When behavior is profoundly at odds with ideals, then the behavior or the ideals or both need to be reviewed. My aim in this essay is to review the ideal that the single fundamental duty* of the physician is to secure the individual patient's best interests and wishes.1,2 I present an alternative view by asserting that physicians have two fundamental duties: they must balance the interests and wishes of the patient with the welfare of the health care system in which they practice. In fact, many physicians actually act in accordance with this dual approach, but they sense an inconsistency between their behaviors, which aim at both goals, and their ideals, which focus on the patient only. In this dual stewardship model the physician is both the agent of the patient and the agent of the health care system. As a consequence of this more complex role, I explain why medicine must stop viewing the case manager as the unavoidable but unwanted child within the family of medicine and must incorporate case management within the very identity of the physician. Finally, I offer some “promissory notes.” That is, when the physician practices according to the dual model just described, at least three beneficial consequences will emerge. First, the abysmal quality of end-of-life care in the United States may improve. Second, restraints on funding for lastchance or experimental treatments may, surprisingly, advance patients' interests. Third, an admission that much is not known about many mental illnesses and an acknowledgment that the money spent in treatment often yields limited results may encourage the health system to resist providing unproven treatments. This admission could be the impetus for developing more scientifically effective approaches to those bewildering problems.
THE SINGLE-STEWARDSHIP MODEL
Until the late 1960s the traditional role of the physician was to secure the medical welfare of his or her patient. The modern notion that the physician's stewardship extends not only to the medical welfare but also to the wishes of the patient is very new. During the long history that preceded the advent of the modern bioethical movement, patients' wishes were treated as secondary to their medical welfare. But this ancient view crashed into the wall of the contemporary bioethics movement. The core of that movement is the notion that the doctor has a deep commitment to secure the wishes and rights of the patient. Patients' autonomy is portrayed not only as valuable in itself but also as necessary for determining what is in the best interests of the individual.3 However, what is crucial to recognize is that from this perspective, wishes are relevant to the actions of a faithful physician whether they are consistent with the patient's welfare or not. In short, the autonomous individual has a legitimate claim to define what is best for himself or herself as a patient even if the doctor disagrees.
It was difficult for the medical practitioners of the 1960s to admit this criterion into their everyday practice. Nevertheless, it is now common to acknowledge that both clinical science and the autonomy of the individual patient are relevant to the duty of the physician. I refer to this approach as the single-stewardship model because in this view, the individual patient represents the intrinsic value or the highest good of the medical steward. Using this model, the ethical responsibilities of the good physician—and the basis for his or her decision making—are defined exclusively in terms of one consideration: the individual patient. This is at odds with nearly all other areas of life, where individuals—patients and non-patients alike—are recognized as having a variety of considerations influencing their decision making. Should we as a society continue with this single-focus, individual-centered ideal?
The answer is No! Good physicians have an additional obligation. They must balance their duty to the patient with their duty to other stakeholders in the health care system. Physicians, as a matter of fact, already follow the dualstewardship model I am proposing, but many feel guilty about it, and this needs to stop. The argument begins by recognizing a fundamental contradiction between managed health care, in which over 75% of U.S. physicians practice, and the single-stewardship model. Most managed care organizations admit that they do not offer the highest-quality packages of health care services. They do not pretend to offer the best. They do not promise to meet all the wishes of all patients. Rather, most offer decent levels of quality, and decent is not equivalent to the best or to all that is wished. The contradiction emerges here. For if the physician who follows the single-stewardship model must seek only the best or what the patient desires, then it would follow that the majority of physicians are intrinsically unethical. This is a tough pill to swallow. When the ethic of an organization implies that the overwhelming number of practitioners are immoral, then either the ethic of the organization must go or practitioners must be eliminated from the profession. In this article I propose a way to revise the ethic.
One possible revision involves changing the physician's duty into a promise to provide what is best for the patient within the limits of the system. In this revised version the system clearly sets the limits for what is a decent level of health care, and the doctor merely administers these limits. The doctor is innocent of any violation of his or her stewardship if the limits of the health care system do not maximize the health interests of the patient. He or she is, after all, a mere administrator of limits that are set by others.
However, this retreat to defining the physician as an administrative bureaucrat is catastrophic. For example, if a patient needs itraconazole for a fungal infection, and if that medication is not part of the preferred drug formulary, then the physician's duty is discharged merely by writing the prescription and informing the patient that it is not covered. In this scenario, the doctor is a mere administrator of the managed care formulary, not a patient advocate.4
What is troublesome about this revised view of the physician as administrator is that it is exactly like the single-stewardship model. In both models the physician is unidimensional. In the first model, all the physician needs to know is the latest from the clinical journals that can help the patient; questions of cost are treated as ethically irrelevant. In the second model, all the physician needs to know is clinical science plus information regarding coverage. In this view, the doctor is nothing but an administrator.
THE DUAL-STEWARDSHIP MODEL
The dual-stewardship model I advocate rejects both of the above models not only because they are unidimensional but also because they too narrowly define the courage, commitment, and trustworthiness that are essential to medicine. To be an effective advocate of the patient who needs itraconozole involves being a trustworthy agent of both the patient and the managed care system. Both the patient and the managed care system must trust that the physician will not prescribe itraconozole without necessity. Single-stewardship physicians are not and never have been organizationally trustworthy, and consequently patients suffer because they often need the organization to pay for medically necessary but expensive treatment. Organizations have duties to hosts of patients, and these duties need to be discharged by individual physicians who are organizationally trustworthy. (On the other hand, physicians singly devoted to being mere administrators are not trustworthy agents of patients.) A thought experiment can clarify the need to include organizational trust within the identity of the doctor. Imagine that physicians were to generally “game the system”5,6 to secure all the medical wishes and the absolute best care for their patients. The result of such “gaming” would be that the health care system becomes economically unstable. That result would, in turn, harm the patients. In this scenario, the ideal or perfect doctor (practicing single-stewardship medicine) becomes the enemy of the truly good doctor (practicing dual-stewardship medicine).
Practicing dual-stewardship medicine involves a shift away from seeking to fulfill the patient's interests and wishes at any cost toward seeking the best health-care value for the patient.7 The argument for practicing cost-effective medicine is grounded in the notion that a physician should neither try to secure the absolute best care for the patient nor try to satisfy all the wishes of the patient. These goals are too high to be meaningful, and when goals are meaningless they are ignored. Rather, as just stated, the duty of the physician is to secure the best value for the patient. By best value I mean something that is a function of both cost and quality. A generic antibiotic may have the same outcome as the newest antibiotic off the research tables. But if they differ in cost, they differ in value. If the generic is much less expensive while therapeutically equivalent to the new antibiotic, then it has a greater value for the patient. Of course, if the generic is less expensive but is also less effective, then it also falls in value. Focusing on the best value for the patient rather than on the patient's absolute best interest and autonomy is to transform the health care system into something that inflates normally because it is bounded by limits rather than mistakenly being considered limitless.
This dual-stewardship model replaces the idea that doctors must treat their patients as their only concern with the idea that there are two equally high duties of the physician. The physician has a duty to secure both his or her patient's welfare and his or her organization's welfare. The two values compete, and balancing them requires as much artistic skill as that required by the surgeon. This balancing skill is not novel for the physician. When a patient's best interest is at odds with that patient's wishes, physicians sometimes resort to medical art to balance these opposing considerations in order to secure an overall good outcome. Similarly, the physician's duties to the health care system are not always commensurable with the maximum well-being or preferences of the patient. Patients often autonomously wish for things that are incompatible with the best interests of the system as a whole. The easy examples are the cosmetic surgeries or the unnecessary diagnostic procedures that may psychologically benefit the patient but threaten the health care system. Harder examples involve last-chance therapies and experimental protocols.
We can further explore this idea of dual stewardship by addressing two criticisms of it. The first criticism is that such a change will undermine patients' confidence in physicians. That is, if we as a society permit physicians to take the system into consideration when making health care decisions, then patients will lose trust in their physicians. One can imagine a patient responding to his or her doctor's prescription for an antibiotic with the following question: “Doctor, is this the cheapest or the best antibiotic?”
This criticism contains a hidden assumption. It assumes that any increase in distrust would harm overall patient outcomes. But there is a growing body of evidence that indicates that a little more distrust may have had an overall beneficial result on patient outcomes. For example, one report8 indicates that nearly 30% of the surgeries carried out in fee-for-service health care were unnecessary. A little increase in patient distrust might have diminished the appalling harm that flowed from these unnecessary and often iatrogenic services. Furthermore, physicians often ignore the large number of unnecessary, expensive diagnostic procedures9 because they hurt only payers.
Surely trust in the physician is the most effective placebo in the arsenal of medicine, but a reasonable level of trust is compatible with a reasonable level of distrust. Trust is not like pregnancy. It comes in degrees, and patients understand this. Patients' expectations of the physician are simply not as childlike as the single-stewardship model presumes. Many health care services are beneficial but too expensive, and refusing to cover these services is permissible so long as one has not promised otherwise. Health maintenance organizations (HMOs) create formularies to clarify their promises. These indicate that they will not pay for some pharmaceutical products even though they may offer clinical advantages. Single-stewardship physicians are deeply troubled by this, since they are ethically forbidden to do anything but what is best for their patients.
For physicians who have this patient-only focus, it is not surprising that payers' needs are not even on their radar screens. But it is precisely this attitude of treating the individual patient as more valuable than the health care system that has brought the fee-for-service health care system to the end of its institutional life. When payers feel that their legitimate needs for resources to sustain their institutions must be sacrificed to the needs of the health care provider, then reason requires revolution. In short, what is most troubling about the single-stewardship model is that it perfectly justifies spending unlimited amounts of resources to eliminate even the smallest risk. This follows because the physician, the central figure in medicine, is structurally forbidden by his or her ethics to be concerned about the health of the system.
In response to this reluctance on the part of single-stewardship physicians to take costs seriously and play the role of rational distributors of scarce resources, the health care system has expanded to include an entirely new health profession, the case manager. This new and expensive profession does not provide any patient care and adds to the cost of the system. Its only job is to resist the single-stewardship physicians! Because physicians, who logically should play the case-manager role, have steadfastly balked at playing it, the bureaucracy of medicine has grown into a nightmare of inefficiency and often delayed necessary treatment. Acceptance of the dual-stewardship model will lead to the elimination of these administratively needed but medically unnecessary professions. To be a doctor is to serve two inextricably related masters. The case manager's role should be incorporated into the very identity of the physician.
The second criticism of the dual-stewardship model is that once it is accepted, the scope of the physician's commitment will be so extensive that it will permit and justify gag clauses to inhibit free and open communication between patients and physicians. Physicians' contracts with HMOs have often contained these gag clauses, which prohibit the physician from engaging in conversation that may economically harm the HMO. Here is an actual example of such a clause: “The physician shall not say anything to the enrollees that could weaken the confidence of the enrollees in the managed care plan.”10 From the perspective of the plan, the goal of these clauses is to protect the economic interests and proprietary information of the plan.
Both public and private managed care plans tend to view themselves as businesses, and as such they claim the traditional right to expect that their employees or other agents will be loyal to the economic interests of stakeholders (the citizenry, if public; the investors, if private). From the perspective of the patient and the provider, such clauses often increase nondisclosure, thereby undermining patients' understanding of their options, and may weaken the capacity of physicians to secure the medical welfare and wishes of their patients. Because gag clauses inhibit free and informed consent, they undermine the patient's welfare and autonomy.11
Shifting from a single- to a dual-stewardship model does not, however, require that we support the gag clause. The central theme of the dual-stewardship model is that the physician will value the health system as well as the patient. With such an approach, the physician will not feel he has a right to “game the system” to satisfy the wishes or interests of the patient. The tendency to “massage” a diagnosis in order to accomplish what the patient wants is something that weakens the power of the HMO to accomplish its fundamental goal. That goal, as stated in much of the managed care legislation of the 1970s and 1980s, was to maximize distribution of a decent level of health care by reducing medical inflation. The physician who is a dual steward understands and supports that goal, and while the dual-stewardship role prohibits the massaging of diagnoses and other such deceptions, this does not require that the physician deliberately withhold relevant information. Alternative treatments may not be covered or fully covered by an HMO, and learning this may indeed disappoint the patient. Furthermore, this limitation on types of treatments may diminsih a patient's satisfaction, but that is common whenever treatments, their costs, and their availability are being weighed. Few patients believe that signing an HMO contract involves a promise to fund everything. Patients are consumers, and as such they select among competing HMOs. Such tradeoffs are compatible with free communication among physicians and patients.
A Difficult Question
Now we come to the difficult question of what it means to say that the duty of the physician is to secure the best value for a patient within a third-party payer system such as an HMO. HMOs come in a variety of forms. Some are “for-profit,” some are private “not-for-profit,” while others are “government sponsored.” Often critics believe that it is only the “for-profit” HMOs that need to be opposed, as if profit were something unethical that soils everything that it touches. What is important to recognize is that the dualstewardship model does not aim to defend “for-profit” managed care. Rather, in this model the question of profit is irrelevant, since the determination of best value is the central goal of the physician. This is something to be determined by multiplying the quality of the medical service by its cost. It does not matter whether the HMO is a for-profit or a not-for-profit or a government agency. What matters is to increase value for the patient. This is a function not only of the physician's taking dual-stewardship duties seriously to achieve the best value, but also a function of how well the particular HMO strives for the best value. If stockholders can produce the best value, then dual-stewardship physicians should side with them. If “not-for-profit” approaches secure maximum value, then they win the value war. If the government can best decrease costs and increase quality, then it is the best means of securing value. My own bias is that a mixture of all three types of HMOs will yield maximum medical value over time. The jury is still out on the solution to this challenging problem. But in all of these scenarios, physicians have a dual rather than a single stewardship role.
Of course, patients do often want their physicians to treat them as if money were irrelevant. What is comforting to recognize is that all three HMO types refuse to accommodate this desire. All three reject the adolescent idea that the doctor's role is to secure the absolute best interests and wishes of the patient, period. All three hold that achieving the best value for the patient is the duty of the physician.
Are the hundreds of thousands of physicians who interact with HMOs immoral for taking HMO costs seriously? If we accept the single-stewardship view of the physician, then answer is yes! But if we reject that view and accept the new vision of the physician's duty, then the common behavior of hundreds of thousands of physicians becomes consistent with our image of the good physician. The good physician is not someone for whom cost of care is irrelevant, but neither is he or she unconcerned about the practices of the HMO. Reasonable medicine is often sacrificed for short-term benefits to the HMO, and this is why physicians' commitment to patients requires that they constantly monitor their HMOs to determine that their practice guidelines and drug formularies secure a reasonably decent level of health care. For while such physicians are legitimate agents of the HMO, they are also agents of the patients they serve, and this dual relationship means that they both support and resist the HMOs in which they practice. Ideally, they would seek to transform a problem HMO into one where both it and the physician could find optimal ways to achieve best value for their patients.
The dual-stewardship role is consistent with the idea that conflicts among the HMO, the physician, and the patient are inevitable. What are the duties of the physician in such conflicts? The first duty is avoid “gaming the system” by distorting the medical facts to secure additional services for their patients. That is the physician's primary commitment to the HMO. The second duty is to share with patients information about the full range of procedures, medications, and services available within and without the HMO. A physician who allows himself or herself to be gagged by the HMO and thus prevented from sharing medically relevant information with the patient has given up the ideal of balancing one's duty to the patient with one's duty to the HMO and instead treats the commitment to the patient as little more than a commitment to the financial integrity of the HMO. A gagged physician becomes the exclusive agent of the HMO rather than someone who faces the more daunting challenge of balancing the good of the patient with the good of the HMO. If a physician believes that an HMO could offer a better value to a patient, then he is obligated to advocate this better value. If an experimental protocol, for example, offers the best chance for life to a patient, and the HMO does not cover this protocol, then there is a clear duty on the part of the physician to inform the patient of this alternative. But there should be no attempt on the part of the physician to secure this protocol by any form of deceptive interpretation of the disease.
Before I turn to the three “promissory notes” mentioned earlier, let me state what dual-stewardship physicians cannot do. They cannot guarantee that a decent minimum of health care will be available to everyone. Because a plumber is efficiently offering plumbing services for sale, he or she does not guarantee that plumbing will be available to everyone. All the plumber promises are values to those who can afford his or her services. HMOs and the demands that they make on physicians to practice dual-stewardship medicine have not provided a decent level of health care to all U.S. citizens. Forty-five million people in our nation still have no access to medical coverage. As of now, the United States has not decided to treat health care as a right. However, what the dual-stewardship model makes possible is an opportunity to provide a reasonable level of medical care at a cost that does not irrationally inflate the health care system beyond the inflation occurring within other typical sectors of the economy. If the government ever decides to extend coverage to the uninsured, then this extension will not involve offering to everyone the kind of entitlement that could dangerously hyper-inflate the health care system.
THREE “PROMISSORY NOTES”
Let us now turn to the heart of the argument that shifting to a dual-stewardship model is compatible with patients' welfare by looking at three beneficial consequences of using that model.
Improvement of End-of-life Care
More decisively than any other study, the SUPPORT study, a controlled trial to improve care for seriously ill hospitalized patients,12 has found that end-of-life care in the United States is still falling very short of our moral expectations of such care. (The acronym “SUPPORT” stands for “Study to Understand Prognoses and proferences for Outcomes and Risks of Treatment.”) The study revealed in stark numerical terms that the overwhelming number of dying patients, who desire to avoid highly technologic, death-prolonging, hospital-based services, continue to receive these services against their express wishes.
Moving toward a dual-stewardship understanding of the relationships among the patient, physician, and health care system will increase the possibility that such bad deaths will diminish. The physician who has that understanding will find those points where the interests of the patient and the interests of the health care system merge to secure the best value for the dying patient. For example, home care following surgery is an effective way to help patients while also increasing the efficiency of the system as a whole by avoiding the vast health-care efforts used in “death-prolonging” strategies that harm both patients and the system. The physician is uniquely fulfilling his dual role when he or she develops and participates in this and other strategies that secure the welfare of the patient while also contributing to cost control.
The transition to home health care after surgery has been revolutionary. Care for dying patients needs to be changed to require home care as the standard of care for the dying. Here is where patients' and the system's interests merge. To view the physician as a caregiver with a dual stewardship is to portray the physician as having a strong duty to avoid both prolonged deaths and wasteful expenditures.
The following are additional suggestions of how the implementation of the dual-stewardship approach would alter how people die in this country. First, in the single-stewardship setting the patient's opposition to non-coercive treatment at the end of life is largely informal and verbal. There is little effort on the part of the health care system to secure living wills or respect those that are filled out, since there is no financial incentive to do so. However, once physicians become committed to the health care system as well as to the patient, the pressure to secure and obey these directives is increased because there is a convergence of the patient's wishes and the system's welfare. In the HMO setting, whether it is public, or “private-for-profit,” or “private-not-for-profit,” the physician's commitment to the patient's desire to avoid technologic death merges with the interests of the system. Penalties for violating these directives as well as bonuses for respecting them are instances of responsible dual stewardship.
Second, the tendency to uncritically give dying patients all the care they want, no matter how unlikely the chance that it will help, needs to be reexamined. Within the single-stewardship system, satisfying patient wishes is equated with the physician's duty, and value is treated as irrelevant. But in the dual-stewardship system, physicians are professionally skeptical of admitting patients into death-prolonging intensive care units (ICUs) in the absence of clear practice guidelines that identify which patients are likely to benefit from such intensive treatment. Practice guidelines are not built on the idea that there is a slight or slim chance that a treatment will be effective. Rather they are built on outcome-based evidence that establishes that a treatment is likely to advance the recovery of the patient. These practice guidelines are tools aimed at screening out services that are unlikely to help. In developing and adhering to end-of-life practice guidelines, dual-stewardship physicians will avoid many of the harms associated with providing unnecessary and expensive end-of-life services that only extend the dying process. Such physicians offer hospice and/or home care nursing as standards of care, since those merge the interests of the patient with those of the system. Within these contexts, aggressive palliative care can be provided in an atmosphere conducive to patient dignity at a fraction of the cost of traditional ICU deaths. This is the kind of change in direction in end-of-life care that can redefine the standard of care for the dying patient.
Third, because the dual-stewardship physician is committed to providing the best care value to the patient, he or she will be a case manager who creates alternatives to viewing cardiopulmonary resuscitation (CPR) as the standing order for those who go into cardiopulmonary arrest. CPR is the first step in the all-too-frequent process of spending vast fortunes with little or no benefit to patients. Such expensive and harmful services are currently provided to hundreds of thousands of terminally ill patients, and the single-stewardship system seems powerless to stop it. This inflexible approach to ordering CPR is similar to instituting universal screening for a disease that affects only a small number. We do our best to avoid these costly screening ventures, not because they provide no benefit, but because they provide so little. Similarly, universal resuscitation for patients who go into arrest remains the standard of care, even though there is a solid body of outcome evidence indicating that this is harmful or useless for very large groups of definable patients. We continue to treat CPR for all as the best care even though we know that it is not the best care value.
When patients' interests merge with the system's interests, the pressure to develop an alternative approach will be increased. One such alternative involves the assessment of patients for CPR before they are admitted to the acute care setting. This means increasing pressure on hospitals to distinguish patients who can benefit from CPR from those who in all probability cannot. Just as information is required to avoid the administration of harmful antibiotics to those patients who are allergic to them, so, too, information is needed to avoid the harmful use of CPR. The dual-stewardship physician seeks such information in attempting to merge the interests of patients with the interests of the system by reinterpreting CPR as a potentially harmful treatment that is standard for some but not all.
Fourth, unlike traditional physicians, the dual-stewardship physician views his or her commitment to patients as strongly bounded by futility. To provide futile services to persons in a permanent coma or a permanent vegetative state undermines the health care system's capacity to help the ill. Even the patient's wish for these services is insufficient to mandate the dual-stewardship physician to provide them. Futile treatment does not merge the interest of the patient with those of the health care system as a whole, it further divides those interests. By continuing to provide futile treatment, single-stewardship physicians inhibit the transfer of vital medical resources to those who can benefit.
Let us now turn to our second promissory note. If the traditional physician alters his or her image of the physician's role to include a responsibility to the payer, what are the implications of this change for clinical research? Research physicians already are troubled by a parallel dual obligation. They are committed to the individual patient, but they are also committed to future patients; in short, they have a dual stewardship. That is, with the informed consent of the patient, they may “use” the competent, adult patient as a human subject. This may help the patient, but it is also aimed at advancing the interests of future patients.13
What would happen if physician researchers took on an additional commitment to those who contribute to such research, namely managed care organizations? The question is somewhat academic, since so much funding for clinical research already flows through the hands of managed care. In fact, there is little doubt that managed care has had an enormous effect on reducing hospital-funded clinical research (explained below). But managed care has not just reduced the quantity of research; it also has altered the kind of research that is carried out. If the physician researcher wants clinical research support from managed care, then he or she must identify with the interests of the managed care organization, just as he or she must identify with the interests of a pharmaceutical firm when accepting its support. Does identifying with such interests lead to harming patients or reducing benefit to them? I will argue that it can in some cases avoid harm. But first we need to look at some of the tools that managed care uses to manage research.
The two main tools that managed care uses to control the costs associated with research are capitation and “the investigational exclusion.” Capitation inserts hospitals into vigorous competition for limited contracts from managed care organizations. It forces hospitals to review traditional cost-shifting strategies that may undermine their competitive efficiency. When hospitals realize that their tendency to support clinical research, through cost shifting, explains their competitive inefficiency, then they tend to reduce support for such research or to significantly reduce or redesign their research so that it comes in line with the hospital's business plan. The business plan of the hospital thus mediates between the pure need for patient research and the hospital's need to operate within a competitive environment.
Hospitals have traditionally viewed themselves as having an academic function in which advancing knowledge was viewed as an intrinsic value, one that could be legitimately funded through cost shifting patient-care revenues to support research and education. However, as mentioned above, under the pressures of competition, hospitals tend to sacrifice this academic approach. Similarly, a business approach replaces this academic approach within managed care. For example, in the traditional fee-for-service setting, if a hospital had a competent physician researcher on its staff, and the physician wished to do research on arthritis, then the traditional hospital might well have supported the research if it were likely to yield publishable results. In short, an academic standard was sufficient to warrant support. But in a business-like model of clinical research, academic or scientific quality remains a necessary virtue, but it is no longer considered sufficient to warrant support. The research must fit into the business plan of the hospital. Thus, the academically qualified arthritis researcher might find little support for his or her research proposals within the hospital precisely because the hospital is not marketing itself as an arthritis “center of excellence.” In other words, representing itself as an arthritis center of excellence is not part of its basic business plan. Since arthritis research does not fit into the overall business or marketing plan of the hospital, then such research is not advancing the competitive goals of the hospital. In this business model of clinical research, the acquisition of scientific truth is viewed as a mere “cost that needs to be justified,” and academic excellence is only one necessary part of a larger process of justification. It is necessary but not sufficient.
What research, then, should the hospital continue? The answers are as varied as the business plans that are adopted by hospitals. Truth and scientific validity remain important in the sense that the scientific method is still employed to establish results, but the business plan of the hospital or other funding agency becomes a critical voice in determining what research gets done. In most market-dependent businesses, some form of research is vital because in its absence a business forsakes its competitive edge. Because the decision to support a hospital's research activity has shifted toward market-driven considerations, it is not surprising that the amount of hospital-based research has declined. What remains often is driven by the desire to seek therapies that allow the hospital to lower costs, rather than by the desire to discover therapies without sufficient attention to costs. In short, the search for less costly therapeutic equivalents has been considered to have a low priority—or none at all—within the single-stewardship model, but it gains in value within the dual-stewardship model.
The investigational exclusion is the second tool that managed care uses to reduce research funding. Managed care contracts follow their forebears in traditional health insurance by excluding investigational research (i.e., clinical research involving human subjects) from coverage. But unlike traditional health insurance, managed care organizations, through the ERISA law, have gained increased legal protection for such exclusions. Furthermore, organizations offering traditional medical indemnity insurance did not press the exclusion rule as forcefully as managed care has, since they were not involved in the detailed management of health care costs. They were not being judged on the basis of competitive efficiency. Instead, they were largely providing financial services and staying clear of controlling the actual nature of medical services, since there were little or no market pressures to manage medical behavior. This independence, of course, contributed to the inflationary character of medicine during the 1980s. Indeed, many felt that unless medical research was itself managed by cost-conscious payers, any hope of controlling overall medical inflation was doomed.
How should we respond to these new forces that limit and control the direction of clinical research? At the most abstract level, managed care aims to fully integrate research into a system of market forces. This integration is aimed at preventing medicine from continually spiraling out of inflationary control while still allowing for steady, albeit slower, progress. Scientifically sound research progress is present throughout the economy, and while improving health care is important to everyone, there are limits in the extent to which the society can support clinical research. My point here is that if physicians support these efforts to integrate research into larger market forces, they will be supporting steady clinical development. This integration will allow clinical research to continue without the harmful medical inflation that was the primary force motivating the emergence of managed health care.
An illustration may help. Many expensive “on-patent” drugs offer little substantial improvements over therapeutically equivalent “off-patent” drugs. When scientifically solid studies determine this therapeutic equivalence or even near-equivalence, then dual-stewardship physicians are justified in using medications that advance the benefit of their organizations, especially if they are entitled to switch to an “on-patent” drug if the less expensive first choice is found to be ineffective. Redirecting clinical research toward “therapeutic equivalence” as well as “outcome-or evidence-based studies” alters the nature of research by moving it toward economically sustainable forms of treatment. This redirection advances patients' welfare by creating a form of medicine that can be sustained within a competitive environment.
The dual-stewardship physician is someone who realizes that for the health care system to survive over the long run, new forms of clinical research are needed. For all of its defects, managed care has certainly changed the course of medicine in the direction of market efficiency. This change was needed. We must not continue to view the task of research as simply to develop new drugs or procedures that offer the slightest clinical advantage, with little or no regard for overall cost implications of this research. Doing so isolates clinical research from the cost constraints imposed by market forces, and this makes medicine as a whole vulnerable to skyrocketing inflation.
With the above in mind, it is clear why managed care responds as it does to the problem of paying for medical experimentation. Economically, the most important thing about research is that it is risky; much of it fails despite our attempts to put positive spins on research failure. Realizing this is vital because the conflict between research funds and therapeutic funds is a critical issue within managed medicine. Both research and therapeutic funds come from a common source. A choice to fund research is, therefore, a choice to reduce funds for other things, including proven therapy. In effect, managed care makes health care a “zero-sum” game in which sacrifices for research become painful, which means that research costs must be treated as being very risky indeed. This zero-sum approach to research is one of the reasons for managed care's successes at controlling medical inflation.
Another implication of this business approach to clinical research is its unwillingness to cede authority over research questions to neutral panels whose members do not bear the responsibility to maintain the financial integrity of a health care system. There is little hope that such panels can determine when a research proposal should be funded and when it should not.14 Many have argued that such research panels are the way out of the cost dilemma, as if these panels can do what trained scientists cannot do, namely, pick the successful research proposals prior to the trial-and-error process. Managed care firms should not assume that neutral panels can make the right choices and thereby avoid the risks associated with trial-and-error research.
I recommend that the neutral panels be replaced by scientifically informed researchers who advise but do not control the final decisions regarding support. Merely because scientifically informed persons advise an institution to spend its limited resources on a research protocol does not mean that the institution is obligated to do so or that it makes business sense to do so. Final decisions rest with the managed care organization to decide whether to support a particular research project and to what extent it will support such research. The organization's managers are at liberty to allow all relevant clinical and economic forces to determine their course of action with respect to funding. Such an approach may decrease overall research funding and shift funds to investor profits, or extend proven therapies to larger populations, or reduce federal or state medical budgets. None of these goals is by nature ethically inferior to advancing clinical knowledge.
One argument in favor of this business approach to clinical research is that such an approach would have reduced the harm associated with autologous bone marrow research. Let us illustrate this claim using a case.
The Case of Mary Carson
Mrs. Mary Carson had had a recurrence of breast cancer, and has been invited to participate in a research protocol involving autologous bone marrow transplantation and the use of high-dose chemotherapy. The idea is that high doses of chemotherapy will destroy recurring cancerous tumors that are spreading throughout her body.15 These high doses will also destroy her immune system, and the transplant of such marrow will hopefully revive her immune system to fight off the deadly infections. This particular protocol is in its second stage, which means that there is insufficient evidence of its efficacy but that there is some reason for optimism. The pharmaceutical company that will produce the chemotherapy will pay for both the costs of the chemotherapeutic agents and the transplant but want Mrs. Carson's hospital costs to be preauthorized for payment. The hospital will accept Mrs. Carson but only if her managed care company pre-authorizes payment for such research. A case manager has resisted pre-authorization, and without that the research team at the community hospital will not enter Mrs. Carson into the protocol.
Avoiding the duty to pay. This type of case is particularly significant, since in recent years much effort was spent attempting to pressure managed care into paying for such lastchance efforts in spite of the fact that the treatment part of the expression “experimental treatment” was overemphasized. Many of the best medical centers throughout the United States dispensed such treatment outside the context of the double-blind method. For a variety of “provider business reasons,” the “random double-blind method” was not applied to verify that bone marrow transplant was in fact superior to the reigning “gold standards” of treatment.15 Managed care organizations throughout the country were pressured into spending millions because patients wanted an experimental treatment that had not been adequately tested. These patients were unable to believe that physicians from the best cancer centers in the world were, in fact, deceived by something that looked good. Managed care attempted to resist the tide to support such research but failed miserably. Indeed, managed care organizations were branded as being run by uncaring profit-driven administrators who were more interested in money than patients' welfare. Political pressure, which is usually more powerful than scientific skepticism, reigned supreme, and fortunes as well as lives were lost for no benefit.
The single-stewardship model is operating in this case, for it explains why both hospitals and physicians have wished to provide this kind of experimental treatment to thousands of people like Mrs. Carson, even if the treatments were not being conducted using appropriate methods and even if fortunes had to be spent to no avail. The single-stewardship model immunizes physician researchers against the need to follow “gold standard” method or to be faithful stewards of limited research budgets, and instead requires them to be committed to the welfare and wishes of their patients alone. Thus the welfare of the health care system is ethically off their radar screens.
What is important to notice about this case is that everyone is trying to avoid payment. One might argue that the pharmaceutical company should pay for the expenses associated with autologous bone marrow research, since they stand to benefit from the research if it is successful. However, most pharmaceutical companies resist such duties. They wish to lay off some of the expenses associated with this kind of research. They are willing to provide the cost of the medication and the medical researcher's fees, but absorbing the associated hospital costs would drive the cost of research—in this case, the experimental protocol for Mary Carson—to astronomical heights. In short, if pharmaceutical companies or other research groups must bear the complete costs, then they will abandon the research.
Perhaps the federal government should pay these kinds of expenses. The government already makes a big commitment to medical research in the form of grants from the National Institutes for Health. Billions of research dollars are poured into health-related research in the form of grants to public and private agencies to do clinical research. But even governments have financial limits. So far, government is not obligated by law to pay for the associated costs of autologous bone marrow transplant and, therefore, government cannot be counted on to absorb the hospital costs associated with experimental treatments like Mrs. Carson's.
Finally, the patient and the patient's family might be expected to absorb the costs. But as one might expect, these costs are potentially staggering. Hospitals attempt to secure payments “up front” prior to the service so that they will not acquire another unpaid debt that could undermine their solvency. But few individuals or families could absorb these associated hospital costs. They could run into hundreds of thousands of dollars, and the family's resources to meet its other expenses would be threatened.
Who should pay when everyone is running away from the payment responsibilities? Should the managed care companies pay?
Managed care and last-chance therapy. Do the managed care case managers have the moral right to refuse to pay for Mrs. Carson's experimental therapy? If they do not have the right, then it would seem that all other parties also lack this moral right. What is clear is that payers, of all stripes, tend to abandon the Mrs. Carsons of the world when last-chance therapies are requested. Why should managed care be held to a standard that government, individuals, and families typically refuse?
I recommend a procedural approach to this question. The decision to refer to a last chance as a therapy or an experiment or a protocol is not a question that can be answered by moral theory. In other words, moral theory cannot tell us how much emphasis to put on “experimental” and how much on “therapy” in the term “experimental therapy.” To say that the question is procedural, however, is to say that research decisions are ones that place patients and payers at significant risk; therefore, the decision-making procedure that is used should not be a neutral one. Since the decision calls primarily for a gamble or a risk, the decision makers should be biased in favor of all those who bear the risk, namely Mrs. Carson and the payers. Since Mrs. Carson takes a risk, she has the right to refuse the invitation to participate in the research. But her consent to participate does not produce a right for her to participate. Others, such as a managed care organizations, must pay for her participation. Neutral panels do not take into consideration the financial well-being of the managed care organization. They do not look at the financial limitations of such organizations, which operate within tight market constraints. In short, neutral panelists do not factor risk into their decisions. They try to reduce the research question to a merely scientific one; unfortunately, it is more complex.
In ordinary business contexts, where research decisions place companies at great risk, such panels can be formed, but they are advisory to those who are at risk, namely managers, who must face stockholders or boards of directors. I recommend this business-like approach for research. If it had been used in the area of autologous bone marrow research, the double-blind method would have been followed and much harm would have been avoided. We would have found out sooner that autologous bone marrow as a treatment for breast cancer, while a lucrative source of income for many centers of research excellence, was a blind alley for patients.
It is surely true that the balance sheet of a managed care organization or of a hospital is of little value to a dying patient or his family. The patient sees only his life; he has trouble realizing that other persons are involved in any experimental research he may wish for. He does not see that others depend on the same common pool of resources that his last-chance therapy would draw upon. But the fact that the critically ill person cannot appreciate the importance of such balance sheets does not diminish their importance. The financial integrity of a managed care organization is terribly important to a nurse or a physician employee or an investor and to the other patients who depend on that organization for future medical care. All may depend on the financial health of the organization for income, retirement, or their future health care.
Whose interest should dominate when decisions about funding experimental research are made? The researcher's? The employee's? The investor's? The patient's? No answers are given by moral theory. These persons' interests are incommensurable. Resolving conflicts among them calls for fair procedures rather than common standards that claim to weigh the importance of incommensurable values against an overarching value. The procedure I recommend is to obtain the consent of all the relevant parties. This should produce clinical research where there is a merger of interests among all parties involved. This merger approach not only will require that clinical investigators design research protocols that meet the highest scientific standards of clinical research but also will require that this research contribute to the welfare of both patients and the organization itself. To restrict support for medical research to those protocols that meet those high standards assures us that a steady increase in research development will continue. In summary, managed care must follow its premier decision-making procedure. It must identify research that merges patients' interests with attempts to constrain the overall cost of care.
Let us now turn to my final promissory note. Physicians operating within managed care frameworks are troubled over patients who are suffering from behavioral or mental illnesses, especially those who need chronic care. Because they practice dual stewardship, they are committed to these patients but also committed to the stability of the health care system as a whole. It is difficult to reduce health care expenses for this poorly defined population and also difficult to gather actuarial data that can serve as good predictors of risk. This risk applies to both the patients and payers; the latter suffer the losses associated with employees and patients with mental illness, who are often less productive.
Moreover, social and family factors are crucial within this context, and this further increases the difficulty of diagnosis and treatment. For example, dysfunctional families make treatment more difficult and positive outcomes harder to achieve. In addition, multiple treatment approaches, which are common within mental health, resist “outcome-based” testing. This increases the degree of difficulty associated with managed care's response to behavioral disorders, since paying for what is of uncertain value is troubling, not only because such payment may waste limited resources on what is ineffective but also because behavioral services can easily drain away limited resources from other effective specialty health care services. Furthermore, the notion of “medical necessity,” which deeply affects decisions within more typical medical contexts, is harder to define within the behavioral context. In the absence of this crucial standard, it is not easy to justify either the approval or the rejection of treatment decisions.16
These difficulties may suggest to some that mental illness needs to be separated from the managed care context and funded with a larger pool of “at-risk funds” so that the risks associated with mental illness can be most widely distributed, perhaps by a governmental program in which the high costs of chronic mental illness are spread throughout the society. While this suggestion has many merits, it faces an important roadblock: Other government-funded pools for health care risk, such as Medicare and Medicaid, are themselves turning to managed care strategies to reduce medical expenditures. Government programs for medical risk are, in the final analysis, simply large pools of money that desperately cry out for management in the face of reduced access to funding. This suggests that paying for the risk associated with behavioral disorders by using universal risk pools does not allow us to avoid the task of rigorously managing these large pools of resources. In short, chronic behavioral care, like ordinary health care, is unlikely to avoid being managed. The question is not whether it should be managed, but how it should be managed.
Classic managed care strategies come into play even within this problematic area, and several of these difficult, culture-changing strategies have already been implemented. First, a host of capitation strategies are in place, and there is no significant evidence that these strategies are epidemiologically inferior to un-managed fee-for-service strategies that opened rather than narrowed access to funds for mental health services. Certainly there are anecdotes where less therapy yields fewer than expected improvements for the patient, but such stories do not weaken the force of the statistical data that suggest that increases of conversational therapy (i.e., psychological counseling) after a certain point do not in general yield justifiable returns. Furthermore, using the family practitioner or the internist, rather than a mental health specialist, as the first line of defense against behavioral illness has not been established to be ineffective. With government funding, we could eliminate these capitation strategies on the number of visits to clinical psychologists. But we would need clear evidence that doing so would bring about clinical improvement over large segments of patients—and this kind of evidence is absent. Continuing therapy sessions may be desired by both the clinician and the patient, but the ability to pay for these additional services is limited. This scarcity requires that defenders of less restrictive access to conversational therapies provide evidence that the increased access will reduce clinical morbidity.
Second, the continuing conflict over conversational versus pharmaceutical approaches to a host of behavioral disorders is something that requires further research. But managed care is skeptical of the idea that “more” of something can do what a “little” has failed to do. Managed care is based on the notion that in cases where patients have not responded to minimum amounts of therapy (in this case, conversational therapy), the chances of securing significantly greater improvements with additional sessions has a low probability.
Third, the use of physician case managers in assessing the medical necessity for mental health services is vital, since some behavioral services can reduce psychological incapacity but others are not likely to do so.
What, then, are managed care's underlying approaches to mental illness? There are several, but the one I want to high-light involves skepticism about how much we know about mental illnesses. It is because we do not know how to manage many patients' behavioral problems that the money we spend in treatment yields such limited results. Mental illness may in some respects mimic ulcer treatment during the fifties. Few ulcer treatments had long-term benefits because the underlying causes were unknown and, therefore, symptomatic treatment had limited efficacy. Wealthy people spent fortunes on managing ulcers, but what they paid for was largely ineffective. Poor families spent far less and often got equal results. Why? Money did not help. Knowledge about ulcers was needed, and its absence made money useless.
Spending money, however, can make us feel as if we are doing something, and it can also be an expression of solidarity with the behaviorally disordered. A more effective approach is one recommended by managed care; it is based on skepticism: Begin the long process of gathering outcome data on various approaches to the bewildering variety of behavioral illnesses and continue to closely monitor the clinical efficacies of widely held pharmaceutical and conversational approaches. For physicians practicing dual stewardship within managed care, their commitment to the health care system as a whole makes it inevitable that they have a high degree of skepticism about how much we know about behavioral disorders. This commitment supports the view that the health system ought to resist providing treatments that have not demonstrated statistically significant results even when they express our solidarity and caring toward patients.
Because of my focus on the physician, I have said little about the stewardship duties of managed care organizations or the administrators that manage these organizations or the investors that may own stock in these organizations. These are large issues that need to be addressed at another time. However, a few closing remarks can serve as both a summary of this essay and as an outline for this larger project. Throughout this essay, I have argued that physicians who view themselves as having ethical duties only to the patient are at odds with the new world of medicine. The structure of managed care makes the single-steward physician a virtual impossibility. Managed care logically requires that physicians become dual stewards. There may be times when physicians are unhappy with this modern medical arrangement, but there is little doubt physicians have dual duties that require the development of new skills aimed at balancing the interests and wishes of patients with the financial integrity of the system in which they practice. I believe that the skills necessary to balance these conflicting commitments are emerging throughout medicine. Furthermore, there is evidence that patients can often benefit from the attempt to merge the interests of the organization with the interests of the patient.
However, to claim that physicians have multiple duties to both patients and organizations and then to claim that the organizations and administrators of these organizations only have single duties to their stockholders or their boards of directors would constitute virtual patient abandonment. If the identity of the physician must change, then so too must the identity of the administrator of a managed care organization. Administrators must also share the dilemmas of being a dual steward. For just as the physician must take on the identity of the case manager, I hope to establish in another setting that managed care organizations and their administrators must incorporate into their identities the traditional concerns of the physician.