An exploratory study on the effectiveness of Pivotal Response Treatment and a robot-based Pivotal Response Treatment on social and communicative skills in children with autism spectrum disorder

Children with autism spectrum disorder (ASD) are characterized by persistent
deficits in social communication and social interaction. Despite the severity
and chronic course of the disorder, the heavy burden on individuals and
families, the high prevalence rate and the high societal costs, no effective
pharmacologic treatments are available for targeting the core symptoms of ASD.
As a behavioural intervention, Pivotal Response Treatment (PRT) has been proven
effective in improving different social and communicative skills in children
with ASD. Additionally, studies focusing on implementing robotics in the
treatment of children with ASD show promising results. Although promising,
earlier studies to the effectiveness of PRT and implementing robots in the
treatment of children with ASD are limited by methodological problems. The
current study will address these problems by conducting a randomized clinical
trial to the effectiveness of PRT with and without the implementation of a
humanoid robot.

The primary objective of the current study is assessing the effectiveness of
robot-based PRT on top of care-as-usual compared to care-as-usual only.
The primary hypothesis is:
1) Robot-based PRT is more effective compared to care-as-usual only in:
-promoting social and communicative skills in children with ASD (within the
natural environment of the child)
-providing a clinically significant improvement on mental health
-decreasing severity of ASD symptoms

Secondary objectives/hypotheses:
2) PRT provided by a human trainer is more effective compared to care-as-usual
in:
-promoting social and communicative skills in children with ASD (within the
natural environment of the child)
-providing a clinically significant improvement on mental health
-decreasing severity of ASD symptoms

3) Robot-based PRT and PRT provided by a human trainer are effective in:
-promoting social and communicative skills in children with ASD (within the
natural environment of the child)
-providing a clinically significant improvement on mental health
-decreasing severity of ASD symptoms

Besides measuring generalization of social and communicative skills into the
natural environment of the child, the interest is in assessing the improvement
in skills during the treatment and assessing the likability of the robot by the
children with ASD and the affect of the child within the robot-based PRT. This
provides more information on the usefulness of implementing a robot in the
treatment of children with ASD.
Hypotheses:
4) Robot-based PRT is more effective compared to PRT provided by a human
trainer in:
-lowering the prompt level (i.e. help that is needed) for communicative skills
in children with ASD during treatment
-heightening the number of learning moments the child shows spontaneous
appropriate behaviour (e.g. initiations) on during the treatment
5) The robot that is used in the robot-based PRT shows a high likability by
children with ASD
6) Children with ASD show positive affect during the robot-based PRT sessions

Also, child rearing pressure on parents is administered, because of the highly
parent-focused PRT.
7) Robot-based PRT and PRT provided by a human trainer are both effective in
decreasing the child rearing pressure experienced by parents.

Additionally, interest is in the relation between questionnaire data about
social behaviour and qualitative reports on robot experiences on the one hand
and possible physical makers of social behaviour and stress on the other.
8) Qualitative affect scores during PRT are related to salivary cortisol levels
9) Social and communicative skills in children with ASD are related to salivary
oxytocin and testosterone levels

The study is designed as an exploratory cluster-randomized (phase IIa - like)
open three-group parallel clinical trial. Subjects are cluster-randomly
assigned to three intervention conditions:
1) Robot-based PRT: Therapy is provided based on Pivotal Response Treatment
with the implementation of a humanoid robot on top of care-as-usual, i.e.
psycho-education and medical management (n = 25)
2) PRT: Pivotal Response Treatment is administered by a human trainer on top of
care-as-usual, i.e. psycho-education and medical management (n = 25)
3) Care-as-usual: includes guidance of parents or intensive psychiatric family
treatment, besides psycho-education of ASD and medical management (n = 25)

The randomization occurs per cluster. For 4 locations of Karakter,
randomization occurs between robot-based PRT and care-as-usual and for 3
locations of Karakter, randomization occurs between PRT and care-as-usual.

The study consists of a selection phase, randomization phase, baseline
measures, a 5-month intervention phase including evaluation measures and
follow-up measures (3 months after the end of the intervention).

The study consists of 3 intervention conditions, to which participants are
cluster-randomly assigned:

1) Robot-based PRT, on top of care-as-usual (robot therapy condition): includes
20 sessions, once a week of 45 minutes. During the sessions, learning moments
are created that provide the child the opportunity to e.g. learning to initiate
within a social conversation. The content of the sessions is focused on
different social and communicative skills and depends highly on the learning
goals, that are determined on forehand. Within all parent-child sessions (14),
a humanoid robot called NAO from Aldebaran robotics is used. The robot can
speak and move, facilitating social interaction with the child. A trained PRT
therapist is present during all robot-based PRT sessions. In addition to the
robot-based PRT, care as usual (psycho-education and medical management if
applicable) is provided.

2) PRT by a human trainer, on top of care-as-usual (PRT condition): includes 20
sessions, once a week of 45 minutes. PRT is provided by a human PRT-therapist,
as is the regular procedure within Karakter. As in the robot-based PRT, the
focus is on improving different social and communicative skills, depending on
the learning goals of each child. The main focus is on teaching parents to
implement PRT principles in social situations with their child, based on the
strengths of parents.

3) Care-as-usual: includes the regular treatment for ASD that is provided in
outpatient departments of Karakter. The care-as-usual condition includes
psycho-education of ASD and medical management, supplemented by guidance of
parents and other primary caregivers or intensive psychiatric family treatment.

The risks of participating in the study are estimated to be very low. No
adverse effects have been described in earlier studies that implemented a
humanoid NAO robot within treatment of children with ASD. Patients assigned to
the three conditions are all expected to benefit from their treatment.
Additional time investement for participation in this therapeutic study is low
compared to regular procedures for treatment of ASD and measuring treatment
effects within Karakter. The robot-basedl PRT and PRT are provided in addition
to the basic care as usual for ASD within the outpatient department of
Karakter. Changes in dosages of medication are not allowed during the study.
When changes in the child's behaviour are noticed, this is reported and the
responsible child psychiatrist is contacted to assure optimal care for each
child. The benefits of participating in the study are expected to highly
outweight the possible disadvantages.