Smoking Cessation Apps for People with Schizophrenia: How Feasible Are m-Health Approaches?

Abstract Background: The large health disparities among those diagnosed with schizophrenia urgently need to be addressed. These disparities are partially caused by adverse health behaviours such as smoking. Smoking cessation apps vary in efficacy across various populations, and there are concerns regarding the accessibility and usability of apps for people with schizophrenia. Objective: This review identifies and examines the feasibility of using apps for smoking cessation in people with schizophrenia. Methods: A non-systematic narrative literature review of smoking cessation apps for individuals with schizophrenia was performed. Results: Eight studies were included in this review. Conclusion: Smoking cessation apps can be acceptable and feasible, but may need to be tailored to the needs of people with schizophrenia. Key messages: (1) Smoking cessation apps could be acceptable and feasible for use in people with schizophrenia; (2) Lack of motivation was perceived as the main potential barrier with regard to people with schizophrenia engaging with smoking cessation apps; (3) In order to improve motivation of people diagnosed with schizophrenia, apps could include games, rewards, and/or social support; (4) Smoking cessation apps with a simple interface seem to be beneficial for this population; (5) Apps may need to be tailored to consider this population’s mental health needs.


Background
People diagnosed with schizophrenia have large disparities in physical health when compared with the general population, as demonstrated by the~15 years reduced life expectancy and increased risk of metabolic syndrome and cardiovascular disorders [1]. These disparities are partially due to side effects of antipsychotic drugs and reduced provision of adequate physical healthcare, alongside adverse health behaviours such as physical inactivity, poor diet and smoking.
While the use of digital health interventions in schizophrenia remains under-researched overall, smoking cessation appears to have gained substantial research interest -with numerous studies already investigating the feasibility and acceptability of using smoking cessation apps in this population. This increased attention is arguably warranted given that people with schizophrenia are more likely to be heavy smokers than the general population, but also more likely to die as a result of tobacco use [12]. Strikingly, in the United States approximately half of all recorded smoking-related premature deaths are attributable to people with severe mental illness (SMI) [13]. Further, around 70% of all deaths are due to physical health conditions such as cardiovascular disease, cancer, and respiratory disease [14]. While motivated to quit smoking, people with schizophrenia face additional barriers to cessation, for example, smoking to cope/manage their mental health symptoms (improving cognitive dysfunction) and side effects [13,15], reducing their confidence in their ability to quit. Therefore, interventions designed for the general population may not be as effective or suitable for this particular group [14][15][16][17]. Additionally, concerns with the accessibility and usability of digital health interventions for SMI populations have been raised [16][17][18][19].
This non-systematic narrative review aims to summarise the existing literature on interventions using smoking cessation apps for individuals with schizophrenia. To do this, evidence from studies in mixed-diagnostic samples of people with schizophrenia in psychiatric care settings was considered. Specifically, this review sought to (i) identify digital smoking cessation techniques that have been used in people with schizophrenia, (ii) examine the feasibility of using apps for smoking cessation in people with schizophrenia, and (iii) explore the potential effects on smoking-related behaviours and mental health.

Methods
To identify the relevant studies for this non-systematic narrative review, literature searches were conducted in January 2022 using keyword search algorithms containing terms relevant to severe mental illness, health behaviour change and digital technologies. The searches were conducted of multiple academic databases, including the Cochrane Central Register of Controlled Trials along with the Health Technology Assessment, Allied and Complementary Medicine, APA PsycInfo, Embase, and MEDLINE(R) databases. Additional searches of Google Scholar and relevant publications' citation lists were also conducted.
From the search results, English language articles reporting on any experimental/ interventional studies examining the feasibility, usability or effects of digital smoking cessation in SMI populations were sought out. This included randomised control trials (RCTs), non-RCTs, quasi-experimental studies, and qualitative research. Of all the relevant studies included in this review, critical information on the studies' sample (including size, mean age, diagnoses and demographics), intervention (including the delivery platform, length of intervention, control/comparator conditions) and findings (be it usage data, outcomes/effects or qualitative feedback) were extracted. Studies were excluded if the mobile component of the intervention only consisted of texts, emails or phone calls. These results are presented in Table 1 and synthesised narratively below.

Overall Studies and Type of Apps Used
Our study identified 15 potential papers; of those, five interventions were delivered online [20][21][22][23][24], one detailed the development of an app [25], and one study only recruited people with mood disorders [26], so they were excluded from this review. The remaining eight studies [27][28][29][30][31][32][33][34] were included and focused on a variety of apps (Quitpal, Kick.it, QuitGuide, Learn To Quit (LTQ), and quitSTART). The key features of these app and apps interfaces, we could detect, are presented in Tables 2 and 3, respectively. Three studies compared the usability of two different smoking cessation apps (LTQ and QuitGuide) in a schizophrenia sample [30,31,33]. Two used the same sample [31,34]. Vilardaga et al. (2019) conducted two case studies with crossover AB interventions, in which participants completed both interventions once, and the order of the intervention (QuitGuide and LTQ) was randomly allocated [30]. The System Usability Scale [35] (SUS; see Appendix A), was used to rate the usability of the interventions, SUS scores range from 0 to 100 and an SUS score above 68 is considered above average. The SUS scores for the two apps were found above the standard cut-off, with QuitGuide slightly outperforming LTQ for Participant 1 (who was diagnosed with schizophrenia), but largely underperforming LTQ in Participant 2 (who was diagnosed as unspecified psychotic disorder) [30]. Interviews were conducted to examine user experiences. The main themes for the LTQ app were "high retention and comprehension of smoking cessation content", "engagement through gamification", "simple to use and understand", and "positive emotional experience". The main themes for QuitGuide were "access to app features was challenging", "serious look and feel", "poor consensus about app content", and the "tracking and charts were perceived useful and desirable" [30].  [31,34]. All participants also received a smartphone with a data plan and 8-weeks supply of Nicotine Replacement Therapy (NRT). Over the 16-week follow-up period, both apps led to a reduction in cigarettes smoked, although the mean reduction in cigarettes per day was higher for LTQ participants (12.3 vs. 5.9 cigarettes) [31,34]. Browne et al. [32] reported that, compared to Quitguide, LTQ was also used on more days (34. Additional randomized control trial (RCT) data presented in Vilardaga et al. (2020) reported that 12% of LTQ participants had biochemically verified 30-day point prevalence abstinence (PPA) from smoking, versus 3% of QuitGuide participants at follow-up [31]. LTQ participant app interaction was significantly longer overall (4.2 vs. 2.1 h; p = 0.044), and rated usability highly (SUS score: 85.2 vs. 78.4, p = 0.046) when compared to QuitGuide participants 29. A small reduction in anxiety, symptom severity, and depression at 4-month follow-up was found for the entire sample [31].
Two studies conducted by Gowarty et al. [28,33] compared the usability and effect on smoking of two apps: quitSTART and QuitGuide. Seventeen participants were randomly assigned to field-test one of the two apps independently for 2-weeks. While both apps have similar features, their design, content and target audience differ: QuitGuide was developed for adults, whereas quitSTART was developed for teenagers and included prompts to record quit attempts.  Both studies found initially that participants using QuitGuide scored overall satisfaction more highly, as measured via SUS, and task completion was slightly higher, with all 9 tasks being completed by 75% of participants, whereas 75% of quitSTART participants completed 8/9 tasks. However, satisfaction and ease of use ratings for quitSTART increased to match QuitGuide at the 2-week follow-up. The interviews supported the notion that after familiarisation with quitSTAR, the app was perceived as easy to use despite the initial navigation difficulties. Further, at follow-up, quitSTART was used for a greater number of days, had a greater number of interactions, and users responded to more notifications, when compared with QuitGuide. Overall, 78% and 75% of QuitGuide and quitSTART users, respectively, self-reported quitting or reducing smoking at follow-up; abstinence was biochemically verified (with breath CO < 7 ppm) in 25% of all quitSTART users [28,33].
Three other studies focused on a single smoking cessation app. The first, Klein et al. (2015), was qualitative and involved a two-stage interview process, with the themes from stage 1 informing the development of the second interview schedule [27]. The first stage interviews aimed to understand participants' smoking-related experiences (smoking behaviour and history, effects of smoking on their mental health, motivation and attempts to quit, use of NRT). In this stage, participants reported that smoking helped them cope with mental illness and that their smoking increased when distressed. The second stage interviews explored experiences of using the Kick.it app (app features, content, functionality, aesthetics, and support). The key theme reported by this study was the importance of personalising the app to users' psychosocial needs, specifically experiences of perceived stigma and social isolation among this population. Participants felt smoking helped them cope with mental illness and their smoking increased when distressed [27]. Therefore, apps could be adapted to consider mental health. Participants also felt that the app should normalise smoking relapses as part of the process of quitting, and that smoking cessation approaches should assist quitting smoking in real time, by providing techniques and support to reduce the urge to smoke when stressors arise in their daily lives. Some users found navigating the Kick.it app overwhelming, leading the authors to conclude that the app interface should be simplified. Finally, the tracking of cigarette use was perceived to be useful, and most participants (83%) were enthusiastic about engaging in a social network, favouring chatrooms to connect to others [27]. Vilardaga et al. (2016) conducted a small-scale (n = 5) usability study of the Quitpal app [32]. The mean SUS score (65.5) indicated usability below the industry standard. A reoccurring barrier found in the interviews was difficulty using the app due to multiple layers. Other themes were: smaller achievable goals, app focus should be reducing smoking, and the use of interactive and motivating features. Most users liked the reminder function [32].
Lastly, Wilson et al. (2019) conducted a mixed-methods study which aimed to refine an already existing smoking cessation intervention for people diagnosed with schizophrenia, schizoaffective disorder and psychotic disorder not otherwise specified [29]. The Multi-Component Mobile-enhanced Treatment for Smoking Cessation (iCOMMIT) consisted of (i) Stay Quit Coach app; (ii) a smartphone-based contingency management application (mCM); (iii) Cognitive Behavioural Therapy (CBT) counselling, which consisted of 5 sessions and a workbook, provided per cohort; and (iv) pharmacotherapy with bupropion, started 2 weeks before quit date. The study consisted of two consecutive cohorts, with five and eight participants, respectively. Changes were made to the app after the first cohort, which included removing the Stay Quit Coach app from the second cohort due to the low self-reported utilisation in the first. Most participants reported that the counselling sessions were the best part of the treatment package. Participants generally noted the mCM was acceptable, but some patients mentioned the difficulty of uploading while filming their CO readings [29].

How Usable Are Smoking Cessation Apps?
Overall, studies demonstrated smoking cessation apps are feasible to use among people diagnosed with an SMI. When comparing overall SUS scores across smoking cessation apps, LTQ had an excellent SUS score (>80), while Quitpal [32], Quitstart [28,33], and in 2/3 studies using QuitGuide [28,33] SUS scores were below average (<70). Although impossible to directly compare with the other apps, it appears that the app designed specifically for patients with schizophrenia and other SMI (LTQ) was reportedly easier to use, possibly due to a simpler interface. Therefore, apps specifically developed and tailored for people with schizophrenia and other SMI may be more accessible and user friendly.
The need for tailoring apps to people's mental health needs was a recurring theme across the interviews. Further, the individual's technology skills (suggesting the simplification of app interfaces and ease of tracking behaviour), their symptomology and the impact that this has on their smoking, and the motivation to use the apps should be taken into consideration. To improve people's motivation and increase engagement with the app, a variety of strategies were suggested, including prompts, social support, progress tracking, rewards (monetary, or virtual through badges), and/or gamification [27,33]. Furthermore, apps that were gamified and included badges and quizzes (quitSTART and LTQ), and had prompts (quitSTART) were associated with longer engagement times and a greater number of interactions when compared with non-gamified apps (such as QuitGuide) [26,28,29,[31][32][33][34]. Hence, gamified apps with rewards such as badges, leader boards, etc., may be beneficial for this group to ensure initial engagement. However, all the studies had a relatively short follow-up, so the impact on sustained app engagement should be explored in future trials.
The desire for apps to have an element of social support, through either chat rooms or social media groups, with other users and/or smoking advisors was also highlighted, particularly when experiencing strong cravings or triggers [27,28,36]. Therefore, a smoking cessation app offering real-time social support and/or social support groups may be beneficial for people with schizophrenia. Alternatively, an app containing games, distraction techniques or quizzes could be useful if real-time support is not feasible.

Do Smoking Cessation Apps Reduce Smoking?
All the studies we included [28,31,33,34] found that smoking cessation apps led to a reduction in self-reported smoking. Two of the four studies found a significant reduction in smoking, determined by 30-day PPA, in patients assigned to LTQ when compared with QuitGuide (12% vs. 3%) [31,34]. The LTQ app also led to a significantly greater reduction in cigarettes compared with QuitGuide, although this was no longer significant when adjusted for the number of cigarettes smoked per day initially [31,34]. Additionally, participants assigned to QuitGuide had significantly more quit attempts (and thus, relapses) when compared with LTQ participants. Collectively, these findings may indicate an app designed specifically for those with schizophrenia and other SMI may be more favourable in this population.
The other two studies found that the proportion of participants who reported they had attempted to quit or reduce their smoking was similar for both apps (78% QuitGuide users vs. 75% quitSTART users) [28,33]. However, these studies only relied on self-reports, which are prone to recall bias, so it would be advisable that future studies use biochemical verification to confirm reduction in cigarette smoking. Due to the small sample sizes and short-term follow-ups, the effectiveness of smoking cessation apps cannot be fully determined, although it does appear that smoking cessation apps have great potential to help people with schizophrenia quit smoking, at least in the short term. Lastly, due to the narrative nature of this review, some smoking cessation apps already existing or in early development may have not been identified. Notwithstanding, this is, to our knowledge, the first comprehensive overview of the current state of affairs in the use of apps for smoking cessation in schizophrenia.
Participants reported that they were more likely to smoke when having worse mental illness symptoms and may be less likely to use an app at such times. However, none of the studies collected data regarding patient's symptomology nor examined if it influenced app use. Further longitudinal RCTs are required to determine the long-term effectiveness of smoking cessation apps among those diagnosed with schizophrenia. Further work is needed to understand the impact of the patient's symptomology on app use, their smoking cessation needs, and the effectiveness of the app. Such knowledge could be used to help people gain insight into their mental health and triggers for smoking relapse, and potentially identify additional support and coping mechanisms to manage their mental health.

Conclusions
Smoking cessation apps are feasible for people with schizophrenia. Apps may benefit from being tailored to the needs of people with schizophrenia, considering both layout (simple interface), content (normalising relapses and integrating symptomology and smoking), and components of the intervention (rewards, social support, etc.) to enhance engagement. There is a dearth of evidence to determine whether apps developed for those with schizophrenia and other SMIs are more effective at smoking cessation than apps designed for the general population. Additionally, there is an absence of evidence to determine if smoking cessation apps are effective at sustaining quit attempts long term. One limitation of these conclusions is the non-systematic narrative review methodology applied. Thus, further systematic reviews and meta-analyses on this topic might provide more definitive insights into the efficacy and utility of using digital devices for smoking cessation in mental healthcare.
Author Contributions: All authors were involved in the writing-review and editing. All authors have read and agreed to the published version of the manuscript.
Funding: J.F. is supported by a University of Manchester Presidential Fellowship (P123958) and a UK Research and Innovation Future Leaders Fellowship (MR/T021780/1) and has received honoraria/consultancy fees from Atheneum, Informa, Nutritional Medicine Institute, ParachuteBH, Richmond Foundation and Nirakara, independent of this work.

Appendix A The System Usability Scale
Five-point Likert scale, ranging from Strongly disagree (1) to Strongly Agree (5) (1) I think that I would like to use this system frequently.
(2) I found the system unnecessarily complex.
(3) I thought the system was easy to use. (4) I think that I would need the support of a technical person to be able to use this system. (5) I found the various functions in this system were well integrated. (6) I thought there was too much inconsistency in this system. (7) I would imagine that most people would learn to use this system very quickly. (8) I found the system very cumbersome to use. (9) I felt very confident using the system. (10) I needed to learn a lot of things before I could get going with this system.