Introduction

Smoking cessation mobile applications are promising tools for supporting cessation [1]. They are comprised of single or multiple interventional features that are designed to affect behavioral change. These features are considered “active ingredients” or behavioral change techniques (BCTs) that intervene on the mechanisms of action or causal pathways of smoking behavior [2,3]. BCTs are rooted in various theoretical frameworks (e.g., social cognitive theory) [4,5]. The translation of the theoretical underpinnings of BCTs to digital platforms involves a design process, often expert-led, that results in different informational structure and aesthetic display of features across smoking cessation mobile apps. The number, type, and aesthetic quality of features included in smoking cessation apps vary. However, several features are foundational given how commonly used they are. These features include self-monitoring, tailored feedback and support, and educational content [6,7].

Self-monitoring is commonly utilized in digital behavioral health change interventions [8–12]. It is the periodic self-recording or passive capturing of a behavior and its outcomes to enhance self-management [13]. In adaptive health interventions, self-monitoring is aimed at providing continuous feedback on health goals and at triggering support whenever and wherever support is needed [14]. Underlying self-monitoring is a self-regulatory mechanism that directs purposeful behaviors [15]. Self-monitoring is a BCT recommended for smoking cessation interventions to elicit cessation [16]. In smoking cessation apps, self-monitoring is often either self-reported (e.g., recording cigarettes smoked) [17] or system-generated (e.g., calculating money saved by not smoking) [18]. Smoking-related physiological parameters (e.g. exhaled carbon monoxide levels) are also candidates for self-monitoring [10] although biomonitoring remains limited in existing cessation apps as is evident by the volume of calculator and calendar/tracker smoking cessation apps versus those that incorporate physiological monitoring [19].

Tailoring, another staple of health interventions, refers to “any combination of information and behavior change strategies intended to reach one specific person, based on characteristics that are unique to that person, related to the outcome of interest, and derived from an individual assessment”[20]. Tailoring affords mobile interventions the ability to cater feedback and support to each individual and their health needs [21]. Evidence shows that tailored health interventions can be more effective than their generalized, one-size-fits-all counterparts [22–24]. Virtually all theoretical frameworks state that behavioral change is dependent on various factors that vary across individuals (e.g., attitudes, skills) and within individuals (e.g., their readiness to adopt or cease a behavior over time) [25]. Specific to smoking cessation, tailored feedback and support (e.g., provide feedback on behavior) are recommended BCTs that facilitate cessation [16]. Smoking cessation apps can approach tailoring in various ways [26]–from basic (e.g., allowing users to set a quit date and list personal reasons for quitting) [27] to advanced tailoring of treatment timing, type, and dose (e.g., just-in-time adaptive stress management strategies to mitigate stress-induced smoking lapses) [28].

Educational content is the backbone of health interventions regardless of the intervention delivery mode (e.g., print, digital). The provision of information is rooted in theoretical frameworks such as the knowledge-attitude-behavior (KAB) model that underscores the importance of relevant knowledge in shaping attitudes, which in turn affects behaviors [29]. Indeed, the inclusion of detailed information in mobile apps has proven to positively affect health behavior [12]. Consistent with BCTs and the clinical guidelines for treating tobacco use and dependence [16,30], cessation apps often include information on tobacco dependence, consequences, and treatment resources (e.g., quitlines, medication) [6,31].

Self-monitoring, tailored feedback, and educational content are perceived as important features among users of smoking cessation apps to achieve cessation and positively influence use [32,33]. However, use of interventional features and their effects on health outcomes differ across populations. Evidence shows vulnerable (vs. affluent) populations have a lower likelihood of using health apps generally or app features specifically [34–36]. For example, a study found that those with lower incomes (vs. those with more economic resources) were less willing to use self-monitoring in diet and fitness mobile apps [35]. Other studies found that provision of information in health interventions had a particular benefit for people who are socioeconomically disadvantaged, possibly because there are greater knowledge gaps among such groups [37]. Studies show that interventions with fewer interventional features were more effective among populations with lower incomes [37], whereas number of features has been associated with better [38] or no difference [39] in outcomes among general populations. Finally, design concepts of health apps, which relate to multiple elements such content input and delivery, navigation, information architecture, typography, and aesthetics of the app, are a necessary determinant of users’ engagement and their seamless interactions with an app [40,41]. However, studies show that some health apps are not designed to accommodate individuals with certain user profiles (e.g., limited health and digital literacy), which can negatively impact user engagement [42].

Capturing users’ perceptions of smoking cessation apps, their features, and their design concepts can inform our understanding of the differential use of interventional features. This is particularly important among high priority populations for smoking cessation efforts. Lower socioeconomic status (SES) is linked with higher prevalence of cigarette smoking across age, race/ethnicity, and U.S. census region [43] People who smoke and are socioeconomically disadvantaged tend to underutilize traditional smoking treatments for several reasons (e.g., lack of access [44], feelings of guilt and shame that accompany admissions of nicotine addiction [45]) but have positive attitudes toward and show interest in mobile app-based cessation support [46,47]. However, there are few mobile interventions that target individuals with low SES [6]. Technology-based interventions targeting disadvantaged populations are also characterized by having low methodological quality and small effects on cessation [48], suggesting a need to engage this target population when developing interventions. Previous studies have cataloged types of smoking cessation apps and features [7,17,19,31]; their alignment with evidence-based nicotine treatment guidelines [19,49,50]; the associations between features’ presence and user engagement and/or quit success [27,51]; and smokers’ general opinions of feature importance/preference [32,33]. There is limited qualitative work on low SES smokers’ perceptions of cessation apps and their features despite the importance of user acceptance as a determinant of use [52]. This qualitative study aimed to gauge perceptions of self-monitoring, tailored feedback and support, and educational content as presented in two smoking cessation apps, QuitGuide, a benchmark app, and Quit Journey, a newly developed app. The study will inform our understanding of the acceptance of individuals with low SES who smoke of these common interventional features and their design as part of the screening and selection of intervention components that are core to the preparation phase of the multi-phase optimization strategy (MOST) for the development and optimization of Quit Journey, a smoking cessation intervention targeting individuals of low SES who smoke [53,54].

Methods

Analysis

Following a deductive thematic analysis approach [68], we identified initial codes in an exploratory phase of the analysis. Then, we identified salient themes using a codebook that included (a) technology acceptance themes based on UTAUT2 (i.e., performance expectancy, effort expectancy, hedonic motivation, facilitating conditions, social influence) [52,67], (b) sentiment (i.e., positive, neutral, negative) (c) interventional feature (i.e., self-monitoring, tailored feedback and support, educational content), (d) design concept, and (e) suggestions. We also identified relevant themes regarding each app overall, app name, and app landing page and were open to emerging themes (e.g., intent/willingness to use a feature).

LT and MW independently coded the transcripts using ATLAS.ti (version 8, ATLAS.ti Scientific Software Development GmbH). Following a multi-coding approach, themes within a semantic domain were mutually exclusive but themes from other semantic domains could be applied to the same quote. Inter-coder agreement was calculated via ATLAS.ti. The Krippendorf c-alpha binary value was 0.937, indicating strong agreement amongst coders when differentiating relevant from irrelevant content [69]. The Krippendorf cu-alpha values, which indicate the reliability of each semantic domain [69], appear in Table 1 alongside themes and their definitions. Discrepancies in coding were resolved through discussions amongst LT, MW, and SEL.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Semantic domains, theme definitions, and inter-coder agreement [64].

https://doi.org/10.1371/journal.pdig.0000658.t001

Results

Participants were 52.63% female, 42.10% non-Hispanic White, 28.94% non-Hispanic Black, and 15.78% Hispanic adults. Most participants had some college education (47.36%), whereas 26.31% were high school graduates and 10.52% had an associate degree. Sample characteristics appear in Table 2 and detailed participant information in S2 Table. Select illustrative quotes appear below with participant ID attribution and the referenced app (QG for QuitGuide, QJ for Quit Journey).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Participant characteristics (N = 38) [64].

https://doi.org/10.1371/journal.pdig.0000658.t002

Perceptions of smoking cessation apps interventional features

Across both apps, 312 quotes were feature-related, of which roughly half focused on self-monitoring (n = 152, 48.71%), followed by tailored feedback and support (n = 112, 35.89%), and educational content (n = 48, 15.38%). Discussions focused primarily on feature performance expectancy that ranged from 39.13% to 70.00% of the extracted quotes for each feature. The remaining technology acceptance themes were discussed to a lesser extent: effort expectancy (16.98%), hedonic motivation (16.34%), and facilitating conditions (1.60%). Quotes that captured participants’ sentiment but did not reflect a technology acceptance theme were coded as “not applicable” and constituted 17.62% of quotes. Distribution of quotes by technology acceptance and sentiment appears in Table 3 for QuitGuide and in Table 4 for Quit Journey.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Distribution of number of quotes by technology acceptance and sentiment toward three interventional features in QuitGuide smoking cessation mobile app.

https://doi.org/10.1371/journal.pdig.0000658.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Distribution of number of quotes by technology acceptance and sentiment toward three interventional features in Quit Journey smoking cessation mobile app.

https://doi.org/10.1371/journal.pdig.0000658.t004

Self-monitoring. Participants had an overall positive sentiment regarding the usefulness of self-monitoring features in helping them quit smoking (QuitGuide: 86.27%, Quit Journey: 100%) such as monitoring aspects of smoking behavior (e.g., slips) and its antecedents (e.g., mood). (S3 Table).

I really like the mood [tracking] because it’s like every day my mood is up and down, and some days I want a cigarette more than others. So I think this … mood scale is really nice and then you can also see how … you felt that day and … see what you can change as far as your mood and you know to help you not want a cigarette. (P25, QJ)

The usefulness of self-monitoring features was owed to the reliability of mobile applications, which allow users to identify behavioral patterns over time to stay motivated and accountable.

I think it would be pretty good … because … keeping track of … what went wrong and … why you went back to smoking or … What is making you smoke is … hard to remember … so an app like this would help a lot. Just keeping track of everything. (P13, QG)

However, participants remarked that self-monitoring could induce unintended negative reactions when users are unable to maintain positive behavioral patterns over time or do not receive appropriate support when reporting lapses.

I feel like if … I say [that] I was not smoking for a week and I’d been pressing the … smoke free [button] all week and then I had to press I slipped and it was like … “you’re starting at zero again” and there was no words of encouragement, I might just go “I’m leaving this app, it made me feel bad.” (P22, QG)

Tailored feedback and support. Participants expressed an overall positive sentiment toward usefulness of tailored features of smoking cessation apps generally (QuitGuide: 82.75%, Quit Journey: 84.61%) and toward pre-specified time- and location-based tailored support specifically (S4 Table).

I think [personalization would be helpful] cause no two people are the same. So, what works for one person, somebody else might need a little adjustment to it for it to work for them or be able to be a better use for them. (P04, QJ)

I like the time of day [notifications] the most. The location [notifications are] nice, and I like that … I like the time [notifications] because certain times make me want to kind of smoke … so [it] would be nice to be able to set that up and have reminders not to, or things that I could do during those times. (P13, QG)

Skepticism about the usefulness of tailored feedback and support focused on the need for active input from users, which could be “too much.” Additionally, participants questioned the usefulness of notifications given the addictive nature of smoking and the unpredictable timing of cravings.

… I smoke because it’s like once [a] craving hits, it hits … and then I just pull out a cigarette … If my phone vibrates and reminds me to not smoke [a] cigarette, it may or may not help me put that …cigarette down, because I already have that cigarette in my hand and I’m already addicted to it. (P09, QG)

I think the notifications [aren’t] that accurate, because you can always have cravings, different places, different times. So, it’s not really a specific time, you have a craving or a specific place, like you can see someone smoking or watch TV and you … have a craving. So, the notifications … doesn’t really do much. (P25, QJ)

Educational content. Participants thought the inclusion of cessation information in both apps was useful with 100.00% of QuitGuide quotes and 58.82% of Quit Journey quotes being positive in sentiment (S5 Table). This sentiment was owed primarily to availability of this content on demand.

I think it’s useful because the app is primarily just for smoking. Whereas if you go like on Google or something you can search [for information] and they give you all these other resources … This is like your own personal journey to … to quit smoking. So, I think it will be helpful. (P16, QJ)

Negative sentiment about the cessation information’s usefulness centered on the substance of the information presented, which was described as elementary and repetitive particularly for established smokers.

I honestly think this [cessation information] would be good for … a beginner person that doesn’t really know all these things, because then they can learn. But most of this … is stuff I’ve already read and seen millions of times … The idea of going back to this one section, every time you need help, is kinda slim because … all you’re doing is reading the same thing over and over. (P28, QJ)

Other technology acceptance themes. Beyond feature performance expectancies, participants generally perceived all features to be easy to use.

I feel like [tracking] could become … part of a daily routine or something like you wake up in the morning, use it once, go again in the evening, and use it again. Like, it doesn’t look too hard to [use] … you don’t really have to … go out of your way to record. So, you can pretty much do it anytime. It’s something … that you can use daily. (P30, QJ)

[The cessation information] seems pretty straightforward and easy to comprehend, find and use. (P10, QJ)

Although most participants agreed that use of features examined here was not engaging nor fun, some stated that enjoyment in using these features was contingent on certain circumstances (e.g., successful quitting) or user predispositions (e.g., having a competitive or analytical nature).

I think [personalization] might be fun if … you’re a couple of days into quitting smoking and feeling successful. But I think in the times you weren’t feeling successful, it wouldn’t be fun. (P22, QG)

When you actually are [seeing] the progress and you’re doing the steps or you can see kind of physically in front of you … how much you’ve cut down, I think then it kind of becomes fun … especially if you’re a competitive person [and] you like to see those results in front of you.. It kind of gives you confidence … So … I think it would be hard at first, because it’s just something you have to get used to. (P07, QG)

Participants identified some facilitating conditions to use the features examined here such as personal commitment to quitting smoking.

But I do think [tracking] would be pretty helpful as far as … it kind of depends on how committed you are to smoking. (P08, QG)

Discussion

This study established positive perceptions of three interventional features of mobile smoking cessation interventions as presented in two apps, Quit Journey and QuitGuide, among individuals with low SES who smoke. They perceived self-monitoring, tailored feedback and support, and educational content as both useful and easy to use features, two technology acceptance themes that have been documented to facilitate technology use [70]. Less emphasis was placed on other technology acceptance themes (e.g., social influence) except for individual predispositions that could facilitate or impede the use of app features [52,67]. Features were generally not perceived as fun or engaging, which could potentially be a barrier to sustained use [52,67]. While Quit Journey generally fared better than QuitGuide in its design concepts (S2 Note), participants made suggestions applicable to both apps such as increasing their interactivity, customization capabilities, and integration with other apps and devices. These recommendations align with users’ expectations of mobile health apps, which are important determinants when aiming to increase app use [32,64,71]. Indeed, health apps that fail to meet users’ expectations are perceived as less useful and possibly deleted shortly after they are downloaded [72]. Given the paucity of studies on users’ acceptance of the “active ingredients” of behavioral interventions, especially among vulnerable populations that smoke, this study fills an important gap in smoking cessation mobile intervention research. Alongside establishing the acceptability of Quit Journey’s components and informing its design concepts, implementing users’ recommendations could potentially improve user engagement, and consequently uptake and sustained use, of smoking cessation apps generally.

As part of MOST preparation phase, this study shows that acceptance of common interventional features of smoking cessation apps by individuals with low SES who smoke is promising [53,54]. They tend to underutilize traditional smoking treatments [44,45] but favor technology-based cessation support, including mobile apps [46,47]. Thus, mobile apps could provide an alternative cessation approach for this vulnerable population. However, there are limited feature-level analyses that have explored the linkage between the presence of certain features and health intervention effectiveness [73]. With evidence being mixed on specific features and number of features that improve health intervention outcomes [38,39,74], further research is needed to understand the effect of individual features on outcomes, especially among vulnerable populations [37]. Furthermore, the link between acceptance of interventional features and uptake of health apps in real-life warrants further investigation. For example, while there is limited data on vulnerable populations’ uptake of smoking cessation apps, they are less likely to use health apps in general and certain interventional features (e.g., self-monitoring) in particular compared to their more affluent counterparts [34–36]. This could potentially be attributed to individual-level factors (e.g., limited digital literacy) [75] or app-specific factors with health apps being too basic [50]. In planned pilot and optimization studies, we will assess the efficacy of Quit Journey and its components on cessation outcomes [53].

Participants’ suggestions centered on making the app features smarter such as by enabling tailored real-time support with minimal user input. Offering smoking cessation support for populations with low SES in real time is a promising tactic [24], will likely satisfy users’ high expectations for smartphone apps [64], and aligns with shifts toward just-in-time adaptive interventions that deliver personalized treatment whenever and wherever it is needed [76]. Personalized support with passive sensing and minimal user input could entice smokers to utilize cessation apps because repeated active tracking of data can be seen as tedious or boring [77]. More importantly, user-initiated support and recording of smoking episodes is often low [78]. A recent meta-regression study found that personalized interventions that relied on system-captured data were more effective than those that relied on user-reported data [79]. Furthermore, suggestions related to design concepts of both apps align with design principles for digital platforms [41]. For example, participants recommended that the educational content be presented in a less text-heavy format with the inclusion of appropriate imagery, videos, and icons. This corresponds with prior work finding that users prefer multiple (vs. single) modes of communicating information [32,40]. Preference for use of coherent color schemes and bright colors was also emphasized in our study (S2 Note), which further aligns with literature on what design components are important to app users [40] and accentuates the importance of both the health content and how it is presented. Collectively, these results emphasize that developers of smoking cessation apps (and health apps generally) must fully capitalize on the advancements in mobile technologies and include their target populations in the app design process using human-centered design methods to maximize engagement and sustained use [80].

The pervasiveness of mobile technologies and health apps contextualizes our results. As U.S. smartphone ownership rates are reaching saturation across multiple sociodemographic populations (including populations with low SES) [81], mobile apps are increasing in both availability and popularity with users. As of 2024 there are roughly five million apps available for download across leading app stores (e.g., Google Play, Apple App Store, Amazon Appstore) [82]. Furthermore, a national 2018 survey found that 49.24% of Americans had apps related to health and wellness downloaded on their smart device [83]. This explains the predominance of perceived usefulness of smoking cessation apps features as the primary technology acceptance theme discussed, where participants intuitively underscored the mechanisms by which each feature could serve as a tool to help them alter their smoking behavior. Additionally, consistent with recent evidence [84], none of the barriers commonly outlined in the literature to negatively affect use of mobile technologies among populations with low SES (e.g., privacy concerns, cost, low digital/health literacy) [75] emerged in our results. Conversely, per users’ suggestions, it was evident that participants had unmet expectations regarding how smoking cessation apps should look and perform. Specifically, most participants either explicitly said the features presented were not fun or stated that enjoyment was contingent on certain predispositions (e.g., being competitive). While these high expectations could be attributed to our participants being young adults, it is important that health apps employ different engagement strategies (e.g., gamification [85], reciprocity [86]) especially when the use of smoking cessation apps may be viewed as a reminder of one’s “addiction”. Prior work has shown that users often abandon health apps after minimal usage for reasons including lack of fun or interest [36,87]. Further research is needed to identify and examine the effects of engagement strategies, especially non-financial ones, on sustained use of smoking cessation apps among disadvantaged groups.

This study had several strengths and limitations. The study gauged perceptions of young adults with low SES who smoke on three interventional features as presented in two smoking cessation apps. Our qualitative approach to elicit feature-based perceptions yielded a wealth of information on features common to smoking cessation apps either as standalone interventions or as part of multi-component ones. These results are highly informative, whereby through a human-centered design approach, we ensured the involvement of our target audience in intervention development and evaluation processes of our new smoking cessation app, Quit Journey [80]. As a high priority population for smoking cessation efforts, the educational attainment requirement for participants recruited for this study was a strength. However, we relied solely on educational attainment as an indicator of SES to the exclusion of other indicators such as income and employment [61–63]. Furthermore, our participants were young adults who are often digitally literate [88] and who wanted to quit, thus were not representative of all people who smoke such as older individuals. All focus groups were conducted virtually, which occasionally introduced technical difficulties and could have dampened their perceptions of enjoyment associated with app use based on viewing screenshots of the apps’ pages. Because of scheduling issues, we had more focus groups for QuitGuide than Quit Journey (i.e., 8 vs. 4). The clearcut boundaries between features as presented in the moderation guide were arbitrary. As implemented in both smoking cessation apps examined here, features often overlap. For example, when a user self-records their mood, they receive a tailored support message depending on their mood. The moderation guide relied on general terms (e.g., tracking, personalization, cessation information) when prompting discussions about self-monitoring, tailored support and feedback, and educational content as they are more accessible to non-researchers. We applied the “suggestions” theme to users’ recommendations even if the suggestion was already implemented in the app. However, users focused on big-picture recommendations to improve smoking cessation apps rather than on recommendations to improve specific features. Finally, the intercoder agreement for sentiment was low. However, all inconsistencies were resolved through discussions.

Conclusions

Self-monitoring, tailored feedback and support, and educational content were perceived to be useful and easy to use features of smoking cessation interventions. Results inform the development of Quit Journey, a multi-component smoking cessation mobile application targeting individuals with low SES smokers who have higher-than-average smoking rates. To increase smoking cessation app acceptance, health professionals should aim to fully capitalize on the capabilities of mobile technologies and employ design elements that can fully meet users’ expectations.

Supporting information

Acknowledgments

ICF services for this manuscript were performed under the support of the National Institute on Minority Health and Health Disparities through contract #75N91019F00132 with the National Cancer Institute. The National Cancer Institute had no input on the development of Quit Journey or its features nor the study conceptualization, data analysis, or interpretation. The efforts of LT, MW, and SEL have been supported by the Division of Intramural Research of The National Institute on Minority Health and Health Disparities. Charmaine Chan, Kristyn Kamke, and Zahra Ansari reviewed the transcripts against audio files. Rahaf Zaza extracted quotes in an exploratory phase of the analysis. The authors thank Yvonne Prutzman and Meredith Grady, co-leads of the National Cancer Institute’s Smokefree.gov Initiative (SFGI), for providing feedback on the manuscript draft regarding SFGI’s programmatic resources.

Disclaimer: The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Institute on Minority Health and Health Disparities, the National Cancer Institute, the National Institutes of Health, or the US Department of Health and Human Services.