The National Institutes of Health Institutional Review Board deemed the study exempt on October 11, 2019. ICF’s Institutional Review Board deemed the study exempt on November 19, 2019 and approved an amendment on February 26, 2020. All participants received information about the study and its procedures and verbally consented to participate in the study before the audio recording began. Userworks, Inc (Silver Spring, MD), a user experience and usability research firm, had access to personally identifiable information for participant recruitment and screening purposes, whereas all authors did not. UserWorks assigned enrolled participants identification numbers (eg, P1), which were used during the focus group discussions, and subsequently in the transcripts and data analysis. Participants received US $150 gift cards as an incentive.

We partnered with Userworks to recruit a convenience sample of 38 participants through its research panels and other commercial platforms (eg, Craigslist). UserWorks contacted candidate participants via email and screened those interested over the phone for enrollment eligibility (Note S1 in Multimedia Appendix 1). Participants were eligible for inclusion in the study if they self-reported being 18‐29 years old, neither 4-year college educated nor enrolled in a 4-year college as an indicator of low socioeconomic status [38-41], current cigarette smokers who smoked at least 100 cigarettes in their lifetime and smoked every day or some days, willing to quit smoking within 6 months, not using smoking cessation aids, not using noncigarette combustible tobacco products (eg, cigars), smartphone owners, and able to speak English. Recruitment occurred between January and April 2020 and no participants withdrew after enrollment. Recruitment was not based on data saturation. However, the sample size was deemed sufficient to achieve data saturation [42]. The 32-item consolidated criteria for reporting qualitative research checklist appears in Table S1 in Multimedia Appendix 1 [43].

We held 12 semistructured focus groups to explore perceptions of a newly developed smoking cessation mobile intervention, Quit Journey, and its features, including the ABM mobile game described here, among young adults who smoked and were socioeconomically disadvantaged. The game was presented as a concept in 8 predesign focus group sessions and as a prototype version in 4 postdesign sessions. A total of 12 participants were involved in both pre and postdesign focus group sessions. Discussions lasted ≈1.5 hours and were held on GoTo Meeting, audio recorded, and transcribed verbatim. Transcripts were autogenerated by GoTo Meeting, verified by 3 members of our team, and were not returned to participants for comment.

Discussions followed a moderation guide based on the technology acceptance models (Note S2 in Multimedia Appendix 1) [34,35]. TG, a user experience strategist and researcher (male), was the primary moderator, while EL, a strategic communications and marketing project director (female), was the backup moderator and note-taker. TG had training and certification in user research, including focus groups. The moderators had no prior relationship with participants who were informed that the moderators were unaffiliated with the research group that commissioned the study. Only the participants, TG, EL, and SEL, were present during the focus group sessions. Before conducting the focus groups, we conducted dry runs.

Following a deductive thematic analysis [44,45], we developed a codebook and corresponding themes using 5 constructs from the Second Unified Theory of Acceptance and Use of Technology: effort expectancy, facilitating conditions, hedonic motivation, performance expectancy, and social influence [34,35]. These constructs were defined as follows: (1) effort expectancy referred to "perceived ease of use or effortfulness with which one can navigate mobile applications and their features and seamlessly integrate them in one's life"; (2) facilitating conditions were defined as "factors that can aid or impede the uptake or use of mobile applications or their features. These include individual-related (eg, skills, predispositions, prior experiences) and technical-related (eg, infrastructure) factors". (3) Hedonic motivation referred to "perceived fun, pleasure, or enjoyment (or lack thereof) associated with the use of mobile applications and their features". (4) Performance expectancy referred to "perceived usefulness or helpfulness of smoking cessation mobile applications and their features in achieving desired health goals and behaviors". (5) Social influence referred to "perceived importance of significant others' recommendations and approval of using mobile applications and their features" [33,46,47, Wakeman M et al, 2024; Tesfaye L et al, 2024, unpublished data]. Quotes that did not fit under a technology acceptance construct were coded as "not applicable". Additionally, we included sentiment (ie, negative, neutral, and positive), design concepts, and users’ suggestions. Positive and negative sentiment captured statements that indicated “a sense of approval, praise, or certainty” or “a sense of disapproval, criticism, or skepticism about any aspect of smoking cessation mobile applications and their features such as their worthiness, utility and impact, time and effort investment, and compatibility with one's life,” respectively. Neutral sentiment captured statements "that were neither positive or negative in tone, contained an equal number of positive and negative remarks, or were conditional (ie, positive in nature but depended on the presence or absence of another factor" [33,46,47, Tesfaye et al, 2024; Wakeman et al, 2024, unpublished data]. The design concepts code captured statements related to "user interface, content input and delivery, information structure, navigation, typography, and aesthetics of smoking cessation applications or their features" [33,46,47, Tesfaye et al, 2024; Wakeman et al, 2024, unpublished data]. Finally, the suggestions code captured statements that were “concerned with improvements, modifications, or additions to smoking cessation applications or their features aimed to improve their functionality or design" [33,46,47, Tesfaye et al, 2024; Wakeman et al, 2024, unpublished data]. We created additional codes for emerging themes (ie, intent/willingness to use the game, novelty). We coded the transcripts for salient themes applying only one theme from each semantic domain to a quote. When multiple themes were present, we selected the predominant or underpinning theme in conditional quotes (eg, a quote was coded as hedonic motivation if the participant considered the game to be useful if it was fun and engaging). We coded the transcripts for all Quit Journey app features, but only present here results related to the smoking cues mobile game. Our research on the use of mobile technologies for smoking cessation and on Quit Journey and its features has been published elsewhere or is currently under review [33,46,47, Tesfaye et al, 2024; Wakeman et al, 2024, unpublished data].

A member of our team extracted quotes for exploratory analysis using Excel (Microsoft Corporation). Two authors (MW and LT) independently coded focus group transcripts using ATLAS.ti (version 8; ATLAS.ti Scientific Software Development GmbH), which was also used to calculate intercoder agreement [48]. We calculated the Krippendorff c-α, a measure of separating relevant and irrelevant content whereby coders code texts of similar locations and lengths, and the Krippendorff cu-α, a measure of semantic domain reliability whereby coders code a given text for the presence or absence of a semantic domain [48]. Discrepancies were resolved through discussions between MW, LT, and SEL. Qualitative results were not returned to participants for feedback.

The Krippendorff c-α was 0.82, whereas the Krippendorff cu-α was 0.66 for technology acceptance, 0.70 for sentiment, and 0.94 for Quit Journey app features. Sample characteristics appear in Table 1, and detailed participant characteristics in Table 2.

We extracted 143 quotes across all semantic domains. A total of 110 quotes captured technology acceptance constructs and sentiment toward the smoking cues mobile game, whereas 33 quotes included suggestions to improve the game. Sentiment toward the smoking cues game was mixed, with negative (45/110, 40.90%) and neutral (36/110, 32.72%) quotes outweighing those with positive sentiment (29/110, 26.36%) (Table 3). Most quotes centered on performance expectancy (34/110, 30.90%), followed by hedonic motivation (17/110, 15.45%), facilitating conditions (10/110, 9.09%), and effort expectancy (6/110, 5.45%). Social influences on game acceptance and use were not discussed. Below we present illustrative quotes, with participant ID attribution, first on performance expectancy, then followed by all other technology acceptance themes as they were discussed to a lesser extent (Table S2 in Multimedia Appendix 1). Last, we present participants’ suggestions to improve the game’s functionality and design (Table S3 in Multimedia Appendix 1).

Perceptions of the smoking cues game’s usefulness were mixed in sentiment with 26.47% (9/34) of the quotes having a negative sentiment, 35.29% (12/34) neutral, and 38.23% (13/34) positive. Some participants attributed the usefulness of the game to its potential role as a distraction tool from cravings, which was rooted in participants’ familiarity with mobile games and their prior use of them as distraction tools.

Few participants who understood that the true purpose of the game was to modify unconscious attention to smoking cues emphasized the novelty of the concept and its potential usefulness in facilitating cessation.

Some participants expressed concerns over the smoking cues imagery, which could trigger cravings and be counterproductive to their quit attempts.

Regardless of what participants thought the purpose of the game was, they wanted evidence of the effectiveness of the game in helping them quit smoking.

A total of 33 quotes included suggestions to improve the smoking cues game (Table S3 in Multimedia Appendix 1). First, participants suggested renaming the game to reference smoking.

Because of its role in helping them quit, some participants suggested that the game playing time and scoring system should be linked to their quitting journey. For example, one participant suggested that players could score more points if they played the game when craving cigarettes.

Because of concerns over cravings when seeing smoking-related images in the game, some suggested changing the premise of the game to enable players to “destroy” smoking stimuli instead of neutral stimuli.

Drawing on their experiences with mobile games, participants suggested adding different levels to the game with varying difficulty (eg, increasing speed and time limits), competitive elements (eg, leaderboards), and the ability to invite friends to play the game as in multiplayer games.

Other participants suggested adding other games to our cessation mobile app (eg, mindfulness games) to help individuals who smoke deal with smoking antecedents such as stress.

This is one of the first studies, to our knowledge, that examines perceptions of a gamified ABM module within a multicomponent smoking cessation mobile app targeting individuals with low socioeconomic status who smoke [27,30]. As a novel concept, participants were willing to try the game provided there was evidence of its efficacy as a tool in aiding them to quit. However, only a few grasped the true purpose of the game as an attentional bias training tool, whereas most viewed the game as a distraction tool during cravings. Skepticism about the usefulness of the game was attributed to the possibility of triggering cravings upon the presentation of smoking cue imagery and the availability of other mobile games, more sophisticated in nature, that could distract individuals who smoke during cravings. Opinions were mixed concerning the game’s enjoyability but were largely positive for ease of use. Our findings underscore that the primary challenge to maximizing user acceptance is to communicate the purpose and benefit of an ABM game and to improve the game’s functionality and design elements to meet users’ expectations and withstand competition from commercially available entertainment games. This is particularly true if an ABM game were to engage users for an extended time that is needed to allow for the delivery of multiple training sessions to positively affect cessation. Beyond our ABM smoking cues game, these results could inform the development and user acceptability of gamified self-help ABM and CBM interventions for other substance use disorders and tobacco products. This is particularly relevant as CBM- and ABM-based mobile games and apps have generally had mixed effects on behavioral change [30]. Finally, our results highlight the importance of user-centered approaches for intervention development. User feedback will help us further improve our ABM mobile game by incorporating participants’ suggestions prior to evaluating its usability and efficacy in field studies.

Our approach to embedding an ABM mobile game within a multicomponent app for smoking cessation is novel. Past attempts to translate ABM techniques to mobile platforms as games have been limited to substance use disorders other than tobacco dependency (eg, opioid use) [49]. Few smartphone-based ABM interventions for smoking cessation have been standalone interventions. Furthermore, these interventions usually deliver the traditional visual probe test on a mobile platform rather than transforming the technique to embrace smartphones as a medium [27-29]. Our approach would allow for user-initiation of the training module at their own will and in their natural environments, eliminating one potential cause of the underperformance of traditional ABM interventions. This could be particularly beneficial to populations that cannot avoid or reduce exposure to smoking-related imagery per clinical guidelines for treating tobacco dependency [11]. Given the scarcity of evidence on ABM mobile games for smoking cessation in general and with disadvantaged groups in particular, research is needed to examine their efficacy in reducing smoking and its antecedents (eg, cravings and reactivity to smoking cues).

Additional strengths of our game-based approach to smoking cues ABM include the accessibility and familiarity of smartphones and mobile games, particularly among young and middle-aged adults [23,50]. Indeed, most participants in our study agreed the game was easy to use and referenced prior experiences with mobile games. The downside to this approach is that it excludes by default individuals with certain predispositions, thus limiting the efficacy of smoking cessation intervention with an ABM component. For example, less digitally literate and older individuals might not be willing or able to use mobile health games [50]. More importantly, high user expectations of ABM game functionality and design present a fundamental challenge. Mobile health games must compete with commercially available, professionally developed entertainment games for users’ attention and engagement, especially as most mobile apps rapidly lose users after their first use [51]. While smoking cessation interventions are on average 6 weeks long, sustaining engagement during this period is a challenge. However, limited evidence suggests that the use of a game was positively associated with the number of smoking cessation app sessions and time spent on the app [52,53].

Indeed, enjoyability emerged as a main factor in user acceptance of the ABM game. Past work has established the important role of enjoyability in users’ adoption of new technologies as well as its impact on user engagement for serious games [35,54]. User engagement, in turn, has been associated with abstinence in pilot testing [55]. Accordingly, negative sentiments in discussing the game’s level of fun and enjoyment are concerning. Some remarked that the game appeared outdated, which may be on account of seeing an early prototype or on account of having experiences with increasingly sophisticated mobile games. Partnerships between academics and mobile game developers may be a beneficial approach to elevate health games [52]. Our game’s level of enjoyability could be improved by implementing user suggestions, which primarily focused on adding new elements or approaches commonplace in other games. Several of the suggested additions (eg, leaderboards, levels, and rewards) have been desired by individuals who smoke in other work and successfully implemented in both smoking cessation and ABM apps [30,56-58].

Our findings show that communicating the intent and benefit of an ABM is a primary requisite to user acceptance. Skepticism of the game’s performance stemmed primarily from concerns that the presentation of cue imagery could trigger cravings and from perceptions of the game as a distraction tool, in which case preferable alternatives are readily available. Suggestions to redesign the game to allow users to “destroy” smoking-related images as well as the common use of mobile apps and games for distractions echo our participants’ perceptions of the ABM game [57,59]. Previous research on an ABM app for substance use disorder documented similar results [56]. Moreover, cravings in response to cues are well-documented among individuals who smoke and provide credence to participants’ concerns [60]. However, the game requires focusing one’s attention on neutral images (ie, fruit slices) to score. For the ABM to be effective, it is necessary to use stimuli that are relevant to and associated with users’ attentional biases [61]. Past qualitative research has identified the need for users to understand the rationale behind CBM and ABM [62,63], which is critical for engagement [54]. Thus, it is imperative to identify appropriate means to communicate the rationale behind the game to users and its potential benefit in helping them quit. This could potentially be accomplished by providing a written rationale, written instructions, or informational tooltips, as some participants suggested. Future research is needed to field test the effects of an ABM game on intended (eg, modifying attentional biases) and unintended (eg, triggering cravings) outcomes.

This study has several strengths, including the recruitment of a diverse sample of participants with low socioeconomic status with roughly ≤50% of the participants from one race or sex. Furthermore, our qualitative methodology provides a wealth of information on user perceptions of the ABM mobile game under investigation. Nonetheless, this study has several limitations. First, although the moderator assured participants that there were no right or wrong answers, our results could have been influenced by social desirability bias [64]. Second, our focus groups were held during the COVID-19 pandemic, which resulted in participants seeing web-based mock pages of the game rather than using it had the discussions been conducted in person. Participants were shown an early prototype of the smoking cues mobile game, which may have affected their perceptions. Our results may have been impacted by the high exposure to and familiarity with mobile games that young adults have, which may not be representative of all individuals who smoke [50]. However, we will field test our smoking cessation app to determine whether the ABM game significantly contributes to quit rates among adults older than 18 years and to shed light on their experiences using the game in their natural environment, which will complement this formative work. Last, although the intercoder agreement for technology acceptance and sentiment domains was low, all inconsistencies were resolved through discussions between MW, LT, and SEL.

This study is one of the first to characterize users’ perceptions of an ABM mobile game for smoking cessation. Participants had mixed perceptions of the game’s usefulness and enjoyability but were willing to use it provided evidence of its efficacy in aiding smoking cessation. These results will help guide the development and improvement of the game to support adults with low socioeconomic status who smoke. Further work should explore the real-life use and effects of gamified ABM techniques in general and for smoking cessation in particular both independently and as part of multicomponent behavioral interventions.