Individuals with serious mental illness experience disparities in lung cancer mortality. Using a two-phase, mixed-methods approach, researchers developed a person-centered lung cancer screening educational intervention (phase 1) for individuals with schizophrenia and bipolar disorder and evaluated acceptability, feasibility, and changes in attitudes toward LCS (phase 2).
Purpose:
Individuals with serious mental illness (SMI) experience disparities in lung cancer mortality. Using a two-phase, mixed-methods approach, we developed a person-centered lung cancer screening (LCS) educational intervention (phase 1) for individuals with SMI (schizophrenia and bipolar disorder) and evaluated acceptability, feasibility, and changes in attitudes toward LCS (phase 2).
Methods:
Phase 1: We conducted three focus groups with mental health, primary care, and radiology clinicians and utilized rapid qualitative analysis to adapt the LCS intervention (LCS walk-through video and smoking cessation handouts) tailored for individuals with SMI. Phase 2: We enrolled LCS-eligible patients with SMI (n = 15) and assessed the feasibility (>50% enrollment; >75% completion) and acceptability (>75% overall satisfaction) of an LCS educational intervention delivered by a radiologist and a mental health clinician at a community mental health clinic. We explored changes in participant attitudes about lung cancer, LCS, and smoking before and after the intervention.
Results:
Phase 1: Focus groups with primary care (n = 5), radiologists (n = 9), and mental health clinicians (n = 6) recommended person-centered language and adapting a video demonstrating the process of LCS to address concerns specific to SMI, including paranoia and concrete thinking. Phase 2: Fifty percent (15 of 30) of eligible patients enrolled in the LCS intervention, 100% (15 of 15) completed the intervention, and 93% (14 of 15) were satisfied with the intervention. Participants reported a significantly greater worry about developing lung cancer postintervention, but there were no other significant differences.
Conclusions:
Radiologists can partner with primary care and community mental health clinics to lead equity efforts in LCS among individuals with SMI.
Lung cancer remains the leading cause of cancer-specific mortality in the United States with an estimated 135 720 deaths in 2020.1 Individuals with serious mental illness (SMI) represent an underserved patient population that experiences significant disparities in lung cancer outcomes and is two to four times more likely to die from lung cancer.2 SMI, specifically schizophrenia and bipolar disorder, affects nearly 13 million US adults who die 15 to 30 years earlier than individuals without SMI. Cancer is the second leading cause of this premature mortality because of a combination of factors that include inequities in cancer care, fragmentation of health care delivery between mental health and primary care services, lack of social support to navigate an increasingly complex health care system, and higher smoking prevalence.2- 7
Lung cancer screening (LCS) has the potential to mitigate existing disparities in lung cancer outcomes among individuals with SMI through early detection. Despite the proven benefits and mandated insurance coverage, uptake remains low.8- 12 Collaborative interventions tailored to individuals with SMI through a trusted mental health clinician and in a familiar setting of a community mental health clinic (CMHC) can increase uptake and reduce existing lung cancer disparities. Prior studies have shown that health preventative services, such as lifestyle changes to decrease cardiovascular disease risk and optimize diabetes management1314, can be integrated into the trusted setting of a CMHC for individuals with SMI and improve health outcomes.15- 17 Radiologists, mental health clinicians, and primary care clinicians have the potential to collaborate to develop LCS outreach interventions for adults with SMI that are embedded within CMHCs.
The required shared decision making to enroll in LCS underscores the importance to tailor educational materials about LCS to fit the health literacy needs and capacities of individuals with SMI who have difficulty with abstract thinking to understand cancer risk and establishing trust and have limited social support to assist in making health decisions. Delivering tailored LCS educational material in a CMHC setting represents an opportune teachable moment for individuals with SMI.
A teachable moment has been defined as a naturally occurring health event that can increase the effectiveness of an educational intervention to promote risk-reducing health behavior change.18 This study leveraged the teachable moment strategy by integrating an LCS educational intervention with monthly mental health encounters, co-delivered by a mental health provider and a radiologist in a CMHC setting that already provided smoking cessation counseling among individuals with SMI. Therefore, the aim of the study was to develop a tailored LCS educational intervention for individuals with SMI and assess the feasibility and acceptability of delivering this intervention in the CMHC.
Study Setting
This study was approved by the institutional review board. The study was conducted at a quaternary, urban academic medical center in collaboration with a CMHC affiliated with the academic medical center. Although primary care was fragmented across more than 30 practices, all individuals attended monthly visits to the CMHC. Patients are seen regularly for psychiatric care in this clinic and receive ongoing smoking cessation and health maintenance counseling, making this a trusted setting. Furthermore, previous research in this clinic showed that 34% of older adults with schizophrenia met LCS eligibility criteria.15
Study Population
Patients with SMI who were eligible for LCS and seen in the clozapine clinic at the CMHC between April and October 2019 were recruited for the study. The clozapine clinic was selected for two reasons: older adults with treatment-resistant schizophrenia and bipolar disorder (1) need tailored approaches and (2) have visits at the CMHC at least every 4 weeks. The inclusion criteria were (1) ages 55 to 77 years, (2) smoking history of ≥30 pack-years, (3) diagnosis of schizophrenia spectrum disorder or bipolar disorder, and (4) capacity to consent for the study and complete a 15-min previously validated survey [19]. Exclusion criteria were (1) cognitive deficits severe enough to preclude a participant's ability to provide consent, (2) presence of a guardian for medical decision making, (3) history of lung cancer, (4) already enrolled in an LCS program.
Participants were identified through electronic health record review for age and smoking history and were confirmed eligible by a member of their mental health care team during a scheduled monthly appointment. To promote trust, the mental health clinician assessed if the participant could be approached by study staff during a visit, introduced the study staff, and assisted in consent for study enrollment (Fig. 1). Study staff used a verbal consent process to promote understanding about the risks and benefits of participating in the study among prospective participants. Participants were given $20 food gift cards as remuneration.
Phase 1: Qualitative Feedback and Educational Intervention Development
Intervention Development
Using departmental media resources, the research team developed a prototype of the educational intervention based on existing institutional materials, literature review, and expert consensus of the interdisciplinary team (radiology, psychiatry). The intervention consisted of two components: (1) a video demonstrating what to expect during their low-dose CT appointment for LCS with a representative patient volunteer and (2) an educational session to discuss the risks of lung cancer and smoking, LCS process, and myths and benefits of smoking cessation.20
To tailor the intervention for SMI, the intervention was reviewed with two patients with SMI during their scheduled monthly CMHC visit and conducted focus groups (October to November 2018) with three provider groups: mental health clinicians (n = 6), primary care physicians (n = 5) that treat individuals from the CMHC, and thoracic radiologists (n = 9). A semistructured interview guide included five domains specific to individuals with SMI: (1) the LCS referral process, (2) communicating lung cancer risk, (3) communicating the benefits and risks of lung and other cancer screening tests, (4) tobacco cessation, and (5) edits to the educational intervention prototype. The interview guide for radiologists was adapted to reflect unique factors related to current practice (eg, “How do you describe LCS to referring providers?”). Focus groups were audio recorded and transcribed verbatim.
Data Analysis
A rapid qualitative analysis was performed by three study team members (E.J.F., C.R.P., and J.M.N.) in two stages.21 First, data were reduced and organized in summary templates that reflected the five a priori domains of inquiry. Second, content from the summary templates was synthesized and displayed in stakeholder-by-domain matrices to allow comparison across stakeholder types. The educational intervention was adapted based on focus group data and feedback from three volunteer patients not included in the study.
Phase 2: Piloting the Intervention
Intervention Delivery and Evaluation
The study was a single-arm pilot study that delivered the intervention in two iterations (Fig. 2). The first iteration included two 30-min educational sessions that leveraged the patient's monthly appointments to increase convenience. The first session, led by a thoracic radiologist (E.J.F.), focused on the LCS process. One month later, the second session, led by a mental health clinician (C.M.), focused on smoking cessation. The educational sessions included 10 min for questions and answers with study staff and a volunteer peer with SMI who had quit smoking and undergone LCS. To promote feasibility and intervention completion, the intervention was adapted to occur in a single educational session that combined both LCS and smoking cessation education (second iteration). This change was prompted to increase patient convenience because of the logistical barriers including missed appointments, urgent changes to scheduled appointments, and transportation barriers.
Feasibility was defined as ≥50% enrollment among eligible participants and at least 75% intervention completion among participants who enrolled in the study.152223 Participants were considered enrolled after consent and completion of the pre-intervention survey. Acceptability was defined as ≥75% overall satisfaction with the intervention as assessed in the postintervention survey.2223 Satisfaction was measured on a one-item Likert scale (“Overall, I was satisfied with the educational sessions”), with responses ranging from “strongly disagree” to “strongly agree.” Participants were coded as satisfied if they responded “strongly agree” or “agree.”
Attitudes toward lung cancer and LCS
Participants were asked about their lung cancer worry (“How often do you worry about getting lung cancer?”), with response options ranging from “rarely or never” to “all the time.” Additionally, participants were asked about their perceived susceptibility to developing lung cancer (“I am in danger of developing lung cancer”), the prevention benefits of LCS (“There is not much you can do to lower your chances of getting lung cancer”), and early detection benefits of LCS (“Getting checked regularly for lung cancer increases the chances of finding cancer when it's easy to treat”). Response options ranged from “strongly disagree” to “strongly agree.”
Attitudes toward smoking cessation
Participants were asked about the benefits of quitting (“How much would quitting smoking reduce your chances of getting lung cancer?”), with response options ranging from “not at all” to “very much.” Participants were asked about their motivation to quit by responding to the prompt “Please tell me which statement best represents what you think right now.” On a 10-point scale, response options ranged from “I enjoy smoking so much I will never consider quitting no matter what happens” to “I have quit, and I am 100% confident that I will never smoke again.”
For phase 2 analysis, descriptive statistics were used to calculate frequencies and percentages for feasibility and acceptability. Differences between pre- and postintervention measures were examined using paired samples t tests using a two-sided significance threshold of .05 and a Cohen's d to measure the effect size of differences. All analyses were conducted with SPSS version 27 (IBM Corp, Armonk, New York).24
Phase 1
Phase 1 results were assessed by provider-driven feedback for the following domains: LCS referral process, communicating lung cancer risk, communicating the benefits and risks of lung and other cancer screening tests, tobacco cessation, educational intervention prototype feedback. Radiologists, primary care providers, and mental health clinicians echoed various barriers and facilitators within the process of care for patients with SMI. The stakeholders addressed common obstacles of establishing eligibility, patient's denial of risk, prioritizing other comorbidities over lung cancer, and the use of tobacco to combat social isolation. Facilitators among the providers included communication of benefits, clarifying the ease of testing and lack of needles or blood samples, and explaining how to keep belongings safe. Additional details of the feedback provided by each provider group are available in Table 1.
Table 1. Provider feedback by domain matrix for phase 1
| LCS Referral Process | Communicating Lung Cancer Risk | Communicating the Benefits and Risks of Lung and Other Cancer Screening Tests | Tobacco Cessation | Educational Intervention Prototype Feedback |
|---|
| Radiology | | | | Overcoming smoking used as a coping strategy Patients with SMI lack support systems Patients have limited access to outpatient cessation services Social isolation fuels tobacco use
| Context with less text and more figures Need for patient-centered language and an appropriate health literacy level Emphasizing short- and long-term outcomes of cessation
|
| Mental health | | | | | Context with less text and more figures Need for patient-centered language and an appropriate health literacy level Emphasizing short- and long-term outcomes of cessation Indicating a lack of needles or blood samples for LCS Explaining how to keep belongings safe Ensuring LCS does not obviate the need to quit
|
| Primary care | | | Prioritizing discussion of cancer screening Difficulty understanding the need for screening with asymptomatic patients and understanding need for follow-up biopsy Benefits of early detection
| Struggle to counter myths Overcoming smoking used as a coping strategy Ambivalence to smoking in mental health care settings with designated outside smoking areas
| Context with less text and more figures Patient-centered language and an appropriate health literacy level Emphasizing short- and long-term outcomes of cessation Indicating a lack of needles or blood samples for LCS Explaining how to keep belongings safe Ensuring LCS does not obviate the need to quit
|
Phase 2
The initial electronic health record screening identified 89 potentially eligible participants based only on age and smoking history; however, 59 patients did not meet eligibility criteria after individual evaluation (Fig. 3). Among the 30 eligible individuals approached to participate in the study, 15 declined. Feasibility was achieved when 15 participants consented and enrolled in the study (50% enrollment rate). All 15 enrolled individuals completed (100%) the intervention and the before and after surveys (Table 2). The average age of participants was 61.3 (± 3.68) years, 60% (9 of 15) male sex, 87% (13 of 15) White non-Hispanic race, and 67% (10 of 15) current smokers (Table 1). A total of 22 educational sessions were conducted: 14 individual LCS and smoking cessation sessions, 1 group LCS and smoking cessation session (two participants in this session), and 6 combined LCS and smoking cessation sessions (Fig. 2). Ninety-three percent (14 of 15) expressed satisfaction with the intervention, meeting the predetermined acceptability threshold.
Table 2. Patient characteristic (n = 15)
| Variable | Mean (SD) or n (%) |
|---|
| Demographics | |
| Age (y) | 61.33 (3.68) |
| Sex | |
| Male | 9 (60) |
| Female | 6 (40) |
| Race | |
| White | 13 (86) |
| American Indian or Alaska Native | 2 (14) |
| Ethnicity (Hispanic or Latino)* | 1 (7) |
| Highest education level* | |
| Did not complete high school | 5 (33) |
| Completed high school or GED | 2 (13) |
| Some college | 6 (40) |
| Insurance type† | |
| Public (Medicaid or Medicare) | 13 (86) |
| Not recorded | 2 (13) |
| Medical characteristics | |
| Perceived overall health | |
| Excellent condition | 2 (13) |
| Very good or good condition | 8 (53) |
| Fair condition | 3 (20) |
| Poor condition | 2 (13) |
| Tobacco use | |
| Current smoker | 10 (67) |
| Former smoker | 5 (33) |
| Cigarettes per day*,‡ | 20.76 (8.62) |
| Time to first cigarette (under 30 min) | 10 (67) |
| Age started smoking | 16.67 (7.19) |
| Recent quit attempt (<6 months) | 4 (27) |
| On a scale of 0 to 10 rate how confident you are that you can quit smoking or stay quit | 8.21 (3.60) |
Nine participants completed a postintervention survey after the second educational session, and six participants completed the postintervention survey after the combined educational session. Table 3 reports the change in pre- and postintervention survey measures. Across almost all measures, participants reported a trend toward positive change in attitudes about lung cancer, LCS, and smoking cessation after intervention. The only statistically significant change was in participants' worry about developing lung cancer between pre- (mean = 1.79, SD = 0.89) and postmeasures (mean = 2.36, SD = 1.22), t(13) = 2.51, P = .03, d = 0.53 (Table 4).
Table 3. Feasibility and acceptability outcomes
| % (n) |
|---|
| Feasibility outcomes | |
| Participants enrolled | 50 (15 of 30) |
| Participants who completed the intervention | 100 (15 of 15) |
| Acceptability outcomes: “Overall, I was satisfied with the educational sessions.” | |
| Strongly agree or agree | 93 (14 of 15) |
| Neither agree nor disagree | 7 (1 of 15) |
Table 4. Pre- and postintervention comparison of attitudes toward lung cancer screening
| Lung Cancer Attitudes | Preintervention, Mean (SD) | Postintervention, Mean (SD) | Mean Diff | t | P | Cohen's d |
|---|
| Perceived susceptibility to lung cancer | 3.36 (1.49) | 4.14 (0.95) | 0.79 | 1.81 | .09 | 0.48 |
| Worry about developing lung cancer | 1.79 (0.89) | 2.36 (1.22) | 0.57 | 2.51 | .03 | 0.53 |
| Lung cancer screening attitudes |
| Prevention benefits of screening for lung cancer | 2.79 (1.58) | 2.71 (1.44) | −0.07 | −0.27 | .79 | −0.07 |
| Early detection benefits of lung cancer screening | 4.07 (1.34) | 4.60 (0.82) | 0.53 | 1.84 | .09 | 0.47 |
| Smoking attitudes |
| Benefits of quitting smoking | 2.93 (1.14) | 3.50 (0.94) | 0.57 | 1.67 | .12 | 0.44 |
| Motivation to quit smoking | 5.93 (2.92) | 6.57 (2.74) | 0.64 | 1.24 | .24 | 0.32 |
Individuals with SMI experience substantial disparities in lung cancer mortality because of a complex multifactorial combination of socio-economic and health care factors that result in higher smoking prevalence and lower rates of cancer prevention, screening, and treatment services.61525 There is a unique opportunity for radiologists to partner with primary care, mental health clinicians, and community mental health agencies to lead collaborative multilevel outreach interventions that address multiple factors and enhance LCS equity among individuals with SMI. A vital first step in these efforts is tailoring educational material to fit the needs of individuals with SMI, and the findings from this pilot intervention indicate that educational interventions designed for and delivered in settings that meet the needs of individuals with SMI are both feasible and acceptable.
A team-based approach to LCS (Fig. 1) that integrates radiology, primary care, and mental health has the potential to bridge gaps related to care fragmentation between mental health and primary care while increasing LCS engagement among individuals with SMI. Integrated care models in radiology, for example, identifying patients undergoing screening mammography who may be eligible for LCS, have demonstrated promise at advancing LCS efforts among eligible patients.26
Currently, we lack educational material for LCS that are tailored to the needs and capacities of individuals with SMI and their clinicians. This gap has the potential to decrease the frequency and effectiveness of provider-patient communication about LCS for older adults with SMI, which may result in missed opportunities to engage patients in LCS, tobacco cessation, and cancer screenings.25 In this study, community-engaged research principles were employed to ensure stakeholders guided the development of the educational intervention to address the needs and capacities of individuals with SMI.27- 30 This approach can be utilized to adapt health decision aids and other radiology educational material to fit the needs of vulnerable populations and promote health equity.
Participants reported a positive change in attitudes toward lung cancer, LCS, and smoking cessation in postintervention surveys, including a significant increase in worry about developing lung cancer. Increased worry (affective risk response) may be a motivator to increase LCS participation in this patient population and has been associated with greater readiness to stop smoking.31- 34 Although dispositional or trait worry has been associated with risk-avoidant behaviors, past studies, including the NELSON Trial, suggest greater affective risk of developing lung cancer is associated with a greater interest in lung screening.35- 38 Ordering physicians should ensure comprehension of the benefits and risks associated with LCS remain paramount; however, eliciting an affective risk response is an important activating mechanism when individuals appraise their need to screen. For example, the Extended Parallel Process Model contends that without sufficient worry about a health threat, individuals will not act to mitigate the threat.39 Crucially, the Extended Parallel Process Model also contends that without sufficient efficacy in how to mitigate the threat, individuals will undergo a counterproductive fear control process and defensively avoid the need to screen. For participants in this study, we believe the change in worry is positive and reflects internalization of risk and indicates the benefits of tailoring the intervention to meet the needs of individuals with SMI. We also believe that the study did not significantly improve understanding of the prevention and early detection benefits of screening, and thus, future development of the intervention should prioritize how best to translate these important benefits of screening to individuals with SMI to maximize likelihood of undergoing LCS.
Limitations of this pilot study include that it was conducted in a single institution in a relatively small sample that was primarily non-Hispanic White and English speaking. Future studies enrolling a larger, more racially diverse population with longitudinal follow-up will allow us to understand the effect of the intervention on knowledge and attitudes about lung cancer, smoking cessation, and LCS and examine barriers to scaling the intervention. Additionally, the team changed the intervention delivery from two sessions to one session, which could potentially impact postintervention survey results. This change underscores the need to adapt interventions in real time to meet patients where they are and to partner with trusted clinicians in a familiar care setting that will promote patient convenience and trust. These iterative changes are important to measure and describe transparently when designing interventions that have the capacity to be implemented and scaled in community settings. Although the results of this study are for a specific population, there are elements that can be adapted to other underserved populations, including community-engaged research principles and multidisciplinary, team-based collaborative approaches to LCS that promote care coordination. Future studies will include evaluating a streamlined radiology referral pathway to undergo LCS.26
In summary, targeted LCS education and team-based approaches to screening are needed to decrease preventable mortality and prevent widening of disparities in lung cancer outcomes among individuals with SMI. Radiology-led collaborative programs with psychiatry, primary care, and community stakeholders are a feasible, acceptable, and promising opportunity for radiologists to advance health equity in LCS.
Collaborative efforts that leverage team-based approach to screening between radiology and other medical specialties to advance equity in LCS uptake can be offered to vulnerable patient populations that experience multilevel barriers to care.
Tailored educational aids for LCS are needed to prevent the potential limitation in the occurrence and effectiveness of provider-patient communication that could result in a missed opportunity to engage patients in LCS and other health preventative services.
Radiology-led collaborative programs to increase equitable LCS participation among individuals with SMI and other underserved populations represent a feasible and acceptable opportunity for radiologists to advance health equity efforts in LCS and bridge disparities in lung cancer outcomes.