AIDA: A Virtual Health Assistant for Persons with TB

You’ve heard of artificial intelligence (AI) being applied to so many aspects of our world today—image generation, software coding, personal assistants—but how can we utilize AI to directly benefit people with tuberculosis (TB)?

In 2024, KNCV Tuberculosis Foundation initiated the in-house development of a virtual TB treatment supporter for people on BPAL TB treatment, accessible via WhatsApp. This AI-powered digital treatment supporter is interactive, multi-language, and educated on TB treatment, common challenges, and overall health concerns. In 2025, we named this virtual TB treatment supporter: Aida.

How does Aida work?

• Aida uses OpenAI’s GPT4o model and local available resources such as patient information folders, guidelines and other materials to respond to questions from people on TB treatment.
• People on treatment can decide themselves to register for this additional service. It can be offered as a support tool but never forced upon.
• After registration and consenting via WhatsApp, the person on TB treatment will receive planned motivational messages from Aida and can engage 24/7 by asking questions or sharing challenges they face at any moment in the language of their choice.
• Registration only requires a name, age, gender and facility or clinic in which the treatment takes place. It does not require staff training or staff monitoring.
• All messages that are sent between Aida and the person on treatment are kept confidential and not shared with health staff nor National TB Program. Whenever needed, the user can be invited to separately message the healthcare worker in case of human follow-up need.
• The use of Aida can be promoted using posters or leaflets at different healthcare facilities.

Our idea is that Aida adds extra support (counselling) to persons on TB treatment, accompanying them on a round-the-clock basis, everyday.

A pilot version of the virtual TB treatment supporter took place in November 2024 and this version is now ready for initial field testing. This initial field testing is the first step toward further evaluation of this digital approach. Field testing will yield valuable experience to further development of this approach. Although data derived from field testing will not be published in scientific journals, it will help us refine our digital initiatives to support people on TB treatment.

Implementation period

Supported by the Gates Foundation, activities within the Chat4TB Project are planned for a period of 4 – 6 months between March– October 2025.

Activities

KNCV will ensure that a localized version of Aida is available for use via WhatsApp, will provide the overall messaging framework, promotional and instructional materials, and will offer guidance for enrollment, monitoring and conducting focus group discussions.

The field testing is planned with partners in-country across different regions. The overall aim of the in-country partner is to manage and oversee field testing of Aida.

In an iterative manner, we will analyze the user interactions for accuracy of the Aida’s responses (content, language) as well as for technical performance (timing, prompting). We will also collect and analyze feedback from participants and partners to understand the user experiences and identify areas for improvement.

Expected outcomes

• Aida will provide an additional layer of care for people currently on TB treatment in selected facilities.
• A locally tailored intervention capable of serving persons globally, with input and experiences from the in-country partners.
• Refinement of a collaborative design of a context-specific AI-powered solution for supporting people on TB treatment.
• Experience in utilizing Large Language Model (LLM) technology, building organizational capacity for AI projects in the future across different regions.
• We are aiming to validate Aida and determine its potential to scale responsibly and relevantly across different settings. This first experience will allow us to draft plans for further evaluation in 2026.