Children and youth with autism spectrum disorder (ASD) have complex primary and specialty medical care needs and additionally may use other community and supportive services. Families of children and youth with ASD navigate multiple, intricate service systems, from early childhood through young adulthood. The American Academy of Pediatrics (AAP) asserts the medical home as a best practice standard of care for all children, and particularly children and youth with ASD, for accessible, continuous, comprehensive, family centered, coordinated, compassionate, and culturally effective care [2].
Prior health services research findings identified benefits of medical homes for primary and specialty care, which is associated with earlier ASD detection and reduced wait times for establishing ASD diagnoses once detected [18, 22] annual preventative medical care visits [27] and decreased emergency department (ED) utilization [16]. High-quality medical homes have been associated with greater likelihood of met service needs for children and youth with ASD (e.g., occupational, physical, and speech therapy) and supportive services (e.g., special equipment, transportation, home health, and respite) [8]. Children and youth receiving family centered and coordinated medical home care had fewer unmet service needs [34]. Yet, children and youth with ASD are still less likely to receive medical home care, which is associated with greater unmet service needs, service gaps, costly care inclusive of ED utilization, and increased family responsibilities [9].
In the US, medical home disparities for children and youth with ASD have been attributed to providers’ limited ASD knowledge and training [38] and in the setting of financial, social, linguistic, and cultural barriers of the families of children and youth with ASD [43]. Racial disparities in medical home care access have been shown for children and youth with ASD (i.e., Black, non- Hispanic, Hispanic, or identified as ‘other race’ (i.e., American Indian, Alaska Native, Asian, Native Hawaiian, or other Pacific Islander) compared to White, non-Hispanic children and youth with ASD [11, 27]. Children and youth with ASD with public insurance have also had lower likelihood of receiving medical home care [44].
Understanding the barriers that children and youth with ASD uniquely encounter may manifest in key medical home functions that should collectively promote access (e.g., addressing parental developmental concerns, adherence to recommended ASD screening schedule, referrals for ASD diagnostic assessments, timely identification, and coordinating with early intervention or special education programs) [2]. Although the AAP’s Bright Futures guidelines endorses periodic developmental and ASD-specific screening [14] there is mixed evidence on age of ASD identification, [20] underscoring challenges with early identification of ASD and lost early intervention referrals through the medical home pathway for those children identified after 3 years. Additionally, children may receive ASD-specific screenings on schedule, but experience delays in ASD diagnostic assessments and early intervention eligibility determination [25]. Families of children and youth with ASD who express challenges in sharing their developmental concerns may delay referrals for ASD diagnostic assessments and early intervention eligibility determination [30]. Sociodemographic factors (e.g., lower educational attainment, unemployment, race/ethnicity, and socioeconomic status (SES)) have been cited as challenges to early intervention access [26].
Children and youth with ASD ages 3 through 21 with qualifying disabilities in public school systems receive services under individualized education programs (IEP) according to US federal law [42]. Medical diagnosis of ASD versus educational ASD classification for special education programming are distinct processes requiring families’ understanding of the differences in the service systems’ requirements. Diagnostic criteria for medical ASD are based on the Diagnostic and Statistical Manual of Mental Disorders (DSM), [3] whereas the US educational ASD classification is based on the Individuals with Disabilities Education Act (IDEA) Part B for special education service eligibility. In addition, skilled therapies outlined in IEPs (e.g., occupational, speech and physical therapy) are educationally based “related services” [42] that are different from medically based therapies, although this distinction can confuse medical home care providers and families [17].
Medical home care providers play an important role in supporting access to educational services for children and youth with ASD and their families, such as by initiating referral into early intervention or school-based services to determine service eligibility [28]. As such, medical homes that are better integrated with educational services are important in promoting children’s overall developmental and cognitive well-being. Children and youth’s health statuses and educational outcomes contribute to children's overall health and well-being, and different providers are integral for addressing the educational needs for children and youth with ASD, including medical home care providers, educational personnel, and families/caregivers. While AAP policy and prior conceptual work support the relationship between medical home care and educational service use [1], there is lack of quantifiable population-based evidence on the association between medical home care and educational service use for the pediatric ASD population specifically [39]. Therefore, the purpose of this study is two-fold: 1) to establish the association between medical home care adequacy and the likelihood of current educational service use for children and youth with ASD; and 2) to examine sociodemographic factors that attenuate the association between medical home care adequacy and educational service use. We hypothesized that reported inadequate medical home care will be associated with decreased current educational service use and attenuated by select sociodemographic factors.
Methods
We analyzed cross-sectional data from the 2016–2017 National Survey of Children’s Health (NSCH) [12] to evaluate the relationship between inadequate medical home care and current educational service use in children and youth with ASD. The NSCH is a nationally representative survey of parents or guardians (caregivers) of children and youth ages 0 to 17 years old in the US and District of Columbia to assess children’s physical and mental health, health care access, family characteristics, neighborhoods, school, and social conditions. The 2016 NSCH was conducted from June 2016 to February 2017, and the 2017 NSCH was conducted from August 2017 to February 2018.
The NSCH uses a complex survey design with a multi-level stratified sampling, primary sampling units (PSU), and survey weights to adjust for the sampling approach [41]. Survey sampling weights were adjusted to control for nonresponse bias in the individual 2016 and 2017 NSCH [41]. Caregivers completed the NSCH screening questionnaire to determine if children resided in their households, and proceeded to identify children and youth with special health care needs (CSHCN) [33] through a validated CSHCN screener [10], by responding affirmatively to at least one of five criteria: 1) need or use of prescription medications, 2) elevated service needs, 3) functional limitations, 4) therapeutic service use, and 5) emotional, developmental, or behavioral problems warranting treatment or counseling. The CSHCN screener further assessed if affirmative response(s) were attributed to a medical, behavioral, or other health condition expected to last ≥ 12 months. Eligible caregivers advanced to age specific NSCH topical questionnaires, yielding data on all current study variables [12].
Sample
Children and youth ages 1–17 whose caregivers reported positive or valid responses to, “Has a doctor or other health care provider EVER told you that this child has ASD or autism spectrum disorder (ASD)?” were in the analytic sample (n = 1,248).
Dependent variable
The dependent variable was current educational service use under an IFSP or an IEP. The 2016/2017 NSCH asked, “Has this child EVER had a special education or early intervention plan?” followed by, “Is this child CURRENTLY receiving services under one of these plans?” Current educational service use was defined as receiving services currently under one of these plans.
Independent variable
The independent variable was inadequate medical home care to reflect medical home care for children and youth with ASD [27]. As shown in Fig. 1, the 2016/2017 NSCH used a composite measure of medical home care based on a series of questions across five components: 1) personal doctor or nurse, 2) usual source of sick care, 3) family centered care, 4) referral problems when needed, and 5) effective care coordination when needed.
The NSCH medical home care question route began by affirming whether caregivers had individuals they considered the child’s personal doctor or nurse. Usual source of sick care was operationalized as having a usual place to go when the child is sick, except the ED. Family centered care used five intermediate variables to describe providers’ interactions: 1) spending enough time with the child, 2) listening carefully, 3) sensitivity to family values, 4) providing specific information, 5) and helping families feel like a partner in the child’s care. Children and youth with ASD received family centered care if their caregivers reported “Always” or “Usually” for each family centered care intermediate variable. Referral problems when needed was defined as caregivers reporting needed referrals and not experiencing problems getting referrals. Having effective care coordination, was defined as caregivers’ reported need for assistance with care coordination, “Usually” receiving assistance, and were “Very Satisfied” with communication between the child’s doctor, other health care providers, schools, child care, or special education program.
We employed a stringent criterion to operationalize adequate medical home care as positive and valid caregiver responses across the five medical home components to understand how this rigorous definition is postured on the issue of coordinated health care and educational services for children and youth with ASD. Since adequacy of medical home care is relative to service need, caregivers who reported children “Did not have health care visit in the past 12 months,” “Did not need referrals during the past 12 months,” or “Did not need coordinated care or had less than 2 services during the past 12 months” were also coded as having adequate medical home care. Inadequate medical home care was operationalized as negative or missing responses to one or more of the five medical home components to methodologically decouple missingness (i.e., missing to all versus missing some or one) and preserve a stringent medical home composite measure. We used a conservative measure of “Missing” if children and youth with ASD had invalid responses to all five medical home components and excluded from analyses.
Conceptual framework and covariates
The Behavioral Model for Vulnerable Populations, an expansion of the Andersen Behavioral Model of Health Services Use, with four domains: predisposing, enabling (including enabling-vulnerability), and need [4, 5, 7, 19] guided this study. The original Andersen model has been commonly used to examine factors affecting health service use [6], and a smaller body of work has also demonstrated its utility with examining factors affecting educational service use [23]. Covariates were identified based on prior research and entered sequentially in the multivariate models based on the four Andersen Behavioral Model of Health Services Use domains.
Sociodemographic information was based on caregiver report for predisposing, enabling (including enabling-vulnerability), and need factors. Predisposing factors expected to affect the likelihood of educational service use included child sex, age (continuous), total number of children in household (1–2; ≥ 3), child race/ethnicity (White, Non-Hispanic; Black, Non-Hispanic; Other), adult education level (< high school; > high school), maternal health status (excellent/very good; < excellent/very good), and family structure (two parents (married/unmarried); single parent, other, none reported).
Enabling factors expected to facilitate current educational service were insurance type (private, public, or unspecified insurance) and family federal poverty level (FPL) (≤ 199%;> 199%) [11, 13, 35, 44]. Vulnerability was assessed using the primary language spoken at home categorized as English, Spanish, and other languages.
Need factors expected to increase the propensity for current educational service use included child age at ASD diagnosis (continuous), ASD severity (mild/moderate; severe), and CSHCN factors (ASD, CSHCN; ASD, non-CSHCN).
Analyses
We used Stata 15.1 to apply survey weights appropriate for the subpopulation, the multi- stage sampling design, and correct variance estimation [40]. Predictor variables were checked for multicollinearity and no assumptions were violated. Bivariate comparison of all dichotomous sociodemographic covariates and current educational service use was performed with chi-square tests. We retained statistically significant covariates (p < 0.05), and all non-significant covariates (p > 0.05) with demonstrated theoretical or clinical significance [21] (i.e., child race/ethnicity, adult education level, family structure, insurance type, and family FPL) from previous health or educational service literature [23].
Predisposing, enabling (including enabling-vulnerability), and need factors were sequentially added during multivariate analyses. We began with a model to predict current educational service use from inadequate medical home care (Model 1), followed by adding the predisposing factors (Model 2). Model 3 added enabling factors, Model 4 tested the enabling-vulnerability factor, and the full model included the need factors. We obtained adjusted odds ratios (aOR), significance levels (p < 0.05), and the 95% confidence interval (CI) to assess statistically significant predictors in multivariate models.
This study was exempt by the University of Wisconsin-Milwaukee’s Institutional Review Board.