Patients With Serious Mental Illness Can Engage in Tobacco Treatment Trials and Tobacco Cessation Interventions During Cancer Treatment

PURPOSE:

More than half of individuals with serious mental illness (SMI) smoke, contributing to premature cancer mortality. A cancer diagnosis provides an opportunity to assist with smoking cessation; however, supportive oncology trials frequently exclude patients with SMI. To fill this gap, we examined differences in engagement and tobacco cessation in a pragmatic clinical trial.

METHODS:

We recruited 303 participants from two National Cancer Institute–designated Comprehensive Cancer Centers, of which 10% had prior diagnoses of SMI (major depressive disorder, bipolar disorder, and schizophrenia spectrum disorders). We compared self-reported smoking behaviors, patient attitudes and beliefs about cessation, and rates of trial completion, engagement, and smoking abstinence among recently diagnosed patients with cancer with and without SMI. Six months after trial completion, we completed qualitative interviews on barriers and facilitators to tobacco cessation in a random sample of participants with SMI.

RESULTS:

Trial participants with SMI had similar motivation to quit smoking as those without SMI. Additionally, participants with SMI had a similar ability to engage in a tobacco treatment trial (6.5 counseling sessions completed v 7.3 sessions) and benefit from tobacco treatment as those without SMI (32.3% v 27.8% 6-month quit rates).

CONCLUSION:

Patients with cancer and SMI were able to engage in and benefit from a tobacco cessation trial integrated into cancer care. A cancer diagnosis provides an opportunity to assist patients with SMI with smoking cessation referrals and treatment. Pragmatic supportive oncology trials that include a diverse population of adults with SMI are needed to inform care delivery and improve cancer outcomes for patients with SMI and cancer.

INTRODUCTION

Combustible tobacco smoking is the leading cause of preventable death in the United States.¹ Although public health efforts have led to population-wide reductions in cigarette use, individuals with serious mental illness (SMI; schizophrenia spectrum disorders, bipolar disorder, and major depressive disorder) continue to smoke at disproportionally higher rates than the general population.² Compared with 17% of adults without mental illness, an alarming 62% of individuals with schizophrenia and 37% of individuals with bipolar disorder report current cigarette use.² Individuals with mental illness are also more likely to smoke more cigarettes, have greater nicotine dependence, and report lower quit rates than the general population.^2,3

Additionally, individuals with SMI die approximately 10-27 years earlier than the general population, and smoking-related illnesses, including cancer, account for 60% of this premature mortality.^4-6 Individuals with schizophrenia are more than twice as likely to die of lung cancer compared with adults without mental illness.⁵ Individuals with SMI also experience disproportionate cancer mortality from other cancers that are associated with smoking such as head and neck cancer.⁷ Furthermore, despite greater incidences of smoking-related illnesses, individuals with SMI receive less screening for tobacco use and lower rates of referrals and follow-up for cessation services within health care systems than individuals without mental illnesses.⁸

A cancer diagnosis provides an opportunity to target smoking cessation among this marginalized population. The ASCO guidelines recommend that oncology providers ask and counsel all patients about tobacco use at every clinic visit.⁹ However, most clinical trials exclude patients with cancer and SMI, leading to inadequate knowledge about best practices for tobacco cessation in this population. Researchers may hesitate to include individuals with SMI in clinical trials because of concerns about their ability to provide informed consent and follow trial procedures. Furthermore, clinicians may hesitate to prescribe tobacco cessation medications to adults with SMI.¹⁰ To promote tobacco cessation in individuals with SMI and inform approaches to broaden inclusion of this population in supportive oncology research, we analyzed the relationship between SMI and smoking cessation among participants with recent cancer diagnoses at two National Cancer Institute–designated Cancer Centers, which were enrolled in a pragmatic trial of a tobacco cessation intervention.¹¹ We aimed to compare (1) smoking behaviors, (2) patient attitudes and beliefs about smoking cessation, and (3) trial and treatment engagement and smoking abstinence between patients with and without SMI.

METHODS

Participants

The sample included 303 participants recruited from two National Cancer Institute–designated Comprehensive Cancer Centers (Massachusetts General Hospital/Dana-Farber/Harvard Cancer Center and Memorial Sloan Kettering Cancer Center) enrolled in a clinical trial for tobacco cessation (ClinicalTrials.gov identifier: NCT01871506, PI E.R.P.). Eligible participants were English- or Spanish-speaking adults (age > 18 years) and current smokers (self-reported smoking ≥ 1 cigarette in the past 30 days) undergoing treatment for a recent diagnosis of breast, gastrointestinal, genitourinary, gynecologic, head and neck, lymphoma, lung, or melanoma cancer.

The study team aimed to include a sample that was generalizable to clinical practice and limit exclusion criteria. Trial participants were not excluded because of a diagnosis of a SMI. A psychiatrist was available for consultation regarding questions about trial eligibility for patients with SMI. Reasons for exclusion included medical and cognitive impairment deemed likely to interfere with study participation (determined by the treating clinician), uncontrolled psychosis or suicide attempt in the past year, insufficient comprehension/English literacy (determined by electronic health record (EHR) and/or treating clinician), and not receiving cancer care at a study site.

Research staff screened cancer clinic schedules to identify and contact potentially eligible patients, confirmed eligibility, and obtained written informed consent from all patients. Study methods and primary outcomes have been previously published.¹¹ Study procedures were approved by the institutional review boards of the participating sites.

Measures

Sociodemographic and cancer history information were collected from the self-reported baseline survey and EHR at the time of enrollment. Sociodemographic data included age, sex, race, ethnicity, education, marital status, employment status, and insurance type. Cancer history information included cancer diagnosis and stage. Cancer diagnoses were further categorized as smoking-related and non–smoking-related by clinical study team members. The trial protocol paper¹² includes full descriptions of the measures used including assessments of validity and reliability.

SMI

SMI was defined as preexisting major depressive disorder, bipolar disorder, or schizophrenia spectrum disorder. SMI diagnoses were identified by review of the EHR including the problem list and medical history. Additionally, the team used a list of commonly used medications for bipolar disorder and schizophrenia to facilitate screening in case the SMI diagnosis was not documented in the EHR.

Prior Smoking and Cessation Behaviors

At baseline, participants self-reported information on smoking behaviors, including the number of cigarettes smoked per day, age of smoking initiation, time to first cigarette after waking, ever use of electronic cigarettes (e-cigarettes; yes/no?), and past 24-hour quit attempts. Participants were also asked about prior cessation efforts, including prior use of cessation medications, cessation counseling, or complementary therapies to quit smoking.

Environmental Influences

Participants reported whether at least one household member smoked and if their oncologist asked, advised, or assisted with tobacco cessation counseling or cessation medications.

Smoking Attitudes and Beliefs

At baseline, motivation to quit was assessed with a self-reported, one-item, 10-point contemplation ladder, ranging from 1 (I enjoy smoking so much I will never consider quitting no matter what happens) to 10 (I have quit, and am 100% confident that I will never smoke again). Perceived importance and cancer treatment benefits of quitting were similarly assessed via one-item, 11-point Likert Scales, ranging from 0 (not at all) to 10 (very much). Confidence in quitting smoking was assessed via an averaged, 11-item scale that ranged from 0 (no confidence at all) to 10 (absolutely confident).

Engagement With Trial and Intervention and Smoking Outcomes

Engagement with the trial and intervention was assessed via number of follow-up surveys completed, self-reported use of cessation medication, and number of counseling sessions attended. Additionally, the study team tracked rates of adverse events and study withdrawals. Our treatment outcome of biochemically confirmed past 7-day point prevalence cigarette smoking abstinence (quit rate) was assessed at 3-month and 6-month follow-ups.

Qualitative Interviews

After completion of 6-month follow-up surveys, a random subset of participants (n = 72) completed in-depth semistructured individual interviews using a priori stratification criteria that included history of SMI and quit status.¹³ The interviews were conducted using a semistructured interview guide developed by study team investigators, and asked about smoking and quitting behaviors, stress and distress, and how the intervention helped or did not help to quit smoking. Interviews were recorded, transcribed, and coded independently by team members using NVivo 11 (QSR International, Burlington, MA).

Statistical Analyses

We examined frequency distributions of all baseline variables and characteristics of participants with (n = 31) and without SMI (n = 272). Missing values were removed for comparison. An inductive approach to qualitative content analysis was used, beginning with open coding of the interview transcripts and further categorization of the codes into broader themes until thematic saturation was reached. High intercoder agreement was reached with kappa > 0.86.

RESULTS

Sociodemographic and Cancer Characteristics

Sociodemographic and cancer characteristics are displayed in Table 1. Overall, 56.1% of participants identified as female and 87.5% identified as White. Participants had a mean age of 58.3 years (standard deviation [SD] = 9.7) and were married (55.6%), privately insured (61.8%), and reported education beyond high school (68.4%). More than half were currently unable to work or unemployed (56.1%).

TABLE 1.

Sociodemographic and Cancer Characteristics Among Study Participants With and Without SMI for the Smokefree Support Study

Approximately 10% of the total sample had a diagnosis of SMI (n = 31). Specifically, 23 (74.2%) had major depression, six (19.4%) had bipolar disorder, and two (6.4%) had schizophrenia spectrum disorders. Participants with SMI were more likely to be female than those without SMI (77.4% v 53.7%). No differences were observed between groups for age, race, ethnicity, education, marital status, employment status, or insurance type.

Thoracic cancer (30.7%) and breast cancer (25.4%) were the most frequently observed cancer diagnoses in the overall sample. Smoking-related cancer diagnoses were observed for 59.7% of the total sample and 51.6% of participants with SMI.

Prior Smoking and Cessation Behaviors

Participants with SMI smoked fewer cigarettes per day (M = 10.6, SD = 7.6) than participants without SMI (M = 14.5, SD = 10.1; Table 2). The groups reported comparable age of smoking initiation, time to first cigarette after waking, and completion of past 24-hour quit attempts. Similar frequencies for self-reported past use of cessation medication, e-cigarettes, counseling, or complimentary therapies were observed.

TABLE 2.

Differences in Prior Tobacco Treatment and Perceptions of Risks at Baseline by SMI Diagnosis

Environmental Influences

Exposure to secondhand smoke from household members was comparable between groups. Seventy-one percent (22/31) of adults with SMI reported being asked by oncology clinicians about smoking compared with 86.9% (225/259) of adults without SMI. 48.4% (15/31) patients with SMI reported that their oncologists recommended cessation counseling compared with 60.6% (151/249) participants without SMI; 32.3% (10/31) patients with SMI were recommended cessation medications versus 47.3% (121/256) patients without SMI.

Smoking Attitudes and Beliefs

Participants with and without SMI reported similar motivations to quit smoking (5.8/10 v 5.9/10, respectively). They also had comparable perceptions about the importance of quitting, cancer treatment benefits, and confidence in quitting (Table 2).

Engagement With Trial and Treatment and Smoking Outcomes

Participants with SMI did not differ significantly from those without SMI in their engagement in the trial or treatment (Table 3). Specifically, individuals with SMI had similar rates of consent, enrollment, random assignment, and completion of study procedures and assessments compared with trial participants without mental illness (see Appendix Fig A1, online only, for full CONSORT diagram). These measures included percentage of follow-up surveys completed (77.4% v 71.3% at 6 months), number of counseling sessions attended (6.5 v. 7.3 intensive treatment group sessions completed), and self-reported cessation medication use at 3 or 6 months. Additionally, there were no differences in adverse events, or rates of withdrawal from the trial. Rates of biochemically confirmed past 7-day cigarette smoking abstinence were similar between groups at 6 months (32.3% v 27.8%).

TABLE 3.

Trial and Treatment Engagement and Smoking Outcomes

Qualitative Interviews

Eight participants with SMI completed the qualitative interviews. Overall, participants with SMI valued participating in the intervention. In particular, participants with SMI emphasized the importance of counselors using frequent and consistent outreach. Additionally, participants with SMI valued that counselors adopted a nonjudgmental approach about smoking behaviors to decrease shame, increase acceptance, and maintain hopefulness. One participant explained, “she (the counselor) was very encouraging and didn't push. She just tried to help you.” Another participant discussed how important it was to “talk to somebody who's not judging me… not saying ‘oh, yeah, smoking is stupid’… kind of putting me down… she also helped me with believing in myself.”

Additionally, consistent with qualitative research conducted with patients with newly diagnosed cancers without mental illness,¹³ participants with SMI discussed the importance of addressing anxiety as an integral part of tobacco cessation. First, patients discussed a longstanding automatic connection between stress and smoking, identifying “I've been smoking since I was 15, so automatically I get stressed out, I light up.” Second, participants recognized the need to learn new strategies to address increased anxiety associated with cancer. “With the cancer diagnosis I was more anxious too, so I kept smoking more even though I knew it wasn't good.”

The intervention helped participants with SMI to cut down on smoking even when not able to quit, which was perceived as a meaningful accomplishment. One patient with SMI shared, “I feel a little better. I don't smoke two packs. It went down to like 12, 13 (cigarettes) which is a good thing.” Another participant explained, “Your program actually. It helped a lot. And my desire, not to quit, but to at least make a change.”

DISCUSSION

The current study examined differences between patients with and without SMI among participants in a trial of tobacco cessation treatments for patients with recent cancer diagnoses. Although this sample size of individuals with SMI was small and included primarily participants with major depressive disorder, patients with cancer and comorbid SMI had comparable levels of interest in smoking cessation and ability to engage and benefit from tobacco treatment as those without SMI. Specifically, there were no group differences in motivation to quit smoking, confidence in quitting, or perceived benefits and importance of quitting. Participants with SMI not only completed surveys and counseling sessions and used cessation medication at similar rates as those without SMI, but also had comparable smoking quit rates at 3 and 6 months.

These findings are particularly relevant as tobacco treatment trials in patients with cancer have systematically excluded patients with SMI in part because of the concern that patients with SMI and cancer will have lower intervention adherence than the general population.¹⁰ Furthermore, randomized trials that include larger samples and broader representations of SMI diagnoses are needed to inform best practices in tobacco cessation for patients with SMI and cancer. This research has potential to build on the converging evidence that supports the benefit and safety of combined counseling and pharmacologic treatment for tobacco cessation among those with SMI and extend those findings to patients with comorbid cancer. Unfortunately, many clinicians hesitate to prescribe tobacco cessation medications to patients with SMI in part because of concerns about tolerability and effectiveness. In particular, concerns were raised about the risk of neuropsychiatric side effects experienced by patients with mental health disorders taking bupropion (an antidepressant) and varenicline. Follow-up studies have demonstrated that varenicline and bupropion are tolerable and effective for patients with stable SMI.¹⁴ In the EAGLES trial, adults with SMI (including bipolar disorder, schizophrenia, and depression) who received varenicline were five times more likely to quit compared with placebo, and there were no differences in neuropsychiatric side effects for trial participants who received varenicline or bupropion compared with placebo.¹⁵ Furthermore, rather than worsen symptoms, tobacco treatment has consistently been shown to improve psychiatric symptoms for individuals with mental illness,¹⁴ lending support to the proposed benefit of integrating tobacco cessation into cancer care.

Although additional research is needed to evaluate engagement and benefits of tobacco treatment for patients with SMI in larger and more diverse supportive oncology trials with greater statistical power, this secondary analysis provides preliminary support for the inclusion of patients with SMI in tobacco cessation trials. Consistent with ASCO's call to decrease barriers to trial participation for underserved populations with poor cancer outcomes,⁹ there is an urgent need to broaden default eligibility language for inclusion and exclusion criteria related to mental illness. Rather than excluding on the basis of psychiatric diagnosis, researchers need to weigh the risks and benefits of the proposed intervention and use targeted recruitment efforts to include participants with bipolar disorder and schizophrenia. Furthermore, systematically tracking rates of mental health disorders among individuals enrolled in oncology trials may provide evidence to support enrollment in clinical trials and guide the design of tobacco cessation interventions for this underserved population.

As assessing smoking characteristics and treatment completion among individuals with SMI was a secondary aim, we did not enroll equivalent numbers of participants with and without SMI. Our findings were consequentially limited by a smaller sample size, which may have reduced the power of our study to detect significant group differences. Future research with targeted recruitment of individuals with SMI is needed to replicate these findings within larger samples.

Notably, this population of individuals with SMI primarily included patients with major depression, a population whose engagement in tobacco cessation and participation in clinical trials may differ from patients with bipolar and schizophrenia spectrum disorders. Additionally, we were limited in our ability to assess the severity of mood and psychotic symptoms for participants with SMI, which may have led to an oversampling of patients with less severe psychiatric illness. Our sample of participants with SMI also had similar or reduced smoking behaviors compared with those without SMI, which is inconsistent with the literature and likely reflects a population that primarily includes MDD. Despite these limitations, this secondary analysis fills a gap in the literature by describing trial participants with SMI and cancer and highlighting their ability to participate, engage, and benefit from a tobacco cessation intervention.

In conclusion, our findings suggest that patients with SMI and cancer are interested in smoking cessation and demonstrate comparable rates of engagement, adherence, and cessation in tobacco trials. Oncology clinicians have an untapped opportunity to mitigate the mortality gap experienced by patients with SMI by asking all patients about their smoking behaviors, particularly those with SMI, and connecting those patients to treatment.

APPENDIX

FIG A1.

CONSORT diagram. BL, baseline; IT, intensive treatment; SMI, serious mental illness; ST, standard treatment.

AUTHOR CONTRIBUTIONS

Conception and design: Elyse R. Park, Jamie S. Ostroff

Administrative support: Colin Ponzani

Provision of study materials or patients: Jamie S. Ostroff

Collection and assembly of data: Elyse R. Park, Jamie S. Ostroff, Colin Ponzani, Susan Regan

Data analysis and interpretation: Eleanor B. Steffens, Elyse R. Park, Alona Muzikansky, Joanna M. Streck, Jamie S. Ostroff, Autumn W. Rasmussen, Giselle K. Perez, Nancy A. Rigotti, Kelly E. Irwin

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Patients With Serious Mental Illness Can Engage in Tobacco Treatment Trials and Tobacco Cessation Interventions During Cancer Treatment

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).