Journal of Child Psychology and Psychiatry **:* (2014), pp **–** doi:10.1111/jcpp.

12354

A randomized controlled trial of Pivotal Response

Treatment Group for parents of children with autism
Antonio Y. Hardan,1 Grace W. Gengoux,1 Kari L. Berquist,1 Robin A. Libove,1 Christina
M. Ardel,1 Jennifer Phillips,1 Thomas W. Frazier,2 and Mendy B. Minjarez3
1
Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA; 2Center for Autism,
Cleveland Clinic, Cleveland, OH; 3Seattle Children’s Hospital/Seattle Children’s Research Institute,
Seattle, WA, USA

Background: With rates of autism diagnosis continuing to rise, there is an urgent need for effective and efficient
service delivery models. Pivotal Response Treatment (PRT) is considered an established treatment for autism
spectrum disorder (ASD); however, there have been few well-controlled studies with adequate sample size. The aim of
this study was to conduct a randomized controlled trial to evaluate PRT parent training group (PRTG) for targeting
language deficits in young children with ASD. Methods: Fifty-three children with autism and significant language
delay between 2 and 6 years old were randomized to PRTG (N = 27) or psychoeducation group (PEG; N = 26) for
12 weeks. The PRTG taught parents behavioral techniques to facilitate language development. The PEG taught
general information about ASD (clinical trial NCT01881750; http://www.clinicaltrials.gov). Results: Analysis of
child utterances during the structured laboratory observation (primary outcome) indicated that, compared with
children in the PEG, children in the PRTG demonstrated greater improvement in frequency of utterances (F(2,
43) = 3.53, p = .038, d = 0.42). Results indicated that parents were able to learn PRT in a group format, as the
majority of parents in the PRTG (84%) met fidelity of implementation criteria after 12 weeks. Children also
demonstrated greater improvement in adaptive communication skills (Vineland-II) following PRTG and baseline
Mullen visual reception scores predicted treatment response to PRTG. Conclusions: This is the first randomized
controlled trial of group-delivered PRT and one of the largest experimental investigations of the PRT model to date.
The findings suggest that specific instruction in PRT results in greater skill acquisition for both parents and children,
especially in functional and adaptive communication skills. Further research in PRT is warranted to replicate the
observed results and address other core ASD symptoms. Keywords: Language deficits, treatment fidelity, group
parent training, social communication, naturalistic developmental behavior intervention.

is an evolution of ABA that also incorporates moti-

Introduction
vational variables into trials and is conducted in the
Autism spectrum disorder (ASD) is characterized by
child’s natural environment. The emerging support
social communication deficits and repetitive behav-
for these models warrants additional large-scale,
iors (American Psychiatric Association, 2013). With
well-controlled studies.
greater than 1% of children carrying an ASD diag-
Parent involvement is critical for effective inter-
nosis (Center for Disease Control, 2014), the devel-
vention with children with ASD (National Research
opment of empirically supported treatments for ASD
Council, 2001). A few treatment models increase
is a critical priority. While early intervention
treatment intensity by training parents to deliver
improves cognitive, language, and adaptive skills
intervention (Steiner, Koegel, Koegel, & Whitney,
for many children (Dawson & Burner, 2011), a broad
2012). Research shows parents can implement behav-
range of interventions are being used with varying
ioral interventions (Koegel, Bimbela, & Schreibman,
empirical support (Vismara & Rogers, 2010). Early
1996) and autism symptoms subsequently improve
studies of Applied Behavior Analysis (ABA) using
(Symon, 2005). Parent education may also promote
Discrete Trial Training (DTT; Lovaas, 1987) and
generalization to natural environments (Steiner
more recent studies of intensive behavioral interven-
et al., 2012) and improve parent psychological
tion have documented significant progress in cogni-
well-being (Minjarez, Mercier, Williams, & Hardan,
tive and language development (Reichow, 2012).
2013). Therefore, research on interventions that
There is also increasing support for treatments,
involve parents in treatment delivery is needed.
recently termed Naturalistic Developmental Behav-
Parent education is a core component of PRT
ior Interventions (NDBIs; Rogers et al., 2014;
(Koegel, Koegel, Harrower, & Carter, 1999), an NDBI
Schreibman et al., 2014) which combine ABA-based
which uses ABA principles to target ‘pivotal’ areas
treatment with a child-led developmental approach,
(e.g. motivation). The focus on pivotal areas is
as in the Early Start Denver Model (ESDM; Dawson
hypothesized to result in improvements in broad
et al., 2010). Pivotal Response Treatment (PRT;
areas of functioning not directly targeted with inter-
Mohammadzaheri, Koegel, Rezaee, & Rafiee, 2014)
vention (e.g. joint attention; Koegel, Koegel, & Brook-
man, 2005). Available evidence has demonstrated
Conflict of interest statement: No conflicts declared. the efficacy of PRT for teaching social communica-

© 2014 Association for Child and Adolescent Mental Health.

Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA
2 Antonio Y. Hardan et al.
tion skills (Bryson et al., 2007). However, the major-
ity of PRT research has involved training small
numbers of families individually and research has
just begun to investigate dissemination through
group therapy models (Baker-Ericz
en, Stahmer, &
Burns, 2007). In an uncontrolled trial of PRT parent
training group (N = 17), the majority of parents
learned PRT within 10 weeks and their children
showed gains in functional communication (Minj-
arez, Williams, Mercier, & Hardan, 2011).
The goal of the present study was to conduct a
randomized controlled 12-week trial to examine the
effectiveness of PRT, taught to parents in a group
format, in targeting functional communication defi-
cits in young children with ASD. A parent psycho-
education control group was selected to examine the
unique contribution of parent training in PRT. The
primary goals of the study were to evaluate: (a)
whether parents learn PRT from a 12-week parent
group, and (b) whether children of parents partici-
pating in the PRT group exhibit greater improvement
in functional communication skills than those
receiving general psychoeducation.
Methods
Study design
This investigation involved a randomized controlled 12-week
trial examining the effectiveness of Pivotal Response Treatment
Group (PRTG) in targeting functional communication deficits
in young children with autism. This study was approved by
Stanford University’s institutional review board, and registered
in the Clinical Trials database (NCT01881750; http://www.
clinicaltrials.gov). The full trial protocol is available upon
request.
Participants
Recruitment occurred over 2 years (July 2010–June 2012).
Participants were recruited through distribution of fliers at
clinics and autism awareness events, referral by local profes-
sionals, and word of mouth. Following informed consent, a
comprehensive evaluation, including review of medical/psy-
chiatric history and completion of psychological assessment
batteries, was conducted to determine study eligibility.
Inclusion and exclusion criteria
Participants included children: (a) 2–6 years old, (b) diagnosed
with autism based on DSM-IV-TR criteria, Autism Diagnostic
Interview-Revised (ADI-R; Lord, Rutter, & Le Couteur,
1994), Autism Diagnostic Observation Schedule (ADOS; Lord
et al., 2000), and expert clinical judgment, (c) communica-
tion-delayed on Preschool Language Scale, 4th Edition (PLS-4;
Zimmerman, Steiner, & Pond, 2002; 2- and 3-year olds were ≥
1 SD below age level, 4-year olds ≥ 2 SDs, and 5–6-year olds ≥ 3
SDs), (d) able to vocalize with intent when prompted by a
clinician during the screening visit, (e) with stable concomitant
interventions (e.g. ABA, special education, medications, bio-
medical treatments) for 1 month prior to participation with no
planned changes, and (f) with at least one parent available to
consistently participate. Exclusion criteria were: (a) participa-
tion in more than 60 min weekly of individual speech therapy,
(b) comorbid severe psychiatric disorder, (c) genetic abnormal-
ity, and (d) active medical problems. No changes in inclusion or
exclusion criteria were applied during the study. The rationale
for limiting concomitant speech therapy was due to our
interest in studying the effects of PRTG on acquisition of
functional communication skills and the concern that the
inclusion of children receiving concurrent intensive speech
therapy would present a confound. Given that the majority of
children in our area get 30–60 min of speech therapy per week
it was not practical to exclude children receiving any speech
therapy; therefore, the decision was made to limit participation
to children receiving routine but not intensive speech therapy.
Procedures
After baseline measures, eligible families were randomized in
parallel (1:1) to the treatment or control group. Stratification
was based on gender, age (2–4:6 vs. 4:7–6:11 years), and
intensity of ABA treatment (<10 vs. ≥ 10 hr/week). The
rationale for these procedures was based on the hypothesis
that younger children might respond better to treatment and
the concern that there might be differences in intensity of
community interventions (i.e. ABA treatment) which could
influence study outcomes within a small sample. Randomiza-
tion was done using a coin flip (www.random.org) by a senior
investigator not involved in the clinical assessment or treat-
ment. Sequence generation was done by the project coordina-
tor who was not involved in the clinical ratings. After
randomization, the senior investigator informed the coordina-
tor about the group assignment. All data were collected during
visits to the autism program offices within a large academic
medical center and outcome measures were administered at
baseline, week 6, and week 12. Data were managed using
REDCap electronic data tools (Harris et al., 2009) hosted at the
Stanford Center for Clinical Informatics.
Pivotal response treatment group (PRTG)
Pivotal response treatment training lasted 12 consecutive
weeks, with one session per week. The group was led by
psychologists specializing in PRT, utilizing the manual, How to
Teach Pivotal Behaviors to Children with Autism (Koegel et al.,
1989) and a standard set of PRT teaching materials and video
examples (Minjarez et al., 2011). Eight 90-min visits were
parent-only group sessions consisting of 4–6 parents and 1–2
leading clinicians. For four of the visits, parent–child dyads
met individually with a clinician (60 min).
Psychoeducation group (PEG)
The PEG also lasted 12 weeks, with a meeting each week. The
curriculum was based on an existing autism parent psycho-
education program at the university and was taught by
clinical psychology graduate students supervised by a
licensed psychologist. Ten sessions were parent-only group
meetings (90 min), and two were individual meetings (60 min)
between parent–child dyads and a psychologist (see online
supplemental material for detailed information on PRTG and
PEG, Appendix S1 and S2).
Measures
Diagnostic instruments and screening. The ADI-R
(Lord et al., 1994) and ADOS (Lord et al., 2000) were adminis-
tered for all study participants. The PLS-4 (Zimmerman et al.,
2002) was used to assess baseline language level and as a
secondary outcome measure at week 12. The Mullen Scales of
Early Learning (Mullen, 1995) assessed cognitive functioning.
Primary outcome measures. Primary outcome mea-
sures included child frequency of utterances and parent
fidelity of treatment implementation. Both measures were
© 2014 Association for Child and Adolescent Mental Health.

obtained from a structured laboratory observation (SLO) at
baseline, week 6, and week 12, and videotaped and scored by
independent raters. SLO included a 10-min period during
which parents were instructed to try getting the child to
communicate as much as possible.
Videos of parent–child interaction across both groups were
scored for parent fidelity of PRT implementation with measures
widely used in PRT research (Bryson et al., 2007; Symon,
2005). At least 80% correct on each of six PRT techniques was
required to meet fidelity criteria. For each 10-min probe, raters
also tallied the child’s functional verbal utterances. Frequency
counts were calculated for total functional utterances and for
each utterance type: (a) unintelligible, (b) imitative, (c) verbally
prompted, (d) nonverbally prompted, and (e) spontaneous.
Nonverbally prompted utterances were those that occurred
following a clear nonverbal prompt from the parent (parent
held item up or deliberately paused the activity and waited
expectantly for a verbal request), whereas spontaneous utter-
ances occurred in the absence of any effort from the adult to
elicit a response.
Raters were bachelors- and masters-level research assis-
tants trained in scoring procedures by three senior investiga-
tors. Initial training was conducted until raters consistently
reached target interobserver agreement (IOA; e.g. 80%). Two
raters then independently scored at least 33% of the videos
randomly selected from the total set. All raters were blind to
group assignment and treatment phase. For parent fidelity,
IOA was calculated, as well as kappa (Cohen, 1960) to correct
for chance agreement. The average IOA across sessions was
88% (range: 70–100%; k = .78). For child utterances, intra-
class correlation coefficients (ICC) were calculated to assess
reliability between raters. Results indicated high levels of
agreement on total number of utterances [ICC (2,1) = 0.993]
and across all utterance types (Unintelligible = 0.947; Imitative
= 0.962; Verbally Prompted = 0.982; Nonverbally Prompted =
0.938; Spontaneous = 0.910; Shrout & Fleiss, 1979).
Secondary outcome measures. The following mea-
sures were obtained: MacArthur-Bates Communicative Devel-
opment Inventories (CDI; Fenson et al., 2007) Words and
Gestures and Words and Sentences versions; Vineland Adap-
tive Behavior Scales, Second Edition (Vineland-II; Sparrow,
Cicchetti, & Balla, 2005) Communication subscale; Social
Responsiveness Scale (SRS; Constantino & Gruber, 2005); and
Clinical Global Impression Scale (Guy, 1976) severity and
improvement subscales (CGI-S and CGI-I). The CGI ratings
were assessed by a psychologist blind to group assignment and
treatment phase, and were specifically focused on social and
communication skills (Appendix S1 for additional detail
regarding outcome measures).
Treatment monitoring measures. At baseline and
week 12, parents in both groups completed a brief question-
naire to report their children’s existing autism interventions
and a weekly concomitant therapies log was collected to
document any changes. In the PRTG, parents were asked to
practice PRT daily and video record at least 10 min each week
for review during group. PEG parents were asked to video
record 10-min parent–child interactions at baseline, week 6,
and week 12.
Statistical analyses
To examine the primary hypothesis that PRT would increase
total child utterances in the SLO, we computed mixed effects
regression models with Treatment Group (two levels: PRTG vs.
PEG), Time (three levels: baseline, week 6, and week 12), and
their interaction as fixed-effects covariates. The interaction of
Treatment Group X Time directly tests the hypothesis by
examining whether treatment groups showed a different
© 2014 Association for Child and Adolescent Mental Health.
RCT of pivotal response treatment group 3
pattern of change in utterances across study time points.
Significant interactions were followed by independent samples
t-tests at each postbaseline time point (Weeks 6 and 12) to
determine the time of onset of treatment effects. Additional
mixed effect regression models were separately computed for
each utterance type. Exploratory analyses using the same
modeling approach were computed for Vineland Communica-
tion standard scores, Vineland Receptive and Expressive
subscale standard scores, SRS total raw scores, number of
words on the CDI out of 396 and 680, CDI mean length of
longest utterance, PLS-4, and CGI-Severity (overall social/
communication symptoms only). The effects of sex, age, and IQ
on the primary outcome (total child utterances) were also
investigated by reestimating the above model and including
these variables as fixed-effects covariates.
Mixed effects regression models are advantageous com-
pared to repeated measures ANOVA in that they accommodate
missing time points, utilize all available data, and therefore
can be considered truly intent-to-treat models. Model fit was
considered by iteratively examining random effects for inter-
cept and slope and alternative repeated measures covariance
structures (Peugh & Enders, 2005). The significance of
random intercept and slope parameters was examined to
determine whether individuals showed significant variability
in outcome measures at baseline and individual differences in
change-over-time. Initial models indicated that an unstruc-
tured repeated measures covariance matrix fit consistently
better than autoregressive and simpler alternative structures.
Final models presented are based on the unstructured
covariance matrix. Type 1 error rate of 0.05 was used for
analyses of the primary outcome, subscales, and all explor-
atory measures. Multiple comparison correction was not
performed for exploratory measures, as this is a small initial
study and the purpose of the exploratory analyses was to
better understand the specificity of the treatment effects for
future study planning.
Results
Study population
One hundred and thirty-five potential subjects were
screened after inquiring about the study (Figure 1);
104 signed a consent form. Forty one did not meet
eligibility criteria and 10 decided not to participate
before baseline measures. Fifty-three subjects (41
males, 12 females; 4.1  1.2 years; range: 2.1–6.9)
were randomized (PRTG = 27; PEG = 26). Target
sample size for ending the trial (52) was determined
by power analysis based on a pilot study (Minjarez
et al., 2011). There were no differences between the
two groups on any child demographic or clinical
baseline measures (Table 1). The majority (94%) of
participating children were in school, primarily in
special day classes (73%) and receiving an average of
19.2 hr of school per week (range = 0–45). Approx-
imately half of the subjects (52%) were receiving ABA
treatment with 34% of children receiving more than
10 hr per week. Almost all children were receiving
speech therapy (91%) with an average of 54 min
individual therapy per week. Groups did not differ in
the number of total outside intervention hours
received (PEG: 26.1  11.1; PRTG: 23.9  10.5;
t = .74; p = .46). Participating parents were primar-
ily female (92%) and all had at least partial college
4 Antonio Y. Hardan et al.

135 subjects screened

104 subjects enrolled in

the study

10 decided not to 41 did not meet inclusion/

participate exclusion criteria

53 randomized

27 subjects assigned 26 subjects assigned

to PRTG to PEG

2 subjects excluded due 3 subjects excluded due to

to changes in changes in concomitant
concomitant therapies 25 subjects (92%) 22 subjects (84%) therapies
completed the study completed the study
1 subject lost to follow-up*

Figure 1 Patient flow diagram for Pivotal Response Treatment Group (PRTG) versus Psychoeducation Group (PEG) in the treatment of
children with autism. *Available data from this subject were included in analyses

education, with 80% college graduates and 35% with

Structured laboratory observation (SLO)
graduate degrees. Median annual household income
was reported in the $125,000–150,000 range. There Table 2 presents results for total, imitative, and
were no significant differences on these variables nonverbally prompted utterances from the SLO.
between groups (p > .1). Five subjects were excluded Irrespective of group, children showed significant
(two PRTG and three PEG) due to changes in improvement in total number of utterances across
concomitant therapies during the trial and their study time points (F(2, 43) = 6.12, p = .005; Fig-
data were not included in the analyses. Forty-seven ure 2). However, individuals receiving PRTG showed
subjects (PRTG = 25; PEG = 22) completed the study greater improvement (F(2, 43) = 3.53, p = .038).
with one participant in the PEG group lost to The treatment effect was most apparent for imita-
follow-up. No adverse effects were noted in either tive (F(2, 43) = 7.67, p = .001) and nonverbally
group. prompted utterances (F(2, 43) = 7.06, p = .002;
Table 2). Spontaneous utterances showed a nonsig-
nificant trend toward greater improvement in PRTG
Table 1 Baseline comparison of participants with autism in
the Pivotal Response Treatment Group (PRTG) and Psychoed- (F(2, 44) = 2.99, p = .060). The treatment effect was
ucation Group (PEG) not significant for unintelligible (F(2, 43) = 0.27,
p = .762) or verbally prompted utterances (F(2, 43) =
PRTG (SD) PEG (SD) 0.32, p = .725).
N 25 23a No parent met fidelity of implementation at base-
Male/Female 19/6 17/6 line. At week 12, 21 of 25 parents in PRTG, and none
Mean age in years 4.1 (1.2) 4.1 (1.3) in the PEG, met fidelity of PRT implementation.
SLO total utterances 45.7 (23.1) 40.8 (22.4) Including treatment fidelity as a covariate in regres-
CGI-severity score 5.2 (0.9) 5.1 (0.8)
Development quotient 52.8 (16.4) 53.5 (14.8)
sion models further qualified the above-described
Social Responsiveness Scale 77.9 (11.4) 78.9 (12.8) treatment effects. Within the PRTG, caregivers who
demonstrated fidelity at week 12 had children with
SD, standard deviation; CGI, Clinical global impression; greater total (F(2, 42) = 6.79, p = .003; Figure S1)
Development quotient measured by Mullen Scales; No statis-
and imitative (F(2, 46) = 6.95, p = .002) utterances
tical differences between the two groups on any of the baseline
clinical characteristics. relative to those who did not meet fidelity. Fidelity
a
Includes one subject who was randomized but did not did not modify the treatment effect for verbally
complete the study. prompted (F(2, 41) = 0.92, p = .405), nonverbally

© 2014 Association for Child and Adolescent Mental Health.

 RCT of pivotal response treatment group 5

Table 2 Treatment responses of participants with autism whose parents were assigned to Pivotal Response Treatment Group
(PRTG) or Psychoeducation Group (PEG)

Group X Time
Mean (SD) Interaction

Baseline Week 12

PRTG PRTG
(n = 25) PEG (n = 23) (n = 25) PEG (n = 22) F p Cohen’s d (Week 12)

SLO
Total utterances 45.7 (23.1) 40.8 (22.4) 64.5 (28.9) 51.4 (33.8) 3.53 .038 0.42
Unintelligible 27.64 (17.8) 22.7 (14.2) 26.1 (14.9) 24.6 (23.3) 0.27 .762 0.08
Imitative 9.7 (7.6) 6.1 (5.4) 19.4 (14.2) 7.1 (7.8) 7.67 .001 1.06
Verbally prompted 7.7 (9.3) 10.7 (10.4) 14.5 (15.2) 16.0 (17.0) 0.32 .725 0.09
Nonverbally prompted 0.2 (0.4) 0.5 (1.1) 2.7 (3.8) 0.1 (0.3) 7.06 .002 0.94
Spontaneous 0.5 (1.4) 0.8 (1.5) 0.9 (1.4) 0.4 (0.6) 2.99 .06 0.46
Vineland PRTG PEG PRTG PEG
Communication standard 69.9 (16.3) 71.6 (15.4) 78.9 (18.9) 72.8 (16.5) 3.80 .041 0.34
score
Expressive raw score 26.7 (12.1) 31.3 (18) 41.7 (14.7) 34 (18.9) 6.95 .004 0.46
Receptive raw score 17.8 (7.1) 18.1 (6.6) 21.5 (14.7) 18.9 (6.5) 4.27 .028 0.23
CDI
Mean length of longest 2.1 (1.7) 2.4 (1.7) 3.7 (2.1) 2.8 (1.4) 3.09 .059 0.50
utterance
Words produced out of 396 137.1 (118.1) 169.5 (134.2) 172.2 (123.6) 215.0 (118.3) 0.213 .736 0.35
Words produced out of 680 229.3 (183.8) 179.1 (163.8) 289.1 (181.9) 239.9 (187.1) 1.30 .284 0.27
PLS
Expressive standard score 62.6 (11.2) 63.9 (11.3) 63.9 (11.6) 63.0 (13.4) 2.00 .165 0.07
SRS total 77.9 (11.4) 78.9 (12.8) 74.9 (12.4) 80.6 (10.7) 0.26 .776 0.49
CGI-Severity 5.2 (0.9) 5.1 (0.8) 4.6 (0.9) 5 (0.8) 6.84 .003 0.47
CGI-Improvement – – 2.4 (0.7) 3.2 (0.7) 15.97 .001 1.14

Vineland: Vineland Adaptive Behavior Scales, Second Edition; CDI: MacArthur-Bates Communicative Development Inventories;
PLS: Preschool Language Scale, 4th Edition; SRS, Social Responsiveness Scale; CGI: Clinical global impression (focused on social
and communication symptoms).

90 trend was observed for CDI mean length of longest

80 PEG utterance (F(2, 32) = 3.09, p = .059) with children in
70 PRTG the PRTG saying an average of 2.1-word (SD1.7)
Total utterances

60 utterances at baseline and an average of 3.7-word

50 (2.1) utterances at week 12, and children in the PEG
40 increasing from an average mean length of utterance
30 of 2.4 (1.7) to 2.8 (1.4). There were no significant
20 treatment effects for SRS total raw scores (F(2,
10 42) = 0.26, p = .776), CDI total words out of 396 (F
0 (2, 42) = .213, p = .736), and out of 680 (F(2,
Baseline Week 6 Week 12 42) = 1.30, p = .284), or PLS-4 expressive language
standard scores (F(1,40) = 2.00, p = .165). The treat-
Figure 2 Effect of Pivotal Response Treatment Group (PRTG) ment effect was observed for CGI-Severity of social
versus Psychoeducation Group (PEG) on total number of utter-
and communication symptoms (F(2, 42) = 6.84,
ances measured during SLO (F(2, 43) = 6.12, p = .005)
p = .003), and CGI-Improvement ratings were signif-
prompted (F(2, 42) = 0.87, p = .425), or spontaneous icantly better in PRTG relative to PEG at weeks 6 and
utterances (F(2, 44) = 0.57, p = .567). 12 (F(1, 44) = 15.97, p < .001).
The pattern of treatment effects for exploratory
measures did not change with fidelity included as a
Exploratory measures
covariate, and fidelity did not significantly modify the
A significant treatment effect was also observed for treatment effect. Finally, excluding participants
the Vineland-II Communication scale (F(2, receiving PRTG who did not meet fidelity did not
19) = 3.80, p = .041), with individuals receiving alter the findings.
PRTG showing larger improvements in communica-
tion (see Table 2). Vineland-II subscales showed a
Predictors of response
significant treatment effect for expressive (F(2,
23) = 6.95, p = .004) and receptive language raw Older children with higher baseline IQ had more
scores (F(2, 21) = 4.27, p = .028). A nonsignificant total utterances, but there were no effects of sex and

© 2014 Association for Child and Adolescent Mental Health.

6 Antonio Y. Hardan et al.
50.00
Change in total utterances from
PEG
40.00
PRTG
baseline to week 12
30.00
20.00
10.00
0.00
–10.00
–20.00
Low visual reception
Figure 3 Total utterances measured during SLO in Pivotal
Response Treatment Group (PRTG) and Psychoeducation Group
(PEG) with low and high Mullen visual reception scores (median
split; F(2, 40) = 3.96, p = .027)
the treatment effect did not differ across age or sex
(largest F(5, 41) = 0.90, p = .492). Baseline Mullen
visual reception (VR) scores were a significant
predictor of treatment response for total and imita-
tive utterances (smallest Group X Time X Mullen VR
interaction F(2, 40) = 3.96, p = .027), accounting for
14% and 18% of the variance in treatment response
for total and imitative utterances, respectively. Fig-
ure 3 displays total utterances in children receiving
PEG and PRTG with low and high VR scores
(median split). Treatment response was not modified
by baseline PLS total language standard scores,
SRS total t-scores, CDI words out of 396 or 680,
or CDI sentence length (largest F(2, 41) = 1.50,
p = .236).
Discussion
Findings from this first randomized controlled trial of
group-delivered PRT support using PRT to improve
functional communication skills for young children
with ASD. PRTG resulted in greater improvement in
frequency of functional utterances compared to PEG.
These findings are consistent with a recent random-
ized controlled trial in older children with ASD
reporting increase in mean length of utterance in
the PRT group compared to controls (Mohammadza-
heri et al., 2014). In addition, PRTG participants
demonstrated greater improvement in adaptive com-
munication skills as reported by parents. Finally,
baseline Mullen VR scores predicted treatment
response to PRTG. While PRT is considered an
established autism treatment (National Autism Cen-
ter, 2009), this trial evaluated one of the largest
groups of children to be treated with PRT in a
research study to date and participants were
well-characterized, allowing for further analysis of
factors contributing to treatment response. The
parent education programs were both provided
against a backdrop of community-based treatments
including school and speech programming, as well
as some ABA programs which were maintained
stable during the study. The majority of parents
(21/25) receiving PRTG demonstrated at least 80%
PRT fidelity of implementation by week 12 and
retention rate was high in PRTG (92%) suggesting
good acceptability of the treatment.
Child improvements were observed in increased
utterances during SLO and in adaptive communica-
tion on the Vineland-II. In contrast, supplemental
measures of language offered mixed results. While
the CDI mean length of longest utterance
approached significance, no differences were
High visual reception observed between groups on CDI number of words
produced, or PLS expressive standard scores. These
findings are consistent with several other studies
which have shown that parent education programs
may not lead to reliable changes on standardized
assessments of language functioning (Smith, Groen,
& Wynn, 2000), though improvements may be
observed on other language measures (Roberts
et al., 2011). These observations also suggest that
as a parent-delivered intervention, PRTG may have a
greater effect on functional use of skills during
everyday interactions. This finding is consistent with
the CGI results showing greater improvement in
PRTG and greater reduction in severity of social
communication impairment when compared with
PEG. Future research including multiple measures
of skill generalization will be needed to investigate
the extent to which child improvement at least in the
short-term reflects the parent’s improved ability to
elicit functional language and whether this potential
mechanism of change has implications for subse-
quent generalization in child skill across interactive
partners and contexts. Longer duration trials may
also be warranted to evaluate changes on standard-
ized language measures.
Observed gains in total frequency of utterances
came primarily from increases in intelligible speech.
The greatest increases in PRTG were in imitative
utterances, which are consistent with findings from
other short-term NDBIs (Ingersoll, Meyer, Bonter, &
Jelinek, 2012). This is likely a result of the emphasis
in PRT training on the use of maintenance tasks to
enhance motivation (Koegel et al., 1989). Nonverbally
prompted speech was also shown to increase as a
result of PRTG, suggesting parents may have begun
incorporating nonverbal prompts as acquisition
tasks. Parents in the PEG did use verbal prompts
during the SLO at a frequency similar to that observed
in the PRTG, but did not naturally use frequent
imitative or nonverbal prompting strategies to scaffold
child language development. Spontaneous utter-
ances were generally low in frequency, though parents
were instructed to actively engage their children,
potentially limiting the opportunity for unprompted
verbalizations. Future studies should consider home
videos or automated measures (e.g. Language ENvi-
ronment Analysis system, LENA; Oller et al., 2010) for
assessing spontaneous speech.
Documented improvements in adaptive communi-
cation for the PRTG were consistent with prior
© 2014 Association for Child and Adolescent Mental Health.

research (Baker-Ericz
en et al., 2007). Increases in
standard scores suggest a rate of improvement over
12 weeks that is faster than normative expectations
and comparable to other controlled trials of early
intervention programs (Dawson et al., 2010).
Improvement on the Vineland-II is particularly
meaningful, as adaptive skills are related to positive
prognosis (Kanne et al., 2011). Increases in receptive
raw scores suggest spread of effect to receptive skills
not explicitly taught (Walton & Ingersoll, 2013). This
pattern appears to indicate that participating chil-
dren increased their understanding of speech as a
result of the intervention. It may also reflect the fact
that parents became more effective at motivating
their children to respond. This finding is consistent
with research in children with developmental dis-
abilities, where progress in expressive language was
associated with improvement in receptive skills
(Walton & Ingersoll, 2013; Wynn & Smith, 2003).
Baseline Mullen VR scores predicted language
gains in both groups, while gender, age, and other
baseline measures did not. Children with high VR
scores receiving PRTG had very large improvements in
total and imitative utterances, while those with low VR
scores had smaller but still meaningful improve-
ments. Children with high VR scores receiving PEG
showed modest improvements at week 12, while those
with low VR scores did not show improvement. These
observations are consistent with reports documenting
positive outcomes associated with higher baseline
cognitive ability and nonverbal IQ in particular (Mag-
iati, Tay, & Howlin, 2014; Turner & Stone, 2007) and
highlight the need to individualize interventions so
that effective treatments are implemented early, when
the brain is most plastic.
Consistent with recommendations regarding the
critical role of parent education in early intervention
(National Research Council, 2001), this study pro-
vided support for the model of training parents to
implement PRT. This mechanism for increasing inter-
vention intensity is promising, particularly given the
need for cost-effective services (Wise, Little, Holliman,
Wise, & Wang, 2010). A recent systematic review of
meta-analyses indicated that parent involvement was
an important predictor of better child outcomes,
especially adaptive behavior gains (Strauss, Mancini,
& Fava, 2013). Given that parent education models
offer opportunity for enhanced generalization (Steiner
et al., 2012), increased service intensity (Kaiser, Han-
cock, & Nietfeld, 2010; Laski, Charlop, & Schreibman,
1988), and improved parent psychological well-being
(Minjarez et al., 2013), additional studies examining
effective parent educational models are needed.
Delivery of parent training in a group format
makes this study unique, as there have been no
randomized controlled trials of group PRT to date.
With limited access to evidence-based treatments for
ASD (Wise et al., 2010), it is especially meaningful to
identify efficacious group models, which allow for
dissemination to large numbers of children with
© 2014 Association for Child and Adolescent Mental Health.
RCT of pivotal response treatment group 7
time-limited clinician involvement. It is also impor-
tant to determine whether group models are more
effective than individual approaches in subgroups of
children or families with specific characteristics.
This study had several limitations. The moderate
sample size meant that power to detect complex
multicomponent patterns was limited. There is also
no information available about longer term effects of
the treatment. Furthermore, children may have got-
ten varying doses of PRT depending on the consis-
tency of parent implementation and parents who did
not meet fidelity may have learned PRT if given more
time. For ethical reasons, it was necessary to allow
children to continue community treatments during
study. There were also a few differences between the
PRTG and PEG including number of individual ses-
sions, qualifications of intervention providers, and
use of homework assignments in the PRTG which
could have influenced results. There were no data
regarding fidelity of treatment implementation by the
group parent educators which would be important for
future studies. This study also targeted language
development only, though there is support for using
PRT to target other skills (Koegel, Koegel, Boettcher,
Harrower, & Openden, 2006; Schreibman, Stahmer,
& Pierce, 1996). Finally, given that parent-report
measures (e.g. Vineland-II) cannot be ‘blind’ to treat-
ment condition in a parent education study, more
objective behavioral or automated measures such as
LENA System (Oller et al., 2010) and measures of
in-home implementation would add important infor-
mation about generalization of skills.
Conclusions
The present study represents a step forward in
applying rigorous experimental design to evaluate
differential treatment effects by comparing an estab-
lished treatment with general parent psychoeduca-
tion. The study also provides support for
group-based parent training, offering a promising
new approach for dissemination to large numbers of
families in an efficient manner. However, additional
large randomized controlled trials will be critical for
determining the optimal intensity of treatment (Rog-
ers & Vismara, 2008). Supplementing parent train-
ing with clinician-delivered intervention and
increasing treatment duration may lead to larger
gains. Comparing established treatments will also be
critical for analysis of predictors of treatment
response. Finally, while there is evidence from this
and other studies to suggest that PRT may have an
impact on social skills (Baker-Ericz
en et al., 2007),
there is a need for more systematic research into the
effects of PRT on social behaviors.
Supporting information
Additional Supporting Information may be found in the
online version of this article:
8 Antonio Y. Hardan et al.
Appendix S1. Pivotal Response Treatment Group Con-
tent.
Table S1. Overview of pivotal response treatment group
content.
Table S2. Overview of psychoeducation group content.
Figure S1. Total child utterances (+/– 95% CI) across
study weeks.
Figure S2. Consort 2010 checklist of information to
include when reporting a randomized trial.
Acknowledgements
This work was supported by an Autism Speaks Treat-
ment Grant (#5773; PI: Hardan); the foundation was
Key points
• Pivotal Response Treatment (PRT) is a naturalistic developmental behavior intervention (NDBI) focused on
motivating children with autism.
• This is the first randomized controlled trial of group parent training in PRT (PRTG).
• PRTG resulted in greater improvement in functional speech and adaptive communication, when compared
with a psychoeducation group.
• Additional large-scale studies are needed to identify optimal intensity and explore effects of PRT on other core
symptoms.
References
American Psychiatric Association (2013). Diagnostic and
statistical manual for mental disorders (5th edn).
Washington, DC: American Psychiatric Association.
Baker-Ericz
en, M.J., Stahmer, A.C., & Burns, A. (2007). Child
demographics associated with outcomes in a community-
based pivotal response training program. Journal of Positive
Behavior Interventions, 9, 52–60.
Bryson, S.E., Koegel, L.K., Koegel, R.L., Openden, D., Smith, I.M.,
& Nefdt, N. (2007). Large scale dissemination and community
implementation of pivotal response treatment: Program
description and preliminary data. Research and Practice for
Persons With Severe Disabilities, 32, 142–153.
Center for Disease Control (2014). Prevalence of autism
spectrum disorder among children aged 8 years-autism
and developmental disabilities monitoring network, 11
sites, United States, 2010. MMWR Surveillance Summaries,
63, 1–21.
Cohen, J. (1960). A coefficient of agreement for nominal scales.
Educational and Psychological Measurement, 20, 37–46.
Constantino, J.N., & Gruber, C.P. (2005). Social respon-
siveness scale. Los Angeles: Western Psychological Services.
Dawson, G., & Burner, K. (2011). Behavioral interventions in
children and adolescents with autism spectrum disorder: A
review of recent findings. Current Opinion in Pediatrics, 23,
616–620.
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, 17–23.
Fenson, L., Marchman, V.A., Thal, D.J., Dale, P.S., Reznick,
J.S., & Bates, E. (2007). MacArthur-bates communicative
development inventories. Baltimore, MD: Paul H. Brookes
Publishing Co.
Guy, W. (1976). ECDEU assessment manual for psycho-
pharmacology. Rockville, MD: U.S. Department of Health,
Education, and Welfare.
not involved in the trial. Data management was sup-
ported by the National Center for Research Resources
and the National Center for Advancing Translational
Sciences, National Institutes of Health, through grant
UL1 RR025744. The authors gratefully acknowledge
the families for their participation.
The authors have declared that they have no com-
peting or potential conflicts of interest.
Correspondence
Grace Gengoux, Department of Psychiatry and Behavioral
Sciences, Stanford University, 401 Quarry Road, Stanford,
CA 94305, USA; Email: ggengoux@stanford.edu
Harris, P.A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., &
Conde, J.G. (2009). Research electronic data capture
(REDCap) – A metadata-driven methodology and workflow
process for providing translational research informatics
support. Journal of Biomedical Informatics, 42, 377–381.
Ingersoll, B., Meyer, K., Bonter, N., & Jelinek, S. (2012).
A comparison of developmental social–pragmatic and
naturalistic behavioral interventions on language use
and social engagement in children with autism. Journal
of Speech, Language, and Hearing Research, 55, 1301–
1313.
Kaiser, A.P., Hancock, T.B., & Nietfeld, J.P. (2010). The effects
of parent-implemented enhanced milieu teaching on the
social communication of children who have autism. Early
Education and Development, 11, 423–446.
Kanne, S.M., Gerber, A.J., Quirmbach, L.M., Sparrow, S.S.,
Cicchetti, D.V., & Saulnier, C.A. (2011). The role of adaptive
behavior in autism spectrum disorders: Implications for
functional outcome. Journal of Autism and Developmental
Disorders, 41, 1007–1018.
Koegel, R.L., Bimbela, A., & Schreibman, L. (1996). Collateral
effects of parent training on family interactions. Journal of
Autism and Developmental Disorders, 26, 347–359.
Koegel, R.L., Koegel, L.K., Boettcher, M.A., Harrower, J., &
Openden, D. (2006). Combining functional assessment and
self-mangagement procedures to rapidly reduce disruptive
behaviors. In R.L. Koegel & L.K. Koegel (Eds.), Pivotal
response treatments for autism (pp. 245–258). Baltimore,
MD: Paul H. Brookes Publishing Company.
Koegel, L.K., Koegel, R.L., & Brookman, L.I. (2005).
Child-initiated interactions that are pivotal in intervention
for children with autism. In E.D. Hibbs & P.S. Jensen (Eds.),
Psychosocial treatments for child and adolescent disorders:
Empirically based strategies for clinical practice (2nd edn,
pp. 633–657). Washington DC: American Psychological
Association.
Koegel, L.K., Koegel, R.L., Harrower, J.K., & Carter, C.M.
(1999). Pivotal response intervention I: Overview of
© 2014 Association for Child and Adolescent Mental Health.

approach. Research and Practice for Persons with Severe
Disabilities, 24, 174–185.
Koegel, R.L., Schreibman, L., Good, A., Cerniglia, L., Murphy,
C., & Koegel, L.K. (1989). How to teach pivotal behaviors to
children with autism: A training manual. Santa Barbara, CA:
University of California.
Laski, K.E., Charlop, M.H., & Schreibman, L. (1988). Training
parents to use the natural language paradigm to increase
their autistic children’s speech. Journal of Applied Behavior
Analysis, 2, 391–400.
Lord, C., Risi, S., Lambrecht, L., Cook, E.H., Jr, Leventhal,
B.L., DiLavore, P.C., . . . & Rutter, M. (2000). The autism
diagnostic observation schedule-generic: A standard
measure of social and communication deficits associated
with the spectrum of autism. Journal of Autism and
Developmental Disorders, 30, 205–223.
Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism
diagnostic interview-revised: A revised version of a
diagnostic interview for caregivers of individuals with
possible pervasive developmental disorders. Journal of
Autism and Developmental Disorders, 24, 659–685.
Lovaas, O.I. (1987). Behavioral treatment and normal educational
and intellectual functioning in young autistic children. Journal
of Consulting and Clinical Psychology, 55, 3–9.
Magiati, I., Tay, X.W., & Howlin, P. (2014). Cognitive, language,
social and behavioural outcomes in adults with autism
spectrum disorders: A systematic review of longitudinal
follow-up studies in adulthood. Clinical Psychology Reviews,
34, 73–86.
Minjarez, M.B., Mercier, E.M., Williams, S.E., & Hardan, A.Y.
(2013). Impact of pivotal response training group therapy on
stress and empowerment in parents of children with autism.
Journal of Positive Behavior Interventions, 15, 71–78.
Minjarez, M.B., Williams, S.E., Mercier, E.M., & Hardan, A.Y.
(2011). Pivotal response group treatment program for
parents of children with autism. Journal of Autism and
Developmental Disorders, 41, 92–101.
Mohammadzaheri, F., Koegel, L.K., Rezaee, M., & Rafiee, S.M.
(2014). A randomized clinical trial comparison between
Pivotal Response Treatment (PRT) and structured Applied
Behavior Analysis (ABA) intervention for children with
autism. Journal of Autism and Developmental Disorders,
44, 2769–2777.
Mullen, E.M. (1995). Mullen scales of early learning: AGS edn.
Circle Pines, MN: American Guidance Service.
National Autism Center (2009). National standards report.
Randolph, MA: National Autism center.
National Research Council (2001). Educating children with
autism. Washington, DC: National Academies Press.
Oller, D., Niyogi, P., Gray, S., Richards, J., Gilkerson, J., Xu,
D., . . . & Warren, S. (2010). Automated vocal analysis of
naturalistic recordings from children with autism, language
delay, and typical development. Proceedings of the National
Academy of Sciences, 107, 13354–13359.
Peugh, J.L., & Enders, C.K. (2005). Using the SPSS mixed
procedure to fit cross-sectional and longitudinal multilevel
models. Educational and Psychological Measurement, 65,
717–741.
Reichow, B. (2012). Overview of meta-analyses on early
intensive behavioral intervention for young children with
autism spectrum disorders. Journal of Autism and
Developmental Disorders, 42, 512–520.
Roberts, J., Williams, K., Carter, M., Evans, D., Parmenter, T.,
Silove, N., . . . & Warren, A. (2011). A randomized controlled
trial of two early intervention programs for young children
with autism: Centre-based with parent program and
© 2014 Association for Child and Adolescent Mental Health.
RCT of pivotal response treatment group 9
home-based. Research in Autism Spectrum Disorders, 5,
1553–1566.
Rogers, S.J., & Vismara, L.A. (2008). Evidence-based
comprehensive treatments for early autism. Journal of
Clinical Child & Adolescent Psychology, 37, 8–38.
Rogers, S.J., Vismara, L., Wagner, A.L., McCornick, C., Young,
G., & Ozonoff, S. (2014). Autism treatment in the first year of
life: A pilot study of Infant Start, a parent-implemented
intervention for symptomatic infants. Journal of Autism and
Developmental Disorders. Advance online publication.
doi:10.1007/s10803-014-2202-y.
Schreibman, L., Dawson, G., Rogers, S.J., Ingersoll, B., Kaiser,
A.P., & Kasari, C., . . . & Halladay, A. (2014). Common
elements of research-based practices for toddlers with ASD.
Paper presented at the Autism Speaks Toddler Treatment
Network Meeting, Atlanta, GA.
Schreibman, L., Stahmer, A., & Pierce, K. (1996). Alternative
applications of pivotal response training: Teaching symbolic
play and social interaction skills. In L.K. Koegel, R.L. Koegel
& G. Dunlap (Eds.), Positive behavioral support: Including
people with difficult behavior in the community (pp. 353–
371). Baltimore, MD: Paul H. Brookes Publishing Co.
Shrout, P.E., & Fleiss, J.L. (1979). Intraclass correlations:
Uses in assessing rater reliability. Psychological Bulletin, 86,
420–428.
Smith, T., Groen, A.D., & Wynn, J.W. (2000). Randomized trial
of intensive early intervention for children with pervasive
developmental disorder. American Journal of Mental
Retardation, 105, 269–285.
Sparrow, S.S., Cicchetti, D.V., & Balla, D.A. (2005). Vineland
adaptive behavior scales (2nd edn). San Antonio, TX:
Pearson.
Steiner, A.M., Koegel, L.K., Koegel, R.L., & Whitney, A.E.
(2012). Issues and theoretical constructs regarding parent
education for autism spectrum disorders. Journal of Autism
and Developmental Disorders, 42, 1218–1227.
Strauss, K., Mancini, F., & Fava, L. (2013). Parent inclusion in
early intensive behavior interventions for young children
with ASD: A synthesis of meta-analyses from 2009 to 2011.
Research in Developmental Disabilities, 34, 2967–2985.
Symon, J.B. (2005). Expanding interventions for children with
autism. Journal of Positive Behavior Interventions, 7, 159–173.
Turner, L.M., & Stone, W.L. (2007). Variability in outcome for
children with an ASD diagnosis at age two. Journal of Child
Psychology and Psychiatry, 48, 793–802.
Vismara, L.A., & Rogers, S.J. (2010). Behavioral treatments in
autism spectrum disorder: What do we know? Annual
Review of Clinical Psychology, 6, 447–468.
Walton, K.M., & Ingersoll, B.R. (2013). Improving social skills
in adolescents and adults with autism and severe to
profound intellectual disability: A review of the literature.
Journal of Autism and Developmental Disorders, 43, 594–
615.
Wise, M.D., Little, A.A., Holliman, J.B., Wise, P.H., & Wang, J.
(2010). Can state early intervention programs meet the
increased demand of children suspected of having autism
spectrum disorders? Journal of Developmental and
Behavioral Pediatrics, 31, 469–476.
Wynn, J.W. & Smith, T. (2003). Generalization between
receptive and expressive language in young children with
autism. Behavioral Interventions, 18, 245–266.
Zimmerman, I.L., Steiner, V.G., & Pond, R.E. (2002). Preschool
language scale (4th edn). San Antonio, TX: The
Psychological Corporation.
Accepted for publication: 20 September 2014