Abstract
INTRODUCTION: Digital cessation support appeals to pregnant smokers. In two pooled RCTs, MiQuit, a pregnancy-specific tailored text messaging intervention, did not show effectiveness for validated prolonged abstinence. However, secondary outcomes and potential moderators and mediators have not been investigated. We aimed to determine, using pooled RCT data: (1) MiQuit effectiveness on a range of smoking outcomes; (2) whether baseline tobacco dependence or quit motivation moderate effectiveness; (3) whether hypothesized mechanisms of action (quitting determination, self-efficacy, baby harm beliefs, lapse prevention strategies) mediate effectiveness. METHODS: Pooled data analysis from two procedurally identical RCTs comparing MiQuit (N = 704) to usual care (N = 705). Participants were smokers, <25 weeks pregnant, recruited from 40 English antenatal clinics. Outcomes included self-reported 7-day abstinence at 4 weeks post-baseline and late pregnancy, and prolonged abstinence. Late pregnancy outcomes were also biochemically validated. We used hierarchical regression and structural equation modeling. RESULTS: MiQuit increased self-reported, 7-day abstinence at 4 weeks (OR = 1.73 [95% CI 1.10-2.74]) and was borderline significant at late pregnancy (OR = 1.34 [0.99-1.82]) but not for prolonged or validated outcomes. Effectiveness was not moderated by baseline dependence (heaviness of smoking "low" vs. "moderate-high") or motivation (planning to quit ≤30 days [high] vs. >30 days [low]), but effects on self-reported outcomes were larger for the high motivation sub-group. MiQuit had a small effect on mean lapse prevention strategies (MiQuit 8.6 [SE 0.17], UC 8.1 [SE 0.17]; P = .030) but not other mechanisms. CONCLUSIONS: MiQuit increased short-term but not prolonged or validated abstinence and may be most effective for those motivated to quit sooner. IMPLICATIONS: Digital cessation support appeals to pregnant smokers. MiQuit, a tailored, theory-guided text messaging program for quitting smoking in pregnancy, has not shown effectiveness for validated prolonged abstinence in two previous RCTs but its impact on other smoking outcomes and potential mechanisms of action are unknown. When pooling trial data, MiQuit increased self-reported short-term abstinence, including making a quit attempt and abstinence at 4-week follow-up, but not late pregnancy, sustained, or validated abstinence. MiQuit appeared effective at late pregnancy for participants with high quitting motivation, but its mechanisms of action remain uncertain. Additional support components are likely required to enhance effectiveness.