Ecological Momentary Assessments in an Incentive-Based Intervention for Smoking Cessation: Predicting Near-Term...

Cigarette smoking is the leading cause of preventable death in the U.S. We evaluate the predictors of intervention engagement and smoking status during a digital incentive-based intervention for tobacco cessation among Medicaid beneficiaries using ecological momentary assessments (EMAs).

Patients (N = 37) self-enrolled from community-based health centers were incentivized monetarily to submit breath samples and complete EMAs twice daily for four weeks. Using unadjusted and adjusted mixed-effect logistic regression (random intercept for patient), we evaluated if EMA responses and time in intervention (i.e., length of participation) predicted subsequent engagement (i.e., breath sample submission) and subsequent smoke status (smoke positive ≥ 6 ppm CO) throughout the 30-day intervention.

In adjusted models, being around others who smoke, cigarette craving, having smoked in the last 4 h, and higher than normal or very high stress were predictive of non-engagement at the next window (odds ratios [ORs]: 0.281-0.577). Being fairly confident (OR: 2.617) and very confident (OR: 5.337) in one's ability to not smoke was the predictive of engagement at the next window. Recent smoking and higher-than-normal stress were predictive of positive smoke status at the next window (ORs: 1.989-2.322). Being very confident in one's ability to not smoke, and at least fairly motivated to not smoke were each predictive of negative smoke status at the next window (ORs: 0.092-0.219).

The results support that EMAs during incentive-based interventions are useful to predict both engagement and smoke status, which could inform future just-in-time adaptive interventions to enhance engagement and response.