Determinants of Patient-Family Caregiver Congruence on Preferred Place of Death in Taiwan

Article Outline

• Abstract
• Introduction
• Methods
  • Study Design and Sample
  • Measurements
    • Outcome Variable
    • Independent Variables
  • Data Collection Procedures
  • Statistical Analysis
• Results
  • Sample Description
  • Patient-Family Caregiver Congruence on the Preferred Place of Death
  • Patient-Family Caregiver Concordance on End-of-Life Care Options
  • Determinants of Patient-Family Caregiver Congruence on Preferred Place of Death
• Discussion
• References
• Copyright

Abstract 

Context

Patient-family caregiver congruence on preferred place of death not only increases the likelihood of dying at home but also contributes significantly to terminally ill cancer patients' quality of life.

Objectives

To examine the determinants of patient-family caregiver congruence on the preferred place of death in Taiwan.

Methods

Patient-family caregiver dyads (n=1,108) were surveyed on preferences and needs for end-of-life (EOL) care. Determinants of congruence on preferences were identified by multivariate logistic regression.

Results

Patient-caregiver dyads achieved 78.1% agreement on the preferred place of death. The kappa coefficient of congruence was 0.55 (95% confidence interval [CI]=0.50, 0.60). The extent of patient-family caregiver congruence on preferred place of death increased with the patient's higher functional dependence (adjusted odds ratio [AOR] and 95% CI=1.04 [1.02, 1.05]), higher patient-rated importance for dying at preferred place of death (AOR [95% CI]=1.60 [1.43, 1.79]), and having a spousal caregiver (AOR [95% CI]=1.62 [1.14, 2.31]). Other determinants of patient-family caregiver congruence included patient age (AOR [95% CI]=1.01 [1.00, 1.03]), patient-family concordance on preferred EOL care options (AOR=1.68–1.73), patient knowledge of prognosis (AOR [95% CI]=0.68 [0.48, 0.97]), and impact of caregiving on the family caregiver's life (AOR [95% CI]=0.98 [0.96, 0.99]).

Conclusion

Increasing patient-family congruence on preferred place of death not only requires knowledge of the patient's prognosis and advance planning by both parties but also depends on family caregivers endorsing patient preferences for EOL care options and ensuring that supporting patients dying at home does not create an intolerable burden for family caregivers.

Key Words: Place of death, end-of-life care, terminally ill cancer patients, family caregivers, preferences for end-of-life care

Back to Article Outline

Introduction 

Terminally ill cancer patients worldwide have a strong preference for dying at home.¹, ², ³, ⁴, ⁵, ⁶ Correspondingly, promoting dying at home has increasingly been regarded as a “gold standard” of end-of-life (EOL) care.⁷, ⁸, ⁹ However, few cancer patients realize their preferences about the place of death.², ⁷, ⁹, ¹⁰, ¹¹ Such a discrepancy between preferred and actual place of death for terminally ill cancer patients highlights the dilemma inherent in making and achieving patient preferences regarding place of death.

The decision about preferred place of death is mediated by a range of complex interpersonal, social, and pragmatic considerations.¹², ¹³, ¹⁴ One of the most consistently reported and influential determinants of home death is social support.¹⁰ Realizing patients' preference for dying at home has been shown to be influenced by family caregivers' clear recognition and endorsement of patients' preferences.¹⁰, ¹⁵, ¹⁶, ¹⁷

Patient-family caregiver congruence on the preferred place of death not only increases the likelihood of dying at a place the patient prefers (either home or hospice)¹⁵, ¹⁷ but also contributes significantly to terminally ill cancer patients' quality of life.¹⁸ Nonetheless, empirical evidence indicates that patients do not always agree with their family caregivers regarding preferences for place of death.⁵, ⁶, ¹⁹ Moreover, no currently available studies have investigated factors that influence patient-family caregiver congruence on the preferred place of death. Therefore, the present study was undertaken to examine the extent of patient-family caregiver congruence on the preferred place of death and factors influencing congruence among a cohort of terminally ill cancer patients and their family caregivers in Taiwan.

Back to Article Outline

Methods 

Study Design and Sample 

A cross-sectional design was used to survey terminally ill cancer patients and their family caregivers from February 2003 through November 2004 about preferences and needs for EOL care. The patients were not under hospice care and were being treated at 24 hospitals throughout Taiwan, which provided care for most of the cancer patients (54.9%) who died during 2000–2006 in Taiwan.²⁰ After the human subject research review committees of the 24 hospitals approved the study, subjects were recruited by convenience sampling. Patients were eligible for the study if they 1) had a malignancy judged by their primary physician to be at a terminal stage, continuing to progress, and unresponsive to curative cancer treatments; 2) were cognitively competent; 3) could communicate coherently with data collectors; and 4) had a designated family caregiver who agreed to participate in the survey.

Primary physicians at each study site were given a detailed explanation of the study and were asked to identify patients without judging their emotional readiness to talk about their preferences for EOL care. After verifying the eligibility of patients, data collectors invited them to participate in this study without seeking their family's permission. This strategy avoided excluding patients because of conflict of opinion between them and the family caregiver about the patient's participation.

Each patient-participant identified one primary family caregiver as the person most involved in the patient's care without receiving financial reimbursement for the care provided. Identified family caregivers were recruited if they were 18 years or older, agreed to participate, and could communicate with data collectors. Patient-caregiver dyads were excluded if either the patient or the caregiver refused to participate. Written informed consent was obtained from all patients and family caregivers.

Measurements 

Patient-family caregiver congruence on the preferred place of death was measured as previously reported.⁶ In brief, terminally ill cancer patients and their family caregivers were independently asked one question regarding where they preferred (or for their family member) to die (at home, hospital, inpatient hospice, or other).

The rationale for selecting independent variables was guided by factors influencing the place of death for terminally ill cancer patients, whether preferred², ¹³, ¹⁴, ¹⁹ or actual.¹⁰, ²¹, ²², ²³

1.Patient-rated importance of dying at the preferred place of death was rated on a 5-point Likert scale (1=not at all important; 5=very important).

Data Collection Procedures 

Terminally ill cancer patients and their family caregivers independently completed the questionnaires in separate rooms and were specifically instructed not to discuss their preferences for the patient's EOL options until both had completed the interviews.

Statistical Analysis 

The extent of patient-caregiver congruence on the issues measured in this study was determined not only by the percentage of agreement but also by kappa coefficients to correct for the amount of agreement that can be expected by chance alone.³⁰ The strength of agreement measured by kappa was determined by the following criteria: kappa≤0.20, poor; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; 0.81–1.00, almost perfect.³¹ Predictors of patient-caregiver congruence on the preferred place of death were identified by multivariate logistic regression with backward selection. All independent variables identified in the current study were included in the initial model. The effects of individual explanatory variables on the outcome variable were measured by adjusted odds ratio (AOR) with 95% confidence interval (CI).

Back to Article Outline

Results 

Sample Description 

From 1,709 eligible terminally ill cancer patients, 1,108 patient-caregiver dyads were recruited (participation rate=64.8%). The primary reasons for declining to participate were fragile physical condition (n=496; 82.5%) and being busy with caregiving (n=105; 17.5%) by patients and caregivers, respectively. Patients who participated and those who did not differed significantly only in age and metastatic status. Participants were younger than the nonparticipants by a mean age of 3.5 years (P<0.0001) and were significantly less likely (80.1%) to have a disease with distant metastasis than nonparticipants (86.6%). Family caregivers who participated in this study did not differ from nonparticipants in terms of age and relationship with patients. However, more male family members than female family members participated (36.3% vs. 28.0%, P<0.0001).

The patients were predominantly (59.5%) males (Table 1), and their median age was 61 years (range: 18–91 years). Most of the patients were married (83.7%) and lived with their spouses (77.6%) and one or more of their children (76.8%). The most common sites of cancer were lung (24.9%), liver and pancreas (15.8%), colon-rectum (11.3%), head and neck (9.1%), and breast (6.6%). The median time since diagnosis was 10.0 months (range: 0.03–144 months). Half the terminally ill cancer patients (50.0%) were aware of their prognosis, and slightly more than one-third (36.2%) had been informed of the prognosis. Terminally ill cancer patients suffered mild to moderate symptom distress (mean: 27.41±8.00) and depended moderately on others to perform daily activities or social functions (mean: 28.05±11.44).

Table 1. Characteristics of the Patient Sample

Characteristic	n	%
Gender
Male	659	59.5
Female	449	40.5
Educational level
Less than high school	754	68.1
High school or higher	354	31.9
Marital status
Married	926	83.7
Single	64	5.8
Separated	9	0.8
Divorced	15	1.4
Widowed	92	8.3
Age (years)
Range	18–91
Mean (SD)	59.75 (14.04)
Median	61.0
Postdiagnosis survival (months)
Range	0.03–144
Mean (standard deviation)	19.34 (28.35)
Median	10.0
Living status
Alone	27	2.4
With spouse	860	77.6
With adult child	851	76.8
Cancer group
Head and neck	101	9.1
Breast	73	6.6
Lung	276	24.9
Gastric	71	6.4
Colon-rectum	125	11.3
Liver/pancreas	175	15.8
Hematological	74	6.7
Other	213	19.2
Metastasis
Yes	887	80.1
No	220	19.9
Comorbidities
Yes	481	43.6
No	622	56.4

The family caregivers were predominantly (63.7%) females (Table 2), with about half (50.4%) being the patient's spouse and another 34.3% being the patient's child. The caregivers' median age was 48 years (range: 18–85 years). Most family caregivers were married (83.0%) and lived with the patient (81.1%). They provided considerable assistance to the patients in every aspect of care measured in this study (mean: 10.76±3.35) and reported a moderate level of caregiving burden (mean: 2.55±0.48).

Table 2. Characteristics of the Family Caregiver Sample

Characteristic	n	%
Gender
Male	400	36.3
Female	702	63.7
Marital status
Married	918	83.0
Single	161	14.6
Divorced/separated/widowed	27	2.4
Living with the patient
Yes	891	81.1
No	217	19.6
Age (years)
Range	18–85
Mean (standard deviation)	48.0 (13.9)
Median	48.0
Relationship with patient
Spouse	558	50.4
Adult child	380	34.3
Parent	44	4.0
Sibling	28	2.5
Friend or relative	31	2.8
Other	77	7.0

Patient-Family Caregiver Congruence on the Preferred Place of Death 

Terminally ill cancer patients and their family caregivers preferred for the patient to die at home (65.8% and 64.0%, respectively) (Table 3). Both sets of respondents identified the second preferred place of death as the hospital. Agreement on the preferred place of death was 78.1%. The kappa coefficient of overall congruence on preferred place of death was 0.55 (95% CI=0.50–0.60), indicating moderate congruence between patients and caregivers on preferences for the place of death. Terminally ill cancer patients and their primary family caregivers achieved a higher congruence on the preferences for home or hospital as the place of death than for the preferences to die at an inpatient hospice unit, other place, or no preference, as indicated by the kappa values for each place.

Table 3. Congruence on End-of-Life Care Options Between Terminally Ill Taiwanese Cancer Patients and Their Family Caregivers

Variable	Patients	Families	Agreement	Kappa (95% CI)
	%	%	%
Preference for place of death (n=1,023)			78.1	0.55 (0.50, 0.60)
Home	65.8	64.0	80.7	0.58 (0.52, 0.63)
Hospital	18.5	20.7	87.4	0.60 (0.54, 0.66)
Inpatient hospice	1.9	2.7	97.4	0.41 (0.23, 0.59)
No preference	9.0	8.4	91.0	0.43 (0.33, 0.53)
Other	4.9	4.1	92.8	−0.01 (−0.06, 0.04)
Preference for cardiac massage (n=1,051)			65.8	0.34 (0.29, 0.39)
Yes	14.6	20.9	80.9	0.35 (0.28, 0.42)
No	66.9	62.8	70.6	0.36 (0.30, 0.42)
Undecided	18.6	16.3	81.9	0.31 (0.23, 0.38)
Preference for mechanical ventilation (n=1,053)			65.8	0.33 (0.28, 0.38)
Yes	14.7	20.7	82.6	0.34 (0.27, 0.41)
No	67.4	63.7	70.5	0.34 (0.28, 0.40)
Undecided	17.9	15.6	80.6	0.30 (0.23, 0.38)
Preference for ICU care (n=1,054)			61.7	0.35 (0.30, 0.40)
Yes	21.4	29.4	94.8	0.37 (0.31, 0.43)
No	58.5	53.4	70.6	0.41 (0.35, 0.46)
Undecided	20.1	17.2	77	0.24 (0.17, 0.31)
Preference for vasopressors (n=1,052)			59.6	0.35 (0.30, 0.40)
Yes	26.7	36.9	73.3	0.39 (0.33, 0.45)
No	53.0	46.8	68.0	0.36 (0.31, 0.42)
Undecided	20.3	16.4	78.0	0.27 (0.20, 0.34)
Preference for hospice care (n=966)			71.7	0.42 (0.37, 0.48)
Yes	52.8	68.6	71.7	0.42 (0.37, 0.48)
No	47.2	31.4	71.7	0.42 (0.37, 0.48)

CI=confidence interval.

Patient-Family Caregiver Concordance on End-of-Life Care Options 

When terminally ill cancer patients and their family caregivers were asked whether they preferred for themselves or for their ill family member to receive life-sustaining treatments if the patient's life was in danger, the concordance between the two parties ranged from 59.6% to 65.8% (Table 3). Only one kappa value (preference for no ICU care: kappa=0.41) for overall (range: 0.33–0.35) and no individual-alternative (range: 0.24-0.39) concordance on each life-sustaining treatment exceeded the threshold for moderate concordance proposed by Landis and Koch.³¹ When there was a discrepancy, family caregivers, without exception, adopted a significantly more aggressive attitude toward each life-sustaining treatment for their ill family member than the patient's stated preferences. However, if the patient's death became unavoidable, family caregivers were also more likely than patients to opt for hospice care (68.6% vs. 52.8%). The two parties showed a moderate agreement on this issue (kappa [95% CI]=0.42 [0.37–0.48]).

Determinants of Patient-Family Caregiver Congruence on Preferred Place of Death 

Variables retained in the final best-fitting multivariate logistic regression model are presented in Table 4. Terminally ill cancer patients' preferred places of death were more likely to be congruent with those identified by their family caregivers if patients were older, had a higher degree of functional dependency, and assigned greater importance to dying at the preferred place of death. For each unit increase in age, in the ESDS score, and in the importance score, the AOR with 95% CI for concordance were 1.01 [1.00, 1.03], P=0.04; 1.04 [1.02, 1.05], P<0.0001; and 1.60 [1.43, 1.79], P<0.0001, respectively. Patient-family caregiver congruence on the preferred place of death was higher (AOR [95% CI]=1.62 [1.14, 2.31], P=0.007) when caregivers were spouses than when they had a different relationship with the patient. However, if terminally ill cancer patients were aware of their prognosis, they were less likely to agree with their family caregivers on the preferred place of death (AOR [95% CI]=0.68 [0.48, 0.97], P=0.03).

Table 4. Best-Fitting Model for Determinants of Congruence on the Preferred Place of Death Between Terminally Ill Cancer Patients and Their Family Caregivers

AOR=adjusted odds ratio; CI=confidence interval.

Variables in the initial model included the following: 1) patient-rated importance of dying at the preferred place of death; 2) demographics and disease-related characteristics of patients; 3) characteristics of family caregivers; 4) patient awareness and experiences of being informed of the prognosis; 5) patient-family caregiver agreement on preferences for end-of-life care options; and 6) caregiving burden, including objective caregiving load and subjective caregiving burden. Details about variables in each category can be found in the Measurements section for independent variables.

Terminally ill cancer patients and their family caregivers were more likely to agree on the preferred place of death if they concurred on the preference for cardiac massage when the patient's life was in danger or receiving hospice care when the patient's death became unavoidable (AOR [95% CI]=1.68 [1.17, 2.41], P=0.005 and AOR [95% CI]=1.73 [1.21, 2.50], P=0.003, respectively). Finally, family caregivers were less likely to agree with their ill family member on the preferred place of death if they rated caregiving as creating a greater negative impact on their life (with each unit increase in the total CRA score, AOR decreased by 0.98 [0.96, 0.99], P=0.004).

Back to Article Outline

Discussion 

This study found that approximately two-thirds of cancer patients and their family caregivers preferred the patient to die at home, confirming previous reports.⁴ However, the extent of congruence between terminally ill cancer patients and their family caregivers on preferred place of death (78.1% agreement, kappa=0.55) was substantially higher than that previously reported.⁵, ³² In earlier studies, 49.3% (kappa=0.16)⁵ and 36% (kappa=0.15)³² of patient-family caregiver dyads agreed on the preferred place of death, respectively. Our higher level of patient-family caregiver congruence on preferred place of death may be explained by two cultural influences. First, dying at home has the special cultural meaning to terminally ill Taiwanese patients and their families of “the fallen leaves can return to their roots.”⁶ Second, Taiwanese people are influenced by the Confucian idea of filial piety and traditional caregiving ideology to see caregiving as repaying their parents or family members. Therefore, Taiwanese family caregivers generally feel obligated to uphold the patient's wish to die at home despite a substantial toll on their quality of life³³ and a higher risk of depressive symptoms.³⁴

The results of this study indicated that, despite the high percentages of patient-family concordance on life-sustaining treatments (59.6%–65.8% and 68.0%–94.8% for overall and individual-alternative, respectively), almost none of the kappa values exceeded the threshold for moderate concordance. Furthermore, because of the low prevalence of preference for receiving life-sustaining treatments (Table 3), the primary source of concordance came from both parties' preferences for not receiving each life-sustaining treatment. When there was a discrepancy, without exception, the family caregiver was more likely to institute a treatment when the patient did not want it than to withhold the treatment when the patient did want it.

These results convey an important message. Disagreements between patients and families about using life-sustaining treatments may result in patients' preferences being overridden at EOL and in terminally ill cancer patients receiving more treatment than they would choose for themselves. Therefore, caution should be applied in holding to the family-centered model rather than exploring the patient's wishes. This caution is particularly worth mentioning in Asian societies, such as Taiwan, where the physician's respect for patient autonomy is frequently subordinated to the family's power.³⁵, ³⁶

This study identified important determinants of patient-family caregiver congruence on preferred place of death. These determinants are consistent with previous findings on predictors of preferred or actual place of death for terminally ill cancer patients, including increased functional dependence,¹⁰, ²¹, ²³ importance to the patient of dying at the preferred place of death,¹⁰, ³⁷ and spousal caregiver.¹⁰, ¹⁹, ²¹, ²³ The present study identified other determinants of patient-family caregiver congruence on preferred place of death, that is, patient age, patient knowledge of prognosis, patient-family concordance on preferred EOL care options, and the impact of caregiving on the family caregiver's life. Because these determinants are embedded within social and cultural contexts, they deserve further discussion.

In our study, the older the terminally ill Taiwanese cancer patient, the greater the extent of patient-family caregiver congruence on preferred place of death, in contrast to inconsistent findings for the influence of age on preferred or actual place of death.², ¹⁰, ¹⁹, ²³ Age represents a biological and a developmental life-trajectory phenomenon with implications for the nature of preferred health care, including preferred place of death.²² Death in older age is more likely to be accepted as “nature taking its course.” Therefore, in contrast to an expectation for “everything” to be done for young adults nearing EOL, avoiding aggressive life-sustaining treatments is a common practice for terminally ill elderly cancer patients in Taiwan. Consequently, patients and their family caregivers are more likely to concur on withholding aggressive treatments and allowing death to take place at home. Culturally, Taiwanese people highly value shou zhong zheng qin (dying naturally of old age at home in one's bed). Such a death is the most glorious and fortunate way of dying.³⁸, ³⁹ Taiwanese elders also are recognized as having unique power and authority in the family; hence, fulfilling their wish of shou zhong zheng qin is one way for Taiwanese people to show the important virtue of filial piety.⁴⁰

The preferred places of death for terminally ill Taiwanese cancer patients who were aware of their own prognosis were less likely to be congruent with their caregivers' choices for them. Awareness and acceptance of death have been associated with “home” as preferred or actual place of death.¹⁰, ³⁷ However, adequate knowledge of prognosis has been suggested as a prerequisite for appropriate treatment planning at EOL without overestimating the efficacy of aggressive but may-be-futile treatments.², ¹², ⁴¹ According to this proposition, terminally ill Taiwanese cancer patients who are aware of their prognosis might more realistically appraise the demands of dying at home and be more concerned about being a burden to family members; consequently, they might shift their preference for place of death to outside the home.

However, we previously showed that knowledge of prognosis was only fairly correlated between terminally ill cancer patients and their family caregivers.²⁶ Therefore, patients' and caregivers' preferences for place of death might not be in accord. This hypothesis was confirmed by further analyses, which indicated that terminally ill cancer patients who were aware of their prognosis were less likely than those who were unaware to prefer to die at home (63.2% vs. 68.6%, respectively), and the “aware” patients were more likely than “unaware” patients to shift their preferred place of death toward a hospital or an inpatient hospice (22.8% vs. 17.6%, respectively). Correspondingly, “aware” patients were less likely to agree with their family caregivers on preferring “home” as the place of death than “unaware” patients (data not shown).

Patient-family caregiver congruence on preferred place of death increased when terminally ill Taiwanese cancer patients agreed with their family caregivers on preferences for cardiac massage if the patient's life was in danger or for hospice care if the patient's death became unavoidable. For most persons, treatment preferences are grounded in a consistent belief and value system.⁴² When both patients and their caregivers opted to withhold cardiac massage or to receive hospice care, they might have valued comfort care more than life-preserving treatments. In this situation, family caregivers were more likely to endorse the patient's preference for dying at home (data not shown). On the other hand, when both parties agreed on cardiac massage if the patient's life was in danger, it was inevitable that the patient should be hospitalized. Therefore, both patients and their caregivers were more likely to choose “hospital” as the preferred place of death (data not shown).

Caregivers in our study who endured a greater negative impact of caregiving were less likely than those with less negative caregiving impact to agree with patients on preferred place of death. This finding echoes mounting evidence that a major barrier to terminally ill cancer patients' preferring to die at home is their concern about burdening family caregivers.², ³, ⁴, ⁶, ¹², ³⁷ This concern is supported by reports that caregiving at home creates significant challenges and serious consequences for family caregivers.⁴³, ⁴⁴, ⁴⁵ The decision to provide EOL care at home is most commonly derived from a promise to honor the patient's wish to die at home.¹² However, when dying at home introduces an extraordinary burden for family caregivers, both terminally ill cancer patients and their caregivers may reevaluate this wish or promise in light of important contextual factors and their own care-receiving or caregiving experiences.³⁷, ⁴⁶ Hospital or inpatient hospice care may be preferred when family caregivers experience overwhelming caregiving demands⁴⁶ or when dying patients try to save relatives the burden of caregiving at home.², ⁴ However, evaluations of care-receiving or caregiving experiences between the two parties are not guaranteed to be in concert. Any mismatches in evaluating care-receiving or caregiving experiences may increase the probability of discrepancy in patients' and caregivers' preferred place of death.

Two strengths of this study are its large sample recruited from multiple hospitals throughout Taiwan and being the first quantitative investigation of factors influencing congruence on preferred place of death between terminally ill cancer patients and their caregivers. However, representation of the targeted population might have been compromised by convenience sampling. The study findings might not be generalizable to terminally ill cancer patients declining to participate, those receiving care at other than the 24 study sites, and those residing in other areas.

The findings might also be limited by not having examined the quality of the patient-family caregiver relationship, the extent of communication about the patient's prognosis and EOL care preferences, and the roles of Taiwanese cultural norms for caregiving (such as the extent of practicing filial piety and of commitment to the family). The cross-sectional design may have limited the ability to capture the dynamic decision-making process regarding preferred place of death that may change because of worsening physical symptoms, diminished family resources, and other variables as the patient's death approaches.³², ⁴⁷, ⁴⁸, ⁴⁹ Furthermore, place of care and place of death are not interchangeable.³² Our study only investigated the extent and determinants of patient-family caregiver congruence on preferred place of death. The extent and determinants of terminally ill patient-family caregiver concordance on preferred place of EOL care warrant further investigation.

Despite these limitations, the study results are informative. Our results underscore a discrepancy between Taiwanese patients and their family caregivers in preferred place of death despite the highly valued cultural meaning of dying at home. Enabling cancer patients to achieve congruence with their family caregivers on preferred place of death not only requires adequate knowledge of the patient's prognosis and advance planning by both parties but also depends on caregivers recognizing and endorsing patient preferences for EOL care options and ensuring that supporting patients dying at home does not create an intolerable burden for caregivers.

Open communication about prognosis and EOL care options not only promotes EOL care decisions that are consistent with the values and wishes of terminally ill cancer patients but will also facilitate family caregivers' understanding of the patient's actual rather than inferred preferences for EOL care options.⁵⁰ Only with adequate knowledge of their ill relative's preferences for place of death and care options can caregivers act on behalf of patients and achieve their best interests. Our results also suggest that EOL care at home be promoted by national health policy and commitment by health care professionals to increase support for family caregivers. This support is vital for the good of both patients and society, because it not only offsets the effects of caregiving stress to allow families to fulfill the patient's wish to die at home but also realizes the potential of cost saving from home deaths of terminally ill cancer patients.⁸

Although empirically testing the predictive model using cross-sectional data is instructive, this study did not definitively test the temporal order of some important variables. Therefore, interpretation of the study findings is limited by causal ambiguity. Future longitudinal research is needed to investigate 1) changes in patient-family caregiver congruence on preferred place of death, 2) factors influencing such changes, and 3) the impact of such changes on caregivers' caregiving burden and patients' actual place of death. Insights from such research will guide the development of interventions to overcome barriers for achieving patient-family caregiver congruence on preferred place of death, to empower families to care for terminally ill cancer patients at home, and to reduce the subjective caregiving burden of family caregivers. Such efforts might enhance the confidence of terminally ill cancer patients that their preferred place of death will be honored without imposing an intolerable burden on their loved ones and might lift the feeling of burdening others. Ultimately, such efforts might facilitate death at a place that is culturally valued and preferred by both patients and their families.

Back to Article Outline

References 

1. Beccaro M, Costantini M, Rossi PG, et al. Actual and preferred place of death of cancer patients. Results from the Italian survey of the dying of cancer (ISDOC). J Epidemiol Community Health. 2006;60:412–416
   • View In Article
2. Foreman LM, Hunt RW, Luke CG, Roder DM. Factors predictive of preferred place of death in the general population of South Australia. Palliat Med. 2006;20:447–453
   • View In Article
   • MEDLINE
   • CrossRef
3. Gott M, Seymour J, Bellamy G, Clark D, Ahmedzai S. Older people's view about home as a place of care at the end of life. Palliat Med. 2004;18:460–467
   • View In Article
   • MEDLINE
   • CrossRef
4. Higginson IJ, Sen-Gupta GJA. Place of care in advanced cancer: a qualitative systematic literature review of patient preferences. J Palliat Med. 2000;3:287–300
   • View In Article
   • MEDLINE
   • CrossRef
5. Stajduhar KI, Allan DE, Cohen SR, Heyland DK. Preferences for location of death of seriously ill hospitalized patients: perspectives from Canadian patients and their family caregivers. Palliat Med. 2008;22:85–88
   • View In Article
   • CrossRef
6. Tang ST, Liu TW, Lai MS, McCorkle R. Discrepancy in the preferences of place of death between terminally ill cancer patients and their primary family caregivers in Taiwan. Soc Sci Med. 2005;61:1560–1566
   • View In Article
   • MEDLINE
   • CrossRef
7. Flory J, Young-Xu Y, Gurol I, et al. Place of death: U.S. trends since 1980. Health Affair. 2004;23:194–200
   • View In Article
8. Mpinga EK, Pennec S, Gomes B, et al. First international symposium on places of death: an agenda for the 21st century. J Palliat Care. 2006;22:293–296
   • View In Article
   • MEDLINE
9. Department of Health . End of life care strategy. London, UK: Department of Health; 2008;
   • View In Article
10. Gomes B, Higginson IJ. Factors influencing death at home in terminally ill patients with cancer: systematic review. Br Med J. 2006;332:515–521
    • View In Article
11. Neutel CI, Bishop ML, Harper SD, Gaudette LA. Proportion of cancer deaths occurring in hospital, Canada, 1994-2000. Can J Public Health. 2005;96:264–268
    • View In Article
    • MEDLINE
12. Exley C, Allen D. A critical examination of home care: end of life care as an illustrative case. Soc Sci Med. 2007;65:2317–2327
    • View In Article
    • CrossRef
13. Seymour J, Payne S, Chapman A, Holloway M. Hospice or home? Expectations of end-of-life care among white and Chinese older people in the UK. Sociol Health Illn. 2007;29:872–890
    • View In Article
    • CrossRef
14. Thomas C, Morris SM, Clark MD. Place of death: preferences among cancer patients and their carers. Soc Sci Med. 2004;58:2431–2444
    • View In Article
    • MEDLINE
    • CrossRef
15. Cantwell P, Turco S, Brenneis C, et al. Predictors of home death in palliative care cancer patients. J Palliat Care. 2000;16:23–28
    • View In Article
16. Fukui S, Fukui N, Kawagoe H. Predictors of place of death for Japanese patients with advanced-stage malignant disease in home care settings: a nationwide survey. Cancer. 2004;101:421–429
    • View In Article
    • MEDLINE
    • CrossRef
17. Grande G, Ewing G. Death at home unlikely if informal carers prefer otherwise: implications for policy. Palliat Med. 2008;22:971–972
    • View In Article
    • CrossRef
18. Tang ST, Liu TW, Tsai CM, et al. Patient awareness of prognosis, patient-family caregiver congruence on the preferred place of death, and caregiving burden of families contribute to the quality of life for terminally ill cancer patients in Taiwan. Psychooncology. 2008;17:1202–1209
    • View In Article
    • CrossRef
19. Choi KS, Chae YM, Lee CG, et al. Factors influencing preferences for place of terminal care and of death among cancer patients and their families in Korea. Support Care Cancer. 2005;13:565–572
    • View In Article
    • MEDLINE
    • CrossRef
20. Department of Health . Statistics of causes of death, volume I. 2007. Available from http://www.doh.gov.tw/EN2006/DM/DM2_p01.aspx?class_no=390&now_fod_list_no=9256&level_no=2&doc_no=51991Accessed December 10, 2009
    • View In Article
21. Grundy E, Mayer D, Young H, Sloggett A. Living arrangements and place of death of older people with cancer in England and Wales: a record linkage study. Br J Cancer. 2004;91:907–912
    • View In Article
    • MEDLINE
22. Gruneir A, Mor V, Weitzen S, et al. Where people die: a multilevel approach to understanding influences on site of death in America. Med Care Res Rev. 2007;64:351–378
    • View In Article
    • CrossRef
23. Wachterman MW, Sommers BD. The impact of gender and marital status on end-of-life care: evidence from the national mortality follow-back survey. J Palliat Med. 2006;9:343–352
    • View In Article
    • MEDLINE
    • CrossRef
24. McCorkle R, Young K. Development of a symptom distress scale. Cancer Nurs. 1978;1:373–378
    • View In Article
    • MEDLINE
25. Benoliel JQ, McCorkle R, Young K. Development of a social dependency scale. Res Nurs Health. 1980;3:3–10
    • View In Article
    • MEDLINE
26. Tang ST, Liu TW, Lai MS, et al. Congruence of knowledge, experiences and preferences for disclosure of diagnosis and prognosis between terminally-ill cancer patients and their family caregivers in Taiwan. Cancer Invest. 2006;24:360–366
    • View In Article
    • MEDLINE
    • CrossRef
27. Tang ST, Liu TW, Lai MS, Liu LN, Chen CH. Concordance of preferences for end-of-life care between terminally-ill cancer patients and their family caregivers in Taiwan. J Pain Symptom Manage. 2005;30:510–518
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (106 KB)
    • CrossRef
28. Emanuel EJ, Fairclough DL, Slutsman J, et al. Assistance from family members, friends, paid care givers and volunteers in the care of terminally ill patients. New Engl J Med. 1999;341:956–963
    • View In Article
29. Given CW, Given B, Stommel M, et al. The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical and mental impairments. Res Nurs Health. 1992;15:271–283
    • View In Article
    • MEDLINE
    • CrossRef
30. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;8476:307–310
    • View In Article
31. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159–174
    • View In Article
    • MEDLINE
    • CrossRef
32. Agar M, Currow DC, Shelby-James TM, et al. Preference for place of care and place of death in palliative care: are these different questions?. Palliat Med. 2008;22:787–795
    • View In Article
    • CrossRef
33. Tang ST, Li CY, Chen CCH. Trajectory and determinants of the quality of life of family caregivers of terminally ill cancer patients in Taiwan. Qual Life Res. 2008;17:387–395
    • View In Article
    • CrossRef
34. Tang ST, Li CY, Liao YC. Factors associated with depressive distress among Taiwanese family caregivers of cancer patients at the end of life. Palliat Med. 2007;21:249–257
    • View In Article
    • CrossRef
35. Back MF, Huak CY. Family centred decision making and non-disclosure of diagnosis in a South East Asian oncology practice. Psychooncology. 2005;14:1052–1059
    • View In Article
    • MEDLINE
    • CrossRef
36. Blackhall LJ, Frank G, Murphy S, et al. Bioethics in a different tongue: the case of truth-telling. J Urban Health. 2001;78:59–71
    • View In Article
    • MEDLINE
    • CrossRef
37. Thomas C. The place of death of cancer patients: can qualitative data add to known factors?. Soc Sci Med. 2004;60:2597–2607
    • View In Article
    • MEDLINE
    • CrossRef
38. Thompson LG. Chinese religion: An introduction. 3rd ed.. Belmont, CA: Wadsworth Pub. Co.; 1979;34–56
    • View In Article
39. Wu YT. Chinese folklore. [in Chinese] Taipei, Taiwan: Chung Wen Publishing Company; 1987;145–162
    • View In Article
40. Dai YT, Dimond MF. Filial piety: a cross-cultural comparison and its implications for the well-being of older parents. J Gerontol Nurs. 1998;24:13–18
    • View In Article
    • MEDLINE
41. Weeks JC, Cook EF, O'Day SJ, et al. Relationship between cancer patients' predictions of prognosis and their treatment preferences. JAMA. 1998;279:1709–1714
    • View In Article
    • MEDLINE
    • CrossRef
42. Patrick DL, Pearlman RA, Starks HE, et al. Validation of preferences for life-sustaining treatment: implications for advance care planning. Ann Intern Med. 1998;127:509–517
    • View In Article
    • MEDLINE
43. Christakis NA, Allison PD. Mortality after the hospitalization of a spouse. N Engl J Med. 2006;354:719–730
    • View In Article
    • CrossRef
44. Emanuel EJ, Fairclough DL, Slutsman J, Emanuel LL. Understanding economic and other burdens of terminal illness: the experience of patients and their caregivers. Ann Intern Med. 2000;132:451–459
    • View In Article
    • MEDLINE
45. Pitceathly C, Maguire P. The psychological impact of cancer on patients' partners and other key relatives: a review. Eur J Cancer. 2006;39:1517–1524
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (162 KB)
    • CrossRef
46. Evans WG, Cutson TM, Steinhauser KE, Tulsky JA. Is there no place like home? Caregivers recall reasons for and experience upon transfer from home hospice to inpatient facilities. J Palliat Med. 2006;9:100–110
    • View In Article
    • MEDLINE
    • CrossRef
47. Fried TR, O'Leary J, Van Ness P, Fraenkel L. Inconsistency over time in the preferences of older persons with advanced illness for life-sustaining treatment. J Am Geriatr Soc. 2007;55:1007–1014
    • View In Article
    • CrossRef
48. Hinton J. Can home care maintain an acceptable quality of life for patients with terminal cancer and their relatives?. Palliat Med. 1994;8:183–196
    • View In Article
    • MEDLINE
    • CrossRef
49. Klinkenberg M, Visser G, van Groenou MIB, et al. The last 3 months of life: care, transitions and the place of death of older people. Health Soc Care Community. 2005;13:420–430
    • View In Article
    • MEDLINE
    • CrossRef
50. Engelberg RA, Patrick DL, Curtis JR. Correspondence between patients' preferences and surrogates' understandings for dying and death. J Pain Symptom Manage. 2005;30:498–509
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (141 KB)
    • CrossRef