Why Should Medical Care Be Family-Centered?: Understanding Ethical Responsibilities for Patients’ Family Members
The movement toward family-centered care in medicine has brought more attention to the importance of patients’ families in clinical practice. However, an issue central to the movement—why families matter—remains underexplored in both the family-centered care and the wider biomedical ethics literature. Given the possibility—and reality—of conflicts between patient and family interests, the need to think about how far providers should go in attending to a family’s welfare and wishes, and the focus that family-centered care has placed on further inclusion of families into care, a clearer understanding of providers’ responsibilities for family members is needed. In this paper, I argue for a framework in which providers have non-instrumental responsibilities for their patients’ family members but their responsibilities for their patients remain primary. To reach this conclusion, I first identify both role- and relationship-based reasons for thinking providers have non-instrumental responsibilities for family members. I then explain why clinicians should still largely prioritize the wishes and interests of their patients over those of their families and provide a reasoning structure for use in situations where wishes and interests conflict.
In recent years, hospitals, clinics, and professional organizations have with increasing frequency pledged their commitment to “patient-and family-centered care” (PFCC). The movement toward PFCC is especially pronounced in pediatrics, where the American Academy of Pediatrics (AAP) has a long-held, explicit commitment to PFCC (AAP 2012). However, the unified movement toward PFCC obscures differing conceptions of its purpose. First, patient-centered care, as opposed to provider- or disease-centered care (Davidson et al. 2007, 605), focuses [End Page 159] on increasing patient involvement in care to accomplish two related, but distinct objectives: improving health outcomes and better respecting patients’ wishes. The Institute of Medicine’s definition of patient-centered care as “care that is respectful of and responsive to individual patient preferences, needs, and values” is representative (IOM 2001, 40).1 For instance, rather than simply stating what she thinks is the best treatment option for a patient, a provider committed to giving patient-centered care should in many cases seek out the patient’s input on different treatment options. Doing so respects the patient’s wishes—the respect objective—and may also result in a higher likelihood that the patient will adhere to the treatment plan—the health outcome objective. Although these two objectives are importantly different, they share a common focus: the patient. It is clear that the intention of patient-centered care is to shift the focus of care from provider to patient.
In discussions of family-centered care, the ultimate focus of the care can be less clear. Increasingly, discussions of family-centered care link it to patient-centered care, as in the AAP’s most recent statement (AAP 2012).2 This connection highlights the ways in which increasing the involvement a patient’s family has in his care can accomplish the same objectives as increasing the patient’s own involvement. When a family serves as a patient’s surrogate, heeding the family’s wishes just is to respect the patient’s wishes (so long as they act as a surrogate should), and when they serve as a patient’s caregivers, considering their wishes helps ensure that the patient receives the best care. For instance, if a provider has a choice of two treatment plans and the family expresses a preference for the care plan that will be much easier for them to implement, the provider ignores the family’s preference at the patient’s peril (Brosco 2015).
Family-centered care, understood in this way, fits easily within an enlightened understanding of physicians’ fiduciary duty to act in their patients’ best interests and follow their wishes. Indeed, an abundance of evidence exists that family-centered care is beneficial for patients and accords with their wishes. To take just two pieces of the evidence cited by the AAP in its most recent statement in support of PFCC, children recover faster and are discharged earlier after tonsillectomies when their mothers are involved in their care, and children who undergo surgery cry less, are less restless, and require less medication when their parents are present and assist in pain assessment and management (AAP 2012, 397).3 Additionally, through its inclusion of families in the decision-making process, family-centered care squares with an understanding of patient autonomy that [End Page 160] recognizes that patients often do not and do not want to make decisions in isolation from their families (Doukas and Hardwig 2003; Kuczewski 1996; Jecker 1990).4
However, many statements in support of PFCC extol its benefits for both patients and families without marking the benefits to families as of solely instrumental importance. For instance, the Institute for Patient-and Family-Centered Care, a leading proponent of the practice, states, “A key goal [of PFCC] is to promote the health and well-being of individuals and families and to maintain their control” (IPFCC 2018, italics added). Similarly, the AAP’s 2012 statement begins by describing PFCC as “grounded in a mutually beneficial partnership between patients, families, and providers” and goes on to summarize the outcomes of PFCC as follows: “Patient- and family-centered care can improve patient and family outcomes, improve the patient and family’s experience, increase patient and family satisfaction. . . .” (AAP 2012, 394; 397, italics added). While the benefits for family members that the statement goes on to mention, such as mental health benefits to mothers of children with chronic illnesses (AAP 2012, 397), have clear benefits for patients, failing to clearly state the rationale for the inclusion of families into care invites differing interpretations of what PFCC actually involves.
Indeed, some have taken family-centered care to be committed to the idea that families’ interests and wishes matter for their own sake, that is, apart from any instrumental benefits to the patient. For instance, Kuo et al. (2012, 298) claim that “FCC moves beyond patient-clinician interaction by considering the needs of all family members, not just the child.” Shields, Pratt, and Hunter (2006, 1318) include the idea that family members are care recipients in their definition of FCC.5 And Fiester (2014, 313) goes so far as to claim that an “obligation to (sometimes) prioritize the needs of family members over patients . . . follows directly from the insight of the US family-centered care movement that there is more than one person in the room who matters morally, more than one who can be harmed.”
Cases in which patients’ and families’ interests conflict present one important reason for distinguishing instrumental and non-instrumental reasons for helping family members. But guidance is also needed in cases in which a health care provider can provide a benefit to family members that has no effect on the patient. For instance, several states, including California, require health care providers to make reasonable accommodations for family members when a patient is declared brain dead before removing cardiopulmonary support. In California, these [End Page 161] accommodations need to be made for “religious or cultural practices or concerns”—in which case “the hospital shall make reasonable efforts to accommodate” them—and/or to give the family a “reasonably brief period” “to gather family and next of kin at the patient’s bedside.”6
Reasonable accommodation laws show the attraction of the idea that health care providers should take the interests of family members into account at least to some extent, apart from any effect these benefits have for patients. If one can provide a significant benefit to a family member at little to no cost to the patient, other patients,7 oneself, or the hospital, it seems clear that one should do so. Many other common practices, such as being considerate of how one breaks bad news to family members, reflect the widespread acceptance of this sentiment.
Though the intuition is widespread, there is not clear ethical guidance about how exactly we ought to understand the nature and extent of providers’ responsibilities for family members (Olson 2017). Given the possibility—and reality—of conflicts between patient and family interests, the need to think about how far providers should go in attending to a family’s welfare and wishes, and the focus that family-centered care has placed on further inclusion of families into care, a clearer understanding of providers’ responsibilities for family members is needed.8 In this paper, I argue for a framework in which providers have non-instrumental responsibilities for their patients’ family members but their responsibilities for their patients remain primary. To reach this conclusion, I first identify both role- and relationship-based reasons for thinking providers have non-instrumental responsibilities for family members. Section 1 details role-based reasons and section 2 relationship-based ones. In section 3, I explain why clinicians should still largely prioritize the wishes and interests of their patients over those of their families and I provide a reasoning structure for use in situations where wishes and interests conflict.
In presenting my argument, I focus on two types of cases. In the first, helping family members or following their wishes is neutral for the patient qua patient: doing so neither helps nor harms the patient’s health and is not against the patient’s wishes. The reasonable accommodation case mentioned above is one case of this sort. Jones, Contro, and Koch (2014) present another: providing some degree of ongoing emotional support for family members after a patient has died.
In the second, helping family members or following their wishes is negative for the patient qua patient in some respect: it is either harmful to [End Page 162] the patient’s health or against the patient’s wishes (or both). For example, increasing a family’s involvement in a patient’s care could be against the patient’s wishes or even possibly harmful to the patient’s health. Beck et al. (2009) examine parents and hospital personnel’s experiences as the Neonatal Intensive Care Unit (NICU) at Copenhagen University Hospital began experimenting with private rooms instead of housing all infants in one shared room. While parents reported much better experiences with the private rooms and hospital personnel preferred them as well, the personnel also reported that the private rooms posed challenges. For instance, one nurse noted that “[i]n highly intensive situations, [she] felt trapped in the smaller room and could not leave the infant even if she was also responsible for an infant in another room” (Beck et al. 2009, 96). The nursing staff intensified their coordination with each other to make sure that all infants did receive the care they needed (Beck et al. 2009, 96), but this study provides an example of how increased family involvement could conflict with optimal care for a patient if it is not implemented correctly or if the resources for proper implementation are not available.
In the next two sections of the paper, I use the first kind of case to highlight role- and relationship-based reasons for helping family members. And, in section 3, I use the second kind of case to argue that clinicians should still generally prioritize patients’ needs.9
1. ROLE RESPONSIBILITIES
The first kind of case—where attending to family members is neutral for the patient’s health and wishes—provides the clearest vantage point for seeing the non-instrumental reasons for helping family members. One main reason, put simply, is that one can. For instance, health care providers can aid the bereavement process of parents of deceased children by taking measures as simple as a phone call. In a study by Back et al. (2009, 477), one such parent stated, “I think it was really important that he [the physician] did call . . . it just showed me that he cared and that [5-second silence] she just wasn’t a . . . just wasn’t a patient . . . that he treated and then she didn’t make it, so, ‘Oh well’.” Beyond being a situation where one can provide help, this situation has several features that make the case for providing help particularly strong: the physician can provide significant aid to the parent (magnitude), only the physician can provide it (dependence), and the physician can do so at minimal cost to himself or others (cost). Many have argued that when these three features are present, individuals have an ethical duty to provide help (Richardson and Belsky 2004; [End Page 163] Dickert and Wendler 2009; Merritt 2011). This duty is often called a duty of easy rescue since philosophers have tended to focus on scenarios where one can save someone from death or significant harm at little or no cost to oneself.
Given the three features mentioned, a strong case can be made that the physician in this example has a duty to call the parent based on this duty alone. However, some have recently expressed concerns about the common employment of the duty of easy rescue in contemporary bioethics (MacKay and Rulli 2017; Rulli and Millum 2016; Garrett 2015). Indeed, there are good reasons to look beyond the duty of easy rescue. For one, not all cases are easy rescue ones. Full bereavement care for grieving parents involves more than a phone call. Second, broadening one’s gaze provides a more complete picture of why health care providers ought to take family members’ needs into account. In the rest of this section, I argue that health care providers have role-based reasons for providing help to family members,10 and in the next section, I argue that they also have relationship-based reasons for providing such help.
One might be tempted to say that the only role-based obligations health care providers have are to their patients. The fiduciary duty to act in a patient’s best interest has long been held to form the backbone of clinicians’ codes of ethics, and one might take this commitment to mean that one ought to focus on only one’s patients. I call this position the Strict Fiduciary View: clinicians have a fiduciary responsibility to their patients but no responsibilities qua clinician to aid anyone else.
While, as I will argue in section 3, there are strong reasons for prioritizing one’s own patients’ needs, this view interprets clinicians’ role responsibilities far too narrowly. For instance, it is commonly accepted that clinicians have a role-based obligation to treat individuals in emergencies, as the familiar scenario of a person in distress on an airplane illustrates. Relatedly, the AMA (American Medical Association) Code of Medical Ethics (2016, 8.3) includes an obligation to “provide urgent medical care during disasters” and goes so far as to state that “[t]his obligation holds even in the face of greater than usual risks to physicians’ own safety, health, or life.”11 The obligation to treat in emergencies has also become enshrined in law. In the US, EMTALA (Emergency Medical Treatment and Labor Act) requires health care providers to provide treatment to individuals in emergencies regardless of their ability to pay.
Emergency situations show that clinicians’ obligations extend beyond their current patients. In these situations, they have an obligation to make [End Page 164] someone their patient.12 This obligation represents a seemingly small, but in fact significant step away from the Strict Fiduciary View. For one thing, emergency situations demonstrate that clinicians can bear responsibilities to provide care even absent a formal clinician-patient relationship. This opens the door to considering whether there are other individuals for whom clinicians need to provide care.13 For another, emergency situations underscore the role-based nature of the duty to provide care. Two of the reasons that support the duty of easy rescue are present in emergency situations, the magnitude of help clinicians can provide and patients’ dependency on them to receive that help, but the third reason, the low cost of helping, is noticeably absent. Health care providers need to provide help in emergencies even when that help is difficult to provide and, moreover, comes with “greater than usual” risk to themselves, as the AMA Code of Medical Ethics states. In this way, health care providers shoulder a special obligation to help others. While we all carry a duty of easy rescue, health care providers must assist others even when it is not easy.
Two important role-based reasons underpin this special responsibility. First, health care providers agree to assume responsibilities to help others when they accept their care-giving role. Such a responsibility is written into professional codes of ethics and included in the oaths that are often spoken at white coat ceremonies. By involving an explicit commitment to the duties of a clinician and the symbolic donning of a white coat, this sort of ceremony emphasizes the distinctiveness of the role of a clinician.
The role comes with distinct privileges as well as responsibilities. Most notably, it is a privilege to be licensed to practice medicine. In a white coat ceremony, a representative of the medical community bestows this privilege on a student, and the privilege can also be revoked by the same community. From this privilege come many sorts of benefits, including financial security, social status, and, for many, the ability to fulfill a long-held dream. So, the second role-based reason comes from a notion of reciprocity. Namely, in exchange for the benefits of practicing medicine, clinicians must provide medical assistance to others, including both their existing patients and, as the emergency example illustrates, others in need. Furthermore, in addition to the personal benefits, the privilege to practice medicine is ethically significant from the perspective of a provider’s future patients in at least two ways. These patients, first, place their well-being in the hands of medical professionals and, second, grant them access to their bodies and protected health information. Both of these ideas can be put under the more general heading of the trust that is placed in health care [End Page 165] providers, and together they supply much of the rationale for thinking health care providers have a fiduciary duty to act in their patients’ best interests, a point I return to in section 3. But when placed alongside the personal benefits of the privilege of practicing medicine, they also provide a fuller picture of why health care providers have a more generalized responsibility to help those in need. The idea of reciprocity here is that in return for the personal benefits and privileged access granted licensed health care providers, they bear a special obligation to provide medical assistance to others. This special obligation is a role obligation because it stems from the privileges given to licensed health care providers.14
In the case of medical researchers, Richardson has argued that their special obligations are tied to the informed consent process rather than their professional role, since the informed consent process is when research participants waive their right to privacy (Richardson 2012a, 63). One might argue the same for clinicians. However, while the authorization to perform medical procedures on one patient in particular is given during the informed consent process, it is important to recognize that informed consent happens within the context of a broader trust that is placed in health care providers. A patient trusts that as a licensed member of the profession, the provider is capable of providing effective care and will also live up to professional standards for privacy. This sort of trust is starkly apparent in emergency situations in which informed consent cannot be given. So, we ought to understand the professional role itself as carrying reasons to provide help to others—reasons based on, first, providers’ voluntary acceptance of their professional role and, second, reciprocity for the benefits and privileges of being a licensed health care provider.
Although we can now see why the Strict Fiduciary View has a far too narrow view of clinicians’ role responsibilities, my argument for expanding the scope of clinicians’ concern beyond their patients has not yet included any reason for clinicians to have special concern for the family members of their patients. Indeed, the emergency example shows that clinicians could be obligated to help anyone in need. However, upon closer inspection, it also illustrates why clinicians often ought to have special consideration for their patients’ family members.
To see why, recall that the case for being required to help in emergencies rests on two of the same reasons that support the duty of easy rescue: the magnitude of benefit and the person’s dependency on a specific individual to provide that benefit. To determine who clinicians ought to help more generally, these same two factors should be employed. Priority should [End Page 166] be given to those who would benefit the most and cannot be helped by others. Once these standard principles for allocating help are accepted, it becomes clear that health care providers often ought to have special concern for their patients’ family members. In many cases, clinicians can provide significant benefits to their patients’ family members that they cannot receive from others, as illustrated by several examples I have already given: the practice of allowing members to gather at a brain-dead patient’s bedside, the hospital that reorganized its NICU to give families more privacy, and the phone call to a deceased child’s parents. The wealth of literature supporting family-centered care contains many others.
Additionally, the third factor from the duty of easy rescue—the cost of providing the help to oneself and others—both reinforces and shows the limit of clinicians’ obligation to care for family members. Although, as the emergency example shows, clinicians can be obligated to shoulder greater costs for helping than others can be, there is a cap to what the profession can require clinicians to bear. So, while clinicians are obligated to do more than just easy rescues, this obligation is still limited in its extent. The above quoted statement from the AMA Code of Medical Ethics mentions risks to clinicians’ “safety, health, or life,” but personal costs to clinicians can also take the form of intrusions on their personal lives. If a medical emergency occurs on the flight to a vacation in Bermuda, such an intrusion is justified, but clinicians do not need to offer the same amount of care to their fellow passengers they would offer their patients during their regular practice. It’s fine to refuse someone’s mid-flight, non-urgent request to examine a troublesome rash, for instance. When considering clinicians’ responsibilities for family members, we should be careful to not make what can be an already consuming occupation even more so.
That means that space for giving attention to family members’ needs must be created within the current parameters of clinicians’ jobs, and the concern, of course, then, is that creating such space would come at the expense of other patients. However, the Danish NICU example shows that is not necessarily the case. There, nurses found that a more family-centered arrangement of the NICU increased their workload, but they developed methods to ensure that care for other patients was not compromised (Beck et al. 2009, 95–97). Likewise, giving families some time to gather at the bedside of a brain-dead patient needn’t come at the expense of care for other patients so long as, among other things, a bed is available. Of course, this will not always be the case. My main aim in this article is to show that clinicians have ethical responsibilities to not just their patients, but their [End Page 167] patients’ family members too. Once that responsibility is acknowledged, clinicians and the institutions for which they work can determine to what extent they can meet it without compromising care for other patients. In some cases, they will not be able to do so, but establishing a responsibility for family members can also result in restructuring of practices—or, more dramatically, facilities themselves, as in the NICU example—to meet this responsibility.15
Despite the ways in which the cost consideration can attenuate a clinician’s obligation to care for family members, it also reinforces the obligation to provide such care. Given the cap that the cost consideration places on the amount and nature of care we can require of clinicians, the criteria for prioritizing which care they should provide are all the more important, and as we have seen, care for family members often ought to be prioritized under the criteria of magnitude of benefit and dependency of the individual.
In this section, we have seen that a clinician’s role responsibilities ought to be understood to be broader than the Strict Fiduciary View would have it. Due to their voluntary assumption of responsibilities to care for others and the benefits of their professional role, clinicians can have responsibilities to care for those who are not their patients, and given the magnitude of benefit to family members and the dependency they often have on clinicians to receive it, care for family members often ought to be prioritized. In the next section, I argue that the relationships family members have with clinicians provides an additional reason for taking family members’ needs into account.
2. RELATIONSHIP-BASED RESPONSIBILITES
While clinicians’ responsibilities are partially defined by their institutional roles, their particular relationships with patients and families can also be ethically significant. In some situations, such as when personnel changes are frequent and interactions are brief, clinicians may seem like interchangeable wearers of white coats to patients and family members. But in other situations, as when a patient and his family have shared harrowing and joyous moments with a clinician, the relationship with that particular clinician can become significant. Recall the quote from the mother who received a phone call from her daughter’s physician after her daughter died: “I think it was really important that he [the physician] did call . . . it just showed me that he cared and that [5-second silence] she just wasn’t a . . . just wasn’t a patient . . . that he treated and then she didn’t make it, so, [End Page 168] ‘Oh well’” (Back et al. 2009, 477). This parent needed to hear that the physician cared for her daughter not just as a patient but as an individual.
Her statement captures an important insight that is usually overlooked in discussions of professional responsibilities: these responsibilities aren’t just institutional; they’re also interpersonal. One bears them in virtue of not only one’s generic institutional role but also one’s particular relationship with a patient. And, as this example illustrates, a strong case can be made that, in some cases, these responsibilities can extend outward from a patient to providing support or assistance to her family members. In this section, I make that case. I discuss the kinds of relationships that can create ethical reasons to help family members and argue that duties of acknowledgment and non-abandonment can require clinicians to offer help to them.
As I have argued elsewhere, relationships gain ethical significance as they grow in length and intensity (Olson 2016).16 In my view, the ethical significance is due, in part, to the meaningfulness of such relationships. Meaningfulness, in this sense, refers to the non-instrumental significance things or people can have for individuals’ lives, apart from contributing to their happiness.17 The core of the mother’s appeal in the above quote is for the physician to see her daughter as important in herself, not simply a patient to be treated and then immediately forgotten. The idea of the meaningfulness of a relationship captures this sentiment well. Back et al. do not give more information about the nature of the interactions between the patient, her family, and the physician in this example, but given that the child died in the physician’s care, we can expect that the interactions were intensely emotional for at least the patient and her mother, and in many such cases, the relationship would have had an extended duration as well. We tend to think that both these factors give relationships meaning, and for good reason. Extended relationships take up a greater portion of our lives than shorter ones, and intense relationships shape our lives more than mundane ones. Not all subjective attributions of meaning are justified—a patient could, for instance, attribute an unjustified amount of significance to a brief interaction with a clinician—but clinicians should not dismiss the human side of their interactions with patients and their families, and attributing meaning to a relationship is a common and reasonable human response to having extended and intense interactions with another. That is to say, patients and their families do not just think such relationships have ethical significance; they actually do.
When one is in a meaningful relationship, one ought to act in ways that acknowledge its meaningfulness.18 The norms of how individuals should [End Page 169] interact when in meaningful relationships differ depending on time and place, but one common—and, perhaps, universal—practice is for the individuals to take a heightened interest in one another’s welfare. Health care providers are already bound to look after their patients’ welfare, but as the mother’s appeal indicates, acknowledging meaningfulness requires more than the provision of clinical care (at least under a narrow definition). It also includes showing concern for the patient as an individual, where that might mean expressing joy when a treatment plan works as hoped or offering consolation when it does not. But, more notably for my argument here, clinicians should also take a heightened interest in their patients’ families’ welfare when their relationships with them are meaningful. By being co-participants in, let’s say, a series of intense interactions spread over an extended period of time, the physician and the patient’s mother in this example have forged a meaningful connection that should not be forgotten by the physician as soon as the patient dies. The physician must certainly turn his attention to other patients and other families, but some degree of heightened interest in the mother’s welfare for at least a short period of time appropriately acknowledges the connection they have shared. As we can put it, the physician develops responsibilities to the patient’s mother that “unwind,” rather than instantly dissipate. The responsibility in question here is a duty to acknowledge their relationship’s meaning, a duty that can be met, in this case, by providing a follow-up phone call and, possibly, offering other forms of bereavement care. Since the relationship with the individual provider is what is meaningful in this case, the physician himself needs to take actions that acknowledge the relationship’s meaning, but these actions could also include referring the mother to other providers who can provide more bereavement care.19
While recognizing the importance of the particular relationships between clinicians and families, we should also remember that these relationships remain professional in nature. It would be inappropriate for a clinician to provide more comprehensive care for a family she especially likes or one with which she shares a long-standing connection outside work. Therefore, when discussing the meaningfulness of a clinician-family relationship, we ought to focus on the professional relationship, measured in terms of the duration and intensity of the interactions they have in the course of giving and receiving medical care.20 Doing so recognizes the importance of interpersonal relationships, while still treating all families fairly.
One might object that my position creates a perverse incentive: health care providers might try to avoid close relationships with patients’ family [End Page 170] members so as to avoid the concomitant responsibilities. Given that my aim is to draw greater attention to the interests and wishes of family members, this is an important objection to address. I offer three points in response. First, a relationship between a family member and a clinical provider need not be a one-to-one relationship. Especially when the members of a care team change frequently, it may be the relationship with the whole team that is meaningful for a family member. In this kind of case, the team is the unit that shares a meaningful relationship with the patient, so the whole team would bear responsibilities for the patient. So, one cannot avoid creating responsibilities by simply switching care providers. Second, other responsibilities clinicians have, such as acting in their patients’ best interests and treating both patients and their families with respect, rule out some other methods for avoiding close relationships. For instance, it would be unethical for a clinician to deliver bad news to a family member via memo rather than in person in order to avoid a more intense interaction with the family member. Finally, as I argued earlier, we should not make an already demanding occupation even more so. These responsibilities should be fit within the existing parameters of the job. I am arguing for a change to clinicians’ responsibilities, not an augmentation of them. The incentive to avoid responsibilities to family members, consequently, should not be great.
Jones, Contro, and Koch (2014) highlight a different interpersonal basis for ongoing responsibilities to family members: what they, following Quill and Cassel (1995),21 call a duty of non-abandonment. They state that “families often report feeling abandoned by their health care team once the death occurs, creating an additional burden of experiencing the simultaneous loss of their child and their physician whom they have come to rely on not only for the medical care, but also for comfort, continuity, and information” (Jones, Contro, and Koch 2014, S12, citing Back et al. 2009). The sense of abandonment is particularly important in this context since “[b]ereaved parents are at risk for anxiety, depression, suicidal ideation, prolonged grief, decreased quality of life, relationship struggles, and social decline” (Jones, Contro, and Koch 2014, S12, citing Hendrickson 2009). In short, bereaved parents are in a psychologically vulnerable state, and continued contact with a care team can prevent the compounding of their experience of loss and, moreover, help them in their grieving process.
Jones, Contro, and Koch’s above statements reveal several factors that combine to create a duty of non-abandonment. First, as I mentioned when [End Page 171] discussing role responsibilities, the magnitude of help that can be provided and the family’s dependence on the care team to receive it are both high. But an additional factor is at play here as well, as indicated by the statement that a family has “come to rely” on a physician. It’s not just that only the physician can provide the comfort, continuity, and information that the family seeks—the dependency criterion—but that this dependency is a result of the relationship that has developed between them. The situation here is different than the paradigmatic duty of easy rescue scenario wherein someone can save a stranger and no one else is around to do so. Instead, this is a situation of induced reliance. The family has come to rely on the physician because of the physician’s past actions—or, more accurately, their past interactions. When a person intentionally induces another to rely on her by, for instance, promising to meet the other person for lunch at a particular restaurant at noon, she incurs an ethical obligation to not leave the other person in the lurch. The situation in Jones, Contro, and Koch’s clinical cases, however, is more complex. In providing information and comfort to a family, a physician’s intention likely is not to cause the family to come to rely on her for such support. Yet, the family’s coming to rely on her is not an unexpected development; indeed, it should be expected. When reliance is an expected byproduct—not the aim—of one’s action(s), some have held that one still incurs an ethical obligation (Scanlon 1998, 300–301; MacCormick 1972, 66–68; Alonso 2009).22 If we add a reasonability qualifier as I did for meaningfulness—that is, the reliance must be reasonable—then I think we ought to accept this principle. When as a result of your interactions with another, that person comes to be vulnerable to how you treat him, then you bear some responsibility for his well-being. This responsibility does not persist indefinitely, but you should not simply abandon the other in a time of need. The earlier mentioned notion of a responsibility unwinding illustrates how the duty of non-abandonment functions. Health care providers can, and often must, end relationships with family members, but when families have reasonably come to rely on them, they should not do so abruptly. Provision of some bereavement care, for instance, can sometimes be required.
So, in addition to their role obligations, health care providers also have two types of relationship-based obligations to their patients’ families: a duty to acknowledge the meaningfulness of their relationships and a duty of non-abandonment. These obligations can require health care providers to offer some forms of care to family members. My arguments in the last two sections provide the complete picture of why the Strict Fiduciary [End Page 172] View is mistaken: clinicians have both role- and relationship-based ethical reasons for attending to the needs of individuals who are not their patients. However, as I argue in the next section, we should not go to the opposite extreme and place patients and family members’ wishes and needs on equal footing. The primary concern of health care providers should still be for their patients.
3. THE CASE FOR PRIORITIZING PATIENTS
As I indicated in the Introduction, one of the great accomplishments of the family-centered care movement has been to better illuminate how the wishes and needs of patients and their family members are intertwined. In many cases, including family members in care creates positive outcomes for patients. However, this is not always the case. Attending to family members’ needs may make it more difficult to provide optimal care for patients, and patients’ wishes may conflict with those of family members. For instance, consider a situation in which a terminally ill, mentally competent adult wishes to end painful, but life-prolonging treatment, while her family wants to continue.23 In such situations, clinicians are faced with the question of how to weigh competing interests and wishes. In an early paper that calls for the inclusion of family members in care, Hardwig (1990, 7) argues for a “presumption of equality” between the interests of patients and their families. For him, this is an issue of fairness.24 This position represents the opposite extreme from the Strict Fiduciary View: what I will call the Equal Consideration View.
However, there are several reasons to resist going to this other extreme, including patients’ greater vulnerability, the importance of patient autonomy, and the desirability of a division of labor among health care providers. In the previous section, I discussed how families’ relationships with clinicians make them vulnerable to clinicians’ actions and inactions. The same points carry over to patients. Like their families, patients come to rely on health care providers for “comfort, continuity, and information,” and their need for these things is typically greater. In other words, the same type of relationship-induced vulnerability exists for patients as for families, and patients typically experience it to a higher degree. But, of course, it’s not just patients’ relationships with their clinicians that place them in a vulnerable position. To state the obvious, patients’ physical and mental health are dependent upon the care they receive from their health care providers. A clinician’s choices about a patient’s treatment plan can have residual effects on the physical and mental health of the patient’s family [End Page 173] members, but it is safe to say that in most cases, the effect on a patient will be greater. The need to provide sensitive health information and be placed in situations that can compromise one’s bodily privacy adds an additional layer of vulnerability for patients that their family members largely do not share.25 Notably, clinicians intentionally induce this kind of vulnerability when they request sensitive health information and place patients in physically vulnerable positions, and patients grant such information and allow themselves to be placed in such positions with the expectation that their health will benefit from their doing so. The case for prioritizing patients, here, then, is not just because of their greater vulnerability. It is also to respect a patient’s wishes, an autonomy-based point.
Moreover, as illustrated by the case involving the patient who wishes to stop life-prolonging treatment, a patient’s ability to make decisions about her health—and in especially significant decisions like this one, the direction her life will take—often depends on the actions of health care providers. Patients are more likely than their family members to be in physically and/or mentally compromised states wherein their control over their life choices is significantly diminished. This, again, is a point about patient autonomy; however, it shows that clinicians need to not only respect patient autonomy but also recognize that it must often be protected. A patient’s voice can be lost in the din of a clinical setting, drowned out by others, including sometimes family members, who can better make their voices heard. The precarious state of their autonomy is another way in which patients are vulnerable. So, clinicians have both vulnerability and autonomy reasons for prioritizing patients, and sometimes these reasons are entwined.
It should be said that the above example also illustrates how the shape of a family member’s life can be tied to the care a patient does or, as the case may be, does not receive. When a loved one is fighting a serious illness, it becomes very clear that you are not the only one steering the direction of your life. However, even in such cases, we ought to accord patients ultimate say in what happens with their bodies. The right to bodily integrity, rightfully, is a bedrock principle of medical ethics. In short, my main point in this section can be summed up by stating that patients are placed in a position where they must trust their health care providers to have their wishes followed and their welfare protected. As noted before, the centrality of trust to the patient-clinician relationship is why clinicians have traditionally been held to possess fiduciary responsibilities for their patients.26 My argument here reaffirms the need for a fiduciary-type [End Page 174] responsibility for patients, even when practicing family-centered care.
Finally, positive consequences follow from a division of labor in which each person’s care providers focus on giving him or her optimal care. One of the most important benefits of this arrangement is the trust that patients can place in care providers to act in their best interest. If care providers were to take Hardwig’s presumption of equality to its logical conclusion and put the needs of all individuals, whether patient, family member, or otherwise, on equal footing, many patients likely would be wary of seeking treatment or disclosing sensitive information. It’s for good reason that hospitals take such strong measures to separate organ transplant from care teams, for instance. If one were to restrict the presumption of equality to just patients and their family members, the loss of trust would likely be less severe, but given the possibilities for differing wishes and interests of patients and their family members in even close-knit families, the loss would remain real nonetheless.
A second type of positive consequence of this arrangement is that individuals benefit, on the whole, from having care providers assigned to them in particular, rather than having their access to care depend on a connection to a family member.27 Even when one is being seen by multiple care providers in a hospital setting, the system of having officially designated patients carries value. It provides a way for the hospital’s care providers to focus the lion’s share of their limited time and energy on specific individuals within that facility, rather than needing to provide equal consideration to the health needs, even non-urgent ones, of all those who walk through their doors. By and large, this system results in better care for patients.
Relatedly, having designated patients helps avoid overwhelming any one care provider, something in the interests of both patients and care providers. By referring family members to other clinicians who have more availability, clinicians can help patients receive more timely care while making their workload more manageable. We ought to be concerned about fairness in the distribution of health care benefits, as Hardwig is, but his presumption of equality is not the way to achieve it.
Rather, the reasons enumerated here point toward a presumption of priority for patients: when the wishes and interests of a patient conflict with those of her family, clinicians ought to prioritize the patient.28 As a presumption, this general rule can be defeated in exceptional cases. For instance, the division of labor points given here are predicated on the idea that family members have separable needs—one health care provider can [End Page 175] treat one family member’s needs, and another can treat another’s. However, one of the insights motivating the family-centered care movement is that family members’ needs are interconnected. Often, this means that their interests align, but when they conflict, it may mean that a family member’s need cannot be met by simply going elsewhere. As mentioned earlier, the conflict may be that one course of treatment, while what the patients wants, is particularly onerous for a family member who serves as his primary care provider. In this kind of situation, a patient’s care itself is the source of the conflict; it is not a question of meeting two independent needs of a patient and a family member.
However, when a family member could reasonably be expected to have her need or wish met by a different health care provider, the division of labor points go through, and the presumption in favor of patients holds. This is the dependency criterion, discussed earlier. When a family member is not dependent on a particular health care provider to have a need or wish met, a patient’s needs and wishes ought to take priority. We should interpret dependency broadly, though, to take into account real-world facts about access to care. If a family member could have a need met by a different health care provider but is highly unlikely to actually be able to get such care, she is still dependent on the provider in question.
Yet even when family members are dependent on a provider to receive care, the points about patient vulnerability and autonomy still generally keep the presumption in favor of patients in place. Health care providers should deviate from this presumption only when, first, the risks and benefits to a family member’s welfare and autonomy significantly outweigh the risks and benefits to a patient and, second, acting in favor of the family member would cause no harm to the patient. For instance, consider a scenario in which a mentally competent patient expresses a strong preference for a treatment plan that would allow him to be discharged from an in-patient facility almost immediately but that would require significant round-the-clock care by his spouse. The spouse expresses her strong preference for a different treatment plan that would keep the patient in the in-patient facility for another few days but that would then result in significantly less burdensome care at home.29 Even if the patient’s care team is confident in the spouse’s ability and resolve to provide the more significant care if she had to do so, this is a case in which the presumption could be overturned. Looking at all the details of the situation, the care team may judge that, first, the additional burden on her significantly outweighs the limitation [End Page 176] on his autonomy and that, second, following her wishes would not cause him harm.30
By allowing the presumption to be overturned only when these two conditions are met, health care providers can avoid the need to continually weigh patient and family interests against each other. Instead, the presumption in favor of patients would function as what Raz (1990, 35–40) has called an “exclusionary reason.” Any options that are contrary to her patient’s interests or wishes would be excluded from a clinician’s reasoning unless the “significantly outweigh” and no harm conditions seemed to be met.31 Such a reasoning structure would help preserve the important fiduciary aspect of the patient-clinician relationship and avoid saddling clinicians with the burden of adjudicating between the merits of different claims to care, a burden that they are sometimes not well-positioned to discharge. Indeed, Hardwig’s justice-based argument shortchanges the reasons we have for wanting to take determinations of whether a medically indicated treatment should be given to a patient out of the hands of clinicians.32 Clinicians cannot avoid such determinations altogether, but a presumption of priority for patients considerably reduces their prevalence.
In the Introduction, I noted that the movement toward family-centered care has been particularly pronounced in pediatrics, so it is important to note the additional complexities involved in cases in which a family member is serving as a surrogate decision-maker. If a conflict seems to exist between the interests of the patient and the family member in such a case, my main argument remains the same: the clinician ought to prioritize the patient’s interests unless the two conditions I have discussed are met. However, what it means to prioritize a patient’s interests when a family member is the surrogate decision-maker is a complicated issue that can include the ethically fraught matter of how far surrogate decision-making authority should extend. When the conflict is between the wishes of the patient and the family member, we cannot assume the patient’s wishes generally should be prioritized, given the family member’s status as the surrogate decision-maker. A full treatment of this type of situation would need to address, inter alia, the issue of the ethical status of juvenile consent. I cannot adequately address these complex issues here, but they have been much discussed in the wider literature (AAP 2016; Navin and Wasserman 2017a; Navin and Wasserman 2017b).33 However, my argument in this article establishes a framework that ought to guide these ongoing [End Page 177] discussions. First, clinicians ought to take into account family members’ interests and wishes, and not just as they relate to a patient’s interests and wishes. But, second, their primary responsibility remains with the patient.
In sum, my arguments in the previous two sections showed the reasons for abandoning the Strict Fiduciary View, the idea that clinicians should act for their patients’ and only their patients’ interests. Here, we have seen several reasons for still prioritizing patients. I have thus shown the way between the two extremes of the Strict Fiduciary View and the Equal Consideration View: a position that maintains the fiduciary idea of acting in a patient’s best interest but broadens a clinician’s concern to take in others’ interests.
The recent focus on family-centered care has brought needed attention to the many ways in which both patients and their families benefit when families are given a more central place in medical care. My argument here shows why the standing families have gained should not be treated as conditional. Even when the instrumental benefits to patients are unclear or absent, family-centered care should still be practiced. The clinician-patient relationship is not the only ethically significant one, and a clinician’s role is not oriented solely to his or her patients. Yet in broadening our understanding of who matters in medicine, we should also not forget the reasons why patients still, generally, come first.
Nate W. Olson, PhD, is Assistant Professor of Philosophy and Associate Director of the Kegley Institute of Ethics at California State University, Bakersfield. His research focuses on ethical issues related to personal and professional relationships, especially in clinical practice and research.
Thanks to David Magnus, Christopher Meyers, Henry Richardson, Stephen Campagna-Pinto, Steven Gamboa, Debra Jackson, Jacquelyn Kegley, Maria Palaiologou, Senem Saner, two anonymous reviewers for this journal, and audiences at the Society for Ethics Across the Curriculum and Association for Practical and Professional Ethics annual conferences for helpful comments on previous versions of this paper. I am especially indebted to David Magnus for the initial idea to write on this topic and guidance as I began work on it.
1. As of 2016, the IOM is called the Health and Medicine Division of the National Academies of Sciences, Engineering, and Medicine.
4. Family-centered care is distinct from “family-oriented care,” in which all members of a family may be a provider’s patients, as in family practice or family therapy. (See an earlier AAP statement (2003, 1542) for more on this distinction.) However, both types of care are motivated by a shared insight: the wishes and interests of family members are intertwined. (See Lindemann Nelson and Lindemann Nelson 1995 for a thoughtful and sustained discussion of how family members’ wishes and interests are interconnected.) I focus on family-centered care in this article in order to explore the issue of whether providers have ethical responsibilities for individuals who are not their patients.
5. “family centred care is a way of caring for children and their families within health services which ensures that care is planned around the whole family, not just the individual child/person, and in which all the family members are recognized as care recipients” (Shields, Pratt, and Hunter 2006, 1318).
6. California Health and Safety Code § 1254.4
7. The California Health and Safety Code §1254.4 specifically mentions that “in determining what is reasonable, a hospital shall consider the needs of other patients and prospective patients in urgent need of care.”
8. I will use both “responsibility” and “obligation” to describe what clinicians owe family members. I believe the notion of a forward-looking moral responsibility (Richardson 1999) best captures the open-ended and forward-looking nature of what clinicians owe family members, so “responsibility” is my preferred term. However, I will also use “obligation” since it is the most commonly used term in the literature. By “obligation,” I mean something that an agent is required to fulfill. According to my usage, a responsibility is a type of obligation.
9. My argument that clinicians bear responsibilities for individuals outside their primary scope of concern has some structural similarities to arguments about ancillary care in research ethics, which concern researchers’ responsibilities to provide care for study participants that is ancillary to research objectives. (See, for instance, Olson 2016, Richardson 2012a, and Merritt 2011.) Just as researchers’ devotion to their research objectives can blind them to the health needs of their participants, clinicians’ commitment to their patients’ health can cause them to overlook the health needs of others. Like arguments in the ancillary care debate, my argument here calls for a broadening of perspective: for clinicians, both patients and families should be subjects of ethical concern. My argument in this paper engages with ideas from the ancillary care debate and brings them into the clinical context.
10. In order to move the debate forward on the duty of easy rescue, Rulli and Millum (2016, 261) call for more investigation of, first, institutional (as opposed to individual) duties of rescue and, second, role obligations. This paper answers the second part of their call. MacKay and Rulli (2017) and Garrett (2015) focus on the first part. As I mention in notes 15 and 19, the issue of whether institutions have responsibilities for patients’ family members is an important one that should be addressed in the future.
11. Opinion 1.1.2 of the AMA Code of Medical Ethics provides more detail on the conditions for when physicians can and cannot refuse to provide care. See also Principle VI, which unlike Opinion 1.1.2, names emergencies as the only factor limiting physicians’ freedom to choose whom to serve. Thanks to an anonymous reviewer for prompting me to consider these other aspects of the code.
12. Thanks to Christopher Meyers and Henry Richardson for pressing me on this point.
13. Physicians’ responsibilities to promote immunization represent another example. For some immunizations, the primary expected benefit is to individuals other than the one receiving the vaccine. Rubella provides an interesting example for the issue of family-centered care. Rubella is most dangerous for fetuses (CDC 2017), so the primary beneficiary of a child’s rubella vaccine may be his or her future sibling. Thanks to an anonymous reviewer for this example.
15. Although my focus in this paper is on clinicians’ responsibilities for family members, this example demonstrates the importance of considering institutional responsibilities. I cannot satisfactorily address this issue here, but to start I would note that at least my reciprocity argument seems to carry over to institutions. Like individual clinicians, medical institutions garner benefits from being licensed to deliver medical care. And so, like individual clinicians, institutions too have a special obligation to provide medical assistance to those who need it.
19. Like the NICU example, this example shows the importance of considering both individual and institutional responsibilities to family members. If the institution itself has a responsibility to this mother, then it would carry a responsibility to ensure she receives some sort of bereavement care through the appropriate providers at the institution.
22. Notably, all these authors focus on cases of reliance where an individual performs an action because she expects that another will perform some other action. The example that has been my focus in this section is not such a case. The daughter’s mother likely would not have changed her actions—or not changed them significantly—had she known that the care team would not contact her after her daughter died. Indeed, she may not have expected that they would provide any form of bereavement care. The ethical situation, therefore, is quite different. The care team’s obligation to provide bereavement care cannot be grounded in the family’s expectations (contrast Scanlon). As my argument in the main text suggests, I think a family’s vulnerability is a stronger basis for this sort of reliance-based obligation than their expectations.
23. Thanks to David Magnus for prompting me to consider to this sort of example.
25. See Richardson 2012a and 2012b for detailed treatments of how individuals’ interactions with each other can involve waivers of privacy rights. Richardson’s theory of ancillary care obligations is built around the idea that such obligations arise when one party waives his right to privacy and so enters a “moral entanglement” with another. I take a broader view of the source of such obligations.
27. This situation differs from one in which members of a family all see a common family practitioner.
30. The in-patient facility would still be legally obligated to follow the patient’s wishes, but they could, for instance, slow-walk the discharge process.
31. In Raz’s terminology, the scope of the exclusionary reason would be restricted to considerations that do not significantly disadvantage family members visà-vis the benefits to the patient (1990, 46).
32. The three reasons given here (clinicians’ special obligations, epistemological limitations, and restricted authority) parallel the three examples Raz uses to illustrate exclusionary reasons (1990, 37–39).