Use of a Rapid Qualitative Method to Inform the Development of a Text Messaging Intervention for People With Serious Mental Illness Who Smoke: Formative Research Study

Background People with serious mental illness are disproportionately affected by smoking and face barriers to accessing smoking cessation treatments in mental health treatment settings. Text-based interventions are cost-effective and represent a widely accessible approach to providing smoking cessation support. Objective We aimed to identify key factors for adapting text-based cessation interventions for people with serious mental illness who smoke. Methods We recruited 24 adults from mental health programs who had a serious mental illness and currently smoked cigarettes or had quit smoking within the past 5 years. We then conducted virtual qualitative interviews between November 2020 and August 2021. Data were analyzed using the rapid thematic analytic approach. Results We identified the following 3 major themes: (1) interplay between smoking and having a serious mental illness, (2) social contextual factors of smoking in adults with serious mental illness, and (3) smoking and quitting behaviors similar to the general population. Participants reported barriers and facilitators to quitting across the 3 themes. Within the “interplay between smoking and having a serious mental illness” theme, barriers included smoking to manage stress and mental health symptoms, and facilitators to quitting included the awareness of the harm of smoking on mental health and patient-provider discussions on smoking and mental health. In the “social contextual factors of smoking in adults with serious mental illness” theme, barriers included high social acceptability of smoking among peers. Positive support and the combined social stigma of smoking and having a mental health condition outside of peer groups motivated individuals to quit. Some participants indicated that low exposure to other smokers during the COVID-19 pandemic helped them to engage in cessation efforts. In the “smoking and quitting behaviors similar to the general population” theme, barriers included smoking after eating, having coffee, drinking alcohol, and experiencing negative social support, and facilitators included health concerns, improvement in the general quality of life, and use of evidence-based tobacco treatments when available. Conclusions People with serious mental illness often smoke to cope with intense emotional states, manage mental health symptoms, or maintain social bonds. Text message content emphasizing equally effective and less harmful ways for stress reduction and mental health symptom management may improve quit rates in individuals with serious mental illness.


Introduction
People with serious mental illness report a higher smoking prevalence (41%) than people without serious mental illness (14%) [1,2]. Unfortunately, most people with serious mental illness who smoke have little or no access to pharmacotherapeutic cessation aids and behavioral approaches, which are the recommended treatments for smoking cessation [3]. While mental health treatment settings offer a possibility for engaging people with serious mental illness in smoking cessation, only a minority of them are currently receiving treatment [4]. In 2019, 46% of the 52.9 million US adults with any mental illness had received mental health services in the past year [4]. Further, many mental health settings do not routinely offer tobacco treatment [5]. Alternative approaches to engage and treat people with serious mental illness for smoking cessation are needed.
Text-based interventions are cost-effective and eliminate many barriers to accessing traditional treatments [6,7]. Ninety percent of individuals with a mental health diagnosis own more than one mobile device, including a mobile phone, and many report high use of multiple devices [8]. In the general population, text-based cessation interventions significantly improve quit rates and use of telephone counseling (ie, Quitline) services [9]. Studies have shown the feasibility of text-based cessation interventions in people with serious mental illness who smoke [10][11][12]. However, text-based cessation interventions need to address barriers to quitting that are specific to people with serious mental illness.
In the literature, many factors have been noted for increased tobacco use rates among people with serious mental illness. For instance, people with mental health conditions report using cigarettes to alleviate emotional problems, such as depression and anxiety, stabilize their mood, and relieve stress [13][14][15]. Many of these factors are shared with the general population of people who smoke. However, studies have not fully explored how these factors and mental health conditions intersect to increase tobacco use rates and quitting challenges in people with mental illness. Adapting text-based cessation interventions to address this intersection directly may increase their relevance and effectiveness, and improve engagement among people with serious mental illness who smoke [16].
The effectiveness of the text-based messaging intervention being adapted has been demonstrated in a large randomized trial conducted among individuals without serious mental illness [17,18]. Our intervention included use of expert and peer-written messages [19]. We developed the expert-written messages through an iterative group review process guided by theoretical frameworks and existing smoking cessation guidelines [3]. Peer-written messages were written by current and former smokers responding to an online survey. The content of expert messages was more "biomedical" in nature (avoidance, behavioral strategies, and health), while the content of peer messages focused on "social" and "real-life" aspects of smoking (expectations, money, quality of life, attitudes, and friends) [19]. The intervention group received motivational text messages weekly for 6 months, and compared with the control group, receiving text messages increased the odds of 7-day point prevalence cessation at 6 months (odds ratio 1.69, 95% CI 1.03-2.8) [17,18]. However, these messages do not address barriers associated with serious mental illness.
In this study, we explored tobacco use behaviors and barriers along with facilitators of quitting among 24 people with serious mental illness to identify key factors for adapting a text messaging intervention that addresses their cessation needs. We also detailed a 4-step approach to identify themes from the qualitative interviews that could be a structured process for future text-based intervention developers.

Study Overview
We conducted an in-depth qualitative exploration of smoking behaviors, barriers, and facilitators among adults with serious mental illness who currently smoke or quit within the past 5 years.
We used a 4-step approach to identify key themes that will inform the development of text messages targeted at people with mental health conditions. Briefly, the 4-step approach included identifying domains that aligned with key interview questions, summarizing transcript data under each domain, consolidating the data into a matrix, and interactively identifying themes under each domain. We provide details in the data analysis section.

Study Setting and Participant Recruitment
Participants were recruited nationally from 3 mental health agencies, most of which are publicly funded (Clubhouses accredited by the International Center for Clubhouse Development [20], Thresholds Agency, and Massachusetts Department of Mental Health agencies such as Open Sky). These mental health programs provide services to individuals with various mental health conditions. Services include the opportunity for friendship, employment, housing, education, and access to medical and psychiatric services. Clubhouses are accessed by about 100,000 people with mental health conditions [21]. Thresholds is a large community mental health center in Chicago that serves 12,500 adults and youth each year. The Massachusetts Department of Mental Health programs and Clubhouses affiliated with Clubhouse International provide various clinical and nonclinical services to adults with serious mental illness.
We recruited 24 adults with serious mental illness through virtual information sessions administered by a research coordinator. Individuals were eligible to participate if they (1) were 18 years or older, (2) were currently smoking cigarettes or had quit within the past 5 years, (3) were willing and able to provide consent, and (4) were currently receiving services from mental health agencies (ie, Clubhouses, Thresholds Agency, and Massachusetts Department of Mental Health agencies such as Open Sky). We excluded pregnant individuals or those attending correctional programs (ie, prisoners).

Ethical Considerations
Potential participants provided contact information to the research coordinator, who set up one-on-one virtual meetings to provide more information about the study. Those interested in participating provided informed electronic consent (e-consent) before enrollment into the study. Electronic signatures were captured using the Research Electronic Data Capture e-consent framework. Participants received a US $15 Amazon electronic gift card for the study. This project was approved by the ethics review boards at the Department of Mental Health in Massachusetts and the University of Massachusetts Chan Medical School (reference number: H00019687_3).

Interview Guide Development
The interview guide was developed by researchers with expertise in mental health and tobacco use based on prior literature [22][23][24]. The interview guide was structured using domains that aligned with the interview questions and the overall study objective of developing text messages for people with serious mental illness. We chose the approach outlined in Rapid qualitative inquiry: A field guide to team-based assessment by James Beebe [25] to generate the domains explored in our study. Informed by the literature, the research team identified the key factors that would be the most meaningful in informing the implementation aspects of a smoking cessation study. Such factors included the context in which individuals with serious mental illness smoked, barriers and facilitators to quitting, and how one's mental health diagnosis may influence their smoking habits. These factors guided the formulation of the key research questions/interview guide, which shaped our study domains.
The final interview questions broadly focused on motivations, beliefs, and barriers and facilitators to smoking or quitting. We also asked about smoking triggers or cues; symptoms of tobacco dependence; previous experiences reducing or stopping smoking, including use of and barriers to evidence-based cessation; and the role of interpersonal relationships in smoking cessation. During the interviews, we prompted participants to reflect on how their diagnosis has influenced their smoking behaviors and quitting experiences. The specified 9 domains that aligned with these questions included the context in which participants smoked, smoking triggers, reasons for smoking, barriers to active quitting, facilitators of active quitting, mental health-specific quitting strategies, mental health-specific facilitators of quitting maintenance, and the influence of the COVID-19 pandemic on smoking.
We refined the interview guide based on feedback from a stakeholder advisory panel of 9 adults with serious mental illness who also currently smoked or had quit within the past 5 years. Panel members were recruited from the same mental health agencies as study participants to ensure the representativeness of the study population.

Data Collection
Two trained researchers conducted the qualitative interviews using a semistructured interview guide via Zoom video conference software (Zoom Video Communications Inc). Interviews were conducted between November 2020 and August 2021, and lasted for 30 to 40 minutes. All interviews were audio recorded, deidentified, and professionally transcribed. In addition to the qualitative interview, participants completed a baseline questionnaire that captured demographic information (eg, age, sex, race/ethnicity, and educational level) and smoking behavior or history of those who recently quit.

Data Analysis
We used the rapid thematic qualitative analytic approach to analyze qualitative data. This rapid approach is a team-based qualitative inquiry that uses triangulation and iterative data analysis to quickly develop a preliminary understanding of a situation from the insider's perspective [25][26][27]. This approach to data analysis provided insights into the context in which participants smoked and their attitudes and beliefs toward quitting smoking.
The 4-step process is as follows: We performed a "test drive" to assess the domains' usability, relevance, and consistency. In this step, 2 coders (CSN and IAL) used the template to summarize the same transcript to evaluate whether the specified domains were identifiable in the data and to check for consistency across coders in capturing the domains. We modified the template based on feedback from the rest of the research team (RSS, MD, AK, and CEM) before testing with a second transcript. Once consistency was established, we divided the transcripts equally across the 2 coders and summarized the transcripts using the modified template. 3. We transferred transcript summaries into a matrix. We placed information from the 24 transcript summaries into a Microsoft Excel matrix. The transcript summaries were put into a matrix to analyze each domain's depth and breadth of data [25]. Each column in the Excel document represented a prespecified domain, and the rows represented study participants. The cells in the Excel document contained summarized responses from each study participant that corresponded to a specific domain. 4. We identified study themes and subthemes using the matrix.
Themes were identified within each domain. The research team collaboratively and iteratively reviewed, discussed, and sorted the data to refine the initial themes and subthemes, and highlight the most salient quotes. The research team members provided multiple perspectives; a triangulation (investigator triangulation [28]) approach was designed to increase the findings' reliability. In addition to investigator triangulation, we used the peer-review approach [29], in which a researcher (MMK) who was not involved in data collection reviewed the evidence that supported the interpretation of the data and conclusions. The final themes highlight the views and experiences with using family or peer support for smoking cessation and considerations for support of cessation interventions. A brief descriptor (subtheme) of what participants reported is provided within each theme. Each of these themes will inform the text message content.

Participant Characteristics
Participant characteristics are presented in Table 1. Most participants were between 36 and 54 years old (11/24, 46%) and male (15/24, 63%). About half self-identified as non-Hispanic White (13/24, 54%) and had a high school education (13/24, 54%). More than half (10/17, 59%) of the participants who were currently smoking had made a quit attempt in the past year. A high proportion (20/24, 83%) of study participants owned a cell phone or smartphone, and a majority (19/20, 95%) of those who owned a cell phone or smartphone often used it to send and receive text messages ( Table 1).

(83%) Yes
Besides yourself, does anyone who lives with you smoke cigarettes now? a

(30%) Yes
Do you own a cell phone or smartphone?

(83%) Yes
Do you ever use your cell phone to send or receive text messages? (N=20) 19 (95%) Yes a Information is missing for 1 participant.

Factors for Adapting Text-Based Interventions for People With Serious Mental Illness Who Smoke
We found 3 major themes that provided insights into the experiences of quitting smoking in people with serious mental illness. The themes included interplay between smoking and having a serious mental illness, social context of smoking, and similarities in smoking and quitting behaviors between the participants and the general population (Table 2). There were barriers and facilitators within each theme, as presented below.

Theme 2: Social Contextual Factors of Smoking in Adults With Serious Mental Illness
The social context played a vital role in the smoking and quitting behaviors of people with serious mental illness.

Barriers
Smoking seemed socially acceptable among peers and was viewed as a social bonding activity.

Facilitators
Participants were motivated to stop smoking due to the existing social stigma of smoking outside their peer networks and viewed positive support as a facilitator to quitting smoking. For instance, 1 smoker indicated that the social stigma of smoking, when added to the stigma of having a mental illness, was a motivator to stop smoking.

Theme 3: Smoking and Quitting Behaviors Similar to the General Population
There were similarities in smoking and quitting behaviors between people with serious mental illness and the general population.

Barriers (Smoking Behavior)
Similar to the general population, participants tended to smoke out of habit, such as after eating or having coffee, or when drinking alcohol.

Facilitators (Quitting Experiences)
However, individuals had concerns about the negative impact of smoking on their physical health and general quality of life, which motivated them to quit smoking.

Main Study Findings
Using the 4-step analytical approach, we identified the following 3 key aspects that can inform the development of text-based cessation interventions for people with serious mental illness who smoke: (1) interplay between smoking and having a serious mental illness, (2) social contextual factors of smoking, and (3) smoking and quitting behaviors similar to the general population. Across the 3 themes, participants reported barriers, motivators, and facilitators to quitting smoking. The variations in the content provided by the participants can facilitate the development of messages that can be implemented in text-based interventions targeted at people with serious mental illness who smoke.
Participants strongly believed that smoking helped manage mental health symptoms and was often used to cope with stress. Stress is often cited as a reason for smoking in individuals with and without serious mental illness [30,31]. However, because stress has a strong positive correlation with mental illness [32], people with serious mental illness may not progress through the stages of change at a rate comparable to that in those without serious mental illness. In addition, the term "stress" could be used as a proxy for mental health symptoms in this patient group. In such a case, smoking could be used as self-medication to alleviate feelings of intense emotions or to feel an immediate sense of relaxation even though individuals are aware of the harmful effects of smoking on their mental and physical health.
Individuals found that speaking with their mental health care provider when quitting helped them navigate the quitting process better. In discussions with their providers, they expressed the impact of tobacco medication (ie, Chantix) on their mental health symptoms. This finding implies that patient-provider discussions regarding smoking and mental health may facilitate successful quitting. Therefore, engaging health care providers when making a quit attempt is essential in this population. Patient-provider discussions regarding smoking also provide a critical opportunity for health care providers to monitor for smoking/quitting-drug interactions [33] as patients may not be aware of these interactions.
Adapting text-based cessation interventions to address barriers to smoking among people with serious mental illness may increase the intervention's relevance, engagement, and effectiveness. Past text-based cessation interventions have not included content specific to mental health conditions [34]. A systematic review of text-based cessation interventions found that message content focused on increasing self-efficacy and encouraging smokers to quit or maintain their quit status by providing quit tips [34]. This content is very similar to the one we used in our own past studies [35,36]. Currently, text messages used in cessation interventions lack content that highlights mental health symptoms and stress management, and content that emphasizes the need to engage mental health care providers when quitting. Therefore, text messages that offer tips on evidence-based strategies for managing stress, regulating emotions, managing negative affect, and dealing with unhelpful thoughts/anxiety can be a meaningful inclusion in text-based interventions. Messages that serve as reminders to engage mental health care providers when quitting may increase the relevance of text-based cessation interventions in this patient population.
We found a strong tendency to smoke in people with mental health conditions and in the mental health environments in which they were embedded. From our observations, smoking seemed ingrained in all aspects of their social lives. For instance, smoking was perceived as a social bonding activity so much that individuals often perceived smoking as a way to be socially included or form friendships. Given the high level of social acceptance observed in this patient population [24,37], these barriers to quitting smoking extend beyond those experienced by people in the general population who smoke. Recognizing the broader factors that drive smoking rates within the living and social environments of people with serious mental illness can complement existing efforts aimed at helping them quit successfully. A few past text messaging interventions have included messages with content relevant to the social context [38]. The messages are about seeking social support or how to deal with a partner who smokes. Additional content could include messages that denormalize smoking [39], provide alternatives to smoking as a group bonding activity that could counteract the perceived social rewards of smoking, and provide information on how to identify cessation role models.
In addition to the distinct challenges identified by people with serious mental illness, they also face challenges commonly observed among people in the general population who smoke. Participants often smoked out of habit, such as after a meal or while drinking alcohol, due to boredom and having access to cigarettes. They reported similar reasons for quitting, including health benefits, the financial cost of cigarettes, and the general improvement in quality of life. Participants also used evidence-based strategies to quit smoking, including nicotine replacement therapy and tobacco treatment medications. These findings are similar to past study findings [40], which indicate that people with serious mental illness are motivated to quit and can engage in cessation using evidence-based resources, yet they often cannot quit. It is important to modify existing text-based interventions to provide information on mental health service access and to include content that increases antismoking attitudes.
Other researchers have discussed the need for a structured process to support the development of text messages [41]. The approach we used in this paper provided an efficient way to identify key factors that can inform text message development from formative interviews. We chose the rapid qualitative approach over other qualitative data analysis approaches because it provided a systematic way to describe the lived experiences of people with serious mental illness who smoke within each domain (eg, the context in which participants smoked, smoking triggers, reasons for smoking, barriers to active quitting, and facilitators of active quitting).

Limitations
This study had limitations. Our study recruited participants from mental health agencies, who had access to mental health services. In many ways, individuals with access to mental health services could differ from those without access to mental health services regarding smoking behaviors, attitudes toward quitting, and cessation experiences. Targeting people with serious mental illness without access to health services could provide additional insights into these patients' challenges. Additionally, we relied on self-reported information when assessing smoking status.

Conclusion
People with serious mental illness may experience life stressors more intensely than those without serious mental illness. As a result, they often use smoking to cope with intense emotional states, manage mental health symptoms, or maintain social bonds. Based on our study findings, text message content should be modified to (1) provide information on the effect of smoking on mental health and alternative approaches to managing mental health symptoms, (2) incorporate smoking denormalization strategies that increase antismoking attitudes, and (3) provide healthier alternatives to smoking for social bonding activities. Adapted text-based interventions can ensure that people with serious mental health illness receive the appropriate support to stop smoking.

Data Availability
This article's data can be shared on reasonable request to the project principal investigators (MD and RSS).