665792

research-article2016
AUT0010.1177/1362361316665792AutismColombi et al.

Original Article

Autism

Implementation of the Early Start 1­–8

© The Author(s) 2016
Reprints and permissions:
Denver Model in an Italian community sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1362361316665792
aut.sagepub.com

Costanza Colombi1, Antonio Narzisi2, Liliana Ruta3,

Virginia Cigala4, Antonella Gagliano4, Giovanni Pioggia3,
Rosamaria Siracusano4, Sally J Rogers5,
Filippo Muratori2,6 and Prima Pietra Team

Abstract
Identifying effective, community-based specialized interventions for young children with autism spectrum disorder is an
international clinical and research priority. We evaluated the effectiveness of the Early Start Denver Model intervention
in a group of young children with autism spectrum disorder living in an Italian community compared to a group of
Italian children who received treatment as usual. A total of 22 young children diagnosed with autism spectrum disorder
received the Early Start Denver Model in a center-based context for 6 h per week over 6 months. The Early Start Denver
Model group was compared to a group of 70 young children diagnosed with autism spectrum disorder who received
treatment as usual for an average of 5.2 h over 6 months. Children in both groups improved in cognitive, adaptive, and
social skills after 3 months and 6 months of treatment. Children in the Early Start Denver Model group made larger gains
in cognitive and social skills after 3 and 6 months of treatment. The Early Start Denver Model group made larger gains in
adaptive skills after 3 months of treatment. Our results are discussed in terms of implications for intervention research
and clinical practice. Our study supports the positive impact of the Early Start Denver Model in a non-English-speaking
community.

Keywords
autism, autism spectrum disorder, early intervention, Early Start Denver Model, preschoolers

Introduction
There is strong evidence of the positive impact of early (Smith et al., 2007). Studies of generalizability to commu-
intervention that starts immediately after a child receives a nity practice are limited due to the difficulties involved in
diagnosis of autism spectrum disorder (ASD) (Koegel selection of participants and clinicians. Community effec-
et al., 2014; Reichow et al., 2008). The body of evidence tiveness studies test whether an intervention is effective
supporting the positive effects of early intervention for when administered in community settings without rigorous
children with ASD mainly comes from university-based research oversight (Smith et al., 2007). In effectiveness
efficacy studies involving high levels of staff training and studies, it is not possible, nor desired, to maintain the level
supervision and high-intensity services (Vivanti et al.,
2014). Additionally, efficacy studies usually require rand-
omization of participants into experimental and control 1University of Michigan, USA
2IRCCS Stella Maris Foundation, Italy
conditions as well as blind evaluation of treatment out-
3National Research Council, Italy
comes (Bowen et al., 2009). Randomization is often 4University of Messina, Italy
impossible to conduct in community intervention due to 5University of California, Davis, USA
the policies regulating the delivery of services. Moreover, 6University of Pisa, Italy

blind evaluations are not usually conducted in clinical

Corresponding author:
practice outside randomized control trials (RCTs). Costanza Colombi, University of Michigan, 4250 Plymouth Rd, Ann
Efficacy studies employing RCTs test whether a given Arbor, MI 48103, USA.
intervention is efficacious under controlled conditions Email: [email protected]

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

2 Autism

of experimental control that is characteristic of randomized implementation of the various components of the ESDM
clinical trials such as randomization of participants and intervention, such as systematic assessment, writing objec-
strict inclusion and exclusion criteria since the goal is tives, and monitoring progress based on data, may pose
to determine whether an intervention would be effective obstacles in intervention contexts where professionals are
as practiced by community clinicians. Due to the com- not used to implementing manualized treatments.
plexity involved in conducting good quality effectiveness In most areas of Italy, children with ASD receive low-
studies, very few ASD-specialized early interventions intensive, non-specialized treatment. In general, the Italian
have been evaluated through community studies (Divan Public Health System offers between 2 and 6 h of interven-
et al., 2015). tion composed of speech therapy, educational services,
A priority in the ASD field is to evaluate whether spe- and psychomotricity. Psychomotricity is a therapeutic
cialized interventions proven efficacious in university approach that aims to support an individual’s personal
RCTs are effective when implemented in communities development (Aucouturier, 2005). It is based on a holistic
around the world. Many communities lack the resources view of human beings that considers each individual as a
needed for the implementation of rigorous efficacy studies. unity of physical, emotional, and cognitive actualities,
An additional challenge for many non-English-speaking which interact with each other and the surrounding social
communities is that most evidence-based early intervention environment (Aucouturier, 2005). It is a system of physi-
models for young children with ASD have been developed cal education which uses motion as an educational means
in English-speaking countries. Using an intervention devel- (Barruezzo, 2000). Psychomotricians are licensed profes-
oped in a different country with a different language and sionals in Italy and in other European countries. However,
different cultural contexts implies the translation of the psychomotricity’s efficacy and effectiveness for young
intervention manual, as well as adaptations of procedures children with ASD have not been evaluated through
to the specific culture of a given community. research studies.
This study evaluates the effectiveness in an Italian com- The primary goal of our study was to evaluate the effec-
munity of the Early Start Denver Model (ESDM; Rogers tiveness of the ESDM delivered in an Italian community
and Dawson, 2010). The ESDM is an empirically based, by comparing the outcomes of the children who received
manualized, play- and routine-based intervention that the ESDM and those of children who received treatment as
fuses developmental and relationship-based methods with usual (TAU) over a 6-month period. Our study attempts to
principles and practices of applied behavior analysis. The evaluate whether ESDM is an effective, evidence-based
ESDM is delivered by adults within the context of play treatment for ASD in young Italian children who lack local
and daily routines in which highly precise naturalistic access to high-quality resources specific for ASD. We
behavioral teaching is embedded, making this one of hypothesized that children in the ESDM group would
the Naturalistic Developmental Behavioral Interventions demonstrate higher gains in language, cognitive, and daily
(NDBI; Schreibman et al., 2015). Dawson et al. (2010) living skills.
conducted an RCT to investigate the effects of the ESDM The secondary aim of our study was to evaluate feasibil-
in young children with ASD. Children who received the ity of ESDM implementation in the context of the Italian
ESDM, compared to children who received available treat- community. Feasibility was evaluated based on guidelines
ment in the community, demonstrated greater gains in IQ, outlined in Bowen et al. (2009) and used in Vivanti et al.
receptive and expressive language, social skills, and adap- (2014). The following indicators were included: acceptabil-
tive behavior, and greater decreases in parent-reported ity (is the intervention judged suitable by the individuals
symptoms of ASD (Dawson et al., 2012). involved in the program?), demand (to what extent is the
The ESDM has been implemented in a variety of program likely to be chosen?), implementation (the degree
settings, including intensive autism-specialty delivery of execution of the intervention based on manualized proce-
(Dawson et al., 2010), day care centers and preschools dures), practicality (the extent to which delivery of the pro-
(Vivanti et al., 2013, 2016), parent education (Rogers et al., gram can be implemented in a specific context), adaptation
2012; Vismara et al., 2009), and telehealth (Vismara et al., and integration (the extent to which the intervention can be
2013, 2016). Moreover, the ESDM has been implemented implemented within the existing system). Documentation of
by a variety of professionals such as psychologists, occupa- feasibility has been conducted through analysis of service
tional therapists, early childhood educators, behavioral utilization and external evaluation of the program.
analysts, and speech therapists (Rogers and Dawson, 2010).
Because it can be implemented in a variety of contexts and
by a variety of professionals, the flexibility of the ESDM Methods
may help to respond to the challenges of translating an evi-
Design
dence-based intervention into a community with a different
culture, language, intervention setting, and professionals’ This is a quasi-experimental treatment study of the
backgrounds. On the other hand, the rigor required in the ESDM in an underserved population: children in the

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

Colombi et al. 3

Messina area (Sicily, Italy) who have access to limited, Treatment teams.  Treatment teams for both groups com-
non-specialized intervention (2–6 h per week). Although prised professionals with background and credentials
standards of treatment for ASD slightly vary across Italy, representative of the services across Italy. Each treat-
health and social care services provided in Sicily (2–6 h a ment team included the following professionals: (a)
week of speech and psychomotor intervention) are repre- child neuropsychiatrists, medical doctors with special-
sentatives of the existing services nationwide. The vast ized training in psychiatric and neurological disorders in
majority of Italian children with ASD are underserved children aged between 0 and 18 years; usually in Italy,
based on guidelines of the National Research Council child neuropsychiatrists lead assessment teams, manage
(2001) recommending 25 h per week of specialized inter- medical conditions, and prescribe and monitor interven-
vention for young children with ASD. tion; (b) clinical psychologists, master-level licensed
The design involved recruiting the first children who psychologists with specialized training in assessment
qualified in the Messina area for the experimental group, and intervention; (c) psychomotricians, bachelor-level
and then recruiting a well-matched comparison group therapists who practice psychomotricity; (d) speech and
through public health centers in Italy. Both groups were language pathologists, bachelor-level therapists who
assessed prior to intervention and at 3 and 6 months after focus on speech development; (e) special educators,
the start of intervention. In both groups, baseline assess- bachelor- or master-level therapists.
ments were administered approximately 2 weeks before
the start of the intervention. The decision to employ a Participants.  Inclusion criteria for children enrolled in the
non-randomized comparison group was made for ethical study for both groups were as follows: (a) 18–48 months of
reasons, given the lack of access to specialized services age at the time of enrollment, (b) diagnosis of ASD made
of these families. The aim of the study was to test the by a qualified professional using standard valid and relia-
effectiveness of ESDM, an already-established effica- ble assessment tools including the Autism Diagnostic
cious intervention, for young children with ASD, and the Observation Scale, (c) agreement to participate in the
feasibility of its implementation by professionals and intervention for 6 months, (d) Italian as the primary lan-
families living in a different culture and speaking a dif- guage spoken at home (as therapists had not been trained
ferent language. to deliver the treatments in another language), and (e)
Families who participated in the ESDM intervention hearing and vision screen within the normal range, ability
were recruited through the Neuropsychiatric Department to use hands, and ambulatory. Exclusion criteria included
of the University of Messina, Italy, which provides diag- the following: (a) any other identifiable genetic condition
nostic services to all children and adolescents with sus- associated with autism (e.g. Fragile X syndrome and Down
pected neurological or neurodevelopmental disorders, syndrome), (b) head trauma, (c) known neurological dis-
after pediatricians’ referral. Additionally, the Department ease (e.g. encephalitis), (d) epilepsy, and (e) significant
provides intervention services to local children with sensory or motor impairment (e.g. cerebral palsy).
developmental and psychiatric disorders. The first 22 The ESDM group comprised 22 children between the
children diagnosed with ASD between the age of 18 and ages of 18 and 48 months who had received an ASD diag-
48 months, who qualified for the study and were inter- nosis. The TAU group comprised 70 children with similar
ested in participating, were recruited and referred to the inclusion and exclusion criteria of those in the ESDM
treatment team. The children who received TAU were group as outlined above. In each group, one child with-
recruited through centers located in five Italian regions. drew from the study prior the beginning of the intervention
Each child and family received an assessment battery due to relocation. The two groups did not differ on base-
prior the beginning of the intervention, after 3 months, line characteristics (see Table 1) on any variables except
and at the end of the intervention after 6 months. The chronological age, for which the TAU group (mean
ESDM was administered 1:1 for 6 h per week in the age = 35.2 months; standard deviation (SD) = 7.6) was
center, the Neuropsychiatric Department of the University significantly older than the ESDM group (mean
of Messina, by professionals trained to implement the age = 31.1 months; SD = 8.0; t = −2.15; p = 0.03).
intervention with fidelity. For the ESDM group, thera-
pists’ fidelity to treatment procedures was monitored
Measures
throughout the intervention. The TAU group received
intervention for an average of 5.2 h per week. Intervention In both groups, senior psychologists experienced in clini-
in the TAU group was administered by professionals with cal evaluations administered the assessment measures.
background similar to those in the experimental group, There were two assessors at each site. The assessors were
except for training in ESDM. The two groups did not dif- not blind to the intervention. However, they were blind to
fer statistically for number of hours received per week the fact that the two groups would be compared. The asses-
(t = 1.30; p = 0.19). Children in both groups did not sors were not part of the intervention team. We did not
receive any services prior the start of the study. have the resources to conduct regular reliability and

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

4 Autism

Table 1.  Baseline measures for control and ESDM groups.

Control group ESDM group F p

  Mean SD N Mean SD N
GMDS
  Total Scale (GQ) 62.1 13.3 70 66.0 10.3 22 1.52 0.26
  Locomotor Development 81.7 16.4 70 80.7 18.9 22 0.37 0.81
  Personal Social Development 55.1 18.3 70 60.5 20.7 22 0.10 0.23
  Hearing and Speech 40.6 16.0 70 48.6 24.6 22 1.72 0.09
  Hand and Eye Coordination 63.4 16.6 70 66.0 19.9 22 1.56 0.55
  Performance Test 75.6 25.2 70 70.8 25.2 22 3.27 9.35
VABS
 Composite 67.4 11.8 70 69.9 10.4 22 0.14 0.38
 Communication 60.3 14.9 70 66.3 12.7 22 0.91 0.09
  Daily Living 71.8 13.5 70 73.3 11.4 22 0.61 0.64
 Socialization 65.7 12.7 70 64.1 11.4 22 0.47 0.61

ESDM: Early Start Denver Model; SD: standard deviation; GQ: General Quotient; GMDS: Griffiths Mental Development Scales; VABS: Vineland
Adaptive Behavior Scales.

fidelity monitor of the tests across sites throughout the every day settings. It provides an assessment of functional
study. However, all assessors had similar training in clini- skills in the domains of Communication, Daily living
cal assessment. Moreover, prior the beginning of the study, skills, and Socialization, as well as an overall composite
all assessors participated in a workshop conducted by the score, which served as a primary outcome variable. The
second author (A.N.) in which all tests and procedures VABS-II has excellent psychometric properties and, for
were reviewed. The following measures were adminis- young children, requires 30 min or less to administer or
tered to the participants at baseline, after 3 months and complete. The variables of interest are the standardized
after 6 months of intervention. score for each of these domains and an overall composite.

The Griffiths Mental Development Scales. The Griffiths Autism Diagnostic Observation Schedule-2 (ADOS-2). The
Mental Development Scales (Griffiths, 1984) is a stand- Autism Diagnostic Observation Schedule (ADOS) is a
ardized developmental test for children from birth to semi-structured interaction that measures symptoms of
96 months of age previously used in children with ASD autism through a standard set of probes. It provides an
(e.g. Chakrabarti and Fombonne, 2005). This test is trans- empirically derived algorithm that differentiates children
lated and standardized in Italian. In Italy, the Griffiths with ASDs from those with other delays or typical devel-
Scales are widely used for assessing children with devel- opment. The ADOS was used for eligibility.
opmental disorders including ASD. This test comprises The primary child’s outcome measures included the
scores in five subscales: Locomotor (assesses gross motor Griffiths Scales General Quotient (GQ) and the VABS
skills including the ability to balance, coordinate, and con- composite score. The secondary child’s outcome meas-
trol movements), Personal Social (measures the develop- ures included the Griffiths subscales’ and the VABS sub-
ing abilities that contribute to independence and social scales’ scores.
development), Hearing and Language (assesses receptive
and expressive languages), Eye and Hand Coordination
(focuses on fine motor skills, manual dexterity, and visual
Intervention
monitoring skills), and Performance (assesses the child’s ESDM. Each child participating in the ESDM group
visuospatial skills, speed of working, and precision). These received three 2-h sessions per week of ESDM treatment
subscales generate a raw score that is converted to Devel- delivered individually in a center-based context for
opmental Quotient score (developmental age divided by 6 months. The children did not receive additional services
chronological), a primary outcome variable. The mean during the study. The treatment team included fully cre-
score for each subscale is 100, and the SD is 15. dentialed professionals with a background in behavioral
principles, clinical, and developmental psychology. Prior
Vineland Adaptive Behavior Scales, Second Edition. Vineland to the beginning of the intervention, all professionals
Adaptive Behavior Scales, Second Edition (VABS-II; participated in the Introductory and Advanced ESDM
Sparrow et al., 2005) is a semi-standardized parent ques- Training modules on site. The first author (C.C.), a certi-
tionnaire designed to assess children’s behavior in real life, fied ESDM Trainer, provided the initial training and

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

Colombi et al. 5

supervised the team on a weekly basis either by video Results

review or by direct observation of treatment delivery. The
core staff members reached fidelity during the intervention Children’s outcomes
according to procedures described in Rogers and Dawson The distributions of the study’s variables did not violate
(2010). Treatment adherence was monitored using the normality. Therefore, we implemented parametric analy-
ESDM fidelity tool, a five-point Likert scale that evaluates ses to test the effect of ESDM on amount of child gain
13 therapist behaviors (Rogers and Dawson, 2010). After across the 6 months of intervention using a repeated-
reaching fidelity, core staff members demonstrated a score measures analysis of variance, with age in months
of at least 4 on the five-point ranking system at each included as a covariate in the model due to the age differ-
monthly fidelity check. ence between the groups. A priori contrasts compared
The ESDM curriculum assessment was administered baseline scores with 3- and 6-month outcome scores.
every 12 weeks by a trained therapist. Based on the results First, we ran the analyses for the primary outcomes meas-
of the curriculum assessment, individualized learning ures: the Griffiths GQ and the VABS composite. Because
objectives were developed for each child and monitored the control groups comprised five different regions, site
during each therapy session. The learning objectives main effects and interactions with time for the primary
focused on verbal and nonverbal communication, joint outcomes in the control group were analyzed and found
attention, social engagement, imitation, play, cognition, nonstatistically significant. For this reason, site was
motor development, and adaptive skills. omitted in the final model. Then, we analyzed the differ-
According to procedures described by Rogers and ences relative to the secondary measures, the Griffiths
Dawson (2010), the ESDM was delivered by the therapists subscales, and the VABS subscales.
within the context of play and daily routines in which
highly precise teaching is embedded. The therapists shared Primary analysis. Table 2 shows outcomes at 3 and
treatment objectives with the parents and discussed with 6 months for the ESDM and the TAU groups. At
them ESDM strategies at the end of each session. Parents 3 months, the groups differ significantly on the Griffiths
observed the therapy, but they were not formally trained in GQ. The ESDM group had an average increase of
delivering ESDM. 10.5 points, while the TAU group had an increase of
4.9 points (F = 2.0, p = 0.03). After 6 months of interven-
TAU.  The TAU consisted of community-based interven- tion, the ESDM showed an increase of 14.8 points, and
tions delivered by child neuropsychiatric services. The the TAU group showed an increase of 7.6 points (F = 2.3,
children did not receive additional services during the p = 0.02).
study. Children participated in TAU for a mean of 5.2 h Significant group differences at 3 months were also
per week. The TAU group comprises children who found in the VABS composite. The ESDM group showed
received intervention in five different Italian regions. an improvement of 7.9 points in comparison to baseline,
As explained above, all treatment teams were similar in while the TAU group showed an improvement of 3.5 points
terms of professional background. In none of the services (F = 2.1, p = 0.03). No significant differences were found at
had professionals received high-level training or certifi- 6 months in the VABS’ composite.
cation in a specific ASD-specialized intervention. For all
centers, ASD intervention training occurred through par- Secondary analyses. A significant group difference was
ticipation in introductory trainings and conferences and found on two Griffiths subscales: at 3 months (F = 2.0,
reading of articles and books. None of the professionals p = 0.04) and at 6 months (F = 4.0, p < 0.01) in the Per-
delivering intervention to the TAU children were certi- sonal Social subscale and at 3 months (F = 22.4, p = 0.01)
fied in a specific intervention and none were Board and at 6 months (F = 2.1, p = 0.03) in the Hearing and
Certified Behavioral Analysts. Two regions implemented Speech subscales. In addition, at 6-month group differ-
intervention inspired by Discrete Trial Teaching (Lovaas, ences approached significance on two additional sub-
1987; Sallows and Graupner, 2005). Two regions imple- scales, Locomotor and Hand and Eye Coordination. On
mented intervention inspired by Treatment and Educa- all these comparisons, the ESDM group made more gains
tion of Autistic and Related Communication Handicapped than the TAU group.
Children (TEACCH; Venter et al., 1992). One region On the VABS, significant group differences at 3 months
implemented intervention inspired by the Developmen- were found for the Communication subscale (F = 2.8,
tal Individual Difference, Relationship-Based (DIR) p < 0.01), with trends toward significance as well for
Model (Wieder and Greenspan, 2001). The TAU group is Daily Living Skills and Socialization subscales, all favor-
representative of the services usually delivered in Italy, ing the ESDM group. At 6 months, no significant group
where licensed professionals administer ASD interven- differences were found on any VABS subscales, though
tions without ASD intervention formal training or the ESDM group made more gains than the TAU group on
certification. all 10 scales.

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

6 Autism

Table 2.  Child outcome after 3 and 6 months of study participation.

Measure 3-month 6-month Group × time Group × Time

outcome outcome (baseline vs (baseline vs
3 months) 6 months)

  Mean Δ Mean Δ Mean Δ Mean Δ F p F p

ESDM (22) Control (70) ESDM (22) Control (70)
GMDS
  Total Scale (GQ) 10.5 4.9 14.8 7.6 2.0 0.03 2.3 0.02
  Locomotor Development 7.8 1.4 10.1 1.2 1.1 0.24 1.7 0.08
  Personal Social Development 14.7 4.4 24.5 9.0 2.0 0.04 4.0 0.00
  Hearing and Speech 12.0 5.8 17.6 11.4 2.4 0.01 2.1 0.03
  Hand and Eye Coordination 9.3 3.6 15.9 7.8 1.6 0.10 1.8 0.06
  Performance Test 12.3 5.7 15.3 6.3 2.9 0.76 1.7 0.49
VABS
 Composite 7.9 3.5 7.7 5.8 2.1 0.03 1.3 0.18
 Communication 10.6 6.5 5.6 8.6 2.8 0.00 0.6 0.49
  Daily Living 6.7 3.1 7.0 6.5 1.6 0.09 0.5 0.55
 Socialization 9.5 3.1 10.5 4.9 1.7 0.08 1.3 0.19

ESDM: Early Start Denver Model; GQ: General Quotient; GMDS: Griffiths Mental Development Scales; VABS: Vineland Adaptive Behavior Scales.

Feasibility treatment fidelity. However, after the acquisition of the

model, the staff appreciated the positive changes of the
Acceptability.  The retention rate of children enrolled in the children and recognized the importance of implementing
program was 95%. The retention rate among the staff was the intervention rigorously.
90%.

Demand.  Since the beginning of the ESDM implementa- Discussion

tion in 2011, more than 200 families have applied for the The increase in early identification of ASD in the world
ESDM program. Thus, demand exceeded the program brings the need for early interventions that are not only
capacity. Moreover, 10 professionals asked to volunteer efficacious in rigorous research settings but also effective
for the project so they could observe the ESDM therapy. and feasible in community settings that have far fewer
More than 50 professionals applied to work as ESDM resources than those available in university centers. This is
therapists. the first effort to demonstrate the applicability and the
effectiveness of the ESDM in an Italian community.
Implementation. All components of the program were Despite the fact that ESDM was administrated for only 6 h
implemented accurately and consistently. C.C., an ESDM per week, the ESDM group demonstrated significant gains
trainer, monitored implementation of components and at both 3 and 6 months of intervention in overall develop-
treatment fidelity throughout the program (see interven- mental quotient and language and at 3 months in adaptive
tion session). Additionally, based on the experience of the skills. Moreover, the ESDM group showed consistent
Messina center, five more Italian regions are planning to higher gains at both time points, in comparison to the TAU
implement the ESDM model within the public health sys- group, in overall cognitive skills and language. Moreover,
tem and/or the educational system. the ESDM showed higher gains at 3 months in adaptive
skills. This is the first study in the literature to report sig-
Practicality.  Audits from the Sicilian region determined that nificant gains in favor of the ESDM group detectable
the program exceeded the regional quality standard. As a within 12 weeks of beginning treatment.
consequence, at the end of the project, the Sicilian region Our results are in line with previous efficacy studies
decided to provide funding for three additional years. demonstrating changes in overall developmental quotient
and adaptive skills in children receiving the ESDM
Adaptation and integration. The staff were not used to (Dawson et al., 2010; Vivanti et al., 2014). The amount of
implementing a specific model with manualized strategies gain shown by the ESDM group in this study compares
and procedures. Thus, the main obstacles involved the favorably with that reported by Dawson et al. (2010), even
implementation of all components of the intervention though those children received many more hours of treat-
including writing objectives every 12 weeks and imple- ment than did those children in this study. Interestingly, the
menting specific intervention strategies according to TAU group in this study also made similar gains to those

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

Colombi et al. 7

reported in the Dawson et al. (2010) study for the children difference in age. The TAU group was slightly older in
in the control group. Similarly to Vivanti et al. (2014) comparison to the ESDM group. Despite the fact that age
effectiveness studies of ESDM, children in the ESDM was controlled in the statistical analysis, a randomized
made significant larger gains in developmental quotient in design with high control of all baseline variables would
comparison to the control groups. Our results, in line with lead to more solid results. However, our limited resources
those described in the Vivanti et al. (2014) study, suggest did not allow for such rigor.
that ESDM may have positive impact on young children Given these initial findings, replication with a more rig-
with ASD even when implemented in communities, with- orous design, including RCTs, will be important in deter-
out the resources and the rigor of universities-based RCTs. mining the effectiveness of ESDM in the Italian culture.
Our study suggests that the ESDM even delivered While efficacy studies at university-based institutions rep-
at low intensity may promote better outcomes in com- resent a first step in developing and testing intervention,
parison to TAU implemented in Italy. The change in effectiveness is established when interventions can be
language and developmental rates are quite important in implemented in the community and can demonstrate
that studies consistently report that IQ and language improved outcomes of young children with ASD. Our
ability at the age of 5 years, rather than autism severity, study represents an important step in testing effectiveness
are the best predictors of outcomes in adulthood (Lord of early intervention for ASD for the Italian community.
et al., 2012; Luyster et al., 2007). Thus, developmental
rates and language may well be the most important tar- Acknowledgement
gets to focus on. We wish to acknowledge the work of the clinicians involved in
The results of our study should not be used to support this work. We would like to thank all children and parents who
low-intensity intervention as gold standard in clinical work participated in this study.
for children with ASD. However, in many communities
Funding
with limited resources, only low-intensity interventions
are available. Our results support the implementation of The author(s) disclosed receipt of the following financial support
high-quality, evidence-based treatment, even at low inten- for the research, authorship, and/or publication of this article:
This work was supported by Assessorato Regionale alla Salute
sity, to promote outcome changes in young children with
della Regione Sicilia (Health Administration of Sicily), Azienda
ASD. Moreover, they suggest that professionals in another
Ospedaliera Universitaria – Policlinico G. Martino – Messina,
culture can master the American-based ESDM and deliver and National Research Council of Italy.
it with acceptable fidelity of implementation, and that chil-
dren in another culture can make gains similar to American References
children receiving this intervention. Aucouturier B (2005) La method Aucouturier. Fantasmes
The feasibility of the ESDM in the Italian community d’action e pratique psychomotrice. De Boeck Supérieur.
was supported by several indicators including implemen- Berruezzo P (2000) El contenido de la psicomotricidad. In:
tation of the treatment at fidelity, acceptability, demand, Bottini P (ed.) Psicomotricidad: prácticas y conceptos.
practicality, adaptation, and integration. The feasibility Madrid: Miño y Dávila, pp.43–99.
analysis suggests that the ESDM is an intervention that can Bowen DJ, Kreuter M, Spring B, et al. (2009) How we design
be implemented successfully in the Italian community. feasibility studies. American Journal of Preventative
Given our encouraging results, it seems appropriate to rec- Medecine 36(5): 452–457.
ommend that the Italian Public Health System allocates Chakrabarti S and Fombonne E (2005) Pervasive developmental
disorders in preschool children: confirmation of high preva-
resources to train health professionals in evidence-based
lence. American Journal of Psychiatry 162(6): 1133–1134.
interventions for young children with ASD. Despite the Dawson G, Jones EJ, Merkle K, et al. (2012) Early behavioral
initial difficulties in learning a manualized intervention, intervention is associated with normalized brain activity
many Italian professionals implemented the intervention in young children with autism. Journal of the American
with fidelity. Academy of Child and Adolescent Psychiatry 51: 1150–1159.
Our results should be interpreted with caution for sev- Dawson G, Rogers S, Munson J, et al. (2010) Randomized, con-
eral reasons. Weaknesses of our study included lack of a trolled trial of an intervention for toddlers with autism: the
randomized control group, lack of fidelity monitoring of Early Start Denver Model. Pediatrics 125(1): e17–e23.
the TAU intervention, lack of blind assessors, and a Divan G, Hamdani SU, Vajartkar V, et al. (2015) Adapting an
slightly older control group in comparison to the ESDM evidence-based intervention for autism spectrum disorder
for scaling up in resource-constrained settings: the develop-
group. Precise descriptions of TAU across the various
ment of the PASS intervention in South Asia. Global Health
regions were not available. Therefore, similarities and dif- Action 8: 10.3402/gha.v8.27278.
ferences between TAU and ESDM cannot be analyzed. Griffiths R (1984) The Abilities of Young Children: A
However, TAU was representative of the services usually Comprehensive System of Mental Measurement for the First
provided in Italy. Moreover, assessors were not blind to Eight Years of Life (revised edition). Bucks: A.R.I.C.D. Test
the intervention. Another limitation to consider is the Agency Limited.

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016

8 Autism

Koegel LK, Koegel RL, Ashbaugh K, et al. (2014) The impor- Smith T, Scahill L, Dawson G, et al. (2007) Designing research
tance of early identification and intervention for children studies on psychosocial interventions in autism. Journal of
with or at risk for autism. International Journal of Speech- Autism Development Disorder 37: 354.
Language Pathology 16(1): 50–56. Venter A, Lord C and Schopler E (1992) A follow-up study
Lord C, Rutter M, DiLavore PC, et al. (2012) Autism Diagnostic of high-functioning autistic children. Journal of Child
Observation Schedule, Second Edition (ADOS–2) (Modules Psychology and Psychiatry 33(3): 489–507.
1–4). Los Angeles, CA: Western Psychological Services. Vismara LA, Colombi C and Rogers SJ (2009) Can one hour per
Lovaas OO (1987) Behavioral treatment and normal educational week of therapy lead to lasting changes in young children
and intellectual functioning in young autistic children. with autism? Autism 13(1): 93–115.
Journal of Consulting and Clinical Psychology 55(1): 3–9. Vismara LA, McCormick CE, Wagner AL, et al. (2016)
Luyster R, Qiu S, Lopez K, et al. (2007) Predicting outcomes Telehealth parent training in the Early Start Denver Model
of children referred for autism using the MacArthur-Bates results from a randomized controlled study. Focus on
Communicative Development Inventory. Journal of Speech, Autism and Other Developmental Disabilities. Epub ahead
Language, and Hearing Research 50(3): 667–681. of print 26 May. DOI: 1088357616651064.
Reichow B, Volkmar FR and Cicchetti DV (2008) Development Vismara LA, Rogers SL, Wagner AL, et al. (2013) Autism treat-
of the evaluative method for evaluating and determining ment in the first year of life: a pilot study of infant start, a
evidence-based practices in autism. Journal of Autism and parent-implemented intervention for symptomatic infants.
Developmental Disorders 38(7): 1311–1319. Journal of Autism and Developmental Disorders 44(12):
Rogers SJ and Dawson G (2010) The Early Start Denver Model 2981–2995.
for Young Children with Autism: Promoting Language, Vivanti G, Dissanayake C and Victorian ASELCC Team (2016)
Learning, and Engagement. New York: Guilford Press. Outcome for children receiving the Early Start Denver
Rogers SJ, Estes A, Lord C, et al. (2012) Effects of a brief Early Model before and after 48 months. Journal of Autism and
Start Denver Model (ESDM) based parent intervention on Developmental Disorders 46: 2441–2449.
toddlers at risk for autism spectrum disorders: a randomized Vivanti G, Dissanayake C, Zierhut C, et al. (2013) Brief
controlled trial. Journal of the American Academy of Child report: predictors of outcomes in the Early Start Denver
and Adolescent Psychiatry 51(10): 1052–1065. Model delivered in a group setting. Journal of Autism and
Sallows G and Graupner T (2005) Intensive behavioral treatment Developmental Disorders 43(7): 1717–1724.
for children with autism: four-year outcome and predictors. Vivanti G, Paynter J, Duncan E, et al. (2014) Effectiveness
American Journal on Mental Retardation 110(6): 417–438. and feasibility of the Early Start Denver Model imple-
Schreibman L, Dawson G, Stahmer AC, et al. (2015) Naturalistic mented in a group-based community childcare setting.
developmental behavioral interventions: empirically vali- Journal of Autism and Developmental Disorders 44(12):
dated treatments for autism spectrum disorder. Journal of 3140–3153.
Autism and Developmental Disorders 45(8): 2411–2428. Wieder S and Greenspan (2001) Climbing the symbolic ladder in
Sparrow SS, Cicchetti DV and Balla DA (2005) Vineland the DIR model through floor time/interactive play. Autism:
Adaptive Behavior Scales, (Vineland-II). Circle Pines, MN: The International Journal of Research and Practice 7(4):
American Guidance Services. 425–435.

Downloaded from aut.sagepub.com at La Trobe University on October 27, 2016