539835

research-article2014

FOAXXX10.1177/1088357614539835Focus on Autism and Other Developmental DisabilitiesMorgan et al.

Article

Impact of Social Communication

Interventions on Infants and
Toddlers With or At-Risk for
Autism: A Systematic Review

Focus on Autism and Other

Developmental Disabilities
2014, Vol. 29(4) 246256
Hammill Institute on Disabilities 2014
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1088357614539835
focus.sagepub.com

Lindee J. Morgan, PhD1, Emily Rubin, MS2, Jaumeiko J. Coleman, PhD3,

Tobi Frymark, MA3, Beverly P. Wang, MLIS3, and Laura J. Cannon, BA4

Abstract
This is a systematic review of the impact of communication interventions on the social communication skills of infants and
toddlers with or at-risk for autism spectrum disorder (ASD). A priori clinical questions accompanied by specific inclusion
and exclusion criteria informed the extensive literature search that was conducted in multiple databases (e.g., PubMed).
Twenty-six studies were accepted for this review. Outcomes were reported by social communication category (i.e.,
joint attention, social reciprocity, and language and related cognitive skills) and communication developmental stage (i.e.,
prelinguistic, emerging language). Primarily positive treatment effects were revealed in the majority of outcome categories
for which social communication data were available. However, the presence of intervention and outcome measure
heterogeneity precluded a clear determination of intervention effects. Future research should address these issues while
also evaluating multiple outcomes and adding a strong family component designed to enhance child active engagement.
Keywords
autism, autism spectrum disorder, social communication, pervasive developmental disorder, speech-language pathology,
intervention
In 2012, the American Speech-Language-Hearing Associ
ations (ASHAs) National Center for Evidence-based
Practice was charged with developing an evidence-based
systematic review (EBSR) on the impact of communication
interventions for infants and toddlers with or at-risk for
ASD in collaboration with experts in the field. An EBSR
addresses unambiguous and specific questions on a particular topic; clearly explains the methods and criteria used to
locate and select studies for inclusion; and entails reviewing, critiquing, and integrating pertinent information from
the selected studies in an effort to provide a synthesis of the
current best evidence (Dollaghan, 2007).
Previous systematic reviews have examined the impact
of communication treatments on various skill areas in wide
age groups of children with ASD (e.g., National Research
Council [NRC], 2001; Schertz, Reichow, Tan, Vaiouli, &
Yildirim, 2012; Wallace & Rogers, 2010). In reviews that
addressed infants and toddlers, lack of empirically validated
treatments for infants and toddlers with ASD (Wallace &
Rogers, 2010) and great heterogeneity in findings (e.g.,
Schertz et al., 2012) precluded ascertaining generalized
treatment effects. Clearly, a closer examination of the evidence pertaining to social communication interventions
used with young children is warranted.

To better elucidate the treatment effects of social communication interventions, a framework devised by ASHA
was adopted (ASHA, 2006). The framework groups intervention goals by social communication outcome categories
(i.e., joint attention, social reciprocity, language and related
cognitive skills, behavior and emotional regulation) across
communication developmental stages (i.e., prelinguistic,
emerging language). The social communication outcome
categories represent core areas of difficulty for individuals
with ASD, whereas the selected developmental stages
reflect the age of the population discussed in this EBSR.
Joint attention is establishing shared attention, social reciprocity entails maintaining interactions by taking turns, language and related cognitive skills applies to the use and
1

Florida State University Autism Institute, Tallahassee, USA

Marcus Autism Center, Atlanta, GA, USA
3
American Speech-Language-Hearing Association, Rockville, MD, USA
4
University of Maryland, College Park, USA
2

Corresponding Author:
Jaumeiko J. Coleman, American Speech-Language-Hearing Association,
National Center for Evidence-Based Practice, 2200 Research Blvd.,
#245, Rockville, MD 20850, USA.
Email: jcoleman@asha.org

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

247

Morgan et al.
understanding of nonverbal and verbal communication, and
behavioral and emotional regulation is the successful regulation of ones emotions and behaviors (ASHA, 2006). In
regard to the communication developmental stages applicable to infants and toddlers, prelinguistic pertains to the
use of nonverbal communicative strategies, such as gestures; and emerging language is the burgeoning use of verbal language (ASHA, 2006). The aim of this EBSR was to
further evaluate the impact of communication interventions
on social communication skills of infants and toddlers with
ASD aged 36 months or less. The clinical questions for this
EBSR follow:
Clinical Question 1: What are the effects of communication interventions on joint attention outcomes for
children 36 months old or less at-risk for or diagnosed
with ASD?
Clinical Question 2: What are the effects of communication interventions on social reciprocity outcomes for
children 36 months old or less at-risk for or diagnosed
with ASD?
Clinical Question 3: What are the effects of communication interventions on language and related cognitive
skill outcomes for children 36 months old or less at-risk
for or diagnosed with ASD?
Clinical Question 4: What are the effects of communication interventions on behavioral and emotional regulation outcomes for children 36 months old or less at-risk
for or diagnosed with ASD?

Method
To complete this EBSR, a multi-step approach was taken
including (a) identification of peer-reviewed articles that
address the population of interest and clinical questions;
(b) evaluation of the methodological rigor of accepted studies; (c) grouping outcomes as prelinguistic or emerging language within one of the following areas: joint attention,
social reciprocity, language and related cognitive skills, or
behavioral and emotional regulation; (d) computing effect
sizes and assigning associated magnitude of effect descriptors; and (e) assessing the findings in relation to the clinical
questions. One author conducted a literature search in 25
electronic databases (e.g., PubMed, ERIC, Research
Autism) using key words related to autism, autism spectrum
disorder (ASD), pervasive developmental disorder (PDD),
speech-language pathology, and treatment (the complete
list of databases, key words, and search strategy is available
on request). Two authors independently assessed the titles
and abstracts of all articles. The references of all full-text
articles were scanned to identify additional relevant studies
and a search for prolific authors was also completed. Two
authors also independently assessed each included study for
methodological rigor; any disagreements about critical

appraisal ratings were resolved via consensus. Singlesubject design studies were assessed using an adapted version of the Single Case Experimental Design (SCED) scale
(Tate et al., 2008) and group studies were assessed using
ASHAs critical appraisal scheme (Cherney, Patterson,
Raymer, Frymark, & Schooling, 2008; Mullen, 2007). See
Supplemental Materials Tables 1 and 2 for further information regarding the critical appraisal processes.
Interrater reliability associated with the sifting of titles
and abstracts as well as the critical appraisal process were
determined using Cohens kappa (; Cohen, 1988) and
weighted . Cohens was used in instances in which only
two rating options equal in weight were available for selection. Weighted was applied to critical appraisal items that
had hierarchical rating options (i.e., sampling process, random allocation, controlling for order effects, precision).
The following is Landis and Kochs (1977) scale for interpreting which was used to categorize the strength of the
agreement: poor agreement (<0.00), slight agreement
(0.000.20), fair agreement (0.210.40), moderate agreement (0.410.60), substantial agreement (0.610.80), and
almost perfect agreement (0.811.00). Percent agreement
was reported when could not be computed or when the
kappa value was zero.
For inclusion in this EBSR, studies had to be peerreviewed and experimental or quasi-experimental.
Furthermore, studies had to examine the impact of communication interventions on social communication skills of
children 36 months old or younger at-risk for or diagnosed
with ASD. For the purposes of this review, the ASD category
included the following diagnostic labels: Asperger syndrome, autism, autistic disorder, PDD, and pervasive development disordernot otherwise specified (PDD-NOS).
Accepted studies were written in English and published after
1970. Studies with mixed populations or mixed ages were
excluded unless data could be separated for analyses. In
addition, only participant data from single-subject design
studies that integrated a control mechanism (i.e., studies
with a withdrawal or reversal phase and/or multiple baseline
design studies) were included; consequently, multiple baseline design studies across participants without a withdrawal
or reversal phase which included only one participant who
met our age criterion were not accepted as they became an
AB design. Both single-subject and group design studies
were included as the former provides information about the
impact of an intervention in consideration of an individuals
unique abilities, whereas the latter are used to evaluate the
generalizability (i.e., external validity) of a treatments
effects.
Interventions in included studies were required to contain at a minimum a description of the training method(s) or
techniques from which they were comprised. Study findings were classified by social communication outcome categories (i.e., joint attention, social reciprocity, language and

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

248

Focus on Autism and Other Developmental Disabilities 29(4)

Table 1. Clinical Question 1: Prelinguistic-Joint Attention Findings.

Single-subject study outcomes
Single-subject studies

Group study outcomes

ES (NAP)

ES magnitude

p value

Group studies

Joint attention intervention

Jones, Carr, and Feeley (2006)
Krstovska-Guerrero and Jones
(2013)
Rocha, Schreibman, and
Stahmer (2007)
Joint attention-mediated learning
Schertz and Odom (2007)
Reciprocal imitation training

0.391.00
1.00

Mediumlarge
Large

NR
NR

Autism 1-2-3 project

Wong and Kwan (2010)
Brief Early Start Denver Model

0.720.97

Mediumlarge

NR

Rogers et al. (2012)

0.610.76

Ingersoll and Schreibman

(2006)
Simultaneous communication
Kouri (1988)
Social engagement intervention
Vernon, Koegel, Dauterman,
and Stolen (2012)

0.000.98

1.00

0.810.99

ES (d)

ES magnitude

p value

3.13

Large

.001

0.000.13 (2/3
CI:NS)

Small

NR

Hanens More than Wordsa

Carter et al. (2011)

0.80.24 (CI:NS)
Small
NR
Interpersonal synchrony vs. non
interpersonal synchrony
Medium
NR Landa, Holman, ONeill, and
0.420.89
Mediumlarge .08.41
Stuart (2011)
Joint attention-mediated learning

Large
p < .05 Schertz, Odom, Baggett, and
0.701.39
Large
1/3 NS
Sideris (2013)

Mediumlarge
NR

Smalllarge

NR

Note. ES = effect size; NAP = non-overlap of all pairs; CI = confidence interval; NR = not reported; NS = not significant.
a
Only treatment follow-up data (i.e., maintenance data) were reported in this study.

Table 2. Clinical Question 2: Prelinguistic-Social Reciprocity Findings.

Single-subject study outcomes
Single-subject studies
Joint attention-mediated learning
Schertz and Odom (2007)
Peer-mediated intervention
Goldstein, Kaczmarek, Pennington,
and Shafer (1992)
Simultaneous communication
Kouri (1988)
Social engagement intervention
Vernon, Koegel, Dauterman, and
Stolen (2012)

ES (NAP)

ES magnitude

p value

0.350.74

Medium

NR

0.220.57 Smallmedium

0.921.00

Large

0.940.98

Large

NR

Group study outcomes

Group studies
Autism 1-2-3 project
Wong and Kwan (2010)
Brief Early Start Denver Model
Rogers et al. (2012)

Joint attention-mediated learning

p < .05 Schertz, Odom, Baggett, and
Sideris (2013)

ES (d)

ES magnitude p value

NR

NR

0.07 (CI:NS)

Small

.008

NR

0.55

Medium

>.05

NR

Note. ES = effect size; NAP = non-overlap of all pairs; CI = confidence interval; NR = not reported; NS = not significant.

related cognitive skills, and behavioral and emotional regulation), and then further categorized into communication
developmental stages (i.e., prelinguistic, emerging language) using the framework devised by ASHA (2006). So,
examples of outcomes classification labels include prelinguistic-joint attention, prelinguistic-social reciprocity, and
emerging languagelanguage and related cognitive skills.
Key participant information (e.g., age), intervention variables (e.g., duration) and outcomes data (e.g., joint attention), including maintenance and generalization findings,
were extracted from each study. Given the importance of
the ecological validity of findings, qualitative data gathered
from surveys and observations completed by caregivers and

non-participating clinical professionals was extracted, if

reported, to evaluate social validity.
Statistical significance and effect size were reported if
available in the study or calculated if applicable raw data
were provided. For single-subject design studies, linear
graphs were visually analyzed to compute the non-overlap
of all pairs (NAP; Parker & Vannest, 2009) effect size for
intervention outcomes. Conventions for applying NAP,
including classification of the magnitude of the effect size
(i.e., small = 00.31, medium = 0.320.84, and large =
0.851.00), were adapted from Parker and Vannest (2009).
For group studies, the calculated effect size metric was
Cohens d (Cohen, 1988). For the purpose of assigning

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

249

Morgan et al.
Table 3. Clinical Question 3: Prelinguistic-Language and Related Cognitive Skills Findings.
Single-subject study outcomes
Single-subject studies

ES (NAP)

ES magnitude

Group study outcomes

p value

Group studies

NR

Brief Early Start

Denver Model
Rogers et al. (2012)

Hanens More than

Words a
Carter et al. (2011) 0.15 to 0 (CI:NS)

Early Start Denver Model

Vismara, Colombi, and Rogers
(2009)

0.201.00

Smalllarge

Pivotal response training

Steiner, Gengoux, Klin, and
Chawarska (2013)
Reciprocal imitation training
Ingersoll and Schreibman (2006)
Reciprocal imitation training vs. video
modeling
Cardon and Wilcox (2011)

Simultaneous communication
Kouri (1988)
UCLA treatment model
Smith, Buch, and Gamby (2000)
Video modeling imitation training
Cardon (2012)

1.00

Large

NR

0.001.00

Smalllarge

NR

0.710.85 (RIT)
0.730.98 (VM)

Medium
Mediumlarge

NR
NR

0.831.00

Large

0.711.00

Mediumlarge

NR

0.931.00

Large

NR

ES (d)

ES magnitude p value

0.14 (CI:NS)

Small

NR

Small

NR

Note. ES = effect size; NAP = non-overlap of all pairs; RIT = reciprocal imitation training; VM = video modeling; M = mixture of statistically significant and non-statistically
significant findings; CI = confidence interval; NR = not reported; NS = not significant.
a
Only treatment follow-up data (i.e., maintenance data) were reported in this study.

descriptive labels to effect sizes reported in group studies,

the following modified version of Cohens classification of
effect size magnitude was used: small = 0 to 0.34; medium
= 0.35 to 0.64; and large = 0.65 or greater with positive
effect sizes favoring the intervention. When provided or
calculable, confidence intervals were also reported for
group study effect sizes. For both single-subject and group
studies, results were considered statistically significant if
the p value was less than .05. For group studies, effect size
confidence intervals that contained the null effect (i.e., d =
0) were not considered to be statistically significant. Both
effect sizes (and their confidence intervals when calculable)
and p values were reported because sample size has a
greater impact on p-value calculation than effect size computation (Borenstein, Hedges, Higgins, & Rothstein, 2009).
Therefore, a treatment effect may be present, but not appear
statistically significant based on the p value due to a sample
size that is too small to detect a treatment effect (i.e., the
study does not have enough power).

Results
Twenty-six studies (n = 19 single-subject and 7 group studies) from 1,379 identified citations were accepted for this
EBSR. The full list of excluded studies and reason(s) for
ineligibility (e.g., not population or age of interest) is available on request. Substantial interrater reliability ( = 0.69;

Cohen, 1988) was noted between the authors who sifted

abstracts and full-text articles. Kappa ratings for critical
appraisal items from both study designs ranged from slight
agreement ( = 0.13) to substantial agreement ( = 0.77),
whereas percent agreement ranged from 84% to 100%.
Most critical appraisal items were adequately addressed in
accepted single-subject design and group studies. See
Supplemental Materials Tables 1 and 2 for more information about study quality.
A total of 427 participants, who were aged 10 to 36 months
and diagnosed with autism, ASD, PDD-NOS, or Asperger
syndrome, were included in this EBSR. Participant race/ethnicity, which was reported in a few studies (Carter et al.,
2011; Dawson et al., 2010; Landa, Holman, ONeill, &
Stuart, 2011; Rogers et al., 2012; Vernon, Koegel, Dauterman,
& Stolen, 2012; Vismara, Colombi, & Rogers, 2009),
included American Indian, Alaskan Native, Asian, Black,
Hispanic or Latino, Multiracial, and White. See Supplemental
Materials Table 3 for additional participant and service delivery data as well as descriptions of interventions.

Study Outcomes
Included studies examined various social communication
outcomes across prelinguistic and emerging language developmental stages to address three of four a priori clinical questions. Study outcomes were classified as prelinguistic-joint

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

250

Focus on Autism and Other Developmental Disabilities 29(4)

attention (Clinical Question 1; n = 11 studies), prelinguisticsocial reciprocity (Clinical Question 2; n = 7 studies), prelinguistic-language and related cognitive skills (Clinical
Question 3; n = 9 studies), and emerging languagelanguage
and related cognitive outcomes (Clinical Question 3; n = 18
studies). No outcomes in the accepted studies fell into the
behavior and emotional regulation category (Clinical
Question 4). For most clinical questions, conclusions about
the relationship between study quality and study outcomes
could not be drawn because study quality was similar across
studies. For Clinical Question 2, however, one of the singlesubject studies (Kouri, 1988) had larger treatment effects but
lower study quality than the other two single-subject studies
in this category (Goldstein, Kaczmarek, Pennington, &
Shafer, 1992; Schertz & Odom, 2007). As indicated in at least
one study of each of the four aforementioned outcome categories, caregivers overwhelmingly were satisfied with the
interventions and associated outcomes. Supplemental
Materials Table 4 provides a summary of findings across
social communication outcome goals and communication
developmental stages. Additional information about study
findings by communication developmental stage and social
communication categories is elucidated below.
Prelinguistic-joint attention findings.Table 1 provides a
detailed list of interventions and associated outcomes for
the 11 studies of prelinguistic-joint attention (Clinical Question 1). Findings from single-subject studies overwhelmingly indicated improvement in prelinguistic-joint attention
skills across the variety of treatment categories. The associated effect sizes ranged from small to large in magnitude
(NAP = 01.00), with the bulk being in the medium to large
range (NAP = 0.391.00). The greatest variability in treatment effect (NAP = 01.00) was noted across participants
who received various joint attention interventions (Jones,
Carr, & Feeley, 2006; Krstovska-Guerrero & Jones, 2013;
Rocha, Schreibman, & Stahmer, 2007; Schertz & Odom,
2007). No effect size confidence intervals were provided or
calculable and no p values were reported. All studies except
Kouri (1988) reported maintenance and generalization findings; overall, target behaviors were maintained following
treatment and skills were demonstrated across a variety of
people and settings.
Group study effect sizes were mainly medium to large in
magnitude (d = 3.13 to 1.39), with the exception of the findings associated with the Brief Early Start Denver Model (d =
00.13: Rogers et al., 2012). In most instances, the findings
were not statistically significant. However, in the case of
Wong and Kwan (2010), the treatment effect associated with
the Autism 1-2-3 Project intervention was in favor of the control group. Growth rate difference effect sizes were medium
to large for a comparison of interpersonal synchrony and
non-interpersonal synchrony interventions (Landa et al.,
2011); yet, they were accompanied by non-statistically

significant p values. In addition, within-group findings

revealed gains from pre- to post-test only for the intervention
group (Landa et al., 2011). The majority of maintenance findings as reported in studies of Hanens More than Words
(Carter et al., 2011) and a joint attention-mediated learning
(Schertz, Odom, Baggett, & Sideris, 2013) intervention were
small to large in magnitude (d = 0.081.18) and accompanied
by primarily non-statistically significant p values; the largest
effect sizes were associated with the joint attention intervention. Maintenance findings from the Landa et al. (2011) comparative study revealed large effect sizes (d = 0.811.56) at
post-treatment and medium to large effect sizes (d = 0.41
0.68) for growth rate differences between the groups; p values were not statistically different for the post-treatment or
growth rate difference data. No generalization findings were
reported.
Prelinguistic-social reciprocity findings.Findings from singlesubject studies that addressed prelinguistic-social reciprocity
outcomes (Clinical Question 2; see Table 2) were small to
large in effect size magnitude (NAP = 0.221.00), with findings from the simultaneous communication (Kouri, 1988)
and social engagement (Vernon et al., 2012) interventions
falling solely in the large range (i.e., NAP = 0.921.00). No p
value or effect size confidence interval information was
reported. Maintenance and generalization of prelinguisticsocial reciprocity findings were limited to a study of a joint
attention-mediated learning (Schertz & Odom, 2007) intervention. Findings revealed that the skills were maintained at
levels higher than what was seen during baseline and that
generalization occurred across a variety of settings.
Much variability existed in group study findings (see
Table 2), with a significant p value (p = .008) associated with
the Autism 1-2-3 Project intervention (Wong & Kwan, 2010),
a negligible effect size (d = 0.07) accompanied by a confidence interval containing the null effect reported in the study
of Brief Early Start Denver Model (Rogers et al., 2012), and
a medium (d = 0.55), but not statistically significant finding
from a joint attention-mediated learning intervention (Schertz
et al., 2013). A small (d = 0.10) and non-statistically significant maintenance finding was reported in Schertz et al.
(2013). No generalization findings were reported.
Prelinguistic-language and related cognitive skills findings.Table
3 provides a detailed list of interventions and associated outcomes in the prelinguistic-language and related cognitive
skills category (Clinical Question 3). The effect sizes from
single-subject studies were in the small to large range (NAP
= 01.00), with most falling in the medium to large range for
the following interventions: pivotal response training
(Steiner, Gengoux, Klin, & Chawarska, 2013), video modeling imitation training (Cardon, 2012), University of California, Los Angeles (UCLA) treatment model (Smith, Buch, &
Gamby, 2000), and simultaneous communication (Kouri,

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

251

Morgan et al.
Table 4. Clinical Question 3: Emerging LanguageLanguage and Related Cognitive Skills Findings.
Single-subject study outcomes
Single-subject studies
Behavioral intervention
Williams, PrezGonzlez, and Vogt
(2003)
Developmental, social
pragmatic intervention
Ingersoll, Dvortcsak,
Whalen, and Sikora
(2005)
Discrete trial training vs.
mand training
Jennett, Harris, and
Delmolino (2008)

Group study outcomes

ES (NAP)

ES magnitude

p value

1.00

Large

NR

0.710.86

Mediumlarge

NR

Early Start Denver

Model
Dawson et al. (2010)

0.941.00 (DTI)

Large

NR

1.00 (MT)

Large

NR

0.201.00

Smalllarge

NR

0.760.87

Mediumlarge

NR

Peer-mediated intervention

Goldstein, Kaczmarek,
Pennington, and Shafer
(1992)
Pivotal response training vs.
discrete trial training
Schreibman, Stahmer,
Cestone Barlett, and
Dufek (2009)
Reciprocal imitation
training
Ingersoll and Schreibman
(2006)
Simultaneous
communication
Kouri (1988)
Social engagement
intervention
Vernon, Koegel,
Dauterman, and Stolen
(2012)
Teaching strategy
intervention
Kashinath, Woods, and
Goldstein (2006)
UCLA treatment model
Smith, Buch, and Gamby
(2000)

0.240.61

Smallmedium

Autism 1-2-3 Project

Wong and Kwan
(2010)
Brief Early Start Denver
Model
Rogers et al. (2012)

Early Start Denver Model

Vismara, Colombi, and
Rogers (2009)
Enhanced milieu teaching
Kaiser, Hancock, and
Nietfeld (2000)

Group studies

Eclectic intervention
vs. applied behavioral
analysis
Zachor and Itzchak
(2010)
Hanens More than
Wordsa
Carter et al. (2011)
Interpersonal vs.
non-interpersonal
synchrony
Landa, Holman,
ONeill, and Stuart
(2011)

ES (d)

ES magnitude

p value

NR

NR

.01.04

0.24 to 0.24 (CI:NS)

Small

NR

0.580.66 (CI: 2/4 NS)

Mediumlarge .03.06

0.18 to 0.59 (CI: 3/4 NS) Smallmedium

NR

0.26 to 0.42 (CI:NS)

Smallmedium

NR

0.49

Medium

.18

NR

0.470.97
0.28 to 0.59

Mediumlarge
Smallmedium

NR
NR

0.000.86

Smalllarge

NR

0.001.00

Smalllarge

0.810.94

Mediumlarge

NR

0.480.96

Mediumlarge

NR

0.001.00

Smalllarge

NR

Note. ES = effect size; NAP = non-overlap of all pairs; NR = not reported; DTI = discrete trial instruction; MT = mand training; M = mixture of
statistically significant and non-statistically significant findings; CI = confidence interval; NS = not significant.
a
Only treatment follow-up data (i.e., maintenance data) were reported in this study.

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

252

Focus on Autism and Other Developmental Disabilities 29(4)

1988). However, no effect size confidence intervals or p

value data were provided. Effect sizes from a comparative
study (Cardon & Wilcox, 2011) of reciprocal imitation and
video modeling were in the medium range (NAP = 0.71
0.85) with the exception of one finding for video modeling
which was large (NAP = 0.98); no p values or effect size
confidence intervals were reported. Maintenance and generalization findings were quite variable among the studies that
reported those data (Cardon, 2012; Cardon & Wilcox, 2011;
Ingersoll & Schreibman, 2006).
In regard to group studies, the Brief Early Start Denver
Model had a negligible post-treatment effect (d = 0.14)
that was associated with a confidence interval that contained the null effect (Rogers et al., 2012). Maintenance
findings, provided in a study of Hanens More than Words
(Carter et al., 2011), were negligible (i.e., d = 0) and accompanied by a confidence interval that contained the null
effect. No generalization findings were reported.
Emerging languagelanguage and related cognitive skills findings.
The majority of studies (n = 18) examined interventions that
addressed emerging languagelanguage and related cognitive skills (see Table 4 for a list of interventions and associated outcomes). Findings from single-subject studies were
in the small to large magnitude range (NAP = 01.00), with
effect sizes from the following studies falling in the medium
to large range: developmental, social pragmatic intervention (Ingersoll, Dvortcsak, Whalen, & Sikora, 2005); treatment strategy intervention (Kashinath, Woods, & Goldstein,
2006); behavioral intervention (Williams, Prez-Gonzlez,
& Vogt, 2003); and enhanced milieu teaching (Kaiser, Hancock, & Nietfeld, 2000). No p values or effect size confidence intervals were reported. Effect sizes from a
comparative study (Jennett, Harris, & Delmolino, 2008) of
discrete trial training and mand training were large (NAP =
0.941.00). In another comparative study (Schreibman,
Stahmer, Cestone Barlett, & Dufek, 2009), effect sizes for
the pivotal response training were medium to large (NAP =
0.470.97), whereas those for discrete trial training were
small to medium (NAP = 0.28 to 0.59). No p values or
effect size confidence intervals were reported in either comparative study. There was notable variability in maintenance
and generalization findings in the four studies (Ingersoll et
al., 2005; Ingersoll & Schreibman, 2006; Kaiser et al.,
2000; Vernon et al., 2012) that reported those data.
The effect sizes from group studies spanned from small
to large in magnitude (d = 0.24 to 0.66). Most findings
were not statistically significant with the exception of some
or all from the Autism 1-2-3 Project study (Wong & Kwan,
2010) and the Early Start Denver Model study (Dawson et
al., 2010). Effect sizes from comparative studies of an
eclectic intervention versus an applied behavioral analysis
intervention (Zachor & Itzchak, 2010) and an interpersonal
synchrony versus non-interpersonal synchrony intervention

(Landa et al., 2011) were small to medium in magnitude and

not statistically significant. All statistically significant
within-group pre- to post-test gains as well as gain scores
between groups were in favor of the intervention group
(Landa et al., 2011). Growth rate effect size was small (d =
0.09) and not statistically significant (p = .83; Landa et al.,
2011). Maintenance findings from a study of Hanens More
than Words (Carter et al., 2011) were small to medium in
magnitude (d = 0.16 to 0.42) and accompanied by confidence intervals that contained the null effect. In Landa and
colleagues (2011) comparative study, the maintenance
effect size was medium in magnitude (d = 0.57) and accompanied by a non-statistically significant p value (p = .24),
whereas the growth rate data collected during the maintenance period translated into a small effect size (d = 0.09)
and non-statistically significant p value (p = .83). No generalization findings were reported.

Discussion
This review of 26 intervention studies including 427 toddlers with ASD and spanning across a broad range of intervention categories indicated primarily positive treatment
effects on social communication skills in terms of both
growth rates and gain scores for all outcome categories for
which social communication data were available, with the
exception of emerging languagelanguage and related cognitive skills, which showed variable and mixed results.
Maintenance results were also variable across all outcome
categories and reporting of generalization results was limited. As a whole, caregivers were satisfied with the interventions and their associated outcomes.
The overall body of literature included in this review
was of appropriate scientific rigor with 24 of the 26 studies
sufficiently meeting the majority of the critical appraisal
points. Yet, patterns of weakness in research design were
noted for both single-subject design (e.g., assessors were
not blind to treatment) and group studies (e.g., use of convenience sampling). Despite these areas of weakness, no
distinct patterns were detected between study quality indicators and reporting of outcomes with the exception of the
Kouri (1988) study. Although the Kouri study reported
large effects for the three prelinguistic outcome areas, this
study met only 4 of the 12 quality indicator appraisal items.
Given potential weaknesses in the design of this particular
study, it is possible that effect sizes for this study could be
inflated and should be interpreted with caution. Following
is a discussion of treatment effects on social communication
within the context of study design.

Single-Subject Designs
Single-subject intervention studies targeting social communication outcomes for toddlers with ASD generally reported

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

253

Morgan et al.
improvements across outcome categories. Medium to large
effects were present for the bulk of studies reporting outcomes for the three prelinguistic outcome categories. For
the joint attention outcomes, the largest variability in effects
was reported for the interventions in the joint attention category. Variability in improvement was reported for the
emerging languagelanguage and related cognitive skills
category as evidenced by effects that ranged from small to
large. For the proportion of studies that reported maintenance and generalization findings, outcomes were highly
variable.

Group Designs
The seven group design studies included in this review
showed positive trends in growth with respect to social
communication outcomes and a range of effect sizes were
reported; however, very few results indicated statistical significance. In summarizing the social communication outcomes reported for group studies, the most promising
effects appear to be in favor of clinician-implemented interventions providing the greatest intensity (Dawson et al.,
2010; Landa et al., 2011). Schertz and colleagues (2013)
provided the single exception to this by reporting on a parent-implemented intervention of brief duration and intensity having an effect on joint attention as well as an effect on
emerging language and related cognitive skills in favor of
the treatment group. Because this particular study, however,
did not blind assessors to the treatment condition, the findings may be biased and must be viewed with caution.

Overarching Implications of Findings

Both focused interventions, which are directed toward
improving targeted symptoms or needs of the child with
ASD, and comprehensive interventions, which are developed to broadly reduce autism symptoms and improve
overall functioning, have been used to improve social communication functioning in children with ASD. Singlesubject design studies included in this review were typically
associated with positive changes in social communication
outcomes, whereas interventions assessed within-group
studies mainly resulted in mixed findings. However, a summary of findings relative to our research questions across
study designs indicates a preponderance of promising
effects for prelinguistic-joint attention and social reciprocity outcomes, with the bulk of interventions associated with
moderate to large treatment effects. With respect to language and related cognitive outcomes at both the prelinguistic and emerging language stages, an inconsistent
picture emerges given the broad range of treatment effects
reported. Finally, the studies reviewed provided no indication of effects on behavioral and emotional regulation outcomes. Although the positive results reported for social

communication outcomes for toddlers with ASD are

encouraging, the mixed nature of these results raises concerns with respect to the focus of treatment outcomes, treatment intensity, and agent of delivery (e.g., clinician versus
parent-mediated), as well as the types of measures utilized
to document intervention outcomes.

Focus of Treatment
The application of a developmental framework ensures that
prelinguistic-social communication skills are addressed
prior to symbolic language (ASHA, 2006; NRC, 2001;
Prizant, Wetherby, Rubin, & Laurent, 2003). In stark contrast to that developmental emphasis is the finding that the
majority of studies in this EBSR were focused on emerging
languagelanguage and related cognitive skills. With relatively few studies evaluating the effects of earlier-emerging,
foundational social communication skills, such as joint
attention and social reciprocity, we question whether normal developmental trajectories are being overlooked in the
bulk of toddler interventions for ASD. Because longitudinal
research has shown clearly the link between early social
communication skills such as joint attention and long-term
linguistic outcomes such as initiating bids and sharing emotions (Wetherby, Watt, Morgan, & Shumway, 2007) a stronger impetus for selecting outcome goals related to the core
challenges of ASD seems warranted. Of additional concern
is the fact that no studies reported behavior and emotional
regulation findings. It may be that these collective skills
have not been targeted due to challenges with measurement
or other practical reasons.

Intensity
Although several systematic review panels have recommended active engagement in intensive instruction for a
minimum of 5 hr per day (Maglione, Gans, Das, Timbie, &
Kasari, 2012; NRC, 2001), none of the included studies in
this review evaluated intensity of instruction to that extent.
One study, however, did report an average of 20 hr per week
(Dawson et al., 2010). Although the findings of this EBSR
suggest that fewer hours of treatment may be sufficient to
ameliorate social communication deficits and support
growth in social communication skills, it is unclear what the
critical mass is for maximizing long-term outcomes for
children with ASD.
That the most promising effects on social communication
outcomes appear to be in favor of clinician-implemented
interventions providing the greatest intensity (Dawson et al.,
2010; Landa et al., 2011), raises concerns for future research.
First, it is important to determine whether comparable treatment effects can be achieved with reduced professional time
since practical, sustainable application in community settings may be compromised due to limited access to and the

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

254

Focus on Autism and Other Developmental Disabilities 29(4)

high costs of qualified and trained professionals. Second,

while current studies highlight potential limitations of parent-mediated approaches, these interventions have greater
longevity given the number of hours a child spends with his
or her parents versus clinicians both in these early stages and
beyond (Woods & Brown, 2011). Thus, it is important to
evaluate innovative ways to maximize dosage and intensity
of parent-implemented interventions and to incorporate
strategies for accurate measurement of the intensity with
which parents implement intervention techniques during
everyday activities.

Agent of Delivery and Setting

The majority of studies included in this review described
delivery of treatment as a shared effort between the professional and the caregiver, with interventions being carried
out in clinical, classroom, and home settings. Several studies did not clearly indicate these methodological details.
Given that purely parent-implemented studies have not
reported significant effects on child social communication
outcomes, it is critical for parent-implemented interventions to be evaluated while controlling for density of parent
implementation, including measures of treatment fidelity
and evaluation of parent learning.
As a result of the documented challenges with generalization of learning for children with ASD, natural environments have been recommended as preferred intervention
contexts. Because treatment provided by caregivers in a
natural environment could be utilized to address issues
related to dosage and intensity, future research should
address the relative effects of providing intervention in natural versus more clinically oriented settings. Related to
issues of context and whether caregivers or professionals
are delivering the treatment is that new skill maintenance
and generalization in natural contexts should be carefully
evaluated. Only a handful of studies reported evaluating
generalization and maintenance, and those findings were
not assessed across a variety of contexts and/or communication partners, with the exception of the KrstovskaGuerrero and Jones (2013) study which evaluated
generalization across communication partners (i.e., interventionist and caregiver) and with novel materials.

Measurement
The challenge of comparing results across study designs
was further compounded by types of outcome measures
administered. Proximal measures were typically used in
single-subject designs to detect specific, incremental
changes in target behaviors (e.g., frequency of parentchild
interactions; Green et al., 2010) that directly correspond to
what was addressed in treatment via direct observations,
whereas distal measures, assessments used to ascertain the

transfer of change from intervention targets, were often

used in group designs (e.g., Vineland Adaptive Behavior
Scales [VABS]; Sparrow, Cicchetti, & Balla, 2005).
Differences in the degree of sensitivity of the two outcome
measures varies depending on the type of outcome (i.e.,
proximal or distal) measured. In concert, these findings
warrant more single-subject and group design studies that
include a combination of proximal and distal measures of
outcome.

Limitations of This EBSR

This EBSR has two primary limitations that are worthy of
note. To ensure that experts in the field vetted study quality,
only peer-reviewed research was accepted in this EBSR.
However, since the likelihood of publishing studies with
significant findings is higher than the likelihood of publishing studies with non-statistically significant results, the risk
of publication bias is high. Only studies written in English
were accepted, which limited the scope of the search for
relevant articles for this EBSR. As is the case with only
accepting peer-reviewed studies, studies on this topic written in other languages could provide another dimension to
the understanding of this topic and/or may contain results
that are principally contrary to the findings in this EBSR.

Conclusion
Although the mixed results described above prevent definitive statements about the efficacy of social communication
interventions for the infant and toddler population with
ASD, the positive findings from this review and previous
reviews (e.g., NRC, 2001; Schertz et al., 2012) suggest benefit from interventions focusing on social communication.
Limited research for this population is available with
respect to some of the specific intervention domains identified as critical by the ASHA (2006) guideline that addresses
diagnosis, assessment, and intervention of difficulties associated with ASD across the life span. As longitudinal
research provides additional evidence as to long-term outcomes associated with acquisition of early social communication skills, a stronger emphasis on outcome goals related
to the core challenges of ASD in prelinguistic-social communication may be revealed. Interventions that have the
potential to be implemented by early intervention systems
that address multiple outcomes and can provide a strong
family component designed to maximize child active
engagement are identified as critical priorities for the next
phase of intervention research.
Authors Note
This systematic review was conducted under the auspices of the
American Speech-Language-Hearing Association; however, this
is not an official position statement of the Association.

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

255

Morgan et al.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest
with respect to the research, authorship, and/or publication of this
article: Three of the authors are salaried employees of the
American Speech-Language-Hearing Association, the organization through which this systematic review was completed.

Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.

Supplemental Material
Supplemental Materials Tables 14 are available at focus.sagepub.
com/supplemental.

References
References marked with an asterisk indicate studies included in
the systematic review.
American Speech-Language-Hearing Association. (2006).
Guidelines for speech-language pathologists in diagnosis,
assessment, and intervention of autism spectrum disorders
across the life span (Guidelines). Retrieved from www.asha.
org/policy
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H.
R. (2009). Introduction to meta-analysis. West Sussex, UK:
John Wiley.
*Cardon, T. A. (2012). Teaching caregivers to implement video
modeling imitation training via iPad for their children with
autism. Research in Autism Spectrum Disorders, 6, 13891400.
*Cardon, T. A., & Wilcox, M. J. (2011). Promoting imitation in
young children with autism: A comparison of reciprocal imitation training and video modeling. Journal of Autism and
Developmental Disorders, 41, 654666.
*Carter, A. S., Messinger, D. S., Stone, W. L., Celimli, S.,
Nahmias, A. S., & Yoder, P. (2011). A randomized controlled trial of Hanens more than words in toddlers with
early autism symptoms. Journal of Child Psychology and
Psychiatry, and Allied Disciplines, 52, 741752.
Cherney, L. R., Patterson, J. P., Raymer, A., Frymark, T., &
Schooling, T. (2008). Evidence-based systematic review:
Effects of intensity of treatment and constraint-induced language therapy for individuals with stroke-induced aphasia.
Journal of Speech, Language, and Hearing Research, 51,
12821299.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
*Dawson, G., Rogers, S., Munson, J., Smith, M., Winter, J.,
Greenson, J., . . .Varley, J. (2010). Randomized, controlled
trial of an intervention for toddlers with autism: The Early
Start Denver Model. Pediatrics, 125, e17e23.
Dollaghan, C. A. (2007). The handbook for evidence-based practice in communication disorders. Baltimore, MD: Paul H.
Brookes.
*Goldstein, H., Kaczmarek, L., Pennington, R., & Shafer, K.
(1992). Peer-mediated intervention: Attending to, commenting on, and acknowledging the behavior of preschoolers with
autism. Journal of Applied Behavior Analysis, 25, 289305.

Green, J., Charman, T., McConachie, H., Aldred, C., Slonims, V.,
Howlin, P., . . . PACT Consortium. (2010). Parent-mediated
communication-focused intervention in children with autism
(PACT): A randomized controlled trial. The Lancet, 375,
21522160.
*Ingersoll, B., Dvortcsak, A., Whalen, C., & Sikora, D. (2005).
The effects of a developmental, social-pragmatic language
intervention on rate of expressive language production in
young children with autistic spectrum disorders. Focus on
Autism and Other Developmental Disabilities, 20, 213222.
*Ingersoll, B., & Schreibman, L. (2006). Teaching reciprocal imitation skills to young children with autism using a naturalistic behavioral approach: Effects on language, pretend play,
and joint attention. Journal of Autism and Developmental
Disorders, 36, 487505.
*Jennett, H. K., Harris, S. L., & Delmolino, L. (2008). Discrete
trial instruction vs. mand training for teaching children with
autism to make requests. Analysis of Verbal Behavior, 24,
6985.
*Jones, E. A., Carr, E. G., & Feeley, K. M. (2006). Multiple
effects of joint attention intervention for children with autism.
Behavior Modification, 30, 782834.
*Kaiser, A. P., Hancock, T. B., & Nietfeld, J. P. (2000). The
effects of parent-implemented Enhanced Milieu Teaching on
the social communication of children who have autism. Early
Education and Development, 11, 423446.
*Kashinath, S., Woods, J., & Goldstein, H. (2006). Enhancing
generalized teaching strategy use in daily routines by parents
of children with autism. Journal of Speech, Language, and
Hearing Research, 49, 466485.
*Kouri, T. A. (1988). Effects of simultaneous communication
in a child-directed intervention approach with preschoolers with severe disabilities. Augmentative and Alternative
Communication, 4, 222232.
*Krstovska-Guerrero, I., & Jones, E. A. (2013). Joint attention in
autism: Teaching smiling coordinated with gaze to respond to
joint attention bids. Research in Autism Spectrum Disorders,
7, 93108.
*Landa, R. J., Holman, K. C., ONeill, A. H., & Stuart, E. A. (2011).
Intervention targeting development of socially synchronous
engagement in toddlers with autism spectrum disorder: A
randomized controlled trial. Journal of Child Psychology and
Psychiatry, and Allied Disciplines, 52, 1321.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer
agreement for categorical data. Biometrics, 33, 159174.
Maglione, M., Gans, D., Das, L., Timbie, J., & Kasari, C.
(2012). Nonmedical interventions for children with ASD:
Recommended guidelines and further research needs.
Pediatrics, 130, S169S178.
Mullen, R. (2007, March 6). The state of the evidence: ASHA
develops levels of evidence for communication sciences and
disorders. The ASHA Leader, 12(3), 89, 2425.
National Research Council. (2001). Educating children with
autism (Committee on Educational Interventions for Children
with Autism, Division of Behavioral and Social Sciences and
Education). Washington, DC: National Academy Press.
Parker, R. I., & Vannest, K. (2009). An improved effect size
for single-case research: Nonoverlap of all pairs. Behavior
Therapy, 40, 357367.

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015

256

Focus on Autism and Other Developmental Disabilities 29(4)

Prizant, B. M., Wetherby, A. M., Rubin, E., & Laurent, A. C.

(2003). The SCERTS model: A transactional, family-centered
approach to enhancing communication and socioemotional
abilities of children with autism spectrum disorder. Journal of
Infants and Young Children, 16, 296316.
*Rocha, M. L., Schreibman, L., & Stahmer, A. (2007). Effectiveness
of training parents to teach joint attention in children with
autism. Journal of Early Intervention, 29, 154172.
*Rogers, S. J., Estes, A., Lord, C., Vismara, L., Winter, J.,
Ftizpatrick, A., & Dawson, G. (2012). Effects of a Brief Early
Start Denver Model (ESDM)Based parent intervention on
toddlers at risk for autism spectrum disorders: A randomized
controlled trial. Journal of the American Academy of Child &
Adolescent Psychiatry, 51, 10521065.
*Schertz, H. H., & Odom, S. L. (2007). Promoting joint attention
in toddlers with autism: A parent-mediated developmental
model. Journal of Autism and Developmental Disorders, 37,
15621575.
*Schertz, H. H., Odom, S. L., Baggett, K. M., & Sideris, J. H.
(2013). Effects of Joint Attention Mediated Learning for toddlers with autism spectrum disorders: An initial randomized
controlled study. Early Childhood Research Quarterly, 28,
249258.
Schertz, H. H., Reichow, B., Tan, P., Vaiouli, P., & Yildirim, E.
(2012). Interventions for toddlers with autism spectrum disorders: An evaluation of research evidence. Journal of Early
Intervention, 34, 166189.
*Schreibman, L., Stahmer, A. C., Cestone Barlett, V., & Dufek,
S. (2009). Brief report: Toward refinement of a predictive
behavioral profile for intervention outcome in children with
autism. Research in Autism Spectrum Disorders, 3, 163172.
*Smith, T., Buch, G. A., & Gamby, T. E. (2000). Parent-directed,
intensive early intervention for children with pervasive developmental disorder. Research in Developmental Disabilities,
21, 297309.
Sparrow, S. S., Cicchetti, D. V., & Balla, D. A. (2005). Vineland
Adaptive Behavior Scales (2nd ed.). Circle Pines, MN:
American Guidance Service.

*Steiner, A. M., Gengoux, G. W., Klin, A., & Chawarska, K.

(2013). Pivotal response treatment for infants at-risk for
autism spectrum disorders: A pilot study. Journal of Autism
and Developmental Disorders, 43, 91102.
Tate, R., McDonald, S., Perdices, M., Togher, L., Schultz, R., &
Savage, S. (2008). Rating the methodological quality of singlesubject designs and n-of-1 trials: Introducing the Single Case
Experimental Design (SCED) Scale. Neuropsychological
Rehabilitation, 18(4), 385401.
*Vernon, T. W., Koegel, R. L., Dauterman, H., & Stolen, K.
(2012). An early social engagement intervention for young
children with autism and their parents. Journal of Autism and
Developmental Disorders, 42, 27022717.
*Vismara, L. A., Colombi, C., & Rogers, S. J. (2009). Can one
hour per week of therapy lead to lasting changes in young
children with autism? Autism, 13, 93115.
Wallace, K. S., & Rogers, S. J. (2010). Intervening in infancy:
Implications for autism spectrum disorders. Journal of Child
Psychology and Psychiatry, 51, 13001320.
Wetherby, A. M., Watt, N., Morgan, L., & Shumway, S. (2007).
Social communication profiles of children with autism spectrum disorders in the second year of life. Journal of Autism
and Developmental Disorders, 37, 960975.
*Williams, G., Perez-Gonzalez, L. A., & Vogt, K. (2003). The
role of specific consequences in the maintenance of three
types of questions. Journal of Applied Behavior Analysis, 36,
285296.
*Wong, V. C. N., & Kwan, Q. K. (2010). Randomized controlled
trial for early intervention for autism: A pilot study of the
Autism 1-2-3 Project. Journal of Autism and Developmental
Disorders, 40, 677688.
Woods, J., & Brown, J. (2011). Integrating family capacitybuilding and child outcomes to support social communication
development in young children with autism spectrum disorders. Topics in Language Disorders, 31, 235246.
*Zachor, D. A., & Itzchak, E. B. (2010). Intervention approach,
autism severity and intervention outcomes in young children.
Research in Autism Spectrum Disorders, 4, 425432.

Downloaded from foa.sagepub.com at Universitas Gadjah Mada on April 10, 2015