A group of young research talents from the Laboratory of Data Discovery for Health (D24H) teamed up with renowned scientists and health communication specialists from Thailand, Hong Kong and Singapore conducted an investigative study on the impact of Artificial Intelligence (AI) chatbots on COVID-19 vaccine hesitancy and acceptance. Focusing on adult guardians of children and seniors in the three regions, the study revealed improved vaccine acceptance and confidence among some chatbot users across different study groups, providing further evidence that well-designed chatbots can positively influence health behaviours. The detailed findings were published in the latest issue of npj Digital Medicine [link to the publication] on May 25, 2023.

Since the COVID-19 outbreak, vaccine hesitancy has been a major challenge in increasing vaccine uptake, particularly among seniors and parents of children during the COVID-19 pandemic. An additional challenge in the course of the pandemic was the infodemic brought forth by an overabundant amount of online and offline information, including deliberate attempts to spread misinformation and undermine pandemic response. As a result, chatbots have gradually gained popularity and showed great potential in becoming an efficient and effective tool for disseminating accurate health-related information. The D24H is privileged to receive funding from the Vaccine Confidence Fund (VCF), a highly competitive grant that supports this multisite parallel randomised control trial (RCT) research on the best use of social media and online platforms to enhance vaccine confidence and uptake.

This study used two chatbots – (i) the D24H chatbot in Hong Kong and Singapore; and (ii) the Thai ChatSure chatbot in Thailand. The team compared both chatbot’s intervention data with the control group data and concluded mixed results on these conversational agent’s effectiveness in improving the study groups’ overall vaccine confidence and acceptance. However, positive results collected from this study showed:

1. The Thailand child group intervened by the Thai ChatSure chatbot demonstrated some capacity to improve vaccine confidence and acceptance, measured by the Vaccine Confidence IndexTM (VCI), through knowledge enhancement; and

2. Chatbots were more effective in improving vaccine confidence and acceptance among minorities and people with lower level of education.

Ms Shirley Kwok, one of the young research talents leading this study, expressed, “Beyond our statistical findings, we found personal stories to be more persuasive for online communication campaigns. These interactive digital campaigns require constant innovation and advancements in AI technology, Natural Language Processing (NLP), and machine learning. What’s more, human behavioural analysis is crucial for chatbot development and operation to ensure accurate responses, especially in free text conversations. Based on the data collected for this study, our team aims to develop a more effective chatbot to address vaccine hesitancy across various vaccinations. By expanding the content scope to cover all vaccines, we hope to leverage the capabilities of tools such as ChatGPT to create a chatbot that addresses most vaccine-related concerns and information needs of diverse communities.”

Echoing Shirley’s comment on the future development and importance of chatbots within the medical communication space, Dr Leesa Lin, Research Director of the D24H, shared, “To more effectively serve the needs of Hong Kong and Greater China users, our team will collaborate with local partners to explore options and transition to a reliable Chinese GPT model. This study and our continued chatbot R&D initiative are part of our ‘Trust in Technology’ strategy to boost confidence and belief in chatbots’ reliability, security, and effectiveness. Trust is an essential component for technology adoption, and our goal is, with the help of advanced technology, to understand how the general public perceives and views different health technologies, including mRNA, disease screening, and AI applications, by taking an all-encompassing, interdisciplinary approach through behavioural, communication, and medical science integrations.”

Acknowledgements: This study was supported by the Vaccine Confidence Fund (VCF) and AIR@InnoHK administered by Innovation and Technology Commission.

About the research team
The study was conceived by Dr Leesa Lin and Professor Joseph Wu. The study was designed by Dr Leesa Lin and Professor Joseph Wu, with important inputs from Saudamini Vishwanath Dabak, Dr Minah Park, Dr Alex Cook, Dr Ed Pertwee, Dr Javier Elkin, Professor Heidi Larson, and Dr Kathy Leung. The RCT was conducted by Kristi Lee, Saudamini Vishwanath Dabak, Dr Vivian Kong, Dr Minah Park, Madison Silzle, Dr Chayapat Rachatan, Dr Alex Cook, Aly Passanante, Dr Kathy Leung and Dr Leesa Lin. The data were collated by Kristi Lee, Dr Saudamini Vishwanath Dabak, Dr Vivian Kong, Dr Minah Park, Madison Silzle, Chayapat Rachatan, Dr Alex Cook, Aly Passanante and Dr Kathy Leung, and analysed by Kristi Lee, Shirley Kwok, Zhengdong Wu, Dr Eric Lau, Dr Kathy Leung and Dr Leesa Lin.

About the Laboratory of Data Discovery for Health (D²4H)
The Laboratory of Data Discovery for Health aims to gather and curate massive, unique data resources and develop deep, frontier analytics to protect global public health while improving individual healthcare through precision medicine. The D24H brings together a multi-disciplinary team of the world’s leading scientists to apply Artificial Intelligence and big data in ways that will transform our approaches to understanding and treating disease.

Spearheaded by the University of Hong Kong with support from world-renowned academic institutions nationally and internationally, the D24H is keen to work with health authorities such as the World Health Organization and China Center for Disease Control and Prevention. Through harnessing collaboration across multiple disciplines and sectors, the D24H advances the frontiers of healthcare technology in Hong Kong, the Greater Bay Area, and beyond to produce ‘moonshots’ that will have tremendous healthcare benefits for global health.

For details of the VCF Award: https://charity.org/give-global-blog/launching-vaccine-confidence-fund-facebook-and-merck