Effectiveness of eHealth Smoking Cessation Interventions: A Systematic Review and Meta-Analysis (Preprint)

BACKGROUND The rapid advancement of electronic health (eHealth) and mobile health (mHealth) technologies has driven researchers to design and evaluate numerous technology-based interventions to promote smoking cessation. The evolving nature of cessation interventions emphasizes a strong need for knowledge syntheses. OBJECTIVE This systematic review and meta-analysis aims to (1) summarize recent evidence from randomized controlled trials (RCTs) regarding the effectiveness of eHealth-based smoking cessation interventions in promoting abstinence; and (2) assess non-abstinence outcome indicators, such as cigarette consumption and user satisfaction, via narrative synthesis. METHODS We searched for studies published in English between 2017 to June 30, 2022 in four databases: PubMed (including MEDLINE), PsycINFO, Embase, and Cochrane Library. Two independent reviewers performed study screening, data extraction, and quality assessment based on the GRADE framework. We pooled comparable studies based on population, follow-up time, intervention, and control characteristics. Two researchers performed independent meta-analysis on smoking abstinence using the Sidik-Jonkman random-effects model and log Risk Ratio (log RR) as the effect measurement. For studies not included in the meta-analysis, the outcomes were synthesized narratively. RESULTS A total of 464 studies were identified through the initial database search after removing duplicates. Following screening and full-text assessments, we deemed 39 studies (n=37,341 participants) eligible for this review. Of these, 28 studies were shortlisted for the meta-analysis. According to the meta-analysis, SMS/App text messaging can significantly increase both short-term (3-month) abstinence (log RR=0.50, 95% CI 0.25 to 0.75, I2=0.72%) and long-term(6-month) abstinence (log RR=0.77, 95% CI 0.49 to 1.04, I2=8.65%), relative to minimal cessation support. The frequency of texting did not significantly influence treatment outcome. mHealth apps may significantly increase abstinence in short term (log RR=0.76, 95% CI 0.09 to 1.42, I2=88.02%) but not in long term (log RR=0.15, 95% CI -0.18 to 0.48, I2=80.06%) in contrast to less intensive cessation support. Additionally, personalized/interactive interventions showed a moderate increase in cessation for both the short-term (log RR= 0.62, 95% CI 0.30 to 0.94, I2= 66.50%) and long-term (log RR= 0.28, 95% CI 0.04 to 0.53, I2= 73.42%). In contrast, studies without any personalized or interactive feature found no significant impact. Finally, the treatment effect was similar between trials using biochemically verified or self-reported abstinence. Among studies reporting outcomes besides abstinence (n=20), a total of 11 studies reported significantly improved non-abstinence outcomes in cigarette consumption (3/14, 21.43%) or user satisfaction (8/19, 42.11%). CONCLUSIONS Our review of 39 RCTs found that recent eHealth interventions may promote smoking cessation, with mHealth being the dominant approach. Despite their success, the effectiveness of these interventions may diminish with time. The design of more personalized intervention designs could potentially benefit future studies. CLINICALTRIAL PROSPERO International Prospective Register of Systematic Reviews CRD42022347104