AI Conversational Agent to Improve Varenicline Adherence: Protocol for a Mixed Methods Feasibility Study

Area of Science:

• Digital health interventions
• Artificial intelligence in healthcare
• Smoking cessation research

Background:

• Varenicline is effective for smoking cessation, but adherence is a challenge, with one-third nonadherent by week two.
• Behavioral support enhances medication adherence, leading to the development of ChatV, an AI health bot.
• ChatV provides medication reminders, answers questions, and tracks intake to improve varenicline adherence.

Purpose of the Study:

• To assess the feasibility of the ChatV health bot for improving varenicline adherence.
• To determine if ChatV is used as intended by participants.
• To evaluate the appropriateness of proceeding with a randomized controlled trial for ChatV.

Main Methods:

• A mixed methods feasibility study involving 40 participants using ChatV with a 12-week varenicline regimen.
• Data collection includes chatbot usage metrics, questionnaires on adherence and usability, and interviews.
• Progression criteria will guide decisions on a future randomized controlled trial, with a health equity lens applied throughout.

Main Results:

• Participant enrollment for the study is scheduled to commence in January 2024.
• The study will collect data on ChatV adoption, acceptability, appropriateness, fidelity, and usability.
• Preliminary data on medication adherence and smoking cessation will be gathered to assess treatment effects.

Conclusions:

• Study results will inform the decision to advance to a larger randomized controlled trial based on predefined progression criteria.
• The study will explore key implementation outcomes: acceptability, appropriateness, fidelity, adoption, and usability of the AI health bot.
• Findings will be crucial for refining the ChatV intervention and optimizing its effectiveness in smoking cessation support.