Background: The association between greater utilization of Web-assisted tobacco interventions and increased abstinence rates is well recognized. However, there is little information on how utilization of specific website features influences quitting.
Objective: To determine the association between utilization of informational, interactive, and online community resources (eg. bulletin boards) and abstinence rates, with the broader objective to identify potential strategies for improving outcomes for Web-assisted tobacco interventions.
Methods: In Spring 2004, a cohort of 607 quitplan.com users consented to participate in an evaluation of quitplan.com, a Minnesota branded version of QuitNet.com. We developed utilization measures for different site features: general information, interactive diagnostic tools and quit planning tools, online expert counseling, passive (ie, reading of bulletin boards) and active (ie, public posting) online community engagement, and one-to-one messaging with other virtual community members. Using bivariate, multivariate, and path analyses, we examined the relationship between utilization of specific site features and 30-day abstinence at 6 months.
Results: The most commonly used resources were the interactive quit planning tools (used by 77% of site users). Other informational resources (ie, quitting guides) were used more commonly (60% of users) than passive (38%) or active (24%) community features. Online community engagement through one-to-one messaging was low (11%) as was use of online counseling (5%). The 30-day abstinence rate among study participants at 6 months was 9.7% (95% Confidence Interval [CI] 7.3% - 12.1%). In the logistic regression model, neither the demographic data (eg, age, gender, education level, employment, or insurance status) nor the smoking-related data (eg, cigarettes per day, time to first morning cigarette, baseline readiness to quit) nor use of smoking cessation medications entered the model as significant predictors of abstinence. Individuals who used the interactive quit planning tools once, two to three times, or four or more times had an odds of abstinence of 0.65 (95% Confidence Interval [CI] 0.22 - 1.94), 1.87 (95% CI 0.77 - 4.56), and 2.35 (95% CI 1.0 - 5.58), respectively. The use of one-to-one messages (reference = none vs 1 or more) entered the final model as potential predictor for abstinence, though the significance of this measure was marginal (OR = 1.91, 95% CI 0.92 - 3.97, P = .083). In the path analysis, an apparent association between active online community engagement and abstinence was accounted for in large part by increased use of interactive quitting tools and one-to-one messaging.
Conclusions: Use of interactive quitting tools, and perhaps one-to-one messaging with other members of the online community, was associated with increased abstinence rates among quitplan.com users. Designs that facilitate use of these features should be considered.