Q&A

O R I G I N A L P A P E R

Access to Care for Autism-Related Services

Kathleen C. Thomas Æ Alan R. Ellis Æ Carolyn McLaurin Æ Julie Daniels Æ Joseph P. Morrissey

Published online: 19 March 2007 � Springer Science+Business Media, LLC 2007

Abstract This paper identifies family characteristics

associated with use of autism-related services. A

telephone or in-person survey was completed during

2003–2005 by 383 North Carolina families with a child

11 years old or younger with ASD. Access to care is

limited for racial and ethnic minority families, with low

parental education, living in nonmetropolitan areas,

and not following a major treatment approach. Service

use is more likely when parents have higher stress.

Families use a broad array of services; the mix varies

with child ASD diagnosis and age group. Disparities in

service use associated with race, residence and educa-

tion point to the need to develop policy, practice and

family-level interventions that can address barriers to

services for children with ASD.

Keywords Autism � Services � Access

Children with autism spectrum disorder (ASD) have a

heterogeneous set of skills and deficits. A successful

treatment protocol must take into account the individ-

ual characteristics of the child and that child’s family in

order to define meaningful treatment goals and strat-

egies to move toward them (Hurth, Shaw, Izeman,

Whaley, & Rogers, 1999; Rogers, 1998; Dawson &

Osterling, 1997). Families are aware that ‘one size does

not fit all,’ and they commonly express interest in a

broad array of treatments in their search for the

one(s) right for their child (Green, Pituch, Itochon,

Choi, & Sigafoos, 2005; Carey, 2004; Gross, 2004;

Bodfish, 2004; Smith & Antolovich, 2000; Levy,

Mandell, Merher, Ittenbach, & Pinto-Martin, 2003).

Several studies have found that families with a child

with ASD experience difficulties accessing services

(Ruble, Heflinger, Renfrew, & Saunders, 2005; Kraus,

Gulley, Sciegaii, & Wells, 2003; Kohler, 2000). How-

ever, studies of child and family characteristics associ-

ated with use of services for ASD have been limited.

Children with ASD of minority race and ethnicity have

been found to receive services at a later age and

receive a different mix of services from white children

(Mandell, Listerus, Levy, & Pinto-Martin, 2002; Levy

et al., 2003). Younger children have been found to use

more services, and families were less likely to express

difficulties accessing care when children were covered

by public or private insurance (Green et al., 2005;

Kraus et al., 2003). Families with higher education,

with more than one child with special health care

needs, and those not following a major treatment

approach for ASD (e.g. TEACCH: Marcus, Garfinkle,

& Wolery, 2001; Lovaas: McEachin, Smith, & Lovaas,

1993; Floortime: Wieder & Greenspan, 2003) have

been found to express problems accessing care or use

K. C. Thomas (&) � A. R. Ellis � C. McLaurin � J. P. Morrissey Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, 725 Martin Luther King Jr. Blvd. CB#7590, Chapel Hill, NC 27599- 7590, USA e-mail: [email protected]

J. Daniels Department of Epidemiology, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

J. P. Morrissey Department of Health Policy and Administration, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

J Autism Dev Disord (2007) 37:1902–1912

DOI 10.1007/s10803-006-0323-7

123

fewer services (Thomas, Morrissey, & McLaurin, 2006;

Kraus et al., 2003). Specific relationships between

service use and child and parent behavioral health

needs have not been consistently demonstrated

(Luther, Canham & Cureton, 2005; Green et al.,

2005; Hare, Pratt, Burton, Bromley, & Emerson,

2004; Kraus et al. 2003; Dunn, Burbine, Bowers, &

Tantleff-Dunn, 2001; Hastings & Johnson, 2001; Fac-

tor, Perry, & Freeman, 1990).

This study sought to expand the literature on access

to services for ASD by studying the characteristics of a

community sample of families and their child with ASD

that are associated with use of a wide range of autism-

related services. The classical Andersen and Aday

model of access to health care provided the framework

for identifying families who have limited access to care

for ASD. Family characteristics that impact service use

were divided into predisposing, enabling and need

factors (Aday, 1989; Aday & Andersen, 1975). A

number of different autism-related services were

examined in order to identify any differences in the

patterns of utilization across type of service.

Methods

Sample

The study was based on a community sample of 383

families with a child with ASD, aged 11 years or

younger, living in North Carolina. Although the sample

was based on a single state, North Carolina is widely

considered to have a comprehensive service system for

young children with ASD and can provide an under-

standing of what utilization looks like when families

have many choices. At the same time, there is variation

in service use in North Carolina (Thomas et al., 2006),

so it should provide a conservative view of the family

and child characteristics that account for variation.

The sample was recruited through a two-pronged

approach. Sixty percent of the sample was obtained

through a subject registry and the remainder through

direct recruitment. The study used the Neurodevelop-

mental Disorders Research Center Subject Registry at

the University of North Carolina at Chapel Hill (UNC).

The Subject Registry enrolls families served at Division

TEACCH (Treatment and Education of Autistic and

related Communication handicapped CHildren) regio-

nal centers who are interested in participating in

research on ASD. When a child is diagnosed, or

suspected of a diagnosis of ASD, either by a private

practitioner or by a state Children’s Developmental

Services Agency, the child can be referred to a

TEACCH Regional Center to confirm that diagnosis.

For those with confirmed diagnoses, TEACCH serves

as a source of training and referral for the family and for

teachers (Campbell, Schopler, Cueva, & Hallin, 1996).

The Registry sent a mailing to families; interested

families provided contact information to the study team

or gave consent to the Registry to do so. Participants

were also recruited directly through mailings to all

elementary school principals and directors of excep-

tional children’s divisions, TEACCH Regional Centers,

and state Children’s Developmental Services Agencies

within North Carolina. In addition, the Autism Society

of North Carolina and Families for Early Autism

Treatment of North Carolina (NCFEAT), two major

advocacy groups, assisted in recruitment by sharing

study information and an invitation to participate. The

referring agency provided potential participants with a

description of the study and contact information for the

study team, so that potential participants, rather than

the research team, initiated contact.

Families who expressed an interest in participating

in the study provided an address and were mailed and

returned written informed consent to participate. The

consent form was approved by the UNC Office of

Human Research Ethics. The majority (97 percent) of

respondents were mothers.

Data Collection

Interviews were conducted in two phases. Families of

children aged 8 and younger were interviewed by

telephone during the winter of 2003–2004 (Thomas

et al., 2006). Families of children aged 9–11 years were

interviewed in person during the winter of 2004–2005.

Each interview was conducted in two segments, with

the phone or in-person interview supplemented with a

self-administered segment. The self-administered seg-

ment was sent by mail to the families of younger

children. For families of older children, the respondent

was given the self-administered survey booklet during

a break in the in-person interview. These efforts

yielded a 91 percent response rate amongst families

who initially expressed an interest in participating.

Measures

All measures were derived from the survey and based

on parent’s recall and report. Thomas et al. (2006)

provides a description of instrument development.

A list of autism-related services was developed from

those described in the literature and listed on web sites

maintained by advocacy groups such as the Autism

123

J Autism Dev Disord (2007) 37:1902–1912 1903

Society of America as well as various state agencies

across the country. Service use was catalogued, regard-

less of the extent to which any given therapy might be

recommended by professionals or supported in the

literature. Services were listed in groups, beginning

with those in school and those outside of school

(including diet therapies, other autism-specific thera-

pies, and general therapies not specific to ASD). The

goal was to distinguish different sources of autism-

related service use. For example, a child might receive

speech/language therapy at school, and outside school,

targeted to children with ASD, or targeted to children

with a variety of diagnoses. The survey also catalogued

parent and sibling services, child care services, and

other special service providers. These sections served

to identify any additional service use that might be

more readily identified by the type of provider than by

the name of the service. For example, after receiving a

set of services in clinic settings, a family might have

a behavioral specialist come to their home to set up a

structured learning environment for the child which

could be maintained with the help of a therapeutic

support person. Brief descriptions of the services were

offered if requested. Medication and supplement use

was measured using a modified version of the Brief

Medication Questionnaire (BMQ; Svarstad, Chewning,

Sleath, & Claesson, 1999).

We used a measure we developed to identify the

major treatment approach for ASD used by the family

(Thomas et al., 2006). After piloting the instrument,

parents recommended that we distinguish between this

‘major treatment approach’ and routine autism-related

services. The major treatment approach was assessed

with a set of questions about approach utilization

(referred to by approach name(s) and acronym). The

survey asked about Applied Behavior Analysis (or

ABA, but not the Lovaas approach), Defeat Autism

Now (DAN), Floor Time (or Greenspan), the Denver

Model, Lovaas and TEACCH (or Treatment and

Education of Autistic and related Communication

handicapped CHildren). In this way, familiarity was

assessed based on name recognition.

Family stress was measured with the Questionnaire

on Resources and Stress (QRS), using the short form

developed by Friedrich and colleagues (Friedrich,

Greenberg, & Crnic, 1983). The survey contains the

subscale on Parent and Family Problems. This is an 18

item scale of true/false questions which provide a

summed score. The norm for children with mental

retardation is around 6. The short form of the QRS has

a high correlation (q = 0.997) with the long form and internal and external validity (Rousey, Best, & Blach-

er, 1992; Friedrich et al., 1983).

The instrument assessed health insurance coverage

of several different types: private, Medicaid, Health

Choice (North Carolina’s state children’s health insur-

ance plan), coverage from any other public program, or

none. Families were asked to code all that apply, so

that any secondary coverage could be noted.

ASD diagnosis was based on parent’s report of

Asperger syndrome, classical autism, or pervasive

developmental disorder not otherwise specified

(PDD-NOS). The instrument also asked about Rhett’s

and childhood disintegrative disorder, but responses

were too few to distinguish, so they were combined

with the classical autism category.

Racial and ethnic minority status was measured

from report of race (up to five separate races could be

identified) and Hispanic culture. All families who were

not white and all who were Hispanic were categorized

as racial and ethnic minority.

Nonmetropolitan residence is based on Rural-Urban

Continuum Code (RUCC) a classification scheme that

distinguishes metropolitan counties by the population

size of their metro area, and nonmetropolitan counties

by degree of urbanization and adjacency to a metro-

politan area (USDA, 2004). RUCC codes are devel-

oped at the zip code level and applied to each family

based on address of residence. (USDA, 2004)

Statistical Methods

Chi square tests (df = 1) were used to measure the

difference in sample demographic characteristics and

use of major treatment approach by age group of the

child with ASD (4 years or less, 5–8 years, and 9–

11 years). Logit models were used to measure the

predisposing, enabling and need characteristics associ-

ated with use of each autism-related service obtained

outside school. This allowed us to see variation in the

factors associated with use of each type of service.

Findings are presented only for those services used by

at least 10 percent of families (n = 38) and models that

produced a significant fit (according to a change in the

likelihood ratio statistic, a = 0.05) in order to assure their robustness.

Results

Sample Characteristics

Table 1 presents characteristics of sample families and

their child with ASD. In comparison to North Carolina

and national populations, the sample was slightly less

likely to be of minority race or ethnicity, had slightly

123

1904 J Autism Dev Disord (2007) 37:1902–1912

higher education and income, and was more likely to

have health insurance. In North Carolina, 21.6 percent

of families are African American and 4.7 percent are

Hispanic (U.S. Census, 2001). Seventy-eight percent of

individuals aged 25 or over in North Carolina have a

high school degree or higher (U.S. Census, 2000). The

state median income in 2003 was $46,000 (U.S. Census,

2003). Among children, age 18 or younger in the

general population, 56 percent have private health

insurance, 23 percent have Medicaid, and 7 percent

have no insurance (national estimates from the Med-

ical Expenditure Panel Survey in 2002; Rhoades &

Cohen, 2004).

The older elementary children (9–11 years) were

more likely to have a diagnosis of Asperger syndrome

or mental retardation, compared to younger children.

Studies of the rates of classical autism compared to

Asperger syndrome are sparse, but Fombonne suggests

a mean ratio of classical autism to Asperger syndrome

of roughly 4:1 (Fombonne, 2003). The rate was 3.5:1 in

the study sample. However, this rate was higher (5.5:1)

among younger children (4 years old or less), but much

lower (1.1:1) for older children. The sample prevalence

of mental retardation was much lower than the median

prevalence of 70 percent reported in a review of the

literature (Fombonne, 2002).

The study sample appears similar to those of other

epidemiological studies of ASD with respect to gender.

The proportion of boys in the sample was very close to

the mean ratio of boys to girls with ASD of 4.3 to 1 (or

86 percent boys) reported in a review of epidemiolog-

ical studies (Fombonne, 2003).

Major Treatment Approach for ASD

Three quarters of the families of the youngest (4 or

younger) children and two thirds of remaining families

reported using one or more of the listed major

treatment approaches for ASD (Table 2). No families

reported using the Denver model, so it was dropped

from the analysis. Thirty percent of the families of the

youngest children were using more than one approach;

use of multiple approaches diminished to 11 percent of

the older children. Just over half (55–62 percent) of

families used a TEACCH approach. One quarter to

one third of families said that they did not use any of

the listed major treatment approaches for ASD.

Service Use

Families were using a broad array of services. Table 3

lists the 56 services included in the survey instrument,

grouped by whether they are received in school or

outside school, and by category of service. The table

gives the percentage of families who were currently

using each service at the time of the survey by age of

the child (4 years or less, 5–8, or 9–11 years old).

Altogether, families were using 46 of the 56 services

listed. Families with children in the middle age group

(5–8 years old) were using a wider range of services (44

out of 56) compared to those with older (9–11 years

old) children (35 out of 56) or younger (4 years old or

less) children (31 out of 56). Cells with fewer than 3

families are reported as zero.

Predictors of Service Use

Table 4 shows logit models of any use for 16

different services that families use outside school.

Table 1 Sample characteristics: Child with ASD and family

Characteristic Percent Mean (SD)

Child characteristics: Age in years – 7 (2.4)

Age 4 years or less 26 Age 5–8 years 52 Age 9–11 years 22

Male 87 Race

White 76 African American 16 Other race 5

Hispanic 4 Autism diagnosis

Asperger syndrome 21 Classical autism 71 Pervasive developmental

disorder not otherwise specified 8

Mental retardation 20 Insurance coverage

Private insurance only 58 Medicaid but no private insurance 21 Medicaid and private insurance 8 Public insurance

(other than Medicaid) only 7

No major treatment approach 32

Family characteristics: Household composition

Two-parent household 77 Single parent household 22 Extended family household 9 Siblings – 1 (0.8)

Education Less than high school 1 High school degree 37 College degree 35 Graduate degree 27

Annual household income Less than $35,000 28 $35,000–$49,999 20 $50,000–$74,999 26 $75,000 or above 26

Well-being Family stress score – 10 (5.5)

123

J Autism Dev Disord (2007) 37:1902–1912 1905

Each model had a significant fit according to a

likelihood ratio test (a = 0.05). Table 4 omits models for 4 services (after school care, psychiatrist, parent

support groups, and occupational therapy) used

outside school by over 10 percent of the sample

because the overall fit was not significant or the

model did not converge.

Predisposing Characteristics

Racial and ethnic minority families had half the odds

(OR = 0.48) of using a case manager, and only a

quarter the odds of using a psychologist (OR = 27),

developmental pediatrician (OR = 28), and sensory

integration (OR = 25). When parents had a college or

graduate degree, families had 2 to nearly 4 times the

odds of using a neurologist, the Picture Exchange

Communication System (PECS) and hippotherapy/

therapeutic horseback riding. When parents had higher

levels of stress, families had slightly higher odds

(OR = 1.1) of using a number of services: summer

camp or respite care, a case manager, medication and

supplements, PECS and hippotherapy/therapeutic

horseback riding. When families did not identify any

major treatment approach for ASD, they had between

one half and one fifth the odds of using care from

family or friends, PECS, parent training classes, sen-

sory integration and casein/gluten free diets. When

families were living in nonmetropolitan areas, they had

lower odds of using summer camp (OR = 0.33) and

respite care (OR = 0.21).

Enabling Characteristics

When children were covered by Medicaid or other

public insurance, families had from 2 to 11 times the

odds of using services that could be considered

medically necessary (medication) as well as therapeutic

support services (respite care, case manager, PECS,

speech/language therapy) compared to families whose

children were covered by private insurance. Con-

versely, when children were covered by Medicaid or

other public insurance, families had only one quarter

the odds of using supplements. When children did not

have health insurance, families had much higher odds

of using a case manager (OR = 4.94) and a develop-

mental pediatrician (OR = 16.60). When children did

not have health insurance, families did not have lower

odds of using any other services. Families that had an

annual income above $50,000 had higher odds of using

a developmental pediatrician and speech/language

therapy.

Need Characteristics

Families of children with Asperger syndrome had twice

the odds of those of children with classical autism of

using medication and one third or less the odds of using

PECS or casein/gluten free diets. Families of children

with mental retardation had twice the odds of using

respite care, a case manager, and sensory integration

therapy, and one quarter the odds of using a psychol-

ogist. Families of children 4 years old or less had at

least twice the odds of families whose children were

5–8 years old of using services that help define a

diagnosis and plan of treatment (case manager, devel-

opmental pediatrician), that help with communication

(PECS, speech/language therapy), as well as supple-

ments. They had half the odds or less of using services

for school-age children (special summer camp) a

psychologist, medications, and social skills training.

Families whose children were 9–11 years old had over

twice the odds of using respite care and a third or less

the odds of using PECS and sensory integration.

Table 2 Major treatment approaches for ASD

Age group:

£4 years 5–8 years 9–11 years

Approach Percent usinga: Anyc Alonec Any Alone Anyc Alonec

TEACCH 55 36 59 44 62 51 Applied Behavioral Analysis (ABA)b 22 5 16 3 9 4 Floor Time 17* 5* 7 0 5 0 Defeat Autism Now (DAN) 13 0 7 0 5 0 Lovaas 11 0 5 0 0 0 None of the above 24 na 35 na 33 na

a Percents do not add to 100 because families may use more than one approach; cells with fewer than 5 families coded as zero b Any approach based on ABA other than the Lovaas approach c v2 test of difference in percent of 5–8 yr olds using given approach compared to younger and older groups, df = 1, * significant difference a = 0.01

123

1906 J Autism Dev Disord (2007) 37:1902–1912

Discussion

In sum, families and their children with ASD used a

broad array of services; the breadth of services used was

highest for children aged 5–8 years. Access to care was

limited for racial and ethnic minority families, those with

low levels of education, those who were not using a

major treatment approach, and those living in nonme-

tropolitan areas. Family use of a major treatment

approach or multiple approaches for ASD was lower

for older age groups of children, although the majority of

families of children in the older age group (9–11 years

old) still used one approach. When parents had higher

levels of stress, the odds of service use were higher.

Medicaid and more family income also increased the

odds of use. When children did not have any health

Table 3 Autism-related service use: In school and outside school by age group

Age in years Age in years

Service Percent using b : £4 5–8 9–11 Service Percent usingb: £4 5–8 9–11

In school a

Speech/language therapy 91 79 65 Occupational therapy 60 66 42 Social skills training 29 28 46 Physical therapy 9 11 6 Adaptive physical education 4 13 16 Music therapy 6 8 6 Audiologist 0 2 0

Outside school Child care services Social therapies

Care from family or friends 64 58 55 Social skills training 6 16 24 Special summer camp 6 16 24 Hippotherapy/therapeutic riding 7 13 7 Respite care 14 10 22 Play therapy 16 7 6 After school care 0 12 13 Music therapy 8 6 0 Day care 12 3 0 Holding therapy 0 2 0 Residential placement 0 0 0 Dog therapy 0 2 0

Dolphin therapy 0 0 0 Other specialist providers Aversive 0 0 0

Case manager 41 24 35 Neurologist 15 18 16 Sensory/motor therapies Developmental pediatrician 20 10 9 Sensory integration therapy 22 22 12 Psychologist 5 13 23 Occupational therapy 21 11 11 Psychiatrist 0 14 24 Auditory integration 0 3 4 Behavioral specialist 9 8 5 Physical therapy 4 0 0 Therapeutic Support Person 0 6 9 Craniosacral trt, myofacial release 0 0 0 Personal Care Assistant 0 4 18 Squeeze machine 0 0 0 Audiologist 0 2 0

CAM therapies Medications & supplements Casein free diet 19 9 6

Medication 36 52 68 Gluten free diet 18 7 6 Supplements 26 13 15 Feingold diet 6 2 0

Specialized eye glasses 0 3 5 Communication therapies/systems Enzyme potentiated desensitization 0 2 0

Picture exchange communication 35 18 10 Immune system therapy 0 0 0 Speech/language therapy 29 15 10 Secretin 0 0 0 Facilitated communication 0 2 0 Acupuncture 0 0 0 Fast for word computer program 0 2 0 Cranial electrical stimulation 0 0 0

Flexyx neurotherapy system 0 0 0 Family services

Parent support groups 36 29 31 Parent training classes 7 12 15 Family counseling 0 6 11 Sibling support groups 0 0 4

a n = 85, 186, 81 children in preschool or school ages £4, 5–8, 9–11 years respectively b Cells with fewer than three families rounded to 0

CAM = complementary and alternative medicine

123

J Autism Dev Disord (2007) 37:1902–1912 1907

insurance, the odds of their family receiving services that

facilitated entry into the system of care was increased.

The mix of services families use varied with the child’s

ASD diagnosis and age, and findings indicate a higher

rate of Asperger syndrome in older children. These

findings are consistent with prior work (Ruble et al.,

2005; Kraus et al., 2003; Birenbaum & Cohen, 1993;

Birenbaum, Guyot & Cohen, 1990), but expand the view

of access to incorporate a wide range of autism-related

services. Disparities in service use associated with race,

residence and education point to the need to develop

policy, practice and family-level interventions that can

address barriers to services for all children with ASD.

Generalizability

This study was based on a community sample of

families in North Carolina which is widely considered

to have a comprehensive service system for young

children with ASD. This service system is supported by

Division TEACCH which provides assessment and

referral services through nine regional centers across

the state. TEACCH also provides training for parents

and teachers. To the extent that TEACCH facilitates

access to ASD services, analysis of access to care in

North Carolina may hide certain barriers to care that

are important in non TEACCH environments. From

this perspective, this analysis provides a conservative

view of barriers to care for ASD. On the other hand, it

is important to note that about 40 percent of sample

families reported that they were not using TEACCH.

Some of these families were using other major treat-

ment approaches, but most were not using any

approach. Study recruitment with the help of schools

and state assessment agencies was effective in bringing

in families outside of TEACCH. Although these

families may have benefited from TEACCH influence

throughout the state, they provided variation in the

parent perspective on major treatment approach and

how that approach impacts decision-making about

patterns of service use.

It is important to keep in mind that these findings

were derived from a sample of families who volun-

teered to participate and were identified because of

their use of services or connection to ASD information

resources. These were families who had the time to

devote energy to completing a survey. Reports from a

treated sample are likely to overestimate the use of

services compared to the general ASD population.

This sample was likely to be missing families with few

disposable resources, families who might be unsure if

their child had ASD, who were not well-connected

Table 4 Factors associated with use of services, by service

Variable C a re

f ro

m F

a m

ily

o r

F ri e n d s

S p e ci

a l S

u m

m e r

C a m

p

R e sp

ite C

a re

C a se

M a n a g e r

N e u ro

lo g is

t

D e ve

lo p m

e n ta

l P

e d ia

tr ic

ia n

P sy

ch o lo

g is

t

M e d ic

a tio

n

S u p p le

m e n ts

P ic

tu re

E xc

h a n g e

C o m

m u n ic

a tio

n

S p e e ch

/L a n g u a g e

T h e ra

p y

P a re

n t

T ra

in in

g

C la

ss e s

S o ci

a l S

ki lls

T

ra in

in g

H ip

p o th

e ra

p y/

T h e ra

p e u tic

R id

in g

S e n so

ry I

n te

g ra

tio n

T h e ra

p y

C a se

in /G

lu te

n F

re e

D ie

ts

Predisposing Characteristics

Racial, ethnic minority 0.48 (0.24, 0.94)

0.27 (0.08, 0.86)

0.28 (0.10, 0.80)

0.25 (0.10, 0.62)

College/grad degree 2.76 (1.28, 5.99)

2.19 (1.00, 4.78)

3.93 (1.31, 11.77)

Family stress score 1.07 (1.01, 1.14)

1.08 (1.01, 1.16)

1.07 (1.02, 1.13)

1.08 (1.03, 1.13)

1.07 (1.01, 1.13)

1.07 (1.01, 1.14)

1.10 (1.02, 1.19)

No major trt approach 0.55 (0.33, 0.89)

0.19 (0.07, 0.48)

0.29 (0.11, 0.78)

0.27 (0.12, 0.62)

0.34 (0.12, 0.91)

Nonmetropolitan area 0.33 (0.11, 0.97)

0.21 (0.06, 0.78)

Enabling Characteristics

Medicaid/public 5.65 (2.20, 14.53)

11.26 (5.20, 24.42)

2.09 (1.09, 3.99)

0.25 (0.09, 0.68)

3.60 (1.56, 8.31)

3.55 (1.54, 8.18)

No health insurance 4.94 (1.85, 13.19)

16.60 (5.18, 53.21)

Income ≥ $50,000 3.53 (1.27, 9.83)

2.49 (1.10, 5.62)

Need Characteristics

Asperger syndrome 2.11 (1.12, 3.98)

0.32 (0.10, 0.98)

0.26 (0.07, 0.94)

Mental retardation 2.48 (1.11, 5.53)

2.23 (1.13, 4.40)

0.24 (0.08, 0.75)

2.09 (1.01, 4.32)

Age 4 years or less 0.33 (0.13, 0.86)

2.83 (1.48, 5.41)

2.78 (1.19, 6.52)

0.33 (0.12, 0.95)

0.53 (0.31, 0.93)

2.24 (1.11, 4.50)

2.09 (1.08, 4.04)

2.49 (1.29, 4.83)

0.38 (0.15, 0.97)

Age 9-11 years 2.44 (1.01, 5.90)

0.24 (0.08, 0.69)

0.38 (0.15, 0.92)

Table contains odds ratios and 95% confidence intervals from significant predictors in logit models predicting the likelihood of any service use. Each model was estimated with the full set of predictors listed; single parent, extended family, and PDD-NOS were also included but not presented for lack of effect.

The reference category includes children who are white, have private health insurance, have classical autism, no mental retardation, and are aged 5-8 years old, with parents with a high school degree or less, who use one of the major treatment approaches, live in a metro area, and have income below $50,000

123

1908 J Autism Dev Disord (2007) 37:1902–1912

with a system of providers, and/or whose child with

ASD was so needy that they did not have the time to

volunteer to participate in a survey. For example, the

sample rate of mental retardation was low (Fombonne,

2002). When children have mental retardation their

needs are greater, and their families may be less apt to

participate in research. To the extent that the sample

had variation in family and child predisposing, enabling

and need characteristics, it was still useful for identi-

fying families and children who did not enjoy full

access to autism-related services. Barriers to care may

be greater in the population.

Future Research

These data showed differences in service use by age

group of the child with ASD. Family use of major

treatment approach(es) for ASD decreased as age of

the group increased, while the breadth of services used

was highest for children aged 5–8 years, the middle age

group. With cross-sectional data, it is not possible to

tell if this represents an age or cohort effect. Future

prospective study of children as they age and transition

through different stages of school into adulthood are

needed to begin to understand the patterns of service

use as well as how those patterns impact child and

family outcomes.

Findings describe significantly lower odds of service

use among racial and ethnic minority families. The

issues related to access and use of mental health

services for minority children are multifaceted involv-

ing economic factors, service sector factors and cultural

factors. Lack of appropriate outreach and cultural

competency of providers are likely to contribute

significantly to low service use (e.g. Lau et al., 2004;

Edwards, 2003). Institutionalized discrimination lead-

ing to a general mistrust of the system, religious and/or

spiritual beliefs, and stigma are also likely to be

important (Schnittker, 2003; Edwards, 2003). Since

the caregivers of racial and ethnic minority children act

as the initial gatekeepers for services, reducing the

barriers to access and use of services that minority

families perceive as especially important, and that they

can control, will be key to bridging racial disparities in

service use. Future work that is able to articulate racial

and ethnic minority families’ perceptions of, attitudes

toward and experiences within the system of care for

ASD, and that relates those attitudes and beliefs to

levels of service use, will be crucial.

Families with more education and higher levels of

stress had higher odds of using services. Education may

equip families with the tools they need to advocate

successfully for services for their child. Higher levels of

family stress may be a result of having a child with more

severe disability (Tobing & Glenwick, 2002; Hastings &

Johnson, 2001) who requires more services. Or, high

levels of family stress may lead families to use more

parent-oriented support services such as respite care. It

is likely that both factors are at play in these findings,

since family stress was associated with use of child-

oriented services such as medications, as well as parent-

oriented services such as respite care. Families would

benefit from future work that identifies both the skills

parents need to advocate and successfully navigate the

complex system of social and health services in order to

obtain services for their children with ASD as well as the

supports families need to maintain their own emotional

health as they devote their energies to their child.

Families that did not use a major treatment approach

had more limited access to services. Previous analysis of

these data indicated that treatment approach was

associated with the constellation of services that a

family uses (Thomas et al., 2006). Future work needs to

develop our understanding of the pathways to diagnosis

and care for ASD, how families end up choosing one

fork or treatment approach in that path, and the

ramifications of those choices and barriers faced in the

help-seeking process (Osgood et al., 2005). For exam-

ple, future research could explore the association

between major treatment approach and service referral

patterns to see how referrals vary by approach and how

those impact the pattern of services used.

Findings identify disparities in service use associated

with residence in a nonmetropolitan area. Future

research needs to assess variation in the supply of

autism-related services across urban and rural areas,

determine the impact of shortages and strategies to

alleviate them.

Higher income increases the odds of service use, but

only for two services (developmental pediatrician,

speech/language therapy). Data are needed to assess

how much families pay out-of-pocket for ASD services,

which services they pay for primarily out-of-pocket,

and following this, which services appear to be out of

reach of families with lower income. Future research

on the cost of ASD and the family share of that cost

will help to develop strategies to remove disparities in

service use based on ability to pay.

Findings here indicate that Medicaid did a good job in

improving access to care. Better information is needed

to understand how children with ASD achieve Medicaid

eligibility across states, which eligible children with ASD

are enrolled, and how to reach eligible children who are

not enrolled (e.g. Stuber & Bradley, 2005; Davidoff,

Yemane. & Hill, 2004). Surprisingly, children with no

health insurance had higher odds of using a case

123

J Autism Dev Disord (2007) 37:1902–1912 1909

manager and developmental pediatrician. These may be

children who were new to services and appropriately

connected with a case manager. Ideally, the case

manager would facilitate enrollment in Medicaid and

access to services. Future analysis of the pathways into

and through the service system would help to develop an

understanding of how families with few financial

resources fare in the system of care for ASD.

The association of measures of child need and

service use showed a general pattern suggesting that

service use fits children’s needs. For example, children

with mental retardation had higher odds of using

support services (respite, case manager). Younger

children had higher odds of using services that help

with communication (PECS, speech/language therapy),

while older children had higher odds of using services

that provided social activities (special summer camp,

social skills training). One finding that stands out from

this pattern is that children with Asperger syndrome

had higher odds of using medications. Future work that

clarifies the variation in needs and service supports of

children at different places on the ASD spectrum, and

how those needs change as children age and develop,

will provide an important foundation from which to

assess problems of barriers to care and unmet needs.

A multi-state sample of families is needed to address

a number of issues raised here. As ASD subject

registries develop across the country (e.g. UNC, the

Waisman Center, Virginia Commonwealth University),

combining these samples may provide a viable way to

construct a multi-state sample. For example, such a

sample could make possible recruitment of sufficient

numbers of racial and ethnic minority families to study

barriers to care; it could support analysis of the

variation in use of major treatment approaches across

the U.S. and associated patterns of service use; and it

would allow exploration of the impact of variation in

state Medicaid policies on Medicaid-reimbursed ser-

vices for children with ASD. Recruitment strategies

must balance concerns about family burden, recruiter

burden (for example when schools, state agencies or

advocacy groups assist with recruitment) and costs

against the goal of obtaining a quality sample. Explo-

ration of recruitment burden for various stake-holders

and target populations as well as the yield from

differing recruitment strategies would provide a foun-

dation for pursuing wider recruitment efforts.

Policy Implications

Findings from this study underscore the importance of

recommendations of the Interagency Autism Coordi-

nating Committee to take into account changes in

needs as children develop across their lifespan (IACC,

2005). Findings suggest that children received a diag-

nosis of Asperger syndrome at later ages. This is

consistent with prior work (Howlin & Asgharian,

1999). These data do not indicate if this is a fine-

tuning of an earlier ASD diagnosis, or if some children

do not receive any ASD diagnosis until they are close

to middle school. If this latter scenario is true, this

points to a need for increased efforts to identify

children with high functioning ASD earlier. The

finding that children with Asperger syndrome had

higher odds than those with classical autism of using

medication suggests that high functioning children with

ASD may need more or different types of services to

support them in the most inclusive settings. Policy-

makers need to keep in mind that the need for services

is not necessarily proportional to the severity of the

ASD diagnosis. Altogether, these findings reinforce the

importance of developing a better understanding of

how children’s needs and service use change as they

age, develop and encounter life’s transitions.

These analyses highlight how racial and ethnic

minority families, those in living in nonmetropolitan

areas and those with limited education achieve only

limited access to care for ASD. Unfortunately, these

disparities are typical of all children with and without

special health care needs (e.g. Newacheck, Hung &

Wright, 2002; Stevens & Shi, 2003). The findings

reported here serve to underscore the importance of

including children with ASD in efforts to develop

policy, practice and family-level interventions that can

address these barriers to care.

Acknowledgments Support for this research was provided by a grant from the National Institute of Mental Health (R21 MH066143) and by funding from the Centers for Disease Control through the North Carolina Center for Autism and Developmental Disabilities Research and Epidemiology. The authors wish to express their appreciation for the comments of three anonymous reviewers and the support and insight of Robin McWilliam, Vanderbilt University; Division TEACCH, especially Gary Mesibov and Ann Palmer; and the Neurodevelopmental Disorders Research Center, especially Joseph Piven and Renee Clark, at the University of North Carolina at Chapel Hill. We thank the North Carolina Public Schools, the Autism Society of North Carolina, North Carolina Families for Early Autism Treatment, the Children’s Developmental Services Assessment Centers, and the Survey Research Unit, University of North Carolina at Chapel Hill for their assistance. We especially wish to thank the families who participated in the study.

References

Aday, L., & Andersen, R. (1975). Development of indices of access to medical care. Ann Arbor, MI: Health Administra- tion Press.

123

1910 J Autism Dev Disord (2007) 37:1902–1912

Aday, L. (1989). Designing and conducting health surveys. San Francisco, CA: Jossey-Bass.

Birenbaum, A., Guyot, D., & Cohen, H. (1990). Health care financing for severe developmental disabilities (Vol. 14). Washington, DC: American Association on Mental Retar- dation.

Birenbaum, A., & Cohen, H. (1993). On the importance of helping families: policy implications from a national study. Mental Retardation, 31(2), 67–74.

Bodfish, J. (2004). Treating the core features of autism: are we there yet? Mental Retardation and Developmental Disabil- ities Research Reviews, 10(318–326).

Campbell, M., Schopler, E., Cueva, J., & Hallin, A. (1996). Treatment of autistic disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 35(134–143).

Carey, B. (2004, December 27). To treat autism, parents take a leap of faith. The New York Times.

Davidoff, A., Yemane, A., & Hill, I. (2004). Public insurance eligibility and enrollment for special health care needs children. Health Care Financing Review, 26(1), 119–135.

Dawson, G., & Osterling, J. (1997). Early intervention in autism. In Guralnick (Ed.), The Effectiveness of early intervention (pp. 307–326). Baltimore, MD: Paul H. Brookes.

Dunn, M., Burbine, R., Bowers, C., & Tantleff-Dunn, S. (2001). Moderators of stress in parents of children with autism. Community Mental Health Journal, 37(1), 39–52.

Edwards, C. (2003) Diversity and cultural competence, Part 2: Issues of practical conceptualization and implementation. The North Carolina Psychologist, November-December 2003.

Factor, D., Perry, A., & Freeman, N. (1990). Brief report: social support, and respite care use in families with autistic children. Journal of Autism and Developmental Disability, 20(1), 139–146.

Fombonne, E. (2002). Epidemiological trends in rates of autism. Molecular Psychiatry, 7, S4–S6.

Fombonne, E. (2003). Epidemiological surveys of autism and other pervasive developmental disorders: an update. Journal of Autism and Developmental Disability, 33, 365–382.

Friedrich, W., Greenberg, M., & Crnic, K. (1983). A short-form of the questionnaire on resources and stress. American Journal of Mental Deficiency, 88, 41–48.

Green, V., Pituch, K., Itchon, J., Choi, A., O’Reilly, M., & Sigafoos, J. (2006). Internet survey of treatments used by parents of children with autism. Research in Developmental Disabilities, 27, 70–84.

Gross, J. (2004, January 30). As autism cases rise, parents run frenzied race to get help. The New York Times.

Hare, D., Pratt, C., Burton, M., Bromley, J., & Emerson, E. (2004). The health and social care needs of family carers supporting adults with autistic spectrum disorders. Autism, 8(4), 425–444.

Hastings, R., & Johnson, E. (2001). Stress in UK families conducting intensive home-based behavioral intervention for their young child with autism. Journal of Autism and Developmental Disability, 31(3), 327–336.

Howlin, P., & Asgharian, A. (1999) The diagnosis of autism and asperger syndrome: findings from a survey of 770 families. Developmental Medicine and Child Neurology, 41(12), 834– 839.

Hurth, J., Shaw, E., Izeman, S., Whaley, K., & Rogers, S. (1999). Areas of agreement about effective practices among pro- grams serving young children with autism spectrum disor- ders. Infants and Young Children, 12, 17–26.

Interagency Autism Coordinating Committee. (2005). Autism Spectrum Disorders (ASD) Roadmap.

Kohler, F. (2000). Examining the services received by young children with autism and their families: a survey of parent responses. Focus on Autism and Other Developmental Disabilities, 14, 150–158.

Kraus, M., Gulley, S., Sciegaii, M., & Wells, N. (2003). Access to specialty medical care for children with mental retardation, autism, and other special health care needs. Mental Retar- dation, 41(5), 329–339.

Lau, A., Garland, A., Yeh, M., et al. (2004). Race/ethnicity and inter-informant agreement in assessing adolescent psyhopa- thology. Journal of Emotional and Behavioral Discorders 12, 145–156.

Levy, S., Mandell, D., Merher, S., Ittenbach, R., & Pinto-Martin, J. (2003). Use of complementary and alternative medicine among children recently diagnosed with autistic spectrum disorder. Developmental and Behavioral Pediatrics, 24, 418–423.

Luther, E., Canham, D., & Cureton, V. (2005). Coping and social support for parents of children with autism. Journal of School Nursing, 21(1), 40–47.

Mandell, D., Listerus, J., Levy, S., & Pinto-Martin, J. (2002). Race differences in the age at diagnosis among medicaid- eligible children with autism. Journal of the American Academy of Child and Adolescent Psychiatry, 41(12), 1447–1453.

Marcus, L., Garfinkle, A., & Wolery, M. (2001). Issues in early diagnosis and intervention with young children with autism. In E., Schopler, N., Yirmiya, C., Schlman, & L., Marcus (Eds.), The Research Basis for Autism Intervention. New York: Kluwer Academic/Plenum.

McEachin, J., Smith, T., & Lovaas, O. (1993). Long-term outcome for children with autism who received early intensive behavioral treatment. American Journal of Mental Retardation, 97, 359–372.

Newacheck, P., Hung, Y., & Wright, K. (2002). Racial and ethnic disparities in access to care for children with special health care needs. Ambulatory Pediatrics, 2(4), 247–254.

Osgood, D., Foster, F., Flanagan, C., & Ruth, G. (2005). On your own without a net: the transition to adulthood for vulnerable populations. Chicago: University of Chicago Press.

Rhoades, J., & Cohen, J. (2004). The uninsured in America, 1996–2003: Estimates for the U.S. population under age 65 (Medical Expenditures Panel Survey Statistical Brief #45). Rockville, MD: Agency for Healthcare Research and Quality.

Rogers, S. (1998). Neuropsychology of autism in young children and its implications for early intervention. Mental Retarda- tion and Developmental Disabilities Research Reviews, 4, 104–112.

Rousey, A., Best, S., & Blacher, J. (1992). Mothers’ and fathers’ perceptions of stress and coping with children who have severe disabilities. American Journal of Mental Retardation, 97(1), 99–110.

Ruble, L., Heflinger, C., Renfrew, J., & Saunders, R. (2005). Access and service use by children with autism spectrum disorders in medicaid managed care. Journal of Autism and Developmental Disorders, 35(1), 3–13.

Schnittker, J. (2003) Misgivings or medicine?: african Americans’ skepticism of psychiatric medication. Journal of Health and Social Behavior, 44, 506–524.

Smith, T., & Antolovich, M. (2000). Parental perceptions of supplemental interventions received by young children with autism in intensive behavior analytic treatment. Behavioral Interventions, 15, 83–97.

Stevens, G., & Shi, L. (2003). Racial and ethnic disparities in the primary care experiences of children: a review of the liter- ature. Medical Care Research and Review, 60(1), 3–30.

123

J Autism Dev Disord (2007) 37:1902–1912 1911

Stuber, J., & Bradley, E. (2005). Barriers to medicaid enrollment: who is at risk? American Journal of Public Health, 95(2), 292–298.

Svarstad, B., Chewning, B., Sleath, B., & Claesson, C. (1999). The brief medication questionnaire: a tool for screening patient adherence and barriers to adherence. Patient Edu- cation and Counseling, 37(2), 113–124.

Thomas, K., Morrissey, J., & McLaurin, C. (2006). Use of autism- related services by families and children. Journal of Autism and Developmental Disorders, forthcoming.

Tobing, L., & Glenwick, D. (2002). Relation of the child- hood autism rating scale-parent version to diagnosis, stress and age. Research in Developmental Disabilities, 23, 211–223.

U.S. Census. (2000). Profile of Selected Economic Characteris- tics: 2000 (Census 2000 Summary File 3). Washington, DC: U.S. Census.

U.S. Census. (2001). Persons by race, age and sex 2000: North Carolina. Washington, DC: U.S. Census.

U.S. Census. (2003). Median Family Income in the Past 12 months (2003 Inflation-Adjusted Dollars) By Family Size: North Carolina. Washington, DC: U.S. Census.

USDA (2004) Measuring rurality: rural-urban continuum codes, http://www.ers.usda.gov/Briefing/Rurality/RuralUrbCon/, accessed 8–2–06.

Wieder, S., & Greenspan, S. (2003). Climbing the symbolic ladder in the DIR model through floor time/interactive play. Autism, 7, 425–435.

123

1912 J Autism Dev Disord (2007) 37:1902–1912

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.