| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Center on Aging, School of Medicine, University of Hawaii, Honolulu, HI
b School of Medicine, University of Hawaii, Honolulu, HI
c College of Education, University of Hawaii, Honolulu, HI
Correspondence: Kathryn L. Braun, DrPH, Center on Aging, John A. Burns School of Medicine, University of Hawaii, 1960 East-West Road, Biomed C-106, Honolulu, HI 96822. E-mail: kbraun{at}hawaii.edu.
Vernon L. Greene, PhD
 |
Abstract |
|---|
 TOP Abstract Literature ReviewMethodsResultsDiscussionReferences |
|---|
Key Words: Asian Americans • Beliefs • Culture • Death • Euthanasia • Fatalism • Religiosity
Research has suggested that support for physician-assisted suicide (PAS) varies by ethnicity and that these variations may be associated with the cultural values and sociohistorical experiences of an ethnic group, which, in turn, influence attitudes toward planning for death (Braun, Pietsch, and Blanchette 2000
). The purpose of this study was to identify these attitudinal constructs and test their usefulness in explaining documented differences in support for PAS among ethnic groups in Honolulu.
|
Literature Review |
|---|
|
TOPAbstractLiterature ReviewMethodsResultsDiscussionReferences |
|---|
; Benson 1999; Cohen, Fihn, Boyko, Jonsen, and Wood 1994; Lee and associates 1996; Meier, Emmons, Wallenstein, Quill, Morrison, and Cassel 1998).
In a review of data from national polls, Benson 1999 noted that 45% to 59% of Americans now support PAS. Several states have taken the issue to their polling places and courts; Oregon legalized PAS in 1997 (Chin, Hedberg, Higginson, and Flemming 1999). Despite this trend, it is important to note that PAS does not carry the same appeal in all ethnic groups. For example, Singh 1979 found that 40% of African Americans felt that a physician should be allowed to help a terminally ill patient die if he or she so requested compared with 65% of Caucasians. In the same survey, only 21% of African Americans (compared with 42% of Caucasians) felt a person with an incurable illness had a right to end his or her own life. Similarly, in a survey of Michigan residents, researchers found that 76% of Caucasian respondents supported the legalization of PAS compared with only 56% of African American respondents (Lichtenstein, Alcser, Corning, Bachman, and Doukas 1997). Ethnic variation may, in some cases, reflect the religious traditions of an ethnic group. For example, the majority of Hispanic and Filipino Americans are Catholic. Catholic doctrine affirms the dignity and sanctity of human life and the sovereignty of God over it, and thus prohibits suicide (John Paul II 1995; National Conference of Catholic Bishops 1994; Vatican 1980).
Still, one's specific ethnicity or religion is not a sure predictor of support, or nonsupport, for PAS. For example, a survey by the Tarrance Group 1994 found that only 73% of Catholics who attend mass weekly opposed PAS because of religious beliefs, and Singh 1979 found that strong religious commitment and frequent religious service attendance, regardless of religion, were associated with low support of PAS. In fact, once Lichtenstein and associates 1997 included a measure of religiosity in their model, "race, age, and sex no longer had a significant effect on support for legalization of physician-assisted suicide" (pp. 129–130).
Attitudes Toward Planning for Death
Distinctions among values, attitudes, and behaviors have been articulated by Rokeach 1975. Individual behavior is influenced by attitudes, which are described as "relatively enduring organization[s] of interrelated beliefs that describe, evaluate, and advocate action with respect to an object or situation" (p.132). Attitudes provide individuals with a frame of reference; attitudes may be selective, biased, and resistant to change, but they can be modified in the face of new evidence. Values and beliefs, more basic than attitudes because they are abstract ideals unrelated to a specific object or situation, often underlie attitudes. An individual's value system is ordered out of cultural, religious, institutional, and social influences (Rokeach 1975). Thus, attitudes are based on values and underlie behaviors.
The current study attempted to articulate and measure attitudes toward planning for death. This exploration is important, as empirical data and everyday experience verify that individuals are unique and that within-ethnic-group differences in end-of-life attitudes and practices may vary as much as, or more than, between-ethnic-group differences. To learn more about ethnic differences in end-of-life planning and the underlying constructs that may affect end-of-life choices, several studies were conducted in Honolulu on a multiethnic sample. Early work (conducted in 1995 through 1996) was qualitative, using focus groups and key informant interviews. This was followed in 1997 by structured interviews with 245 adults; partial findings from this study are reported in this article.
The 1995–1996 qualitative study involved respondents from five ethnic groups—Chinese, Filipino, Japanese, Hawaiian, and Vietnamese—and found differences in attitudes toward planning for death among, and within, these groups. Differences reflected respondents' varying cultural and religious traditions, degree of religiosity, educational attainment, experience with artificial prolongation of life of family members, trust in family and the health care system, and number of generations the individuals' families had lived in the United States (Braun and Nichols 1997).
In response to questions about euthanasia (participants were not asked specifically about PAS), this investigation revealed that all of the respondents' religions had prohibitions against killing. But this was interpreted differently by participants depending on their own experience with death and dying. For example, many Filipino participants who were first-generation immigrants and primarily Catholic said that euthanasia and suicide would go against God's commandments and show a distrust of God's plan for one's life. Yet those Filipino respondents who worked in health care, including a first-generation Filipino Catholic nun, expressed an acceptance of withholding futile treatment and of giving adequate medication to relieve pain even if this act might fatally suppress respiration. Japanese participants, most of whom were born in Hawaii, told stories of very long-lived relatives who had required years of care and had died with "too many tubes." Although they noted that Buddhism was against killing, they also acknowledged two concepts from the Japanese language that would support the withholding or withdrawing of futile treatment: shikata ga nai, which means "nothing can be done" or "it cannot be helped," and akirameru, meaning "to leave things as they are" or "to let things proceed naturally." Many Native Hawaiian participants (those who trace their ancestry back before first contact with Caucasians) expressed distrust of the government and the health care system, voicing concern that legalizing euthanasia may result in poorer care for economically disadvantaged members of this ethnic group. Chinese respondents referred to a traditional view of suicide as something that would bring "bad luck" on the family. But one related a story about a relative who had been on life support for several years and how this experience helped the family distinguish between refusing, withholding, or withdrawing technology and suicide; all family members had since completed advance directives. Across ethnic groups were concerns about being in pain, being dependent, being a burden, and being impoverished at life's end, and these individuals seemed supportive of euthanasia as a way to gain control over the dying process. Other participants noted a strong cultural norm that planning or talking about death could make it happen or would suggest disregard for God's plan or one's karma. Others were strong advocates of discussing and documenting wishes regarding the end of life. Others were not concerned about their own wishes but trusted that family and doctor would make the right decisions when the time came. Still others said they did not know how they felt about planning for death or aid in dying but that they would want to seek guidance from their church or minister on these issues (Braun and Nichols 1997).
On the basis of the findings from this qualitative study, we hypothesized that one's end-of-life behaviors would be influenced by several sets of variables. Ethnic membership, certainly, is associated with cultural and religious values that comprise the value system that underlies individuals' attitudes toward end-of-life planning. One's lived experience—immigrant status, education, exposure to life-prolonging technology—may cause one to modify one's attitudes. We hypothesized six potential attitude factors: Advocacy to Discuss/Document Wishes, Wanting to Control One's Death, Trust in Others to Make the Right Decisions, Reliance on Religious Guidance at the End of Life, Fears and Anxieties About Life's End, and Fatalism About Death's Timing. We hypothesized that these factors, rather than ethnic or religious membership per se, would account for variance in support for PAS and that this finding could help minimize ethnic and religious stereotyping in favor of a more sophisticated modeling of how culturally and religiously influenced attitudes affect end-of-life decision making.
|
Methods |
|---|
|
TOPAbstractLiterature ReviewMethodsResultsDiscussionReferences |
|---|
Although interviews were conducted in English, student interviewers were of the same ethnic background as the group they interviewed except for the student assigned to interview the Hawaiian group, who was a young Caucasian born and raised in a Hawaiian community. Interviewers first identified senior participants through senior centers, churches, and social organizations in Honolulu that tended to attract members of a specific ethnic group. Participating seniors then identified an adult child willing to be interviewed; children were of the same ethnicity as their parents. This sampling strategy was pursued to allow testing of generational differences in end-of-life practices, as previous work suggested that adult children differ in beliefs and practices from their parents (Braun and Nichols 1997; Kalish and Reynolds 1976). Also, seniors tend to congregate in groups that are ethnically homogeneous, whereas middle-aged Hawaii residents are more likely to be living and working in integrated settings. Although Hawaii is an ethnically integrated state where 60% of newborns are of mixed race (Hawaii Department of Health 1998), care was taken to select participants for this study who were not of mixed race. The exception was for Native Hawaiians, because less than half a percent of this ethnic group is full Hawaiian. In this case, participants were recruited from senior centers and service programs that serve Native Hawaiian homestead communities.
Caucasian, Japanese, and Filipino participants were easily identified and interviewed. Hawaiian and Chinese participants were harder to recruit; the two students interviewing these groups estimated that they asked 4 adults for each 1 who agreed. The Chinese-group interviewer reported that the high refusal among Chinese participants was due to discomfort with the topic of death, expressing concerns that talking about death would bring bad luck. Thus, Chinese participants likely represent the more acculturated members of this ethnic group. The student interviewing Native Hawaiians reported high levels of distrust, which took time to overcome. Even though he had been born and raised in Hawaii, he needed to spend time "talking story" with potential interviewees before consent would be provided. Once given, however, Hawaiian respondents appeared to share their feelings openly. In nine cases, a direct parent–child pair could not be interviewed, sometimes because the adult child did not have time to participate or lived out of state; instead, another close relative (e.g., a niece, nephew, or adult grandchild) of the older adult was interviewed. Interviews were completed within 5 months for a final sample of 245 adults (125 seniors and 120 adult children) in the five ethnic groups (48 Caucasians, 48 Chinese, 50 Filipinos, 50 Native Hawaiians, and 49 Japanese). The limitation of the sample for generalizing results is acknowledged. But given the exploratory nature of the study—that is, to identify and measure attitudes toward planning for death—we felt it to be adequate. Of the 245 study participants, 202 said either "yes" or "no" (rather than "perhaps") to the question about support for PAS, and analysis related to PAS was limited to this group.
Measures
The full four-part interview schedule included 85 questions. Respondent characteristics included age, gender, birthplace, educational attainment, ethnicity, religious affiliation, and experience with life-threatening illness and family caregiving. A series of questions asked about advance directive knowledge and completion and reasons for completing one or not. The third section included 25 items designed to capture attitudes toward end-of-life planning, each scored on a 5-point Likert scale; these were developed directly from findings from the 1995–1996 qualitative study. The final section focused on PAS. Our definition (a doctor providing a patient with the means to end his or her own life, such as a drug that can be lethal in certain doses) contained all the elements from definitions used by others—prescription of medication (many definitions specify "lethal drug" or "lethal dose"), by a physician, to be used by the patient, with the intent to end his or her own life. Respondents were asked, "Is there any condition under which this should be allowed?" If the participant answered "no," questioning was concluded. If the participant answered "yes" or "perhaps," another 29 questions were asked about possible conditions (e.g., should the requester be over 18? have a terminal illness? be in pain? need a second opinion? need agreement from family? Braun 1998). At the conclusion of the interview, the participant was thanked and offered a $10 gift certificate.
Analysis
Many social phenomena are conceived of as processes operating among underlying constructs. Factor analysis is a statistical procedure commonly used to test one's conceptualization of these constructs or theoretical abstractions (Heck 1998), and in this study we used it to obtain more clarity about attitudes toward planning for death. We hypothesized that underlying factors would likely represent constructs such as advocacy for discussing/documenting wishes, wanting to control one's death, trust in others, reliance on religious guidance, fears and anxieties about life's end, and fatalism about death's timing; 25 items were included to measure these constructs. Once data collection was completed, we performed a factor analysis with SPSS (SPSS, Inc 1999) using principal-axis factoring with oblimin rotation. Several criteria were used to judge the model's fit: (a) the percentage of variance explained was reasonably high (>50%), (b) factors had eigenvalues greater than 1.0, (c) factors consisted of four or more items, and (d) factors were conceptually sound with low or low–moderate levels of intercorrelation. Once the best solution was obtained, the reliability of each factor was determined by using Cronbach's alpha (a measure of internal consistency) and factor scores were produced (with a mean of 0 and a standard deviation of 1).
A correlation matrix developed with SPSS was used to examine relationships among ethnicity, other characteristics, attitudes, and PAS support. Ethnicities were Caucasian, Chinese, Filipino, Hawaiian, and Japanese, and other characteristics were generation (senior vs adult child), religion (Catholic or not), immigrant status (yes or no), educational attainment (measured on an ordinal scale ranging from 1 = less than 9th grade to 7 = graduate school), and personal or family experience with life-threatening illness, caregiving, intensive care unit (ICU), or hospice (from 0 to 4).
We tested the path model with Mplus (Muthen and Muthen 1998), a comprehensive statistical modeling program that offers a diverse set of statistical models incorporating categorical and continuous observed and latent variables. In path analysis, the observed dependent variables may be continuous, categorical, or a combination of continuous and categorical. In this study, the dependent variable, PAS support, was binary; those who said "yes" when asked if there was any condition under which PAS should be allowed were coded 1; those who said "no" were coded 0. (As noted earlier, the 43 participants who said "perhaps" were excluded from this analysis.) In the general modeling framework of Mplus, binary dependent variables are reformulated as latent, continuous response variables (y*). This makes it possible to maintain linear relationships among the variables in the model (i.e., as opposed to a nonlinear relationship between x and y, using logistic regression with a binary dependent variable). In the binary case, the latent response variable formulation defines a threshold 
on y* so that y = 1 is observed when y* exceeds ; otherwise, y = 0 is observed. A linear regression equation is used to relate y* to x: y* = 
x + 
, where is a slope parameter and is a residual that is uncorrelated with x. An intercept term is not needed because of the threshold parameter . The y* formulation for the binary dependent variable, PAS support, in our model therefore focuses on changes in the values of the continuous latent response variable y*. An R2 may also be defined, which is slightly different from an R2 of regression with a continuous variable because the variance of the error term in the latent response formulation is not a free parameter. The path coefficients reported are standardized, and the significance of each parameter was determined (i.e., for p < .05, the ratio of the parameter to its standard error was 1.96 or greater, and for p < .10, the ratio of the parameter to its standard error was 1.65).
|
Results |
|---|
|
TOPAbstractLiterature ReviewMethodsResultsDiscussionReferences |
|---|
|
2 (df = 4, N = 202) = 36.5, p < .001. Although the sample was not randomly selected, it is important to note that a similar ethnic distribution was found in response to a question on ending one's life included in a 1998 statewide, random-sample survey ("Do you think that a person has a moral right to end his/her life, when faced with an incurable illness?), lowest among Filipinos (45%) and Hawaiians (54%) and higher among Caucasians (70%), Chinese (71%), and Japanese (75%; A. T. Onaka, personal communication, March 1, 2000).
Factor Analysis
Using principal-axis factoring with oblimin rotation, the 25 items first yielded six factors, explaining 60% of the variance. Although all factors had eigenvalues of greater than 1.0, they did not make clear conceptual sense. For example, the sixth factor had only one substantive loading, eight items loaded on two factors, and one item did not load on any factor. Several items were eliminated, and the next analysis, using 21 items, yielded five conceptually clear factors explaining 59% of the variance. The solution is shown in Table 2 under these factor labels: Discuss/Document, Trust, Fears, Religious Guidance, and Fatalism. The Cronbach's alphas for the factors ranged from .66 to .78.
|
|
Path Model
Results of the path model are shown in Table 4 . The significant paths are summarized in Fig. 1. First, the findings support the notion that respondent characteristics, including ethnicity, were associated with attitudinal factors. Specifically, those more likely to advocate discussing/documenting wishes were Catholic and had more years of education. The Trust factor was positively associated with Filipino or Japanese ancestry and senior generation and negatively with immigrant status. Senior generation and Hawaiian or Filipino ancestry were associated with the Religious Guidance factor. The Fears factor was negatively associated with Filipino ancestry and education. The Fatalism factor was positively associated with Chinese or Filipino ancestry and negatively associated with education and immigrant status. The R2s for the Discuss/Document, Trust, Religious Guidance, Fears, and Fatalism factors were .08, .14, .14, .13, and .15, respectively. Second, the full model supports the notion that attitudinal factors influence PAS support. Specifically, PAS support was associated with four attitudinal factors: Discuss/Document (positive), Religious Guidance (negative), Fears (positive), and Fatalism (negative).
|
|
|
Discussion |
|---|
|
TOPAbstractLiterature ReviewMethodsResultsDiscussionReferences |
|---|
First, what is known about the Filipino American group that may explain the direct effect seen on PAS support by Filipino ancestry? Similarly to Hispanic Americans, Filipino Americans are from a country once under Spanish rule (for about 300 years), and the majority are Catholic and relatively recent immigrants to the United States. Within this study, the Filipino group had the largest proportion of first-generation immigrants, had no respondents with graduate degrees (compared with 18%–37% of other groups), and had the least amount of experience with life-threatening illness and family caregiving. Thus, they were probably the least acculturated of the groups to the ways of dying in the United States and the most likely to consider their traditional values in response to questions about end-of-life planning. Important Filipino values include bahala na (the belief that most things are outside individual control and in the hands of God), utang na loob (debt of gratitude within relationships, e.g., of children toward their parents), and an avoidance of hiya (shame caused by losing face; Ponce and Forman 1980). Some of these values were reflected in the model's attitudinal factors. For example, the positive association between Filipino ancestry and fatalism about death's timing fits with the concept of bahala na. The positive association with trust in others to make decisions and the negative association with fears/anxieties at life's end fit with the concept of utang na loob (i.e., why be anxious when one's children are indebted to do their best by you?) The positive association with religious guidance is in line with the important role of church and Catholic doctrine among Filipinos. Not captured by the model, perhaps, were items that reflected the strong belief that "suicide is a sin," the shame (hiya) that suicide would bring to the family, and fears about the implications of suicide on funeral and burial practices. These notions were expressed by Filipino American participants in our earlier qualitative studies.
The other group that did not fit the proposed model was Native Hawaiians. Similarly to other Native Americans, Native Hawaiians have experienced a loss of land and power and several centuries of disenfranchisement and discrimination within their traditional homeland. Although Native Hawaiians live in communities throughout the state, "Hawaiian homestead" communities (those in which land and housing are available at a reduced price) are located in rural areas that are far from major medical centers. Perhaps most applicable to this study are the facts that Native Hawaiians have the worst health status, are least likely to have comprehensive health insurance, and have the shortest life expectancy of the major ethnic groups in the state (Blaisdell 1993; Blaisdell and Mokuau 1991). The fact that few Native Hawaiians live to be seniors is reflected in the age structure of this group; whereas Native Hawaiians comprise 19% of the general population in the state, they comprise only 7% of the population age 60 and older (Hawaii Executive Office on Aging 1995). Thus, the model may not have adequately captured a sense of forced extinction that is likely felt by many Native Hawaiians and that would logically lead them to oppose PAS.
Despite this, we believe that effort to articulate attitudes related to end-of-life planning should be pursued. The factor analysis procedure was successful in identifying five relatively distinct constructs—Advocacy to Discuss/Document End-of-Life Wishes, Trust in Others to Make Decisions, Reliance on Religious Guidance at the End of Life, Fears and Anxieties About Life's End, and Fatalism About Death's Timing—from items that emerged from qualitative research in this area. We believe that this work, if expanded and refined, may lead to a mechanism for measuring attitudes related to end-of-life practices that are sometimes attributed to ethnic membership, and may in fact stem from specific cultural and religious traditions, but that have been modified through individuals' encounters with various social, political, educational, and health care systems.
Since this study was conceptualized, other investigators have published work on attitudes and end-of-life issues. For example, research has suggested that individuals across ethnic groups may feel that talking about something or taking an action might bring "bad luck." This is reflected in the adage "Don't count your chickens before they hatch." Specifically related to end-of-life practices, in focus groups conducted in New York (with separate groups for Caucasian, African, and Hispanic Americans), some members of each group believed that "completing a health care proxy form could actually bring on an adverse medical outcome" (Morrison, Zayas, Mulvihill, Baskin, and Meier 1998, p. 122). Carrese and Rhodes 1995 also found concern that completing an advance directive might chance fate in a study of Navajo. In the current study, fatalism was negatively associated with PAS support. Two of the ethnic groups—Chinese and Filipinos—were associated with a more fatalistic attitude toward the timing of death, either in terms of tempting early death by talking about it or trusting that the timing of one's death was in control of larger forces. It makes sense that education was negatively associated with the Fatalism factor. Immigrant status was negatively associated as well, which is in line with Brislin 1981, who found that people who emigrate often have a greater internal locus of control than their counterparts who stay home and that successful immigrants show an increase in the belief that "problems can be addressed through individual effort rather than left to fate or chance" (p. 306).
Also explored by other investigators was the relationship between support for PAS and the fear of becoming dependent or burdensome at life's end. For example, Chin and associates 1999 found that Oregon residents who requested PAS expressed more concern about loss of autonomy and control of bodily functions than did the study's control patients who died from similar diseases but without PAS. Blendon, Szalay, and Knox 1992 found that the fear of burdening families was the main reason that survey respondents considered alternatives to end their lives if terminally ill. These authors also noted that other common end-of-life fears for Americans include becoming a financial burden, living in pain, being dependent on machines, and ending up in a nursing home. Similarly, the current study found a positive association between support for PAS and fears and anxieties about life's end. This factor was negatively associated with education and with Filipino ancestry, likely for reasons of high intrafamily reliance and utang na loob, as described earlier.
A measure of religiosity is often included in studies of end-of-life attitudes and practices. Singh 1979 measured religiosity in terms of frequency of church attendance and found that strong religious commitment and frequent religious attendance, regardless of religion, were associated with low support of PAS. Lichtenstein and colleagues 1997 operationalized religiosity as "importance of religion in your life" and found it to be the only independent variable in their study significantly associated with support for PAS. The current study included four attitudinal items about the importance of religious guidance at the end of life (Table 2 ) and found that the measure was negatively associated with support for PAS. As others have found, senior generation was positively associated with religiosity (Kalish and Reynolds 1976). Filipino and Hawaiian ancestry were also both positively associated with seeking religious guidance at life's end. As noted earlier, Catholic church and doctrine strongly influence the Filipino worldview. Spirituality is important among Hawaiians as well, as it is closely integrated with the care of the land (aina) and living in harmony (lokahi) with the environment (Dudley 1990; Mokuau and Browne 1994).
The Discuss/Document factor was positively associated with support for PAS. However, it was the least well-explained factor in the model (R2 = .08). That it was positively associated with education was not surprising. Its association with Catholic faith was not expected, however, as within the sample, smaller proportions of Catholics reported having a will (53% vs 65% of non-Catholics), living will (45% vs 59% of non-Catholics), or durable power of attorney (24% vs 38% of non-Catholics). Also, there was a significant negative correlation between the Fatalism and Discuss/Document factors (r = -.30, p < .01). This may suggest that these two factors may tap opposite ends of a continuum between internal and external locus of control regarding end-of-life planning. Specifically, individuals with a high internal locus of control would be more likely to believe in taking actions to influence the quality or timing of their deaths, whereas individuals with a high external locus of control would be less likely to plan for death, leaving this in the hands of a higher power. This hypothesis could be tested by co-administering a locus-of-control instrument (Rotter 1973; Wallston, Wallston, Kaplan, and Maides 1976).
The current study found no relationship between trust in others to make decisions and PAS support. In a similar exploration, McKinley, Garrett, Evans, and Danis 1996 hypothesized that a cultural history of discrimination and unequal access to care might result in feelings of distrust in the health care system by African Americans and might explain why African Americans, compared with Caucasians, were more likely to desire life-sustaining treatments and less likely to have plans to complete a living will. They found, however, that trust was not correlated with these outcomes. In fact, 96% of both groups said they trusted the health care system and only 16%–19% of both groups feared inadequate medical care. As suggested by McKinley and associates 1996, the operationalization of this factor needs more refinement. Both studies used a combination of global questions (e.g., "Generally speaking, doctors can be trusted") and personal questions (e.g., "If fatally ill, I would rely on my doctor to make the right decisions for me"), and perhaps the two need to be separated. Also, the current study mixed questions about trust in physicians and trust in family members, which should be separated. The former is linked to trust in the health care system and degree to which one feels it will honor one's wishes. The latter may be tapping how individuals make decisions (e.g., by themselves, with the help of an intimate partner, by deferring to someone else, with concern for the good of the whole family).
Strengthening this line of work on attitudes toward end-of-life planning may be a measure of death anxiety. For example, the Death Attitude Profile—Revised includes measures of death acceptance (neutral acceptance, approach acceptance, and escape acceptance), fear of death, and death avoidance (Wong, Reker, and Gesser 1994). It is likely that all five of these constructs are influenced by cultural traditions and, in turn, influence end-of-life choices. Relevant items from the Wong and associates' scale include "I avoid thinking about death altogether" and "I try to have nothing to do with the subject of death."
Other variables included in the study could benefit from additional refinement as well. The operationalization of the Experience variable was perhaps too broad, including personal and family experience with serious illness, caregiving, ICU, and hospice. More promising may be to measure negativity associated with these experiences, as study participants reported in response to open-ended questions that having a negative experience watching a family member die spurred them to take a more active role in thinking about and planning for death. Another relevant predictor might be current health status, which was not included in this study.
The study was limited by the small size and nonrandom selection of its sample. Participation was voluntary, and recruitment difficulties among Chinese (for reasons of discomfort with the topic) and Hawaiians (who took time to "check out" the interviewer before consenting to an interview) suggest that the study's data reflect the thoughts of individuals within each group who were most open to talking about death and dying with a stranger. However, as noted earlier, a question on an individual's right to end his or her life if terminally ill included in a random sample, statewide survey found a similar distribution of support by ethnicity as this study found for PAS support—lower among Filipinos and Hawaiians and higher among Caucasians, Chinese, and Japanese.
For practitioners, the findings underscore the importance of two things: (a) learning about the values and experiences of the different ethnic groups with which one works and (b) taking a good history. Numerous books and articles are appearing that provide information on different cultures, identify important cultural values, and examine how these values may influence health-related behaviors (Braun et al. 2000; Kalish and Reynolds 1976; Koenig 1997). Taking a good history includes asking open-ended questions about a person's education and experience but, more important, about their level of religiosity, their propensity to plan ahead, support for discussing/documenting wishes, the extent to which they feel that their death is part of a larger plan or outside of their control, if they think discussing death will bring it on, and how much they would like to share or give away decision-making power. We agree with other investigators who recommend that practitioners not delay talking with patients about end-of-life issues. Such discussions should be initiated after getting to know a patient but before the patient becomes ill (Miles, Koepp, and Weber 1996) and can be facilitated by the use of values history instruments (Doukas and McCullough 1991; McIver-Gibson 1990). Although the current emphasis on cultural competence is important and is aided by gaining knowledge about the history and values of a specific culture, it is imperative that practitioners not use information about cultural norms to stereotype individual patients. Rather, by asking open-ended questions related to individual beliefs and attitudes (that may indeed be influenced by that individual's ethnic membership), a practitioner will likely get a more complete and useful picture of the patient on which a good helping relationship at the end of life can be established.
Conclusion
Several investigators have found that many U.S. minority groups feel and act differently than the dominant North American culture when it comes to end-of-life decision making. Although there is clear proof of general ethnic-specific differences, empirical and everyday evidence acknowledge a range of within-group differences as well. This study presents a broader, empirical approach to the issue by trying to identify and measure attitudinal factors related to PAS support that may apply across ethnic groups. Although five attitudinal factors were identified, contrary to the hypothesis ethnic membership was directly associated with PAS support, that is, individuals of Filipino and Hawaiian ancestry were less likely than others to support PAS even after controlling for other respondent characteristics and attitudinal factors. Although further refinement of the conceptual model and its measurement is needed, this study offers a step toward clarifying and enhancing understanding of cultural influences on end-of-life choices. Practitioners are encouraged to learn about the cultural traditions of the people they serve and to ask open-ended questions that will elicit underlying attitudes that influence end-of-life behavior.
|
Acknowledgments |
|---|
Received for publication April 15, 1999. Accepted for publication June 30, 2000.
|
References |
|---|
|
TOPAbstractLiterature ReviewMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M Hall, F Trachtenberg, and E Dugan The impact on patient trust of legalising physician aid in dying J. Med. Ethics, December 1, 2005; 31(12): 693 - 697. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S Frileux, C Lelievre, M T Munoz Sastre, E Mullet, and P C Sorum When is physician assisted suicide or euthanasia acceptable? J. Med. Ethics, December 1, 2003; 29(6): 330 - 336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. WERTH Jr., D. BLEVINS, K. L. TOUSSAINT, and M. R. DURHAM The Influence of Cultural Diversity on End-of-life Care and Decisions American Behavioral Scientist, October 1, 2002; 46(2): 204 - 219. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
