Implementation planning for equitable tobacco treatment services: a mixed methods assessment of contextual facilitators and barriers in a large comprehensive cancer center

Abstract

Tobacco use among cancer patients is associated with an increased mortality and poorer outcomes, yet two-thirds of patients continue using following diagnosis, with disproportionately higher use among racial/ethnic minority and low socioeconomic status patients. Tobacco treatment services that are effectively tailored and adapted to population characteristics and multilevel context specific to settings serving diverse patients are needed to improve tobacco cessation among cancer patients. We examined tobacco use screening and implementation needs for tobacco treatment services to inform equitable and accessible delivery within a large comprehensive cancer center in the greater Los Angeles region.

We conducted a multi-modal, mixed methods assessment using electronic medical records (EMR), and clinic stakeholder surveys and interviews (guided by the Consolidated Framework for Implementation Research).

Approximately 45% of patients (n = 11,827 of 26,030 total) had missing tobacco use history in their EMR. Several demographic characteristics (gender, age, race/ethnicity, insurance) were associated with greater missing data prevalence. In surveys (n = 32), clinic stakeholders endorsed tobacco screening and cessation services, but indicated necessary improvements for screening/referral procedures. During interviews (n = 13), providers/staff reported tobacco screening was important, but level of priority differed as well as how often and who should screen. Several barriers were noted, including patients’ language/cultural barriers, limited time during visits, lack of smoking cessation training, and insurance coverage.

While stakeholders indicated high interest in tobacco use assessment and cessation services, EMR and interview data revealed opportunities to improve tobacco use screening across patient groups. Implementing sustainable system-level tobacco cessation programs at institutions requires leadership support, staff training, on routine screening, and intervention and referral strategies that meet patients’ linguistic/cultural needs.

Lay Summary

Implementation of equitable tobacco cessation services for diverse cancer patients will require understanding the specific needs and referral processes within health care setting context and target populations. In our study, we identified barriers to implementing a tobacco cessation program for diverse cancer patients (e.g., Asian/Asian American, Black/African American, Hispanic/Latino/a). Barriers noted by clinic team members to routine tobacco use screening and treatment included limited time during patient visits, lack of clinic team training on smoking cessation needs, language/cultural barriers for patients, and insurance coverage. Our findings showed health system leaders, providers, and staff agree that both tobacco use screening and providing tobacco cessation services are important, but there is a need for better understanding and improvement of clinic workflows, designated roles, and responsibilities of providers and staff, and increased awareness and training about tobacco use screening, available cessation services, and referral to treatment.

Implications

Practice: Equitable implementation planning can be informed by multi-modal assessments that indicate: (i) where targeted efforts are needed (screening, clinic team training) and (ii) engagement and support by inner (leadership) and outer (community partners) context stakeholders.

Policy: Employing a health equity lens at the implementation planning stage, within systems and communities, is critical to addressing the systemic barriers to tobacco use screening and referral.

Research: Research that informs multilevel approaches to optimizing implementation to achieve health equity is required at the system, provider staff, and patient levels.

BACKGROUND

Persistent cigarette smoking and other tobacco use by cancer patients is causally linked to an increased all-cause and cancer-related mortality, risk for secondary primary cancers, cancer treatment toxicity, and attenuated response to cancer treatment [1–3]. Quitting tobacco use after a cancer diagnosis can greatly reduce these negative outcomes, resulting in reduced risk of new cancers and longer post-treatment survival [4]. However, despite the known risks for smoking and the benefits of quitting, a relatively large proportion of cancer patients who smoked prior to diagnosis continue to smoke after diagnosis [5]. Thus, cancer patients need tobacco treatment; one important opportunity for delivering treatment is in the context of oncology visits, which can be “teachable moments” to encourage lifestyle changes.

Based on a 2009 survey of 58 NCI-Designated Clinical and Comprehensive Cancer Centers, tobacco use treatment was not considered a part of standard cancer care in the majority of cancer centers [6]. Efforts to implement systematic tobacco use screening and treatment significantly increased in 2017, when NCI launched the Cancer Center Cessation Initiative (C3I), a nationwide Cancer Moonshot effort to increase the reach and effectiveness of smoking cessation treatment at cancer centers using the principles and methods of implementation science. Key features of C3I include a population-level approach (i.e., every patient is assessed for tobacco use and offered treatment), a systems-based approach (i.e., using electronic health record technology to facilitate program integration and enhance long-term sustainability), and an evidence-based approach (i.e., providing a range of known effective treatment approaches including referrals to quit lines, pharmacotherapy, behavioral treatment, etc.). Prior cancer centers within C3I reported increasing patient tobacco use screening to approximately 96% and engaging approximately 25% of tobacco-using patients in an evidence-based treatment [7], which was substantially higher than the <15% typically observed in patient settings [8, 9]. Other emerging studies also indicate effective strategies require a multilevel approach, including training providers [10–14]. While effective interventions for tobacco cessation among oncology patients are available [15, 16] and health system-based tobacco cessation referral programs have been established throughout NCI cancer centers in recent years [10, 17–23], few studies to date have examined the implementation needs of routine tobacco cessation services settings with diverse patient race/ethnicity, linguistic, and social needs, where tailoring and adaptation of evidence-based programs are likely required.

Racial/ethnic minority status, low socioeconomic status (SES), and LGBTQ+ status individuals are at high-risk for tobacco use and suffer disproportionately from tobacco-related illness and death [24–26]. While rates of tobacco-related outcomes fare better in California compared to the USA as a whole, gains in tobacco-related outcomes have not been equitably distributed across population groups [27], due in part to structural and systemic inequities. Tobacco is targeted in more low SES communities, and there is greater point-of-sale marketing in neighborhoods with more Black residents [28]. Racial/ethnic minority adult tobacco users (especially Asian/Pacific Islanders and African Americans) are more likely to initiate use in young adulthood [29–34], and use remains disproportionately high among those of low SES compared to high SES [35–37]. Regarding cessation efforts, women have more difficulty maintaining long-term abstinence than men [38], and previous research indicates that a significantly smaller proportion of African Americans had quit for at least 1 year (30.4%) compared with other groups (39.8% for Asian Americans; 36.6% for Hispanics/Latinos; 42.9% for Non-Hispanic Whites) [39]. Successful implementation planning of tobacco cessation services within oncology settings serving diverse populations will require effective tailoring and adaptation to the needs of the target patient population and multilevel context [40–43]. Emerging evidence in non-oncology clinical settings suggests systematic implementation of tobacco cessation interventions is often suboptimal and exposed to a variety of multilevel barriers [44, 45]. Thus, there is a vital need to comprehensively assess the local community health system context to effectively employ a health equity lens in implementation planning for tobacco cessation services [46, 47].

Since 2017, 52 centers have participated in the NCI Cancer Center Cessation Initiative (C3I) Moonshot initiative [48]. The [BLINDED] Cancer Center joined the initiative in 2020 to build capacity and infrastructure for tobacco cessation treatment services for patients and survivors in the greater Los Angeles region. This manuscript describes a mixed methods implementation planning study that examines the feasibility, acceptability, and infrastructure requirements needed to implement a tobacco cessation program within a large comprehensive cancer center serving one of the most diverse patient populations in the country. The goal of the study is to focus on identifying the tailoring and adaptation of tobacco requirements for tobacco use screening and cessation services within this context.

METHODS

Study aim

We conducted a multi-model, mixed methods assessment [49] (concurrent QUANT + QUAL design) of contextual health system-level facilitators and barriers to systematic tobacco use screening and implementation of tobacco cessation services within the [BLINDED] Cancer Center, which serves a diverse cancer patient and survivor population in the greater Los Angeles region. The study included three components: (i) electronic medical records (EMR) assessment of routine tobacco use screening by patient sociodemographic factors; (ii) quantitative survey of oncology providers, clinic staff, and administrators on motivation and perceptions of tobacco service needs; and (iii) qualitative interviews with clinic stakeholders (providers and staff) at two oncology outpatient sites. All procedures were approved by the authorship team’s Institutional Review Board (IRB UP-20-01168).

Tobacco screening data in the EMR

We examined EMR data from all patients who had a least one visit to all [BLINDED] Cancer Center clinics between January and June 2019 (N = 26,030). All data, obtained from the Cerner-based EMR system, were de-identified and provided by the Clinical Research Informatics unit at the [BLINDED] Institute. Tobacco use history is available through the Social History section of the medical records, which includes: (i) type of tobacco product used (e.g., cigarettes, oral, hookah or pipe, vaping or e-cigarette, other type); (ii) current tobacco user status (e.g., never smoker, 4 or less a day, 5–9 a day, 10 or more a day, former smoker); (iii) age of tobacco use onset; (iv) age of successful cessation; and (v) readiness to quit.

Each patient’s smoking status was initially coded as: never smoker, former smoker, current smoker (comprised of anyone who reported any combustible cigarette use), or missing. For the purposes of this study, where the primary interest is whether smoking status was recorded in the EMR, we re-coded smoking status information as: status present (coded as “0”) versus status missing (coded as “1”). Additionally, each patient record provided data on several sociodemographic factors: race/ethnicity (coded as Asian, Black, Hispanic/Latino, White, Other), sex/gender (coded as Female, Male), age group (coded as 65 and older, younger than 65), and primary insurance as measure of socioeconomic status (coded as Private Insurance, Medicare, Medicaid/None).

Clinic survey assessing motivation and perceptions of tobacco service needs

We administered a brief self-administered online survey to assess current procedures and baseline motivation of health system stakeholders (physicians, nursing staff, administrators, other clinic staff members) to screen for tobacco use, as well as perceptions of providers’ role in providing treatment and the need for tobacco treatment as part of cancer care. A total of 45 [BLINDED] oncology providers and clinic staff were identified through staff lists, from a range of clinics (Women’s Health, [BLINDED] Ambulatory Clinic, Radiation Oncology, Other) and roles (Physician, NP/PA, RN, MA/LVN, Patient Navigator, Scheduler, Other) and contacted via email to participate in the 14-item survey. Of the 45 physicians and clinic staff recruited for the quantitative survey, 32 completed the self-administered survey on existing tobacco use screening processes and perspectives on system-level needs for tobacco cessation services (71%). Three participants were drawn at random to receive a $25 Amazon gift card.

The surveys were provided via REDCap link by email in July 2021 to all identified providers and staff. Research team members sent weekly reminders until the survey data collection window closed in October 2021. The survey included items on current tobacco use assessment practices at clinical outpatient units. Current assessment practices were measured by asking about systems put in place, frequency of, and process for assessing patients’ tobacco use, in addition to who is responsible for documenting tobacco use in the EMR and awareness of tobacco cessation referral options. The survey also included ratings (strongly disagree to strongly agree) on the importance of routinely assessing tobacco use and providing tobacco cessation options as a part of cancer care. Participants rated whether their unit is equipped for implementation of tobacco use screening and referral right now. Participants also rated the extent that clinic leadership is supportive of systematic tobacco use assessment and EMR documentation and if leadership expects the participant to advise patients on quitting.

Quantitative data analysis of EMR and survey data

. For analysis of the EMR data, we conducted logistic regression models including each demographic characteristic (gender, age, primary insurance, race) as main effect predictors, as well as all possible two-way interactions between each demographic characteristic. The primary outcome was missingness of tobacco use (smoking status) in the EMR record (status present versus status missing). Results of the logistic regression analysis are presented as odds ratios. Predicted probabilities of smoking status missingness for all main effect and interaction groups were calculated and presented for ease of interpretation. Significance was set at the p < 0.05 level. All analyses were conducted using SPSS 28.0. For the survey data, descriptive results are presented in frequencies and means.

Qualitative interviews with clinic providers and staff

We conducted one-on-one in-depth interviews with providers and staff in two oncology outpatient clinics between June 2021 and September 2021. The research team identified two physician champions, one from medical oncology and one from pulmonary/thoracic surgery, who connected the research team to clinic managers in their outpatient clinics. Managers then provided contact information of potential providers and staff for participation. One graduate-level research team member emailed clinic staff members to provide study information and ask for participation via in-person interviews, with the goal of reaching 6–8 participants per site.

Interview guides consisted of semi-structured, open-ended questions based on domains in the Consolidated Framework for Implementation Research (CFIR), including inner and outer setting factors that impact implementation of tobacco cessation services, planned referral and cessation services, and aspects of the participants’ clinic roles and perspectives on tobacco cessation needs for their patient population. In-person one-on-one interviews, with clinic staff members, within the same clinic took place on consecutive days for approximately 30 min each. Providers were interviewed virtually, via Zoom, by the same graduate-level research team member. Following all interviews, participants were asked the following demographic questions: race, ethnicity, gender, age, and length at organization. All participants received a $25 gift card upon interview completion. Audio files of all interviews were sent to an external service for transcription and then de-identified by the research team.

Qualitative data analysis.

The qualitative data analysis was guided deductively by tobacco treatment topics and CFIR framework domains covered in the interview guide as well as inductively through emergent themes. Core members of the research team, comprised of those with expertise in tobacco control, health services research, social work, and clinical psychology, met bi-weekly to discuss emergent themes and develop appropriate codes through an immersion/crystallization process. Two independent coders from the core team thoroughly read and coded all interview transcripts using the codebook. We used Miles and Huberman’s process of creating matrices to display the data for comparative analyses [50] where one team member extracted all data relevant to each code. The team continued to meet to discuss discrepant coding and modify the codebook accordingly throughout analysis period. Both coders met at the conclusion of coding to discuss major themes and finalize analysis of interviews. Measurement of coding agreement was calculated on a random subset of two interview transcripts, with average agreement of approximately 70%.

Integration of quantitative and qualitative data.

Following completion of the quantitative and qualitative data analysis, the team used a mixed methods approach to integrate the qualitative data (from interviews) with the quantitative data (survey and EMR data) to systematically compare the findings on facilitators and barriers to tobacco use screening and treatment referral across levels of context (patient factors from EMR, motivation by health system stakeholders from surveys, current clinic processes for screening from qualitative interviews) [49, 51] and then used to guide recommendations for future interventions. Qualitative and quantitative findings were merged and compared through ongoing research team meetings [52], presented in a joint display table format [53], and reported to health system/cancer center leadership and the broader research team in Spring 2022 for further interpretation and discussion of intervention tailoring and adaptation needs.

RESULTS

Tobacco use history missingness in the EMR social history form

Approximately, 45% of patients (n = 11,827) with at least one cancer center visit between January and June 2019 had missing tobacco use information (i.e., there was no tobacco use history documented in the Social History form). Logistic regression results revealed tobacco use missingness was associated with several demographic characteristics, including main effects of age and insurance status, as well as several interactions between demographic characteristics (Table 1). For example, there were overall differences in predicted probabilities of missing tobacco use data as function of primary insurance (Table 2). Patients with Medicaid/no insurance were more likely to have missing data relative to those with private insurance (61.6% vs. 44.0%). The significant insurance × race interaction indicated that the gap between Medicaid/No insurance and Private insurance differed by race. While the gap between Medicaid/No insurance and Private Insurance was 9.1% for White patients, the gap was 24.2% for Asian patients, and 28.1% for Black patients. Similarly, there was a significant insurance × gender interaction, indicating that Female patients had a larger insurance-related data missingness gap (19.5%) compared to Male patients (13.5%). There was also a significant race × gender interaction, indicating that the data missingness gap between men and women differed by race. That is, while the gap between White men and women was relatively small (0.2%), there was a substantial gender gap for Asian patients (i.e., 12.3% greater predicted probability for women) and for Black patients (i.e., 6.8% greater predicted probability for men).

Table 1

Logistic regression results (presented as odds ratios) of smoking history data missingness as a function of gender, age, primary insurance, and race/ethnicity

		Odds of missing data
	N	OR	(95% CI)	p
Gender
Male (reference)	12,024	1.00	–	–
Female	14,002	0.92	(0.78, 1.07)	0.28
Age
65 and older (reference)	12,082	1.00	–	–
Younger than 65	13,948	1.48	(1.28, 1.71)	<0.001
Primary insurance
Private insurance (reference)	12,034	1.00	–	–
Medicare	11,400	0.84	(0.73, 0.97)	0.02
Medicaid/none	1,631	1.55	(1.08, 2.24)	0.02
Race/ethnicity
White (reference)	10,795	1.00	–	–
Asian	3,639	1.06	(0.82, 1.37)	0.67
Black	1,068	0.86	(0.60, 1.25)	0.43
Hispanic/Latino	5,147	0.80	(0.64, 1.00)	0.50
Other	5,381	2.08	(1.66, 2.60)	<0.001
Gender × age
Female, Younger than 65		0.83	(0.71, 0.96)	0.02
Gender × primary insurance
Female, Medicare		1.22	(1.04, 1.42)	0.01
Female, Medicaid/none		1.41	(1.14, 1.75)	<0.01
Gender × race/ethnicity
Female, Asian		1.66	(1.42, 1.95)	<0.001
Female, Black		0.74	(0.57, 0.98)	0.03
Female, Hispanic/Latino		0.96	(0.83, 1.11)	0.53
Female, Other		1.18	(1.02, 1.35)	0.02
Age × primary insurance
Younger than 65, Medicare		0.55	(0.47, 0.65)	<0.001
Younger than 65, Medicaid/None		0.63	(0.45, 0.87)	0.01
Age × Race/Ethnicity
Younger than 65, Asian		0.79	(0.62, 1.02)	0.07
Younger than 65, Black		1.09	(0.77, 1.55)	0.63
Younger than 65, Hispanic/Latino		1.28	(1.04, 1.58)	0.02
Younger than 65, Other		0.88	(0.71, 1.10)	0.27
Primary insurance × race/ethnicity
Medicare, Asian		1.22	(0.95, 1.57)	0.13
Medicare, Black		1.17	(0.82, 1.66)	0.40
Medicare, Hispanic/Latino		1.01	(0.81, 1.25)	0.94
Medicare, Other		0.96	(0.77, 1.20)	0.71
Medicaid/None, Asian		1.92	(1.31, 2.83)	<0.001
Medicaid/None, Black		2.20	(1.30, 3.73)	<0.01
Medicaid/None, Hispanic/Latino		1.32	(0.99, 1.76)	0.06
Medicaid/None, Other		1.09	(0.80, 1.49)	0.58

Table 2

Predicted Probabilities of data missingness for tobacco use in EMR

	Main effect	Two-way interactions
		Gender		Age		Primary insurance
		Male	Female	65 and older	Younger than 65	Private	Medicare	Medicaid/none
Gender
Male	46.7	–	–	–	–	–	–	–
Female	48.3	–	–	–	–	–	–	–
Age
65 and older	48.2	46.3	50.1	–	–	–	–	–
Younger than 65	46.8	47.2	46.5	–	–	–	–	–
Primary insurance
Private insurance	44.0	45.4	42.7	40.6	47.5	–	–	–
Medicare	36.9	35.8	38.0	40.4	33.4	–	–	–
Medicaid/no insurance	61.6	58.9	64.2	63.5	59.6	–	–	–
Race/ethnicity
White	41.3	41.4	41.2	41.8	40.7	41.1	32.5	50.2
Asian	52.6	46.5	58.8	56.1	49.2	45.9	41.9	70.1
Black	43.0	46.4	39.6	42.5	43.5	35.2	30.5	63.3
Hispanic/Latino	40.8	41.4	40.1	38.5	43.1	38.3	30.0	54.1
Other	59.9	58.0	61.8	62.1	57.7	59.7	49.6	70.3

Note: Bolded cells indicate significant associations in the logistic regression

Clinic survey assessing motivation and perceptions of tobacco service needs

Survey respondents (n = 32) were mostly providers (physicians, NPs) (n = 9) and clinic staff (RN/MA/LVN) (n = 13) and majority worked in Women’s Health, Dermatology, and Ambulatory clinics. Overall, results indicate that health system stakeholders agreed that it is important to routinely assess patient’s tobacco use (84% agree or strongly agree) and provide tobacco cessation services as part of cancer care (80% agree or strongly agree). However, more than one-third (38%) reported not having or not knowing whether a current system is in place to assess patient’s tobacco use within the clinic settings and more than half (52%) strongly disagreed or disagreed that their unit was equipped to implement tobacco screening and referral services currently. Importantly, more than one-third of participants did not agree that there was support for or expectations by clinic leadership for routine tobacco use screening and documentation.

Qualitative interviews with providers and clinic staff

Thirteen participants completed in-depth interviews (Table 3) from two purposively sampled oncology outpatient clinics, including 4 providers (physicians, PAs) and 9 clinic staff (RN/LVN/LPN, clinic manager, patient care assistant, nurse specialist). Three-fourths of the interview participants were female and half of the respondents (50%) had been working at [BLINDED] for more than 5 years. About one-third identified as Hispanic or Latino/a and half (50%) identified as White. Themes from interviews are presented and organized by CFIR domains: inner setting, outer setting, intervention characteristics, and characteristics of individuals, and compared and integrated with quantitative findings described above in Table 4.

Table 3

Characteristics of participants from qualitative interviews (N = 13)^a

Demographic characteristics	N	%
Total	13	100
Clinic role
Physician	2	15.4
PA	2	15.4
RN	3	23.1
LVN/LPN	3	23.1
Other/Unknown	3	23.1
Self-reported race
Asian American/Pacific Islander	3	23.1
Middle Eastern	1	7.7
White	6	46.2
Unknown	3	23.1
Ethnicity
Hispanic or Latino/a	4	30.8
Non-Hispanic or Latino/a/Unknown	9	69.3
Gender
Female/feminine	9	69.3
Male/masculine	3	23.1
Other/unknown	1	7.7
Age (years)
18–29	1	8.3
30–39	2	16.7
40–49	2	16.7
50–59	6	50.0
60+	1	7.7
Length at organization (years)	12
<1	1	(8.3)
1–5	5	(41.7)
5–10	5	(41.7)
>10	1	(8.3)

^aMissing demographics from one participant who declined to respond.

Table 4

Key findings/recommendations for intervention, based on CFIR domains, qualitative, and quantitative data

CFIR domain	Qualitative interview data	Quantitative survey and EMR data	Intervention recommendations
Inner setting	Agreement across providers/staff that routine tobacco use screening is important, but unclear perceived roles and responsibilities of who should collect the information. (Theme 1) Many providers/staff believed provider/staff training or education would improve routine tobacco use screening in their clinic setting. (Theme 2)	SURVEY: •84% agree or strongly agree “it is important to routinely assess patient’s tobacco use.” •80% agree or strongly agree “providing tobacco cessation services is part of cancer care.” •35% did not agree “there was support for or expectations by clinic leadership for routine tobacco use screening and documentation.”	Training staff and other team members on evidence-based tobacco screening strategies that are feasible within oncology clinic settings.
Outer setting	Patients’ cultural acceptance in discussing tobacco use and English language proficiency were reported barriers. (Theme 3) Patients’ social needs were cited barriers to accessing tobacco cessation service. (Theme 4)	EMR: •Among Asian patients, there was 12.3% greater predicted probability of data missingness for women compared to men. •Among Black patients, there was 28.1% greater data missingness for uninsured compared to privately insured patients. •Close to half of patients are Hispanic, Black, Asian, or other race and ethnicity.	Engagement with community members and policy partners to identify community needs and appropriate strategies. Providing screening and tobacco cessation materials in multiple languages.
Intervention characteristics	Consistent workflows and system-wide policy could improve tobacco use screening and referral. (Theme 5) Increasing providers/staff awareness of available tobacco cessation services within the system. (Theme 6) Input and partnership with community stakeholders and patient advocates can inform optimal adaptation needs of existing evidence-based programs to fit the current context. (Theme 3)	SURVEY: •38% reported not having or not knowing whether “a current system is in place to assess patient’s tobacco use within the clinic settings.” •52% strongly disagreed or disagreed that “their unit was equipped to implement tobacco screening and referral services currently.”	Increasing awareness among providers and clinic staff that the Cancer Center has existing resources for tobacco treatment. Facilitate established workflows and protocols for screening and referral.
Characteristics of individuals	Providers/staff had limited awareness of smoking cessation guidelines for cancer patients. (Theme 7) Providers/staff indicated established workflows and responsibilities for tobacco use screening and cessation referral would improve their comfortability in tobacco use discussions with patients. (Theme 8)	SURVEY: •50% of participants have worked at cancer center for less than 5 years.	Emphasizing during intervention training that smoking cessation is part of NCCN guidelines Providing information about how tobacco cessation will improve treatment of all cancers and prevent secondary cancers

Inner setting

Two themes emerged for inner setting, or context within the clinic, that inform the need to clearly delineate roles and responsibilities within clinic teams and provide education to team members to optimize routine tobacco use screening and cessation referral.

Theme 1: providers/clinic staff agreed routine tobacco use screening among cancer patients is important, but were unclear about roles and responsibilities

Clinic team members recognized the importance of systematic and routine tobacco use screening. Some providers and staff described how important their role is in assessing tobacco use and supporting patients without support/resources for quitting. Defined workflows for assessment of tobacco use after the first patient encounter were lacking in both clinics. Some providers further recognized their role in the lack of a routine screening process:

“…I think I could do a better job of reassessing, especially postoperatively…I can definitely do a better job of checking in on patients…” (PA, Clinic 2).

Most providers were aware of the need for follow-up with patients to reassess tobacco use. Overall, providers and staff responses illustrated how there was agreement on the need to routinely assess tobacco use among patients but workflow and designated roles across clinic team members were not clear.

Theme 2: most providers/clinic staff believed provider/staff training or education would improve routine tobacco use screening in their clinic setting

In both clinics, providers and staff were asked for what they believed would improve routine screening of tobacco use and were prompted with examples of leadership support, incentives, designated staff time, and staff training or education. Most providers and staff reported a desire for training/education: “…I [want to] know how to successfully—be more convincing, I guess, about what I [want to] share and educate my patients on” (PA, Clinic 2). Providers and staff also shared why clinic staff need more guidance in how to discuss tobacco use with patients: “…I don’t think the nurses know how to engage in a conversation about smoking cessation … they don’t go through the six A’s of smoking cessation. I don’t think [it] comes natural…” (Nurse Specialist, Clinic 1).

Providers expressed wanting to be impactful in their discussions about tobacco use and cessation with patients and showed an openness to learning more, further highlighting the need for additional training and education. Providers and staff described wanting to know the purpose of routine screening and have a clearer workflow in their clinics. Providers and staff also reported a need for additional training in how to use the EMR to document screening.

Outer setting

Two themes for outer setting, or clinic’s external environment, became apparent that highlight the need to address barriers to tobacco use screening and referral and access to cessation services relating to culture, English language proficiency, and social needs of patients.

Theme 3: patients’ cultural acceptance in discussing tobacco use and English language proficiency were reported barriers to address in tobacco screening and referral

Clinic staff and providers discussed barriers to tobacco use assessment and access to cessation treatment for racial/ethnic minority groups (e.g., Armenian, Chinese, Korean, and Hispanic/Latino/a). English language proficiency and cultural differences were factors noted as impacting patients’ comfortability in discussing tobacco use and normalization of smoking behaviors leading some patients to downplay their usage. Providers and staff shared challenges of utilizing a translation service during appointments, and also noted potential barriers to providing linguistically accessible tobacco cessation services to meet patients’ needs. One provider highlighted the impact of provider-patient cultural concordance on receptivity to cessation counseling and need to recognize community-specific, culturally tailored messaging:

I think someone [physician]… who’s Persian sees a Persian patient, understands the context of Persian culture, he can relate to them and perhaps frame it in a way for them to better appreciate the significance of smoking cessation. (Physician, Clinic 2)

Theme 4: patients’ social needs were cited barriers to accessing cessation services

Clinic staff and providers named financial hardship, including type of health insurance, as well as transportation to and from healthcare services, as barriers to accessing tobacco cessation treatment for racial/ethnic minority patients:

Insurance is the big one. Transportation, jobs, just daily life. But I would say the insurance is probably the big one…And finances. Some people just can’t afford [it]. (RN, Clinic 1)

One clinic staff member elaborated on issues related to health insurance and access to treatment specifically for patients with public health insurance: “…when it comes to the smoking cessation stuff, they [public insurance] don’t provide that too often...” (LVN, Clinic 2). Providers and staff stressed the importance of tobacco cessation treatment for low-income patients and older adults with Medicaid (Medi-Cal) and/or Medicare insurance, and discussed the complexity of barriers due to cost and transportation limiting access to treatment. Overall, social needs of racial/ethnic minority and low-income patients emerged as perceived barriers to accessing cessation treatment.

Intervention characteristics

Two themes pertaining to intervention characteristics, or attributes of routine screening and referral, were noted that emphasize the importance of setting up clear clinic workflows for tobacco use screening and referral and making staff aware of available cessation services.

Theme 5: workflow for tobacco use screening varied across clinic settings

While some providers and clinic staff could describe current tobacco use screening processes for patient visits, many could not specify who on the clinic team was a part of the workflow and few mentioned a routine screening process. Responses also varied for how often tobacco use screening occurred during patient visits and what current protocols were for screening. While some staff thought screening was everyone’s job, some providers did not report a formal tobacco use screening process for who was to screen and when. One provider described screening falling to PAs: “…PAs are integral to our practice, and they know to ask about the smoking history…they end up by default screening” (Physician, Clinic 2). Despite having a designated EMR form for ascertaining tobacco use status, roles and agreed upon workflow for tobacco use screening, and identifying patients for referral to tobacco cessation services, were not established across clinics.

Theme 6: few providers/clinic staff were aware of available tobacco cessation services

Providers and clinic staff exhibited some knowledge of tobacco cessation services offered within the cancer center. However, most providers and staff lacked information about the services that were currently offered so many did not refer patients to cessation treatment services already available in their health system. Providers and staff expressed interest in receiving updated and accessible information on cessation services where they could refer patients, which was also discussed when they reported wanting additional training and education to improve screening. Some providers and staff did offer medication as an option for cessation treatment to patients, and some would refer patients to existing programs in primary care or social work:

…we will refer them back to their PCP to discuss the different options…we’ve actually given patients the tobacco cessation, like the patch. That’s pretty much all I know that we would give. I don’t know of any programs … (LVN, Clinic 2)

Characteristics of individuals

Two themes related to characteristics of individuals (team members in the clinic) point to opportunities for training to increase awareness of smoking cessation guidelines specific to cancer patients, and improve comfort discussing tobacco through clearly delineated workflows.

Theme 7: providers/staff were less aware of cessation guidelines for cancer patients

Clinic staff and providers displayed a lack of knowledge about specific guidelines for smoking cessation for cancer patients, such as guidelines from the National Comprehensive Cancer Network. Additionally, providers expressed limited knowledge about the specifics of guidelines for cancer patients: “Only to the extent for lung cancer. That’s it. Even then, I still always need a refresher course…” (Physician, Clinic 2). Overall, providers and staff in both clinics indicated a willingness to learn more about guidelines for smoking cessation for cancer patients, further highlighting the need for additional education and training.

Theme 8: providers/staff indicated established workflows and responsibilities for screening and referral would improve comfortability in discussions with patients

Providers and staff mostly viewed their interactions as positive with patient smokers, yet the level of interaction with patients about tobacco use varied. Few providers or staff were both screening and discussing tobacco use with patients. Some would screen for tobacco use but not discuss use, or some would not screen but did discuss use. One provider described interactions with their patients who refused to quit smoking: “…when they tell me ‘Hell, no.’ They’ll say that in a general good-natured way, and I won’t usually push too far beyond that…” (Physician, Clinic 1). Another provider did report pushing patients to discuss their usage in order to promote cessation, based on the strength of the patient–provider relationship.

A few of the providers and staff described negative interactions with patient smokers and also reported limiting discussions due to discomfort: “Some patients I know not to ask [about tobacco use] …I don’t push too much because the patients don’t like us to…they told me, ‘I’m going to do it,’ and that’s it” (LPN, Clinic 1). The core concerns, however, did not appear to be providers and staff feeling uncomfortable when bringing up tobacco, rather that there was no formal process in place for how to screen and when or how to discuss use in either clinic.

DISCUSSION

This multi-modal assessment of current tobacco use screening informs several opportunities for establishing capacity and infrastructure to best optimize the implementation of tobacco cessation services across contexts serving diverse populations. First, we observed opportunities for improvement in documentation of tobacco use status in patients’ EMR records (only 55% of patients had any smoking status/tobacco use information), and targeted efforts in equitable documentation of tobacco use among patients by gender, race/ethnicity, insurance, and age, which may include provider and clinic team trainings and in-language support services to address patients’ linguistic, cultural, and social needs. While we observed no data missingness differences between women and men, or among race/ethnicity groups, there were several interactions between gender, race/ethnicity, and insurance status that provide insight into how demographic-based discrepancies might be addressed by a clinic-level intervention to improve tobacco use screening and documentation. For example, the finding that there was a substantial gender gap for Asian patients (i.e., 12.3% greater predicted probability of data missingness for women) may indicate that clinic staff have a cognitive bias towards believing that Asian women patients are unlikely to smoke. In addition, insurance gaps among Asian and Black patients (24.2% greater data missingness for uninsured Asian patients and 28.1% for uninsured Black patients) may indicate an institutional bias against patients with relatively few financial resources, which may be exacerbated by persistent implicit racial biases. Use of evidence-based interventions tailored to specific communities could emphasize both quick and efficient tobacco screening strategies, as well as referral to low- to no-cost, effective tobacco treatment resources (e.g., the state quitline). Implementing an equitable tobacco cessation program for cancer patients in a diverse setting, such as the one in this study, will require equitable identification of patient need and referral access to cessation programs.

Second, our study also highlights the importance of addressing cultural and social needs of patients, particularly in settings serving patients from diverse cultural, linguistic, and socioeconomic backgrounds. Language and cultural backgrounds of diverse patient populations may impact provider and staff abilities to successfully screen for tobacco use or communicate the importance of tobacco cessation. Our findings indicate that provider and staff training on recommended guidelines for tobacco cessation among cancer patients is needed, which will likely require adaptations to account for a range of communication strategies, appropriateness of messaging to patients, and addressing community-specific barriers and stigmas, in order to serve patients from multiple racial/ethnic, cultural, and linguistic communities. These adaptations can be optimally developed with outer context stakeholders, including cancer center community advisory boards and community partners, to ensure program fit and sustainability. Tobacco cessation services will need to provide equitable options based on health literacy, transportation, appointment times, and cost/insurance coverage to better serve all patient groups. Settings with patients from diverse cultural backgrounds, languages, and social norms around smoking will require targeted messaging and tobacco cessation counseling that meets the unique needs of communities. Addressing health-related social needs within health care settings has gained focus in recent years and screening for social needs and social risk may be an important consideration when designing implementation of system-level programs that would benefit the most vulnerable, high-risk patient groups [54–58]. In the context of tobacco cessation, potential adaptations to address diverse patient needs can build upon work previous work described in NCI Evidence-Based Cancer Control Programs, such as manualized advice for clinicians on how to build rapport and conduct functional tobacco screening with diverse populations [59], and culturally tailored booklets to educate patients on cessation strategies [60].

Third, prior studies have observed system-level implementation of tobacco cessation programs for cancer patients will entails low-cost, efficient systems that take advantage of EMR tobacco use history to identify all patients who smoke without the need for time-consuming manual referrals [61, 62]. While there was consensus across stakeholders that providing tobacco cessation services is important, findings from the EMR analysis and the qualitative interviews indicate potential to examine and refine clinic workflows, designated roles and responsibilities, and improve awareness and training for tobacco use screening and referral. These findings align with prior studies on implementation of tobacco cessation programs that indicate addressing systemic and organizational barriers are necessary for longer term adoption and maintenance [45, 63]. Several strategies have been previously documented as system- or provider-level interventions that can result in improved tobacco cessation outcomes, including identifying system leadership support, having physician champions engage with peers within the system, implementing standards of care and other system policies, providing incentives for practice change, delivering provider and staff trainings to improve systematic tobacco use screening and referrals, covering dedicated staff to provide tobacco cessation services, and use of EMR to monitor and provide feedback to clinic teams [64–66]. Thus, our findings are consistent with the need to use multiple implementation strategies within the system, across providers, clinic staff, and clinic sites, to maximize routine use of screening and treatment services in diverse patient populations.

Finally, providers and staff in our study indicated the need for increased training to clarify the roles and responsibilities of clinic team members within their workflow to facilitate routine implementation of tobacco screening and service provision for cancer patients. The American Society of Clinical Oncology referenced the need for internal champions to achieve system-wide adoption and implementation of tobacco cessation programs. Having physician or other system-level champions will be important for buy-in across stakeholder groups within a large health system setting, as evidenced in prior studies [67]. Furthermore, establishing roles and responsibilities across departments and within clinic teams for tobacco use screening and referral, as well as having dedicated staff to identify and follow-up with patients who need tobacco treatment services, and system-level informatics and data support for monitoring feedback and uptake of treatment services, will lead to overall system improvements. Ongoing assessment of EMR data for tobacco screening and data missingness will also identify specific targets for equitable uptake.

While our study provides a comprehensive assessment of existing capacity and implementation needs for system-level tobacco cessation services in an oncology setting, some limitations should be noted. First, tobacco use history items are not mandated and enforced to be completed during a patient visit, introducing the possibility that even if tobacco use was assessed during the clinic visit, patient’s tobacco use history may not be stored in their EMR, or may be stored outside of the Social History (e.g., in a free-text clinic note) thus making this information inaccessible to data queries. In addition, prior studies have documented underreporting of tobacco use among cancer patients. Second, our EMR assessment of tobacco use status was based on 2019 data. More recent data may have differing patterns due to disruptions from the COVID-19 pandemic which disproportionately impacted racial/ethnic minority and low-income groups. Furthermore, due to limitations of the EMR data set, we could not examine the reason(s) for visits or patients’ medical histories, and rather just examined whether tobacco use status was present for any patient. Thus, we could not take into account severity of disease or other factors that may impede tobacco use screening. Third, health insurance status of patients and other insurance-based factors were noted as potential barriers to routine tobacco use screening and treatment referral. Managed care penetration and tobacco treatment coverage in the study region may be different from other regions in the USA. Finally, participants in our survey and qualitative interviews may not be representative of all clinic stakeholders as we interviewed clinic team members from only two outpatient settings within a large health system. While we iteratively assessed the qualitative data for thematic saturation, our qualitative data relied on a small sample from two clinics. Nonetheless, our qualitative interviews were purposively sampled from two different outpatient settings and we aimed to compare responses across clinics and within sites.

CONCLUSIONS

Opportunities for implementation of health system-wide sustainable tobacco cessation programs include leadership support for change, clinic staff training on brief, low-burden screening, as well as intervention and referral strategies that meet the needs of diverse patients. Employing a health equity lens at the implementation planning stage is critical to addressing the systemic barriers to tobacco use screening and referral and multiple stages (i.e., EMR documentation, training of providers and staff, perceived and determined roles and responsibilities). Thus, in this assessment, equitable implementation planning can be informed by multi-modal assessments that indicate: (i) where targeted efforts are needed (screening, clinic team training) and (ii) engagement and support by inner (leadership) and outer context stakeholders (community partners). It is critical to ensure equity with implementation to avoid worsening disparities [68, 69]. Our findings indicate that multilevel approaches to optimizing implementation to achieve health equity is required at the system, provider staff, and patient levels.

ACKNOWLEDGEMENTS

The authors thank the USC Norris Comprehensive Cancer Center Data Science Core (David Birtwell) and the USC Keck School of Medicine (Neil Bahroos) for informatics support. The authors also thank the support and guidance of the tobacco treatment team at the USC Chan Division of Occupational Science and Occupational Therapy.

FUNDING

This work was funded by a supplement to the USC Norris Comprehensive Cancer Center CCSG grant (5P30CA014089-45l; PI: Caryn Lerman).

COMPLIANCE WITH ETHICAL STANDARDS

Conflicts of interest: The authors declare that they have no conflicts of interest.

Human rights: All procedures were approved by the authorship team’s Institutional Review Board (IRB UP-20-01168).

Informed consent: Informed consent of participants was waived. All data were de-identified.

Welfare of Animals: This article does not contain any studies with animals.

TRANSPARENCY STATEMENTS

Study registration: This study was not formally registered.

Analytic plan registration: The analysis plan was not formally pre-registered.

Availability of data: De-identified data from this study will be made available by emailing the corresponding author.

Availability of analytic code: Analytic code used in this study may be available by emailing the corresponding author.

Availability of materials: Materials used to conduct this study are not publicly available.