ORIGINAL ARTICLE Print ISSN 1738-3684 / On-line ISSN 1976-3026

https://doi.org/10.30773/pi.2019.0229 OPEN ACCESS

Efficacy of Interventions Based on Applied Behavior Analysis

for Autism Spectrum Disorder: A Meta-Analysis
Qian Yu1 , Enyao Li1, Liguo Li1, and Weiyi Liang2
Department of Child Rehabilitation, The Fifth Affiliated Hospital of Zhengzhou University, Zhengzhou, China
1

Department of Rehabilitation, Peking University Shenzhen Hospital, Shenzhen, China

2

Objective To systematically evaluate evidence for the use of interventions based on appied behavior analysis (ABA) to manage various
symptoms of children with autism spectrum disorder (ASD).
Methods Sensitivity analyses were conducted by removing any outlying studies and subgroup analyses were performed to compare the ef-
fectiveness of ABA and early start denver model (ESDM), picture exchange communication systems (PECS) and discrete trial training (DTT).
Results 14 randomized control trials of 555 participants were included in this meta-analysis. The overall standardized mean difference
was d=-0.36 (95% CI -1.31, 0.58; Z=0.75, p=0.45) for autism general symptoms, d=0.11 (95% CI -0.31, 0.54; Z=0.52, p=0.60) for social-
ization, d=0.30 (95% CI -0.02, 0.61; Z=1.84, p=0.07) for communication and d=-3.52 (95% CI -6.31, -0.72; Z=2.47, p=0.01) for expres-
sive language, d=-0.04 (95% CI -0.44, 0.36; Z=0.20, p=0.84) for receptive language. Those results suggested outcomes of socialization,
communication and expressive language may be promising targets for ABA-based interventions involving children with ASD. However,
significant effects for the outcomes of autism general symptoms, receptive language, adaptive behavior, daily living skills, IQ, verbal IQ,
nenverbal IQ, restricted and repetitive behavior, motor and cognition were not observed.
Conclusion The small number of studies included in the present study limited the ability to make inferences when comparing ABA,
ESDM, PECS and DTT interventions for children with ASD. Psychiatry Investig 2020;17(5):432-443

Key Words Autism, Children, Applied behavior analysis, Meta-analysis.

INTRODUCTION to the core symptoms of ASD, but also to a range of co-exist-

ing conditions that individuals with ASD often experience,
Autism spectrum disorder (ASD) is a neurodevelopmental including emotional and behavioral problems (i.e., anxiety,
disorder characterized by early impairments in socialization compulsions, aggression destruction and uncooperative be-
and communication, as well as restricted interests and repeti- havior), sleep problems (i.e., difficulty in falling asleep, super-
tive behaviors.1 Currently, the Centers for Disease Control ficial sleep, early awakening and low sleep efficiency), feeding
and Prevention (CDC) estimates that one in every 59 children and eating problems, gastrointestinal problems, sensory sen-
has ASD.2 Although most children are diagnosed at the age of sitivities, learning and intellectual disabilities, as well as co-
3 years old, approximately 39% are not evaluated for the first morbid health and mental health diagnoses.4 Compared with
time until after 4 years old.2 the core features of ASD, these co-existing conditions can be
ASD is recognized as a major public health concern be- equal or greater for parents and teachers of children with ASD
cause of its early onset, long duration, and high levels of asso- than the core, and have a significant impact on behavior man-
ciated impairments.3 This impairment is attributable not only agement, learning acquisition, and the development of social
Received: September 5, 2019 Revised: November 7, 2019 relationships.5 There are many intervention approaches for
Accepted: February 22, 2020 treating ASD, including applied behavior analysis (ABA), di-
 Correspondence: Qian Yu, MSc
Department of Child Rehabilitation, The Fifth Affiliated Hospital of Zheng-
ets and vitamins, floor time, holding, medication, sensory in-
zhou University, No.3 Kangfuqian Street, Erqi District, Zhengzhou 450000, tegration, speech and music therapy, special education and
China
Tel: +86-15517568226, E-mail: [email protected]
visual schedules.6,7 However, there is little empirical evidence
cc This is an Open Access article distributed under the terms of the Creative Commons to prove the effectiveness of these approaches and the avail-
Attribution Non-Commercial License (https://creativecommons.org/licenses/by- able evidence shows mixed results.8-11
nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduc-
tion in any medium, provided the original work is properly cited. ABA is a scientific approach in which procedures based on

432 Copyright © 2020 Korean Neuropsychiatric Association

 Q Yu et al.

the principles of behavior are systematically applied to iden- based interventions in the improvement of ASD symptoms.
tify environmental variables that influence socially signifi-
cant behavior and are used to develop individualized and METHODS
practical interventions.12,13 This methodology is highly effec-
tive in teaching basic communication, games, sports, social Protocol registration and PRISMA guidelines
interaction, daily living and self- help skills. As the increasing The procedures for this meta-analysis have been registered
number of service providers and certified professionals in in the PROSPERO International prospective register of sys-
the field have suggested, the ABA field has shown even more tematic reviews (No. CRD42018118487), which published
significant growth in the field of behavioral interventions for protocols from systematic reviews prior to the initiation of
children with autism.14,15 Since the mid-1980s, there has been data extraction in an effort to reduce reporting bias.23,24 The
evidence that ABA has contributed to the steady accumula- methods used to conduct this study were in accordance with
tion of intelligence, language and social functions in children the Cochrane Handbook for Systematic Reviews.25 This study
with ASD.16-18 was designed in accordance with the PRISMA guidelines.26
Nowadays, there are also several types of interventions
which are based on ABA and share a common set of core fea- Information sources and search strategy
tures, such as Early Start Denver Model (ESDM), Picture Ex- Two independent researchers identified studies by search-
change Communication Systems (PECS), Discrete Trial Train- ing electronic databases and manually searching for appropri-
ing (DTT) and Pivotal Response Treatment (PRT). ESDM ate published studies and published system reviews. The fol-
uses the teaching strategies which involve interpersonal ex- lowing databases were searched: Pubmed, Embase, Web of
change and positive affect, shared engagement with real-life Science, Cochrane Library, Wanfang and Weipu. The main
materials and activities, adult responsivity and sensitivity to keywords utilized in the article searches included the follow-
child cues, and focus on verbal and nonverbal communica- ing: autism spectrum disorder, autism, autistic disorder, ASD;
tion, based on a developmentally informed curriculum that applied behavior analysis, ABA; discrete trial teaching, DTT;
addresses all developmental domains.19 pivotal response treatment, PRT; picture exchange communi-
PECS is a manualized program that guides children to use cation system, PECS; early start denver model, ESDM; psy-
an exchange-based communication system, which has been chotherap* and cognitive behavi* therap*. It was limited to
a common intervention choice for nonverbal children with the title, abstract or topic, depending on the availability of
ASD in clinical and school settings.20 DTT consists of a series search options in each database. The search was limited to
of direct and systematic instruction methods, which are used journals in English and Chinese. Additionally, the search was
repeatedly until the children acquires the skills and focuses on not limited by date. Thus, all databases were searched from
analyzing the skills into small elements and units.21 PRT is an the earliest indexed date to December 24, 2018.
intervention that focuses on arranging the environment to
promote the use of target structures and then provides oppor- Eligibility criteria and study selection
tunities for children to use the target structures in natural Studies were included if they met the following criteria: 1)
game interactions.22 Even though these interventions have the study should be a randomized controlled trial (RCT); 2)
their own designs and performance forms, they are all consis- participants were between the ages of 0 and 18 years old; 3)
tent with the principles of ABA and show effectiveness in dif- participants were diagnosed with ASD; 4) the treatment used
ferent functions of children with ASD.19-22 in experimental group was based on / derived from applied
The literature on ABA-based interventions for children behavior analysis (DTT, PST, PECS, ESDM and so on); 5)
with ASD has been constantly growing over the past decade. the treatment used in control group was conventional inter-
At present, there are quite a number of studies on psychoso- vention; 6) the study included at least one standardized con-
cial interventions based on ABA in children with ASD. Fur- tinuous outcome measure related to autistic symptom. The
thermore, comparable outcome measures were used in the final selection of studies was performed using tools provided
study to make meta-analysis possible. This meta-analysis in the Cochrane Collaboration Handbook.27
would include ABA-based interventions like ABA, ESDM,
PECS, DTT, PRT and so on. Selection of outcome measures
The primary purpose of this meta-analysis was to systemati- Outcome measures were selected depending on their validi-
cally review the evidence for the use of ABA-based interven- ty and frequency of use. Judgement of the validity of autistic
tions to manage dysfunction in children with ASD. In addition, symptoms measures in the ASD population was based on two
we would also examine the differences among types of ABA- methodologically rigorous reviews which were recently pub-

www.psychiatryinvestigation.org 433
Applied Behavior Analysis for Autism

lished. This study mainly selected outcomes related to high- categories according to our judgment of each area or potential
frequency autistic symptoms (used more than 3 times in all risk of bias: low risk of bias, unclear risk of bias and high risk
included researches). Therefore, the general symptomatic out- of bias. Only methodological strengths and weaknesses associ-
comes of ASD, including socialization outcomes, communica- ated with the results of this meta-analysis were considered
tion outcomes, expressive language outcomes, receptive lan- when assessing the risk of bias. Whether the study should be
guage outcomes, adaptive behavior outcomes, daily living included in the meta-analysis is judged individually based on
skills outcomes and intelligence quotient (IQ) outcomes, were the results of the risk of bias assessments, excluding those with
finally selected in this study. The selected indicators of general higher bias risk.
symptom outcomes for ASD were Mullen Scales of Early Selection bias was assessed based on adequate description of
Learning (MSEL), Autism Diagnostic Observation Schedule random sequence generation and concealment of treatment
(ADOS), Assessment of Basic Language and Learning Skills group allocation. In order to maintain the highest level of sci-
(ABLLS), Aberrant Behavior Checklist (ABC), The Autism entific and methodological rigor, it was determined that only
Diagnostic Interview-Revised (ADI-R), Vineland Adaptive RCTs would be included in this review. Thus, selection bias
Behavior Scales (VABS), Autism Treatment Evaluation Check- would only come from treatment allocation. Due to the nature
list (ATEC) and Childhood Autism Rating Scale (CARS). The of interventions, blinding of participants and personnel was
selected measures for socialization outcome were ADI-R and not feasible in any of the included studies. Thus, all studies had
VABS. The selected measures for communication outcome a high risk of performance bias. Attrition bias was assessed by
were VABS and Psychoeducational Profile (C-PEP). The se- examining the reports of withdrawals and drop-outs. Out-
lected measures for expressive language outcome were MSEL, come data were considered complete if there were no missing
ADOS and Reynell Developmental Language Scales (RDLS). pre- or post-treatment data, or if the study authors had carried
The selected measures for receptive language outcome were out an intent-to-treat analysis. Therefore, reporting bias was
RDLS and MSEL. The selected measures for adaptive behavior evaluated purely based on evidence of selective outcome re-
outcome were VABS and C-PEP. VABS was also used for daily porting provided in the study reports. There was no exclusion
living skills outcome measure. In addition, Differential Ability study based on bias risk assessment.
Scales (DAS) and Stanford-Binet Intelligence Scale (SBIS) were
chosen as measures of IQ outcomes. Summary measures and syntheses of results
If two of the selected outcome measures were used in a study, Data syntheses were performed using Review Manager ver-
one of them was chosen for analysis. sion 5.3 (Cochrane Collaboration software). We assessed con-
tinuous data and analyzed continuous data based on the basis
Data collection process and risk of bias within studies of the available means and standard deviations. There was no
Data extraction and risk of bias assessment were performed clear evidence that the distribution was biased. Assuming that
according to the Cochrane Collaboration Guidelines. All refer- two or more studies were found to be suitable for inclusion
ences found by the search strategy were gathered by the refer- and that those studies were considered to be satisfactory, a
ence management program EndNote X6 (Thomson Reuters, meta-analysis of the results was performed. Since the studies
New York City, USA). All citation sourced from the search measured several outcomes in a nonuniform manner, out-
strategy were transferred to EndNote X6. The first author con- come data were synthesized using standardized mean differ-
ducted the systematic search and the second author verified ence (mean/standard deviation) for both intervention and
inclusion/ exclusion of a subset of studies. The two authors in- control group.
dependently screened the originally selected studies and Higgin’s I2 test was used to describe the impact of heteroge-
agreed on which studies should be selected for the review. neity on the effect estimates in percentage terms. It was cho-
Data extraction and risk of bias assessment were conducted sen over Cochrane’s Q Test (a value of 0.10 used as a cut-off
independently by the first and second authors. In the event of for significance) since the latter had low power when there
a disagreement, resolutions were reached in discussion with were few studies. Higgins et al.28 proposed a tentative classifi-
the third referees, if necessary following inspection of the full cation of I2 values with the purpose of helping to interpret its
paper. The Cochrane Collaborative tool was used to assess the magnitude. Thus, percentages of around 25% (I2=25), 50%
risk of bias in each included study. The tool included the fol- (I2=50), and 75% (I2=75) would mean low, medium, and high
lowing domains: sequence generation, allocation concealment, heterogeneity, respectively. And a random-effect model was
blinding of participants and personnel, blinding of outcome chosen to estimate the effect of intervention.29
assessment, complete outcome data, selective outcome report-
ing, and other sources of bias. Studies were allocated to three

434 Psychiatry Investig 2020;17(5):432-443

 Q Yu et al.

records were screened based on the title and abstract, 1,242 of

1,421 of records identified 2 of additional records which were excluded. 33 of full-text articles were assessed for
through database searching identified through other sources
eligibility and 19 of them were excluded for the following rea-
sons: eight studies were not RCTs; one study could not pro-
1,117 of records after duplicates removed vide full text after contacting the author; three studies only
provided abstracts of conference articles; four studies did not
meet the requirement for participants; three studies did not
1,117 of records screened 1,084 of records excluded meet the requirement of interventions in the control group;
one study did not include relevant outcomes. Finally, 14 RCTs
19 of full-text articles
were included in this review and meta-analysis.32-45
33 of full-text articles
assessed for eligibility excluded, with reasons
Study characteristics
A summary of study characteristics could be found in Table
14 of studies included
in qualitative synthesis
1. A total of 555 participants (278 of experimental groups and
277 of control group) aged 6 to 102 months were included.
Participants were composed of American, European, Latino,
14 of studies included in quantitative Asian, African and multiracial people. All participants in 14
synthesis (meta-analysis)
studies had diagnosis of ASD by clinicians with the ADOS,
Figure 1. PRISMA flow diagram of study selection. the Autism Diagnostic Interview-Revised (ADI-R), the Diag-
nostic and Statistical Manual of Mental Disorders, Fourth
Additional analysis Edition (DSM-VI) or the Diagnostic and Statistical Manual of
Due to the relatively limited research addressing treatment Mental Disorders, Fifth Edition (DSM-5).32-45 Each study in-
options based on ABA for children and adolescents with ASD, cluded at least one standardized continuous outcome measure
it was deemed appropriate to include studies that used applied related to autistic symptoms, such as socialization, communi-
behavior analysis (ABA), discrete trial teaching (DTT), pivotal cation, adaptive behavior, language, verbal IQ, non-verbal IQ,
response treatment (PRT), picture exchange communication inappropriate speech, response, imitation, irritability, noncom-
system (PECS) or early start denver model (ESDM) as inter- pliance, motor, body use, activity level, daily living skills, self-
vention in experimental group. To compare the effectiveness help, IQ, cognitive, early-learning, visual reception, general
of these delivery methods, a subgroup analysis was conducted impression and so on. The scales with higher using frequency
by comparing the confidence intervals of the summary esti- were ADI-R, ADOS, MSEL, VABS, ABC, CARS, ABLLS,
mates in these subgroups (ABA group, DTT group, PRT ATEC and C-PEP.
group, PECS group, and ESDM group). No overlap or mini- Five studies used ABA-based intervention,36,38,40-42 one study
mal overlap between the confidence intervals was considered used DTT,37 five studies used ESDM32-34,39,43 and three studies
statistically significant. Only subgroup analysis of the result were found to use PECS.35,44,45 Eight studies were administrat-
measurements was performed if the overall summary esti- ed by trained therapists,32,33,36,40-42,44,45 while five by teach-
mates were significant. ers35,37-39,43 and one by parents.34 Seven studies encouraged
Because of the small number of studies in each review cate- parents or caregivers to assist with generalization of acquired
gory, it is not possible to formally assess publication bias skills to the home environment and one of them also needed
through funnel plots or statistical tests.30,31 In order to analyze parents or caregivers to cooperate with therapists on home
the impact of outlying studies on summary estimates, sensitiv- visit and supervision.32,33,36,39-41,44 Dawson et al.33 provided con-
ity analysis was carried out by removing each type of outlier tinuous training for parents during semimonthly meeting to
studies. help them use the ESDM strategy in their daily activities.32
The duration of each session was 30 to 120 minutes and the
RESULTS duration of the intervention was between 2 and 36 months.
Intervention settings varied in different studies, such as cen-
Study selection ter, elementary school, mainstream school, institution, kin-
Flow diagram of the search results was shown in Figure 1. dergarten, department of developmental-behavioral pediat-
1,421 records were identified through database searching and rics in hospital and home. All studies were approved by local
2 of additional records were identified from published sys- Institutional Review Board and informed consents were ob-
tematic review. After removing 306 duplicated records, 1,117 tained from the participants’ parents. Gordon et al.35 assigned

www.psychiatryinvestigation.org 435
 Table 1. A summary of study characteristics

436
Exp. Cont. Age Type Exp. Exp. Caregiver Outcome
Study Location
(N) (N) (months) of Exp. dosage administrator involvement measurement
Dawson et al.32 48 21 54.1 ESDM 2 h each time, twice a day, Trained Parents used ESDM strategies Center ADI-R, ADOS, MSEL,
5 days a week, for 2 years therapists during daily activities PDD-BI, VABS
Dawson et al.33 39 21 23.9 ESDM 2 h per session, twice a day, Trained Parent gave family intervention Center ADI-R, ADOS, MSEL,
5 days per week, for 2 years therapists above 5 hours per week VABS, RBS
Estes et al.34 39 21 24 ESDM A high level of intensity, Parents Children only got in-home Home DAS, VABS, ADOS,
for 2 years intervention ABC, ADI-R
Applied Behavior Analysis for Autism

Gordon et al.35 26 28 81.6 PECS Once a day, for 9 months Teachers No involvement Elementary school ADOS, COSMIC,
Videotaped observation
Grindle et al.36 11 18 58.2 ABA 6 h each day, 5 days per week, Therapists Generalization of acquired Mainstream school SBIS-4, LIPS-R, VABS,

Psychiatry Investig 2020;17(5):432-443

for 2 years skills at home ABLLS, ABLLS-R
Hamdan37 13 13 102.0 DTT 35 min per session, Teacher No involvement Institution NVCSS
3 sessions a week, for 12 weeks students
Leaf et al.38 8 7 6.0 ABA 2 h per session, Teachers No involvement Kindergarten GARS-2, SSIS, SRS, WM,
32 sessions in total ABC, SVS
Li et al.39 17 18 38.0 ESDM 2 h a session, 1 session per day, Teachers 2-h ESDM intervention Department of developmental- ABC, CARS, CGI-S
6 days a week, for 12 weeks at home per day behavioral pediatrics in hospital
Sallows et al.40 13 10 33.0 ABA 39 h a week in year 1 Therapists Weekly supervision School VABS
37 h a week in year 2 at home
Smith et al.41 15 13 36.0 ABA 30 h a week, for 2–3 years Student Cooperation with home School SBIS, RDLS, VABS, ACBC,
therapists visit and group visit WIAT, ELM, FSQ
Yan42 42 42 38.0 ABA 30–40 h a week, for 6 months Therapist No involvement Hospital ATEC, C-PEP, ASSS,
PedsQLTM, ABC
Xu et al.43 16 20 47.5 ESDM 30 min a session, 2 sessions a day, Teachers No involvement Center CARS
5 days a week, for 8 weeks
Kong et al.44 20 20 54.0 PECS 30 min a session, 1 session a day, Therapists Cooperation with Department of developmental- C-PEP
5 days a week, for 6 months therapists on PECS behavioral pediatrics in hospital
Liu et al.45 25 25 55.0 PECS 1 session a day, 5 days a week, Therapists No involvement Department of developmental- C-PEP
for 6 months behavioral pediatrics in hospital
Exp.: experimental group, Cont.: control group, ADI-R: Autism Diagnostic Interview-Revised, ADOS: Autism Diagnostic Observation Schedule, MSEL: Mullen Scales of Early Learning,
PDD-BI: Pervasive Developmental Disorder-Behavioral Inventory, VABS: Vineland Adaptive Behavior Scales, RBS: Repetitive Behavior Scale, DAS: Differential Ability Scales, ABC: Aberrant
Behavior Checklist, SBIS-4: Stanford-Binet Intelligence Scale-Fourth Edition, LIPS-R: Leiter International Performance Scale-Revised, ABLLS: Assessment of Basic Language and Learning
Skills, ABLLS-R: Assessment of Basic Language and Learning Skills-Revised, NVCSS: Non-verbal Communication Skills Scale, GARS-2: Gilliam Autism Rating Scale-Second Edition, SSIS:
Social Skills Improvement System, SRS: Social Responsiveness Scale, WM: Walker-McConnell Scale of Social Competence and School Adjustment, SVS: Social Validity Survey, CARS: Child-
hood Autism Rating Scale, CGI-S: Clinical Global Impression-Severity, RDLS: Reynell Developmental Language Scales, ACBC: Achenbach Child Behavior Checklist, WIAT: Wechsler Indi-
vidualized Achievement Test, ELM: Early Learning Measure, FSQ: Family Satisfaction Questionnaire, ATEC: Autism Treatment Evaluation Checklist, C-PEP: Psychoeducational Profile,
ASSS: Autism Social Skills Scale, PedsQLTM: Pediatric Quality of Life Inventory Measurement Model
 Q Yu et al.

all participants into three intervention groups. The patients in

Blinding of participants and personnel (performance bias)

immediate treatment group (ITG; five class groups, 26 chil-
dren) received training immediately after baseline assessment.

Blinding of outcome assessment (detection bias)

The patients in delayed treatment group (DTG; six class

Random sequence generation (selection bias)

groups, 30 children) received training about 9 months later

Incomplete outcome data (attrition bias)

Allocation concealment (selection bias)
and immediately after the second assessment. The patients in

Selective reporting (reporting bias)

no-treatment group (NTG; six class groups, 28 children) re-
ceived no training. In this review, we only selected the pa-
tients in immediate treatment group and no-treatment group.
A summary of study characteristics can be found in Table 1.

Other bias
Risk of bias within studies

Selection bias (random sequence generation and allocation

Estes et al.34
concealment)
All of the included studies were performed with adequate Grindle et al.36

random sequence generation, either manually generated or Dawson et al.33

computer-generated. Dawson et al.32 used random permuted Dawson et al.32
blocks (Fourth Edition), while Li et al.,39 Yan et al.42 and
Li et al.39
Kong et al.44 used randomized digital table. Dawson et al.,32
Yan42
Gordon et al.,35 Hamdan et al.,37 Leaf et al.,38 Li et al.,39 Sal-
lows et al.,40 Smith et al.,41 Yan et al.,42 Xu et al.43 and Kong et Leaf et al.38
al.44 performed adequate allocation concealment. The re- Gordon et al.35
mainder of the included studies indicated that allocation Hamdan37
concealment was implemented, but did not provide sufficient
Sallows et al.40
information about the concealment method.
Smith et al.41

Performance and detection bias (blinding of participants, Liu et al.45

personnel and outcome assessment) Kong et al.44
As previously stated, blinding of participants and person-
Xu et al.43
nel was not possible in any of the included studies. And in all
studies, clinicians rating scales were blind to treatment allo- Figure 2. Risk of bias within studies.
cation so these outcome measures were considered to have a
low risk of detection bias. was d=-0.36 (95% CI -1.31, 0.58; Z=0.75, p=0.45) with no
significant difference between the experimental and control
Attrition and reporting bias (incomplete outcome data conditions. There were high levels of heterogeneity across in-
and selective outcome reporting) cluded studies (I2=94%). A forest plot illustrating these re-
Dawson et al.32 was considered to have a high risk of attri- sults was included in Figure 3.
tion bias due to the deletion of missing data from the study A subgroup analysis was carried out on ABA-based inter-
analysis. The remainder of the included studies were deemed vention36,38,40-42 versus PECS intervention35,44,45 versus ESDM
to have complete outcome data. There was no evidence of se- intervention34,39,43 to compared the outcome of general symp-
lective outcome reporting in any of the studies included. Risk toms of ASD. There was no significant difference in the effec-
of bias within studies is shown in Figure 2. tiveness of interventions among ABA subgroup, ESDM sub-
group and ESDM subgroup. In the ABA-based intervention
Outcome of general symptoms of ASD subgroup, the overall SMD was d=-0.12 (95% CI -1.34, 1.10;
Eleven studies reported the outcome of general symptoms Z=0.19, p=0.85) with on significant difference between ex-
of ASD and we rated the overall quality of the evidence as perimental and control conditions.36,38,40-42 As the high levels
moderate.34-36,38-45 These studies reported 434 participants of heterogeneity existed, we found that the SDM reported by
(215 in the experimental condition and 219 in the control Sallows et al.40 was an outlier. Therefore, we carried out a sen-
condition). The overall standardized mean difference (SMD) sitivity analysis by removing the study of Sallows et al.40 So

www.psychiatryinvestigation.org 437
Applied Behavior Analysis for Autism

-4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4 -50 -25 0 25 50
Favours Favours Favours Favours Favours Favours Favours Favours
(experimental) (control) (experimental) (control) (experimental) (control) (experimental) (control)
General symptoms Socialization Communication Expressive language

-4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4 -4 -2 0 2 4
Favours Favours Favours Favours Favours Favours Favours Favours
(experimental) (control) (experimental) (control) (experimental) (control) (experimental) (control)
Receptive language Adaptive behavior Daily living skills Intelligence quotient

Figure 3. Forest plots.

that the overall SMD in ABA-based intervention subgroup ing this study. The sensitivity analysis altered the results in
was changed to 0.67 (95% CI -0.06, 1.29; Z=2.14, p=0.03) and terms of statistical significance between experimental group
the difference between intervention and control conditions and control group (Z=2.01, p=0.04) and heterogeneity among
was significant. studies (I2=0%, p=0.94). A forest plot illustrating these results
In the PECS intervention subgroup, the overall SMD was was included in Figure 3.
d=-3.67 (95% CI -7.88, 0.54; Z=1.71, p=0.09).35,44,45 There were Subgroup analysis was also conducted to compare ABA-
no significant differences between experimental and control based intervention36,40,41 and ESDM intervention.32-34 It was
conditions and the levels of heterogeneity among studies were noted that there was no study used PECS or DTT intervention
high (I2=98%). After conducting sensitivity analysis by remov- to report outcomes of socialization. There was no significance
ing the studies of Gordon et al.,35 Liu et al.45 and Kong et al.44 in the effectiveness of interventions between ABA subgroup
respectively, the insignificance between experimental and con- and ESDM subgroup. In the ABA-based intervention sub-
trol groups remained (p=0.16; p=0.32; p=0.33). group, there was no significant difference (p=0.60) between
In the ESDM intervention subgroup, the overall SMD was experimental and control groups and its heterogeneity was
d=-0.55 (95% CI -0.92, -0.17; Z=2.86, p=0.04) with signifi- insignificant (I2=0%, p=0.81).36,40,41 In the ESDM intervention
cant difference between the experimental and control condi- subgroup, there was still no significance (p=0.90) between ex-
tions.34,39,43 There was no significant heterogeneity across perimental and control conditions and there were moderate
studies (I2=0%, p=0.76). levels of heterogeneity across studies (I2=78%, p=0.01).32-34 Al-
though the study of Dawson et al.32 was regarded as an outlier
Outcome of socialization and was removed to conduct a sensitivity analysis, the insig-
Six studies reported the outcome of socialization and we nificance between experimental and control conditions re-
rated the overall quality of the evidence as moderate.32-34,36,40,41 mained (p=0.13) with no significant heterogeneity among
These studies reported 200 participants (101 in the experi- studies (I2=0%, p=0.54).
mental condition and 99 in the control condition). The over-
all SMD was d=0.11 (95% CI -0.31, 0.54; Z=0.52, p=0.60) and Outcome of communication
there were moderate levels of heterogeneity across studies Seven studies reported the outcome of communication and
(I2=55%, p=0.05). Since there was significant baseline imbal- we rated the overall quality of evidence as moderate.33,34,36,40,41,44,45
ance [Mean (SD): Experimental group=-0.8 (4.7), Control These studies reported 246 participants (122 in the experimen-
group=3.4 (5.7); d=-0.80, 95% CI -1.41, -0.18] in the study of tal condition and 124 in the control condition). The overall
Dawson et al.,32 we performed sensitivity analysis by remov- SMD was d=0.30 (95% CI -0.02, 0.61; Z=1.84, p=0.07) with no

438 Psychiatry Investig 2020;17(5):432-443

 Q Yu et al.

significance between experimental and control conditions. we rated the overall quality of evidence as moderate.34,40,41,43-45
There were low levels of heterogeneity across studies (I2=33%, These studies reported 210 participants (106 in the experi-
p=0.18). mental condition and 104 in the control condition). Among
Since Estes et al.’s34 study had significant baseline imbal- the six studies, two used ABA-based intervention,40,41 two
ance [Mean (SD): the experimental group=5.3 (20.2), the used ESDM34,43 and two used PECS.44,45 The overall synthesis
control group=10 (17.2); d=-0.24, 95% CI -0.92, 0.43], we indicated insignificance between experimental and control
performed a sensitivity analysis by removing Estes et al.’s34 conditions (p=0.93) with insignificant heterogeneity (I2=0%,
study. The heterogeneity among studies decreased (I2=16%, p=0.70). A subgroup analysis was conducted to compare the
p=0.31) and the difference between experimental and con- effectiveness of ABA, ESDM and PECS and it showed that
trol groups changed to be significant. there were no significant differences between experimental
Among the seven studies that reported outcome of com- and control conditions in each subgroup (p=0.78, p=0.29,
munication, three used ABA-based intervention,36,40,41 two p=0.39). A forest plot illustrating these results was included
used ESDM33,34 and two used PECS.44,45 A subgroup analysis in Figure 3.
to compare ABA versus ESDM versus PECS interventions
was conducted and differences among subgroups were insig- Outcome of daily living skills
nificant (p=1.00). In ABA, ESDM and PECS subgroups, the Five studies reported the outcome of daily living skills and
differences between experimental and control groups were we rated the overall quality of evidence as moderate.33,34,36,40,41
all insignificant (p=0.31, p=0.16, p=0.07). A forest plot illus- These studies reported 77 participants (36 in the experimental
trating these results was included in Figure 3. condition and 41 in the control condition). Among the five
studies, three used ABA-based intervention36,40,41 and two used
Outcome of expressive language ESDM.33,34 The overall SMD was d=0.31 (95% CI -0.22, 0.84;
Four studies reported the outcome of expressive language Z=1.14, p=0.26) with no significant difference between the ex-
and we rated the overall quality of evidence as moder- perimental and control conditions. As there were moderate
ate.33,35,40,41 These studies reported 150 participants (78 in the levels of heterogeneity among studies and Dawson et al.’s33
experimental condition and 72 in the control condition). study showed significant baseline imbalance [Mean (SD): ex-
Among the four studies, two used ABA-based interven- perimental group=-22.6 (11.9), control group=-28.8 (9.2)], we
tion,40,41 one used ESDM33 and one used PECS.35 Thus, we conduct a sensitivity analysis by removing it. It was noted that
did not have adequate studies to carry out a subgroup analy- the heterogeneity among studies decreased (I2=18%) and the
sis. Significant improvement was shown in the overall synthe- differences between experimental and control groups re-
sis (p=0.01). The heterogeneity was significant (p<0.00001, mained insignificant (p=0.57). The subgroup analysis was
I2=97%). Since Gordon et al.’s35 study had significant baseline used to compare the effectiveness of ABA and ESDM, and
imbalance [Mean (SD): experimental group=0.9 (0.2), control there were no significant differences between experimental
group=7.5 (0.1); d=-41.62, 95% CI -49.79, -33.45], we per- and control conditions in each subgroup (p=0.43, p=0.47). A
formed sensitivity analysis by removing it. The sensitivity forest plot illustrating these results was included in Figure 3.
analysis altered the result of heterogeneity (I2=0%, p=0.39). A
forest plot illustrating these results was included in Figure 3. Outcome of IQ
Four studies reported the outcome of IQ and we rated the
Outcome of receptive language overall quality of evidence as moderate.34,36,40,41 These studies
Three studies reported the outcome of receptive language reported 116 participants (57 in the experimental condition
and we rated the overall quality of evidence as moderate.33,40,41 and 59 in the control condition). Three of these studies chose
These studies reported 96 participants (52 in the experimental ABA-based intervention36,40,41 and one chose ESDM interven-
condition and 44 in the control condition). Two of these stud- tion.34 The heterogeneity test showed insignificant differences
ies chose ABA-based intervention40,41 and one chose ESDM across studies (I2=0%, p=0.78) and no significant effectiveness
intervention.33 There was not significant heterogeneity across were found between experimental and control conditions in
studies (I2=0%, p=0.52) and we did not find significant effec- the overall synthesis (p=0.87). A forest plot illustrating these
tiveness in the overall synthesis (p=0.84). A forest plot illus- results was included in Figure 3.
trating these results was included in Figure 3.
Other outcomes
Outcome of adaptive behavior In terms of verbal IQ, nonverbal IQ, restricted and repeti-
Six studies reported the outcome of adaptive behavior and tive behavior, and motor and cognition, there was no signifi-

www.psychiatryinvestigation.org 439
Applied Behavior Analysis for Autism

cant difference in the effectiveness of interventions between suggested that there was possibility of effectiveness in ABA
experimental and control conditions. We did not carry on subgroup and ESDM subgroup. A previous systematic review
the following analysis because only two studies reported in also showed the similar result.46 In the previous review, three
each outcome measure (p=0.56, p=0.65, p=0.30, p=0.32, types of interventions were targeted: 1) behavioral interven-
p=0.57). For the outcomes that only one study recorded, we tions-based essentially on learning theory and on ABA (limit-
did not conduct any test with inadequate studies. ed to not only early intensive behavioral intervention, but also
included ABA programs derived from it; 2) social-communi-
DISCUSSION cation focused interventions, targeting social communication
impairment, as the core symptom of ASD; 3) multimodal de-
We performed a meta-analysis of ABA-based studies (ABA, velopmental interventions targeting a comprehensive range of
ESDM, PECS and DTT) in this study to investigate the over- children’s development. In the subgroup analysis, the behav-
all effectiveness of the intervention programmers for children ioral intervention subgroup included two studies that chose
with ADS, and we observed no significant effects for the out- ABA-based interventions and suggested that there was not
comes of general symptoms of ASD, receptive language, enough evidence to support the treatment effectiveness of
adaptive behavior, daily living skills, IQ, verbal IQ, nonverbal ABA-based interventions.47,48 One of the two studies was in-
IQ, restricted and repetitive behavior, motor and cognition. cluded in our meta-analysis41 and the other study was not in-
However, significant effects were shown on socialization, cluded because its participants had no definite diagnosis of
communication and expressive language. ASD. Even though both the previous study and our study have
This study compared three types of ABA-based interventions consistent conclusion, further study is still needed to accumu-
(ABA, ESDM and PECS). Only one study reported relevant late evidence on the effect on general symptoms of ASD be-
outcomes of DTT so that we could not include it in subgroup cause of limited researches.
analysis. In ABA-based intervention subgroup and ESDM in- Regarding the outcomes of socialization, communication
tervention subgroup, there were significant differences in the and expressive language in this study, we concluded that there
effectiveness between experimental and control conditions was significant effectiveness of ABA-based interventions. The
while PECS intervention not. ABA and ESDM did not have results on daily living skills did not show significant effective-
significant differences in the effectiveness on socialization and ness of ABA-based interventions on this outcome. The results
daily living skills. Additionally, all of ABA, ESDM, and PECS of Makrygianni et al.’s49 study were consistent with our study:
had no significant differences in effects on communication and ABA-based interventions were moderately to very effective in
adaptive behavior. As for other outcomes, there were not avail- improving communication skills (Effect Size: g=0.650) and
able studies to include in the analyses. expressive language skills (Effect Size: g=0.742), moderately
This study conducted a meta-analysis of ABA-based inter- effective in improving socialization (Effect Size: g=0.444),
ventions (ABA, ESDM, PECS, and DTT) for children with lowly effective in improving daily living skills (Effect Size:
ASD. Although several meta-analyses have assessed interven- g=0.138). The effect size (ES) in previous studies was one of
tion programs related to ABA, all of them only chose one type the indexes of magnitude and direction of the treatment ef-
of ABA-based interventions that could not comprehensively fect.50-52 Specifically, ES constituted a quantitative assessment
reflect the effectiveness and some of them included non-ran- of the magnitude and the power of a phenomenon.53 The type
domized controlled trials so that it would introduce signifi- of ES, used in the present study, was the standardized mean
cant bias in the data analysis. Moreover, the quality of evi- change (ESchange) which expressed the difference between pre-
dence for all outcomes were moderate, resulting in more and post-treatment measures. Hedges54 g was used to calcu-
reliable evidence than that produced by previous studies. late the standardized mean change because it constituted a
The small number of available studies has been limited in conservative estimate. For the interpretation of ES, Cohen
the ability to make inferences in comparing the four types of provided Rules-of-Thumb suggesting that 0.2 represented a
ABA-based interventions and investigating each type of inter- small ES, 0.5 represented a medium ES and 0.8 represented a
vention’s strengths and weaknesses in terms of important out- large ES.51
comes. This review also neglected the influences of partici- Studies used quasi-experimental, within-subjects, and pre-
pant baseline levels and parent participation. post design to evaluate the efficacy of ABA-based interventions
Regarding the outcome of autism general symptoms of ASD on ASD. The remaining studies used a quasi-experimental be-
in this study, we concluded that there was not enough evi- tween-groups pre-post design, comparing the performance of
dence to support the effectiveness of ABA-based interventions an experimental group, receiving ABA-based intervention
for treating ASD. However, the results of subgroup analysis while a control group received an eclectic or “treatment-as-

440 Psychiatry Investig 2020;17(5):432-443

 Q Yu et al.

usual” intervention. Only two studies used a random experi- medium-to-strong effect sizes. Medium-to-high confidence
mental between- groups pre-post design and were included in in findings was noted for 81% of the studies in the meta-anal-
our meta-analysis.40,41 The outcomes in some studies were ysis; however, three-fourths of the reviewed studies did not
based on an assessment of larger number of studies and used include treatment integrity, which may affect the ability to
more rigorous analyses to estimate mean effect sizes of each draw conclusion about the effectiveness of the interventions.
outcome. However, due to the selection bias of the included Therefore, it is necessary for children with ASD to ensure
studies, larger sample randomized control trials are still needed. long-term adherence to treatment, for ABA-based interven-
For the outcomes of IQ, verbal IQ, nonverbal IQ, restricted and tions may have slower effect.
repetitive behavior and motor, we did not find relevant studies It was also noted that parental synchrony and sensitivity
to compare and analyze. played a role in helping mediators enhance the communica-
The present study also demonstrated the insignificant ef- tion and social interaction of children with ASD60 and in the
fectiveness of ABA-based interventions for children with effectiveness of enhancing children’s reciprocity of social in-
ASD on receptive language, adaptive behavior and cognition, teraction toward others not only in Aldred et al.61 and Green
which was consistent with the previous study.55 In the previ- et al.,62 but also in the other studies.63-67 However, our study
ous study, thirteen studies met the inclusion criteria and six did not consider the influences of parental synchrony and
of them were randomized comparison trials with adequate sensitivity, which should be improved in the further study.
methodologic quality. Meta-analysis of 4 studies concluded it This review suggested that the outcomes of socialization,
when compared with standard care. ABA intervention pro- communication and expressive language may be promising
grams did not significantly improve the cognitive outcomes targets for ABA-based interventions involving children with
of children in the experimental group who scored a SMD of ASD. However, significant effects for the outcomes of general
0.38 (95% CI 0.09 to 0.84; p=0.11), for receptive language; symptoms of ASD, receptive language, adaptive behavior,
SMD of 0.29 (95% CI 0.17 to 0.74; p=0.22) or adaptive be- daily living skills, IQ, verbal IQ, nonverbal IQ, restricted and
havior; SMD of 0.39 (95% CI 0.16 to 0.77; p=0.20). Among repetitive behavior, motor and cognition were not observed.
the four included studies, two studies were not eligible for The small number of studies included in the present study
this review because they excluded children with ASD who were limited in the ability to make inferences when compar-
had an IQ score less than 5056,57 and the others were included ing ABA, ESDM, PECS and DTT interventions for children
in this review.40,41 Thus, currently, there is inadequate evi- with ASD and investigating the strengths and weaknesses of
dence that ABA-based interventions have better outcomes each type of intervention in terms of important outcomes.
than standard care for children with ASD on receptive lan- More methodologically rigorous researches will be necessary
guage, adaptive behavior and cognition. to ascertain the precise potential of ABA-based interventions
Additionally, we found that long-term, comprehensive for children with ASD.
ABA-based interventions were beneficial to lifelong develop-
ment of children with ASD. In Virués-Ortega’s study, the re- Acknowledgments
This study was supported by the Joint Construction Project of Henan
sults suggested that long-term, comprehensive ABA-based in- Medical Science and Technology Research Plan (No.2018020223).
tervention led to (positive) medium to large effects in terms
of intellectual functioning, language development, acquisition Conflicts of Interest
of daily living skills and social functioning in children with The authors have no potential conflicts of interest to disclose.

ASD.58 Although favorable effects were apparent across all Author Contributions
outcomes, language-related outcomes (IQ, receptive and ex- Conceptualization: Qian Yu. Data curation: Enyao Li, Liguo Li, Weiyi
pressive language, communication) were superior to non- Liang. Formal analysis: Weiyi Liang. Investigation: Weiyi Liang. Methodolo-
verbal IQ, social functioning and daily living skills, with effect gy: Qian Yu. Project administration: Qian Yu. Resources: Qian Yu. Software:
Weiyi Liang. Supervision: Qian Yu. Validation: Qian Yu. Visualization:
sizes approaching 1.5 for receptive and expressive language Qian Yu. Writing—original draft: Weiyi Liang, Qian Yu. Writing—review
and communication skills. Dose-dependent effect sizes were & editing: Qian Yu.
apparent by levels of total treatment hours for language and
ORCID iDs
adaptation composite scores. In Roth et al.’s59 study, adoles-
Qian Yu https://orcid.org/0000-0001-8882-1994
cents and adults with ASD were included and the results sug- Enyao Li https://orcid.org/0000-0001-7804-4984
gested that the behavioral interventions in the areas of aca- Liguo Li https://orcid.org/0000-0003-2794-6587
demic skills, adaptive skills, problem behavior interventions Weiyi Liang https://orcid.org/0000-0002-8278-355X

in the areas of academic skills, adaptive skills, problem behav-

ior, phobic avoidance, social skills, and vocational skills had

www.psychiatryinvestigation.org 441
Applied Behavior Analysis for Autism

REFERENCES Spectrum Disord 2008;2:430-446.

1. American Psychiatric Association. Diagnostic and Statistical Manual 21. Smith T. Discrete trial training in the treatment of autism. Focus Au-
of Mental Disorders (5th Edition). Washington, DC: American Psy- tism Dev Disabil 2001;16:86-92.
chiatric Publishing; 2013. 22. Mohammadzaheri FM, Koegel LK, Rezaee M, Rafiee SM. A random-
2. Centers for Disease Control and Prevention. Prevalence of autism ized clinical trial comparison between Pivotal Response Treatment
spectrum disorder among children aged 8 years – Autism and devel- (PRT) and structured Applied Behavior Analysis (ABA) intervention
opmental disabilities monitoring network, 11 sites, United States, 2014. for children with autism. J Autism Dev Disord 2014;44:2769-2777.
Morbidity and Mortality Weekly Report Surveillance Summaries, 23. Schiavo JH. PROSPERO: an international register of systematic review
April 27th, 2018. protocols. Med Ref Serv Q 2019;38:171-180.
3. Simonoff E, Pickles A, Charman T, Chandler S, Loucas T, Baird G. 24. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et
Psychiatric disorders in children with autism spectrum disorders: al. Preferred reporting items for systematic review and meta-analysis
prevalence, comorbidity, and associated factors in a population-de- protocols (PRISMAP) 2015 statement. Syst Rev 2015;4:1.
rived sample. J Am Acad Child Adolesc Psychiatry 2008;47:921-929. 25. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of
4. Maskey M, Warnell F, Parr JR, Le Couteur A, McConachie H. Emo- Interventions. Chichester: Wiley Online Library; 2008.
tional and behavioural problems in children with autism spectrum 26. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis
disorder. J Autism Dev Disord 2012;43:851-859. JPA, et al. The PRISMA statement for reporting systematic reviews
5. Pearson DA, Loveland KA, Lachar D, Lane DM, Reddoch SL, Man- and meta-analyses of studies that evaluate health care interventions:
sour R, et al. A comparison of behavioral and emotional functioning explanation and elaboration. PLos Med 2009;6:e1000100.
in children and adolescents with autistic disorder and PDD-NOS. 27. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of
Child Neuropsychol 2006;12:321-333. Interventions Version 5.1.0. Available at: http://handbook.cochrane.
6. Green VA, Pituch KA, Itchon J, Choi A, O’Reilly M, Sigafoos J. Inter- org/. Accessed March 20, 2011.
net survey of treatments used by parents of children with autism. Res 28. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsis-
Dev Disabil 2006;27:70-84. tency in meta-analyses. BMJ 2003;327:557-560.
7. Hess KL, Morrier MJ, Heflin LJ, Ivey ML. Autism treatment survey: 29. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
services received by children with autism spectrum disorders in public Trials 1986;7:177-188.
school classrooms. J Autism Dev Disord 2008;38:961-971. 30. Cohen J. Statistical Power Analysis in the Behavioral Sciences. 2nd
8. Foxx RM. Applied behavior analysis treatment of autism: the state of Edition. Hillsdale: Lawrence Erlbaum Associates, Inc; 1988.
art. Child Adolesc Psychiatr Clin N Am 2008;17:821-834. 31. Macaskill P, Walter SD, Irwig L. A comparison of methods to detect
9. Howlin P. Interventions for people with autism: recent advances. Adv publication bias in meta-analysis. Stat Med 2001;20:641-654.
Psychiatr Treat 1997;3:94-102. 32. Dawson G, Jones EJ, Merkle K, Venema K, Lowy R, Faja S, et al. Early
10. Schechtman MA. Scientifically unsupported therapies in the treatment behavioral intervention is associated with normalized brain activity in
of young children with Autism Spectrum Disorders. Pediatr Ann young children with autism. J Am Acad Child Adolesc Psychiatry
2007;36:497-505. 2012;51:1150-1159.
11. Smith T. Outcome of early intervention for children with autism. Clin 33. Dawson G, Rogers S, Munson J, Smith M, Winter J, Greenson J, et al.
Psychol Sci Pract 1999;6:33-49. Randomized, controlled trial of an intervention for toddlers with au-
12. Baer DM, Wolf MM. Some still-current dimensions of applied behav- tism: the Early Start Denver Model. Pediatrics 2009;125:e17-e23.
ior analysis. J Appl Behav Anal 1987;20:313-327. 34. Estes A, Munson J, Rogers SJ, Greenson J, Winter J, Dawson G. Long-
13. Baer DM, Wolf MM, Risley TR. Some current dimensions of applied term outcomes of early intervention in 6-year- old children with autism
behavior analysis. J Appl Behav Anal 1968;1:91-97. spectrum disorder. J Am Acad Child Adolesc Psychiatry 2015;54:580-
14. Cooper JO, Heron TE, Heward WL. Definition and Characteristics of 587.
Applied Behavior Analysis. In: Cooper JO, Heron TE, Heward WL, 35. Gordon K, Pasco G, McElduff F, Wade A, Howlin P, Charman T. A
Editors. Applied Behavior Analysis. 2nd Edition. Upper Saddle River, communication-based intervention for nonverbal children with au-
NJ: Pearson, 2007, p.2-23. tism: what changes? Who benefits? J Consult Clin Psychol 2011;79:447-
15. Shook GL. An examination of the integrity and future of the Behavior 457.
Analyst Certification Board credentials. Behav Modif 2005;29:562- 36. Grindle CF, Hastings RP, Saville M, Hughes JC, Huxley K, Kovshoff H,
574. et al. Outcomes of a behavioral education model for children with au-
16. Fenske EC, Zalenski S, Krantz PJ, McClannahan LE. Age at interven- tism in a mainstream school setting. Behav Modif 2012;36:298-319.
tion and treatment outcome for autistic children in a comprehensive 37. Hamdan MA. Developing a proposed training program based on dis-
intervention program. Res Dev Disabil 1985;549-558. crete trial training (DTT) to improve the non-verbal communication
17. Lovaas OI. Behavioral treatment and normal educational and intellec- skills in children with autism spectrum disorder (ASD). Int J Spec
tual functioning in young autistic children. J Consult Clin Psychol Educ 2018;33:579-591.
1987;55:3-9. 38. Leaf JB, Leaf, JA, Milne C, Taubman M, Oppenheim-Leaf M, Torres N,
18. Remington B, Hastings RP, Kovshoff H, degli Espinosa F, Jahr E, et al. An evaluation of a behaviorally based social skills group for indi-
Brown T, et al. Early intensive behavioral intervention: outcomes for viduals diagnosed with autism spectrum disorder. J Autism Dev Dis-
children with autism and their parents after two years. Am J Ment Re- ord 2017;47:243-259.
tard 2007;112:418-438. 39. Li HH, Li CL, Gao D, Pan XY, DU L, Jia FY. Preliminary application of
19. Smith M, Rogers S, Dawson G. The Early Start Denver Model: A early start denver model in children with autism spectrum disorder.
Comprehensive Early Intervention Approach for Toddlers with Au- Chin J Contemp Pediatr 2018;20:793-798.
tism. In: Handleman JS, Harris SL, Editors. Preschool Education Pro- 40. Sallows GO, Graupner TD. Intensive behavioral treatment for children
grams for Children with Autism. 3. Austin, TX: Pro-Ed Corporation with autism: four- year outcome and predictors. Am J Ment Retard
Inc, 2008, p.65-101. 2005;110:417-438.
20. Angermeier K, Schlosser RW, Luiselli JK, Harrington C, Carter B. Ef- 41. Smith T, Groen AD, Wynn JW. Randomized trial of intensive early in-
fects of iconicity on requesting with the Picture Exchange Communi- tervention for children with Pervasive Developmental Disorder. Am J
cation System in children with autism spectrum disorder. Res Autism Ment Retard 2000;105:269-285.
42. Yan HH. The effect of behavioral analysis therapy on children with au-

442 Psychiatry Investig 2020;17(5):432-443

 Q Yu et al.

tism and its influence on social function. Chin J Convalescent Med 2009;154:338-344.
2018;27:580-582. 56. Eikeseth S, Smith T, Jahr E, Eldevik S. Intensive behavioral treatment
43. Xu Y, Yao J, Yang J. Application of early start denver model for early at school for 4- to 7- year-old children with autism: a one-year com-
intervention on autistic children. Chin J Clin Psychol 2017;25:188-191. parison controlled study. Behav Modif 2002;26:49-68.
44. Kong XY, Song FX, Li H, Xu HW, Wang YN, Li Y, et al. Effect of pic- 57. Eikeseth S, Smith T, Jahr E, Eldevik S. Outcome for children with au-
ture exchange communication system on children with autism. Chin J tism who began intensive behavioral treatment between ages 4 and 7:
Rehabil Theory Pract 2014;11:1086-1088. a comparison controlled study. Behav Modif 2007;31:264-278.
45. Liu XL. The application value of picture exchange communication sys- 58. Virués-Ortega J. Applied behavior analytic intervention for autism in
tem in the rehabilitation and nursing of children with autism. Mod J early childhood: meta-analysis, meta-regression and dose–response
Integr Tradit Chin West Med 2016;25:3634-3636. meta-analysis of multiple outcomes. Clin Psychol Rev 2010;30:387-399.
46. Tachibana Y, Miyazaki C, Ota E, Mori R, Hwang Y, Kobayashi E, et al. 59. Roth ME, Gillis JM, DiGennaro Reed FD. DiGennaro Reed. A meta-
A systematic review and meta-analysis of comprehensive interventions analysis of behavioral interventions for adolescents and adults with
for pre-school children with autism spectrum disorder (ASD). PLoS autism spectrum disorders. J Behav Educ 2014;23:258-286.
One 2017;12:e0186502. 60. Aldred C, Green J, Emsley R, McConachie H. Brief report: Mediation
47. Smith T, Groen AD, Wynn JW. Randomized trial of intensive early in- of treatment effect in a communication intervention for pre-school
tervention for children with pervasive developmental disorder. Am J children with autism. J Autism Dev Disord 2012;42:447-454.
Ment Retard 2000;105:269-285. 61. Aldred C, Green J, Adams C. A new social communication interven-
48. Reitzel J, Summers J, Lorv B, Szatmari P, Zwaigenbaum L, Georgiades tion for children with autism: pilot randomised controlled treatment
S, et al. Pilot randomized controlled trial of a Functional Behavior study suggesting effectiveness. J Child Psychol Psychiatry 2004;45:1420-
Skills Training program for young children with autism spectrum dis- 1430.
order who have significant early learning skill impairments and their 62. Green J, Charman T, McConachie H, Aldred C, Slonims V, Howlin P, et
families. Res Autism Spectr Disord 2013;7:1418-1432. al. Parent-mediated communication-focused treatment in children
49. Makrygianni MK, Gena A, Galanis P, Katoudi S. The effectiveness of ap- with autism (PACT): a randomised controlled trial. Lancet 2010;375:
plied behavior analytic interventions for children with Autism Spectrum 2152-2160.
Disorder: a meta-analytic study. Res Autism Spectr Disord 2018;51: 63. Casenhiser DM, Shanker SG, Stieben J. Learning through interaction
18-31. in children with autism: preliminary data from asocial-communica-
50. Hedges LV, Olkin I. Statistical Methods for Meta-Analysis. San Diego, tion-based intervention. Autism 2013;17:220-241.
CA: Academic Press; 1985. 64. Ichikawa K, Takahashi Y, Ando M, Anme T, Ishizaki T, Yamaguchi H,
51. Lipsey MW, Wilson DB. Practical Meta-Analysis. Los Angeles: Sage et al. TEACCH-based group social skills training for children with
Publications; 2001. high-functioning autism: a pilot randomized controlled trial. Biopsy-
52. O’Mara AJ, Marsh HW, Craven RG. Meta-Analytic methods in educa- chosoc Med 2013;7:14.
tional research: issues and their solutions. Australia: Paper presented at 65. Kaale A, Smith L, Sponheim E. A randomized controlled trial of pre-
Australian Association for research in education annual conference; school-based joint attention intervention for children with autism. J
2005. Child Psychol Psychiatry 2012;53:97-105.
53. Kelley K, Preacher KJ. On effect size. Psychol Methods 2002;17:137- 66. Kasari C, Gulsrud AC, Wong C, Kwon S, Locke J. Randomized con-
152. trolled caregiver mediated joint engagement intervention for toddlers
54. Hedges LV. Distribution theory for Glass’s estimator of effect size and with autism. J Autism Dev Disord 2010;40:1045-1056.
related estimators. J Educ Behav Stat 1981;6:107-128. 67. Lawton K, Kasari C. Brief report: longitudinal improvements in the
55. Spreckley M, Boyd R. Efficacy of applied behavioral intervention in quality of joint attention in preschool children with autism. J Autism
preschool children with autism for improving cognitive, language, and Dev Disord 2012;42:307-312.
adaptive behavior: a systematic review and meta-analysis. J Pediatr

www.psychiatryinvestigation.org 443