Barriers to Receiving Applied Behavior Analysis Services in Children With Autism Spectrum Disorder

Introduction Applied behavior analysis (ABA) is commonly used to treat children with autism spectrum disorder (ASD). The objective of this study is to evaluate barriers to ABA treatment in ASD. Methods A voluntary 51-question survey, including demographics, socioeconomic status, parental assertiveness/self-advocacy, and parent perceptions, was provided to caregivers of children aged one to eight years old diagnosed with ASD. The survey consisted of a series of yes/no, five-point Likert scale, multiple-choice, and text field questions. Results A total of 540 surveys were completed. The median time since ASD diagnosis was three to five years ago. Respondents were identified as receiving ABA (r-ABA) vs. not receiving ABA (n-ABA). Respondents were from Florida (60%; r-ABA = 61.7%; n-ABA = 57.0%), Pennsylvania (18%; r-ABA = 21.3%; n-ABA = 12.8%), Delaware (17%; r-ABA = 11.7%; n-ABA = 26.8%), and New Jersey (5%; r-ABA = 5.3%; n-ABA = 3.4%) (p < 0.001). Caregiver belief in ABA treatment, empowerment, and comfort level were greater for r-ABA (4.20 ± 0.72; 3.19 ± 0.93; 4.35 ± 0.72) compared to n-ABA (4.06 ± 0.83; 2.90 ± 1.00; 4.03 ± 0.91), respectively (mean ± SD). Conclusion Barriers to accessing ABA services are a multifactorial issue. Location, income, education of the caregiver, time since ASD diagnosis for the child, as well as caregiver empowerment, belief in treatment, and level of comfort in accessing services most likely contribute to children with ASD not receiving ABA. Empowering parents on their impact on their child’s treatment may improve ABA use.


Introduction
Autism spectrum disorder (ASD) is a neurodevelopmental disorder that has a prevalence of one in 36 in the United States and is characterized by social communication deficits and repetitive sensory and motor behaviors [1][2][3][4].Diagnosis is challenging and is often not made until age four; prompt diagnosis and therapy are essential in promoting optimal outcomes.Diagnostic barriers have been identified previously, including those with low socioeconomic status (SES) and members of minority groups, such as African Americans, who are diagnosed with ASD 1.5 years later than their peers [5][6][7].
It is well known that children with ASD who receive early diagnosis and intervention are more likely to have long-term positive outcomes [8,9].Children receiving intervention before age five are more likely to successfully integrate in normal classroom settings and perform well [10].Children receiving long-term applied behavior analysis (ABA) therapy experienced 47% improvement in intellectual function [11,12].In addition to its effects on children, families also experience caregiver strain, emotional distress, financial burden, and higher divorce rates [13][14][15].While approaches for ASD treatment may vary, ABA is the most common treatment due to abundant evidence of efficacy [13].
ABA utilizes one-on-one interactions with board-certified behavior analysts (BCBAs) and group-oriented

Inclusion/exclusion criteria
The survey was sent to caregivers if they had a child who had been diagnosed with ASD and had received care through Nemours Children's Hospital, consented to receive surveys, and reported English as their primary language.Of survey respondents, children were included if they were between ages one and eight years old.

Measures
The survey was created and adapted from a variety of validated surveys in the field addressing socioeconomic status, parental assertiveness, self-advocacy, and parental perception.The survey was developed by a research team comprised of a developmental and behavioral pediatrician, a clinical child psychologist, a doctorally prepared BCBA, a parent of a child with ASD, and a medical student.A 51-item survey was developed that consisted of a series of yes/no, five-point Likert scale, multiple-choice, and text field questions.The questions addressed factors believed to contribute to caregivers of ASD children receiving ABA, including demographics, socioeconomic status, parental assertiveness/self-advocacy, and parent perceptions (Appendix).

Data collection
Study data were collected, managed, and stored using the secure, de-identified, electronic data capture tool, REDCap (Vanderbilt University, Nashville, TN), hosted at Nemours Children's Health [21,22].Eligible participants could complete the 20-minute survey at their leisure using a secure REDCap survey link, which expired 14 days after receipt.

Statistical analysis
Data output was generated from REDCap.Survey responses were collected predominantly as Likert scale questions and converted into dichotomous variables.Responses "Strongly Agree" and "Agree" were coded as "Yes" and "Disagree" and "Strongly Disagree" were coded as "No."The survey response "Neutral" was excluded from the data analysis.Descriptive statistics were performed on caregiver demographics and responses to ABA barriers.Pearson's chi-square test (χ2) and independent Student's t-test were performed comparing those receiving ABA (r-ABA) and those who have never received ABA (n-ABA) on response variables as indicated by the variable type.Multivariate logistic regression analysis was performed on responses to ABA barriers comparing ABA utilization.The demographic variables of race, income, and insurance status were used as controls for potential co-variability.Statistical analysis was performed using IBM SPSS Statistics 28 (IBM Corp., Armonk, NY), SAS 9.4 (SAS Institute Inc., Cary, NC), and MathWorks MATLAB_R2020a (MathWorks, Natick, MA).Significance was defined as p < 0.050.

Factor analysis
Twelve questions relating to caregiver beliefs in treatment, empowerment, and comfort level were examined for factorability.Outliers were not an issue since all the questions existed on a scale from one to five.The Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett's test of sphericity were used to see if the data were appropriate for factor analysis.The entire analysis was done using SAS 9.4.

Participant demographics
A total of 540 surveys were received, with 444 completed surveys, and 73 incomplete surveys.A total of 23 participants whose child was diagnosed within the last six months due to not having proper time to receive services were excluded.For each question, participants had the option of answering "other," "don't know," or to omit the question.Of the respondents, 92.5% were female and 73.1% were aged 30-44 years.Ethnicities of participants included 56% (n = 234) Caucasian, 13% (n = 55) African American, 4.8% (n = 20) Asian, and 17% (n = 73) Latino or Hispanic.If respondents selected multiple ethnicities, they were reclassified as "multiracial."No significant differences in caregiver age, marital status, and ethnicity between their child receiving ABA and not receiving ABA were noted (Table 1).

Variable
Have received ABA therapy, N (%)

Time since ASD diagnosis
When comparing raw survey responses for time since diagnosis, there was no statistical significance among groups (p = 0.067).However, when subgroup analysis was performed, caregivers with r-ABA children were more likely to be diagnosed at least three years ago (r-ABA = 73.7%;n-ABA = 62.9%) (p = 0.018) than those diagnosed within the last two years (r-ABA = 26.3%;n-ABA = 37.1%) (Table 1 and Figure 1).When comparing r-ABA vs. n-ABA for each time since diagnosis, those who were more newly diagnosed were less likely to have received ABA, while those who had more time since their diagnosis were more likely to receive ABA.

FIGURE 1: Length of time since ASD diagnosis of those who have never received ABA
ASD: autism spectrum disorder; ABA: applied behavior analysis.

FIGURE 2: Income level of those who have never received ABA
ABA: applied behavior analysis.

FIGURE 3: Education status of those who have never received ABA
ABA: applied behavior analysis.
Those who did not agree that proper treatment will improve the behavior of children with ASD were more likely to state they were not currently getting ABA (11.06 (2.61-46.89),p < 0.010) and were more likely to state they had to stop getting ABA because their insurance did not cover the therapy compared to those who did agree that treatment would improve behavior (OR = 7.05, 95% CI = 2.11-23.6,p < 0.010).

Never received ABA
Caregivers whose child was not receiving ABA cited the most common reasons their child has never received ABA include: long waitlist times (33.7%), they have heard bad things about ABA (13.7%), they feel overwhelmed and do not know where to start (12.6%),ABA was not covered by their insurance (10.3%), or there was no ABA in their area (10.3%).Additional reasonings included that the caregiver did not want the therapy at this time (10.3%), they have never heard of ABA therapy (9.1%), or did not have enough time to get their child to all their appointments (8.6%).

Factor analysis
Twelve questions were examined for factorability.Outliers were not an issue due to the fact all the questions exist on a scale from one to five.Eleven of these questions had at least one correlation with a magnitude of at least 0.3, which suggests reasonable factorability.The Kaiser-Meyer-Olkin measure of sampling adequacy was 0.82, above the recommended value of 0.60, and Bartlett's test of sphericity was significant implying that these data are appropriate for factor analysis.It can be inferred that there is shared variance among the 12 questions due to the fact the final communality estimates were all above 0.30.
The question "I can always find the time to make all the phone calls I need to get my child into therapies" was eliminated at the start because it had large factor loadings on multiple factors.Three factors were selected because this produced a simple structure where each variable loaded highly only onto a single factor.The proportion of the variance explained by factor 1 was 55.09%, 32.42% for factor 2, and 12.49% for factor 3. The inter-factor correlation between factor 1 and factor 2 was 0.08, between factor 1 and factor 3 was 0.19, and between factor 2 and factor 3 was 0.41; due to the correlation between factor 2 and factor 3 being above 0.30, an oblique varimax rotation was done to account for this correlation (    3).

Factor 3: caregiver comfort level
Factor 3 measures caregiver comfort level and contains the variables comfortable sending emails, comfortable making calls, and taking control to get services for their child.Each question had a loading on this factor of at least 0.55 and loadings below 0.20 on all other factors.Factor 3 has a Cronbach's alpha of 0.82.The average of caregiver responses, on a five-point Likert scale, for factor 3 was greater in those who received ABA (M = 4.35; SD = 0.72) than those who never received ABA (M = 4.03; SD = 0.91) (t(466) = 4.289; p < 0.001) (Table 3).

Discussion
There are many barriers to accessing adequate ABA and many children with ASD do not receive the services they need.By surveying caregivers, we identified factors preventing children with ASD from receiving ABA.The biggest barriers identified in this study were location, income, caregiver education, and time since the child's ASD diagnosis as well as caregiver empowerment, belief in treatment, and comfort level accessing services.

Caregiver identity
Prior studies have shown that a gender imbalance among primary caregivers exists [19].Our study echoes previous findings, as the majority of respondents were female between ages 30 and 44 years.Previous studies have also identified it as a barrier, showing African American children with ASD to be diagnosed at a later age and African American and Hispanic populations with decreased accessibility to ASD treatment when compared to Caucasian children [7,[23][24][25].Interestingly, race was not noted to be a significant ABA access barrier.We speculate the patient demographics of Nemours Children's Hospital locations are contributory despite comparability to US Census racial rates [26].

Caregiver location
Survey data suggest location is contributory to ABA disparities.Data demonstrate an uneven number of BCBAs between US counties; more than half of all counties have none [27].In our study, respondents who reported they did not know how to find the right services cited a lack of ABA in their area as a barrier to obtaining therapy for their child.Delaware had a higher rate of n-ABA children.Florida, New Jersey, and Pennsylvania, however, each had more children receiving services than not.Notably, the amount of BCBAs in each state is as follows: Delaware = 135, Florida = 7,306, New Jersey = 2,810, and Pennsylvania = 2315 [28].
The ratio of BCBA/10,000 people is as follows: Delaware = 1,311, Florida = 3,354, New Jersey = 3,041, and Pennsylvania = 1,793.While the reasons for these differences remain unclear, our study illuminates a need for further research to explore the impact of location on ABA initiation.

Caregiver income, education, and work activity
Cost is a barrier to receiving ABA [17].Horlin et al. found that the median annual cost of having a child with ASD was $34,900, with nearly 90% of that cost due to loss of wages [17].Our study showed that those with higher household incomes and higher educations were more likely to receive ABA for their child.This may be due to having better access to resources to facilitate treatment, such as time, transportation, or additional help outside the family.We found that of those who worked full-time, part-time, were homemakers, or were unemployed, a greater percentage of their children with ASD have received ABA.Those who were disabled were less likely to receive ABA.This may be due to the high cost of having a child with ASD even before paying for ABA; requiring that caregivers take on more hours to provide financially.Caregivers who work less or are unable to work may not be able to shoulder ABA costs.This may be due to lower income due to fewer hours or an inability to qualify for insurance benefits to allay treatment expenses.Lack of time may be another ABA barrier and suggests that those with more free time are better equipped to engage in intensive therapies for up to seven days a week.

Insurance
Over the past few decades, there has been an increase in both Medicaid and private insurance plans covering ABA.However, it is generally believed that caregivers with Medicaid or other government-sponsored plans have a longer wait time to access ABA.There are also differences in reimbursement rates that vary widely across plans [29].Low reimbursement could deter qualified providers from contracting with particular insurance and, thus, limit available resources for families.Our study found that insurance status was not significant in caregivers securing ABA for their child.Additionally, our data suggest that insurance coverage, or lack thereof, seems to influence family perceptions regarding treatment benefits.Caregivers who did not agree that proper treatment would improve the behavior of children with ASD were more likely to state they were not currently getting ABA and were more likely to state they had to stop getting ABA because their insurance did not cover the therapy.We noted a positive correlation between the number of children receiving ABA and the amount of time since ASD diagnosis.This is likely attributed to the well-recognized long wait times across the country [30].

Caregiver belief in treatment
We found caregiver belief in treatment was greater in those who received ABA than those who never received ABA.Caregivers whose child had never received ABA were less likely to believe that the earlier a child gets treatment for ASD the more progress the child will make and also less likely to believe that with proper treatment, behavior and development will improve.A previous study reported that 48% of families did not know anything about ASD before their child was diagnosed [6].This emphasizes the need for providers to educate caregivers on the significant impact that these behavioral therapies, such as ABA, can have on their child's behavior, development, and overall long-term outcomes.

Caregiver empowerment
There was a significant difference in caregivers' level of empowerment seen in those who received ABA as compared to those who did not.Families who never received ABA were less likely to know how to find the right services, less likely to know what to do if their child was not getting the right services, and less likely to know which agencies, therapists, and doctors they needed to contact to get proper services for their child.However, caregiver empowerment regarding ABA has not yet been fully studied and further research is warranted.

Caregiver comfort level
Our data showed there was a significant difference in caregivers' comfort levels between those who had received ABA and those who had not, with comfort levels greater in those who received ABA than those who never received ABA.Caregivers surveyed who had not received ABA for their children were less comfortable contacting child services through phone or email.This shows that caregiver comfort level is a barrier to receiving ABA.This underscores the need for greater caregiver education on ASD treatment options to promote greater comfort in utilizing the resources available.

Limitations
While this study successfully used a multisite, multistate children's hospital system, the data collected for this study consisted of a survey, which inherently has limitations.Performing a survey analysis includes biases of each survey participant, subjectivity of answers, and the ability to recall specific barriers that they faced in the past.Additionally, this was a long 51-question survey, which could have created some fatigue during survey completion and could have led to participants not completing it in its entirety.Another limitation is that the survey was only provided in English and there may be language barriers that could contribute to accessing ABA that were not addressed in our study.

Conclusions
A total of 540 caregivers of patients with ASD completed a 51-question survey to identify the barriers to children with ASD receiving ABA.Caregivers from Delaware noted a significant disparity in ABA accessibility.Caregivers who were more assertive had a positive outlook on treatment, and caregivers who had more knowledge about ABA were more likely to obtain it for their children.ABA accessibility is a multifactorial issue.Location, income, education of the caregiver, time since ASD diagnosis for the child, as well as caregiver empowerment, belief in treatment, and level of comfort in accessing services most likely contribute to children with ASD not receiving ABA.This emphasizes the need for providers to educate caregivers on the significant impact that these behavioral therapies, such as ABA, can have on their child's behavior, development, and overall long-term outcomes.We did not like the ABA provider e.We did not like ABA therapy for our child f.ABA is not our priority right now g.ABA providers do not speak our language h.We had insurance issues that kept us from getting ABA i.We changed insurance providers j.My insurance doesn't cover ABA and we can't pay for it ourselves k.We can't find a provider that matches our schedule l.We had transportation issues that kept us from getting ABA m.We were physically distancing due to COVID-19

Factor 2 :
caregiver empowermentFactor 2 measures caregiver empowerment for getting treatment for their child and contains the variables of knowing who to contact, finding the right services, easy-to-get services, and knowing what to do if their child is not getting the right services.Each question had a loading on this factor of at least 0.60 and loadings below 0.20 on all other factors.Factor 2 has a Cronbach's alpha of 0.84.The average of caregiver responses, on a five-point Likert scale, for factor 2 was greater in r-ABA (M = 3.19; SD = 0.93) than in n-ABA (M = 2.90; SD = 1.00) (t(467) = 3.159; p < 0.001) (Table

Survey 1 . 2 . 4 . 6 . 7 . 10 . 20 . 22 . 23 . 24 .
Are you the primary caregiver of the child who has autism?If other, please explain: ______ 3. What state do you currently live in?What is your relationship to the child who has autism?a. Biological parent b.Step-parent c. Foster parent d.Adoptive parent e.Family member (grandparent/aunt/uncle) f.Friend of the family 5. Has your child been formally evaluated and diagnosed with autism by a specialist?(Outside of your primary care provider) How long ago was your child diagnosed with autism?What is the primary health insurance for your child with autism?a.Current or former employer or union (of the person completing this survey or another family member) b.Insurance purchased directly from an insurance company (by the person completing this form or another family member) c.Medicaid, Medical Assistance, CHIP, or any kind of government assistance plan for those with low incomes or a disability d.TRICARE, VA, or other military health care e. Indian Health Insurance f.No health insurance g.Other (please specify below)8.If other, please specify the primary form of health insurance for your child with autism.9. How many ADULTS are currently living in your house, including yourself?How many children UNDER the age of 18 years old are currently living in your house (including your child with autism)?know how to find the right services for my child with AUTISM.take control to get the services that my child needs to make progress.know what to do if my child is NOT getting the right therapy services.am comfortable making phone calls to get my child services.am comfortable sending emails to get my child services.like to connect with other parents on social media to find resources for my child.can always find the time to make all the phone calls I need to get my child into therapy.Disagree 18.I know which agencies, therapists, and doctors I need to contact to get my child the help he or she needs.2023Littman et al.Cureus 15(11): e48585.DOI 10.7759/cureus.48585find it easy to get the services my child needs to make progress.If selected disagree or strongly disagree, please specify why: ______ 21.The earlier a child gets TREATMENT for autism the more progress the child will make.With proper treatment, the BEHAVIOR of children with autism will improve.With proper treatment, the DEVELOPMENT of children with autism will improve.With proper treatment, children with autism can be more like children who do not have autism.
yes, how long did your child go without ABA? is the HIGHEST level of school you completed?a. Didn't finish high school b.Didn't finish high school, but completed a technical/vocational program c. Completed high school and a technical/vocational program d. High school diploma or equivalency (GED) e. Less than 2 years of college f.Associate's degree (junior college) or equivalent g.Bachelor's degree h.Master's degree or other postgraduate training i. Doctorate or other professional degree (PhD, EdD, DVM, MD, JD, DDS, etc.) j.Other 46.If didn't finish high school, what was the highest grade you completed?______ 47.Which of the following best describes your current daily activities and/or responsibilities?a. Working full-time b.Working part-time c.Unemployed or laid off, not looking for work d.Unemployed or laid off, looking for work e. Homemaking or raising children full-time f.Retired g.Disabled (not working because of permanent or temporary disability) 48.What is your total household income in the past year, before taxes and other deductions?a. Less than $5,000 b. $5,000 through $24,999 c. $25,000 through $49,999 d. $50,000 through $74,999 e. $75,000 through $99,999 f. $100,000 through $149,999 g. $150,000 + h.Do not know 49.Is the home where you live: a. OWNED by you or someone in the house (WITH a mortgage or loan)?b.OWNED by you or someone in the house free and clear (WITHOUT a mortgage or loan) c.RENTING a home or apartment d.LIVING in a home/apartment without having to pay money or rent?e. LIVING with family or friends (giving money for rent/expenses) f.LIVING with family or friends (NOT giving money for rent/expenses) g.Homeless h.Other (please specify your living situation below) 50.If other, please specify your living situation.______ 51.Is this survey complete?a. Yes b.No

TABLE 2 : Factor loadings in factor analysis Factor 1: caregiver belief in treatment
Factor 1 measures the concept of caregiver belief in treatment and contains the variables child development information, behavior information, early treatment information, and treatment improvement information.Each question had a loading on this factor of at least 0.55 and loadings below 0.10 on all other factors.Factor 1 has a Cronbach's alpha of 0.811.The average of caregiver responses, on a five-point Likert scale, for factor 1 was greater in r-ABA (M = 4.20; SD = 0.72) than n-ABA (M = 4.06; SD =0.83) (t(467) = 1.91, p = 0.029) (Table3).

TABLE 3 : Caregiver belief, empowerment, and comfort level based on ABA status
* = Statistically significant difference between subgroups defined as p < 0.050.ABA: applied behavior analysis.