Continued smoking after the diagnosis of cancer is associated with decreased survival, quality of life, and functional status and with increased cancer recurrence and second primary malignancies.1-5 Moreover, continued smoking interferes with the efficacy of cancer therapies.6-9 Thus, promoting smoking cessation after the diagnosis of cancer is essential to enhance clinical outcomes.
The United States Public Health Service has a clinical practice guideline available, Treating Tobacco Use and Dependence, which identifies the gold standard for delivery of smoking-cessation treatment as the combined use of pharmacotherapy and behavioral counseling.10 In fact, the use of these treatments can double or triple the likelihood of long-term cessation.10 Despite the availability of effective smoking-cessation interventions and strong evidence for their efficacy, there tends to be low dissemination of these types of treatments.11, 12
Evidence exists that a significant proportion (54%–69%) of smokers with cancer do not receive formal assistance with their quit attempts.13-15 The few studies that have been conducted within the context of cancer indicate that patients often prefer to quit on their own and that healthcare provider attitudes toward smoking cessation after the diagnosis of cancer and lack of knowledge about how to provide optimal smoking-cessation treatment influences the provision of smoking-cessation services.13-18 Given the paucity of data, however, further studies are needed to determine whether smoking-cessation treatments are being offered to patients after the diagnosis of cancer and to define the factors associated with increasing the application of evidenced-based smoking-cessation interventions among cancer patients. This study extends the literature by identifying patients' reports of whether they received advice and assistance from their healthcare providers to quit smoking and their subsequent interest in participating in a smoking-cessation treatment program. The specific objectives of these analyses were to describe the type of smoking-cessation interventions that smokers and recent quitters with cancer report having received from their healthcare providers, to identify patient interest and preferences for participating in a cessation program, and to identify the factors (age, sex, education, type of cancer, stage of disease, level of nicotine dependence, partner smoking) associated with interest in participating in a cessation program among smokers with cancer. This information can be used to inform the development of programs to recruit patients who continue smoking after a diagnosis of lung or head and neck cancer and to develop future effective smoking-cessation programs tailored for cancer patients.
MATERIALS AND METHODS
This study was part of a larger prospective clinical study that described smoking abstinence rates in patients with lung cancer and head and neck cancer after their diagnosis and examined the factors related to smoking abstinence at 3 months and 6 months after entry into the study. Data collected at entry to the study were used in the analyses presented in this report. Consecutive patients who were seen in ambulatory thoracic and head and neck oncology programs at a comprehensive cancer center in the Northeast United States were screened and invited to participate in an institutional review board approved study. Written informed consent was obtained and data were collected at a mutually agreeable time. Inclusion criteria were smokers (defined as those who smoked at least 1 cigarette within the last 7 days) or recent quitters (defined as those who quit smoking within the last 6 months) who were recently diagnosed patients (defined as diagnosed within the last 120 days) with nonsmall cell lung cancer or head and neck cancer. Recent quitters were included in this study because previous research has identified that cancer patients who have quit within the last 6 months are at high risk for return to smoking.15, 19
Standardized questionnaires were used to collect demographic, disease-related, and tobacco-related data. Demographic information was collected at baseline using a self-report questionnaire and included: age, sex, marital status, and race. Disease-related variables of interest for this study were type of cancer and disease stage and were collected from a medical chart record review by a trained research coordinator. The American Joint Committee on Cancer Staging Manual20 was used to classify stage. Stages I through III were defined as early stage disease, and stage IV was defined as late stage disease.
The tobacco-related variables of interest for this study that were collected from smokers and recent quitters included: age at which smoking started, the number of years smoking, whether the participant's spouse/partner was smoking, and the type of assistance that was given to participants by healthcare providers related to smoking cessation. Among smokers, additional variables were collected and included the number of cigarettes smoked per day, the number of quit attempts in the last year, and the level of nicotine dependence. These variables were collected through a self-report, standardized questionnaire that has been used in previous studies.21, 22
Smoking was defined by using a 7-day point-prevalence abstinence measure, which asked patients whether they had smoked any cigarettes within the last 7 days before the study. This measure was chosen because it is 1 of the most common measures used in previous studies and in meta-analyses.23, 24
The Heaviness of Smoking Index was used to measure nicotine dependence. This is a 2-item questionnaire that is a shortened form of the Fagerstrom Test for Nicotine Dependence. It is a reliable and valid measure of the level of nicotine dependence and has been used extensively in previous research.25, 26 Previous studies have demonstrated that this measure has adequate reliability (Cronbach alpha coefficient [α] = .72), and test-retest reliability (α = .82) and is 1 of the best predictors of smoking cessation among the nicotine-dependence scales.25-28 Scores range from 0 to 6, and a score ≥4 is considered high dependence.
A self-report, standardized questionnaire that had been used in previous research was used to gather information related to the type of smoking-cessation interventions that participants received from healthcare providers.15, 29, 30 The variables of interest for this study were collected by patient self-report and included: 1) whether advice to quit smoking was provided, 2) whether assistance with smoking cessation was offered, and 3) what type of assistance was offered to aid in the cessation attempt. Advice to quit smoking was defined as whether the healthcare provider (physician or nurse practitioner) advised the patient to quit smoking (yes vs no). Assistance with smoking cessation was defined as whether the healthcare provider recommended any smoking-cessation treatment options (yes vs no). If the patient responded yes, then they were asked to indicate what type of assistance was offered by the healthcare provider. The types of assistance that could be chosen by participants included counseling, information, and/or medication used for smoking cessation (ie, nicotine replacement, bupropion, varenicline). Participants could select more than 1 method that was recommended for cessation. We also asked patients which specific pharmacotherapy cessation aides they were using at the time they entered the study.
A self-report, standardized questionnaire was used to collect information about patient interest in a smoking-cessation program and about their preference for the type of program. Patient interest in a smoking-cessation program was measured by asking patients whether they were interested in a smoking-cessation program (yes or no) and, if they were yes, then which type or components of a program would be of interest. The choices for the various types of programs included none, individual treatment (one-on-one treatment with a tobacco-treatment specialist), group counseling, referral to community-based program, telephone counseling, or other. Patients could choose more than 1 type of program.
Descriptive statistics were used to characterize patients at study entry. Logistic regression was used to model smokers' interest in participating in a smoking-cessation program. Potential predictors included age (<57 years vs ≥57 years), sex, education level (some college vs less than a high school education), type of cancer (lung vs head and neck), disease stage (early/regional vs late), nicotine dependence (a score ≥4 on the Heaviness of Smoking Index vs a score <4), marital status (married/living together vs single/separated/divorced/widowed), and smoking status of partner (yes vs no). To balance the number of patients in different age groups among smokers, patients were categorized into 2 age groups using the median age (57 years).
Univariate logistic regression was used to examine the odds ratios (OR) for predictors of interest in a smoking-cessation program at entry to the study. A stepwise multiple logistic regression model was used to identify which variables remained significant when considered in combination. The significance level that was used for entering effects was.15, and the significance level of the Wald chi-square test for an effect to stay in the model was.10. All P values were 2-tailed, and a P value <.05 was considered significant. The Hosmer and Lemeshow test did not provide evidence of a lack of fit to the data (P = .49). All analyses were conducted using the statistical software SAS, version 9.2 (SAS Institute, Inc., Cary, NC).
The sample consisted of 160 patients with lung cancer or head and neck cancer. Of 1783 participants who were screened for eligibility, 282 were eligible, 180 agreed to participate, and 160 had complete data available for this analysis. The reasons that the 102 participants gave for not enrolling in the study were health limitations (n = 4), no time (n = 19), not interested (n = 58), other (n = 9), and no response (n = 12). Before data collection, 17 participants who signed consent cancelled because of health limitations (n = 5), died (n = 1), changed their mind (n = 5), had limited time (n = 4), or had other reasons (n = 2). Three additional participants were excluded for missing data.
Table 1 provides information about characteristics of the sample by cancer site and for the entire sample. Overall, the median age of participants was 57 years (range, 25-81 years). Most participants were men (63%; n = 101), white (93%; n = 149), and had greater than a high school education (70%; n = 112). Fifty-seven percent of the sample (n = 91) had lung cancer, and 12% (n = 19) had stage I or II disease, 40% (n = 64) had stage III disease, and 47% (n = 75) had stage IV disease at diagnosis. At the time they entered the study, 32% (n = 51) of participants were smokers. The median number of cigarettes smoked per day among smokers was 10. The frequency of smoking among partners was 24% (n = 39).
Table 1. Characteristics of the Sample (n=160)
|Age: Median [range], y||58 [28-81]||54 [25-77]||57 [25-81]|
| Men||42 (46.2)||59 (85.5)||101 (63.1)|
| Women||49 (53.8)||10 (14.5)||59 (36.9)|
| White||84 (92.3)||65 (94.2)||149 (93.1)|
| Black||6 (6.6)||2 (2.9)||8 (5)|
| Other||1 (1.1)||2 (2.9)||3 (1.9)|
| Missing||2 (2.2)||0 (0)||2 (1.3)|
| ≤High school||25 (27.5)||21 (30.4)||46 (28.8)|
| >High school||64 (70.3)||48 (69.6)||112 (70)|
| Single/never married||12 (13.2)||10 (14.5)||22 (13.8)|
| Married/partnered||59 (64.8)||47 (68.1)||106 (66.3)|
| Separated/divorced/widowed||20 (22)||12 (17.4)||32 (20)|
|Stage at diagnosis|
| Missing||0 (0)||2 (2.9)||2 (1.3)|
| Early||15 (16.5)||4 (5.8)||19 (11.9)|
| Regional||37 (40.7)||27 (39.1)||64 (40)|
| Late||39 (42.9)||36 (52.2)||75 (46.9)|
Reliability of Measures
The Cronbach α coefficient was measured for the Heaviness of Smoking Index and was α = .55 at baseline for this measure.
Provider-Recommended Smoking-Cessation Interventions
Seventy-nine percent (n = 126) of participants in the sample reported that a healthcare provider advised them to quit smoking; of these, 86% (n = 44) were smokers, and 75% (n = 82) were recent quitters. In addition, 53% (n = 84) of the total sample reported that a healthcare provider recommended a specific smoking-cessation treatment to assist them in their quit attempt; of these, 65% (n = 33) were smokers, and 47% (n = 51) were recent quitters.
Participants reported that the most common type of assistance offered by their healthcare providers was pharmacotherapy alone (57.5%; n = 19 of 33) or combined pharmacotherapy and behavioral treatments (27.2%; n = 9 of 33). In total, 19.4% (31 of 160) of participants were using a pharmacologic cessation aide at the time of study entry, including 27.5% (14 of 51) of smokers and 15.6% (17 of 109) of recent quitters.
Interest and Preferences for a Smoking-Cessation Program
Thirty percent (n = 48) of participants expressed interest in a smoking-cessation program, including 51% (n = 26) of smokers and 20.2% (n = 22) of recent quitters. We examined the type of smoking-cessation program that smokers and recent quitters would be interested in if an ideal program were available. An individualized smoking-cessation program alone was endorsed most frequently by participants. Approximately 23% (n = 11 of 48) of participants indicated an interest in participating in more than 1 type of behavioral support program, including 19.2% (n = 5 of 26) of smokers and 27.3% (n = 6 of 22) of recent quitters (see Table 2). Group counseling, telephone counseling, and other were the types of behavioral treatments that were of most interest to smokers; whereas group counseling, telephone counseling, community-referral, and other were of interest to recent quitters.
Table 2. Preferences for Type of Smoking-Cessation Programs Among Cancer Patients
|Patients interested in program||22||20.2||26||51||48||30|
|Preference of program type|
| No treatment||2||9.1||0||0||2||4.2|
| Individual treatment only||9||40.9||14||53.8||23||47.9|
| Group counseling only||1||4.5||1||3.8||2||4.2|
| Referral community program only||1||4.5||0||0||1||2.1|
| Telephone counseling only||0||0||0||0||0||0|
| Other onlya||2||9.1||6||23.1||8||16.7|
| ≥2 Programs||6||27.3||5||19.2||11||22.9|
The Factors (Age, Sex, Type of Cancer, Stage of Disease, Education, Level of Nicotine Dependence) Associated With Interest in Participating in a Smoking-Cessation Program
Potential factors that were related to interest in participating in a formal smoking-cessation program were examined at study entry among smokers by using logistic regression models (see Table 3). In a multivariate model, among smokers, younger patients and patients with early stage disease were significantly more interested in participating in smoking-cessation programs, such that the OR for younger patients was 3.9 (95% confidence interval [CI], 1.0-15.5), and the OR for patients with early stage disease was 6.7 (95% CI, 1.4-31.0). In addition, the presence of a smoking partner was associated significantly with increased interest in a smoking-cessation program (OR, 5.5; 95% CI, 1.0-29.6).
Table 3. Factors Related to Interest in Participating in a Smoking-Cessation Program Among Smokers (n=51)
| Age, y||<57 vs ≥57a||3.2 (1.0-10.7)||.06|
| Sex||Men vs womena||0.4 (0.1-1.4)||.16|
| Education||College vs ≤high schoola||0.5 (0.1-1.6)||.22|
| Type of cancer||Lung vs head and necka||1.5 (0.5-4.7)||.53|
| Nicotine dependence||Index score ≥4 vs othera||2.3 (0.7-8.4)||.19|
| Stage||Early/regional vs latea||3.2 (1.0-10.7)||.06|
| Marital status||Married/living together vs othera||1.0 (0.3-3.2)||.98|
| Partner smoking status||Smoking vs othera||0.4 (0.1-1.3)||.11|
| Age, y||<57 vs ≥57a||3.9 (1.0-15.5)||.054|
| Stage||Early/regional vs latea||6.7 (1.4-31.0)||.02|
| Partner smoking status||Smoking vs othera||5.5 (1.0-29.6)||.049|
The results from the current study suggest that healthcare providers in this setting were trying to actively engage smokers with cancer in smoking-cessation treatment. Our study indicates that 86% of smokers received advice to quit smoking from their healthcare providers compared with 72% who received advice to quit smoking in a study of cancer survivors by Coups and colleagues.31 We observed that 65% of patients in the current study were offered assistance with pharmacotherapy. However, only approximately 1 of 3 smokers and 1 of 5 recent quitters reported using a pharmacologic cessation aide at the time they entered the study. It is important to recognize that the use of pharmacotherapy along with behavioral counseling is essential to optimize success with long-term smoking abstinence, especially in highly dependent smokers.10 There are 4 main types of first-line pharmacotherapy that can be used to enhance smoking cessation among cancer patients: long-acting nicotine-replacement treatment, short-acting nicotine-replacement treatment, bupropion, and varenicline.10 Although all of these agents are effective, their efficacy varies compared with placebo, and varenicline appears to be more effective compared with nicotine-replacement treatment and bupropion.10 All of these drugs may be used in patients with cancer, and their use depends on the patient's coexisting comorbidities, past experience with use of these drugs, preferences, and cost.10, 28, 32, 33 Readers are referred to other sources for a more in-depth discussion of pharmacotherapy options in cancer patients.32-34
Another significant finding from our study is that approximately 33% of the sample indicated that they would be interested in participating in a smoking-cessation program, representing 50% of smokers and 20% of recent quitters. The majority of patients indicated an interest in individualized programs that incorporated the use of pharmacotherapy. Our findings are similar to those from 3 other studies that assessed the use of evidenced-based treatment among cancer patients and reported that many cancer patients attempted to quit on their own and 33% to 50% used evidenced-based smoking-cessation treatment.13-15 It is somewhat surprising that patient-reported use of pharmacologic cessation aides in our study was similar to that from studies conducted more than 5 years ago, especially because quitting smoking is essential to improve clinical outcomes associated with cancer treatment.2, 5, 8, 13, 14 Substantial evidence indicates that increased intensity of treatment for tobacco dependence is associated with improved cessation rates.10
Almost 25% of participants were interested in the use of multiple types of behavioral treatments (ie, group counseling, use of telephone counseling) as part of a smoking-cessation program to enhance their cessation efforts. The overall level of interest in using behavioral treatment in our study was higher than that reported by Coups and colleagues,31 who reported that only 3% of cancer patients actually used a behavioral treatment in their quit attempt. It is noteworthy that patients in our study were interested in having multiple types of behavioral interventions available to assist them during their quit attempts, and this suggests that the integration of various options for behavioral support during cancer treatment may be an important component of a smoking-cessation treatment program.
The factors associated with smokers' interest in participating in a smoking-cessation program included younger age, early stage disease, and having a partner who was a smoker. The effects of age on continued smoking are mixed, with some studies suggesting that younger age is associated with continued smoking, and other studies suggesting that older age is associated with continued smoking.29, 35 Younger smokers were 4 times more likely to be interested in participating in a smoking-cessation program compared with older smokers. It is encouraging that younger smokers were interested in cessation programs, because patients who are younger often experience a greater intensity of cancer treatment (ie, surgery and/or combined chemotherapy and radiation) and have competing demands of managing a family, employment, and a life-threatening illness, which makes smoking cessation an especially difficult task.36, 37 However, smoking cessation is essential regardless of age. This finding suggests that age may be an important factor to consider when developing future smoking-cessation programs. Interventions targeted toward increasing interest in participation of smoking-cessation programs appear to be needed for older smokers.38
Patients with lung cancer or with head and neck cancer who have early stage disease reportedly are at high risk for continued smoking after their diagnosis.13, 19 Often, patients with early stage disease have fewer comorbidities, are not symptomatic from their illness, and may not feel compelled to quit smoking in the same manner as others who are receiving active cancer treatment and/or who have advanced disease. Thus, the finding that patients with early stage disease were almost 7 times more likely to be interested in participating in a smoking-cessation program is noteworthy and suggests that patients in this particular group are very interested in seeking assistance in their effort to quit smoking.
Patients with lung cancer or head and neck cancer who live with other smokers were identified as more likely to remain smokers after their diagnosis compared with those who lived with nonsmokers.29, 39 It is encouraging that smokers with partners who were smokers were almost 6 times more likely to be interested in participating in a smoking-cessation program. Often, living with a smoker presents as a barrier and affects motivational readiness to change.40 There is some initial evidence that partner support can enhance smoking-cessation outcomes. Park and colleagues41 conducted a meta-analysis to determine whether an intervention to enhance partner support helped as an adjunct to a smoking-cessation program. Results from their study suggested that interventions focused on enhancing supportive behaviors while minimizing, criticizing, or blaming behaviors were most promising, especially when they were implemented with live-in, married, or equivalent to married partners.
Taken together, these findings suggest that there may be developmental, cognitive, and/or affective factors that interfere with smokers' intention to effectively access smoking-cessation programs after a diagnosis of cancer. We identified some factors that may be used to develop future smoking-cessation interventions, such as targeting patients with early stage disease for smoking-cessation programs and tailoring programs based on age and the presence of partner support. However, 50% of our sample indicated a lack of interest in participating in a formal smoking-cessation program. Some factors that may present as barriers are low motivation, disbelief of the efficacy of treatment, low confidence in the ability to quit smoking, stigma associated with seeking treatment for smoking cessation, and/or a belief that treatment is not needed and that willpower alone is enough to enable cessation.39, 42 Because all previous studies used quantitative methods, the use of qualitative research methods, such as narrative research, may be the best way to help understand patient experiences with smoking cessation after the diagnosis of cancer and to uncover the factors that impede accessing the effective use of evidence-based smoking-cessation treatments so that interventions can be developed that increase interest and motivation to use evidenced-based treatments.43-45
In addition, relapse-prevention programs geared toward recent quitters are needed. Fifteen percent of recent quitters in the current study were using pharmacologic cessation aides; and 27% indicated that a variety of behavioral interventions, such as group or community-based counseling programs, would be of interest to enhance their cessation efforts. Thus, these findings provide support for the development and testing of smoking-cessation interventions and relapse-prevention programs for patients with lung cancer and head and neck cancer. To our knowledge, there have not been any effective relapse-prevention strategies developed to date for cancer patients. However, Hajeck and colleagues46 conducted a meta-analysis that examined relapse-prevention interventions for smoking cessation among the general population and observed that the extended use of varenicline and nicotine-replacement treatment may prevent relapse, but further studies are needed to confirm their efficacy.
To increase smokers' and recent quitters' use of effective treatment, innovative methods are needed to make smoking-cessation treatments more engaging. To stimulate more creative approaches with which to engage smokers in cessation treatment, a consumer-oriented approach to reach smokers has been advocated.47 This approach is potentially promising, especially during times of medical illness, when awareness of the ill effects of tobacco are raised among smokers. Backinger and colleagues11 suggested that viewing smokers as consumers and taking a fresh look at quitting from their perspective, redesigning smoking-cessation treatment programs so they are more engaging, and then marketing cessation products in ways that reach smokers are core strategies for increasing the demand for evidenced-based smoking-cessation treatment.
An additional finding that warrants discussion is that the Cronbach α coefficient for the Heaviness of Smoking Index was much lower than expected in the current study, which calls into question the reliability of this measure for use in patients with cancer. The majority of studies that used this measure were conducted in the general population.25-27 This finding suggests that the Heaviness of Smoking Index may not tap into the essential components of nicotine dependence that are experienced by patients with cancer and underscores the need for further research focused on understanding smoking behaviors among patients with cancer.48, 49
The current study had several limitations. In particular, we used self-report to define smoking status. Biochemical verification of smoking status with cotinine or carbon monoxide is recommended to enhance precision in determining smoking-cessation outcomes in clinic-based studies and in studies that involve special populations, including cancer patients.50 Future studies should incorporate the collection of both self-report and biochemical verification of smoking status to further support an empirical basis of the utility of biochemical verification across a range of studies, treatment modalities, and populations.
Although 1 of the strengths of this study is that we collected information about whether patients were advised to quit smoking, whether they received assistance, and, if so, then what type of assistance they received from their healthcare providers, we did not collect information about whether smokers who were not ready to quit smoking received motivational messages that encouraged them to quit smoking. The US Public Health Service clinical practice guideline, Treating Tobacco Use and Dependence, recommends that all patients receive advice and assistance to quit smoking.10 For patients who are not ready to make a quit attempt, however, the use of the “5-R” model is recommended to enhance motivation to quit smoking.10 The “5-R” model provides specific strategies that healthcare providers can use to personalize the message to their patients about why quitting smoking is important to their health. This model suggests the use of 5 simple steps, which include: 1) explain the relevance of quitting smoking to their current health situation, 2) highlight the risks associated with current smoking, 3) identify the rewards associated with quitting smoking, 4) help patients identify the roadblocks associated with quitting smoking at this time, and 5) repeat the message at regular intervals. This model may be helpful in guiding future interventions that aim to motivate smokers to quit smoking and encourage them to use evidence-based smoking-cessation treatments. Further testing of health communication messages that increase motivation and encourage use of evidence-based smoking-cessation treatments is warranted among smokers with cancer.10, 51, 52