My Hospital Buddy Ida: AR-CBT for Children With Cancer

Background and rationale. Childhood cancer is among the leading causes of disease burden in children worldwide, with more than 400,000 children and adolescents estimated to develop cancer each year. Although survival has improved markedly in higher-resource settings, the diagnosis and subsequent treatment impose a profound and often traumatic psychological burden. Repeated hospitalizations, an unfamiliar and frequently frightening medical environment, and disruption of everyday routines contribute to clinically meaningful anxiety, low mood, and substantial impairment in health-related quality of life that can persist into survivorship. Left unaddressed, these difficulties are associated with poorer treatment adherence, prolonged adjustment problems, and elevated rates of post-traumatic stress and internalizing symptoms extending well beyond the acute treatment period.

Two features of pediatric cancer care are particularly distressing for children. First, children repeatedly undergo invasive procedures, including venipuncture, central line or port access, and painful diagnostic evaluations, that provoke fear, pain, and anticipatory distress with each subsequent encounter. Second, prolonged chemotherapy produces aversive symptoms such as nausea, vomiting, and fatigue that frequently become conditioned and anticipatory over the course of treatment, further eroding wellbeing and treatment tolerance. Psychosocial standards of care in pediatric oncology designate evidence-based psychological support as an essential, rather than optional, component of comprehensive cancer care, and cognitive behavioral therapy (CBT) has the strongest evidence base among psychosocial interventions for reducing distress and improving adjustment in children facing serious illness.

In practice, access to trained pediatric mental health specialists is severely constrained, particularly in low- and middle-resource settings, and conventional face-to-face therapy places additional burden on children who are already fatigued and immunosuppressed. Digital delivery offers one route to scalability, and app-based interventions have been shown to reduce anxiety and depressive symptoms in youth more broadly. Immersive technologies are especially promising in this context: virtual reality (VR) reliably reduces procedural pain and anxiety in pediatric patients, including children with cancer. Augmented reality (AR), which overlays interactive digital content onto the real environment rather than replacing it, has been used for pediatric health education and, increasingly, for mental health applications, but had not previously been combined with a structured, manualized psychotherapeutic framework in pediatric oncology. Prior digital and play-based interventions in this population have been largely limited to distraction or education without an explicit therapeutic model, and resilience-focused programs have most often targeted parents, adolescents, or young adults rather than school-aged children (6 to 12 years) -- a group old enough to experience and voice fear, yet too young for most adolescent- or parent-directed programs, and who typically benefit from concrete, play-based, developmentally tailored delivery rather than verbal therapy alone.

To address this gap, the investigators developed "Hospital Buddy Ida" ("Hastane Arkadaşım İda"), an augmented reality-supported, manualized CBT program in which a three-dimensional virtual companion guides children through psychoeducation, emotion-regulation, and coping modules integrated with Child Life principles. This trial was designed to evaluate the efficacy, feasibility, and acceptability of this AR-supported CBT program (AR-CBT) relative to an identical CBT program delivered without the AR application (standard CBT), and relative to usual oncologic care, so as to isolate the incremental contribution of the augmented reality delivery layer from the therapeutic content itself.

Development context and reporting. This trial was conducted as the efficacy-testing phase of a multi-phase intervention development program that progressed sequentially from user-experience research and application prototyping, through pilot feasibility testing, to the present three-arm randomized controlled trial. Reporting follows the CONSORT statement and its extension for e-health interventions (CONSORT-EHEALTH), and participant sex and sociodemographic characteristics are reported in accordance with the Sex and Gender Equity in Research (SAGER) guidelines. The trial was conducted across multiple pediatric oncology centers spanning university, training-and-research, and private hospital settings in Istanbul and Antalya, Türkiye. Written informed consent was obtained from caregivers and assent was obtained from children; study data were anonymized by protocol number.

Intervention content and delivery. The AR-CBT and standard-CBT programs share identical manualized content: a twelve-module curriculum (three psychoeducational modules, five psychotherapeutic modules, and four relaxation-based modules) developed by certified CBT specialists in collaboration with Child Life methodology, and consolidated into nine clinician-delivered sessions to reduce burden on medically fragile participants. Session content progresses from rapport-building and orientation to the body and hospital environment, through emotion regulation and cognitive restructuring, to problem-solving, and incorporates diaphragmatic breathing, progressive muscle relaxation, and guided imagery, supported by between-session homework and a closing "graduation" session. Children randomized to AR-CBT additionally used the Hospital Buddy Ida application, in which the same three-dimensional virtual companion overlays the CBT and Child Life module content onto the child's real surroundings via augmented reality; children randomized to standard CBT received identical content through a printed workbook without the AR application. Both active arms were delivered at the bedside by clinical psychologists and were matched on number of sessions and scheduled session duration, so that the augmented reality layer constituted the only systematic difference between the two active treatment arms. Treatment fidelity was supported by a standardized treatment manual and session-by-session fidelity checklists.

Randomization, allocation concealment, and blinding. Eligible children were randomized using a computer-generated dynamic (minimization) allocation procedure balanced across participating sites, with allocation concealed from enrolling staff through a central electronic randomization system and implemented only after completion of baseline assessment. Because the active interventions involved direct interaction with an application or workbook, children, caregivers, and treating clinicians could not be masked to group assignment; however, all outcome assessments were administered by trained researchers who did not deliver the intervention and remained blinded to group allocation throughout, and statistical analyses were likewise conducted blind to group labels.

Sample size and analytic approach. An a priori power analysis (three-group comparison, power = 0.80, two-sided alpha = 0.05) indicated that approximately 129 participants would be required to detect a medium effect size on the primary outcome; allowing for anticipated attrition, the trial targeted a larger enrollment. The primary efficacy analysis was pre-specified as an intention-to-treat linear mixed-effects model (with group, time, and group-by-time interaction as fixed effects, participant as a random effect, and baseline score and child age as covariates), which accommodates all available data under a missing-at-random assumption. A completer-based analysis of covariance adjusted for baseline score and age was pre-specified as a sensitivity analysis, together with an additional sensitivity analysis stratified by age band to address possible baseline imbalance. Because a large number of secondary and subscale outcomes were examined, these were treated as exploratory a priori, and their significance levels were adjusted for multiplicity within outcome families using the Benjamini-Hochberg false-discovery-rate procedure.

Rationale for outcome domains. Outcome domains were selected to capture the constructs most directly threatened by cancer treatment in this age group: overall health-related quality of life as the pre-specified primary construct, alongside secondary domains of psychological resilience, caregiver-observed behavioral and emotional problems, treatment-related anxiety, and the conditioned, anticipatory symptoms of fatigue, nausea, and distress that are characteristic of prolonged chemotherapy. Perceived knowledge about the hospital environment, staff, illness, and procedures was assessed as a secondary domain because understanding of what will happen is an established route to reduced procedural fear and improved cooperation. Acceptability was evaluated from the perspective of children, caregivers, and nursing staff, given that real-world scalability depends on acceptability across all stakeholders, not efficacy alone. In a subset of participants, physiological indices (heart rate and electrodermal activity) were recorded as exploratory measures to provide a preliminary, objective complement to self- and parent-reported outcomes.

Innovation and significance. This trial provides the first randomized evaluation of a complete, manualized cognitive behavioral therapy protocol embedded within an augmented reality companion for children with cancer. Its principal methodological innovation is a three-arm design deliberately configured to disentangle two elements that digital mental health studies typically confound: the therapeutic content and the technology used to deliver it. By holding the CBT and Child Life curriculum, the number of sessions, and scheduled clinician contact time constant across the two active arms, the trial isolates the incremental contribution of the augmented reality layer itself. This directly addresses a question that has remained unresolved across the broader immersive-technology literature, namely whether such technology confers therapeutic benefit beyond engagement, novelty, or short-lived procedural distraction.

The clinical and public health significance of the study is anchored in a specific, well-documented gap in care. School-aged children (6 to 12 years) undergoing cancer treatment are underserved by existing psychosocial programs, which have predominantly targeted parents, adolescents, or young adults, and access to specialist child mental health support is scarcest in precisely the low- and middle-income settings that bear the largest share of the global childhood-cancer burden. Because the therapeutic content of the program is standardized and delivered in part through the application, an effective augmented reality-supported CBT program offers a concrete mechanism for narrowing this gap: it could extend guideline-concordant psychosocial care to resource-limited settings without a proportional increase in specialist clinician time, reduce site-to-site variation in the quality of psychosocial care, and support supervised delivery by trained non-specialist staff through a consistent, reproducible protocol.

More broadly, demonstrating that a structured, multi-session psychotherapy can be delivered through an augmented reality companion, rather than using the technology solely for procedural distraction, would help reframe immersive technology as a credible delivery vehicle for evidence-based mental health care. The implications extend beyond pediatric oncology to other pediatric populations that face repeated, distressing medical procedures and would benefit from engaging, developmentally appropriate, and scalable psychological support. The findings of this trial are intended to establish the feasibility, acceptability, and preliminary efficacy required to justify larger, prospectively registered, multinational confirmatory trials incorporating longer-term follow-up, masked clinician-rated and objective outcomes, and evaluation of downstream endpoints such as treatment adherence and healthcare utilization, as well as the development and testing of a cross-platform version of the intervention to maximize real-world reach.

Inclusion criteria

• Diagnosis of any childhood cancer
• Currently within the first 4 months of chemotherapy treatment
• Has completed at least 1 prior chemotherapy cycle
• Sufficient Turkish language ability to complete study measures
• Presence of a literate caregiver able to provide consent and assist with study procedures

Exclusion criteria

• Chronic illness other than cancer
• Diagnosed brain disease
• Epilepsy
• Severe neurodevelopmental or neurological disorder
• Clouded consciousness or delirium
• History of motion sickness
• Sensory or cognitive impairment precluding use of the AR application