CONFERENCE PROCEEDING
Design and evaluation of a personalized mHealth intervention system based on machine learning to promote smoking cessation in China
,
 
,
 
,
 
,
 
 
 
More details
Hide details
1
School of Public Health, Fudan University, Shanghai, China
 
 
Publication date: 2025-06-23
 
 
Tob. Induc. Dis. 2025;23(Suppl 1):A2
 
KEYWORDS
TOPICS
ABSTRACT
BACKGROUND: In China, over 50 million smokers want to quit, yet most lack access to cessation services. This study aims to design a personalized smoking cessation mobile intervention system integrating professional perspectives and smokers' opinions using machine learning, and to evaluate its effectiveness.
METHODS: First, we developed a personalized mHealth intervention for smoking cessation. Based on Health Action Process Approach (HAPA), we identified the Behavior Change Techniques (BCTs) required by users. For the system function, the BCTs guided the design of the interactions, like games helping users weigh pros and cons. For the cessation messages, we added the insights from peer quitters on the basis of traditional scientific guidance to form a comprehensive message database. For each message, we identified the corresponding quitting stage according to HAPA. Through machine learning, the system could recommend messages based on user’s quitting stage and preferences.
We conducted a randomized controlled trial from September 2024 in China. Current smokers aged 18-65 were recruited online and randomized to intervention group (mHealth program) or to the control group (electronic cessation handbook) in a 1:1 ratio. After 3-month intervention, we conducted the follow-up survey, the primary outcome was 7-day point prevalence abstinence (PPA) biologically validated using saliva nicotine test strips.
RESULTS: The baseline characteristics of participants between the intervention group (n=136) and control group (n=136) were comparable. Follow-up was completed by 113 and 116 participants, respectively. By intention-to-treat analysis, the biologically validated 7-day PPA rate was 17.6% in the intervention group and 7.4% in the control group (OR=2.70, p=0.010). Compared with baseline, the average daily cigarette consumption of participants in intervention group decreased from 8.7 to 2.2 (P<0.001). Among intervention group users, 83.1% reported satisfaction with the intervention.
CONCLUSIONS: The personalized mHealth intervention system may help smokers to quit, which can therefore be considered for large-scale implementation in China.
eISSN:1617-9625
Journals System - logo
Scroll to top