The Role of Physicians in the Allocation of Health Care: Is Some Justice Better than None?
Physicians’ advocacy obligations are best understood as going beyond advocacy on behalf of individual patients, which I call the “individualistic view,” to include advocacy for intelligent research-based allocation schemes that promote good outcomes and cost-effective care for all patients, which I call the “systemic view.” This systemic view includes moving beyond self-interest to promote less-wasteful and more cost-conscious allocation decisions and the setting of priorities at all levels to expand health care access. It includes physician involvement in discussions with patients in the context of clinical care, involvement in the formulation and administration of benefit structures and other allocation policies, and, finally, involvement in promoting public dialogue about health care priorities. This involvement is based on a concept of a deliberative process that can result in “just enough” decisions within systems for the preservation and promotion of health care and other societal goods.
Physicians traditionally have been given role-specific obligations to promote the well-being of their individual patients, one patient at a time. They are not expected to be concerned with how health care is best allocated between patients, or with how health-care allocations compare to other social goods and services. The assumption seems to be that our society’s health-care allocation should be the cumulative result of individual clinical decisions made on behalf of individual patients. In this view, physicians are the gatekeepers of health-care allocation, because they are responsible for making clinical decisions about prescriptions and admissions to health-care institutions and other services. But they should be only the kind of gatekeepers that keep the gates wide open for their patients, in the face of the vagaries of bureaucracies and misguided [End Page 1] policies. Physicians should be advocates for the needs of their individual patients and should resist pressures to be gatekeepers who deny access to some health-care products and services. Physicians should not be involved in what is known as “bedside or clinical rationing or allocation.” To the extent that clinical rationing or allocation is necessary at all, it should be done by others and imposed on physicians and their patients from outside the physician–patient relationship (Pellegrino 1986; Rosoff 2017). Contemporary criticisms of this view tend to appeal to clinical realities to argue that, unfortunately, physicians cannot avoid making trade-off decisions between patients, given the structural constraints of our health-care system (Hall and Berenson 1998). In this paper I will extend this critique to argue that those realities reflect an important moral truth that flies in the face of the traditional approach. I will argue that physicians’ advocacy obligations are best understood as going beyond advocacy on behalf of individual patients, which I call the “individualistic view,” to include advocacy for intelligent research-based allocation schemes that promote good outcomes and cost-effective care for all patients, which I call the “systemic view.”
The paper is divided into three sections. The first section is a critique of two arguments against the systemic view I endorse: 1) that physician involvement in rationing or allocation decisions is not necessary and 2) that, in any case, our society does not have a coherent sense of justice to guide such allocation decisions. In fact, as others have pointed out, physician involvement in allocation decisions is inevitable, and the fact that there are competing approaches to distributive justice is no more problematic in health-care settings than in other social policy domains. It just means that it is necessary to turn to a fair process orientation that emphasizes the processes and procedures for how allocation decisions are made, rather than a fair final distribution orientation that emphasizes whether or not the actual allocation outcomes are just.
The second section addresses a third argument for the individualistic view: that the United States currently lacks a sufficiently closed health-care system that would allow for fair trade-offs where savings from some services that are not provided could be better allocated for other services, making it impossible to be just in such an unjust system. There are numerous health-care systems in the US, most at the local level (e.g., health-care institutions like hospitals or systems of hospitals, insurance providers, or state governments, and Medicaid), but some at the national level (e.g., Medicare, the Veterans Administration, the Indian Health Service). Some [End Page 2] may be more closed than others and better equipped to redistribute cost savings within their own system. But certainly such redistribution does not help others who are in other systems or outside of any system. However, Leonard Fleck’s notion of “just enough” shows the practical wisdom of taking an incremental approach to such imperfections, to improve fairness even in the absence of universal health-care policies (Fleck 2009).
The third and final section addresses the individualists’ fallback argument: that even if there were sufficiently closed health-care systems to permit the redistribution of resources, physicians should not be involved in making such allocation decisions because of the necessary fiduciary relationship between physicians and their individual patients that relies on the trust patients put in physicians to privilege their individual well-being. I will argue that physicians have to be involved at the systemic level as a necessary part of being an effective advocate for individual patients. The section concludes with a discussion of the implications for clinical practice and identifies various ways in which physicians can balance their obligations to include obligations of justice.
CRITIQUE OF VIEW THAT CLINICAL ALLOCATION/RATIONING IS UNNECESSARY AND UNJUST
The first argument against physician involvement in allocation decisions is that there is no need for rationing that would deny some patients access to beneficial services. This is because there is enough “waste” in any given system to cover health care services for everyone in the system who needs them, so the only need is to increase the efficient use of existing resources. There were earlier permutations of this argument in the 1990s (Relman 1990; Califano 1992), and more recent discussions from 2012 of a turn from an ethics of rationing to an ethics of waste avoidance (Brody 2012; Berwick and Hackbarth 2012). There is also important and appropriate debate about what is meant by the term “rationing” in the first place, since the “R word” conjures up different images for different people and “the R word’s power to stop conversation reflects the popular belief that cost should be no object at the bedside” (Bloche 2012). Some definitions of rationing require the rationing to be explicit. Other definitions require absolute scarcity or rationing of only necessary services (Ubel and Goold 1998). Ubel and Goold argue for a broad definition of rationing to include “any explicit or implicit measures that allow people to go without beneficial health care services” (1998). They note that one important advantage of the broad definition is that it highlights the frequency with [End Page 3] which patients already go without beneficial services because of the costs and that makes for a better starting point for discussions. The AMA’s Ethical Force Project prefers the term “setting limits” because “rationing” implies that the only way to set limits is to predetermine each person’s fair share of resources—their ‘ration’ of health care—and they prefer to outline specific expectations for an ethical process, not a predetermined result (Levine et al. 2007). As Ubel and Goold point out, the use of the word “rationing” can enable some to deny that it is already happening in the form of implicit rationing by ability to pay and that has implications for fairness. Even though there are political and psychological advantages to keeping rationing decisions implicit and invisible, this is, as Calabresi and Bobbit pointed out as early as 1978, fundamentally dishonest and can conceal real injustices from both recognition and correction (1978; Fleck 2009, 10).
For a wider interpretation of “rationing” to be just it is necessary to establish allocation criteria that can accommodate the less explicit structural determinants of allocation (Calabresi and Bobbitt 1978; Callahan 1987; Churchill 1987). We have only limited resources to meet virtually unlimited health care needs and wants. Unlimited spending in health care puts at risk spending for other social goods (Fleck 2009). Unlimited spending on health care by some also puts at risk the availability of health care for others. Some people are unable to access necessary care due to cost (Kaiser Family Foundation 2012, 3). This scarcity is not necessarily a material one of actual hospital beds or therapies like drugs, although it can be (Mariner 1995, 1440). This scarcity is primarily a financial one. As taxpayers and insurance premium payers, we want health care costs to be controlled. Cost containment is increasingly challenging in this era of increasingly expensive medical innovations and technology (Callahan 1990). Take the rising cost of prescription drugs, for example. Although comprising only 10% of total health expenditures, spending on prescription drugs is receiving considerable attention because of its rapid growth (114% from 2000 to 2010, Kaiser Family Foundation 2012, 1) and the fact that costs are rising without being linked to improved health outcomes that expenditures are meant to promote. The United States spends more on health care than most other nations, and has some of the lowest health outcomes (Kaiser Family Foundation 2012, 7). This combination of rising costs, limited access, and poor health outcomes leads many to call for a more just setting of priorities and more just allocation schemes (Callahan 1987; Daniels 2001; Fleck 2009) But as Fleck points [End Page 4] out in Just Caring, establishing a more just allocation scheme is made more difficult when there is no widely shared conception of health care justice in our society (2009, 15).
Competing Conceptions of Justice
As the advocates for the individualistic view point out, there is considerable disagreement in our society about how best to understand justice. Without guidance on the societal level about just allocation decisions, what is to guide physician judgment on the local level? There are numerous contenders for theories of justice including utilitarian, libertarian, social contract, egalitarian, and virtue based. Each has its benefits and challenges. Michael Sandel summarizes these disagreements when he says that our society’s concept of justice “revolves around three ideas: maximizing welfare, respecting freedom, and promoting virtue” (2009, 6).
To ask whether a society is just is to ask how it distributes the things we prize—income and wealth, duties and rights, powers and opportunities, offices and honors. A just society distributes these goods in the right way; it gives each person his or her due. The hard questions begin when we ask what people are due, and why.(2009, 19)
To understand the implications of this pluralism, it is important to discuss and illustrate the tensions among these three ideas for health-care allocation.
One approach to answering Sandel’s “hard questions” would be to prioritize the first idea he implicates in our multidimensional concept of justice, and argue that what people are due is what maximizes the general welfare—a utilitarian form of justice. Health care is among the goods that promote opportunity or are a necessary part of a person’s ability to make a living. Being healthy is an important part of not only personal well-being, but public health is also an important part of maximizing the overall well-being of a population.
Sandel points out that the most glaring weakness of simple utilitarian concepts of justice that promote overall welfare is that they fail to respect individual rights (2009, 37). Utilitarian concepts of justice also depend on being able to measure, aggregate, and calculate happiness or well-being. But not everyone values their health or health care equally. Sandel considers the application of a typical utilitarian mechanism for comparing value—namely—cost-benefit analysis. “Cost-benefit analysis tries to bring [End Page 5] rationality and rigor to complex social choices by translating all costs and benefits into monetary terms—and then comparing them” (2009, 41). Another example of a utilitarian based calculation would include using Quality Adjusted Life Years (QALYs) in an attempt to refine concepts of benefit. Sandel asks whether a utilitarian concept of justice can give adequate weight to human dignity and individual rights or equality of opportunity. How compelling is the argument that it reduces everything of moral importance to a single scale of pleasure and pain (2009, 48)? For example, narrowly defined cost-benefit analyses of treatments for people with disabilities have been shown to underestimate the benefits and overestimate the burdens in the calculation of net benefits and costs (Neumann and Greenberg 2009). Members of the disabilities community would argue that a simple utilitarian calculus would disenfranchise its members further by denying them the equality they are due as moral agents (Neumann and Greenberg 2009). In that light, Fleck supports the use of utilitarian tools such as QALYs and cost-benefit analysis, but as only one kind of resource for democratic deliberations (as defined below) that give equal voice to the people with disabilities (2009, 58–62).
Utilitarian concepts of justice have some intuitive appeal and can be useful in some ways, but they also present challenges, particularly regarding individual rights. Promotion of individual rights leads to a conception of justice based on the protection of freedom or liberty, the second idea described by Sandel. Libertarian theories of justice would require these three things of a state: No paternalism; No morals legislation; and No redistribution of income or wealth (Sandel 2009, 60). For example, H. Tristram Engelhardt defends the libertarian view that essentially restricts the role of the state to police powers in preventing individuals from using violence and fraud against one another, to protect persons and their property from unconsented interference by others (1996). For Engelhardt, there are no positive rights or welfare rights to anything that must be provided by the state (Fleck 2009, 125). On this theory, providing health care to others could be regarded as an act of charity, but never a requirement of justice. Robert Nozick agrees, and defends libertarian principles as the best foundation for a theory of distributive justice (1974). His view is that efforts to maximize a certain level of health care for all are really unjust in that they redistribute wealth that already belongs to someone. Engelhardt and Nozick argue that health care is a private good since different individuals value health and health care quite differently. Additionally, libertarian proponents like Sade would argue that the [End Page 6] expertise and skills of physicians are their personal property and cannot be “conscripted” for public use (1971). In a capitalist economic structure like in the US, the market is viewed as the best mechanism to distribute private goods and services that people value differently. As a result, these theorists turn to the privatization of health care as a best way to insure a fair allocation of health care.
Fleck disagrees: “I reject the view that health care can be distributed in a morally acceptable manner through some combination of individual ability to pay and charitable impulse” (2009, 12). Fleck argues that market proponents falsely claim that markets are impersonal and impartial and thus the appropriate morally neutral distributional instruments. Markets reflect, rather than correct, broader societal inequities and discriminations. Moreover, they fail to distinguish between health desires (tummy tucks, whole-body CT scans each year) and health needs (cancer chemotherapy). “In short, markets in health care fail to control escalating health care costs, fail to establish morally defensible priorities for health spending or priority setting, tend to be entirely unresponsive to meeting the health needs of the financially and medically least well-off, and tend to generate wasteful and inequitable inefficiencies of their own” (Fleck 2009, 148). As Sandel argues, we have drifted from having a market economy to being a market society where some of the good things in life (health, education, family life, nature, art, and civic duties) are corrupted or degraded if turned into commodities (2012). Pellegrino further argues that health care is not a commodity and treating it as such is deleterious to the ethics of patient care and that health is a human good that a good society has an obligation to protect from the market ethos” (1999).
Moreover, the protection of freedom is not only associated with libertarian concepts of justice and their accompanying arguments against utilitarianism or rights-based approaches to health-care justice. Liberty is just one dimension of the larger principle of respect for persons, which can also undergird so-called social contract theorists like John Rawls and bring the element of equality into the discussion of justice. In Rawls’s A Theory of Justice, he argues that the best way to think about justice is to ask what principles we would agree to in an initial situation of equality, if we did not already know the resources available to us (1971). Rawls outlines two principles of justice that would follow from this original position of equality. First, we would not choose utilitarianism. Because we would be behind “a veil of ignorance,” not knowing where we would end up in society, we would want to pursue our ends and be treated with [End Page 7] respect. We would not want to be oppressed, even if it adds to the general welfare. We would want a principle of equal basic liberties for all citizens. Second, we would tolerate certain inequalities, but only if they benefit the least well-off. This “difference principle” means that we would allow only those social and economic inequalities that would work to the benefit of the least advantaged in society.
Norman Daniels extends Rawls’s approach to “justice as fairness” explicitly to health-care distribution and fair equality of opportunity (2001). Daniels argues that health care should be privileged among social goods because
Preventing and treating disease and disability with effective healthcare services (construed broadly to include public health and environmental measures, as well as personal medical services) derives from the way in which protecting normal functioning contributes to protecting opportunity.(2001)
He also argues that justice requires that all societies meet health-care needs fairly under reasonable resource constraints.
I mention Daniels in particular because he is the first modern philosopher to move beyond trying to establish a unified theory of justice and describe the need for fair processes. Since there is no consensus on principled solutions to allocation problems, we need a fair process that meets several constraints referred to as “accountability for reasonableness” (Daniels 2001). Daniels and Sabin outline the following four conditions for just allocation schemes: 1) Publicity Condition—decisions and their rationale must be publically accessible; 2) Relevance Condition—decisions should aim to provide a reasonable construal of how the organization or society should provide “value for money” in meeting the varied health needs of a defined population; 3) Appeals Condition—there is a mechanism for challenge and dispute resolution; and 4) Enforcement Condition—there is either voluntary or public regulation of the process to ensure the first three conditions (Daniels and Sabin 1997). Daniels’s approach lays the basis for other process approaches to justice that will be described later.
The final element in Sandel’s pluralistic view of our society’s concept of justice is virtue. He points out that modern theories of justice try to separate questions of fairness and rights from arguments about honor, virtue, and moral dessert. But Aristotle does not think justice can be neutral in this way (Sandel 2009, 187). For Aristotle, the purpose of politics or government is not to set up a framework of rights that is neutral among ends. It is to form good citizens and to formulate good character. Sandel is talking [End Page 8] about both the virtues that constitute a good society and those that belong to individual character. For Aristotle, justice means giving people what they deserve, giving each person his or her due. Many modern theorists would argue that as human beings of equal worth, every person is entitled to a fair share of health care. But how is a “fair share” to be determined?
Modern virtue theorists might approach social justice and health care from a different perspective other than the appropriate ends of a good life. Martha Nussbaum writes about Amartya Sen and the concept of capabilities as fundamental entitlements (2003). She argues that an adequate theory of social justice rests on claims about specific entitlements that are to some extent independent of the preferences that people happen to have (2003, 34). Fundamental rights are best understood in terms of capabilities—best thought of as secured to people only when the relevant capabilities to function are present. Nussbaum argues against a “neoliberalism” that views rights in negative terms. The capabilities approach emphasizes the central importance of what people are actually able to do and to be. She writes,
The basic idea of my version of the capabilities approach, in Women and Human Development, is that we begin with a conception of the dignity of the human being, and of a life that is worthy of that dignity—a life that has available in it “truly human functioning” in the sense described by Marx.(2003, 40)
She outlines ten central human capabilities: 1) Life; 2) Bodily Health; 3) Bodily Integrity; 4) Senses, Imagination, and Thought; 5) Emotions; 6) Practical Reason; 7) Affiliation; 8) Other Species; 9) Play; and 10) Control Over One’s Environment (2003, 41–42). She would diverge from the social contract tradition of Rawls by arguing that Rawls’s theory would be better able to give an account of the relevant social equalities and inequalities if the list of primary goods were formulated as a list of capabilities and not things (2003, 51). Perhaps the most important critique of contract theory is that it imagines contracting parties as rough equals, none able to dominate other, and none asymmetrically dependent upon others. But life is not like that.
Real people begin their lives as helpless infants, and remain in a state of extreme, asymmetrical dependency, both physical and mental, for anywhere from ten to twenty years. At the other end of life, those who are lucky enough to live on into old age are likely to encounter another period of extreme [End Page 9] dependency, either physical or mental or both, which may itself continue in some form for as much as twenty years.(2003, 51)
This discussion of a virtue basis for justice has a direct bearing on the professional character and obligations of physicians as healers who strive to promote the flourishing of their patients. In some sense, we have come full circle back to discussing well-being.
Modern virtue-based theories of justice, as proposed by Nussbaum and Sen, have been criticized for focusing on abilities and opportunities as requisites for well-being instead of focusing on well-being itself (Powers and Faden 2006). Madison Powers and Ruth Faden begin with the basic premise that “social justice is concerned with human well-being.” They identify six core dimensions of well-being: 1) Health; 2) Personal security; 3) Reasoning; 4) Respect; 5) Attachment; and 6) Self-determination (2006). Thomas Beauchamp and James Childress describe their views as arguing that, “Being healthy, being secure, and being respected are desirable states of being, not merely capabilities or functionings.” For example, we want not merely the capability to be well-nourished, but to be well-nourished. “Justice is concerned with the achievement of well-being, not merely the capabilities to pursue it” (2013). Justice would consist of securing a sufficient level of each dimension for each person. They see the basic principle and focus as egalitarian justice, not merely beneficence or social utility.
Each of these competing conceptions of justice described by Sandel and others has some relevance for health-care allocation, and philosophers combine insights from many into new theories. Some try to reconcile these differences by looking for common ground between them. For example, Beauchamp and Childress summarize their survey of theories of justice by pointing to the concept of a decent minimum of health care as a possible compromise because it incorporates some moral premises that most theories stress. It guarantees basic health care for all on the premise of equal access while allowing for additional access by individual choice. An egalitarian could see the opportunity to use an equal opportunity principle. A utilitarian could see that it minimizes public dissatisfaction and maximizes social utility and permits allocation decisions based on cost-benefit analyses. Supporters of a capabilities theory or a well-being theory could see that it increases the likelihood of many to afford better quality of care and to achieve better states of health. Libertarians may dislike the outcomes approach but could find substantial opportunity for the free market since one tier of health care is left completely to individual [End Page 10] choice (Beauchamp and Childress 2013). “A health care system that finds pockets of support from each of these accounts could also turn out to be the fairest approach to democratic reform of the system. We do not now have—and are not likely ever to have—a single viable theory of justice” (2013, 273). But as they point out, this approach to compromise depends on having a process for “democratic reform of the system.” They do not unpack that requirement, but others like Daniels and Fleck have gone further down that road by developing a “process” theory of justice.
Although never citing Sandel’s work, another contemporary philosopher, Leonard Fleck, describes the many competing concepts of justice in his 2009 work, Just Caring, and, like Sandel, does not end up supporting a particular theory of justice (2009, 15). Fleck argues that none of these accounts of health-care justice is really adequate by itself. His take-home message is that “our theories of justice have their limits; some problems of health care justice are too complex, too ambiguous, or too open to reasonable but conflicting responses for any one theory to yield the clearly rationally dominant response” (2009, 140).
Instead of applying one theory of justice, Fleck argues for what he calls democratic deliberation as a process that can result in fair allocation. Fleck argues that fair allocation decisions require a system of rational democratic deliberation that is impartial, a conversation among undominated equals, honest and transparent, mutually respectful, and committed to giving good reasons as the primary mechanism for moving the conversation toward resolution (2009, 140). Fleck claims that both David Eddy (Eddy 1996) and Norman Daniels (Daniels and Sabin 1998):
have identified the essential moral insight that makes rational democratic deliberation fair and feasible. Any fair approach must be a product of patient choice, freely and rationally self-imposed, not simply at an instant of time but over the course of a life, and not simply from the point of view of my actual health-needy self.(Fleck 2009, 152)
Similarly, Daniels argues in his 1985 book, Just Heath Care, that allocation processes should be guided by rules: that decision-making processes must be transparent with the inclusion of key stakeholders, based on good data, with conflicts of interest managed, and open for challenge or review. Similar ethical and practical guidance on fair coverage decisions has been made by the multi-stakeholder Ethical Force Program (Levine et al. 2007). [End Page 11] Levine et al. outline an ethical process for setting limits and ensuring fairness in coverage decisions that includes the following expectations for the processes of designing and administering health benefits: Transparent to those affected, Participatory by stakeholders, Equitable and Consistent (similar decisions under similar circumstances), Sensitive to Value in the weighing of net health outcomes and resources required to achieve them, and Compassionate in the flexibility to be responsive to individual values and priorities and attentive to the most vulnerable individuals and those with critical needs. The Institute of Medicine, in its report on Essential Health Benefits under the Affordable Care Act, includes many of the above criteria to guide methods for defining and updating Essential Health Benefits with a few additions: Transparent, Participatory, Equitable and Consistent, Sensitive to Value, Responsive to New Information, Attentive to Stewardship, Encouraging of Innovation, and Data-Driven (National Academy of Sciences 2011).
Our society has competing conceptions of justice, making determinations of just allocations problematic. How do we justify claims that the status quo rationing is unjust? I agree with authors who have turned to a more practical and process orientation for determining just allocation that tries to include and balance insights from various conceptions of justice. Although these processes can, and perhaps ultimately should, be applied on a national level to set priorities for everyone in our society, it is more likely that we can and should apply them at a more local level to begin. As I will argue, physicians are a necessary part of these localized deliberative practices. Because local allocation decisions are likely to fall prey to accusations of injustice between systems, that is why an argument for a concept of “just enough” is necessary and is included in the following section.
CRITIQUE OF THE VIEW THAT LOCAL ALLOCATION SCHEMES ARE IMPOSSIBLE AND UNJUST
Even though our society has very different concepts of justice, evidence suggests that more people, and perhaps most people, regard health care as an important good that should be available to all in our society (Kaiser Family Foundation 2012). Access to health care is an important part of current political debates and there is an increased interest in expanding access to health care. The recent call for “Medicare for All” is one example (www.mforall.org). But universal access to health care is by no means [End Page 12] assured and there is no reasonable likelihood that it will happen soon. The first argument in favor of local allocation is that it is necessary for the foreseeable future. To simply ignore local-level allocation questions is to fail to do the good that is possible in the pursuit of some unattainable perfectly just allocation.
The second argument is that there are many sufficiently closed systems in which decisions to eliminate waste, be more cost effective, and set priorities will reliably improve access for people in those systems. It is necessary and important to also assume that these measures of eliminating waste, being more cost effective, and establishing priorities are primarily in service of increased access and not only other things like increased profits or executive compensation. What is learned in the process of pursuing justice in local resource allocation decisions should be useful in framing and advancing larger debates about justice nationally and internationally in the future.
What’s Just Enough?
Fleck argues in Just Caring for a process of rational democratic deliberation in addressing a broad range of problems of health-care justice. Echoing Daniels, he outlines three core moral considerations that are brought together in this process. The first is that for a rationing decision to be fair, it must be rationally justified to those who are most likely to be affected. Secondly, the deliberative process must be public, visible, and transparent. Thirdly, and perhaps most importantly, is the requirement that all rationing decisions be self-imposed. The moral weight of the justification rests on this very important feature that members of a certain community have agreed to the rationing or allocation scheme (2009, 15). Fleck reaches back to John Rawls’s distinction between an ideal theory of justice intended as a critical tool for assessing the basic structures of society and the non-ideal theories that become necessary in practice. He quotes Rawls,
In practice, we must usually choose between several unjust, or second best, arrangements; and then we look to non-ideal theory to find the least unjust scheme. Sometimes this scheme will include measures and policies that a perfectly just system would reject.(Fleck 2009, 113)
Fleck has referred to this as the domain of “interstitial justice” as contrasted with the basic structures of society (1987). “Interstitial justice refers to all the nooks and crannies, all the intricate social spaces, in which mundane problems of micro-justice occur” (Fleck 2009, 113). He illustrates the use of the non-ideal justice theory in considering the [End Page 13] problem of the allocation of ICU beds where decisions need to be made that might constitute “an incremental improvement in the justness of social practices.” He argues that having a democratic deliberative process among future possible ICU patients that generates rationing criteria for those limited ICU beds is “just enough” and to be preferred over the judgments of individual ICU directors (2009, 119).
Although Fleck seems to link the use of his non-ideal theories back to ideal theories of justice, non-ideal theories of justice often take as their starting point the rejection of ideal principles. There is a robust social and political philosophical literature that argues that theories of perfect justice are not helpful as a starting point for alleviating social inequalities. John Arras argues that the methodological approach that is associated with ideal theories is an impossible exercise (2010). Others have argued that ideal theories cannot address actual injustices (Schmidtz 2011). But non-ideal theories are not without their critics. Tessman argues that although ideal theorizing should be rejected, non-ideal theories may result in conditions that are the best possible because normative expectations are lowered and adapted, and still they may be morally unacceptable (2010). One of her central claims is that moral wrongdoing and failure may not always be avoidable.
Fleck also speaks to the issue of physicians and rationing. He argues that a deliberative process would not matter, practically speaking, unless there were also professional habits and social policies that faithfully translate these allocation decisions into practice (2009, 31).
Following Daniels and Fleck, I have argued that given the variegated nature of our health-care system and our competing views of justice, local allocation and “just enough” processes will be necessary to achieve the goals of efforts like the AMA Ethical Force Program and the ACA benefit structure guidelines described above. What remains to be discussed are why the inclusion of physicians is necessary and the features of an intelligent allocation scheme in health care.
PHYSICIANS AS ADVOCATES AT THE INDIVIDUALISTIC AND SYSTEMIC LEVELS / IMPLICATIONS FOR CLINICAL PRACTICE
In allocation decisions, there is an inescapable need for medical judgment because medical care is always delivered to individual patients. [End Page 14]
Physicians and other health professionals are more likely to protect their own moral integrity as loyal advocates of their patients’ just interests and as fair rationers of societal resources if they make their rationing decisions within the framework of a fixed global budget and a system of health care priorities that have been freely agreed to by all (representatively) who are part of that health care system.(Fleck 2009, 401)
What is the basis for the moral integrity of physicians? The traditional view of physicians as advocates for their patients one at a time is well-articulated in the work of Pellegrino, who argues that the central ethical commitment of the physician is to the patient’s good (Pellegrino and Thomasma 1988). He argues that there are three ways in which a physician can function as a gatekeeper. One is morally mandatory (the responsibility to practice rational medicine and use only those diagnostic and therapeutic modalities that are beneficial and effective for the patient). One is morally questionable (negative gatekeeping to limit the use of health resources, which creates a conflict of interest between the physician’s primary responsibility as advocate of the patient and a more circumscribed responsibility as guardian of society’s resources). Pellegrino writes that “under certain carefully defined conditions of economic necessity and moral monitoring, a negative gatekeeping role might be morally justified” (Pellegrino 1986, 44). The third form of gatekeeping is morally indefensible (positive gatekeeping in which the physician encourages the use of health-care services for personal or corporate profit) (Pellegrino 1986, 44). Pellegrino and others argue that physicians have primary obligations to their individual patients, one at a time, and that commitment has priority over some larger obligation to community or public health or justice (Pellegrino 1986; Tilburt and Cassel 2013; Rosoff 2017).
Despite the professional popularity of the individualistic view, it is a relatively new feature of medical morality (Wynia 2005). Historically, physicians were viewed as having such larger public health obligations, and many modern professional Codes of Ethics, including those of the American Medical Association (AMA) and the American College of Physicians (ACP), include obligations for public health and also obligations of justice (AMA Council on Ethical and Judicial Affairs 2016; Sulmasy and Bledsoe 2019). The newest version of the ACP Ethics Manual includes a discussion of Resource Allocation. It states, “Medical care is delivered within social and institutional systems that must take overall health care resources into account. Increasingly, decisions about resource allocations challenge the physician’s primary role as patient advocate. This advocacy role has [End Page 15] always had limits” (Sulmasy and Bledsoe 2019, S19). The manual further states that the physician should practice responsibly using resources in an appropriate and efficient manner and that resource allocation decisions are best made at the policy level “when all stakeholders consider and deliberate about the resources that exist, to what extent they are limited, costs, and how to balance these factors for equitable distribution” (S19). Similarly, a recent empirical study of attitudes towards cost-conscious care among US physicians and medical students concludes that
Most US practicing physicians and medical students believe physicians have a responsibility to reduce health-care costs and limit unnecessary testing while also protecting and promoting the well-being of individual patients. However, medical students are more accepting of cost-conscious care, even when compared to physicians most proximate to them in age. This may be due to the combined effects of generational differences, new medical school curricula, students’ relative inexperience providing cost-conscious care within complex health-care systems, and the rapidly evolving U.S. health-care environment. Medical students may thus have the potential to serve as positive change agents . . . and medical educators should ensure that students are equipped with the knowledge, skills, and support they need to play this important role.
A more recent study of medical students’ views of health-care reform and professional responsibility had a similar result with an overwhelming majority believing that it is a professional responsibility to engage in health policy, suggesting that tomorrow’s physicians are likely to participate in future health-care reform efforts (Rook et al. 2019).
In light of the critiques of the traditional individualistic position, this suggests a fourth way to understand medical gatekeeping that reconciles these historical concerns for broader issues of justice with the clinical role. This would be the kind of gatekeeping that expands the clinician’s role in clinical decision making to include contextual factors such as availability, costs, and ability to pay. The moral argument for this expanded view is straightforward. Since all physicians have multiple patients, their fiduciary obligations to patient welfare always extend across a cohort of patients under their care. As a result, as part of their core obligations to the community of patients they serve, physicians and other health-care professionals have an obligation to use their expertise to make sure allocation policies are developed at the institutional level (be they hospital systems or insurers) that make the most sense based on current research about beneficial outcomes and cost-effectiveness. If physicians [End Page 16] are protected from allocation decisions at the institutional level because it is not part of their role-specific duty toward only individual patients, they will end up being less effective patient advocates at the bed-side level. The obligation of advocacy is best understood not only in terms of advocating for individual patients against the vagaries of bureaucracies and misguided policies, but also in terms of advocating for good policies and procedures that promote good outcomes and cost-effectiveness for all the patients under a health professional’s care (Glover and Povar 1991).
Stated simply, the positive moral argument for the systemic view does not appear terribly controversial. Nevertheless, it has provocative implications for changes in the scope of content of clinical practice. Five such implications will be particularly important for those who embrace the systemic view. The first implication is that physicians should not be guided by self-interest in the development and participation in reimbursement schemes, and physician expertise should not be used in service of the promotion and flourishing of particular medical specialties but in service of better outcomes and more cost-effective care. The second implication is that physicians should discuss costs with patients as part of informed consent. The third implication is that physicians should promote and participate in gathering good data upon which conversations and allocation decisions can be based. The fourth implication is that physicians should participate in the formulation of benefit structures and allocation policies. The fifth implication is that physicians should participate in promoting a national conversation about the need to establish health-care priorities.
Implications for Clinical Practice
First implication: not guided by self-interest
Like the individualistic approach, the first principle of a viable systemic approach is that health-care professionals have an obligation not to be guided primarily by self-interest in their development and participation in reimbursement schemes. As Pellegrino and other have argued, physicians have obligations primarily to patient well-being and they shouldn’t use their position to advance their own financial gain (Pellegrino 1986; Brody 2010). Fiduciary concern for the welfare of patients is at the very heart of common notions of professionalism (Veatch 1991). There is much literature about the problems with managed care plans in which physicians receive personal incentives to provide less care to patients to control costs (Mechanic 2000; Morreim 1995). But there are also problems with fee-for-service [End Page 17] arrangements where physicians have incentives to provide too much care to maximize personal income (Schroeder and Frist 2013). In an ideal world, physicians would be incentivized to provide exactly the right care for patients. But we live in a far from ideal world, highlighting the need for physicians to be the kind of gatekeepers who don’t count personal income as the primary goal.
Just as physicians should not use their expertise in service of promoting self-interest as described above, they should avoid compromising their decision making in the service of promoting the existence and flourishing of particular medical specialties. Physician expertise should be in service of better outcomes and more cost-effective care for individuals and communities. Howard Brody called for each specialty society to commit to appointing a blue-ribbon study panel to report that specialty’s “Top Five” list. This list would consist of five diagnostic tests or treatments that are very commonly ordered, are among the most expensive services, and which have been shown not to provide any meaningful benefit to at least some major categories of patients for whom they are commonly ordered. Each specialty society would also be tasked with coming up with a plan for educating its members about these lists (Brody 2010). Brody argues that this kind of approach is necessary to begin to recapture the roughly 20% to 30% of health-care expenditures estimated to be wasted on nonbeneficial measures (Berwick and Hackbarth 2012; Fisher, Bynum, and Skinner 2009). Brody’s idea has been put into action by the Choosing Wisely® campaign initiated in 2012 by the American Board of Internal Medicine Foundation and Consumer Reports to promote more thoughtful ordering of health care services (Health Affairs Blog 2017). The Robert Wood Johnson Foundation sponsored an event in October of 2017 to discuss what has been accomplished in the five years since the initiation of the Choosing Wisely® campaign. The event noted “the movement’s growth since its founding, with seventy new societies signing on, more than 400 recommendations issued, and a steady increase in the number of studies testing interventions to reduce low-value care” (Health Affairs Blog 2017). Although progress has been made, caution is warranted. Challenges remain for better education about the campaign and its goals. A roadmap for increasing the impact of the Choosing Wisely® campaign in the next five years includes recommendations to specifically target the drivers of different types of low-value services with more rigorous study designs, and to incentivize professional societies to collaborate on approaches with the American Board of Internal Medicine (ABIM) [End Page 18] Foundation serving as the convener (Health Affairs Blog 2017). The Choosing Wisely® campaign exemplifies the positive moral argument for going beyond the individualistic view of clinical obligations to consider the welfare of patients beyond their own.
Second implication: informed consent and costs
One of the most important and provocative clinical implications of accepting a broader view of physician obligations for health-care allocation decisions is what it means for the conversations that physicians have with their patients about the costs of care. In March, 1982, Dr. MaryAnn Bailey wrote a memo to the other staff of the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research that she shared with me commenting on a draft of the Commission’s report on “Shared Decision-Making” (Personal communication). In that memo, Dr. Bailey noted that a failure to talk about cost as part of informed consent would be inconsistent with the access report that was also being published by the President’s Commission (President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research 1983). Although the issue of cost was skillfully raised by Dr. Bailey, it was never included in the final Informed Consent report (1982). Physicians have traditionally shied away from discussing costs with patients for at least three reasons. First, a more traditional account of a physician’s role-specific obligations would posit that physicians should not talk about costs because to do so would be to associate with the business side of health care and not the solely professional assessment of what would benefit the patient (Rosenbaum and Lamas 2012). A second and similar reason is that physicians are very influential over the choices of their patients, and therefore physicians might influence patients in service of the physician’s own self-interest or the interests of the institutions in which they practice (Hall and Berenson 1998). The physician would therefore have a conflict of interest in discussions of cost. That is the same kind of concern evidenced in reimbursement schemes discussed above. The trust necessary in the patient/physician relationship could be violated with discussions of payment. A third reason is that physicians just do not know about the costs of the services that they recommend and provide (Allan, Lexchin, and Wiebe 2007).
But there is evidence to suggest that patients want to know about their out-of-pocket costs when making decisions about their health-care choices (Alexander et al. 2004). A physician’s primary obligation to promote an [End Page 19] individual’s well-being necessarily includes such contextual features as the timing of an intervention, location, and also out-of-pocket costs. Patients’ preferences and values include these nonmedical considerations about timing relative to other personal events, location if it means challenges to their ability to travel, and costs if it means a choice between paying for the drug and paying the rent. A recent study indicates that in spite of the ACA, about two-thirds of all bankruptcy filings in the US are related to illness and medical bills (Himmelstein et al. 2019). By taking into consideration the patient’s nonmedical preferences, the physician indicates that he or she recognizes the patient as something other than his or her medical condition and recognizes the obligation to treat the patient as a whole person (Goold 2013). Conscientious physicians necessarily discuss such things as cheaper medications, more limited testing, and less frequent doctors’ visits with people who are self-pay because they realize that patients may not be able to afford everything that seems to be “standard of care” (Goold 2013). It is a matter of promoting patient benefit in an expanded sense of benefit that includes the contextual features of care that are necessary to promote adherence.
In addition to promoting patient benefit, discussing costs could also be an important feature in avoiding bad outcomes as a matter of reducing risk of harm. If cost is not discussed, patients may not disclose that they cannot afford the prescription and just never fill it; or not go to a specialist for fear that they cannot afford what could be recommended such as an expensive surgery. There is evidence that the high costs of drugs in general, especially with the high costs of co-pays, lead to exactly those types of bad outcomes due to prescriptions not being filled or refilled (Prasad, De Jesus, and Mailankody 2017, 3). Many states have so-called “dispense as written” laws that require pharmacists to dispense exactly what the physician has written on the prescription, including name brands even if there is a less expensive generic alternative (Shrank et al. 2011). To help patients, it is necessary for physicians to know about less expensive options, especially in these states, so they can write the less-expensive prescription. A recent example would be an alternative to the EpiPen, (Auvi-Q®) since the price of the EpiPen has increased by over 400% in the last decade.
But to argue that it is important to discuss costs does not necessarily translate into requiring only the physician, in particular, to have these discussions, since the physician may not be knowledgeable about actual costs. Many larger institutions and hospitals have billing offices whose job it is to discuss costs with patients before treatment and many would [End Page 20] prefer that physicians and health-care staff not discuss these issues with patients. Also, other members of the team, like pharmacists, may be in a better positon for this discussion about cost. But I am arguing that it is always the physician’s responsibility to normalize the discussion of costs by establishing that costs are important and should be discussed, and to provide access to that information from knowledgeable others if the physician does not have it. For example, colleagues have disclosed that, as Kaiser physicians, they are instructed to give all available options and the pros and cons of each and also the information that not all plans might cover all options, and then to provide the patient the number to call for coverage information, including co-pay information (Personal Communication). Because costs are important to patients, physicians, who have information about what trade-offs in medical interventions may mean to patients, should be able to talk about them (Ubel et al. 2016).
However, there is little in the literature about how physicians should actually talk about costs. There is one article in the Journal of Oncology entitled “SPIKE$: A Six-Step Protocol for Delivering Bad News About the Cost of Medical Care” (McFarlane, Riggins, and Smith 2008). The authors modify the well-known method for breaking bad news known as the SPIKES method. SPIKES stands for S (setting and listening skills); P (patient perception); I (information about what the patient currently understands or wants to know); K (knowledge required to be able to pass on to patients); E (explore emotions and options) and S (strategy and summary). (McFarlane, Riggins, and Smith 2008). With increasingly expensive oncology drugs in particular, these authors are proposing the addition of financial factors into SPIKE$. The additions include such things as being clear about who is available to address financial concerns in the practice setting; from a patient perspective, inviting others to be present who know more about the patient’s finances; inquiring about other avenues the patient has already tried for financial relief; and being clear about who will be responsible for working with organizations that manage co-pays or free access to drugs. Another article emphasizes the need for empathic communication when discussing costs with patients (Hardee, Platt, and Kasper 2005).
Perhaps the most difficult part of these conversations is the emotional toll on health care professionals of hearing patient stories about their financial losses, including lost homes, and the current financial stresses they are under, and working with patients and families who want to discuss whether the benefits of the drugs are worth the financial and other [End Page 21] burdens. It is no wonder that financial issues are not routinely addressed in the setting of oncology care (McFarlane, Riggins, and Smith 2008). It is very challenging to build the capacity to have full and specific discussions about cost in hurried practice settings. Various tactics are being tried by hospitals to increase transparency of costs to physicians and patients, including hiring pharmacists to specifically research patient’s co-pays and provide that information to patients and embedding information into Electronic Health Records that makes co-pay information visible to physicians in real time to have the information available for patients. In general, there is a call for increasing transparency in health-care costs, as evidenced by the new federal mandate as part of the ACA that began on January 1, 2019 and requires hospitals to provide the prices of treatment and procedures online so that they are readily available to the public. Many major medical insurance companies also have made information about costs and co-pays associated with various treatments available through their patient portals. But there is a difference between the price of the treatment, what the patient will pay as a co-pay, and the actual cost of the treatment for the institution. Information available online is only so helpful and can be quite confusing (Mehrotra, Chernew, and Sinaiko 2018). Physicians have an obligation to help provide more meaningful information to their patients to help in individual decision making.
Third implication: physician participation in promoting research and good data
The third principle of a systemic view is that allocation decisions should be based on good data about effectiveness and cost. This is extremely difficult in the current environment. Whether it is the physician or pharmacist trying to counsel patients on a cost-effective choice or the formulary committee of the local hospital, private insurer, or Medicaid trying to make good choices about how best to expend scarce resources, the available data about efficacy of new drugs are deficient. Many drug trials are funded by pharmaceutical companies and are designed with surrogate endpoints (other than overall survival) aimed to shorten trial duration, minimize the number of trial participants, and enable rapid access of drugs to market (Wise 2016). Many drugs are me-too drugs, where head-to-head trials of old and new agents do not need to prove superiority of drug A over drug B, just non-inferiority (Wise 2016). And finally, the FDAs decision to introduce a “breakthrough” category in 2012 has increased the risk of premature approval based on limited evidence (Wise 2016). [End Page 22]
The recent FDA approval of Nusinersin in December 2016 for all patients with any form of spinal muscular atrophy (SMA) is a case in point. The drug costs $125,000 per dose for 6 treatments in the first year for a total of $750,000 (not including costs of administration) and has to be provided for life. The costs of administration are relevant because the drug has to be administered via lumbar puncture, and for some SMA patients it would require sedation or general anesthesia. The cost of the drug for maintenance is $350,000/year, not including the costs of administration. Studies show a clear response in SMA1 in the short term, but other forms of SMA and long-term results are not well studied. And there are currently no studies in any adult patients, but they are expected in the near future. Pediatric hospitals struggle with how to install the capacity to provide the drug and how to establish priorities for who gets the drug when not all can (Burgart et al. 2018). Hospitals have determined different mechanisms to help address allocation decisions, ranging from the medical director alone to a specific multidisciplinary committee that includes physicians, administrators, nurses, social workers, and ethicists (Burgart et al. 2018).
State Medicaid offices struggle with the fact that providing this drug to a small number of patients would require reducing spending elsewhere by some percentage to cover the costs. Most Medicaid budgets are flat and some are being reduced. So to pay for Nusinersin, it is likely that doctors, hospitals, pharmacies, labs, behavioral health providers, and nursing homes would need to experience a reimbursement cut. To be clear, Medicaid rules do not allow states to simply not pay for “medically necessary” drugs; though it is possible under some Medicaid block-grant proposals that state Medicaid plans could find themselves in the same boat as other insurers and providers, weighing the benefits of drugs against their costs and making trade-offs in the presence of scarce resources.
Since the impact of paying for Nusinersin on providing other services is considerable, and since little is known about the effectiveness of Nusinersin in many populations, the first necessary step to making ethically justifiable decisions about coverage is to push for better research that yields better data on which to base such important decisions about benefits and costs.
Fourth implication: physician participation in the formulation of benefit structures and allocation policies
A fourth principle of the systemic view is that physicians have obligations to participate with others in the design and administration of benefit structures and in the design and administration of other allocation [End Page 23] policies and procedures. As the AMA’s Ethical Force Program’s Consensus Framework notes “the contents and limits of health care benefits must be established through an ethical process” (Levine et al. 2007, 17). “All stakeholders in health care—including patients—should be held to a balanced system of accountability for the effective and efficient operation of the system” (2007, 18). Most health-care systems have medical directors who are responsible for making ultimate determinations of whether a particular patient receives a particular service. Understandably, they struggle to apply pre-determined polices if they even exist. It makes more sense, ethically and practically, if the processes to establish these benefits and allocation policies are more inclusive in the first place. Take the examples of the allocation of ICU beds or the use of the drug Nusinersin described earlier: physicians’ participation in the development of systemwide policies would be more important and necessary than only their participation at the bedside. Physician participation in whether to include certain procedures like transplants in a benefits structure would also be important. The more inclusive the procedures to establish benefits and allocation schemes that include a wide array of stakeholders, the better the chances that the decisions at the bedside would be just enough (Fleck 2009).
Fifth implication: physician participation in promoting a national conversation
The fifth principle of a viable systemic view is the requirement that physicians not only engage in discussions within systems and at the bedside, but they also should advance the national public dialogue about health-care access and allocation and the need to set priorities. Physicians should use their considerable power and influence to advance the dialogue about shared health-care needs, as opposed to acting like professional unions to advance the power and privilege of the medical profession (Brody 2010, 283). This conversation is particularly difficult in our individualistic society where there is little general discussion of the common good and the medical commons (Cassel and Brennan 2007). When there is public dialogue around the Affordable Care Act, for example, there is also considerable commitment to personal choice of insurance plans or whether to purchase health insurance at all if it does not make financial sense for the individual (Kaiser Family Foundation 2012). On a policy level, it is clearly more cost-effective if everyone purchases insurance (Berwick, Nolan, and Whittington 2008). The cost of paying for care for people [End Page 24] without insurance is born by others in the form of cost-sharing among payers for emergency departments and hospital stays (Porter 2009). However, there is evidence that people are interested only in their own costs, not the costs borne by others (Sommers et al. 2013). This leads some to conclude that people may be most interested in their own well-being, not the well-being of their neighbors. This would seem to undermine the very possibility of the rational democratic deliberative processes put forward by Fleck in Just Caring. He writes elsewhere that he believes the public is capable of learning about the problems of health-care cost escalation, the role of advancing technology, and marginal benefits and trade-offs. “Oncologists can endorse expensive cancer drugs because they think only about cancer. The broader healthy public has to struggle with whether to improve five dysfunctional years of arthritis sufferers in order to provide two more months of life to someone with cancer” (Fleck 2006, 16). As Allen Buchanan argues, if members of managed care plans (and presumably their physicians) are unhappy with the current state of affairs, then they need to engage in the public deliberations that can create and legitimize a shared conception of health-care justice (Buchanan 1998). The problem is not simply lack of information, but also motivation. Reaching back to the discussion of virtue-based theories of justice, it is centrally important that we care not only about our own capabilities and well-being, but that we care about what happens to all of us together. As responsible members of society, everyone has a stake in promoting just processes. Physicians have obligations of justice because their patients do as well. Public dialogue about health-care allocation requires the participation of all of us and especially physicians and their expertise and ethical commitments to a systemic view of advocacy.
I have argued in this paper that physicians’ advocacy obligations are best understood as going beyond advocacy on behalf of individual patients, which I call the “individualistic view,” to include advocacy for intelligent research-based allocation schemes that promote good outcomes and cost-effective care for all patients, which I call the “systemic view.” This systemic view includes moving beyond self-interest to promote less-wasteful and more cost conscious allocation decisions and the setting of priorities at all levels to expand health-care access. I have argued that the setting of health-care priorities and allocation decisions is necessary, as is the involvement of physicians in this difficult task. This involvement is based on a concept [End Page 25] of a deliberative process that can result in “just enough” decisions within systems. These decisions won’t be perfectly just, but they are a necessary beginning for the preservation and promotion of health-care and other societal goods. There are five implications for physicians of this systemic view, all of which run counter to the profession’s traditional views on the clinical significance of medical economics. The first implication is already widely recognized: that physicians should not act out of self-interest when developing or participating in reimbursement schemes. The second implication goes further to prescribe that, contra the received wisdom, physicians should include cost in informed consent discussions in the context of clinical care. Thirdly, to give those discussions an evidence base, physicians should be involved in promoting research for sound data to support allocation decisions. Fourthly, the systemic view implies that physicians should be involved in the formulation and administration of benefit structures and other allocation policies. Finally, as a part of that involvement, the fifth implication is that physicians should be involved in promoting public dialogue about health-care priorities. Only by embracing these five implications will physicians be able to more fully satisfy their professional obligations to their patients.
Jacqueline J. Glover, Ph.D., is a Professor in the Department of Pediatrics and at the Center for Bioethics and Humanities at the University of Colorado Anschutz Medical Campus where she directs efforts in clinical ethics consultation. She is the co-medical director of the Ethics Consultation Service at the University of Colorado Hospital and an ethics consult faculty lead at the Children’s Hospital Colorado. Her research interests include pediatric ethics, ethics consultation, rural bioethics, interprofessional education, professionalism, and health policy.