Currently available evidenced-based treatments are specifically designed for smokers who are ready to quit within 30 days, which constitute only 12% of smokers [1]. Evidence-based strategies are needed to reach the remaining population of unmotivated smokers and accelerate their motivation to quit before smoking-related illness occurs (or worsens).

Because unmotivated smokers are not likely to seek treatment, the majority of research has focused on integrating smoking cessation into infrastructures that are frequented by smokers, such as health care or worksites [2-6]. Mobile apps are another channel to reach smokers who are unmotivated to quit. However, while there are studies examining the effectiveness of mobile apps for smoking cessation [7], to our knowledge, there are no studies that focus on using mobile apps for motivating smokers who are not ready to quit. This is likely because it is difficult to encourage unmotivated smokers to use a smoking cessation app because the content is not personally relevant to them. Given prior research showing that the use of mobile apps increases with greater personal relevancy and perceived utility [8], it is important to find out what is important to unmotivated smokers, provide information on that topic using a mobile app, and then gradually weave in a smoking cessation intervention to increase intervention engagement and motivate quitting. This is consistent with the “foot-in-the-door” technique from social psychology [9]. For example, in 1 smoking cessation induction trial, smokers were offered asthma education on how to prevent asthma attacks in their children with asthma who were recently admitted to urgent care for an asthma exacerbation. The intervention largely focused on what they cared about most (preventing future asthma attacks in their children), but smoking cessation and how to quit were gradually woven into the intervention over time [10-13].

We previously conducted a series of focus groups with smokers who were unmotivated to quit and found that the most important thing on their minds was managing mood and stress, and that they would be therefore interested in a mobile app that could help them with mood management [14]. This is consistent with results from our large prior quantitative study with smokers unmotivated to quit, which showed a high prevalence of depression and depressed mood [15]. Therefore, focusing on things that are most important to unmotivated smokers (such as stress and mood) may increase app engagement and prime them to consider quitting smoking. To date, smoking cessation interventions using smartphone apps alone have not been successful [7]. One reason for this could be the lack of focus on content that would most interest unmotivated smokers.

Therefore, we developed a mobile app based on behavioral activation (BA) [16-18], hypothesizing that increases in positive affect and decreases in negative affect could prime unmotivated smokers to consider quitting. BA is an evidenced-based treatment for depression and depressed mood that has been successfully used in both psychiatric and medical populations [19-22] and has been shown to be superior to medication [19]. BA has also been shown to be effective for depression when delivered via a web-based platform [23]. BA purports that depression results from chronically low levels of, or loss of, positive environmental reinforcement. The lack of both pleasant events and positive reinforcement is associated with depression and negative avoidant coping behaviors (eg, smoking) [17]. BA improves mood through strategies to increase activation (eg, setting goals to engage in activities that are consistent with personal values and goals), which in turn increases access to natural sources of positive reinforcement (eg, other people, hobbies, and activities) [24].

In a series of prior studies, we iteratively and collaboratively developed a BA-based mobile app with smokers who were unmotivated to quit, through focus groups, a “think aloud study,” and a 1-month single-arm usability trial [14]. This current study aimed to conduct a pilot randomized trial of the BA app with smokers who were unmotivated to quit in order to assess (1) the feasibility (eg, enrollment) of the BA app, (2) satisfaction and engagement with the BA app, (3) if unmotivated smokers complete BA app intervention activities despite risk factors for continued smoking (low motivation to quit, low confidence to quit and high negative affect), and (4) preliminary effectiveness of the BA app on smoking behavior and on motivation and confidence to quit smoking. We hypothesized that (1) smokers who are unmotivated to quit will engage with the BA app, (2) engagement with the BA app will be associated with lower negative affect and greater positive affect, and (3) those who are randomized to the BA app will show positive changes in smoking (eg, set quit date and reduce the number of cigarettes) and in smoking-related variables (greater motivation and confidence to quit).

To our knowledge, there are no mobile apps that focus on motivating unmotivated smokers to quit, which comprise the majority of smokers. Our study filled this gap by developing and pilot-testing a theory-based mobile app that focuses on motivating smokers to quit by directly targeting mood-based barriers to quitting. Our main findings were (1) smokers who were unmotivated to quit were willing to enroll in our study (of those eligible to participate, 57/66, 86%, were randomized), (2) participants reported high levels of satisfaction and engagement with the intervention (ie, scheduling and completing activities), (3) those at highest risk for continued smoking (low motivation to quit, low confidence to quit, and high negative affect) demonstrated high levels of intervention engagement, and (4) the hypothesized associations between theory-based BA constructs (eg, activity scheduling and completion) and outcomes (positive and negative affect and smoking behavior) were in the expected direction. As expected, due to the small sample size, there were no significant differences between the groups on smoking behavior, but there were promising trends in the hypothesized direction regarding the number of cigarettes smoked, setting a quit date, and motivation and confidence to quit. The above findings suggest that mobile app interventions can be made appealing to smokers who are unmotivated to quit and warrant further efficacy testing.

Feasibility studies are used to determine whether an intervention is appropriate for further testing, particularly when there are few or no published studies on a particular intervention technique [35]. Our study met the criteria for intervention feasibility outlined by Bowen et al [35]. Specifically, we demonstrated acceptability (satisfaction, engagement, and no dropout), demand (enrollment rate and use of app), implementation (successful execution), practicality (ease and quality of implementation, and low burden on patients), integration (fit into daily life), and limited efficacy (promising effects of the program on key variables). Therefore, we believe that there is sufficient evidence to warrant further testing in a larger trial.

The novel aspect of the app is that it takes a “foot-in-the-door” approach by focusing on aspects that we found matter most to unmotivated smokers (stress and mood management) with the idea that addressing these risk factors for continued smoking will increase smokers’ receptivity to pushed content on smoking cessation. While it may seem counterintuitive to enroll unmotivated smokers into a smoking cessation app, we were able to demonstrate that our approach is feasible, given our high rate of enrollment and no dropout. One other study has also demonstrated high enrollment of unmotivated smokers by focusing on aspects that matter most to them, and then weaving in smoking cessation messages [12]; however, there was no digital component in this study. Additionally, our sample was comprised of smokers with high levels of depressed mood (21/56, 38% scored above the cutoff for depression on the CES-D) and low motivation and confidence to quit, indicating that we were able to attract smokers who are at risk for continued smoking.

Our results showed moderate to high levels of satisfaction and high levels of engagement with the intervention (planning and completing activities). High levels of engagement with the app also mean that these smokers, who were unmotivated to quit, received our intermittent, pushed messages about smoking cessation; messages that they would not otherwise receive. The messages focused on key barriers to quitting and were based on state-of-the-art motivational approaches [12,36] as well as our previous mixed methods research with the target population [14]. We believe that this approach was successful because there were no study dropouts.

Although counterintuitive, those with greater risk factors for continued smoking (low motivation to quit, low confidence to quit, and high negative affect) completed a greater number of a variety of activities over the course of the study (ie, engaged with the hypothesized active ingredients of the intervention). These findings have several implications. First, just because smokers are unmotivated to quit, or have high risk factors for continued smoking, does not mean that they are not willing to engage with an intervention. This finding lends support to continue with future research in developing digital behavior change solutions for those who are unmotivated to change. Second, these associations were prospective and provided support for the theoretical model underpinning the intervention. Finally, this finding dovetails with a previous study with smokers, where smokers who were unmotivated to quit were more likely to quit with an intensive motivational interviewing intervention than a less intensive one [13].

Consistent with BA theory, we hypothesized that completing a greater number of activities would lead to higher positive affect and lower negative affect and that positive changes in mood would be associated with less smoking. We found support for some but not all of these associations. We found that pleasant activities were associated with greater positive affect and that positive affect predicted smoking fewer cigarettes the next day. Although significance was borderline due to the small sample, the fact that these trends were based on a theoretical model specified a priori give support for future study of these relationships in a larger sample. Planning more activities was also associated with lower levels of negative affect (although only a trend), but lower levels of negative affect were significantly and prospectively associated with smoking fewer cigarettes. This set of findings indicates support for key relationships with the BA model and therefore warrants further investigation.

As with any pilot feasibility study, there are limitations in what can be concluded from the study. For example, the effectiveness of the app on smoking behavior cannot be determined due to lack of power, although there were promising trends in the hypothesized direction regarding intentions to quit. While we can conclude that the results are promising and warrant a larger trial, we cannot conclude that the cross-sectional and prospective relationships between hypothesized variables will be demonstrated in a larger trial. In general, our effects were small in size and should be interpreted with caution. The trial was also conducted in the United Kingdom with a relatively racially and ethnically homogeneous population, so it is unclear if smokers who are unmotivated to quit in other countries and smokers of other races and ethnicities would find the app equally satisfactory or if they would engage with the app to the same extent. Additionally, participants and study staff could not be masked to treatment conditions, given the study design and digital platform. Finally, although satisfaction scores were in the “moderate to high” range, future studies should do qualitative work postintervention to learn about how to increase satisfaction with the app.

These limitations should be viewed in the context of innovation. Innovation regarding intervention approaches for unmotivated smokers has stalled. The vast majority of digital interventions include content and features that are tailored for those who are motivated to quit. On a broader scale, digital interventions in general have not been sufficiently leveraged to target those who are not motivated to change. This study presents progress in this direction and could potentially serve as a paradigm for others to blend social psychology principles (eg, foot-in-the-door approach) with evidenced-based treatments (BA) to motivate change across different areas of health using digital platforms.