Effect of Palliative Care Services on the Aggressiveness of End-of-Life Care in the Veteran's Affairs Cancer Population

Abstract

Background:

Cancer care near the end of life (EOL) has become more aggressive over the years. Palliative care services (PCS) may decrease this aggressive cancer care in terminally ill cancer patients. Our objective was to observe the aggressiveness of cancer care near the EOL among Veterans Affairs cancer patients before and after the institution of a PCS team. We also assessed the time taken prior to death to initiate a PCS consultation and its effect on the aggressiveness of cancer care near the EOL.

Methods:

This is a retrospective chart review analysis performed at the local Veterans Affairs hospital looking at the last 100 patients in each of the years, 2002 and 2008, who died with active cancer. Only patients in 2008 had access to a PCS team.

Results:

In the last 30 days of life, compared to 2002, patients in 2008 had a higher incidence of: chemotherapy administration, more than one hospital admission, more than 14 days of hospital stay, intensive care unit admissions, and in-hospital deaths. Patients with timely PCS consults in 2008 appeared to have a lower incidence of: chemotherapy administration, more than one emergency department visit, more than one hospital admission, more than 14-day hospital stays, intensive care unit admissions, and deaths in the hospital. Timely PCS consults were associated with earlier and more frequent hospice referral.

Conclusions:

Cancer care near the EOL has become more aggressive with time at one of the hospitals in the Veterans Affairs healthcare system (VAHS). Institution of a PCS service was unable to completely decrease this trend of increasing aggressiveness of cancer care near the EOL. However, timely PCS consults may help attenuate this aggressiveness.

Introduction

Cancer mortality is a major cause of death despite advances in the early detection and treatment of cancer.¹ A large number of cancer patients each year face challenging choices about issues pertaining to cancer care near the end of life (EOL).² One issue that frequently arises near the EOL involves choosing between aggressive cancer care such as the administration of chemotherapy and treating malignancy-related complications with hospital admissions or emergency department visits, all of which is done with the hope of prolonging life. Such aggressive cancer care may have little therapeutic benefit toward the EOL³ versus an approach that focuses on symptom management, comfort, and hospice. Hospice has been shown to improve the quality of EOL care⁴ whereas aggressive cancer care near the EOL has been associated with increased mortality as well as lower quality of care due to unmet patient emotional and symptom needs as well as decreased family satisfaction.^5,6 Nevertheless, several observational studies have demonstrated an increased tendency over the past decade to use aggressive cancer care measures near the EOL among the private sector of health care.^7–9 The development of palliative care services (PCS) to help ameliorate the suffering in dying patients and anguished family members has been shown to improve overall quality of care.¹⁰ The effect of PCS on aggressiveness of cancer care near the EOL is unknown in the Veterans Affairs health care system (VAHS). The objectives of this study were to observe the aggressiveness of cancer care near the EOL in a cohort of cancer patients before and after the employment of a PCS team in a Veterans Affairs Hospital. We also wanted to assess the time taken prior to death to initiate a PCS consultation and analyze its effect on the aggressiveness of cancer care near the EOL.

Methods

Setting

The Veterans Affairs Hospital in Omaha, Nebraska, is a 200-bed acute care, academic hospital. It serves the entire state of Nebraska and western Iowa. The PCS began in October 2003. In 2008, it consisted of five members: a medical physician, a social worker, a pharmacist, a chaplain, and a program support coordinator. The PCS team is an inpatient and outpatient consulting team primarily consulted for advanced care planning, goals of care discussions, EOL issues, and pain or symptom management. The hospital does not have an inpatient palliative care unit.

Study design and variables analyzed

This is a single institution, retrospective analysis of data obtained from the computerized patient record system (CPRS) provided by the VAHS. This analysis included the last 100 patients in 2002 as well as the last 100 patients in 2008 who had a known diagnosis of malignancy and died with active cancer while receiving care from the Veterans Affairs Hospital in Omaha. Patients were excluded from this cohort if they had not spend at least the last 30 days of their life being followed up or registered with the Veterans Affairs Hospital in Omaha. The 2 years were chosen as representative of the times before and after the establishment of PCS as it was available only to patients in the 2008 group. Although the PCS team was set up in 2003, it was not until 2008 that it comprised all five essential members. Hence, data from 2008 ensures that the PCS team was able to establish consistent care practices by that time as well as provide enough time for a culture shift to occur amongst the consulting services regarding the utility and availability of the PCS team. The Veterans Affairs Hospital of Omaha's Institutional Review Board approved the project protocol before the initiation of the study.

We performed individual patient chart reviews using CPRS and recorded the following variables: age, gender, race, site and stage of cancer, administration of chemotherapy, date of death, date of PCS consult, date of hospice initiation, hospice duration, and location of patient in the last 30 days of life. Race was evaluated as white, black, or other. Based on the date of the PCS consult, patients in 2008 were categorized into the following groups: PCS consultations more than 2 weeks before death, PCS consultations less than 2 weeks before death, or no PCS consultation. We defined any consultation at least 2 weeks prior to death as a timely consultation since it allowed the PCS team, as per their consensus, to have a minimum required time to establish rapport with the patient and their family. Moreover, it also ensured that the PCS team was available for at least half of the intended outcome measurement period which was the last 30 days before death.

Aggressiveness of cancer care near the EOL was examined by a composite measure adapted from Earle et al.¹¹ that has been validated and used as a point of reference for many palliative care studies. This variable was assessed by six indicators in the last month of life: use of chemotherapy, more than one emergency department visit, more than one hospital admission, more than 14 days of hospitalization, an intensive care unit (ICU) admission, and death in a hospital.

Statistical analysis

This study involved a descriptive analysis of several variables of aggressiveness of cancer care near the EOL in different groups. χ² analysis was used to compare the differences among categorical variables regarding the aggressiveness of cancer care experienced between the different groups of patients. Differences between patient groups in terms of cumulative indicators of aggressive cancer care were initially performed using a univariate analyses via χ² method for all categorical variables. On the basis of variables found to be significantly associated with the composite measure of aggressiveness of cancer care on univariate results, logistic regression was used to examine the simultaneous influence of multiple predictors. A two-sided p value <0.05 was considered to be statistically significant.

Results

Patient characteristics

Median age of patients in 2002 was 74 years and in 2008 was 72 years. Almost all patients were male and about 90% were Caucasians in both groups. Lung cancer and prostate cancer were the two most common tumor types in both years. There were 86 patients in the 2002 group that had metastatic cancer at time of death versus 69 in the 2008 group. The remaining patients had active locally advanced cancer. Patient characteristics are displayed in Table 1.

Table 1.

Patient Characteristics

Patient Characteristics	2002	2008
No. of patients	100	100
Median Age (years)	74	72
Male	96	100
Race
Caucasian	88	91
African American	8	7
Other	4	2
Primary tumor site
Lung	55	35
Prostate	10	14
Pancreas	3	8
Colon	9	7
Other gastrointestinal	2	9
Head and Neck	1	5
Hematologic malignancy	10	8
Other	10	14
Stage
Regional	14	31
Metastatic	86	69
Palliative care consults	N/A	58
Hospice referral	69	60

Aggressiveness of care in 2002 versus 2008

The group of patients in 2008 received more aggressive cancer care measures near the EOL when compared to patients in 2002 (Fig. 1). In the last 30 days of life, cancer patients in 2008 had a higher incidence of chemotherapy administration (18% versus 5%, p=0.004), more than one hospital admission (16% versus 6%, p=0.009), more than 14 days of hospital stay (38% versus 8%, p<0.001), more number of intensive care unit admissions (33% versus 6%, p<0.001) and more number of deaths in an acute care hospital (38% versus 18%, p=0.001). However, there was no statistical significant difference between the number of patients in both groups experiencing more than one emergency department visit (9% versus 6%, p=0.42). Finally, there were 69 hospice referrals among the patients in 2002 versus 67 among the 2008 patients (p=0.762). Of the 67 hospice referrals received in 2008, 33 (49%) were initiated less than three days before the patient's death compared to 25 (36%) of the 69 hospice referrals in 2002 (p=0.124).

FIG. 1.

Trends of aggressive end-of-life (EOL) care in the last month of life (2002 vs. 2008).

Effects of PCS on EOL care

There were 58 patients in the 2008 group who received PCS consultations. Of these 58 patients, 25 (43%) had PCS consults 2 or more weeks before their death and were considered to be timely in nature. Table 2 delineates the aggressiveness of cancer care near the EOL among the patients in 2008 based on the timeliness of their PCS consults. In the last 30 days of life, compared to patients without PCS consults or consults less than 2 weeks before death, patients with timely PCS consults had a lower incidence of: chemotherapy administration (12% versus 20%), more than one emergency department visit (0% versus 12%), more than one hospital admission (8% versus 19%), more than 14-day hospital stay (24% versus 43%), intensive care unit admissions (24% versus 36%), and in-hospital deaths (24% versus 43%). However, these observed differences were not statistically significant. Patients with timely PCS consults also appeared to have more number of hospice referrals (76% versus 64%, p=0.06). Furthermore, only 4 (16%) of the 25 patients in this group received hospice referral less than 3 days before death compared to 29 (39%) of the 75 patients with no PCS consults or delayed PCS consults (p=0.035). Interestingly, regardless of time to consult, 76% of the patients with PCS consults had hospice referrals compared to 55% of the patients without PCS consults (p=0.027).

Table 2.

Aggressiveness of End-of-Life Care among the Patients Based on the Timeliness of the Palliative Care Services Consult

	No. of pts based on time from PCS consult to death
	2002	2008
Last 30 days of life	No PCS	>2 weeks	<2 weeks	No PCS
Number of patients	100	25	33	42
Received chemotherapy	5 (5%)	3 (12%)	8 (24%)	7 (17%)
>1 emergency department visit	6 (6%)	0 (0%)	5 (15%)	4 (10%)
Hospitalized >14 days	8 (8%)	6 (24%)	16 (48%)	16 (38%)
>1 hospital admission	6 (6%)	2 (8%)	10 (30%)	4 (10%)
ICU admission	6 (6%)	6 (24%)	14 (42%)	13 (31%)
<3 days in hospice	25 (25%)	4 (16%)	13 (39%)	16 (38%)
Hospice consult	69 (69%)	19 (76%)	25 (76%)	23 (55%)
Death in acute care hospital	18 (18%)	6 (24%)	12 (36%)	20 (48%)

PCS, palliative care services; ICU, intensive care unit.

Patients in 2008 with no PCS consult or delayed PCS consult had a higher cumulative number of aggressive cancer care indicators near the EOL when compared to patients in 2008 with timely PCS consults (Fig. 2). Upon performing a logistic regression analysis, receiving timely PCS consults was the only significant predictor of having a less likelihood of two or more aggressive cancer care indicators near the EOL in the 2008 group (Table 3). Interestingly, patients in 2002 were also found to be less likely of having two or more aggressive cancer care indicators near the EOL when compared to patients in 2008; this was also found to be significant in a logistic regression analysis (Table 3).

FIG. 2.

Effect of time and palliative care services (PCS) on aggressive end-of-life (EOL) care.

Table 3.

Multivariate Analysis of Factors Predicting the Presence of Two or More Indicators of Aggressive Cancer Care Near the End of Life

Factors	Univariate analysis	Multivariate analysis
Patients from 2008
Timely PCS^a	p=0.028	OR=0.280 95% CI=0.090–0.869
Stage	p=0.628	N/A
Primary site	p=0.699	N/A
Race	p=0.291	N/A
Age	p=0.903	N/A
Patients from 2002 and 2008
Year^b	p=0.003	OR=0.254 95% CI=0.121–0.534
Stage	p=0.417	N/A
Primary site	p=0.881	N/A
Race	p=0.609	N/A
Age^c	p=0.003
<60		OR=3.540 95% CI=1.488–8.426
61–79		OR=3.715 95% CI=1.613–8.556

Compared to patients with no PCS consult or delayed PCS consult.

Compared to patient in 2008.

Compared to patients 80 years and older.

PCS, palliative care services; OR, odds ratio; CI, confidence interval; N/A, not available.

Discussion

In this study, we show an increased propensity to use aggressive measures of cancer care during the last 30 days of life in 2008 when compared to 2002. The institution of a PCS service was unable to decrease this use of aggressive cancer care near the EOL. However, there was a trend toward less aggressive cancer care near the EOL among patients who received timely PCS consultations, which was not statistically significant.

Studies have demonstrated that the cancer treatment of patients at the EOL have become increasingly aggressive over time.^7–9 Aggressive cancer care near EOL is linked to increased mortality, lower quality of patient care, and decreased family satisfaction.^5,6 Moreover, it also results in disproportionately greater health expenditures.³ Keating et al.¹² compared the aggressiveness of cancer care near the EOL in the VHA to the private sector using the SEER-Medicare database. Compared to patients enrolled in fee-for-service Medicare, patients in the VHA were less likely to receive aggressive cancer care near the EOL. However, our study is the first to assess the trend of aggressiveness of cancer care near the EOL in a single Veteran's Affairs hospital before and after the institution of a PCS team. Even though the aggressiveness of care is shown to be generally less amongst VHA hospitals versus the private sector,¹² we demonstrate that perhaps the VHA system is not immune from the global phenomenon of increasing trend of aggressive cancer care near the EOL with time. With the last decade seeing an expansion in the therapeutic possibilities for patients with advanced cancer, it is plausible that the Veterans Affairs oncologists and cancer care teams have several more therapeutic options to offer leading to more aggressive EOL care.

With the increasing aggressiveness of cancer care near the EOL among cancer patients, improving cancer care near the EOL has become a priority and is viewed as integral to high-quality chronic care management.¹³ The National Comprehensive Cancer Network (NCCN) recommends the introduction of palliative care to all cancer patients at initial diagnosis and its integration during treatment to address symptoms and psychosocial issues.¹⁴ With its focus on management of symptoms, psychosocial support and assistance with decision making, PCS improves the quality of care and reduces cost of health care.^9,15 However, the time of initiation of the PCS consult is also of significant importance in terms of cancer care near the EOL. Temel et al¹⁶ found that among patients with metastatic non–small–cell lung cancer, patients who received early PCS consultation upon diagnosis had less aggressive cancer care near the EOL, increased overall survival and better quality of life and mood when compared to patients who did not receive an automated early PCS consult on diagnosis. The same study also observed that late referrals to PCS are inadequate to alter the quality and delivery of care provided to patients with cancer. In order to have a meaningful effect on patients' quality of life and cancer care near the EOL, PCS must be provided earlier in the course of the disease.¹⁷ Similarly, in our study, there were lower incidences of aggressive cancer care indicators near the EOL in patients in 2008 with timely PCS consults versus patients in 2008 with no PCS or delayed PCS consults (Fig. 2); however, this was not statistically significant. This could be due to a small sample size of only a hundred patients in each group. Nevertheless, we did demonstrate that timely PCS consultations are associated with earlier and more frequent hospice referrals avoiding aggressive life-prolonging cancer care. In fact, even patients receiving a PCS consult irrespective of the time to consultation were more likely to get a hospice consult than those patients who never had a PCS consult.

There were several limitations in our study. Apart from its retrospective nature, the patients from 2002 and 2008 were not matched for their demographic and disease characteristics perhaps leading to selection bias. Moreover, we were unable to control for any sort of referral-based selection bias based on the medical conditions and performance statuses of the patients in either groups or even the practice patterns of the individual consulting practitioners taking care of the patients. The small sample size of our patient cohort made it difficult to provide valuable statistical analysis on differences between groups of patients. Moreover, we are hopeful that the 5-year interval between analysis points was long enough to provide consistent PCS care practices along with providing a culture change in the acknowledgement of the utility of a PCS team. If not, then this may too have confounded our results. Finally, the lack of socioeconomic condition, religion, marriage status and other social demographics makes it hard to uncover all the factors responsible for aggressive cancer care near the EOL.

In conclusion, we believe that our study demonstrates that just like the rest of the private sector, cancer care near the EOL has become more aggressive with time at one Veterans Affairs hospital. The institution of a PCS team did not appear to decrease the aggressiveness of cancer care near the EOL. Nevertheless, if consulted in a timely fashion, PCS consults may potentially decrease the extent of aggressive cancer care near the EOL.

Footnotes

Acknowledgment

Part of this study was presented at the 2010 national ASCO meeting in Chicago, Illinois.

Author Disclosure Statement

No competing financial interests exist.

References

Centers for Disease Control and Prevention: Cancer survivorship: United States, 1971–2001. MMWR, 2004; 53:526–529.

Arias

, Anderson

, Kung

, Murphy

, Kochanek

. Deaths: Final data for 2001. Natl Vital Stat Rep, 2003; 52:1–100.

Fisher

, Wennberg

, Stukel

, Gottlieb

, Lucas

, Pinder

. The implications of regional variations in Medicare spending: Part 2: Health outcomes and satisfaction with care. Ann Intern Med, 2003; 138:288–298.

Von Gunten

. Interventions to manage symptoms at the end of life. J Palliat Med, 2005; 8:S88–S94.

Teno

, Mor

, Ward

et al. Bereaved family member perceptions of quality of end-of-life care in U.S. regions with high and low usage of intensive care unit care. J Am Geriatr Soc, 2005; 53:1905–1911.

Norris

, Merriman

, Curtis

, Clarridge

, Wennberg

, Fisher

. Next of kin perspectives on the experience of end-of life care in a community setting. J Palliat Med, 2007; 10:1101–1115.

Earle

, Neville

, Landrum

, Ayanian

, Block

, Weeks

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

Tang

, Wu

, Hung

, Chen

, Huang

, Liu

. Determinants of aggressive end-of-life care for Taiwanese cancer decedents, 2001 to 2006. J Clin Oncol, 2009; 27:4613–4618.

Barnato

, McClellan

, Kagay

, Garber

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

10.

Bakitas

, Lyons

, Hegel

, Balan

, Brokaw

, Seville

, Hull

, Li

, Tosteson

, Byock

, Ahles

. Effects of a palliative care intervention on clinical outcomes in patients with advanced cancer. JAMA, 2009; 302:741–749.

11.

Earle

, Park

, Lai

, Weeks

, Ayanian

, Block

. Identifying potential indicators of the quality of end-of-life cancer care from administrative data. J Clin Oncol, 2003; 21:1133–1138.

12.

Keating

, Landrum

, Lamont

, Earle

, Bozeman

, McNeil

. End-of-life care for older cancer patients in the veterans health administration versus the private sector. Cancer, 2010; 116:3732–3739.

13.

Meier

, Casarett

, von Gunten

, Smith

, Storey

. Palliative medicine: Politics and policy. J Palliat Med, 2010; 13:1–6.

14.

National Comprehensive Cancer Network. Clinical practice guidelines in oncology. Palliat Care, 2009. www.nccn.org/professionals/physician_gls/PDF/palliative.pdf. 2010 November 1.

15.

Morrison

, Penrod

, Cassel

, Caust-Ellenbogen

, Litke

, Spragens

, Meier

. Cost savings associated with US hospital palliative care consultation programs. Arch Intern Med, 2008; 168:1783–1790.

16.

Temel

, Greer

, Muzikansky

, Gallagher

, Admane

, Jackson

, Dahlin

, Blinderman

, Jacobsen

, Pirl

, Billings

, Lynch

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

17.

Morita

, Akechi

, Ikenaga

, Kizawa

, Kohara

, Mukaiyama

, Nakaho

, Nakashima

, Shima

, Matsubara

, Uchitomi

. Late referrals to specialized palliative care service in Japan. J Clin Oncol, 2005; 23:2637–2644.