism or Autism Spectrum Disorder, that is, ASD?”

States' Medicaid expenditures were drawn from publicly available data from the Centers for Medicaid and Medicare Services (U.S. Department of Health and Human Services, n.d.) for total state spending for children with disabilities, which included an array of different services in 2005. State median income values in 2006 for families with children were obtained from the Annie E. Casey Foundation (2008). State median income was included as an indicator of state wealth (Annie E. Casey Foundation, 2008; Carroll & Zhou, 2010), which is related to the proportion of a state's population with chronic health conditions, as well as an indicator of available resources that can be devoted to social welfare spending (Annie E. Casey Foundation, 2008). The proportion of people in each state living in nonmetropolitan areas was obtained from the U.S. Census, to adjust for barriers to care faced by rural families (U.S. Census Bureau, Population Division, 2006, 2010).

Measures

Dependent variables

The National Survey of Children with Special Health Care Needs contained questions asking families to report out-of-pocket expenses associated with their child's medical care over the 12 months prior to their interview: $0, $1–$249, $250–$500, $501–$999, $1,000–$5,000, and $5,001 or more. Eligible expenses that counted as out-of-pocket costs included a variety of health-related needs including co-payments, medications, special foods, and durable equipment. Insurance premiums and reimbursable costs were excluded.

Four dependent variables of financial burden were developed from variables in the National Survey of Children with Special Health Care Needs: (a) a measure identifying families who incurred any out-of-pocket medical expenses, (b) an indicator of the level of expenses or absolute burden (less than $250, $250–$500, and greater than or equal to $500), (c) an indicator of the level of out-of-pocket spending as a proportion of family income or relative burden (less than 1%, 1%–3%, and greater than 3%), and (d) family-needed additional income to care for the child. Family income was imputed through an algorithm reported elsewhere (Parish, Shattuck, & Rose, 2009).

Independent variable

In the regressions, the total state Medicaid spending for children with disabilities was entered in $100s per capita (of state population) in 2005, centered at its mean.

Individual covariates

Individual and family covariates obtained from the National Survey of Children with Special Health Care Needs included family income relative to the federal poverty level (less than 200%, as compared to 200% or higher); race/ethnicity (minority, which included Black, Hispanic, Native American and Aleutian, Pacific Islander, Asian and multiracial, as compared to White); parental rating of the severity of the child's condition (severe as compared to moderate and mild); and whether the child had public health insurance (relative to having only private insurance).

State covariates

Median state income for families with children in 2005 was measured in $10,000s (Annie E. Casey Foundation, 2008). The percentage of families living in nonmetropolitan areas, as defined by the U.S. Census Bureau, was coded so that a one-unit change in this variable corresponded to a 1% change in percentage living in nonmetropolitan areas. Both variables were centered at their means. At the state level, the number of covariates was limited by the small “sample” of states available. That is, all states (n  =  50) were included, but in multilevel modeling, the number of covariates at the second (state, in this case) level was constrained by the number of units at the second level (Raudenbush & Bryk, 2002).

Analysis Sample

Starting with 2,123 children with autism, we reduced the sample further due to the following conditions: First, we discarded data on children not having health insurance, which eliminated the records for 79 children. Second, we removed data on 33 children living in the state of Connecticut. The data for Connecticut were found to be unreliable, with per capita Medicaid spending reported as unrealistically low. The final analysis sample consisted of 2,011 children having either private or public insurance living in the District of Columbia and any state (except Connecticut; see Table 1). Sensitivity analyses that included Connecticut did not substantively change these findings.

Table 1

Description of the Sample

Description of the Sample
Description of the Sample

Analytic Strategy

Logistic regression models were needed to handle the binary outcomes. Further, multilevel modeling was needed to account for the nesting of children within states and to properly estimate the standard errors of the Medicaid expenditure variable. To accommodate both, hierarchical generalized linear model (HGLM) was used for all four dependent variables (Raudenbush & Bryk, 2002). For the analysis of variables related to out-of-pocket expenses, a two-part model was used. We first estimated the effect of Medicaid expenditures on whether or not families with children with autism had any out-of-pocket costs. Subsequently, among families reporting that they had out-of-pocket expenses, we estimated the effect of Medicaid expenditures on how burdensome out-of-pocket costs were in absolute dollars and relative to family income (Blumberg, personal communication, 2007). The coefficient estimates are log-odds that can be transformed into odds ratios via exponentiation as in standard logistic regression.

Missing data

Multiple imputation with SAS was employed for missing data, with Mplus handling post-estimation procedures (Graham, 2009; Graham, Olchowski, & Gilreath, 2007; Schaefer, 1997; Rose & Fraser, 2008; von Hippel, 2007). There were also valid nonresponses on the dependent variables that were not imputed.

Weighting and variance adjustment

Mplus handled the complex sampling weights of the stratified random sample according to the specification in the National Survey of Children with Special Health Care Needs, but neither Mplus nor any multilevel software adjusted the variance for the complex sample (Carle, 2009). Our simulations showed that variance adjustments for multilevel data and complex survey data resulted in similar standard errors for individual-level covariates, while the complex survey variance adjustments insufficiently corrected the standard errors for state-level variables. Because the state-level variables were more salient to this investigation than the individual-level variables, we used the multilevel data analysis.

Results

Table 1 presents the unadjusted description of the sample. Table 2 presents the percentage of families of insured children with autism in each state with any out-of-pocket spending, total spending greater than $500 for the year, and total spending exceeding 3% of household income for the year, respectively. The rankings of the states in the percentage of families with spending greater than $500, and 3% of household income are also presented. The states are listed by the ranking on relative burden or the percentage of families with spending greater than 3% of household income. As is evident from Table 2, there was significant variability in the percentage of families with spending greater than 3%. This ranged from a low of 11% of insured children in Colorado and Rhode Island to a high of 44% in Utah and Alabama.

Table 2

Per Capita Medicaid Spending for Disabled Children and Percentage of Families of Insured Children with Autism with Financial Burden by State, State Rankings

Per Capita Medicaid Spending for Disabled Children and Percentage of Families of Insured Children with Autism with Financial Burden by State, State Rankings
Per Capita Medicaid Spending for Disabled Children and Percentage of Families of Insured Children with Autism with Financial Burden by State, State Rankings
Table 2

Continued

Continued
Continued

The results of the multilevel regressions are presented in Table 3. Higher per capita Medicaid expenditures for children with disabilities are associated with reduced odds of several measures of financial burden among families of children with autism. A difference of $100 in per capita Medicaid expenditures was associated with 43% lower odds of having any out-of-pocket costs (p < .01); 48% lower odds of having expenses between $250 and $500; and 60% lower odds of having expenses over $500 (both relative to under $250, p < .001); 42% lower odds of having expenses constituting 3% of income or greater (relative to less than 1% of income, p < .001); 29% lower odds of needing additional income to care for the child (p < .01).

Table 3

Multilevel Regression Results Predicting Annual Measures of Family Financial Burden

Multilevel Regression Results Predicting Annual Measures of Family Financial Burden
Multilevel Regression Results Predicting Annual Measures of Family Financial Burden

Discussion

Corroborating findings of financial burden among families of children with special health care needs reported elsewhere, we found significant state-level variability in the absolute and relative financial burden reported by families of children with autism (Shattuck & Parish, 2008). This variability persisted after controlling for state wealth; population living in nonmetropolitan areas; characteristics of the family and of the individual child, including impairment severity; and whether or not the child was insured. This study's important contribution is, after controlling for these covariates, its robust and consistent finding that families who lived in states with greater levels of Medicaid spending for children with disabilities incurred fewer expenses, both in absolute (actual dollar amounts) and relative (to family income) terms for their children with autism. Further, these results indicated that a substantial amount of state variability was associated with per capita state Medicaid spending for children with disabilities. Although these results should not be interpreted as evidence of a causal relationship, the findings offered preliminary evidence that underscored the critical role of Medicaid as a safety-net insurer for families of children with autism, particularly if families have depleted financial resources in their efforts to obtain services for their child.

The significance of the relationship between financial burden and levels of state Medicaid spending for children with disabilities is troubling in the context of the current economic situation. The so-called Great Recession has ended, but the National Association of State Budget Officers (2011), among others, continues to predict a bleak outlook for state budgets. Slow economic growth and continued high levels of unemployment create more demand for Medicaid services at the same time as they reduce state tax revenues, which in turn constrains states' abilities to fund their Medicaid programs. The $87 billion in federal revenue pumped into state Medicaid programs via the American Reinvestment and Recovery Act served as an important resource for states (National Association of State Budget Officers, 2011). However, this revenue stream ended in fiscal year 2011, thus increasing states' Medicaid responsibilities contemporaneously with ongoing state budget crises (National Association of State Budget Officers, 2012). Notably, the findings reported here analyzed data from a strong economic period. As such, the current situation for families of children with autism is likely to worsen as states take action to cut their Medicaid spending. Given other evidence that families raising children with disabilities face elevated levels of material hardship and deprivation, which likely has a deleterious effect on the children's well-being, policymakers should consider ways to strengthen Medicaid and to reduce the financial burdens these families incur (Parish, Rose, Andrews, Grinstein-Weiss, & Richman, 2008).

Most importantly, these findings indicate that policymakers who opt to cut Medicaid spending are likely shifting costs of the care for children with autism to their families. These findings may signal particularly dire consequences for children with autism, given broad state cutbacks to Medicaid (National Association of State Budget Officers, 2012). Such cuts would only serve to further erode critical support for these vulnerable children and their families, because the cuts would have a major effect on states' abilities to sustain their programs (Holahan, Buettgens, Chen, Carroll, & Lawton, 2011). State and federal policymakers should find other ways to address budget deficits, rather than enacting Medicaid cuts that penalize parents of children with autism.

This study's limitations must be considered to fairly interpret its findings. The analyses were correlational, and we cannot infer causality between state programs and family financial expenditures. Second, the ordinal measures of household income and families' expenditures could not fully capture the level of nuance in families' financial burdens. Third, the children's diagnosis of autism was based on parental self-report and is not clinically verified. Fourth, small samples within each state precluded us from analyzing the effects of Medicaid spending on the subset of Medicaid or privately insured children with autism independently. However, we noted that Medicaid spending here may well be a proxy for the adequacy of state support for children with autism, generally, and the effects may extend beyond publicly insured children (Baker & Royalty, 2000; Mitchell, 1991). In addition, there is ample evidence that the large role Medicaid plays in financing health care exerts influence over the care delivered to privately insured children as well (DeWalt, Oberlander, Carey, & Roper, 2005). Fifth, further research should be conducted to understand the financial burden incurred by families of uninsured children. Doing so was not possible with the current dataset because only 79 children lacked insurance. In addition, the magnitude of Medicaid is so large that it likely influences the larger health care system in each state. Further research to understand these relationships is necessary, but beyond the scope of the present analyses. We were not able to analyze the insurance premiums paid by families, because these data were not available in the National Survey of Children with Special Health Care Needs. Finally, some children with autism received services from their schools, and some Medicaid services were provided through the Home and Community Based Waiver program. Because of limitations in the data, we were unable to model the effects or contribution of these services, but further research to understand them would be useful. Further research should investigate the extent to which specific Medicaid program spending is associated with families' financial burden.

Numerous important strengths counterbalance the above limitations. First, the sampling design of the National Survey of Children with Special Health Care Needs resulted in a representative sample of children with autism from each state. Second, the use of multilevel regression enabled us to examine both individual-level and state-level public health program characteristics that were correlated with families' out-of-pocket spending for their children with autism. As far as we know, this study is the first of its kind to analyze the relationship between state Medicaid spending and the financial burden experienced by families raising children with autism.

Our findings highlight that families living in states with greater Medicaid spending for children with disabilities, measured here in per capita spending, reported reduced out-of-pocket health care costs incurred to care for their children with autism. If the pattern of findings reported here is replicated in other research, we speculate that states with less Medicaid program spending for children with disabilities will appear to be shifting health care costs to families that are caring for children with disabilities.

Conclusion

This research examined the relationship between state Medicaid spending and the financial burdens families face in caring for their children with autism. The consistent and robust relationship found between state Medicaid spending levels and the level of financial burden incurred by families of children with autism indicated that Medicaid buffered the effects of raising children whose care needs are typically expensive.

The current grim economic picture for the states raises the risk that state Medicaid cuts will pose significant burdens for families caring for children with autism. The deleterious magnitude of such cuts is underscored by these results, which analyzed data collected during a strong economy. Policymakers concerned about promoting the well-being of vulnerable children with autism and their families should weigh the effects of cutting Medicaid, as these cuts will likely result in uniquely deleterious effects for these children. As our findings demonstrate, families raising children with autism are uniquely vulnerable because of the markedly higher levels of financial burden they incur, even compared to families raising children with other special health care needs. Their vulnerability is heightened because they typically need not only intensive medical care, but also other expensive services, including intensive educational services, supervision, and often a range of therapies. State policymakers need to work to protect families of children with autism during these challenging economic times.

References

References
Andersen
,
R
.
(
1995
) .
Revisiting the behavior model and access to care: Does it matter
?
Journal of Health and Social Behavior
,
36
,
1
10
.
Annie Casey Foundation
.
E
.
(
2008
) .
Median family (with child) income: 2005
.
Autism Speaks
. (
2007
) .
Arguments in support of private insurance coverage of autism-related services
.
Baker
,
L. C
.,
&
Royalty
,
A. B
.
(
2000
) .
Medicaid policy, physician behavior, and health care for the low-income population
.
The Journal of Human Resources
,
35
(
3
) ,
480
502
.
Blumberg
,
S. J
.,
Welch
,
E. M
.,
Chowdhury
,
S. R
.,
Upchurch
,
H. L
.,
Parker
,
E. K
.,
&
Skalland
,
B. J
.
(
2008
) .
Design and operation of the national survey of children with special health care needs, 2005–2006
.
Vital Health Statistics
,
1
(
45
) ,
1
188
.
Carle
,
A. C
.
(
2009
) .
Fitting multilevel models in complex survey data with design weights: Recommendations
.
BMC Medical Research Methodology
,
9
(
49
) ,
1
13
.
Carroll
,
C. D
.,
&
Zhou
,
X
.
(
2010
) .
Measuring wealth effects using U.S. state data
.
Chiri
,
G
.,
&
Warfield
,
M. E
.
(
2012
) .
Unmet need and problems accessing core health care services for children with autism spectrum disorder
.
Maternal and Child Health Journal
,
16
,
1081
1091
.
Dawson
,
G
.,
&
Osterling
,
J
.
(
1997
) .
Early intervention in autism
.
In
M
.
Guralnick
(
Ed.
) ,
The effectiveness of early intervention
(pp.
307
326
).
Baltimore, MD
:
Paul H. Brookes
.
DeWalt
,
D. A
.,
Oberlander
,
J
.,
Carey
,
T. S
.,
&
Roper
,
W. L
.
(
2005
) .
Significance of Medicare and Medicaid programs for the practice of medicine
.
Health Care Financing Review
,
27
,
79
90
.
Fox
,
H
.,
McManus
,
M
.,
&
Limb
,
S
.
(
2000
) .
An examination of state Medicaid financing arrangements for early childhood development services
.
Maternal and Child Health Journal
,
4
,
19
27
.
Ganz
,
M
.
(
2007
) .
The lifetime distribution of the incremental social costs of autism
.
Archives of Pediatric and Adolescent Medicine
,
161
,
343
349
.
Gelberg
,
L
.,
Andersen
,
R. M
.,
&
Leake
,
B. D
.
(
2000
) .
The behavioral model for vulnerable populations: Application to medical care use and outcomes for homeless people
.
Health Services Research
,
34
(
6
) ,
1273
1302
.
Graham
,
J. W
.
(
2009
) .
Missing data analysis: Making it work in the real world
.
Annual Review of Psychology
,
60
,
549
576
.
Graham
,
J. W
.,
Olchowski
,
A. E
.,
&
Gilreath
,
T. D
.
(
2007
) .
How many imputations are really needed? Some practical clarifications of multiple imputation theory
.
Prevention Science
,
8
(
3
) ,
206
213
.
Holahan
,
J
.,
Buettgens
,
M
.,
Chen
,
V
.,
Carroll
,
C
.,
&
Lawton
,
E
.
(
2011
) .
House Republican budget plan: State-by-state impact of changes in Medicaid financing
.
Washington, DC
:
Kaiser Commission on Medicaid and the Uninsured
.
Honberg
,
L. E
.,
Kogan
,
M. D
.,
Allen
,
D
.,
Strickland
,
B. B
.,
&
Newacheck
,
P. W
.
(
2009
) .
Progress in ensuring adequate health insurance for children with special health care needs
.
Pediatrics
,
124
,
1273
1280
.
Hurth
,
J
.,
Shaw
,
E
.,
Izeman
,
S. G
.,
Whaley
,
K
.,
&
Rogers
,
S. J
.
(
1999
) .
Areas of agreement about effective practices among programs serving young children with autism spectrum disorders
.
Infants and Young Children
,
12
,
17
26
.
Kogan
,
M
.,
Strickland
,
B
.,
Blumberg
,
S
.,
Singh
,
G
.,
Perrin
,
J
.,
&
van Dyck
,
P
.
(
2008
) .
A national profile of health care experiences and family impact of autism spectrum disorder among children in the United States, 2005–2006
.
Pediatrics
,
122
,
e1149
e1158
.
Limb
,
S
.,
McManus
,
M
.,
&
Fox
,
H
.
(
2001
) .
Pediatric provider capacity for children with special health care needs: Results from a national survey of state Title V directors
.
Washington, DC
:
Maternal and Child Health Policy Research Center
.
Mitchell
,
J. B
.
(
1991
) .
Physician participation in medicaid revisited
.
Medical Care
,
29
(
7
) ,
645
653
.
Montes
,
G
.,
&
Haterman
,
J. S
.
(
2008
) .
Association of childhood autism spectrum disorders and loss of family income
.
Pediatrics
,
121
(
4
) ,
e821
e826
.
National Association of State Budget Officers
. (
2012
) .
Summary: Spring 2012 fiscal survey of states
.
National Association of State Budget Officers
. (
2011
) .
Summary: Fall 2011 fiscal survey of states
.
Newacheck
,
P. W
.,
&
Kim
,
S. E
.
(
2005
) .
A national profile of health care utilization and expenditures for children with special health care needs
.
Archives of Pediatric and Adolescent Medicine
,
159
,
10
17
.
Oakumura
,
M. J
.,
VanCleave
,
J
.,
Gnanasekaran
,
S
.,
&
Houtrow
,
A
.
(
2009
) .
Understanding factors associated with work loss for families caring for children with special health care needs
.
Pediatrics
,
124
,
392
398
.
Parish
,
S. L
.,
Rose
,
R. A
.,
Andrews
,
M. E
.,
Grinstein-Weiss
,
M
.,
&
Richman
,
E. L
.
(
2008
) .
Material hardship among US families raising children with disabilities
.
Exceptional Children
,
75
,
71
92
.
Parish
,
S. L
.,
Shattuck
,
P. T
.,
&
Rose
,
R. A
.
(
2009
) .
Financial burden of raising children with special health care needs: Association with state policy choices
.
Pediatrics
,
124
,
435
442
.
Peele
,
P. B
.,
Lave
,
J. R
.,
&
Kelleher
,
K. J
.
(
2002
) .
Exclusions and limitations in children's behavioral health care coverage
.
Psychiatric Services
,
53
,
591
594
.
Raudenbush
,
S. W
.,
&
Bryk
,
A. S
.
(
2002
) .
Hierarchical linear models: applications and data analysis methods
.
Thousand Oaks, CA
:
Sage
.
Rose
,
R
.,
&
Fraser
,
M
.
(
2008
) .
A simplified framework for using multiple imputation in social work research
.
Social Work Research
,
32
(
3
) ,
171
178
.
Ruble
,
L. A
.,
Heflinger
,
C. A
.,
Renfrew
,
J. W
.,
&
Saunders
,
R. C
.
(
2005
) .
Access and service use by children with autism spectrum disorders in Medicaid managed care
.
Journal of Autism and Developmental Disorders
,
35
,
3
13
.
Schafer
,
J
.
(
1997
) .
Analysis of incomplete multivariate data
.
Boca Raton, FL
:
Chapman & Hall/CRC
.
Shattuck
,
P. T
.,
&
Parish
,
S. L
.
(
2008
) .
Financial burden in families of children with special health care needs: Variability among states
.
Pediatrics
,
122
(
1
) ,
13
18
.
Thomas
,
K. C
.,
Ellis
,
A. R
.,
McLaurin
,
C
.,
Daniels
,
J
.,
&
Morrissey
,
J. P
.
(
2007
) .
Access to care for autism-related services
.
Journal of Autism and Developmental Disorders
,
37
,
1902
1912
.
Thomas
,
K. C
.,
Morrissey
,
J. P
.,
&
McLaurin
,
C
.
(
2007
) .
Use of autism-related services by families and children
.
Journal of Autism and Developmental Disorders
,
37
,
818
829
.
U.S. Census Bureau, Population Division
. (
2006
) .
Annual estimates of the population for counties: April 1, 2000 to July 1, 2005 (CO-EST2005-01-04)
.
Washington, DC
:
U.S. Census Bureau
.
U.S. Census Bureau, Population Division
. (
2010
) .
Metropolitan and micropolitan statistical areas and components, December 2009, with codes
.
Washington, DC
:
U.S. Census Bureau
.
U.S. Department of Health and Human Services
. (
n.d.
) .
Centers for Medicare and Medicaid services
.
U.S. Department of Health and Human Services, Health Resources and Services Administration, & Maternal and Child Health Bureau
. (
2007
) .
The national survey of children with special health care needs: Chartbook 2005–2006
.
Rockville, MD
:
U.S. Department of Health and Human Services
.
van Dyck
,
P. C
.,
Kogan M
,
D
.,
McPherson
,
M. G
.,
Weissman
,
G. R
.,
&
Newacheck
,
P. W
.
(
2004
) .
Prevalence and characteristics of children with special health care needs
.
Archives of Pediatric and Adolescent Medicine
,
158
(
9
) ,
884
890
.
von Hippel
,
P
.
(
2007
) .
Regression with missing Ys: An improved strategy for analyzing multiply imputed data
.
Sociological Methodology
,
37
(
1
) ,
83
117
.
Young
,
A
.,
Ruble
,
L
.,
&
McGrew
,
J
.
(
2009
) .
Public vs. private insurance: Cost, use, accessibility, and outcomes of services for children with autism spectrum disorders
.
Research in Autism Spectrum Disorders
,
3
,
1023
1033
.

Author notes

Editor-in-Charge: Glenn T. Fujiuri

Susan L. Parish (e-mail: slp@Brandeis.edu), Lurie Institute for Disability Policy, Heller School for Social Policy and Management, Brandeis University, 415 South Street, MS 035, Waltham, MA 02454, USA; Kathleen C. Thomas, Roderick Rose, and Mona Kilany, University of North Carolina at Chapel Hill, and Paul T. Shattuck, Washington University in St. Louis.