Abstract:
Virtual humans are a form of artificial agent that is starting to be used to deliver
e-health
interventions. Several early studies show promising feasibility, but research designs have been
preliminary to date. Therefore, more randomised controlled trials (RCTs) are needed to
evaluate their effectiveness in delivering psychological interventions and improving
health
outcomes.
The aim of this study was to evaluate the effectiveness of a brief relaxation
intervention
delivered by a virtual human recording compared with a real human audiotape and a control group on
wound healing and stress in a healthy community sample.
An RCT with a mixed factorial design was conducted. One hundred and fifty-nine healthy adults
aged between 18 and 60 years (M =26.04) participated in a 90-minute session, during which they
completed questionnaires and underwent a non-invasive tape-stripping procedure to disrupt the
skin barrier, followed by a 20-minute recovery period. Participants were randomised to (1)
relaxation delivered by a Digital Human, (2) relaxation delivered by human audiotape (3) reading
magazines during the recovery period. A series of psychological questionnaires collected
self-reported stress, relaxation, anxiety and pain, satisfaction, engagement, and
open-ended responses; a wrist sensor wrist measured heart rate (HR) and electrodermal activity
(EDA) and saliva samples collected stress hormones salivary cortisol (sCort) and alpha-amylase
(sAA).
Results indicated no significant differences in skin barrier recovery (SBR), psychological or
physiological stress measures between the delivery methods. However, the Digital Human and
audiotape groups had significantly greater relaxation following the relaxation session, superior to
the control group. Anxiety was significantly more reduced, and satisfaction and engagement levels
were greater in the audiotape group. Additional open-ended responses indicated that the appearance
and mismatch in lip-syncing of the Digital Human made some people feel uneasy and could be
improved.
The findings demonstrate a virtual human’s ability to relax people, comparative to traditional
audiotapes in an RCT. The study provides suggestions to improve virtual humans’ delivery of
relaxation interventions and provides future experimental design recommendations.
Ultimately, these findings encourage future studies to explore further the effects of
an interactive virtual human to deliver relaxation and improve health outcomes, particularly in
highly stressed or clinical wound samples.