There is greater identification of children with autism spectrum disorders (ASD) and, as a result, more attention to specialty services to address the challenges children with ASD face. Along with the growth in identification of ASD is a growth in the population of Latino children, yet there is some evidence that disparities exist in diagnosis and services between Latino and non-Latino White children. This study further documents these disparities and investigates the mechanisms that may contribute to them. Diagnosis and specialty services were compared between 48 Latino and 56 non-Latino White children diagnosed with ASD, and factors that contribute to differences are explored. Results show that Latino children were diagnosed almost one year later than White children, received fewer specialty services, and had higher unmet service needs. Factors that accounted for differences in the number of services received were maternal level of education and the number of sources of knowledge about autism. Findings suggest that service providers need to work to provide greater awareness and knowledge about autism, and make services more accessible to Latino families.
Autism spectrum disorders (ASD) are a growing concern among families, policymakers, and service providers in the United States (Newschaffer et al., 2007). According to the most recent Centers for Disease Control and Prevention (CDC) surveillance study, one in every 88 children has an ASD (CDC, 2012). The substantial increase in prevalence has created a public health priority to implement policies and programs that foster proper assessment, treatment, and support services for individuals with ASD and their families. Early diagnosis and subsequent treatment are critical to optimizing outcomes for children with ASDs (Anderson, Avery, DiPietro, Edwards, & Christian, 1987; Howard, Sparkman, Cohen, Green, & Stanislaw, 2005; National Research Council, 2001).
There is no evidence to suggest that the chance of autism occurring is dependent upon characteristics of the family, such as race or ethnicity, income, lifestyle, and education (Dyches, Wilder, Sudweeks, Obiakor, & Algozzine, 2004). Yet studies exploring the relationship among sociodemographic characteristics, diagnosis, and health care services for children with autism have found socioeconomic and racial and ethnic disparities (CDC, 2012; Fontaine & Bearman, 2011; Liptak et al., 2008; Mandell et al., 2009). The most recent CDC study reported an increase in the detection of ASD among Black and Latino children; however, disparities remain for these populations compared to non-Latino White children in the diagnosis of ASD.
The focus of this paper is on Latino children and families because they are the fastest growing minority population (U.S. Census Bureau, 2011a) and, as a result, are coming to the attention of autism treatment and service providers. The Latino population increased by 43% between 2000 and 2010 (Passel, Cohn, & Lopez, 2011). This growth rate accounts for more than half of the entire population growth in the United States. Latino children and youth are the largest culturally distinct ethnic group of children in the United States; 23.1% of children under the age of 18 are Latino (Lopez & Velasco, 2011).
Studies have consistently reported lower prevalence rates of autism diagnosis for Latinos compared to non-Latino Whites (CDC, 2012; Liptak et al., 2008; Mandell et al., 2009). However, differential prevalence rates may actually represent differences in access to diagnostic services, advocacy, and support as well as the effects of state and federal policies regarding immigrants (Fontaine & Bearman, 2011; Mandell, Listerud, Levy, & Pinto-Martin, 2002). In a study using population-level data in California, Latino children were found to have similar rates of autism diagnosis compared to White children in some years and lower rates compared to Whites in other years, depending on the presence or absence of anti-immigrant policy (Fontaine & Bearman, 2011). Latino children may also be more likely to receive an alternate diagnosis, such as specific language impairment, and may not be evaluated for an ASD when an intellectual impairment is observed (Mandell, Ittenbach, Levy, & Pinto-Martin, 2007; Mandell et al., 2009).
Lower levels of service use and less access are well documented among Latinos across different service systems for various reasons, including lack of information, language barriers, fear of exposing immigration status, and few services that take into account cultural differences (Alegría et al., 2007). There is emerging evidence that once they are diagnosed with an ASD Latino children have less access and lower utilization and quality of health care compared to White children (Liptak et al., 2008; Magaña, Parish, Rose, Timberlake, & Swaine, 2012; Parish, Magaña, Rose, Timberlake, & Swaine, 2012). However, little research has examined the use of autism-specific specialty services among Latino children with autism.
In addition to the limited research on Latinos, little work has been done to theoretically approach disparities in ASD diagnosis and care. In this study, we use Fundamental Cause Theory as our conceptual framework. Fundamental Cause Theory was originally proposed by Link and Phelan in 1995 to call attention to broader social conditions as the major cause of health inequalities rather than viewing these conditions as merely indicators of health. Since that time, it has been used to explain disparities in health outcomes, interventions, and policies in numerous studies (e.g., Chang & Lauderdale, 2009; Herd, Schoeni, & House, 2008; Lufty & Freese, 2005; Peng, 2009; Tehranifar et al., 2009). Link and Phelan proposed that fundamental social causes affect disease outcomes through multiple risk factors. Flexible resources, such as money, knowledge, power, prestige, and beneficial social connections, can be used to avoid risks or to minimize the consequences of illness once it occurs. Individuals who have access to and are able to use the resources flexibly over time and contexts have more favorable outcomes. For example, Peng (2009) found higher SES to be related to higher levels of personal resources, including money, power, social capital, knowledge, and educational attainment. People with higher levels of resources have been shown to have greater access to new technologies and treatments, which then affect outcomes (Chang & Lauderdale, 2009). Treatments that are more difficult to access and implement can lead to greater inequalities because those with more resources are better able to access and implement them (Phelan, Link, & Tehranifar, 2010). This point relates directly to children with ASD because autism treatments and services are difficult to access, and evidence-based treatments for autism, such as intensive therapy based on applied behavioral analysis (ABA), can be difficult to implement, thus potentially contributing to disparities in access and use of these services.
In the current study, we do not propose to use this theory to examine the outcomes for children with autism but rather children's access to and use of autism treatments and services, which presumably lead to better outcomes in the adaptive functioning of children with autism. Applying this perspective to analyzing disparities in early diagnosis and treatment of autism may contribute to understanding the mechanisms for these disparities as well as recommendations for practice to address these disparities.
In this study, we explored disparities in age of diagnosis and service utilization among Latino children compared to non-Latino White children with ASDs and their families in Wisconsin. We are particularly interested in whether Latino children with ASD are adequately accessing publicly funded services, such as Birth to Three and the Medicaid Waiver funded Intensive Treatment Services for Children with Autism. In response to the increased identification of children with autism, the state of Wisconsin determined that it would authorize payment only for services that are evidence based. Therefore, it has designated ABA-based therapy as the main service it will fund for children with autism through the Medicaid Waiver program. There are many eligibility requirements to receive this service, including meeting a disability listing according to SSA rules; having received a diagnostic evaluation by a qualified clinician stating the child has a diagnosis of autism, Asperger syndrome, or pervasive developmental disorder not otherwise specified (PDD-NOS); and the child being under eight years old, and the family must locate a provider who will provide treatment when funding becomes available. Because this is a publicly funded service, it is particularly important that the children from disadvantaged backgrounds are able to access it. However, some of the requirements themselves may create barriers for families to access the service. For example, if the child is diagnosed with autism later, he or she may not be eligible; finding a qualified clinician could be a challenge, and having the knowledge and connections to locate a provider could be difficult. All of these requirements are compounded for children whose parents are Spanish speaking as very few clinicians and providers in the state of Wisconsin are bilingual.
In order to understand the mechanisms by which disparities occur, we investigated whether flexible resources were related to service utilization and unmet service needs consistent with Fundamental Cause Theory (Link & Phelan, 2005). We conceptualized level of education and number of sources of knowledge about autism as resources that give parents flexibility and advantages in being able to advocate for and access services. We selected education level as a variable because higher levels of education may allow families more access to knowledge about disabilities and autism, and it is highly linked to higher income and SES overall. Higher SES means greater access to financial resources that might allow families to pay for second opinions and specialty services that are not covered by insurance. Education also allows them better access to employment with good insurance coverage. A greater number of sources to obtain information about autism was also selected as an indicator of flexible resources. These sources, including the Internet, books and magazines, other parents, and professionals, may be linked to networks that are important in obtaining services. Further, they can help families obtain important knowledge about treatments and how to get them for their children with autism.
Our research questions are (1) Do Latino and non-Latino White children differ in the following diagnostic information: (a) the age when parents notice something different about their child, age of expressing concerns to a doctor, and age of diagnosis; (b) doctor's response to parental concerns; and (c) sources of knowledge about autism once diagnosed? (2) Are there disparities between Latino and non-Latino White children in services, including (a) ever receiving key public services, including Birth to Three and Medicaid Waiver–funded autism therapy; (b) the total number of services received at the time of interview and the number of unmet service needs? (3) Do flexible resources (as measured by maternal education level and the number of sources of knowledge about autism) account for the relationship between race or ethnicity and services?
Participants and Procedure
This study was approved by the Institutional Review Board on Human Subjects at the University of Wisconsin-Madison. Latino and non-Latino White families were recruited with the help of service agencies, support groups, and organizations that provide services or advocacy for children with developmental disabilities and their families in Wisconsin. Inclusion criteria were that the respondent was the main caregiver (all mothers in this study) and the child had a diagnosis of an ASD and was between the ages of 2 and 22 years old. The diagnosis was reported by the parent and was not confirmed by an outside source. However, the Latino parents were administered the Autism Diagnostic Interview as part of a larger, in-depth study that focused on treatment and services, parental stress and coping, and autism symptoms among Latino children with ASD and their families. In order to recruit sufficient numbers of Latino families, Spanish-speaking support groups and service agencies that offered bilingual services were engaged in the recruitment process. They were asked to send study materials in Spanish to potential participants. Participants mailed in a reply sheet if interested and were called back by study staff. Fifty-four Latina mothers of children with ASD were interviewed in their homes. However, seven of them did not complete the treatment and services questionnaire. As a result, 47 were included in these analyses. Interviews were conducted in the language of preference of the participants by bilingual, bicultural interviewers. The White mothers filled out the same treatment and services questionnaire as the Latina mothers through a mail survey. Advertisement for the study was placed on organizational LISTSERVs (e.g., Autism Society) and through service agency mailers. Initially, 88 mothers of children with ASD expressed interest in the study, and, of these, a total of 61 completed the treatment and services questionnaire, which is a 75% completion rate. All mail surveys were in English. Fifty-six of these mothers reported their race/ethnicity as White, four mothers reported their race/ethnicity as Asian, and one reported herself as being Latina. Due to the small number of mothers of Asian descent, they were excluded from the analysis. The mother who was Latina was retained in the study and included in the Latino group. Our final sample included a total of 56 non-Latina White mothers and 48 Latina mothers.
Table 1 shows similarities and differences in demographic characteristics between the two groups. White children were similar to Latino children in mean age at the time of interview and in gender. However, White children had significantly more behavioral problems than Latino children as reported by the mothers. The White and Latina mothers differed on most of the demographic characteristics. The White mothers were older, had higher incomes and higher levels of education, and were more likely to be married or living with a partner than the Latina mothers.
Most of the Latina mothers (88%) were either born in a Latin American country or Puerto Rico and were of Mexican (62%) or Puerto Rican (17%) descent. The remaining 21% were from Central America (Honduras and Costa Rica) or South America (Columbia and Peru) or were of mixed ethnicities. Half of the mothers reported that they spoke English either “good” or “excellent” while the other half did not feel confident speaking English.
Diagnosis and Service-Related Variables
Mothers responded to a questionnaire that included 52 questions adapted from a survey created by David Mandell (see Mandell, Maytali, Novak, & Zubritsky, 2005) and from a measure of services currently received and needed developed by Marsha Seltzer and Marty Krauss for their Aging Family Caregiver study (see Magaña, Seltzer, & Krauss, 2002). Specific variables from this questionnaire used for the current analysis included the following.
Age of concern, report, and ASD diagnosis
Mothers were asked the age of the child when the parents noticed something different, the age of the child when the parents reported their concern to a medical provider, and the age of the child when he or she first received an ASD diagnosis. These are reported in years.
Provider proactive response
Mothers were asked how the medical provider responded to their concerns and were asked to check all that applied: conducted developmental tests, told you it was too early, made a referral to a specialist, told you that your child would grow out of it, and told you your child's development was normal. Those who responded that their provider conducted developmental tests and/or referred them to a specialist were recoded as receiving a proactive response from the provider.
Key public services
Parents were asked whether their child ever received Birth to Three services, early childhood educational services, special education services with an individualized education program, and the autism waiver therapy service. Responses were coded (1) yes and (0) no for each of these services.
Sources of knowledge about autism
Parents were asked where they learned about autism once their child received a diagnosis. They were asked to check all that apply among the following: health care professionals, educational professionals, support group, Internet, parents of children with autism, books and magazines, and friends. Responses were coded (1) yes, (0) no. Values of 1 were counted to compute the number of sources of knowledge variable.
Number of services currently received and number of unmet services
Mothers were asked whether their child was currently receiving one of the seven following services: physical, occupational, or speech therapy; psychological services; social/recreational activities; respite; and intensive autism therapy. Responses were coded (1) yes, (0) no. Values of 1 were counted to compute the number of services currently received variable. If the mother responded no, she was asked if her child needed the services. If the respondent indicated that the child was not receiving the service but needed it, the service was given a code of 1, indicating an unmet service need. All other responses were coded as 0. Values of 1 were counted to compute the number of unmet services.
The number of maladaptive behaviors was assessed by the Scales of Independent Behavior-Revised (SIB-R; Bruininks, Woodcock, Weatherman, & Hill, 1986). The SIB-R consists of eight categories of behavior problems, including internalizing and externalizing behaviors that are common among children with developmental disabilities. Mothers were asked whether their child exhibited any of the maladaptive behaviors, and responses were coded (1) yes, (0) no. Values of 1 were counted to compute the number of child maladaptive behaviors. Reliability and validity have been established by Bruininks et al. (1986). The SIB-R has been used reliably in studies of Latino populations (Blacher & McIntyre, 2006; Magaña & Smith, 2006).
Gender and age were also assessed and reported. To determine diagnosis of the child, mothers were asked whether the child had ever received a diagnosis of autism, Asperger's syndrome, or PDD-NOS. If they responded yes to any of these categories, they were asked who gave the diagnosis and whether it was still considered a current diagnosis.
Maternal demographic variables reported in Table 1 include age, family income (three income categories: $0–$29,999, $30,000–$49,999, and $50,000 and up), education (three educational categories: high school degree or lower, some college, and college graduate), and employment and marital status.
Chi-square or t tests were used to determine differences between groups in categorical and continuous demographic characteristics, respectively. Statistical techniques outlined by Baron and Kenny (1985) were used to test mediating effects of the flexible resource variables within hierarchical regression models. An α level of .05 was used for all statistical tests. In the regression models, we adjusted for the number of behavioral problems, marital status, and child's age. We then added the flexible resource variables: number of sources of knowledge about autism and maternal education level. We adjusted for the number of child behavior problems and marital status because these were both significantly different between the groups and could be related to service use and need. We decided to adjust for child age rather than maternal age because it is more likely to affect services for the child. While income could also be considered a flexible resource, this variable was highly correlated with level of education (r = .66; p = .000), and there was some missing data for this variable. Therefore, we chose not to include income in our regression models. Preliminary analyses were conducted to ensure no violation of the assumptions of normality, linearity, multicollinearity, and homoscedasticity.
For our first research question, we asked whether Latino and non-Latino White children differed in diagnosis-related questions. Table 2 displays the results of t test analyses for the age when parents first noticed something different about their child, the age at which parents told their medical provider about their concern, and the age of ASD diagnosis. The two groups did not differ significantly when asked at what age they noticed something different about their child or at what age the parents told a medical provider about their concerns. However, White children were significantly more likely to be diagnosed with ASD almost one year earlier than Latino children. When asked how the provider responded to their concerns about their child's development, only 31.7% of all mothers reported that their provider responded proactively (defined as taking an action that would lead to assessing ASD). Latina mothers were slightly more likely to report a proactive response from their providers (35.4%) than White mothers (28.6%); however, this difference was not significant. For the majority of families in both groups, medical providers did not respond proactively to parent concerns.
We asked mothers where they learned about autism once their child received an ASD diagnosis, which we refer to as sources of knowledge about autism. As seen in Table 2, Latina and White mothers were found to differ significantly across five sources of knowledge about autism. White mothers reported a higher rate of receiving information from friends, other parents of children with autism, health care professionals, books and magazines, and the Internet than did Latina mothers. The two groups of mothers were similar on receiving information about autism from family members, support groups, and education professionals. Very few parents across both groups reported they did not seek out information. When the eight sources of knowledge about autism were totaled, White mothers (M = 3.46, SD = 2.10) had significantly more sources of knowledge than Latino mothers (M = 1.71, SD = 1.57; t = 4.77, p < .001).
For research question 2, we asked whether Latino and non-Latino White children differed in ever receiving key public services, including the Medicaid Waiver–funded autism therapy. Table 3 illustrates that Latino children were less likely than White children to ever receive Birth to Three services and Medicaid Waiver–funded intensive autism therapy. However, Latino children were just as likely as White children to receive early childhood and special education services.
As part of research question 2, we also asked whether Latino and non-Latino White children differed in the total number of services being used at the time of interview and the total number of unmet services. Table 3 shows that White children were more likely to be receiving recreational programs, psychological services, respite, and intensive autism therapy than Latino children. When the seven services were totaled, White children were receiving significantly more services (M = 3.79, SD = 1.56) compared to Latino children (M = 2.44 services, SD = 1.34; t = 4.70, p < .001).
Regarding unmet service needs, Table 3 shows that Latino children had more unmet service needs for psychological services, respite, and intensive autism therapy than White children. Overall, Latino children had a greater number of unmet service needs (M = 2.23, SD = 1.60) compared to White children (M = 1.23, SD = 1.37; t = −3.42, p < .01).
For research question 3, we asked whether flexible resources (conceptualized as maternal education level and the number of sources of knowledge about autism) accounted for the relationship between ethnicity and services. To test for this, we used hierarchical multiple regression and followed the analytical method outlined by Baron and Kenny (1985) for mediation. We first established a relationship between the independent variable (ethnicity; 1 = Latino, 0 = White) and the mediating variables (level of education, r = −.60, p < .001; and number of sources of knowledge, r = −.43, p < .001). Latino ethnicity was associated with lower levels of education and fewer sources of knowledge about autism. Next, we established a relationship between ethnicity and the outcome variables (number of services, r = −.42, p < .000; number of unmet service needs (r = .32, p < .01). Latino ethnicity was associated with fewer services used and more unmet service needs. Third, we established a relationship between the mediating variables and the outcome variables (number of services and education level, r = .44, p < .001; number of services and sources of knowledge, r = .44, p < .001; number of unmet service needs and level of education, r = −.22, p < .05; and number of unmet service needs and sources of knowledge, r = −21, p = < .05). A higher number of services used was associated with higher levels of education and more sources of knowledge. A higher number of unmet services was associated with lower levels of education and fewer sources of knowledge.
Finally, we created two regression models for each dependent variable with ethnicity as the independent variable (see Table 4). Model 1 adjusted for child age, number of behavior problems, and mothers' marital status. In Model 2, we added the two proposed mediators, maternal education and knowledge about autism to the analyses. For the dependent variable number of services received, being Latino was significantly related to receiving fewer services in Model 1. This model explained 18% of the variance of number of services received using the adjusted R2. When the flexible resource variables were added in Model 2, the relationship between ethnicity and the dependent variable was no longer significant, and the two mediating variables were significantly related to the dependent variable. The Model 2 variables explained 28% of the variance of number of services received. We also tested for mediation for each of the proposed mediators separately and found that entering knowledge about autism and excluding level of education reduced the relationship between ethnicity and services; however, it was still significantly related. When we entered level of education and excluded knowledge about autism, the relationship between ethnicity and services was no longer significant. Nevertheless, the reduction of the relationship of ethnicity and services was stronger when both variables were entered in the model together.
For the dependent variable number of unmet service needs, the flexible resource variables did not account for the relationship between ethnicity and outcome. Being Latino and having a greater number of maladaptive behaviors were related to the number of unmet service needs in both Model 1 and Model 2. The variables in Model 1 explained about 12% of the variance and about 11% of the variance for Model 2.
Previous research has found disparities in the identification of ASD and health care access and use among children who come from racial or ethnic minority backgrounds (Liptak et al., 2008; Magaña, Parish, Rose, Timberlake, & Swaine, 2012; Parish, Magaña, Rose, Timberlake, & Swaine, 2012). The present study sought to further advance research on racial and ethnic disparities by examining the experiences of Latino children with ASD and their families obtaining key autism-related interventions and services and by examining potential mechanisms for these disparities. This focus addresses two demographic trends: Latinos, and especially Latino children, are the fastest growing minority group, and the numbers of children identified with an ASD are increasing. Therefore, research that helps to document disparities and identify factors related to these disparities can further aid in the process of early identification and receipt of intervention programs among this rapidly growing population.
In our first research question, we asked whether there were differences between the two groups in diagnosis-related questions. Consistent with previous research (Mandell et al., 2005), we found a significant difference in the age of diagnosis such that Latino children were more likely to be diagnosed with an ASD almost one year later than White children. Interestingly, there were no significant differences between the two groups in the age when parents noticed something different with their child and the age at which parents told the doctor about their concerns. Furthermore, we found that the majority of parents, irrespective of their ethnicity, reported that the medical providers did not have a proactive response to their concerns. These findings suggest that what happens after expressing concerns to a doctor may be critical to the receipt of a timely diagnosis. Latino immigrant families may have fewer options and resources with respect to researching other possibilities and obtaining second opinions. On the other hand, middle class White families are likely to have more access to the Internet, books, and other resources to explore alternative options as well as resources to pay for specialty clinicians. The attainment of an earlier diagnosis may also be tied to the barriers Latino families experience in terms of the cost and availability of services (Kilbourne, Switzer, Hyman, Crowley-Matoka, & Fine, 2006; Lord & Bishop, 2010), cultural and linguistic competency among providers (Begeer, El Bouk, Boussaid, Terwogt, & Koot, 2009), and biased care systems (Liptak et al., 2008; Mandell et al., 2005).
In our second research question, we asked whether there were disparities between the two groups of children in terms of service use. We found that Latino children were less likely to ever receive Birth to Three services and the Medicaid Waiver–funded intensive autism therapy. Additionally, at the time that questionnaires were filled out, Latino children were receiving fewer services and had more unmet service needs than non-Latino White children. It is important to note there were no differences between the two groups in the use of early childhood services for children from ages 4 to 6 and for K–12 special education services. These programs are run by the school districts, which may be doing a better job at reducing inequalities in receiving services. When asked about their sources of knowledge related to autism, White mothers were more likely to receive information from friends, other parents who have children with autism, health care professionals, books and magazines, and the Internet as compared to Latina mothers. This finding relates directly to the concept of flexible resources as knowledge and beneficial connections are resources that help people navigate health care systems leading to better outcomes (Link & Phelan, 1995).
In our third research question, we sought to explore whether more flexible resources among White families account for the differences between Latino and White children in service utilization. We found that when taking into account differences in education levels and sources of knowledge about autism, the significant difference between White and Latino children was eliminated in the number of services received. These findings support the Fundamental Cause Theory, which proposes that fundamental social factors, such as socioeconomic status and race and ethnicity, are the main causes of disparities in health services and outcomes. These fundamental factors affect outcomes by way of flexible resources in which people who have more of them are more advantaged in accessing important treatments and services that can affect health outcomes (Link & Phelan, 1995). Disparities in age of diagnosis, access to treatment, and consequent treatment utilization may be tied to the availability of flexible resources, such as money, knowledge, power, prestige, and beneficial social connections. As evidenced in this study, White families had more flexible resources than Latino families, which may have led to a greater ability to navigate the autism service system and thus take advantage of more services for their children.
We found that the level of education and sources of knowledge did not account for the differences in the number of unmet service needs. After taking into account these variables, the Latino children continued to have greater unmet service needs than the White families. When asked about whether they received a service and, if not, whether they needed it, many Latino parents indicated that they never knew the service existed and reported they would want the service for their child. While this response relates to a lack of knowledge, it may be that our variable, sources of knowledge about autism, does not capture specific knowledge about what services are available, what they are used for, and how to get them. It is important to note that the unmet needs assessment reflects the parent perspective and not that of health care professionals. Because some Latino parents were hearing about the service for the first time within this assessment with little detail, they may not have an accurate sense of whether their child needed the service while White families may have had more familiarity with the service and whether their child needed it.
An interesting finding is that the White children had more behavioral problems than the Latino children as reported by their parents. It is not clear whether there are real differences in the children or their environments that lead to differences in maladaptive behaviors or whether there are differences in the perception of behavioral problems between the two groups of parents. This finding is consistent with those found in other studies that compare Latinos with ASDs to Whites and is worth further exploration (Magaña & Smith, 2006; Magaña & Smith, 2013).
Study limitations should be noted. First, there may be a lack of generalizability in the study findings because the sample was voluntary. While the demographic variables of the Latino sample were more consistent with those in other studies, the White families in the study were less representative of White families in Wisconsin, especially with respect to level of education (U.S. Census, 2011b). In the state of Wisconsin, 25.8% of adults have a bachelor degree or higher while in our White sample, 75% had a bachelor degree or higher. Second, though we controlled for level of education, because the differences were so great in this variable between the two groups, it is difficult to know whether level of education or ethnicity was the strongest predictor of the number of services. Third, because this was a cross-sectional study, the direction of effects cannot be determined. Fourth, the questionnaires were administered differently between groups: through in-person interviews with the Latina mothers and through a mail survey for the White mothers. This could lead to differences in response effects; however, these differences are less likely in response to factual questions, such as receipt of services and demographic information (Aday & Cornelius, 2006). Last, while the majority of our Latino sample was of Mexican and Puerto Rican descent, these findings may not represent all Latino groups.
Despite these limitations, this study has important implications that should be considered. The results point to significant ethnic disparities in the early identification and diagnosis of ASDs and in access to specialty services. Specifically, Latino children were more likely to receive a later diagnosis of ASD when compared to non-Latino White children with ASD. Latino children received fewer early intervention services, autism waiver services, and fewer services overall than White children. Lower levels of maternal education and fewer sources of knowledge about autism accounted for the differences in the number of services received. Study results highlight the strong role that maternal education plays in the attainment of specialized treatment services for children with ASD. A higher level of education may lead to an increased ability to advocate for their children and greater knowledge about the plethora of services available for children with an ASD (Mandell et al., 2009). Another important factor that may account for differences in service use is the language spoken by the parent, which could also be considered a flexible resource. Only half of the parents in the Latino sample felt they were proficient in English. For parents who were not proficient in English, it may have been challenging to navigate a system that does not include bilingual materials and providers.
As a result of the aforementioned findings, it is imperative to provide educational information for Latino families in their language of choice on what ASD is, such as its early warning signs, treatment options, and how to advocate for these services. Further, it is equally imperative to facilitate access to treatment for all families in order to ensure equal access to autism-related services for those who come from disadvantaged backgrounds. Hiring bilingual and bicultural clinicians and staff to specifically work on reaching underserved populations will go a long way. Because accessing diagnostic and specialty services can be layered and complex, providers could also utilize community-based interventions, such as patient navigation, which is designed to remove barriers to care (Freeman & Rodriguez, 2011; Natale-Pereira, Enard, Nevarez, & Jones, 2011) and educational programs that use promotores de salud or community health workers (WestRasmus, Pineda-Reyes, Tamez, & Westfall, 2012). These types of programs are focused on teaching advocacy skills while helping to guide patients through complex health care systems and can be designed for specific underserved communities to incorporate cultural and language components (Natale-Pereira et al., 2011; WestRasmus et al., 2012). While community-based programs, such as these, are emerging in health care areas, such as cancer detection and diabetes, these types of programs are needed to reach underserved populations with respect to ASD.
Editor-in-Charge: Glenn T. Fujiura
Sandra Magaña, (e-mail: firstname.lastname@example.org), University of Illinois at Chicago, Disability and Human Development, 1640 Roosevelt Road, Chicago, IL 60608, USA; Kristina Lopez, University of Michigan; Arellys Aguinaga, University of Wisconsin-Madison; and Holly Morton, University of Wisconsin-Madison.