Chat-based instant messaging support combined with brief advice and active referral for smoking cessation in proactively recruited community smokers: a pragmatic, cluster randomized controlled trial

Abstract

SIGNIFICANCE: Mobile instant messaging tools (e.g., WhatsApp, Facebook Messenger) are widely used but under-studied as a mHealth modality for tobacco dependence treatments. We developed and evaluated a chat-based cessation intervention delivered through mobile instant messaging for community smokers. METHODS: We conducted a two-arm, pragmatic, cluster randomized controlled trial in which 1185 adult daily cigarette smokers (mean [SD] age=43.4 [15.4] years; 79.7% men) were proactively recruited from and cluster randomized by 68 community sites throughout Hong Kong. Subject in the intervention group (n=591) received a brief advice to quit plus an option of being referred to a preferred smoking cessation service (active referral) at baseline, and three months of chat-based, personalized cessation support grounded in behavioral change theories through instant messaging. The control group (n=594) received very brief advice to quit and a self-help booklet at baseline. The primary outcomes were self-reported past 7-day point prevalence abstinence (PPA) at 3- and 6-month after baseline. Secondary outcomes included abstinence validated by exhaled carbon monoxide (<4ppm) and salivary cotinine (<10ng/ml) at 3- and 6-month. RESULTS: Overall follow-up rates were 75.4%, 70.0% and 69.2% at 1-, 2- and 3-month, respectively, with no significant between-group differences (p=0.60- 0.90). By intention-to-treat, the intervention group had higher PPA than the control group at 1- (10.7% vs 7.7%; p=0.082), 2- (14.4% vs 9.1%; p=0.005) and 3-month (15.2% vs 10.1%; p=0.008). The results were corroborated by biochemically-validated abstinence at 3-month (7.6% vs 4.0%; p=0.009) and did not change materially after adjusting for imbalanced baseline factors and clustering effect.