Family Decision-Making about End-of-Life Care


A family therapist doing research that seeks families and patients deal with the process of making decisions at the end-of-life also serves as a clinician providing palliative care consultation at a local hospital. When a family who gave their informed consent to participate in her study confesses that they only did so because they thought their family member who is dying would not receive needed treatment with their participation, a number of ethical questions arise.


Part I

Dr. Luci Menendez is a licensed family therapist with years of clinical experience helping families cope with the grief associated with both death and non-death related loss as well as a child and family development researcher at a large university.  Luci has been working as part of an interdisciplinary palliative care consultation service at a local hospital.  The palliative care team specializes in providing pain and symptom management for patients and families facing chronic or terminal life-limiting illnesses.  As a family therapist paid through a contract with the hospital, Luci specializes in “relational pain management,” helping to sort through emotional and sometimes conflict-laden family dynamics.  Physicians refer patients for a consultation with the palliative care team when they believe cure is no longer a realistic treatment goal.  All members of the palliative care team then meet with the patient and family to decide if they would like to continue receiving treatment focused on cure or shift care goals to emphasize comfort care.

In addition to clinical work, Luci is interested in research that focuses on the family as the unit of investigation with the goal of eventually developing and testing clinical interventions to assist patients and families through the process of making decisions at the end-of-life.  She approaches her research and practice from the perspective of general systems theory. Luci plans to conduct a study using validated measures of family cohesiveness and adaptability, participant observation of family interactions and decision-making processes, and qualitative interviewing.


  1. What ethical issues arise from Luci’s dual role as a clinician/researcher?
    1. Could these issues be avoided?  Should they be?
    2. How does this dual role facilitate and hinder Luci’s clinical work and research?
  2. Should Luci be asking dying patients and their families to participate in research at all? Why or why not? 
    1. Should dying patients and their families be viewed as “vulnerable populations”? Why or why not?
    2. What additional ethical issues arise if using dying patients and their families is classified as research involving vulnerable populations?

Part II

Luci gains IRB approval for her study.  She recruits research subjects through the network of physicians making referrals to her palliative care team.  After talking with patients and families about the need for a palliative care consultation, the physicians secured permission for Luci to contact the patient and families about the possibility of participating in a research study.  During Luci’s first meeting with patients and their families, she explains both her normal clinical role on the palliative care team and her interest in researching the decision-making processes families use related to palliative care.  Luci describes the study and answers all questions.  She carefully stresses that receiving her clinical services and that of the rest of the palliative care team is not contingent upon their participation in the research.  While everyone is present, Luci then asks the patient and family members to each sign an informed consent document, which includes the statement that research participants are free to stop participating and withdraw consent at any time for the use of any data they have provided.

Thirty minutes prior to the start of each palliative care family conference, Luci meets simultaneously with the patient and his or her family to reconfirm their decision to participate and to administer pre-test measures.  After each palliative care family conference, Luci records participant observation data.  On separate occasions she conducts qualitative interviews, one with all family members present, then one with each individual, followed by a second, follow-up interview with the entire family.  So as to not overly tax the participants, Luci keeps all questionnaires and interviews brief.

These procedures appear to be going well until Luci meets with one particular family. Though the patient and all family members had signed informed consent documentation stating they were freely volunteering for this project, a comment was made by a family member during an individual interview about how strongly the patient’s primary care physician urged the patient to participate in the study.  When Luci follows up on this comment during the second family interview, the same family member explains that they got the impression that the physician thought the family would benefit from extra interaction with a family therapist.  Luci re-explains that receiving clinical services was never dependent upon participation in the research project and the patient and family could have met with her as often as they liked and as time allowed.  Then, several family members shyly confess they had only agreed to participate in the study out of fear that the patient would not receive all the treatment the doctor thought best.


  1. Was undue influence or coercion involved in this case?  By the physician?  By Luci?  By family members? 
    1. How is undue influence defined? 
    2. Is some influence okay or is any amount of influence understood to be coercive?
  2. If some family members would like to participate in the study while others do not, how should Luci proceed given that her research interests are in collecting family-level data? 
    1. Should the desires of any single member of the family carry sufficient moral weight as to override the desires of all others? 
    2. If only one person does not want to participate but all other family members do, does that one person have the right to insist that family-level data not be used?   
    3. Should Luci have obtained the informed consent of each individual member separately before proceeding with the research project?
  3. If Luci uses data only from some members of the family but not others, does this invalidate her quest to gather full family-system data? 
    1. Are the data she has collected so far from this family valid, now that she knows some members were reluctant participants? 
    2. Can Luci trust the validity of data from other families who were referred by this one physician?

Part III

Although Luci is worried about the loss of data to her study (especially since full family participation was hard to come by), she reminds the patient and family that they are free to withdraw their participation and data from the study at any time.  At this point, the dying patient, with whom Luci has developed a close therapeutic relationship, reiterates his interest in participation in the research project and urges his family to “please participate.”

As a family clinician-investigator, Luci has duties not only to see that her research causes minimal harm, but to intervene in harmful family dynamics.  Luci is reminded that much of her interest in collecting family-level data is that a dying patient’s decisions about end-of-life care have enormous impact on family members and that in some cultures duty to family carries more moral weight than individual preference.  Indeed, it is a holistic focus on the family system that distinguishes family researchers and clinicians from others who study and intervene with individuals.  She hopes her research on the family as a whole will lead to clinical interventions that strengthen family relationships during such a vulnerable time.


  1. How should Luci respond to the dying patient’s request that reluctant family members participate in the research? 
    1. If she says nothing, is this a neutral response or does it have the effect of helping to insure the inclusion of these data?
    2. Does the fact that a person is dying automatically add moral weight to what a person wants to the neglect of other family member’s desires?  Is this coercion?  Why or why not?

Note on Teaching this Case

This case was constructed with several pedagogical assumptions and goals in mind.  First, it is geared toward investigators who are already familiar with the fundamentals of ethical research practice, particularly regulatory standards associated with Institutional Review Boards (IRBs). As such, this case study intentionally focuses on ferreting out and potentially challenging assumptions embedded in traditional research ethics protocols.

Second, it is designed to highlight some of the unique challenges faced in clinical research, which those conducting traditional basic research are less likely to face.  Similarly, this case raises issues that are of central concern to family systems researchers, but may seem less relevant to social scientists whose unit of analysis is at the level of the individual, or larger social groups whose members may be anonymous to one another or have less intimate connections. 

Finally, those unfamiliar with the nuances of clinical research or a systemic (vs. reductionistic) approach to science, may find the myriad issues raised in this case to be so multifaceted that the case loses pedagogical efficacy.  While this is a risk, it was decided that the realism associated with simultaneously wrestling with the complexities of this case offered an alternative to case studies frequently found in the ethics literature that make clear philosophical points at the expense of face validity.

. . Family Decision-Making about End-of-Life Care. Online Ethics Center. DOI:.

It is certainly understandable, even commendable, that those who provide palliative care consultation services would want to have a better understanding of variables that may affect the quality of care they provide.  Some of this understanding will be acquired “on the job,” as one provides these services—at least by perceptive consultants.  But systematic research is likely to bring other important matters to light, challenge assumptions that might otherwise adversely affect the services offered, and so on.  In short, the recipients of these services can benefit even more from consultants who effectively incorporate research results into their practice.  The sort of research interests that Dr. Menendez has are consistent with her concern to provide the best services she can — and to help others do so as well.

At the same time, she should have some real concerns about whether it is appropriate for her to serve as therapist and researcher for the same clientele, whether individuals or family units.  The informed consent issues raised in Part II of this case illustrate why.  Apparently some primary care physicians, intentionally or inadvertently, have led family members to believe that the quality of service they can expect will be adversely affected by not agreeing to participate in the study.  If this worry is engendered by their primary care physician, who is not providing the therapeutic services in question, it makes good sense to suppose that the therapist/researcher could be seen as equally, if not more, threatening.  Will the therapist/researcher be less interested in those patients, or their family members, who indicate they do not want to participate?  The very fact that patients or family members might worry about this could have a negative affect on the therapeutic relationship they have with Dr. Menendez. 

So, as researcher, Dr. Menendez might restrict her research to patients and their families who are not under her care.  This means that if her own patients and their families are to be participants in the sort of research project she is interested in, another researcher would be needed.  There is still no assurance, of course, that the problem will be solved.  Patients and their families would be informed that, in addition to consulting with their therapist, there will be someone else involved, albeit for research rather than therapeutic purposes.  Will they worry that the therapist and researcher are cooperating with each other in such a way that their therapeutic services will be affected by not participating in the study?

Adding a researcher to the mix may raise other patient and family concerns.  Not only are they expected to discuss sensitive, private matters with a therapist, they are to be observed by a third party, a researcher with whom they have no other relationship.  Aside from causing them discomfort, this might also have a negative affect on the therapeutic relationship by, say, causing them to be more reticent.  If this factor cannot be ruled out, this could also affect the validity of any claims about what goes on between therapist and patient/family that they are not being observed for research purposes — whether by their therapist or a third party. 

There does not seem to be a sure way of totally eliminating such worries about the possible adverse affects of the research on therapeutic services, or worries about the validity of the research data itself.  A point to be emphasized, however, is that special care needs to be taken in regard to informed consent.  Patients and their families need assurances that participating (or not participating) in the research project will not affect the quality of services they will receive.  The scenario in Part II suggests that very little, if any, monitoring of the informed consent process was done in regard to the role of the primary care physicians.  At the very least, this issue should be addressed.  The concern is that undue pressure might have been used on patients and their families, or that this is their perception.  But the concerns do not stop there.  As just pointed out, additional worries about the quality of care provided by the therapist/researcher remain, even if referring physicians handle things very well.

These problems aside, Dr. Menendez, and perhaps her patients as well, would like to have full participation by both patients and family members.  What if one or more of the family members express reluctance to participate, or outright refuse?  Noncoercive efforts to persuade may be appropriate, although it can be very difficult to determine where to draw the line when attempting to persuade the reluctant to join with their already willing family members.  The researcher already has strong allies in this circumstance, unless, of course, there is a history of significant tensions or divisions among family members.  So, additional efforts to recruit family members can easily become a matter of undue pressure.  Dr. Menendez may be disappointed at not gaining the consent of all patients and family members, but this need not be fatal to her project — especially if she is joined in her research efforts by other researchers, thus enlarging the potential pool of participants.

Ethical Issues and Analysis

Part I of this case study introduces Dr. Luci Menendez as both a researcher and a clinician who seeks to develop an integrative program of research whereby her clinical work informs her research and vice versa. Critical to this case is an understanding of the ways in which general systems theory informs Luci’s research and clinical practice.  General Systems Theory (von Bertalanffy, 1968), the basis of family therapy and many theories of family process, is most readily epitomized as ‘the whole is greater than the sum of its parts.’  Individual parts of the system are interdependent and information feedback loops between parts or between the system and the broader environment function to keep the organization of the system relatively stable.  “This systemic approach has led to a method of treating psychological problems and of posing research questions that is fundamentally different from the traditional, individually based one”(Copeland & White, 1991, p. 8). 

Copeland and White (1991) argue that family researchers, such as Dr. Luci Menendez, not only have the traditionally recognized responsibility to assess the effects of a study on individual research participants, but have a special ethical responsibility to attend to the impact of the study on the family as a whole.  Similarly, family therapists are ethically required to attend to both the well-being of individual family members and the well-being of the family as a whole, a difficult balance to achieve at best. 

One may rightly question from the start whether Luci should have recruited families to her study with whom she would eventually have a clinical relationship.  Whether or not Luci recruits research subjects from the same population that she will be serving clinically is partially influenced by the availability of palliative care consultation services.  These services are still relatively new and not all hospitals or communities have interdisciplinary palliative care teams. The very fact that these services are new may be an argument for the importance of researching currently unexplored issues so as to increase evidence-based clinical practices. However, if Luci’s team is the only one of its kind that is readily accessible to Luci for research, she may be at higher risk for unwittingly pressuring her clients to participate. 

Only interviewing palliative care patients and families not receiving services from Luci’s professional team would ostensibly lessen the complexities of this case by reducing the formal fiduciary relationships Luci has with clients/subjects. However it is debatable whether the absence of formal relationships with specific family members completely eliminates Luci’s more general duty as a socially sanctioned professional to protect the well-being of society’s members.  In other words, even if Luci interviewed research subjects with whom she does not have formal therapeutic relationships, the fact that she is a clinician with a specialized knowledge and skill set may still have some ethical bearing on her research relationships.

Though others may disagree, I would argue that Luci’s two roles are neither 100% separable, nor equally exchangeable.  Luci’s membership in a publicly recognized and regulated clinical profession with all of its attending benefits (e.g., status), obligates her to give priority to her clinical role over her researcher role.  In other words, Luci can be a clinician without assuming a researcher role, but her clinical knowledge must inform her research choices. Her clinical knowledge likely makes her more sensitive to the types of harm that may befall individuals and families participating in her research project which may obligate her to take steps above and beyond those required by federal, state, and institutional research regulations.

Recognizing the complexity of her dual role as a clinician/researcher, Luci took precautions in her research design.  First, she used a two stage recruiting process, whereby patients and families were first invited to consider participation in research by someone other than the researcher, the physician in this case.  Whereas this was intended to increase the autonomy of family members in deciding whether or not to participate in the study, it increased Luci’s risk of having nuances of the study misrepresented.  Furthermore, Luci failed fully to account for perceived power dynamics in the physician’s relationship with the family, leaving them vulnerable to perceived (if not actual) “authoritative persuasion.”

Second, when meeting with families to describe the research opportunity, Luci made explicit the dual nature of her relationship with patients and families, stressing that clinical care is a higher priority than research, and that the decision whether or not to participate in research would not negatively affect the clinical services they received.  During informed consent procedures, Luci also explained the on-going voluntariness of research participation.  While these precautions are commonly required by Institutional Review Boards as means of protecting individual research subjects, additional efforts may be necessary to protect the family system.

For instance, the case does not specify the exact nature of the informed consent document Luci has each family member sign, but it does say the discussion took place with everyone present.  Copeland and White (1991) note that “especially in studies in which families are asked to discuss important, real issues together [e.g., end-of-life care], the promises of anonymity and confidentiality about what they say, usually afforded to research subjects, are limited because the other family members are sitting there and listening” (p.4).  Per most IRB requirements, the informed consent document should discuss the limits of confidentiality.  This is typically understood as delineating the conditions under which the researcher may not keep absolute confidentiality. 

Confidentiality is typically understood as the ethical mechanism through which we respect the right of privacy of individuals.  But does this individual-focused understanding of privacy and confidentiality adequately apply to information about relationships, which by definition involve more than one individual?  Family researchers are faced with the dilemma of gathering and protecting information that from the perspective of individual family members may be considered quasi-private.  There may be a genuine risk of harm to individuals and/or family relationships if some members of the family disclose relational information that the other members did not want disclosed. 

In this case, a fully ethical approach to informed consent in family research might also include a discussion of the fact that data collected from one individual, even during individual interviews, cannot be completely separated from information about other members of the family because the focus of the research is on shared family history and dynamics. One approach is to include a statement on the consent document stating that agreement to participate in family data collection includes giving permission to other family members to disclose potentially private information about one another. 

Having such a statement included in consent procedures allows the researcher to explain the importance of gathering “un-edited” family data, while simultaneously facilitating family members’ discussions about possible limitations on the type of information they will share with the researcher.  Of course, research subjects are always free to edit their responses, but by making this process explicit, the researcher may be able to at least gather information directly from subjects about the limitations of the data rather than solely relying on hindsight speculation about missing data.

Explicitly highlighting interest in the family as a whole also gives the researcher an opportunity overtly to discuss family dynamics in the process of consenting to participate in research.  Families differ in important ways from other groups studied by researchers (Copeland & White, 1991; Greenstein, 2001).  In addition to being interdependent systems of individuals, families “develop private, idiosyncratic norms and meanings about their own activities. . . , [creating unwritten] patterns of, and rules for, behavior” (Larzelere & Klein, 1987, in Greenstein, 2001, p. 11) that are often hidden from public view.  Families have ways of restructuring their view of themselves in order fit these family rules and expectations as a means of managing family tensions and maintaining family stability (Copeland & White, 1991). Family members also have multiple statuses and enact multiple roles simultaneously (e.g., father, son, and brother) requiring researchers to be sensitive to the fact that the kinds of responses offered by family members may depend on the role and status the individual is occupying in the context of gathering family data (Gelles, 1978, in Greenstein, 2001). 

These systemic considerations are not typically considered in the traditional bioethics or research ethics literatures.  Relying on an individualistic approach to research ethics, it is tempting to resolve Luci’s case by simply saying, “If a family member does not want to participate, that’s the end of the story; just collect data from those who agree.”  This response is problematic in at least two ways.  First, the validity of system-level data is likely to be compromised, thereby altering the risk-benefit analysis used by IRB reviewers.  Second, assuming a purely individualized approach to ethics in the context of family dynamics may itself be a morally questionable activity that may increase the risk of harm to the family system

Ivan Boszormenyi-Nagy (1984, 1986, 1987, 1991), a founding family therapy theorist, argues that “relational ethics” is critical to healthy family functioning, such that failure of each family member to give “due consideration” to the interests of other members is seen as the heart of family dysfunction.  Nagy (1991) claims that family functioning is enhanced when members of the family can trust that the family system as a whole will facilitate the process of balancing considerations of the well-being of oneself with considerations of the well-being of others. 

In this case study, some family members acknowledged during data collection that their motivation for participation had been out of a perceived benefit to the dying patient.  From a traditional perspective, subject participation “out of fear” of lost benefits raises questions of voluntariness and possible coercion (both direct and indirect).  Superficially, this circumstance arose due to miscommunication.  At a deeper moral level, however, it could be argued that the situation is also borne of “relational ethics,” in that family members gave “due consideration” to the wishes and interests of other members of the family system. 

Luci’s response is in keeping with traditional research ethics: she reminds family members of their individual freedom to withdraw from the study.  In her attempt to protect the rights of individuals, however, does Luci risk harming the system by challenging the family’s “idiosyncratic norms. . .[and unwritten] patterns of, and rules for, behavior” (Larzelere & Klein, 1987, in Greenstein, 2001, p. 11), which has demonstrably included “due consideration”?  In other words, by highlighting individuals’ rights to withdraw their participation, is Luci, in effect, suggesting that “due consideration” of other family members’ interest in contributing family-level data (e.g., the dying patient) is not relevant?  In doing so, does she undermine the trustworthiness of the family system to support “due consideration” — a key factor in healthy functioning according to Nagy  (1991)?  If this line of reasoning holds, then Luci’s adherence to traditional research ethics protocols may violate her ethical responsibilities as a family clinician and researcher to protect (and enhance when possible) the welfare of the family system.

Biomedical ethics and most approaches to research ethics emphasize individual autonomy in decision-making, but this tends to decontextualize people from their social context, a criticism increasingly explored in feminist ethics.  Recognizing that human beings have autonomous moral status (i.e., their moral worth is not dependent on external considerations) need not automatically be equated with decision-making that is free from the influence of others.  Certainly, the influence of the researcher on the consent process needs to be kept to a minimum.  However, it is morally suspect to presume that decision-making itself must always be free of the influence of others. 

While some attention has been given to cultural or societal-level groups (e.g., Native American tribal considerations), little discussion has occurred about the moral relevance to decision-making of intermediate level groups such as the family. Yet in many cultures these more personal groupings impact one’s daily life most, and it is not uncommon for loyalty to one’s family to be given priority over individual interests. If Nagy’s theory of family functioning is correct, it would suggest that being in intimate relationships with others changes the level of influence on ethical decision-making we consider to be appropriate, particularly in contrast to non-intimate relationships. 


  • Bertalanffy, L. von. 1968. General Systems Theory. New York: Guilford Press.
  • Boszormenyi-Nagy, I. 1984. Invisible Loyalties: Reciprocity in Intergeneration Family Therapy. New York: Brunner/Mazel, Publishers.
  • Boszormenyi-Nagy, I. & Krasner, B. R. 1986. Between Give and Take: A Clinical Guide to Contextual Therapy. New York: Brunner/Mazel, Publishers.
  • Boszormenyi-Nagy, I. 1987. Foundations of Contextual Therapy: Collected Papers of Ivan Boszormenyi-Nagy, M.D. New York: Brunner/Mazel, Publishers.
  • Boszormenyi-Nagy, I.,  Grunebaum, J. & Ulrich, D. 1991. Contextual Therapy. In A. S. Gurman & D. P. Kniskern (Eds.). Handbook of Family Therapy, Volume II. New York: Brunner/Mazel.
  • Copeland, A. P., & White, K. M. 1991. Studying Families (Applied Social Research Methods Series, Volume 27).  Thousand Oaks, CA: Sage Publications.
  • Gelles, R. J. 1978. Methods for studying sensitive family topics. American Journal of Orthopsychiatry, 48(3), 408-424.
  • Greenstein, T. N. 2001. Methods of family research. Thousand Oaks, CA: Sage
  • Larzelere, R. E., & Klein, D. M. 1987. Methodology. In M. B. Sussman & S. K. Steinmetz (Eds.), Handbook of marriage and the family (pp. 126-156). New York: Plenum.