Combining an Effective Rule-based Chatbot With ChatGPT-supported Real-time Question-answer Functions to Promote Seasonal Influenza Vaccination Among Community-dwelling People Aged 65 Years and Above

This study is a partially blinded (outcome assessors and data analysts) and parallel-group randomized controlled trial. Community-dwelling people aged ≥65 years who have not received the SIV for the approaching flu season will be randomized evenly (1:1) into either the intervention group or the control group. During the intervention period, the intervention group will have access to the hybrid combination of a rule-based chatbot (SIV-Chatbot) with the ChatGPT-supported real-time Question & Answer (QA) functions, while the control group will have access only to the SIV-Chatbot.

The SIV-Chatbot will deliver the same interventions for both groups. The number of intervention sessions (four) and timing (week 0, 2, 4 & 6) will also be the same. In the intervention group, the hybrid Chatbot will invite the participants to raise open-ended questions related to SIV using flexible methods (such as voice messages, typing with keyboards, or handwriting) at each session. The ChatGPT will answer these questions immediately in the format of text messages. The functions can be used repeatedly throughout the intervention period (weeks 0-6), and participants will be free to spend as much time as they want on such functions.

Telephone interviews will be conducted at baseline (T0), after completion of the intervention (T1), and six months after T1 (T2). The primary outcome will be the self-reported SIV uptake measured at T2. The research team will validate that outcome by requesting the participants upload an image of their SIV receipt, while hiding their personal identification, via the WhatsApp account used in this study.