Feasibility randomised controlled trial of the Online Support and Intervention for Child Anxiety – Autism Version (OSI-A) within public health clinical services in Chile

Last registered on June 18, 2026

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0018901

Initial registration date

June 11, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

June 18, 2026, 9:13 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Chile

Region

Santiago

Primary Investigator

Name

Miguel Cordero

Affiliation

Universidad del Desarrollo

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Claudia Martínez

PI Affiliation

Pontificia Universidad Catolica de Chile

Contact Investigator

Additional Trial Information

Status

On going

Start date

2026-06-01

End date

2027-11-30

Keywords

Behavior, Electoral, Environment & Energy, Health

Additional Keywords

Autism, Anxiety, CBT, Online support, Parent led therapy

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Autistic children often experience heightened anxiety due to sensory sensitivities, social challenges, and difficulties with change. While Cognitive Behavioural Therapy (CBT) is an evidence-based treatment for anxiety, access remains limited—particularly in low-resource settings—due to shortages of trained therapists. Parent-led CBT, in which therapists guide parents to support their children, offers an effective and resource-efficient alternative. The Online Support and Intervention for Child Anxiety (OSI) programme was developed to improve access to CBT through structured and interactive digital content, with parents leading the implementation of therapeutic strategies.
OSI has been successfully evaluated in schools and clinical services in the United Kingdom and has demonstrated outcomes comparable to more intensive therapist-led approaches. An adapted version for autistic children (OSI-A) was subsequently developed to support parents in implementing graded exposure, developing step plans, and promoting independence. In the Chilean adaptation, OSI-A is further augmented by an artificial intelligence (AI) chatbot, supervised by therapists, which provides reminders, structured guidance, progress monitoring, and support between therapist contacts.
The primary objective of this study is to assess the feasibility and acceptability of implementing OSI-A within clinical health settings in Chile. Specifically, the study will evaluate recruitment rates, intervention adherence, participant retention, data completeness, and intervention acceptability, while also collecting preliminary clinical and health economic outcome data to inform a future definitive trial.
By evaluating the feasibility of this digitally augmented, parent-led CBT programme for autistic children, the study will generate evidence on implementation requirements, trial procedures, and the potential role of conversational AI in supporting low-intensity mental health interventions. Findings will inform the design of a future fully powered effectiveness and cost-effectiveness trial and contribute to the development of scalable and accessible mental health services for autistic children and their families.

External Link(s)

Registration Citation

Citation

Cordero, Miguel and Claudia Martínez. 2026. "Feasibility randomised controlled trial of the Online Support and Intervention for Child Anxiety – Autism Version (OSI-A) within public health clinical services in Chile." AEA RCT Registry. June 18. https://doi.org/10.1257/rct.18901-1.0

Sponsors & Partners

Partner

Name

Inter-American Development Bank

Type

none

URL

https://www.iadb.org/en

Experimental Details

Interventions

Intervention(s)

The study includes three parallel arms with a total of 180 caregiver–child dyads (60 per arm).
Arm 1: OSI-A (Paper/PDF + Therapist Support)
Participants receive the OSI-A parent-guided cognitive behavioural intervention through structured paper and/or PDF materials. Families are provided with all programme modules at the start of treatment and receive brief weekly support from a trained therapist via telephone or videoconference (approximately 20–30 minutes per week) over an eight-week period. A follow-up review session is conducted approximately four weeks after programme completion.
Arm 2: OSI-A + AI-Supported Delivery
Participants receive the same OSI-A parent-guided cognitive behavioural intervention delivered through structured paper and/or PDF materials, together with the same schedule of weekly therapist support. In addition, families receive access to AILA, a WhatsApp-based conversational AI assistant that provides reminders, psychoeducation, activity guidance, progress monitoring, and support between therapist contacts. AILA is designed to enhance engagement and programme adherence and does not replace therapist support or provide clinical decision-making.
Arm 3: Treatment as Usual (TAU)
Participants continue to receive the standard services available through participating outpatient mental health services, autism support programmes, community health centres, educational services, or other healthcare providers. The nature and intensity of services received may vary according to local practice, service availability, and individual family needs. Participants allocated to this arm do not receive OSI-A materials, therapist support associated with OSI-A, or access to the AILA chatbot during the study period.
The intervention period is eight weeks, with follow-up assessments conducted at 14 and 24 weeks post-randomisation.
Randomisation: Participants are allocated in a 1:1:1 ratio using computer-generated randomisation implemented through REDCap.

Intervention Start Date

2026-06-15

Intervention End Date

2027-11-30

Primary Outcomes

Primary Outcomes (end points)

The primary outcomes focus on feasibility and implementation. Together, these measures will help determine whether OSI-A can be successfully delivered and evaluated in the Chilean context, and whether progression to a future definitive trial is justified.
To support this decision, we will use predefined Traffic Light Criteria covering key aspects of study feasibility, including recruitment, intervention adherence, participant retention, data completeness, acceptability, AI chatbot engagement, therapist fidelity, and safety. These criteria will classify each domain as green (proceed), amber (review and modify if needed), or red (do not proceed without substantial changes).
Rather than testing hypotheses about effectiveness, the purpose of these outcomes is to identify potential implementation challenges, assess the practicality of study procedures, and determine whether any modifications are needed before undertaking a larger effectiveness and cost-effectiveness trial.

Primary Outcomes (explanation)

The primary outcomes focus on feasibility and implementation and are intended to assess whether OSI-A can be successfully delivered and evaluated in Chile. Recruitment rate will be calculated as the number of caregiver–child dyads randomised per month. Intervention adherence will be defined as completion of at least 6 of the 8 OSI-A modules, based on programme records and REDCap logs. Therapist adherence will be assessed using therapist logs and supervision records. Participant retention will be measured through completion of follow-up assessments at 14 and 24 weeks post-randomisation.
Data completeness will be assessed as the proportion of completed responses for key outcome measures, including the Child Anxiety Impact Scale–Parent Version (CAIS-P) and the Child Health Utility 9D (CHU-9D). Acceptability will be explored through semi-structured interviews with caregivers and therapists, focusing on participants’ experiences of the intervention. AI chatbot engagement will be assessed among participants in the OSI-A + AILA arm using chatbot usage data. Safety will be monitored through the recording of adverse events and other safety concerns throughout the study.
These outcomes will be interpreted using predefined Traffic Light Criteria to determine whether progression to a future definitive trial is warranted and to identify any modifications required to the intervention or study procedures.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary outcomes include:

Child anxiety symptoms and anxiety-related impairment.
Child emotional and behavioural functioning.
Child health-related quality of life and quality-adjusted life years (QALYs).
Healthcare, educational, and social service utilisation.
Medication use and family-incurred costs.
Caregiver anxiety symptoms, depressive symptoms, and caregiver burden.
Parenting skills and confidence.
Caregiver labour market participation.

Secondary Outcomes (explanation)

Child anxiety symptoms and anxiety-related impairment will be assessed using the Child Anxiety Impact Scale (CAIS-P and CAIS-C) and the Autism Spectrum Condition–Anxiety Scale (ASC-ASD). Emotional and behavioural functioning will be measured using the Strengths and Difficulties Questionnaire (SDQ). Child health-related quality of life will be assessed using the Child Health Utility 9D (CHU-9D), which will also be used to estimate quality-adjusted life years (QALYs).

Healthcare, educational, and social service utilisation will be measured through parent-reported surveys and the Client Service Receipt Inventory (CSRI), including hospital care, primary care, mental health services, community services, educational support, school absenteeism, medication use, and family-incurred costs. Healthcare costs will be estimated by combining resource utilisation data with relevant unit costs.

Caregiver outcomes will include labour market participation, anxiety symptoms (GAD-7), depressive symptoms (PHQ-9), caregiver burden (Zarit Burden Interview – Short Form), and parenting skills and confidence (PRCS). Unless otherwise specified, outcomes will be measured at baseline, 14 weeks, and 24 weeks post-randomisation.

As this is a feasibility study, analyses of secondary outcomes are exploratory and are intended to provide preliminary estimates of clinical and economic effects, assess measure performance, and inform the design of a future definitive effectiveness and cost-effectiveness trial.

Experimental Design

This study is a three-arm, mixed-methods, parallel-group feasibility randomized controlled trial evaluating the implementation of the Online Support and Intervention for Anxiety in Autistic Children (OSI-A) programme in Chile. The study will recruit 180 caregiver–child dyads with autistic children aged 5–12 years who experience clinically significant anxiety symptoms. Participants will be recruited through public health services in Santiago, Chile, and randomly assigned in equal proportions to one of three study arms. Randomization will be conducted at the caregiver–child dyad level using a computer-generated allocation sequence implemented through REDCap. The study will assess the feasibility, acceptability, and preliminary effectiveness of the intervention through quantitative surveys, administrative data, and qualitative interviews with caregivers and therapists. Follow-up assessments will be conducted approximately 14 and 24 weeks after randomization.

Experimental Design Details

Not available

Randomization Method

Randomization will be conducted using the REDCap randomization module and a computer-generated allocation sequence. Eligible caregiver–child dyads will be randomly assigned in a 1:1:1 ratio to one of the three study arms: (1) OSI-A, (2) OSI-A + AI Chatbot, or (3) Treatment as Usual (TAU). Randomization will be stratified by recruitment site (Hospital Dr. Exequiel González Cortés vs. Corporación Municipal de La Florida) and baseline anxiety severity (high vs. low anxiety) to ensure balance across treatment arms. Allocation will occur after completion of baseline assessments and will be concealed from study personnel until assignment.

Randomization Unit

The unit of randomization is the caregiver–child dyad. Each eligible dyad will be individually randomized to one of the three study arms: OSI-A, OSI-A + AI Chatbot, or Treatment as Usual (TAU). There is no separate cluster-level randomization.
Randomization will be stratified by recruitment site and baseline anxiety severity.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Not applicable. The treatment is not clustered; randomization occurs at the caregiver–child dyad level.

Sample size: planned number of observations

180 caregiver–child dyads (60 per treatment arm)

Sample size (or number of clusters) by treatment arms

A total of 180 caregiver–child dyads will be randomised in a 1:1:1 allocation ratio across three study arms: 60 dyads allocated to OSI-A, 60 dyads allocated to OSI-A + AILA, and 60 dyads allocated to Treatment as Usual (TAU).

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

The study is not powered for definitive hypothesis testing. Power calculations suggest that, with N = 180 caregiver–child dyads allocated equally across three arms, the study has adequate power only under an optimistic standardized effect size of approximately Hedges’ g = 0.74 SD for parent-reported clinical outcomes at 14 weeks. Under more moderate assumptions, power is lower. Therefore, estimates will be interpreted as preliminary and used to inform the design of a future fully powered trial.

Supporting Documents and Materials

IRB

Institutional Review Boards (IRBs)

IRB Name

Servicio de Salud Metropolitano Sur Oriente

IRB Approval Date

2026-03-19

IRB Approval Number

2886

IRB Name

Hospital Dr. Exequiel González Cortés

IRB Approval Date

2025-09-26

IRB Approval Number

83-02092025

IRB Name

Facultad de Medicina Universidad del Desarrollo - Clínica Alemana de Santiago

IRB Approval Date

2025-10-15

IRB Approval Number

2025-65

Analysis Plan

Analysis Plan Documents

OSI_A_rct_analysis_plan_vjune2026

MD5: cca5d7ac20b656009e73a9efbdb95584

SHA1: eb741a619facb156e4eb4c8b953aad8e68cee32e

Uploaded At: June 11, 2026