Despite decades of declining smoking prevalence since the landmark publication of the 1964 Surgeon General’s report, smoking continues to be the leading cause of preventable death in the United States (HHS, 2020). It is estimated that approximately 480,000 deaths per year are attributable to cigarette smoking and secondhand smoke exposure (HHS, 2020; Rostron, 2013). Over a third of those deaths are people with mental illnesses, who tend to have a high prevalence of smoking relative to the general population (Prochaska et al., 2017). According to the Surgeon General’s Office, the economic toll of cigarette smoking exceeds $170 billion in the United States each year (HHS, 2020).

Most individuals who smoke are interested in quitting (Babb et al., 2017). Still, fewer than one-third of individuals who smoke cigarettes attempt to quit using FDA-approved smoking cessation pharmacotherapies (e.g., nicotine replacement therapy and prescription medications) (Babb et al., 2017; HHS, 2020). Notably, the most common quitting method is ‘cold turkey’ (i.e., unaided quit attempts), even though fewer than one in 10 individuals who smoke achieve smoking cessation (i.e., having discontinued smoking for one day or longer with the intention of quitting; (Babb et al., 2017; HHS, 2020). This is worrisome because smoking cessation confers health benefits to all individuals who smoke, regardless of age (HHS, 2020). Furthermore, the underutilization of FDA-approved smoking cessation pharmacotherapies is inconsistent with the US Public Health Service’s Clinical Practice Guideline on Treating Tobacco Use and Dependence, which recommends improving smoking cessation rates via evidence-based treatment modalities (Fiore et al., 2008; HHS, 2020). While substantial evidence suggests these smoking cessation medications are associated with increasing quit rates in individuals who smoke, bolstering their utilization remains an unrealized public health goal (Cox et al., 2011; HHS, 2020; Schlam & Baker, 2013). From a health equity standpoint, this is also concerning given that the rates of smoking cessation pharmacotherapy utilization are even lower among people from marginalized populations, such as African Americans, Hispanic/Latinx Americans, and American Indians/Alaska Natives relative to White Americans due to factors such as financial barriers and medical mistrust (Fu et al., 2007; Fu et al., 2008; Fu et al., 2014; Fu et al., 2005; Scharff et al., 2010). Consequently, these populations experience disproportionately higher rates of smoking-related morbidity and mortality relative to White Americans (Agaku et al., 2020; Braveman et al., 2010; Gone & Trimble, 2012; HHS, 2020; Martell et al., 2016; Mowery et al., 2015).

Smoking cessation apps (SCAs) are computer programs available on mobile devices, such as tablets and smartphones, designed to aid in smoking cessation (Abroms et al., 2013). SCAs are heterogeneous, variously incorporating features such as cigarette trackers, geofencing, educational materials, behavioral therapies, and gamification to maintain smoking cessation progress (Abroms et al., 2013; Abroms et al., 2011; Portelli et al., 2022; Robinson et al., 2020; Seo et al., 2022). Specifically, 29 % of SCAs are only informational (i.e., similar to an electronic book or e-Book), 30 % are multifunctional (i.e., they have assistive features to facilitate smoking cessation), and 42 % are a combination of informational and multifunctional (Seo et al., 2022).

In the past decade, there has been an influx of SCAs in the marketplace (Abroms et al., 2013; Abroms et al., 2011; Seo et al., 2022). As of 2020, over one hundred SCAs were available in the Google Play and iPhone App Stores (Seo et al., 2022); 62 % were free, 24 % were initially free with in-app purchases for additional features, and 14 % required upfront payment (Seo et al., 2022). In addition to the commercial options, the National Cancer Institute created two free SCAs (QuitGuide and quitSTART) integrated with their Smokefree.gov initiative (Prutzman et al., 2021). SCAs are a potentially accessible method of disseminating smoking cessation interventions at the population level, given that American smartphone usage is increasingly commonplace (Heffner & Mull, 2017; Smith et al., 2015).

Feasibility/pilot studies of specific SCAs have shown high ratings in usability, acceptability, engagement, and likability (Bricker et al., 2017; Gowarty et al., 2021; Iacoviello et al., 2017; Marler et al., 2019). However, randomized clinical trial (RCT) data regarding the association between SCA utilization and smoking abstinence are mixed. Some studies noted higher rates of smoking abstinence relative to various comparators (e.g., usual care, self-help booklets, web-based interventions, other apps, smoking cessation counseling, or placebos that lack functionality) (BinDhim et al., 2018; Bricker et al., 2022; Bricker et al., 2020; Carrasco-Hernandez et al., 2020; Houston et al., 2022; Masaki et al., 2020; Webb et al., 2022), whereas other studies did not find any significant differences (Affret et al., 2020; Baskerville et al., 2018; Etter & Khazaal, 2022; Garrison et al., 2020). A recent systematic review/meta-analysis of 9 RCTs conducted between 2017 and 2023, which were described as high-quality by Guo and colleagues (2023), did not support the effectiveness of SCAs as standalone interventions. Still, it noted that their effectiveness was augmented when used with pharmacotherapies (Guo et al., 2023). Specifically, participants using SCAs with pharmacotherapy (3 studies, N = 1342) reported greater rates of smoking abstinence relative to participants who only received pharmacotherapy. SCAs have the potential to be utilized as a population-level intervention, but a lack of population-level studies limits the current literature. Population-level studies are important to consider given that RCTs tend to be limited by samples that are not always generalizable due stringent eligibility criteria and other barriers that prevent underrepresented populations from enrolling in the studies (King et al., 2011; Scharff et al., 2010; Webb Hooper et al., 2019). This limitation must be addressed to more clearly understand if this type of intervention is applicable to the general population, especially people from marginalized populations. Furthermore, many apps available in the iPhone and Android marketplace remained unstudied.

Using population-level data, the present study aimed to examine relationships among SCA use initiation, pharmacotherapy utilization, and quit attempts longitudinally in individuals who regularly smoke cigarettes. The specific aims are to longitudinally examine, among individuals who regularly smoke cigarettes, the relationship between 1) SCA use initiation and NRT utilization, 2) SCA use initiation and prescription medication utilization, and 3) SCA use initiation and quit attempts.

The sample was predominantly female (56 %), 25 to 34 years and 55 or older (24 % & 25 % respectively), non-Hispanic White (71 %), with incomes of $10,000 to $24,999 and $25,000 to $49,999 (27 % & 25 %), and high nicotine dependence (62 %). Overall, 4 % of the sample reported the initiation of SCA use. The descriptive characteristics of this sample are shown in Table 1.

In the present study, we used data from the PATH Study to longitudinally examine the relationship between SCA use initiation, NRT utilization, prescription medication utilization, and quit attempts in individuals who regularly smoked cigarettes and planned to quit within a year. The results showed an association between past 12-month SCA use initiation and prescription medication utilization at follow-up. However, we did not observe this relationship for past 12-month NRT utilization. Additionally, we observed an association between past 12-month SCA use initiation and past 12-month quit attempts at follow-up. Notably, these findings persisted even when controlling for socioeconomic status and internalizing mental health factors, both of which have a significant association with smoking. This study extends the literature by providing a longitudinal population-level analysis of SCAs.

Inconsistent with previous SCA studies, we did not observe an association between SCA use initiation and NRT utilization (Masaki et al., 2020; Webb et al., 2022). This is notable given that, unlike other studies focusing exclusively on the nicotine patch and gum, our study included all NRT modalities (Masaki et al., 2020; Webb et al., 2022). These discrepant findings may be attributable to whether some SCAs emphasized other treatment modalities instead of NRT (Abroms et al., 2013). Furthermore, some participants may have been skeptical of the effectiveness of NRT relative to prescription medications since the former contains nicotine, especially some racial/ethnic minority participants (Carpenter et al., 2011; Fu et al., 2007; Fu et al., 2008; Fu et al., 2014; Fu et al., 2005). This may partly explain why we observed an association between SCA use initiation and prescription medication utilization. These findings are consistent with the results of a previous longitudinal SCA study that examined the relationship between an SCA and varenicline utilization to see if it predicted smoking cessation (Carrasco-Hernandez et al., 2020). SCA users’ prescription medication utilization may be related to the informational components of their respective apps, which dispel misconceptions about the commercially available medications. While SCAs cannot address the financial barriers that often prevent prescription medication utilization, they may contribute to less negative attitudes about their safety and effectiveness (HHS, 2020; Zeng et al., 2011). Given this potential explanation for the findings, SCAs may be a novel intervention for increasing positive attitudes about pharmacotherapies among marginalized populations, who have tended to underutilize them (Fu et al., 2007; Fu et al., 2008; Fu et al., 2005; HHS, 2020). Furthermore, the SCAs can be utilized to provide information about pharmacotherapies that are available for free via state-sponsored quitlines (Prutzman et al., 2021). However, there remains the need to study attitudes among marginalized populations regarding the efficacy and safety of SCAs as a smoking cessation intervention.

Overall, these findings partially correspond with the recent meta-analysis, which noted that combining SCAs and pharmacotherapies (i.e., NRT and prescription medications) may be more effective than standalone SCA utilization (Guo et al., 2023). Specifically, SCA users are likely to benefit from app features such as providing accurate and accessible information about pharmacotherapies, which leads to their utilization (Heffner et al., 2015; Hoeppner et al., 2015). Given that pharmacotherapies are heavily underutilized by individuals who smoke, SCAs may be a population-level intervention that merits further investigation. Furthermore, the informational components of SCAs may also be related to increased pharmacotherapy utilization by minoritized groups that have historically underutilized them due to systemic barriers as well as misconceptions about their effectiveness, but it may not be effectively addressing NRT misconceptions (Avila et al., 2022; Fagan et al., 2004; Fu et al., 2007; HHS, 2020).

Some studies have shown that smoking abstinence and quit attempts are linked to SCA utilization (Bricker et al., 2020; Carrasco-Hernandez et al., 2020; Danaher et al., 2019; Houston et al., 2022; Masaki et al., 2020; Webb et al., 2022). However, systematic review/meta-analysis research suggest this relationship may depend on the inclusion of pharmacotherapies (Guo et al., 2023). Our study, however, supports that notion of using SCAs as a standalone intervention is associated with quit attempts in individuals who regularly smoked cigarettes, regardless of their race/ethnicity, sex, or mental health status. Furthermore, earlier research has noted that using an acceptance and commitment therapy-based SCA was associated with higher odds of quitting smoking relative to a comparator app, even in marginalized groups (e.g., Black, Hispanic/Latinx, and low-income adults) (Hayes et al., 2009; Santiago-Torres, Mull, Sullivan, Kendzor, et al., 2022; Santiago-Torres, Mull, Sullivan, Zvolensky, et al., 2022; Santiago-Torres et al., 2022). However, some studies have not found differences in smoking abstinence rates or quit attempts between an SCA group and a control group (Affret et al., 2020; Baskerville et al., 2018; Etter & Khazaal, 2022; Garrison et al., 2020). The discrepancies in results could be due to differences in the study’s sample characteristics, SCA type examined, or study duration. Our study used data from a nationally representative cohort study, while three of the four studies just mentioned (Baskerville et al., 2018; Etter & Khazaal, 2022; Garrison et al., 2020) did not use nationally representative samples. The study that included a nationally representative sample (Affret et al., 2020) was conducted in France, which has a different tobacco regulatory environment than the United States. Additionally, the four other studies (Affret et al., 2020; Baskerville et al., 2018; Etter & Khazaal, 2022; Garrison et al., 2020) each focused on a specific type of SCA, whereas our study examined all types of SCAs. Finally, our study had a longer timeframe than these four RCT studies due to the prospective cohort design of the PATH Study (Affret et al., 2020; Baskerville et al., 2018; Etter & Khazaal, 2022; Garrison et al., 2020; Hyland et al., 2017).

Our study had several noteworthy strengths. To our knowledge, it is the first study to investigate use of SCAs in a nationally representative sample from the United States. Many studies on SCAs tend to have small sample sizes of minoritized participants, whereas our study had a substantial number of such participants. Moreover, the study was bolstered by a longitudinal design that spanned five years. Another aspect of our study that differed from the existing literature is that we evaluated SCAs more broadly instead of focusing on a specific type, making our findings more applicable to the real-world experiences of SCA users.

Along with the strengths, we also note some limitations of the present study. We measured past 12-month SCA use initiation without posing any specific questions about the type of SCAs used by the participants. Therefore, we cannot determine what types of SCAs, such as cognitive-behavioral therapy-based, acceptance and commitment therapy-based, and others, are driving the relationship between the primary outcomes. This limitation is notable because many commercially available SCAs are not supported by scientific data, thus raising concerns about their efficacy. (Abroms et al., 2013; Heffner et al., 2015). Additionally, there is the possibility that some participants may be unknowingly using pro-tobacco apps developed by the tobacco industry (BinDhim et al., 2014, 2015). Another limitation of our study is that it only focused on individuals who regularly smoked cigarettes and planned to quit within a year. As such, it may not generalize to people who smoke cigarettes infrequently and are less certain about their quit plans. Our pharmacotherapy treatment utilization variable included Wellbutrin, Zyban, and Bupropion, which are medications used for the treatment of smoking as well as treatment of depression. This variable also included Chantix, which is exclusively used for smoking cessation. Given that depression and smoking are correlated, it may be difficult to tease apart who is taking the medication to treat depression vs. smoking cessation. However, this issue is partly addressed by our sample being limited to people who indicated a plan to quit smoking/tobacco products within a year. Lastly, we did not examine sex differences in relation to the outcome variables. These analyses may have yielded notable differences given that prior research has noted that men and women have disparate quit rates depending on what type of smoking cessation treatment is utilized (Smith, Bessette, et al., 2016; Smith, Weinberger, et al., 2016).

In our study, we longitudinally examined the relationship between SCA use initiation, NRT/prescription medication utilization, and quit attempts using a nationally representative sample of individuals who regularly smoked and planned to quit smoking within a year. Our findings suggest that SCA use initiation is associated with prescription medication utilization as well as quit attempts. Future studies should examine which types of SCAs drive this relationship and which sociodemographic factors relate to their use, and if they relate to quitting smoking. Qualitative research is also needed to better understand the relationship between SCA use initiation and prescription medication utilization.