Support of cancer patients' spiritual needs and associations with medical care costs at the end of lifec

Authors

  • Tracy Balboni MD, MPH,

    Corresponding author
    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    4. Department of Radiation Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
    • DFCI, 44 Binney Street, Boston, MA 01225
    Search for more papers by this author
    • Fax: (617) 632-4274

  • Michael Balboni PhD, ThM,

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    Search for more papers by this author
  • M. Elizabeth Paulk MD,

    1. Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas
    Search for more papers by this author
  • Andrea Phelps MD,

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    Search for more papers by this author
  • Alexi Wright MD,

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    4. Division of Women's Cancers, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author
  • John Peteet MD,

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    4. Department of Psychiatry, Brigham and Women's Hospital, Boston, Massachusetts
    Search for more papers by this author
  • Susan Block MD,

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    4. Department of Psychiatry, Brigham and Women's Hospital, Boston, Massachusetts
    Search for more papers by this author
  • Chris Lathan MD,

    1. Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author
  • Tyler VanderWeele PhD,

    1. Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
    2. Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts
    Search for more papers by this author
  • Holly Prigerson PhD

    1. Center for Psycho-Oncology and Palliative Care Research, Dana-Farber Cancer Institute, Boston, Massachusetts
    2. Department of Palliative Care, Dana-Farber Cancer Institute, Boston, Massachusetts
    3. Harvard Medical School Center for Palliative Care, Boston, Massachusetts
    4. Division of Women's Cancers, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts
    Search for more papers by this author

  • c

    We thank Heather Stieglitz, PhD, Maureen Clark, and Richard Timms, MD for material, administrative, and technical support.

Abstract

BACKGROUND:

Although spiritual care is associated with less aggressive medical care at the end of life (EOL), it remains infrequent. It is unclear if the omission of spiritual care impacts EOL costs.

METHODS:

A prospective, multisite study of 339 advanced cancer patients accrued subjects from September 2002 to August 2007 from an outpatient setting and followed them until death. Spiritual care was measured by patients' reports that the health care team supported their religious/spiritual needs. EOL costs in the last week were compared among patients reporting that their spiritual needs were inadequately supported versus those who reported that their needs were well supported. Analyses were adjusted for confounders (eg, EOL discussions).

RESULTS:

Patients reporting that their religious/spiritual needs were inadequately supported by clinic staff were less likely to receive a week or more of hospice (54% vs 72.8%; P = .01) and more likely to die in an intensive care unit (ICU) (5.1% vs 1.0%, P = .03). Among minorities and high religious coping patients, those reporting poorly supported religious/spiritual needs received more ICU care (11.3% vs 1.2%, P = .03 and 13.1% vs 1.6%, P = .02, respectively), received less hospice (43.% vs 75.3% ≥1 week of hospice, P = .01 and 45.3% vs 73.1%, P = .007, respectively), and had increased ICU deaths (11.2% vs 1.2%, P = .03 and 7.7% vs 0.6%, P = .009, respectively). EOL costs were higher when patients reported that their spiritual needs were inadequately supported ($4947 vs $2833, P = .03), particularly among minorities ($6533 vs $2276, P = .02) and high religious copers ($6344 vs $2431, P = .005).

CONCLUSIONS:

Cancer patients reporting that their spiritual needs are not well supported by the health care team have higher EOL costs, particularly among minorities and high religious coping patients. Cancer 2011;. © 2011 American Cancer Society.

Spiritual care is recognition of the role of patients' religion and/or spirituality in illness and attention to religious/spiritual concerns as part of medical care. Spiritual care is an integral component of palliative care 1 supported by data among patients with advanced illness demonstrating: 1) the importance of religion/spirituality in the illness experience, 2-5 including its central role in preserving patient quality of life (QOL) 6, 7; 2) frequent religious/spiritual concerns among these patients 2, 4, 5, 8; and 3) patients' expressed wishes for spiritual care as part of their medical care. 7, 9

Religion/spirituality is associated with patient QOL not only in the setting of advanced illness 6; recent data suggest that religion/spirituality also influences patient end of life (EOL) care. 10-12 For example, religiousness and religious coping (eg, seeking God's support in confronting the stress of illness) are associated with preferences for 3, 11, 13 and receipt of 11 aggressive medical treatment (eg, mechanical ventilation) at the EOL. The receipt of aggressive medical care at the EOL is an outcome with broad implications, given its associations with heightened symptom burden and distress in patients, 14 worse bereavement outcomes in surviving family members, 14 and high costs. 15

Results from the Coping with Cancer study demonstrate that spiritual care is associated with better patient QOL near death and reduced EOL medical care intensity, 10 but remains infrequent. 3 These findings raise the questions: 1) Is the lack of spiritual care by the health care team associated with medical care costs at EOL; and 2) Given higher rates of aggressive EOL care among racial/ethnic minorities 15 and high religious coping patients, 11 is insufficient attention to spiritual needs associated with greater EOL costs among these at-risk patient groups? We hypothesized that patients whose spiritual needs are not well supported by the health care team would have increased EOL medical care costs, and that cost implications would be greatest among racial/ethnic minorities and high religious coping patients.

MATERIALS AND METHODS

Study Sample

Patients were recruited from September 1, 2002 to August 31, 2007 from 7 outpatient sites: Dana-Farber Cancer Institute and Massachusetts General Hospital, Boston, Massachusetts; New Hampshire Oncology-Hematology, Hookset, New Hampshire; Parkland Hospital and Simmons Comprehensive Cancer Center, Dallas, Texas; Veterans Affairs Connecticut Comprehensive Cancer Clinics, West Haven, Connecticut; and Yale University Cancer Center, New Haven, Connecticut. Eligibility criteria included an advanced cancer diagnosis with disease refractory to first-line chemotherapy, age ≥21 years, presence of an informal (nonpaid) caregiver, and adequate stamina to complete the interview. Exclusion criteria for patient-caregiver dyads included patient or caregiver meeting criteria for dementia or delirium by neurocognitive examination or inability to speak English or Spanish. All participants provided written, informed consent according to protocols approved by participating centers' human subjects committees.

Study Protocol

Research staff underwent a 2-day training program in the protocol, chart extraction, and interviewing. Potential participants were identified from outpatient appointment schedules; recruitment did not include specific reference to religion/spirituality. Upon enrollment, patients participated in a baseline interview. Patients' medical records were reviewed to extract disease/treatment information. A second assessment was performed within 3 weeks after the participant's death, including chart extraction to obtain EOL care information and a postmortem interview of a caregiver present during the final week of life.

Of 920 eligible patients approached, 658 (72%) participated. The most common reasons for nonparticipation included “not interested” (n = 109) and “caregiver refuses” (n = 35). There were no significant differences between nonparticipants and participants in sex, age, race, or education. At the time of completion of active study follow-up (August 28, 2008), 366 had died and a postmortem interview was performed; 40 lacked complete postmortem or spiritual care data, resulting in a final sample of 339 (93% of 366).

Baseline Measures

Spiritual care

Patients rated the degree to which their spiritual needs were met by the health care team. Spiritual care was dichotomized to low (spiritual needs supported not at all, to a small extent, and to a moderate extent) versus high (spiritual needs largely or completely supported), consistent with findings from our prior report. 10 A detailed description of measures of spiritual care and religiousness are available in our prior report. 10

Religious variables

Patients rated their religiousness and spiritual support received from religious communities. Pargament's validated religious coping questionnaire 16 measured how patients use their religious beliefs to adapt to the stress of illness. Positive religious coping (eg, “seeking God's love and care”) is the most common form of religious coping. 11, 16 Negative religious coping is less common 11 and often indicates existential distress. 17 Given our previous report indicating greater aggressive EOL care in patients endorsing high (above the median score of 12) positive religious coping, 11 positive religious coping was similarly dichotomized.

Other baseline variables

The previously validated McGill Quality of Life questionnaire 18, 19 was used to assess baseline patient QOL, existential well-being, and social support. The patient-physician relationship, history of an EOL discussion with a physician, presence of advance directives, EOL treatment preferences, and terminal illness awareness were also assessed. Age, sex, race/ethnicity (white; black; Asian American, Pacific Islander, or Indian; Hispanic; or other dichotomized to white vs nonwhite), education (years), and insurance status were patient-reported. Karnofsky performance status was obtained by physician assessment.

EOL Care

Hospice—a predictor of EOL costs 20 and QOL 14 among cancer patients—was recorded as none, <1 week, 1 to 4 weeks, 5 to 8 weeks, 9 to 12 weeks, or >12 weeks, and dichotomized to ≥1 week versus <1 week/no hospice (median split). Days of intensive care unit (ICU) care with and without mechanical ventilation—key determinants of EOL costs 15—were recorded and dichotomized as any versus no ICU care (median split) in the last week. Location of death—another established aggressiveness EOL care measure 21 known to impact costs 22—was recorded as ICU, non-ICU hospital, home, home with hospice, inpatient hospice, or nursing home/long-term care facility and dichotomized to death in an ICU versus other locations. Receipt of resuscitation and chemotherapy in the last week of life was also recorded.

Costs of EOL care

Medical care costs over the last 7 days of life were estimated and adjusted to reflect costs as of January 1, 2010. 23 ICU costs were based on published estimates (nonventilated ICU care, Day 1 = $4646, Day 2 = $4149, Day 3 or more = $3696/day; and ventilated ICU care, Day 1 = $7061, Day 2 = $5666, Day 3 or more = $4946/day). 24 If the number of ICU days was <7 and the patient died in the hospital, the patient was considered to have been in a non-ICU hospital setting for the remaining days (estimated at $1803/day 25), an assumption supported by data indicating that patients with an ICU stay require an average additional 6 days in the non-ICU hospital. 24 Among patients receiving <7 days of ICU care who died in a nonhospital setting, the remaining days and associated costs were split evenly between a non-ICU hospital setting 25 and their death setting (cost estimates of those locations described below). Resuscitation costs were estimated for patients who did not have costs estimated by an ICU admission (as ICU/hospitalization costs include resuscitation costs) and were based on the Agency for Health Research and Quality Healthcare Utilization Project estimate ($10,464, average hospital stay of 4 days). 26 The remaining 3 days were based on average medical costs in the last month of life ($339/day). 15

Hospice costs were based on Centers for Medicare and Medicaid Services (CMS) 2010 reimbursement rates (home hospice = $143/day and inpatient hospice = $636/day). 27 EOL costs for patients receiving 7 or more days of hospice were based on the inpatient or home hospice care daily costs. Patients receiving <1 week of hospice were considered to have 3.5 days of hospice, with the remaining 3.5 days based on average daily costs in the last month of life. 15 Among patients who did not receive hospice or aggressive care, costs were estimated according to their place of death. Patients dying in a hospital were considered to have 2.3 days of hospital stay preceding their death per the Dartmouth Atlas estimates. 28 The remaining 4.7 days were based on average daily costs in the last month. 15 Among patients dying at home, all 7 days were estimated by average daily costs in the last month of life. 15 Patients dying in a long-term care facility were considered to have been in that facility for 7 days given data indicating that median stays are 142 days, 29 with average costs based on average CMS reimbursement rates for 2010 ($399/day). 30 Receipt of chemotherapy in the last week of life was estimated at $821 (25% of estimated $3281 per 4 weeks). 31

Statistical Analysis

Student t tests (continuous variables) and chi-square tests (ordinal and dichotomous variables) were used to compare baseline characteristics among those reporting high versus low spiritual support from the health care team. Logistic regression assessed the relation between patient-reported spiritual care and receipt of ICU care in the last week of life, receipt of at least 1 week of hospice care, and dying in an ICU. Analyses were adjusted for predictors of EOL care, including race, 15, 32-34 EOL treatment preferences, 14, 35 EOL discussion, 14, 36 advance care planning, 37 positive religious coping, 11 and religiousness. 3, 13 EOL care models were also adjusted for variables potentially related to spiritual care and EOL care, including education, health insurance status, recruitment site (Northern vs Southern sites), baseline existential well-being and social support, religious community spiritual support, the patient-physician relationship, and terminal illness awareness. Additional confounds considered were: age, sex, performance status, QOL, and negative religious coping; variables were retained when the P value remained significant (P < .05) after controlling for other confounds. Multivariate models were repeated among racial/ethnic minorities and high religious coping patients. Hierarchical linear modeling with clustering by site was performed to further assess possible confounding by site. Final models were used to generate adjusted rates of any ICU care in the last week of life, receipt of 1 week or more of hospice, and death in an ICU. The same models were used to assess the relationship between patient-reported spiritual care and medical care costs in the last week of life, repeated as an analysis of variance procedure with least square means to obtain adjusted cost estimates for the full sample, racial/ethnic minorities, and high religious coping patients. Statistical analyses were performed with SAS version 9.1 (SAS Institute Inc, Cary, NC). All reported P values are 2-sided and considered significant when <.05.

RESULTS

Sample Characteristics

Baseline sample characteristics are shown in Table 1. Compared with patients reporting low support of their religious/spiritual needs, patients reporting high support of their religious/spiritual needs were more likely to be of nonwhite race/ethnicity, were less likely to have health insurance, were more religious, and endorsed greater positive religious coping. High spiritual support patients also reported greater existential well-being and social support. Recruitment site differed between the 2 groups, with high spiritual support patients being more frequently recruited from Parkland hospital and less frequently from Yale Cancer Center.

Table 1. Baseline Characteristics of the Study Sample by Spiritual Care From the Health Care Team (N = 339)a
CharacteristicTotal Sample, n = 339Low Spiritual Support, n = 252High Spiritual Support, n = 87P
  • Abbreviations: CT, Connecticut; M, mean; MA, Massachusetts; NH, New Hampshire; SD, standard deviation; TX, Texas.

  • a

    There were missing data in <3% of patients for the following variables: health insurance status, Karnofsky performance status, recruitment site, quality of life, existential well-being, social support, religiousness, positive religious coping, negative religious coping, patient-physician relationship, and terminal illness awareness. Data regarding preferences for aggressive treatment measures at end of life were available for 301 patients.

  • b

    Minority race/ethnicity patients include 66 black, 55 Hispanic, and 5 other race patients.

  • c

    Statistically significant.

  • d

    A measure of functional status that is predictive of survival, where 0 is death and 100 is normal health. A validated measure of quality of life with 5 domains: overall, physical, psychological, existential, and social support. Existential items and support items were removed and used as separate predictors.

  • e

    McGill Quality Of Life possible scores range from 0 to 70. Social support possible scores range from 0 to 20. Existential well-being possible scores range from 0 to 60.

  • f

    A measure of positive religious appraisals in coping with illness (eg, “I've been seeking God's love and care”), scale 0 to 21.

  • g

    A measure of negative religious appraisals in coping with illness (eg, “I've been wondering whether God has abandoned me“), scale 0 to 21.

  • h

    Measure of patient-physician relationship (scale 0-5) assessing patient's trust in the physician, sense of being cared for as a whole person, sense of being respected, respect for the physician, and comfort asking questions about care.

Age, M [SD]58.3 [12.6]58.6 [12.5]57.3 [12.8].38
Male, No. (%)183 (54.0)140 (55.6)43 (49.4).32
Nonwhite race, No. (%)b126 (37.2)85 (33.7)41 (47.1).02c
Married, No. (%)200 (59.5)154 (61.6)46 (53.5).19
Education, y, M [SD]12.4 [4.0]12.6 [4.3]11.8 [4.1].09
Health insurance, No. (%)190 (57.1)153 (61.7)38 (44.2).005c
Karnofsky performance status, M [SD]d63.4 [15.8]64.1 [17.6]61 [18.9].13
Recruitment site, No. (%)    
 Yale Cancer Center, CT66 (19.6)61 (24.2)5 (5.9) 
 Veterans Association of Connecticut Cancer Center, CT13 (3.4)12 (4.8)1 (1.2) 
 Simmons Comprehensive Cancer Center,  TX34 (10.1)24 (9.5)10 (11.8) 
 Parkland Hospital, TX152 (45.1)103 (40.9)49 (57.7) 
 Dana-Farber Cancer Institute/ Massachusetts General Hospital, MA7 (2.1)7 (2.8)0 (0) 
 New Hampshire Oncology-Hematology,  NH65 (19.3)45 (17.9)20 (23.5)<.001c
McGill Quality of Life, M [SD]e 45.2 [14.0]45.0 [15.1].90
Existential well-being, M [SD]e 44.9 [10.4]47.8 [(9.7].02c
Social support, M [SD]e 17.0 [3.5]18.1 [2.8].006c
Religiousness, No. (%)    
 Not at all important38 (11.2)36 (14.3)2 (2.3) 
 Somewhat important71 (20.9)64 (25.4)7 (8.0) 
 Very important230 (67.9)152 (60.3)79 (89.8)<.001c
Positive religious coping, M [SD]f11.1 [6.4]10.2 [6.5]13.7 [5.4]<.001c
Negative religious coping, M [SD]g2.0 [3.5]1.8 [3.4]2.5 [3.9].14
Spiritual support from religious communities, No. (%)    
 Not at all110 (32.5)101 (40.1)9 (10.2) 
 To a small extent42 (12.4)39 (15.5)4 (4.6) 
 To a moderate extent43 (12.7)34 (13.5)9 (10.2) 
 To a large extent54 (15.9)35 (13.9)19 (21.6) 
 Completely supported90 (26.6)43 (17.1)47 (53.4)<.001c
End of life discussion with a physician124 (36.6)88 (34.9)37 (42.1).23
Patient-physician relationship, M [SD]h4.8 [0.5]4.7 [0.5]4.9 [0.4].09
Preference for aggressive treatment measures at end of life, No. (%)86 (28.6)66 (30.0)20 (24.4).34
Terminal illness awareness, No. (%)86 (28.6)111 (45.0)31 (37.6).22
Advanced care planning, No. (%)186 (54.9)143 (56.8)43 (49.4).24

Medical Care Received in the Last Week of Life

Patients died a median of 111 (interquartile range, 54-249) days after the baseline interview. In the last week of life, 35 patients (10.3%) received ICU care, and 25 (7.4%) died in an ICU. Most (n = 248, 73.1%) received hospice care at EOL, and 199 (58.7%) received a week or more of hospice care. Adjusted estimates of receipt of ICU care in the last week of life, death in an ICU, and 1 week or more of hospice care according to receipt of spiritual care are shown in Figures 1 and 2. In comparison to those reporting high spiritual support, patients reporting that their religious/spiritual needs were poorly supported were less likely to receive a week or more of hospice care and were more likely to die in an ICU, with the impact being greatest among racial/ethnic minority and high religious coping patients. Furthermore, both racial/ethnic minority and high religious coping patients had higher estimated rates of receiving ICU care in the last week of life when they reported their spiritual needs were poorly supported by the health care team.

Figure 1.

Adjusted estimates of ICU admissions in the last week of life and of ICU deaths by provision of patient-reported spiritual care from the health care team in the complete sample, racial/ethnic minorities, and high religious coping patients (n = 303). aEstimates were adjusted for education, race/ethnicity, baseline existential well-being, baseline social support, advance care planning, end-of-life discussion, health insurance status, patient-physician relationship, degree of positive religious coping, preferences for aggressive care, religiousness, spiritual support from religious communities, Northern versus Southern recruitment site, and terminal illness awareness. Models repeated with cluster analysis by site with findings unchanged. Sample reduced from 339 because of missing data; findings unchanged when analyses repeated with data imputed to mean values.

Figure 2.

Adjusted estimates of receiving a week or more of hospice care by patient-reported receipt of spiritual care from the health care team in the complete sample, among nonwhite race/ethnicity patients and high religious coping patients (N = 303). aEstimates were adjusted for education, race/ethnicity, baseline existential well-being, baseline social support, advance care planning, end-of-life discussion, health insurance status, patient-physician relationship, degree of positive religious coping, preferences for aggressive care, religiousness, spiritual support from religious communities, Northern versus Southern recruitment site, and terminal illness awareness. Models repeated with cluster analysis by site with findings unchanged. Sample reduced from 339 because of missing data; findings unchanged when analysis repeated with data imputed to mean values.

Medical Care Costs in the Last Week of Life

Adjusted estimates of medical care costs in the last week of life according to patient-reported receipt of spiritual care are shown in Figure 3. The estimated difference in EOL care costs was $2441 (95% confidence interval [CI], $488-$4394) for patients reporting high support of their spiritual needs by the health care team as compared with those reporting receiving less spiritual support, with greater differences seen among racial/ethnic minorities ($4206; 95% CI, $544-$7869) and high religious coping patients ($4060; 95% CI, $1331-$6789).

Figure 3.

Costs of medical care in the last week of life by patient-reported receipt of spiritual care from the health care team in the total sample and among nonwhite race/ethnicity and high religious coping patients (N = 291). aLeast square mean estimates of costs were adjusted for education, race/ethnicity, baseline existential well-being, baseline social support, advance care planning, end-of-life discussion, health insurance status, patient-physician relationship, degree of positive religious coping, preferences for aggressive care, religiousness, spiritual support from religious communities, Northern versus Southern recruitment site, and terminal illness awareness. Models repeated with cluster analysis by site with findings unchanged. Sample reduced from 339 because of missing data; findings unchanged when analysis repeated with data imputed to mean values.

DISCUSSION

Spiritual care is a key component of palliative care 1 that has previously been shown to improve the well-being of dying patients and facilitate their transitioning to less-aggressive, QOL-focused care at the EOL. 10 However, spiritual support from the health care team remains infrequent in the care of terminally ill patients. 3 This study demonstrates that advanced cancer patients who report that their religious/spiritual needs are insufficiently supported by the health care team have increased medical costs in the final week of life. EOL costs among advanced cancer patients reporting low support of their religious/spiritual needs by the health care team were $2441 more on average as compared with those who reported themselves well supported. Notably, low religious/spiritual support among racial/ethnic minority patients and high religious coping patients was associated with greater cost differences in the last week of life, on average $4206 more among minorities and $4060 more among high religious coping patients. These findings are robust considering adjustment for multiple potential confounding factors such as EOL discussions and geographic location. Hence, the frequent absence of spiritual care by the medical team 3—culturally competent care mandated by national guidelines 1—is associated with decreased well-being of dying patients 10, 38 and their families, 14 and increased costs because of greater futile, aggressive care at the EOL.

Studies indicate that EOL costs are largely driven by aggressive measures, such as ICU care, 15 whereas increased hospice is associated with lower health care costs. 20 Provision of spiritual care appears to influence costs at the EOL by leading to greater hospice utilization and facilitating earlier referrals to hospice while decreasing ICU care and deaths, particularly among patients known to be at greater risk for aggressive care—racial/ethnic minorities 34, 39, 40 and high religious coping patients. 11 Religion/spirituality has been shown to play a key role in the experience of EOL 2, 3 and in EOL decision making, 12 particularly among minorities. 34, 39, 40 This study's demonstration that low patient-reported support of their spiritual needs is associated with increased aggressive care among minorities, beyond the contribution of religious coping and religiousness, highlights the central role that religion/spirituality often plays in the experience of terminal illness among African American 40 and Latino patients. 34 Attention to patients' religion/spirituality as part of medical care appears to assist terminally ill patients in avoiding burdensome, aggressive medical care at the EOL. Potential causal mechanisms include facilitating resolution of spiritual needs and distress that would otherwise result in more aggressive and QOL-compromising care. Recognition of patients' religion/spirituality as part of EOL care may also assist patients in transitioning away from a focus on extending life to a focus on spiritual priorities at the EOL, such as finding spiritual peace—a factor of primary importance to patients facing death 7 that is associated with less aggressive care at EOL. 41 Involving patients' religion/spirituality in medical decision making may assist patients in recognizing less aggressive EOL care options that remain consistent with their religious/spiritual beliefs.

The implications of the provision of spiritual care to dying patients are noteworthy given that it represents higher quality EOL care 1, 10 that, despite its presence in national care quality standards, 1 frequently remains absent at the EOL. 3 The projected economic impact is approximately $1.4 billion ($2441 × 562,340 annual cancer deaths 42) for care delivered in the last week of life, or approximately 1.5% of direct cancer costs per year ($1.3 billion of $93.2 billion). 42 Furthermore, this study's cost estimates only comprise cancer care in the last week of life; spiritual care could result in greater cost implications if other EOL clinical settings were included and examined over a longer time period. Whereas some hypothesize that ample supply of aggressive technologies creates unnecessary demand, 28 our study suggests that medical demand is impacted by medical system engagement of underlying psychosocial issues that mediate EOL decision making.

This study is limited by the finding that cost calculations were based on estimates present in the literature rather than medical claims data. However, by using national cost averages that include unaccounted for items such as type of chemotherapy or laboratory costs, these estimates provide a justifiable approximation of the overall impact of spiritual care on US EOL medical care costs. Furthermore, cost calculations solely represent costs to medical payers and do not include costs to patients and their families, such as caregiving costs and lost employment revenue. Another limitation is the unclear content of the spiritual care patients reported receiving from their health care team. Data from our prior report assists in characterizing spiritual care, however. First, patients' ratings of support of their religious/spiritual needs are significantly associated with receiving pastoral care visits, 10 supporting the convergent validity of the spiritual care measure and suggesting that pastoral care is a key aspect of spiritual care. Furthermore, although spiritual care from the health care team is associated with less intensive EOL care and reduced costs, receipt of pastoral care visits alone does not predict medical care received at the EOL. 10 This finding is consistent with the finding that medical caregivers such as physicians and nurses are the primary health providers assisting patients in medical decision making, and reinforces the multidisciplinary, collaborative approach to spiritual care outlined in recent guidelines. 43 In addition, patients' ratings of spiritual support from the health care team and their associations with EOL outcomes appear distinct from patients' perceptions of their relationships with their providers and of care communication, as the associations of spiritual support with EOL care and costs adjusted for the patient-physician relationship (5-item measure of trust, mutual respect, feeling viewed as a whole person, and comfort asking questions regarding care) and history of an EOL discussion with a physician. Interestingly, in contrast to patients' overwhelmingly positive perceptions of their medical care and care communication, 44 a minority (26%) reported high support of their religious/spiritual needs by their care providers, further bolstering confidence that patients are identifying spiritual care as absent from their medical care and distinct from other aspects of medical care provision. Future studies are required to describe the specific content of spiritual care from the medical team and its impact on patient and family well-being, medical decision making, and medical costs.

In conclusion, infrequent spiritual care by the health care team is associated with higher medical care costs during the final week of life, particularly among racial/ethnic minorities and high religious coping patients. These findings highlight the importance of spiritual care guidelines 1 and of educating health care providers in their appropriate role in spiritual care provision. 43 The integration of spiritual care into the care of dying patients holds promise to improve patients' 10 and bereaved caregivers' 14 well-being, while also avoiding health care costs because of futile, aggressive care at the EOL.

FUNDING SOURCES

Supported by grant No. MHJ63892 from the National Institute of Mental Health, grant No. CA 106370 from the National Cancer Institute (H.P.), the Fetzer Foundation (H.P., M.E.P.), and an ASCO Young Investigator Award (T.B.). The funding organizations had no role in the design and conduct of the study; collection, analysis, or preparation of the data; or preparation, review, or approval of the manuscript.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary