Codevelopment, Implementation, and Evaluation of an AI-powered Conversational Interface for Tobacco Cessation: Protocol for a Single-Group Pre-Post Study
An Indian multi-site study protocol describes CARE, a multilingual AI conversational interface to be evaluated alongside existing tobacco-cessation services.
TL;DR
- The CARE protocol outlines development, feasibility testing and a six-month pre-post evaluation across three tobacco cessation centers.
- The intervention is planned as a multilingual mobile-app chatbot using retrieval-augmented generation and an LLM.
- The study has no control arm, so its planned effectiveness findings will be preliminary and hypothesis-generating.
The published protocol describes CARE as a conversational interface being developed with tobacco users and cessation providers, followed by feasibility testing and a six-month pre-post assessment. It lists engagement, usability, acceptability, quit attempts and intention to quit among the evaluation areas. No outcome results are reported in this protocol. [PubMed; Public Health Foundation of India.] [1] [2]
The protocol says the tool will use retrieval-augmented generation and a large language model, and will be delivered in a mobile app with multilingual support. It describes implementation across three tobacco cessation centers and explicitly notes that the single-group design has no control arm, making effectiveness results preliminary rather than causal. [PubMed; Public Health Foundation of India.] [1] [2]
Why it matters
The study connects AI product development with existing cessation-service infrastructure in India. Its design and planned measures are relevant to implementation, but the protocol provides no evidence yet that the intervention changes cessation outcomes.
Editor's note
Protocol only; no results are available. Single-group pre-post design has no control arm. No cessation or treatment advice is provided.