Copyright B 2015 Wolters Kluwer Health, Inc. All rights reserved.

Li N. Liu, PhD Chen H. Chen, MSN Tsang W. Liu, MD Yu C. Lin, PhD Shiuyu C.K. Lee, DNSc Siew T. Tang, DNSc

Preferences for Aggressive End-of-life Care and Their Determinants Among Taiwanese Terminally Ill Cancer Patients K E Y

W O R D S

Background: Studies on factors influencing preferences for aggressive end-of-life

Cardiopulmonary resuscitation

(EOL) care have focused predominantly on preferred goals of EOL and seldom

Goals of end-of-life care

their analyses. Objective: The aim of this study was to investigate the determinants

Life-sustaining treatments

of preferences for a wide range of aggressive EOL care from the aforementioned

Preferences for end-of-life care

factors. Methods: A cross-sectional survey was conducted using a convenience

Terminally ill cancer patients

comprehensively incorporate patients’ predisposing, enabling, and need factors into

sample of 2329 terminally ill cancer patients recruited from 23 hospitals throughout Taiwan. Results: Among these Taiwanese terminally ill cancer patients, 8.2% preferred prolonging life as their goal for EOL care. When combining those who wanted and those who were undecided as wanting that specific treatment, 27.9% preferred cardiopulmonary resuscitation when their life was in danger, and 36.0%, 27.3%, 24.3%, and 26.7% preferred to receive care at intensive care unit, cardiac massage, intubation, and mechanical ventilation support, respectively. Those at risk of preferring aggressive EOL care were men, younger than 45 years, married, diagnosed within 6 months, and with comorbidity and their physician had not accurately disclosed their prognosis or discussed EOL care issues to/with them. Conclusions: Few Taiwanese terminally ill cancer patients preferred to prolong life as their goal for EOL care, cardiopulmonary resuscitation when their life was in

Author Affiliations: Department of Nursing, Fu Jen Catholic University, New Taipei City, Taiwan (Dr Liu); Department of Nursing, Kang-Ning Junior College of Medical Care and Management, and School of Nursing, Chang Gung University, Tao-Yuan, Taiwan (Ms Chen); National Institute of Cancer Research, National Health Research Institutes, Taipei, Taiwan (Dr Liu); Department of Nursing, Tzu Chi University, Hualien, Taiwan (Dr Lin); Department of Nursing, National Taipei University of Nursing and Health Science, Taiwan (Dr Lee); School of Nursing, Medical College, Chang Gung University, Tao-Yuan, Taiwan (Dr Tang). Bureau of Health Promotion, Department of Health, Taiwan (DOH 9911020C), with partial support from National Health Research Institute (NHRI-EX101-9906PI and NHRI-EX102-10208PI).

Preferences for Aggressive End-of-life Care

None of the funding sources had any role in designing and conducting the study; collecting, managing, analyzing, and interpreting the data; or preparing, reviewing, or approving the manuscript. The authors have no conflicts of interest to disclose. Correspondence: Siew T. Tang, DNSc, School of Nursing, Medical College, Chang Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan, Tao-Yuan, Taiwan 333, ROC ([email protected]). Accepted for publication March 1, 2014. DOI: 10.1097/NCC.0000000000000155

Cancer NursingTM, Vol. 38, No. 3, 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

n

E9

danger, and life-sustaining treatments at EOL. Preferences for aggressive EOL care are determined by patients’ predisposing, enabling, and need factors. Implications for Practice: Terminally ill cancer patients at risk of preferring aggressive EOL care should receive interventions to help them appropriately weigh the burdens and benefits of such aggressive treatments.

T

reatment preferences of terminally ill cancer patients play an important role in determining the intensity of end-of-life (EOL) care.1Y3 A preference for lifeextending treatment is associated with more aggressive EOL care1Y4 and intensive hospital-resource consumption.5 Eliciting and honoring patients’ treatment preferences are key components of high-quality EOL care.6 However, honoring terminally ill cancer patients’ preferences for such aggressive EOL care is questionable. Indeed, life-sustaining treatments have limited value for these patients7Y9 because the treatments not only cannot improve survival10,11 but can also exact a substantial toll in terms of emotional burden, failed expectations, and financial costs for patients,10,12 family members,12 and society.13,14 To help terminally ill cancer patients avoid suffering from potentially futile life-saving treatments, the mechanisms underlying preferences for aggressive EOL care need to be identified. These mechanisms may be revealed by comprehensively investigating the factors influencing preferences for aggressive EOL care. Such a study could identify a high-risk patient group who would benefit from active interventions to help them weigh the burdens and benefits of aggressive EOL treatments. However, studies exploring factors influencing preferences for aggressive EOL care have predominantly focused on preferred goals for EOL.2,3,16Y19 Relatively less attention has been paid to cardiopulmonary resuscitation (CPR),3,19,20 mechanical ventilation support,16,20,21 and intensive care unit (ICU) care.16 Furthermore, such research seldom comprehensively incorporates patients’ predisposing, enabling, and need factors, as suggested by the behavioral model of health service utilization.22,23 Therefore, the purpose of this study was to incorporate these factors in investigating determinants of Taiwanese terminally ill cancer patients’ preferences for a wide range of treatments considered aggressive EOL care.

n

Methods

Study Design and Sample This study used data collected in a cross-sectional survey from April 2011 through November 2012 with a convenience sample of terminally ill cancer patients recruited from 23 teaching hospitals throughout Taiwan. Cancer patients were eligible to participate in the study if (1) their disease continued to progress and was judged by their oncologists as unresponsive to current curative cancer treatment, (2) they were cognitively competent to communicate with data collectors, and (3) they were 20 years or older. The human subject research review committees of the

23 study hospitals approved this study and all participants signed a written informed consent form. From 2764 eligible patients, 2467 patients were recruited (participation rate, 89.3%). Primary reasons for declining to participate were feeling too weak (n = 191, 64.3%) or being uninterested (n = 73, 24.6%). Patients who declined to participate could not be compared with participants because of restricted access to information about those who declined to participate. Of the 2467 patients participating in the survey, 2329 (94.4%) provided complete information regarding their preferences for EOL care and composed the study sample. Most of study participants were men (56.7%), younger than 65 years (65.3%), and married (78.0%) and had a junior high school education or less (55.2%). Participants’ most common cancer sites were colonrectum (16.4%), liver-pancreas (15.8%), lung (15.7%), head and neck (12.9%), and breast (11.4%). The median postdiagnosis survival at the time of data collection was 13.0 months (mean T SD, 27.1 T 40.2 months).

Outcome Measures PREFERENCES FOR AGGRESSIVE EOL CARE

Participants were asked whether they preferred (1) life-prolonging or comfort-oriented treatments or other as their goal for EOL care, (2) CPR when life was in danger, and (3) life-sustaining treatments, including ICU care, cardiac massage, intubation, and mechanical ventilation support. This interview protocol was adapted from those validated in previous studies24,25 by including more items specifically related to treatments that are common in acute care experiences and have life-prolonging, financial, and emotional consequences. For example, our protocol added preferences for use of intubation and mechanical ventilation support. After the final interview protocol was reviewed by 5 experts in palliative and EOL care, they judged it to be a valid representation of life-prolonging treatments. To elicit participants’ preferred goals for EOL care,24 we asked, ‘‘If you could choose, would you prefer (1) a course of treatment that focused on extending life, (2) a plan of care that focused on relieving pain and discomfort, (3) whatever your physician suggested, (4) maintaining current anti-cancer treatment, or (5) giving up any treatments?’’ Next, we asked about participants’ preferences for each aggressive EOL care treatment, one by one. Before obtaining preferences regarding CPR, participants were told the following: ‘‘If your heart were to stop beating and your life were in danger, your healthcare professionals might provide CPR. CPR consists of electric shocks to the heart, pumping the chest to stimulate the

E10 n Cancer NursingTM, Vol. 38, No. 3, 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Liu et al

heart, help with breathing, and heart medications given through the veins.’’25 Then, participants were asked, ‘‘If someday, when your life were in danger, would you prefer to receive CPR?’’ For life-sustaining treatments, participants were asked the following 3 questions as if they knew they were dying: (1) ‘‘If your heart stopped beating, would you want your chest to be pumped? (2) ‘‘If you were unable to breathe on your own, would you want to be intubated and on a breathing machine? In this situation, a tube would be placed through your mouth or nose into your lungs. This tube would be attached to a breathing machine. During that time, you would have to be continuously on the breathing machine and would be unable to talk and might be sedated.25’’ And (3) ‘‘If you need intensive care, would you like to stay in an ICU? An ICU is an isolated care unit that heavily uses health technology to provide intensive care. If you receive care in an ICU, you could only have contact with your family at specific visiting times.’’ The goal for EOL care was further dichotomized into prolonging life and other. For CPR and each life-sustaining treatment, participants were asked whether they (1) wanted the treatment, (2) did not want the treatment, or (3) were undecided. Responses for each treatment were dichotomized into ‘‘want treatment’’ and ‘‘do not want treatment’’ by combining those who wanted a specific treatment with those who were undecided about wanting it, as previously suggested.3 This decision was based on the clinical default in the United States and in Taiwan generally being to provide CPR or life-sustaining treatments to all patients without their explicit consent unless specifically refused.3,26 In current clinical practice in Taiwan, if patients are undecided about their preference for CPR if their life were in danger or their preferences for life-sustaining treatments if their heart stopped beating or they could not breathe on their own, physicians are required by law to provide cardiac massage, intubation, and ventilation support to rescue patients. Therefore, we combined patients who wanted a specific treatment with those who were undecided and used their combined responses as outcome variables to identify patients’ risk factors for receiving futile and painful aggressive EOL treatments.

dichotomized as junior high school or less and greater than junior high school. Self-reported financial status was divided into financial sufficiency (making ends meet) and financial strain. Participants were asked whether their physician had disclosed prognosis to them, and if so, what the prognosis was. Participants were recognized as having received an accurate prognosis from their physician only if they indicated that the physician had informed them that their disease could not be cured and they would probably die in the very near future. End-of-life care discussions were assessed by asking participants, ‘‘Has your doctor discussed with you what kind of care you would want to receive if your disease continued to progress, your condition continually declined, and you were dying?’’24 Responses were coded 1 (yes) and 0 (no). Need factors are defined as individual perceived healthcare needs and other health status indicators evaluated by healthcare professionals and influencing the likelihood of healthcare use. Need factors measured in this study included diagnosis, postdiagnosis survival, metastatic, and comorbidity status. Postdiagnosis survival was calculated as the interval between diagnosis and data collection and was categorized into 6 or fewer, 7 to 12, 13 to 24, and 25 or more months.

Independent variables

In our sample of Taiwanese terminally ill cancer patients, only 8.2% specifically preferred life-prolonging treatment as their goal for EOL care (Table 1). Comfort-oriented EOL care was preferred by the largest group (48.4%) of participants. Approximately one-third of patients would follow any treatment suggested by their primary physician. Because the current clinical default is generally to provide CPR or life-sustaining treatments to all patients without their explicit consent unless specifically refused, participants who wanted a specific treatment and those who were undecided about wanting it were combined in the category ‘‘want treatment.’’ After combining those who wanted and those who were undecided about wanting each specific treatment, 27.9% preferred CPR if their life was in danger. The proportions of participants who preferred to receive life-sustaining treatments such as ICU care, cardiac massage, intubation, and mechanical ventilation support were 36.0%, 27.3%, 24.3%, and 26.7%, respectively. Only 39.7% of participants indicated that their physician had accurately disclosed prognosis to them,

Based on the behavioral model of health services utilization,22,23 independent variables were conceptualized as patients’ predisposing, enabling, and need factors. Predisposing factors are conceptualized as demographic characteristics and social structures existing before the onset of illness and predisposing or inclining individuals to use healthcare services. Predisposing factors included gender and age. Age was divided into 4 groups (e44, 45Y64, 65Y74, and Q75 years). Enabling factors indicate available family and community resources that facilitate or impede the likelihood of healthcare use. Enabling factors measured in this study included marital status, education, financial sufficiency, and whether their physician had accurately disclosed prognosis to them and had discussed with them about their preferences for EOL care. Participants’ marital status was categorized as married (including cohabiting), single, divorced, separated, and widowed. Educational attainment was

Preferences for Aggressive End-of-life Care

Statistical Analysis Determinants of preferences for each type of aggressive EOL care were examined by multivariate logistic regression using the generalized estimating equation method.27 This method uses robust standard errors that account for correlations in the error term due to clustering of patients in the same hospital and with simultaneous adjustment for all confounding variables in the model. The effects of each independent variable on the outcome variables were measured by adjusted odds ratio (AOR) with 95% confidence interval (CI).

n

Results

Cancer NursingTM, Vol. 38, No. 3, 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

n

E11

Table 1 & Preferences for Aggressive Care at the End of Life (N = 2329)

Variable Goal of EOL care Life-prolonging treatment Comfort-oriented care As originally treated As physician suggests Giving up CPR Yes No Undecided ICU care Yes No Undecided Cardiac massage Yes No Undecided Intubation Yes No Undecided Mechanical ventilation support Yes No Undecided Accurate prognostic disclosure Yes No EOL care discussion Yes No

n

%

191 1128 110 774 125

8.2 48.4 4.8 33.2 5.4

254 1679 396

10.9 72.1 17.0

417 1490 422

17.9 64.0 18.1

287 1693 349

12.3 72.7 15.0

239 1764 326

10.3 75.7 14.0

278 1709 342

12.0 73.4 14.7

911 1382

39.7 60.3

180 2149

7.7 92.3

Abbreviations: CPR, cardiopulmonary resuscitation; EOL, end of life; ICU, intensive care unit.

and 7.7% of participants reported that their physician had discussed EOL care preferences with them. Female Taiwanese terminally ill cancer patients were significantly less likely than men to prefer prolonging life as their goal for EOL care, intubation, and mechanical ventilation support (Tables 2Y4). Preferences for prolonging life as the goal for EOL care and for ICU care were significantly higher among patients younger than 45 years than among those older than 74 years. Among the enabling factors, educational level and financial status did not significantly influence patients’ preferences for aggressive EOL care except that patients with an educational level greater than junior high school were significantly less likely to prefer CPR when their life was in danger (Tables 2Y4). Except for the goal of EOL care, the likelihood of preferences for all aggressive EOL care identified in this study was significantly lower for widowed patients than for married patients. However, single patients had 1.90 times (95% CI, 1.21Y2.96; P = .005) higher odds of preferring to prolong life as their goal for EOL care than for married patients. When physicians accurately disclosed patients’ prognosis to them, they were significantly less

likely to prefer prolonging life as their goal for EOL care, CPR when their life was in danger, ICU care, and cardiac massage (AOR [95% CI], 0.66 [0.51Y0.84] to 0.76 [0.58Y0.99]). If physicians had discussed EOL care preferences with patients, they were significantly less likely to prefer CPR when their life was in danger, cardiac massage, intubation, and mechanical ventilation support (AOR [95% CI], 0.52 [0.38Y0.71] to 0.62 [0.45Y0.87]). Among the need factors examined in this study, only postdiagnosis survival and comorbidity status influenced study participants’ preferences for aggressive EOL care (Tables 2Y4). Taiwanese terminally ill cancer patients diagnosed fewer than 6 months earlier were significantly more likely than those diagnosed more than 25 months earlier to prefer prolonging life as their goal for EOL care, having CPR when their life was in danger, and receiving all life-sustaining treatments examined in this study at EOL. Patients with chronic diseases other than their cancer were significantly more likely than those without comorbidities to prefer receiving cardiac massage, intubation, and mechanical ventilation support at EOL.

n

Discussion

When our sample of Taiwanese terminally ill cancer patients considered their goals for EOL care, only 8.2% of them preferred life-prolonging treatment, which was at the lowest end of worldwide statistics. Among advanced cancer patients in other countries, 19% in USA,15 27.6% in USA,2 28% in USA,28 29.6% in USA,16 31% in Germany,29 and 38% to 47% in USA3 preferred life-prolonging treatment as their goal for EOL care. A minority (24.3%Y36.0%) of our sample of terminally ill cancer patients preferred life-sustaining treatments. Of note, the proportions of patients preferring these life-sustaining treatments were based on those who wanted the specific treatment plus those who were undecided (Table 1). The proportions of our sample preferring life-sustaining treatments were also substantially lower than reports of 47.4% to 66% preferring CPR,3,19,30,31 52.6% to 62.9% preferring ICU care,16,31 and 40% preferring intubation with mechanical ventilation support.30 Our participants’ low prevalence of preferring to prolong life as their goal for EOL care and life-sustaining treatments at EOL may be due to several factors. First, our sample comprised only terminally ill cancer patients rather than heterogeneous samples of advanced cancer patients in previous studies.2,3,15,16,28Y31 Terminally ill patients experience considerable health deterioration over time, which has been significantly linked to declining preferences for life-sustaining treatments.32,33 Second, our sample was Taiwanese, whereas the other study samples were mainly from the United States and Europe. Attitudes toward life-sustaining treatments differ by culture. For example, Asian-Americans were reported to have generally more positive attitudes than EuropeanAmericans do toward life-sustaining treatments but were personally more reluctant to use such treatments.34 Third, the methods used to explore goals for EOL care were different between our study and others. Instead of offering participants a choice between only life-extending and comfort-oriented care as the goal for

E12 n Cancer NursingTM, Vol. 38, No. 3, 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Liu et al

Table 2 & Predictors of Preferences for End of Life (EOL) Care: Prolonging Life as a Goal and Cardiopulmonary Resuscitation (CPR) When Life is in Danger

Prolonging Life as the Goal of EOL Care Participants Parameter Predisposing factors Gender Female Male Age, y e44 45Y64 65Y74 Q75 Enabling factors Marital status Divorced/separated Widowed Single Married Educational level 9Junior high school eJunior high school Financial sufficiency Sufficient/just enough Not enough Prognostic disclosure Yes No EOL care discussion Yes No Healthcare need factors Cancer site Lung Breast Colon-rectum Liver-pancreas Blood/lymph Head and neck Stomach Other Survival postdiagnosis, mo e6 7Y12 13Y24 Q25 Metastasis Yes No With comorbidity Yes No

Adjusted Odds Ratio

95% Confidence Limits

CPR When Life is in Danger Crude Rate, %

Adjusted Odds Ratio

95% Confidence Limits

n

%

Crude Rate, %

1007 1318

43.3 56.7

8.6 12.5

0.70 Ref

0.54

0.92

.009

24.6 30.4

0.83 Ref

0.65

1.06

.138

260 1243 488 312

11.3 54.0 21.2 13.6

19.8 10.9 8.4 6.4

2.23 1.53 1.27 Ref

1.12 0.88 0.74

4.44 2.68 2.16

.023 .131 .387

29.5 25.3 30.0 30.9

1.22 0.86 1.08 Ref

0.80 0.63 0.78

1.85 1.17 1.48

.361 .341 .654

132 193 185 1806

5.7 8.3 8.0 78.0

12.1 5.7 20.8 10.3

0.88 0.79 1.90 Ref

0.48 0.39 1.21

1.62 1.59 2.96

.676 .511 .005

31.1 22.9 26.8 28.3

1.12 0.72 1.06 Ref

0.75 0.52 0.71

1.67 0.98 1.58

.590 .038 .770

1037 1275

44.9 55.2

12.9 9.1

1.25 Ref

0.92

1.69

.154

23.4 31.4

0.73 Ref

0.57

0.95

.017

1609

72.8

11.1

1.12

0.90

1.39

.294

26.6

0.90

0.68

1.17

.422

602

27.2

10.7

Ref

29.3

Ref

911 1382

39.7 60.3

9.2 11.8

0.73 Ref

0.57

0.94

.017

21.4 32.2

0.66 Ref

0.51

0.84

.001

180 2149

7.7 92.3

11.2 10.8

1.18 Ref

0.76

1.82

.470

14.5 29.0

0.52 Ref

0.38

0.71 G.001

362 262 379 364 42 297 176 424

15.7 11.4 16.4 15.8 1.8 12.9 7.6 18.4

8.1 10.3 9.5 11.6 11.9 17.6 6.3 10.6

Ref 1.23 1.03 1.18 0.95 1.49 0.65 1.20

0.70 0.57 0.65 0.40 0.85 0.33 0.77

2.19 1.85 2.12 2.26 2.62 1.28 1.87

.472 .929 .591 .908 .167 .211 .412

27.6 23.0 27.0 31.5 21.4 31.9 26.9 27.2

Ref 0.86 0.84 0.98 0.54 0.97 0.78 0.81

0.58 0.55 0.77 0.27 0.64 0.50 0.56

1.27 1.27 1.26 1.09 1.46 1.23 1.18

.445 .405 .902 .084 .873 .282 .274

729 385 407 751

32.1 17.0 17.9 33.1

13.2 11.5 8.7 9.5

1.39 1.26 1.00 Ref

1.04 0.88 0.71

1.84 1.80 1.41

.024 .203 .997

33.2 27.4 29.4 23.1

1.40 1.08 1.23 Ref

1.12 0.85 0.95

1.76 1.39 1.59

.003 .528 .114

1829 464

79.8 20.2

10.6 11.7

1.08 Ref

0.64

1.83

.776

27.5 29.7

1.04 Ref

0.81

1.34

.769

1341 957

58.4 41.6

10.3 11.5

1.08 Ref

0.78

1.49

.635

28.9 26.4

1.09 Ref

0.92

1.29

.318

P

P

Responses for each treatment were dichotomized to ‘‘want treatment’’ (1) and ‘‘do not want treatment’’ (0) by combining those who wanted and those who were undecided as wanting that specific treatment. Abbreviation: Ref, reference.

Preferences for Aggressive End-of-life Care

Cancer NursingTM, Vol. 38, No. 3, 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

n

E13

Table 3 & Predictors of Preferences for End of Life (EOL) Care: Intensive Care Unit (ICU) Care and Cardiac Massage ICU care Participants Parameter Predisposing factors Gender Female Male Age, y e44 45Y64 65Y74 Q75 Enabling factors Marital status Divorced/separated Widowed Single Married Educational level 9Junior high school eJunior high school Financial sufficiency Sufficient/just enough Not enough Prognostic disclosure Yes No EOL care discussion Yes No Healthcare need factors Cancer site Lung Breast Colon-rectum Liver-pancreas Blood/lymph Head and neck Stomach Other Survival postdiagnosis, mo e6 7Y12 13Y24 Q25 Metastasis Yes No With comorbidity Yes No

Cardiac Massage

95% Crude Adjusted Confidence Limits Rate, % Odds Ratio

P

95% Crude Adjusted Confidence Limits Rate, % Odds Ratio

P

n

%

1007 1318

43.3 56.7

33.1 38.1

0.85 Ref

0.69 1.04

.105

24.0 29.8

0.82 Ref

0.64 1.06

.132

260 1243 488 312

11.3 54.0 21.2 13.6

41.5 33.6 36.4 38.3

1.51 0.95 1.05 Ref

1.04 2.20 0.67 1.36 0.79 1.41

.031 .791 .726

31.0 24.8 27.4 31.5

1.19 0.80 0.95 Ref

0.73 1.94 0.54 1.19 0.65 1.40

.494 .276 .808

193 132 185 1806

8.3 5.7 8.0 78.0

35.6 29.2 36.1 36.9

0.89 0.68 0.97 Ref

0.57 1.40 0.49 0.93 0.71 1.31

.621 .015 .821

26.5 21.4 30.1 27.8

0.89 0.62 1.24 Ref

0.54 1.48 0.44 0.88 0.95 1.62

.660 .007 .117

1037 1275

44.9 55.2

33.7 37.9

0.85 Ref

0.66 1.08

.174

24.2 29.7

0.83 Ref

0.63 1.09

.185

1609 602

72.8 27.2

36.3 33.8

1.08 Ref

0.85 1.38

.529

26.2 28.1

0.92 Ref

0.70 1.21

.531

911 1382

39.7 60.3

29.2 40.3

0.67 Ref

0.55 0.81 G.001

22.4 30.5

0.76 Ref

0.58 0.99

.043

180 2149

7.7 92.3

25.7 36.9

0.75 Ref

0.56 1.01

.057

16.2 28.2

0.61 Ref

0.41 0.92

.018

362 262 379 364 42 297 176 424

15.7 11.4 16.4 15.8 1.8 12.9 7.6 18.4

33.4 33.7 35.7 40.9 31.0 37.3 34.9 35.7

Ref 1.00 0.93 1.11 0.78 0.97 0.85 0.86

0.71 0.67 0.81 0.42 0.64 0.56 0.59

1.41 1.29 1.53 1.44 1.48 1.27 1.25

.996 .648 .504 .426 .889 .418 .422

24.5 24.1 25.7 30.1 16.7 31.2 27.4 29.1

Ref 1.04 0.94 1.01 0.45 1.10 0.91 0.99

0.68 0.62 0.77 0.18 0.84 0.57 0.69

1.57 1.43 1.32 1.17 1.45 1.44 1.42

.871 .774 .966 .103 .489 .679 .961

729 385 407 751

32.1 17.0 17.9 33.1

40.3 36.3 36.8 32.1

1.24 1.06 1.11 Ref

1.04 1.48 0.79 1.43 0.86 1.42

.019 .701 .431

31.9 26.4 27.6 23.7

1.30 1.03 1.11 Ref

1.06 1.59 0.76 1.40 0.87 1.42

.011 .840 .387

1829 464

79.8 20.2

36.0 36.4

1.08 Ref

0.85 1.37

.548

26.9 29.0

1.03 Ref

0.79 1.35

.807

1341 957

58.4 41.6

37.4 34.2

1.12 Ref

0.92 1.36

.255

28.6 25.5

1.20 Ref

1.04 1.38

.011

Responses for each treatment were dichotomized to ‘‘want treatment’’ (1) and ‘‘do not want treatment’’ (0) by combining those who wanted and those who were undecided as wanting that specific treatment. Abbreviation: Ref, reference.

EOL care, our study provided more alternatives than the previous studies did, for example, treatments suggested by their physician, maintaining current anticancer treatment, and forgoing

any treatments. Indeed, one-third of our study participants preferred to follow their physician’s suggestion. This finding is noteworthy because variations in EOL care strongly depend on

E14 n Cancer NursingTM, Vol. 38, No. 3, 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Liu et al

Table 4 & Predictors of Preferences for End of Life (EOL) Care: Intubation and Mechanical Ventilation Support Intubation Participants Parameter Predisposing factors Gender Female Male Age, y e44 45Y64 65Y74 Q75 Enabling factors Marital status Divorced/separated Widowed Single Married Educational level 9Junior high school eJunior high school Financial sufficiency Sufficient/just enough Not enough Prognostic disclosure Yes No EOL care discussion Yes No Healthcare need factors Cancer site Lung Breast Colon-rectum Liver-pancreas Blood/lymph Head and neck Stomach Other Survival postdiagnosis, mo e6 7Y12 13Y24 Q25 Metastasis Yes No With comorbidity Yes No

Mechanical Ventilation Support

95% Adjusted Confidence Limits Odds Ratio

Crude Rate, %

95% Adjusted Confidence Limits Odds Ratio

n

%

Crude Rate, %

1007 1318

43.3 56.7

19.3 27.9

0.70 Ref

0.54 0.91 .007

22.9 29.4

0.76 Ref

0.61 0.95 .017

260 1243 488 312

11.3 54.0 21.2 13.6

29.1 21.6 24.7 27.0

1.39 0.89 1.13 Ref

0.90 2.13 .135 0.63 1.24 .481 0.76 1.68 .543

31.4 24.1 27.2 28.9

1.45 0.94 1.11 Ref

0.98 2.13 .067 0.67 1.31 .704 0.76 1.61 .592

193 132 185 1806

8.3 5.7 8.0 78.0

25.8 18.8 27.9 24.3

0.99 0.65 1.32 Ref

0.60 1.62 .965 0.46 0.92 .016 0.94 1.85 .107

27.3 21.4 29.0 26.8

0.95 0.60 1.22 Ref

0.59 1.53 .834 0.41 0.86 .006 0.87 1.72 .246

1037 1275

44.9 55.2

21.8 26.1

0.85 Ref

0.67 1.08 .175

23.8 28.7

0.82 Ref

0.67 1.02 .071

1609 602

72.8 27.2

22.5 26.3

0.87 Ref

0.65 1.17 .363

25.0 27.8

0.93 Ref

0.70 1.24 .632

911 1382

39.7 60.3

19.4 27.3

0.76 Ref

0.54 1.07 .112

22.3 29.4

0.79 Ref

0.60 1.03 .086

180 2149

7.7 92.3

14.0 25.0

0.62 Ref

0.45 0.87 .006

16.2 27.4

0.59 Ref

0.39 0.90 .015

362 262 379 364 42 297 176 424

15.7 11.4 16.4 15.8 1.8 12.9 7.6 18.4

20.6 18.4 22.8 25.7 16.7 31.9 25.1 25.8

Ref 1.10 1.03 1.01 0.55 1.47 1.04 1.10

24.2 23.0 25.1 29.0 19.1 31.9 27.4 26.5

Ref 1.08 0.90 0.95 0.55 1.19 0.89 0.90

729 385 407 751

32.1 17.0 17.9 33.1

29.4 24.5 23.9 19.6

1.44 1.19 1.17 Ref

1.14 1.83 .003 0.88 1.61 .251 0.95 1.45 .135

31.6 26.4 27.4 21.7

1.41 1.14 1.26 Ref

1.15 1.73 .001 0.90 1.44 .285 1.05 1.49 .011

1829 464

79.8 20.2

23.5 26.8

0.98 Ref

0.75 1.29 .907

26.0 28.6

1.01 Ref

0.79 1.29 .943

1341 957

58.4 41.6

25.7 22.0

1.29 Ref

1.09 1.53 .004

27.9 24.6

1.23 Ref

1.01 1.49 .038

0.74 0.65 0.74 0.20 1.00 0.63 0.77

1.64 1.64 1.39 1.50 2.16 1.71 1.56

P

.630 .899 .935 .242 .053 .880 .608

0.73 0.56 0.68 0.21 0.76 0.50 0.60

1.60 1.44 1.31 1.42 1.86 1.58 1.34

P

.696 .667 .734 .215 .457 .682 .598

Responses for each treatment were dichotomized to ‘‘want treatment’’ (1) and ‘‘do not want treatment’’ (0) by combining those who wanted and those who were undecided as wanting that specific treatment. Abbreviation: Ref, reference.

physician practices29,35 even though physicians often misunderstand patients’ preferences for EOL care29,36,37 and seldom discuss EOL care issues with terminally ill cancer pa-

Preferences for Aggressive End-of-life Care

tients.38,39 To avoid patient suffering from ineffective lifesustaining treatments and to counteract the trend of increasing treatment intensity at EOL in Taiwan40 and worldwide,41,42 Cancer NursingTM, Vol. 38, No. 3, 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

n

E15

we suggest interventions to help terminally ill cancer patients accurately understand the burdens and benefits of aggressive EOL care and to facilitate patient-physician consensus on an appropriate goal for EOL care. We found that female Taiwanese terminally ill cancer patients had a lower likelihood than men did of preferring to prolong life as the goal for EOL care, as reported.2,29 We extend this line of evidence to show that women were also more likely than their male counterparts to prefer to forgo intubation and mechanical ventilation support at EOL. Our results also concur with the literature that greater education level of terminally ill cancer patients reduced their likelihood of preferring CPR when their life was in danger.21,32,34 Similarly, we found that younger age was associated with stronger preferences for life-prolonging treatment as the goal of EOL care, as reported.2,18 We also found that patients younger than 45 years were significantly more likely to prefer ICU care. However, we did not confirm the finding43 that younger patients had a higher probability of preferring CPR than older patients did. Married patients have been reported to be similarly,2 more,33 or 17 less likely than unmarried patients to prefer life-prolonging treatment as their goal of EOL care or to prefer more lifesustaining treatments. These conflicting findings may result from not differentiating unmarried status. In our study, widowed patients were not only significantly less likely to prefer CPR when their life was in danger, as reported,43 but also less likely to prefer all life-sustaining treatments examined in this study. These findings may be due to their older age, which brings more experiences of death and dying among relatives and friends. However, our single patients were significantly more likely than married patients to prefer life-prolonging treatment as their goal of EOL care. To disentangle the impact of marital status on preferences for aggressive EOL care, the nonspecific ‘‘unmarried’’ status should be further categorized into single, widowed, and separated/divorced because these ‘‘unmarried’’ groups may not have the same life experiences and circumstances. A novel finding from our study is that terminally ill cancer patients whose physicians had accurately told them their prognosis were less likely to prefer aggressive EOL care than were patients not told their prognosis. Although this relationship has not been previously reported, our result is consistent with a report that physicians have been documented as the main source of patients’ prognostic knowledge44 and cancer patients who recognized they were terminally ill preferred symptom-directed EOL care over life-extending therapy.10,45 Furthermore, patients whose physician informed them of the life-limiting nature of their illness were more likely to have had EOL care discussions with their physician.28 When patients discuss EOL care preferences with their physician, they can accurately estimate the burdens and benefits of life-sustaining treatments at EOL.21 In contemporary acute care settings with an emphasis on life-saving activities, such EOL care discussions result in patients’ valuing comfort-oriented care over life-extending therapy at EOL and opposing dying in an ICU and doing everything possible to extend life for a few days.12 We extend this line of evidence by finding that terminally ill Taiwanese cancer patients who had discussed EOL care issues with their physician were significantly

less likely than those who had not to prefer receiving cardiac massage, intubation, and mechanical ventilation support at EOL. Taken together, results from these studies support our finding that terminally ill cancer patients whose physician had accurately disclosed prognostic information to them and had discussed EOL care issues with them were less likely to prefer aggressive EOL care. Our study revealed a never explored association between postdiagnosis survival and preferences for aggressive EOL care. The treatment goals for cancer patients diagnosed fewer than 6 months earlier tend to be curative, which may hinder their coming to terms with their fatal illness. This inability to accept their terminal status likely increased their likelihood of preferring life-prolonging treatment as their goal of EOL care, CPR when their life was in danger, and life-sustaining treatments even when their disease continued to progress. In contrast, patients who had lived longer with their cancer were more likely to prefer comfort-oriented care rather than aggressive life-sustaining treatment. These findings suggest that patients recently diagnosed with a terminal disease are at risk for preferring aggressive EOL care. They would benefit from interventions to facilitate adjustment to their new disease and forthcoming death, thereby to help them weigh the burdens and benefits of aggressive EOL treatments. Our study also documented a seldom explored relationship between medical comorbidities and preferences for aggressive EOL care, such as cardiac massage, intubation, and mechanical ventilation support,17 in contrast to a report of no association with treatment preferences.2 Terminally ill cancer patients who have another disease may have learned from their previous illness experiences to prefer more aggressive treatments to relieve their symptom distress. This preference might have evolved from the prospect of dying if they did not receive aggressive care appearing greater than the prospect of poor health from aggressive care, predisposing them to prefer aggressive EOL care.17 For cancer patients without other disease, these 2 prospects would seem similarly distant from their experience.17 Our study identified a high-risk group of terminally ill cancer patients in a large sample recruited throughout Taiwan who preferred aggressive EOL care even though it might not improve either their quantity or quality of life.7Y12 However, our sample was recruited by convenience sampling because no complete list of all prospective terminally ill cancer patients was available. Therefore, our sample might not represent the targeted population, although our study participants had distributions of gender, age, and disease categories similar to those of Taiwanese cancer decedents in 2011,46 except patients with colorectal cancer were overrepresented in our sample. Our findings may also not be generalizable to terminally ill cancer patients residing in other geographic areas or receiving care at nonteaching hospitals. The cross-sectional design of our study might not provide information on dynamic changes in patients’ preferences for EOL care as their death approaches. Despite incorporating a wide range of influencing factors into our analysis, we can never exclude the possibility of unmeasured residuals commonly found in an observational study. Such unmeasured residuals include but are not limited to patients’ experiences and beliefs about aggressive EOL care,21 symptom distress,15,32 psychological state (ie, depressive mood,17 sense of

E16 n Cancer NursingTM, Vol. 38, No. 3, 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Liu et al

burden to others),18,47 and quality of life3; family members’ inability to come to terms with the patient’s death and involvement in EOL care decision making; and physicians’ attitudes toward aggressive and palliative EOL care and clinical practices. Furthermore, despite providing patients information about CPR and life-sustaining treatments when asking their preferences for CPR when their life was in danger and aggressive EOL care, we did not evaluate their understanding and emotional acceptance of those treatments. In conclusion, only a minority of Taiwanese terminally ill cancer patients preferred prolonging life as their goal for EOL care, CPR when their life was in danger, and receiving lifesustaining treatments at EOL. Among our sample of terminally ill cancer patients, those at high risk for preferring aggressive EOL care were men, younger than 45 years, married or single, with a junior high school education or less, diagnosed fewer than 6 months earlier, and with chronic disease other than their cancer and their physician had neither accurately disclosed their prognosis to them nor discussed EOL care issues with them. These high-risk patients should receive interventions to help them weigh the burdens and benefits of aggressive EOL treatments. Research is urgently needed to evaluate the effectiveness of such interventions in clarifying terminally ill cancer patients’ misconceptions and unrealistic expectations of aggressive EOL care and in facilitating EOL care decision making that eventually leads to informed patients receiving EOL care congruent with their preferences. In this way, terminally ill cancer patients can avoid unnecessary suffering from life-sustaining treatments and achieve a good death as recommended by the Institute of Medicine.6 Society may also benefit by avoiding the escalating EOL care expenditures resulting from expensive life-sustaining treatments in patients’ final days of life.5,13,48

References 1. Silveira MJ, Kim SY, Kenneth ML. Advance directives and outcomes of surrogate decision making before death. N Engl J Med. 2010;362: 1211Y1218. 2. Wright AA, Mack JW, Kritek PA, et al. Influence of patients’ preferences and treatment site on cancer patients’ end-of-life care. Cancer. 2010;116: 4656Y4663. 3. Rose JH, O’Toole EE, Dawson NV, et al. Perspectives, preferences, care practices, and outcomes among older and middle-aged patients with latestage cancer. J Clin Oncol. 2004;22:4907Y4917. 4. Cisgriff JA, Pisani M, Bradley EH, et al. The association between treatment preferences and trajectories of care at the end-of-life. J Gen Intern Med. 2007;22:1566Y1571. 5. Teno JM, Fisher ES, Hamel MB, et al. Medical care inconsistent with patients’ treatment goals: association with 1-year Medicare resource use and survival. J Am Geriatr Soc. 2002;50:496Y500. 6. Institute of Medicine. Field MJ, Cassell C, eds. Approaching Death: Improving Care at the End of Life. Washington, DC: National Academy Press; 1997:4. 7. Reisfield GM, Wallace SK, Munsell MF, et al. Survival in cancer patients undergoing in-hospital cardiopulmonary resuscitation: a meta-analysis. Resuscitation. 2006;71:152Y160. 8. Soares M, Depuydt PO, Salluh JIF. Mechanical ventilation in cancer patients: clinical characteristics and outcomes. Crit Care Clin. 2010;26:41Y58. 9. Taccone F, Artigas AA, Sprung CL, et al. Characteristics and outcomes of cancer patients in European ICUs. Crit Care. 2009;13:R15. 10. Mack JW, Weeks JC, Wright AA, et al. End-of-life discussions, goal attainment, and distress at the end of life: predictors and outcomes of receipt of care consistent with preferences. J Clin Oncol. 2010;28:1203Y1208.

Preferences for Aggressive End-of-life Care

11. Barnato AE, Chang CC, Farrell MH, et al. Is survival better at hospitals with higher ‘‘end-of-life’’ treatment intensity? Med Care. 2010;48:125Y132. 12. Wright AA, Zhang B, Ray A, et al. Associations between end-of-life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA. 2008;300:1665Y1673. 13. Nicholas LH, Langa KM, Iwashyna TJ, et al. Regional variation in the association between advance directives and end-of-life Medicare expenditures. JAMA. 2011;306:1447Y1453. 14. Niederman MS, Berger JT. The delivery of futile care is harmful to other patients. Crit Care Med. 2010;38(suppl):518Y522. 15. Bambauer KZ, Gillick MR. The effect of underlying health status on patient or surrogate preferences for end-of-life care: a pilot study. Am J Hosp Palliat Med. 2007;24:185Y190. 16. Parr JD, Zhang B, Nilsson ME, et al. The influence of age on the likelihood of receiving end-of-life care consistent with patient treatment preferences. J Palliat Med. 2010;13:719Y726. 17. Winter L, Parker B. Current health and preferences for life-prolonging treatments: an application of prospect theory to end-of-life decision making. Soc Sci Med. 2007;65:1695Y1707. 18. Winter L, Parks SM. The reluctance to burden others as a value in endof-life decision making: a source of inaccuracy in substituted judgment. J Health Psychol. 2012;17:179Y188. 19. Ackroyd R, Russon L, Newell R. Views of oncology patients, their relatives and oncologists on cardiopulmonary resuscitation (CPR): questionnairebased study. Palliat Med. 2007;21:139Y144. 20. Maida V, Peck J, Ennis M, et al. Preferences for active and aggressive intervention among patients with advanced cancer. BMC Cancer. 2010;10:592. 21. Barnato AE, Anthony DL, Skinner J, et al. Racial and ethnic differences in preferences for end-of-life treatment. J Gen Intern Med. 2009;24:695Y701. 22. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;361:1Y10. 23. Phillips KA, Morrison KR, Andersen R, et al. Understanding the context of healthcare utilization: assessing environmental and provider-related variables in the behavioral model of utilization. Health Serv Res. 1998;33:571Y596. 24. Zhang B, Wright AA, Huskamp HA, et al. Health care costs in the last week of life: associations with end-of-life conversations. Arch Intern Med. 2009;169:480Y488. 25. Stapleton RD, Nielsen EL, Engelberg RA, Patrick DL, Curtis JR. Association of depression and life-sustaining treatment preferences in patients with COPD. Chest. 2005;127:328Y334. 26. Jensen JB, Brown DR. In-hospital resuscitation: history, incidence, outcomes, and opportunities. Contemp Crit Care. 2008;5:1Y10. 27. Twisk JWR. Longitudinal data analysis: a comparison between generalized estimating equations and random coefficient analysis. Eur J Epidemiol. 2004; 19:769Y776. 28. Wagner GJ, Riopelle D, Steckart J, et al. Provider communication and patient understanding of life-limiting illness and their relationship to patient communication of treatment preferences. J Pain Symptom Manage. 2010;39:527Y534. 29. Winkler EC, Reiter-Theil S, Lange-Riess D, et al. Patient involvement in decisions to limit treatment: the crucial role of agreement between physician and patient. J Clin Oncol. 2009;27:2225Y2230. 30. Nelson JE, Gay EB, Berman AR, et al. Patients rate physician communication about lung cancer. Cancer. 2011;117:5212Y5220. 31. Yun YH, You CH, Lee JS, et al. Understanding disparities in aggressive care preferences between patients with terminal illness and their family members. J Pain Symptom Manage. 2006;31:513Y521. 32. Ditto PH, Smucker WD, Danks JH, et al. Stability of older adults’ preferences for life-sustaining medical treatment. Health Psychol. 2003;22: 605Y615. 33. Fried TR, Van Ness PH, Byers AL, et al. Changes in preferences for lifesustaining treatment among older persons with advanced illness. J Gen Intern Med. 2007;22:495Y501. 34. Blackhall LJ, Frank G, Murphy ST, et al. Ethnicity and attitudes towards life sustaining technology. Soc Sci Med. 1999;48:1779Y1789. 35. Barnato AE, Bost JE, Farrell MH, et al. Relationship between staff perceptions of hospital norms and hospital-level end-of-life treatment intensity. J Palliat Med. 2007;10:1093Y1100.

Cancer NursingTM, Vol. 38, No. 3, 2015

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

n

E17

36. Deep KS, Griffith CH, Wilson JF. Discussing preferences for cardiopulmonary resuscitation: what do resident physicians and their hospitalized patients think was decided? Patient Educ Couns. 2008;72:20Y25. 37. DesHarnais S, Carter RE, Hennessy W, et al. Lack of concordance between physician and patient: reports on end-of-life care discussions. J Palliat Med. 2007;10:728Y740. 38. Keating NL, Landrum MB, Rogers SO Jr, et al. Physician factors associated with discussions about end-of-life care. Cancer. 2010;116: 998Y1006. 39. Michiels E, Deschepper R, Bilsen J, et al. Information disclosure to terminally ill patients and their relatives: self-reported practice of Belgian clinical specialists and general practitioners. Palliat Med. 2009;23: 345Y353. 40. Tang ST, Wu SC, Hung YN, et al. Trends in quality of end-of-life care for Taiwanese cancer patients who died in 2000Y2006. Ann Oncol. 2009; 20:343Y348. 41. Earle CC, Neville BA, Landrum MB, et al. Trends in the aggressiveness of cancer care near the end of life. J Clin Oncol. 2004;22:315Y321. 42. Ho TH, Barbera L, Saskin R, et al. Trends in the aggressiveness of end-

43.

44.

45.

46.

47.

48.

of-life cancer care in the universal health care system of Ontario, Canada. J Clin Oncol. 2011;29:1587Y1591. Laakkonen ML, Pitkala KH, Strandberg TE. Older people’s reasoning for resuscitation preferences and their role in the decision-making process. Resuscitation. 2005;65:165Y171. Gattelari M, Voigt KJ, Butow PN, et al. When the treatment goal is not cure: are cancer patients equipped to make informed decisions? J Clin Oncol. 2002;20:503Y513. Weeks JC, Cook EF, O’Day SJ, et al. Relationship between cancer patients’ predictions of prognosis and their treatment preferences. JAMA. 1998;279:1709Y1714. Department of Health. Statistics of causes of death, volume I. 2011. http://www.doh.gov.tw/ufile/doc/2011%20statistics%20of%20causes% 20of%20death.pdf. Accessed on February 8, 2013. McPherson CJ, Wilson KG, Murray MA. Feeling like a burden to others: a systematic review focusing on the end of life. Palliat Med. 2007;21: 115Y128. Riley GF, Lubitz JD. Long-term trends in Medicare payments in the last year of life. Health Serv Res. 2010;45:565Y576.

E18 n Cancer NursingTM, Vol. 38, No. 3, 2015 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Liu et al

Preferences for Aggressive End-of-life Care and Their Determinants Among Taiwanese Terminally Ill Cancer Patients.

Studies on factors influencing preferences for aggressive end-of-life (EOL) care have focused predominantly on preferred goals of EOL and seldom compr...
152KB Sizes 0 Downloads 4 Views