Procedural Aggressiveness in Veterans with Advanced Non-Small-Cell Lung Cancer at the End of Life

Abstract

Background:

Evidence suggests that the aggressiveness of care in cancer patients at the end of life is increasing. We sought to evaluate the use of invasive procedures at the end of life in patients with advanced non-small-cell lung cancer (NSCLC).

Objective:

To evaluate the utilization of invasive procedures at the end of life in Veterans with advanced NSCLC.

Design:

Retrospective cohort study of Veterans with newly diagnosed stage IV NSCLC who died between 2006 and 2012.

Setting/Subjects:

Subjects were identified from the Veterans Affairs Central Cancer Registry.

Measurements:

All Veterans Administration (VA) and Medicare fee-for-service healthcare utilization and expenditure data were assembled for all subjects. The primary outcome was the number of invasive procedures performed in the last month of life. We classified procedures into three categories: minimally invasive, life-sustaining, and major-operative procedures. Logistic regression analysis was used to evaluate factors associated with the receipt of invasive procedures.

Results:

Nineteen thousand nine hundred thirty subjects were included. Three thousand (15.1%) subjects underwent 5523 invasive procedures during the last month of life. The majority of procedures (69.6%) were classified as minimally invasive. The receipt of procedures in the last month of life was associated with receipt of chemotherapy (odds ratio [OR] 3.68, 95% confidence interval [CI] 3.38–4.0) and ICU admission (OR 3.13, 95% CI 2.83–3.45) and was inversely associated with use of hospice services (OR 0.35, 95% CI 0.33–0.38).

Conclusions:

Invasive procedures are commonly performed among Veterans with stage IV NSCLC during their last month of life and are associated with other measures of aggressive end-of-life care.

Introduction

As the number of therapeutic options for patients with advanced cancer has increased over the past few decades, receipt of aggressive care near the end of life has become more common.^1,2 Currently, the use of invasive procedures in terminally ill patients is not specifically considered an indicator of aggressive end-of-life care. However, there is general consensus within the oncology and palliative care communities that limiting the use of aggressive interventions in patients with advanced cancer in the last few weeks of life is an important component of providing high-quality cancer care.^3,4 Despite an increased number of therapeutic options, advanced non-small-cell lung cancer (NSCLC) is still associated with a poor prognosis with a median survival of less than one year.^5,6 The purpose of this study was to evaluate the utilization of invasive procedures at the end of life in Veterans with advanced NSCLC.

Methods

Subjects

We selected a cohort of Veterans from the Veterans Affairs Central Cancer Registry (VACCR) who were identified as having stage IV NSCLC and who died between January 1, 2006 and December 31, 2012. The VACCR identifies all Veterans with cancer who are diagnosed and/or receive their first course of treatment in a Veterans Health Administration (VHA) facility. We subsequently assembled all VA and Medicare Fee-for-Service utilization and expenditure data for all Veterans.

Data sources

Data on each Veteran's healthcare utilization were combined from the VA patient treatment files, VA outpatient and pharmacy claims, the national patient care database, and the fee basis files, which include information regarding care purchased by the VA but received in the community. VA-specific data were merged with Medicare part A and B claims data for all eligible Veterans. These data are available through the VA Information Resource Center with VA Central Institutional Review Board approval.

A daily record of all healthcare utilization was assembled, including location of care (inpatient, nursing facility, home, etc.). Demographic information was obtained from the VA vital status file. Age was recorded as age in years at the time of death. Race was recorded as white, black, or other. Region was recorded as the geographic region of the last VA facility in which a Veteran received care. Information on cancer stage and cell type was obtained from the VACCR. Cell type was recorded as adenocarcinoma, squamous cell, and other NSCLC. Receipt of chemotherapy was defined as identification of a claim for any intravenous chemotherapy. ICU admission was defined as any identification of ICU as the daily location of care, whether in a VHA or Medicare hospitalization. Receipt of hospice care was defined as identification of any claim for either inpatient or outpatient hospice services in the VA inpatient or outpatient care files, VA fee basis file, or Medicare hospice benefit claims files.

Procedures

Our primary outcome was the number of invasive procedures performed in the last month of life. Procedures were identified using International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) procedure codes. The VA medical SAS database includes procedure codes for all procedures performed in both the inpatient and ambulatory surgical setting. Procedures performed outside of the VA were captured using the fee basis files and Medicare fee-for-service files. We were unable to capture procedures performed in younger Veterans outside of the VA paid for by private insurance, so these procedures were not included unless the Veteran was Medicare eligible due to disability.

Procedures were further classified using the Healthcare Cost and Utilization Project (HCUP) procedure classes and clinical classification software (CCS) classes.⁷ We classified procedures into three prespecified categories: minimally invasive (endoscopic and percutaneous procedures typically performed with local anesthesia or conscious sedation), life-sustaining (intubation and/or mechanical ventilation, tracheostomy, feeding tube placement, dialysis, and CPR), and major-operative procedures (procedures classified as HCUP procedure classes 3 and 4).^7–10 Procedures classified within the same CCS class that were recorded on the same day were assumed to represent one procedure (for example a bronchoscopy with bronchoalveolar lavage and a bronchoscopy with transbronchial biopsy recorded on the same day were counted as one procedure). As the purpose of this analysis was to evaluate the use of procedures that might contribute to morbidity, we did not include very minor procedures (e.g., excision of skin lesion). Additional details on the coding strategy can be found in Appendix A1.

Statistical analyses

Statistical analyses were performed using SAS 9.3 statistical software (Cary, NC). We used logistic regression analysis to examine factors associated with invasive procedures performed within the last month of life. All models were adjusted for patient comorbidities as measured by the Deyo adaptation of the Charlson comorbidity index (CCI).¹¹ The CCI was calculated for each individual using VA and/or Medicare inpatient and outpatient claims data from the four months before diagnosis date. The number of comorbidities was subsequently categorized as 0, 1, 2, or ≥3. As all patients carried a diagnosis of advanced malignancy, diagnosis codes for cancer and metastatic malignancy were excluded from the CCI for all patients.¹² Given that receipt of hospice services, chemotherapy, and ICU admissions are closely interrelated, we performed separate analyses entering these variables into the models independently. For all models a p-value <0.05 was considered statistically significant. We performed a sensitivity analysis excluding subjects under the age of 65 who may have had procedures paid by private insurance and not captured in the VA or Medicare data. We also performed a sensitivity analysis excluding subjects who expired within one month of diagnosis of advanced malignancy who may have had procedures performed as part of their initial diagnostic evaluation.

Results

We identified 19,930 Veterans with stage IV NSCLC who expired between 2006 and 2012. Table 1 shows the characteristics of the study population. The majority of subjects was male (98.3%) and white (79.7%). The mean age was 69.0 (standard deviation 9.7) years. Overall 3000 (15.1%) subjects underwent 5523 invasive procedures during the last month of life.

Table 1.

Demographic Characteristics of Decedents with Stage IV Non-Small-Cell Lung Cancer (n = 19,930)

	n	%
Age
<55	1585	7.95
55–64	6747	33.85
65–74	5852	29.36
75–84	4713	23.65
85+	1033	5.18
Region
Mid West	4491	22.53
North East	2450	12.29
South	8952	44.92
West	3400	17.06
Sex
Male	19,589	98.29
Female	341	1.71
Race
White	15,874	79.65
Black	3654	18.33
Other	110	0.55
Invasive procedure in the last month of life	3000	15.05
Chemotherapy in the last month of life	2927	14.69
Hospitalization in the last month of life	5233	26.26
ICU admission in the last month of life	1933	9.70
Hospice in the last month of life	7667	38.47
Histology
Adenocarcinoma	7294	36.60
Squamous cell carcinoma	5889	29.55
Other non-small-cell lung cancer	6747	33.85
Comorbidities
0	8233	41.31
1	6780	34.02
2	3261	16.36
≥3	1656	8.31

Table 2 reveals the frequency of procedures performed within the last month of life. The majority of procedures (69.6%) was classified as minimally invasive. There were 1049 life-sustaining procedures and 600 major-operative procedures performed in the last month of life. Overall, the most common procedures were thoracentesis, bronchoscopy, and central venous catheterization. The most common major-operative procedures were surgical excision of brain tissue, pericardiotomy, and open reduction and internal fixation of hip fractures.

Table 2.

Frequency of Select Invasive Procedures Performed in the Last Month of Life in Veterans with Non-Small-Cell Lung Cancer

	n (Number of procedures)	n (Number of subjects)	% (Percent of subjects)
Minimally invasive procedures
Bronchoscopy	794	723	3.63
Thoracentesis/chest tube	1157	828	4.15
Image-guided biopsies	536	513	2.57
Central venous catheterization	885	776	3.89
Port placement	42	41	0.21
GI endoscopy	221	192	0.96
Angiography (cardiac and noncardiac)	209	185	0.93
Life-sustaining procedures
Intubation and/or invasive mechanical ventilation	765	643	3.23
Tracheostomy	33	29	0.15
Feeding tube placement	96	88	0.44
Dialysis	77	62	0.31
CPR	108	100	0.50
Operative procedures
Lung resection	26	25	0.13
Other thoracic surgeries	92	73	0.37
Other major surgeries	482	357	1.79

Table 3 presents the demographic characteristics associated with receipt of any invasive procedure (minimally invasive, life-sustaining, or major-operative procedures) in the last month of life. Demographic characteristics associated with an increased likelihood of undergoing procedures included younger age, non-white race, and the presence of comorbidities. Year of death was inversely associated with the likelihood of receiving an invasive procedure (i.e., subjects who died later in the study period were less likely to receive a procedure). Geographic region was not associated with risk of undergoing an invasive procedure. ICU admission (odds ratio [OR] 3.13, 95% confidence interval [CI] 2.83–3.45) and receipt of chemotherapy (OR 3.68, 95% CI 3.38–4.0) within the last month of life were strongly and positively associated with risk of undergoing an invasive procedure. Hospice utilization within the last month of life was strongly and negatively associated with receipt of invasive procedures (OR 0.35, 95% CI 0.33–0.38).

Table 3.

Demographic Characteristics Associated with Receipt of Any Invasive Procedure in the Last Month of Life

Effect	Odds ratio	95% Wald confidence limits
Age, years
<55	1.557^*	1.295	1.873
55–64	1.536^*	1.312	1.798
65–74	1.367^*	1.166	1.603
75–84	1.105	0.939	1.301
85+	Reference	Reference	Reference
Female gender	0.907	0.709	1.16
Non-white race	1.244^*	1.150	1.345
Region
Midwest	1.082	0.979	1.195
New England	1.079	0.961	1.212
South	0.986	0.901	1.079
West	Reference	Reference	Reference
Death year^*	0.958^*	0.943	0.974
Comorbidities
0	Reference	Reference	Reference
1	1.444^*	1.342	1.553
2	1.455^*	1.328	1.595
>3	1.415^*	1.256	1.595

p < 0.05.

Table 4 shows the demographic characteristics associated with use of life-sustaining and major-operative procedures in the last month of life (ignoring minimally invasive procedures). Characteristics associated with the receipt of a life-sustaining or major-operative procedure were similar to those associated with receipt of any invasive procedure with the exception that non-white race was no longer statistically significant. ICU admission (OR 5.26, 95% CI 4.26–6.50) and receipt of chemotherapy (OR 2.46, 95% CI 1.98–3.06) within the last month of life were strongly associated with the receipt of life-sustaining and major-operative procedures. Again, hospice use (OR 0.29, 95% CI 0.22–0.38) within the last month of life was strongly and negatively associated with receipt of life-sustaining and major-operative procedures.

Table 4.

Demographic Characteristics Associated with Receipt of a Life-Sustaining or Major Operative Procedure in the Last Month of Life

Effect	Odds ratio	95% Wald confidence limits
Age, years
<55	2.082^*	1.129	3.841
55–64	1.725	0.994	2.995
65–74	1.408	0.805	2.465
75–84	1.257	0.710	2.225
85+	Reference	Reference	Reference
Female gender	0.931	0.434	1.999
Non-white race	1.017	0.792	1.306
Region
Midwest	1.313	0.966	1.784
New England	1.077	0.746	1.556
South	0.954	0.714	1.275
West	Reference	Reference	Reference
Death year^*	0.909^*	0.863	0.957
Comorbidities
0	Reference	Reference	Reference
1	1.608^*	1.277	2.025
2	1.242	0.913	1.690
>3	1.954^*	1.390	2.746

p < 0.05.

We performed sensitivity analyses in patients under the age of 65 and in patients who died within one month of diagnosis of advanced-stage NSCLC. There were 8332 patients under the age of 65, of which 1391 (16.7%) underwent 2697 procedures. Two thousand ninety-nine Veterans died within one month of diagnosis; 81 (3.9%) of these individuals underwent a total of 129 procedures. Risk factors for receipt of invasive procedures in the last month of life were similar in these cohorts compared with the overall cohort.

Discussion

In this study of Veterans with advanced NSCLC, we found that 15.1% of patients underwent at least one invasive procedure during their last month of life. Most procedures (70%) were classified as minimally invasive. A large variety of procedures were performed, ranging from potentially palliative interventions, such as thoracentesis, to life-sustaining therapies such as intubation and mechanical ventilation, to major surgeries. The use of invasive procedures was strongly and positively correlated with other measures of aggressive care in the last month of life, including ICU admission and the receipt of chemotherapy.

While our data suggest that Veterans with advanced lung cancer frequently undergo aggressive interventions at the end of life, the rates of these interventions were overall lower than those reported in other cohort studies. Kwok et al. recently evaluated the use of invasive procedures in elderly Medicare patients with advanced cancer using data from the Surveillance, Epidemiology, and End Results Registry.¹³ They found that 24.6% of advanced lung cancer patients underwent an invasive procedure in the last month of life. As in our study, the most common procedures were pleural drainage, bronchoscopy, and vascular access. Likewise, while our finding that 9.7% of Veterans experienced an ICU admission in the last month of life is similar to that reported by Keating et al. in a cohort of elderly Veterans with lung and colon cancer,¹⁴ it is far lower than estimates of elderly cancer patient admissions in the private sector (>25%).¹⁵

Our study suggests that certain Veterans with advanced lung cancer may be more likely to receive invasive procedures near the end of their lives than others. Veterans, who were younger, of non-white race, and who had more comorbidities, were more likely to receive invasive procedures. It is not surprising that younger patients would be offered and chose more aggressive treatment options than elderly patients. Similarly, racial variation in treatment intensity at the end of life has been previously described, with patients who identify as black being more likely to undergo invasive procedures at the end of their lives than patients who identify as white.^16,17 Interestingly this association did not hold when we limited the analysis to only life-sustaining and major-operative procedures. However, the number of non-white patients undergoing these procedures was low.

An unexpected finding in our study was that patients who died later in the study period were less likely to receive invasive procedures than patients who died earlier in the study period. In general, prior studies suggest that intensity of end-of-life care in patients both with and without cancer is increasing.^1,8,18,19 The reason for our conflicting finding is unclear. One likely contributing factor is that access to hospice and palliative care services increased significantly within the VHA during the study period.^20,21 Indeed, from 2006 to 2012, the proportion of Veteran decedents with cancer who received hospice care increased from 55% to 68%.²¹ In addition, during this time the VHA mandated that all VA inpatient medical facilities provide palliative care consultation teams for Veterans. Early integration of palliative care with standard oncologic care in the management of patients with advanced NSCLC has been shown to improve patient's quality of life as well as survival, while reducing the use of aggressive care at the end of life.^22,23 With the VA policies promoting palliative care and allowing the concurrent receipt of hospice and cancer treatment, the VA has provided Veterans with earlier access to these supportive services. With the additional support of palliative care and hospice specialists in more recent years, Veterans may be better able to clarify their wishes for aggressive procedures and, particularly if their symptoms are managed and palliated medically, they may have less need or desire to undergo aggressive interventions near the end of life. In fact, in our study, we found that Veterans who were referred to hospice were significantly less likely to undergo invasive procedures at the end of their lives.

In this study, the use of invasive procedures was well correlated with other indicators of aggressive end-of-life care, including ICU admissions and receipt of chemotherapy within the last 30 days of life. This suggests that there is a subgroup of patients who choose to receive very aggressive care at the end of their lives. Understanding the treatment preferences of patients with advanced illness is challenging. It is well known that patients with advanced cancer are often willing to accept treatments with major toxicities for only a small perceived benefit. Complicating this matter, advanced NSCLC is often associated with high symptom burden and treatments that are not expected to provide a survival benefit, may still play an important role in symptom palliation. Despite only modest benefits of chemotherapy in end-stage NSCLC in terms of survival, multiple studies have demonstrated that some types of chemotherapy can reduce cancer-related symptoms even in patients with advanced disease.^24–26 Similarly, some invasive procedures have been clearly demonstrated to have palliative benefits, such as the use of rigid bronchoscopy for the relief of central airway obstruction and the placement of tunneled pleural catheters for the palliation of dyspnea associated with malignant pleural effusions.^27–29

Healthcare providers are faced with the task of helping patients with advanced cancer make informed decisions about various treatment options based on the anticipated burdens of such treatments and the likelihood of expected outcomes. It is likely that some of the procedures performed in this study were done either for palliative purposes or in patients who preferred aggressive care and were willing to accept significantly burdensome therapies for what they perceived to be a survival benefit. It is also possible, however, that some of these patients may not have understood either their prognosis or the small likelihood that the procedures and interventions performed would provide either a survival or symptomatic benefit.³⁰ Identifying patients who might choose not to undergo invasive procedures when presented with a realistic understanding of the risks, benefits, and alternative options for managing their disease and symptoms may represent an opportunity to improve end-of-life care. Further research is necessary to better understand the wishes and expectations of patients who receive procedures at the end of life as well as the impact of these procedures on patients' and caregivers' quality of life.

Our study has a number of important limitations. First, without clinical information regarding patient preferences and the circumstances surrounding the use of invasive procedures, caution must be exercised before making judgment about the appropriateness of such interventions at the end of life. While it is likely that some of the procedures were performed for palliative purposes, it is impossible to know how many procedures fall into this category. Second, our study was restricted to Veterans. Like many studies of Veterans, our study population was primarily male and white. Older Veterans with cancer treated within the VHA have been shown to receive less aggressive end-of-life care than similar men treated in the private sector despite having similar or better survival rates.^14,31 These factors may decrease the generalizability of our findings to other patient cohorts. Third, our study relied on ICD-9 procedure codes documented in administrative data to identify use of invasive procedures. While such a strategy is well validated for identifying the use of the life-sustaining therapies evaluated in this study,^10,32 given the lack of financial incentives to bill for minor procedures within the VHA, it is likely that some minor procedures were missed. We were also unable to capture procedures performed in younger Veterans at non-VA facilities paid for using private insurance. However, our sensitivity analysis that excluded younger decedents yielded similar findings giving us confidence that this limitation does not alter our overall study results. Nevertheless, the frequencies of procedures identified in this study should be regarded as minimum estimates. Finally, because we were examining service use in the last month of life and some patients were alive for less than one month, we did not differentiate when during the last month Veterans began to use hospice, only that they had done so. Other research on Veterans' use of hospice during this same era finds a median hospice stay of ∼20 days.²¹

Providing high-quality care at the end of life for patients with cancer is a priority recognized by patients, caregivers, and professional societies alike.^2–4 Patients diagnosed with advanced lung cancer often have high symptom burden and limited survival.^5,6,23,33 Nevertheless, they frequently receive aggressive measures at the end of life and also tend to underuse hospice services.^2,34,35 In this study, we identified that many Veterans with advanced lung cancer received invasive procedures in the last month of life and that receipt of such procedures was strongly correlated with other measures of aggressive end-of-life care. While some of these procedures may have been palliative, it is likely that some failed to provide meaningful survival or symptomatic benefit and thus may have been more burdensome than beneficial. Further investigation is warranted to better understand the use of invasive procedures near the end of life and to ensure that providers and patients fully understand their potential benefits and burdens to provide optimal care for patients with advanced cancer.

Acknowledgments

This work was supported by a Merit Review Award (IIR 12-121) from the Health Services Research and Development Service, United States Department of Veterans Affairs.

A portion of this work was presented at the American Thoracic Society International Conference in the form of an oral abstract, San Francisco, CA, May 2016.

Footnotes

Author Disclosure Statement

The opinions expressed in this study, do not necessarily reflect the views of the Department of Veterans Affairs or the United States Government. No competing financial interests exist.

Appendix A1.

ICD-9 Codes Used to Identify Invasive Procedures

Procedure	ICD-9 CM procedure codes
Minimally invasive procedures
Bronchoscopy	33.21, 33.22, 33.23, 33.24, 33.27
Thoracentesis	34.91
Chest tube placement	34.04
GI endoscopy	42.23, 42.24, 44.13, 44.14, 45.13, 45.14, 45.16, 45.23, 45.25,
Image-guided biopsies (adrenal, lung, lymph node, liver, bone)	07.11, 33.26, 40.11, 50.11, 77.4, 77.4x
Central venous catheterization	38.93, 38.95
Port placement	86.07
Angiography	00.66, 00.50, 00.51, 00.52, 00.53, 00.54, 00.56, 00.57, 17.51, 17.52, 17.55, 36.01, 36.02, 36.05, 37.70, 37.71, 37.72, 37.73, 37.74, 37.75, 37.76, 37.77, 37.78, 37.79, 37.80, 37.81, 37.82, 37.83, 37.85, 37.86, 37.87, 37.89, 37.94, 37.95, 37.96, 37.97, 37.98, 88.40, 88.41, 88.42, 88.43, 88.44, 88.45, 88.46. 88.47, 88.48, 88.49, 88.50, 88.51, 88.58, 88.60, 88.61, 88.62, 88.63, 88.64, 88.65, 88.66, 88.67, 88.68
Life-sustaining procedures
Mechanical ventilation	96.04, 96.70, 96.71, 96.72
Tracheostomy	31.1, 31.21, 31.29
Feeding tube placement	43.1, 43.11, 43.19, 43.2, 44.32
Dialysis	39.95
CPR	99.60, 99.63
Operative procedures
Lung resection	32.20, 32.21, 32.22, 32.23, 32.24, 32.25, 32.26, 32.27, 32.29, 32.3, 32.30, 32.39, 32.4, 32.41, 32.49, 32.5, 32.50, 32.59
Other thoracic surgeries	30.01, 30.09, 30.1, 30.21, 30.22, 30.29, 30.3, 30.4, 31.3, 31.41, 31.45, 31.48, 31.49, 31.5, 31.61, 31.62, 31.63, 31.64, 31.69, 31.71, 31.72, 31.73, 31.74, 31.75, 31.79, 31.91, 31.92, 31.98, 31.99, 32.0, 32.09, 32.1, 32.6, 32.9, 33.0, 33.1, 33.20, 33.25, 33.28, 33.29, 33.34, 33.39, 33.41, 33.42, 33.43, 33.48, 33.49, 33.92, 33.93, 33.98, 33.99, 34.01, 34.02, 34.03, 34.05, 34.06, 34.09, 34.1, 34.20, 34.21, 34.22, 34.23, 34.24, 34.25, 34.26, 34.27, 34.28, 34.29, 34.3, 34.4, 34.51, 34.52, 34.59, 34.6, 34.73, 34.74, 34.79, 34.81, 34.82, 34.83, 34.84, 34.85, 34.89, 34.93, 34.99

CPR, cardiopulmonary resuscitation; GI, gastroenterology; ICD-9 CM, International Classification of Diseases, 9th revision, Clinical Modification.

References

Earle

, Neville

, Landrum

, et al.: Trends in the aggressiveness of cancer care near the end of life. J Clin Oncol, 2004; 22:315–321.

Earle

, Landrum

, Souza

, et al.: Aggressiveness of cancer care near the end of life: Is it a quality-of-care issue?. J Clin Oncol, 2008; 26:3860–3866.

National Quality Forum: Endorsement Summary: Cancer Measures. http://qualityforum.org/News_And_Resources/Endorsement_Summaries/Endorsement_Summaries.aspx (last accessed April 1, 2017).

ASCO Institute for Quality: QOPI 2015 QCDR Measures. http://instituteforquality.org/qopi/measures. 2017. (Last accessed April 1, 2017).

Kato

, Severson

, Schwartz

: Conditional median survival of patients with advanced carcinoma: Surveillance, epidemiology, and end results data. Cancer, 2001; 92:2211–2219.

Blackstock

, Herndon

2nd , Paskett

, et al.: Outcomes among African-American/non-African-American patients with advanced non-small-cell lung carcinoma: Report from the cancer and leukemia group B. J Natl Cancer Inst, 2002; 94:284–290.

Healthcare Cost and Utilization Project (HCUP): Procedure Classes. https://hcup-us.ahrq.gov/toolssoftware/procedure/procedure.jsp. 2017. (Last accessed April 1, 2017).

Barnato

, McClellan

, Kagay

, et al.: Trends in inpatient treatment intensity among Medicare beneficiaries at the end of life. Health Serv Res, 2004; 39:363–375.

Barnato

, Cohen

, Mistovich

, et al.: Hospital end-of-life treatment intensity among cancer and non-cancer cohorts. J Pain Symptom Manage, 2015; 49:521–529.e1–5.

10.

Barnato

, Farrell

, Chang

, et al.: Development and validation of hospital “end-of-life” treatment intensity measures. Med Care, 2009; 47:1098–1105.

11.

Deyo

, Cherkin

, Ciol

: Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol, 1992; 45:613–619.

12.

Wang

, Wong

, Hamilton

, et al.: Impact of age and comorbidity on non-small-cell lung cancer treatment in older veterans. J Clin Oncol, 2012; 30:1447–1455.

13.

Kwok

, Hu

, Dodgion

, et al.: Invasive procedures in the elderly after stage IV cancer diagnosis. J Surg Res, 2015; 193:754–763.

14.

Keating

, Landrum

, Lamont

, et al.: End-of-life care for older cancer patients in the Veterans Health Administration versus the private sector. Cancer, 2010; 116:3732–3739.

15.

Goodman

DCMN

, Chang

, Fisher

, Wennberg

JE.

: Trends in Cancer Care Near the End of Life: Regional and Hospital-Specific Analyses. A Dartmouth Atlas of Health Care Brief. http://dartmouthatlas.org/downloads/reports/Cancer_brief_090413.pdf (last accessed April 1, 2017).

16.

Barnato

, Chang

, Saynina

, et al.: Influence of race on inpatient treatment intensity at the end of life. J Gen Intern Med, 2007; 22:338–345.

17.

Tschirhart

, Du

, Kelley

: Factors influencing the use of intensive procedures at the end of life. J Am Geriatr Soc, 2014; 62:2088–2094.

18.

Cooke

, Feemster

, Wiener

, et al.: Aggressiveness of intensive care use among patients with lung cancer in the surveillance, epidemiology, and end results-medicare registry. Chest, 2014; 146:916–923.

19.

Sharma

, Freeman

, Zhang

, et al.: Trends in end-of-life ICU use among older adults with advanced lung cancer. Chest, 2008; 133:72–78.

20.

Frahm

, Barnett

, Brown

: Trends in hospice utilization across age among the veteran population. Am J Hosp Palliat Care, 2011; 28:424–428.

21.

Mor

, Joyce

, Cote

, et al.: The rise of concurrent care for veterans with advanced cancer at the end of life. Cancer, 2016; 122:782–790.

22.

Saito

, Landrum

, Neville

, et al.: Hospice care and survival among elderly patients with lung cancer. J Palliat Med, 2011; 14:929–939.

23.

Temel

, Greer

, Muzikansky

, et al.: Early palliative care for patients with metastatic non-small-cell lung cancer. N Engl J Med, 2010; 363:733–742.

24.

Shepherd

, Dancey

, Ramlau

, et al.: Prospective randomized trial of docetaxel versus best supportive care in patients with non-small-cell lung cancer previously treated with platinum-based chemotherapy. J Clin Oncol, 2000; 18:2095–2103.

25.

Anderson

, Hopwood

, Stephens

, et al.: Gemcitabine plus best supportive care (BSC) vs BSC in inoperable non-small cell lung cancer—a randomized trial with quality of life as the primary outcome. UK NSCLC gemcitabine group. Non-small cell lung cancer. Br J Cancer, 2000; 83:447–453.

26.

Hardy

, Noble

, Smith

: Symptom relief with moderate dose chemotherapy (mitomycin-C, vinblastine and cisplatin) in advanced non-small cell lung cancer. Br J Cancer, 1989; 60:764–766.

27.

Davies

, Mishra

, Kahan

, et al.: Effect of an indwelling pleural catheter vs chest tube and talc pleurodesis for relieving dyspnea in patients with malignant pleural effusion: The TIME2 randomized controlled trial. JAMA, 2012; 307:2383–2389.

28.

Roberts

, Neville

, Berrisford

, et al.: Management of a malignant pleural effusion: British Thoracic Society pleural disease guideline 2010. Thorax, 2010; 65 Suppl 2:ii32–ii40.

29.

Ernst

, Feller-Kopman

, Becker

, et al.: Central airway obstruction. Am J Respir Crit Care Med, 2004; 169:1278–1297.

30.

Weeks

, Catalano

, Cronin

, et al.: Patients' expectations about effects of chemotherapy for advanced cancer. N Engl J Med, 2012; 367:1616–1625.

31.

Landrum

, Keating

, Lamont

, et al.: Survival of older patients with cancer in the Veterans Health Administration versus fee-for-service medicare. J Clin Oncol, 2012; 30:1072–1079.

32.

Quan

, Parsons

, Ghali

: Validity of procedure codes in International Classification of Diseases, 9th revision, clinical modification administrative data. Med Care, 2004; 42:801–809.

33.

Bakitas

, Lyons

, Hegel

, et al.: Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer: The Project ENABLE II randomized controlled trial. JAMA, 2009; 302:741–749.

34.

Temel

, McCannon

, Greer

, et al.: Aggressiveness of care in a prospective cohort of patients with advanced NSCLC. Cancer, 2008; 113:826–833.

35.

Mack

, Chen

, Boscoe

, et al.: Underuse of hospice care by Medicaid-insured patients with stage IV lung cancer in New York and California. J Clin Oncol, 2013; 31:2569–2579.