Patterns of Tobacco-use Cessation Counseling Services Usage among Elderly Lung Cancer Patients in a Rural State Population - A Cancer Registry-linked Study

Abstract

Objectives

Tobacco-use is common among elderly lung cancer patients and continued tobacco-use can impact prognosis. This study evaluates patterns of receipt of Tobacco-use Cessation Counseling (TCC) services among these patients.

Methods

Using West Virginia Cancer Registry-Medicare linked database (2004–2007), we identified elderly patients with lung cancer (n = 922) and categorized them by receipt of TCC services. Hierarchical generalized logistic model was constructed and survival outcomes were analyzed by Kaplan-Meier analysis, Log-Rank test, and Cox proportional hazards modeling.

Results

Majority of patients (76.7%) received TCC services. Unadjusted analysis showed favorable survival outcomes in patients who received TCC services. However, adjusted lung cancer mortality risk was no different between the groups (HR (95% CI) = 1.78 (0.87–3.64)).

Conclusion

This study highlights the critical need to address disparities in receipt of TCC services among elderly. Although lung cancer preventive services are covered under the Medicare program, underutilization of these services is a concern.

Introduction

Tobacco-use is the leading preventable cause of lung cancer in the United States (US). It accounts for 90% of all lung cancer cases and 87% of all lung cancer deaths in the US.¹ Therefore, efforts to decrease lung cancer mortality have been focused on early detection and treatment of lung cancer, and smoking avoidance and cessation.^2–7

Clinical practice guidelines for preventive care in lung cancer have been published by American Society of Clinical Oncology, Biesalski et al, Cancer Guidance Group, College des Medecins du Quebec, National Cancer Institute (NCI), US Department of Health and Human Services, and US Preventive Services Task Force.^8–13 While these guidelines recommend smoking cessation among asymptomatic individuals, it is strongly encouraged among individuals diagnosed with lung cancer, as growing evidence suggests that continued smoking may compromise the effectiveness of cancer treatment, reduce the tolerance of patients for cancer treatment, and/or increase the risk of complications.¹⁴ Specifically, continued smoking following lung cancer diagnosis can interfere with radiation therapy and chemotherapy, increase risk of infection due to surgery, and decrease post-operative wound healing.^{14, 15}

Prior research has shown smoking to be common among patients at the time of lung cancer diagnosis, and that patients continue to smoke following diagnosis.¹⁶ Continued smoking following diagnosis also increases the risk of metachronous tumors/new primary cancer,^{14,17, 18} and can impact patients’ quality of life.¹⁹ However, impact of continued smoking on survival outcomes remains uncertain. While one study showed positive survival outcomes with smoking cessation,²⁰ another study reported no significant differences in prognosis basis on smoking status.²¹ Specifically, Johnston-Early and colleagues compared the survival outcomes among 112 small cell lung cancer patients (SCLC); 20 of whom had stopped smoking permanently before diagnosis, 35 had stopped at diagnosis, and 57 continued smoking. The authors found significant survival differences among the three groups, the worst outcomes being among those patients that continued to smoke. On the other hand, when Gail and colleagues examined the prognostic factors in 392 patients with resected Stage I Non-Small Cell Lung Cancer (NSCLC); continued smoking was not significantly associated with either increased disease risk, recurrence, or death. Regardless of its impact on survival, promoting smoking cessation among lung cancer patients at the time of diagnosis is much needed. Time of cancer diagnosis has also been described as a teachable moment for intervening with smokers and providing cessation treatment.²² Therefore, many insurance agencies including Medicare cover tobacco-use cessation counseling services. Beginning in March 2005, the Centers for Medicare and Medicaid Services began providing coverage for Tobacco-use Cessation Counseling (TCC) services for outpatient and hospitalized beneficiaries, who were smokers and had a disease/adverse health effect that was tobacco related.²³ The use of such services and its impact on health outcomes among elderly lung cancer patients remains unknown. To that end, the main focus of this study was to determine the patterns of receipt of TCC services among elderly lung cancer patients with a history of tobacco-use. The objectives were to: (1) determine the proportion of patients receiving TCC services; (2) determine the factors associated with receipt of TCC services; (3) to determine survival benefits associated with receipt of TCC services; and (4) to determine lung cancer mortality risk associated with non-receipt of TCC among patients.

Methods

Data source

This study used West Virginia Cancer Registry (WVCR)-Medicare (WVCR-Linked) linked 2004–2007 data files. West Virginia (WV) has the second highest lung cancer death rate and the highest smoking prevalence rate (26.8%) in the nation.²⁴ Therefore, the state served as an ideal population for this study. The WVCR-Linked data files are similar in structure to the NCI’s Surveillance, Epidemiology, and End Results-Medicare linked data files, and represent data from the West Virginia Cancer Registry. Details on the creation of WVCR-Linked data files can be found elsewhere.²⁵ Cancer registry data files provided clinical, demographic, cause of death, initial treatment, and tobacco-use history information for patients. The Medicare administrative data files provided the claims information for care provided by physicians, inpatient hospital stays, hospital outpatient clinics, home health care agencies, skilled nursing facilities, and hospice programs.

Study Cohorts

We identified Medicare beneficiaries aged ≥66 years, with incident lung cancer diagnosis between July 1, 2005 and October 31, 2007, and with a history of tobacco-use in the WVCR-Linked data files.²⁶ Beneficiaries were excluded if they were diagnosed only at death, had a prior malignancy, were enrolled in a managed care plan, or lacked Part A or B of Medicare. The resulting cohort (Cohort A) was used for study objectives 1–2.

Given the limited years of follow-up data, for study objectives 3–4, we limited Cohort A to patients with cancer diagnosis between July 1, 2005 and December 31, 2005, and followed them for two years following cancer diagnosis to determine lung cancer specific mortality.

Accessing receipt of TCC services

Patients were followed for two months following their cancer diagnosis to determine receipt of TCC services. A TCC session refers to face-to-face patient contact by the practitioner following a cancer diagnosis and can be minimal (3 minutes or less), intermediate (3–10 minutes), or intensive (greater than 10 minutes). TCC services were identified from the Medicare claim data files using appropriate Current Procedural Terminology codes (99201–99205,99211–99215,99406,99407,G0375,G0376).

Dependent variables

The primary outcome of interest, receipt of TCC services, was categorized as ‘Receipt’or ‘Non-receipt’. Survival time in days was calculated from the date of cancer diagnosis to the date of death or the two year follow-up cutoff date, which ever came first. To estimate lung cancer specific survival, patients not found to be deceased by the cutoff date, or who died due to causes other than lung cancer were censored at that time and considered to be alive.

Covariates

Based on prognostic significance, covariates included lung cancer type and stage, age at diagnosis, gender, race, urban-rural residence, comorbidity, and measures of socio-economic status. Age at diagnosis was categorized as 66–69 years, 70–74 years, 75–79 years, and ≥80 years. Race was categorized as ‘White’ or ‘Non-white’. Urban-rural residence was categorized as ‘Metro’, ‘Urban’, or ‘Rural’, using the Rural-Urban Continuum codes. Comorbidity was estimated using a modified Charlson comorbidity score, based on inpatient claims from the year preceding the cancer diagnosis.^27–29 Given the lack of individual socio-economic measures in our data source, median household income and the percentage of individuals with some college education in the census tract of residence, were used as markers of socio-economic status.

Data Analysis

Pearson chi-square tests were used to determine unadjusted associations between categorical variables of interest. Hierarchical generalized logistic model was constructed to identify factors associated with receipt of TCC services. The model treated census tract as a random effect to account for potential correlation among patients within the same county. Non-parametric estimates of the survivor function by receipt of TCC services were calculated using the Kaplan-Meier method and compared using log-rank test. Multivariate Cox proportional hazards model was constructed to estimate lung cancer mortality risk associated with non-receipt of TCC services. Variances in models were adjusted to account for patient clustering at the census tract level by the use of the robust inference of Lin and Wei.³⁰ All analysis were performed with SAS 9.2. The study was approved by West Virginia University-Institutional Review Board.

Results

Patient characteristics

We identified 922 patients from the WVCR-Linked database. Table 1 shows the distribution of clinical and socio-demographic characteristics of these patients by type of lung cancer. Majority of the patients had NSCLC and were diagnosed with late-stage disease. Compared to NSCLC patients, majority of SCLC patients were diagnosed at late stages (p≤0.05).

Table 1.

Characteristics of continuously enrolled Medicare Fee-for-service beneficiaries with incident lung cancer diagnosis and with a history of tobacco-use in West Virginia, July 2005 through October 2007.

Characteristics	Proportion (%)
	Non-Small Cell Lung Cancer	Small Cell Lung Cancer
Overall, n (%)	764 (82.8)	158 (17.1)
AJCC-TNM Stageˆ
I	17.9	-
II	8.5	-
III	22	21.5
IV	27.6	38
Un-staged	24	34.8
Age (years)
66–69	23.4	28.5
70–74	29.8	27.2
75–79	24.9	25.3
≥80	21.9	19
Gender
Male	57.3	50
Female	42.7	50
Race
Non-white	2.2	-
White	97.8	100
Urban-rural residence
Metro	55.8	58.2
Urban	38.9	34.8
Rural	5.4	-
Comorbidity, Charlson score
0	20.7	24.7
1	30.1	26.6
≥2	49.2	48.7

AJCC = American Joint Committee on Cancer; TNM = Tumor Node Metastasis.

^-Cell size suppressed to meet privacy guidelines.

^ˆAssociation between characteristic and cancer type among beneficiaries; Chi-square tests (p ≤ 0.05).

Source: West Virginia Cancer Registry-Medicare linked data files, 2004–2007.

Receipt of TCC services

Table 2 shows the descriptive characteristics of patients by receipt of TCC services. Majority of patients (76.7%) received TCC services and the proportions were higher among those with NSCLC diagnosis, as compared to those with SCLC diagnosis. Overall, patients receiving TCC services had early-stage disease, were mostly younger, and resided in non-rural areas (p≤0.05).

Table 2.

Descriptive characteristics by receipt of tobacco-use cessation counseling services among continuously enrolled Medicare Fee-for-service beneficiaries with incident lung cancer diagnosis and with a history or tobacco-use in West Virginia, July 2005 through October 2007.

Characteristics	Receipt of Tobacco-use Cessation Counseling Services
	Receipt		Non-receipt
	No.	%#	No.	%#
Overall, n (%)	707	76.7	215	23.3
Lung cancer type
NSCLC	595	77.9	169	22.1
SCLC	112	70.9	46	29.1
AJCC TNM stageˆ
I	126	87.5	18	12.5
II	58	86.6	-	13.4
III	156	77.2	46	22.8
IV	196	72.3	75	27.7
Un-staged	171	71.8	67	28.2
Age (years)ˆ
66–69	182	81.3	42	18.8
70–74	224	82.7	47	17.3
75–79	172	74.8	58	25.2
≥80	129	65.5	68	34.5
Gender
Male	388	75	129	25
Female	319	78.8	86	21.2
Race
Non-white	12	70.6	-	29.4
White	695	76.8	210	23.2
Urban-rural residenceˆ
Metro	387	74.7	131	25.3
Urban	271	77	81	23
Rural	49	94.2	-	5.8
Comorbidity, Charlson score
0	147	74.6	50	25.4
1	216	79.4	56	20.6
≥2	344	75.9	109	24.1

NSCLC = Non-Small Cell Lung Cancer, SCLC = Small Cell Lung Cancer, AJCC = American Joint Committee on Cancer, TNM = Tumor Node Metastasis.

^#Row percentages

^-Cell size suppressed to meet privacy guidelines.

^ˆAssociation between beneficiary characteristics and receipt of tobacco-use cessation counseling services; Chi-square test (p ≤0.05).

Source: West Virginia Cancer Registry-Medicare linked data files, 2004–2007.

Factors associated with receipt of TCC services

Controlling for socio-demographic characteristics, age remained a strong predictor for receipt of TCC services and the odds gradually decreased with increase in age (Table 3). Cancer stage and urban-rural residence were the only other significant predictors for receipt of TCC services. Specifically, patients with early-stage disease were more than as twice likely to receive TCC services, as compared to those with late-stage disease. Similarly, the likelihood of receipt of TCC services was lower among patients residing in metro areas, as compared to those residing in rural areas.

Table 3.

Factors associated with receipt of tobacco-use cessation counseling services among continuously enrolled Medicare Fee-for-service beneficiaries with incident diagnosis of lung cancer and a history of tobacco-use in West Virginia, July 2005 through October 2007.

Characteristics	Odds Ratio	95% Confidence Interval	p-value
Lung cancer type
NSCLC	1.31	0.86 to 1.99	0.2
SCLC		1 (Reference)
AJCC TNM stage
Un-staged	1.05	0.70 to 1.59	0.81
I	2.65**	1.47 to 4.80	< 0.01
II	2.55*	1.16 to 5.59	0.02
III	1.16	0.74 to 1.81	0.52
IV		1 (Reference)
Age (years)
66–69	2.58***	1.60 to 4.15	< 0.001
70–74	2.69***	1.71 to 4.25	< 0.001
75–79	1.68*	1.08 to 2.61	0.02
≥80		1 (Reference)
Gender
Male	0.83	0.59 to 1.16	0.27
Female		1 (Reference)
Race
Non-white	0.68	0.20 to 2.34	0.52
White		1 (Reference)
Urban-rural residence
Metro	0.16**	0.04 to 0.55	< 0.01
Urban	0.19**	0.05 to 0.67	< 0.01
Rural		1 (Reference)
Comorbidity, Charlson score
0	0.93	0.61 to 1.41	0.74
1	1.27	0.86 to 1.88	0.23
≥2		1 (Reference)
Percentage with some college education (%)ˆ
0.0–0.10	0.34	0.10 to 1.20	0.09
0.11–0.20	1	0.68 to 1.46	0.99
≥0.21		1 (Reference)
Median household income ($)ˆ
0–25,000	0.79	0.27 to 2.35	0.67
25,001–50,000	0.89	0.32 to 2.51	0.83
≥50,001		1 (Reference)

NSCLC = Non-Small Cell Lung Cancer, SCLC = Small Cell Lung Cancer, AJCC = American Joint Committee on Cancer, TNM = Tumor Node Metastasis.

Statistical significance:

^*p ≤ 0.05;

^**p ≤ 0.01;

^***p ≤ 0.001

^ˆCensus tract level measure

Model: N = 922, Fit Statistics: −2 restricted log pseudo-likelihood = 4308.15, Covariance parameter estimates: Intercept = county, estimate = 0.17, standard error = 0.16.

Source: West Virginia Cancer Registry-Medicare linked data files, 2004–2007.

Survival outcomes

Survival rates and median survival times were significantly greater among patients receiving TCC services, compared to those not receiving such services. Specifically, two year median survival time exceeded by 159 days among patients receiving TCC services. However, in the Cox proportional hazards model, the adjusted lung cancer mortality risk was found to be no different among patients by receipt of TCC services (Table 4).

Table 4.

Lung cancer mortality risk associated with non-receipt of tobacco-use cessation counseling services among continuously enrolled Medicare Fee-for-service beneficiaries with incident diagnosis of lung cancer and a history of tobacco-use in West Virginia, July 2005 through December 2007.

Characteristics	Hazard Ratio (95% Confidence Interval)
Tobacco-use cessation counseling services
Non-receipt	1.78 (0.87 to 3.64)
Receipt	1 (Reference)
Lung cancer type
NSCLC	1.04 (0.64 to 0.1.71)
SCLC	1 (Reference)
AJCC TNM stage
I	0.06*** (0.02 to 0.18)
II	0.30** (0.12 to 0.74)
III	0.51** (0.31 to 0.82)
IV	1 (Reference)
Age (years)
66–69	0.46* (0.22 to 0.94)
70–74	0.69 (0.31 to 1.55)
75–79	0.48* (0.24 to 0.96)
≥80	1 (Reference)
Gender
Male	0.62 (0.37 to 1.03)
Female	1 (Reference)
Race
Non-white	0.70 (0.11 to 4.28)
White	1 (Reference)
Urban-rural residence
Metro	3.12* (1.22 to 7.96)
Urban	2.63 (0.96 to 7.21)
Rural	1 (Reference)
Comorbidity, Charlson score
0	0.46 (0.44 to 1.08)
1	0.77 (0.43 to 1.39)
≥2	1 (Reference)
Percentage with some college education (%)ˆ
0.0–0.10	0.47 (0.42 to 1.08)
0.11–0.20	0.67 (0.45 to 1.01)
≥0.21	1 (Reference)
Median household income ($)ˆ
0–25,000	0.76 (0.26 to 2.19)
25,001–50,000	1.27 (0.57 to 2.80)
≥50,001	1 (Reference)

NSCLC = Non-Small Cell Lung Cancer, SCLC = Small Cell Lung Cancer, AJCC = American Joint Committee on Cancer, TNM = Tumor Node Metastasis.

Statistical significance:

^*p ≤ 0.05;

^**p ≤ 0.01;

^***p ≤ 0.001

^ˆCensus tract level measure

Model: N = 140, Fit Statistics: −2 log likelihood = 835.92, Global null hypothesis: Likelihood ratio chi-square test = 77.48 (p ≤ 0.05).

Source: West Virginia Cancer Registry-Medicare linked data files, 2004–2007.

Discussion

Tobacco-use is common among patients diagnosed with lung cancer and if continued can impact treatment outcomes, increase risk of malignancy, and/or decrease quality of life. Therefore, lung cancer diagnosis can be used as a teachable moment for smoking cessation counseling, as a patient’s motivation and interest in smoking cessation may increase after such an event. In this population based study, we determined the patterns of receipt of TCC services among elderly patients with incident lung cancer diagnosis and a history of tobacco-use.

Overall, TCC services were received by more than half of all patients. The proportions was significantly higher among patients with early-stage disease, as compared to those with late-stage disease. This finding is expected, as patients with early-stage disease are good candidates for curative therapy and have better prognosis. Therefore, patients with early-stage disease may be more motivated to quit tobacco-use, as compared to those with late-stage disease. In fact, patients with late-stage disease have been shown to be less likely to enroll in smoking cessation programs, as compared to those with early-stage disease.³¹ Age at diagnosis was also found to be a significant predictor for receipt of TCC services, with the odds gradually decreasing with increase in age. This finding may have resulted from differences in disease severity and comorbid illness burden, variation in physician practice patterns, and/or patient treatment preferences.³¹ Compared to younger patients, poor prognosis is common among older patients, and it is likely that these patients may choose to decline TCC services regardless of its benefits. Surprisingly, receipt of TCC services was found to be higher among patients residing in rural areas, as compared to those from non-rural areas. This finding may have resulted from the fact that prevalence of smoking is higher among elderly from rural areas, and that increased awareness of risks from continued smoking may have motivated individuals to seek and their providers to administer these services.

Prior studies have been inconclusive on the relationship between smoking cessation and prognosis following lung cancer diagnosis.^{20, 21} Although in this study we could not determine the success or failure of TCC attempts, receipt of such services by patients was associated with longer survival times. However, it is very likely that this finding may have resulted from increased disease severity among patients not receiving TCC services. In fact, when controlled for variability in patient clinical and socio-demographic characteristics, the adjusted lung cancer mortality risk was found to be no different among patients by receipt of TCC services. Overall, while findings from this study fail to show an association between receipt of TCC and adjusted lung cancer mortality risk, promoting smoking cessation at any stage of the disease is important.

This study has several limitations. A major limitation is the lack of information on success or failure of TCC attempts in patients. Specifically, administrative claims data used in this study fail to capture information on patients true quit attempts and success rates following receipt of TCC services. Such information is necessary to accurately quantify the benefits of TCC services. Data limitations also prevented us from studying the frequency and intensity of TCC attempts among patients. Furthermore, an inherent limitation of using claims data is the possibility of misclassification as a result of coding errors.³² However, claims data have been evaluated for their utility as a source of epidemiologic or health services information in cancer patients.³² The results of this study are generalizable to the West Virginia Medicare Fee-for-service (FFS) population aged 66 years and older, as data for Medicare beneficiaries enrolled in the managed care plan were not available for this study. Limited data availability at the time of this study also prevented us from doing a long-term follow-up of patients. Future studies can overcome the barriers seen in this study by collecting data on individual-level measures of socio-economic status, success and failure of TCC attempts, physician practice patterns, and patient preferences in using TCC services.

In summary, although preventive care services such as TCC services are covered under the Medicare program, underutilization of these services among beneficiaries is a concern. While some encouraging results have been demonstrated with use of TCC services in this study, more empirical studies that track type, length, adherence, and outcomes of such interventions are needed.