AVATAR therapy implementation and innovation: Testing delivery of standard AVATAR therapy across settings in sub-Saharan Africa and South Asia alongside UK-based automation
Associated organisationsAddis Ababa University · All India Institute of Medical Sciences New Delhi · IND Rakesh Chadda 62307 · IND Yumnam Surbala Devi 237342 · King's College London · University College London · University of GlasgowEurope PMC affiliations are not treated as award recipients or mapped locations.
Funding£2.7M
PeriodAug 2024 — Jul 2028
In plain English
AI plain-English summary
People who hear voices that others do not will soon be able to confront those voices through a digital avatar, delivered in their own language in Ethiopia and India, or via an automated AI system in the UK. Hearing distressing voices (auditory verbal hallucinations) affects millions of people worldwide, yet most evidence-based therapies exist only in high-income countries and require specialist therapists. This project tackles two gaps: the lack of culturally adapted voice-hearing therapy in low- and middle-income countries, and the shortage of trained therapists everywhere. The researchers will establish training hubs in Ethiopia and India, adapting AVATAR therapy—where a patient speaks to a computer-generated representation of their voice—into Amharic, Oromo, Hindi, and other languages spoken by over half a billion people. In parallel, they will build an English-language version where AI generates the avatar’s responses automatically, then test whether non-specialist mental health workers can deliver it effectively. If successful, this work could make a proven therapy accessible to populations currently excluded from it—people in low-resource settings and those on long waiting lists in the UK. It would also demonstrate that AI can handle complex therapeutic dialogues, potentially reshaping how mental health care is delivered at scale.
View original technical description
In this project we will establish AVATAR therapy Implementation and Training Hubs in two distinct LMIC settings (Ethiopia and India). For the first time, we will test the feasibility and cultural acceptability of standard AVATAR therapy delivered in Indian and Ethiopian languages, spoken by over half a billion people. In parallel work, we will develop an English-language version of AVATAR therapy in which automated dialogues are delivered using Artificial Intelligence (AVATAR-AI) and test the feasibility of delivery by a non- specialist UK workforce. We will establish Hub leadership within each LMIC setting, building capacity for sustainable implementation of standard AVATAR therapy (now) and AVATAR-AI (future). This study has 5 key aims, delivered over 5 workpackages: Aim 1: To develop, through diverse stakeholder engagement, ethically responsible and culturally acceptable adaptation of AVATAR therapy across LMIC and UK sites. Aim 2: To establish Implementation Hubs in Ethiopia and India, to deliver and train AVATAR therapy. Aim 3: To develop and conduct user testing of AI-powered conversational agents to deliver AVATAR therapy dialogues (in English). Aim 4: To deliver AVATAR-AI, as a blended digital therapeutic. Aim 5: To test the feasibility, acceptability, and clinical impact of AVATAR-AI (UK), delivered by non-specialist mental health workers.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know