7
The Challenge of AIDS

We have discussed the notion of medical ethics in the liberal state by analyzing two sets of common law developments—informed consent and malpractice litigation. In these discussions we emphasized how developments in the common law represented attempts by courts to enforce the integrity of liberal law in order to align the behavior of physicians with the public morality of the liberal state. Thus medical paternalism has been curbed as the practice of medicine has been forced to respect the patient's negative freedom. The lack of respect shown by traditional medical ethics for patient's rights has been cured, at least partially, by judges enforcing the law. In this regard, the liberal society has told doctors that patients must be accorded the respect due every citizen in the state.

In this chapter, we will review aspects of medical ethics that are not defined by the law, but rather by the physician's "duty." While recognizing that medical ethics in the liberal state requires that physicians acknowledge the negative freedom of its citizens, we have not argued that medical ethics should be restricted to the set of duties or obligations owed by each and every citizen of the liberal state to his fellow citizen. To the contrary, a medical ethics for the liberal state involves notions of duty other than those typically owed by one citizen to another. Medical ethics involves altruism. This altruism enhances the sense of cooperation in the liberal state and helps bring to life the public morality of equal respect and concern for others that is central to liberalism. It helps transform the pluralism of liberalism into an "overlapping consensus," and helps create the sense of community without which the liberal state could not function. In short, medical

ethics encompasses some duties that go beyond activities the law can enforce.

Today, one of the best ways to understand the role of altruism in medical ethics is to consider the physician's moral role in an epidemic. Again, allow me to refer to a recent personal experience. A friend was acting as "ward attending physician" for a month and had to go out of town. He asked if I could cover for him, meaning I would have to make rounds with the house physicians on his team. This involves reviewing the patients admitted to a team of interns and residents throughout the day and helping that team develop a treatment plan. In addition, I would have to deal with any problems in patient care that were troubling the team of physicians. This is usually quite enjoyable as the house staff at our hospital arc very intelligent and caring and the discussions about patients arc stimulating. This time, however, my experience was unexpectedly tense.

On rounds in the morning we did not see every patient. While we were talking about the new admissions, the resident physician in charge of the team told me that they were having trouble with Mrs. D. This middle-aged patient had a long history, of intravenous drug use and had been admitted with a blood clot in the left arm. The thinking of the team was that the blood clot had been formed because the patient had injected some contaminated cocaine. They believed the clot was infected, and so they thought that several weeks of intravenous antibiotics would be needed. The patient had repeatedly taken out the intravenous lines in her arms over the course of the weekend, and had left the hospital once to "shoot up." The patient wanted a central intravenous catheter, that is a catheter placed into a large vein of the neck or shoulder. The house staff was hesitant, as these catheters, while in some ways easier to maintain than peripheral ones, can also readily be used to inject illicit drugs. Since chronic intravenous drug users such as this patient usually scar their veins injecting drugs, making it more and more difficult to find veins to shoot into, a central line is often a great opportunity to "get high." While angry about the team's hesitance to put in the central catheter, the patient clearly wanted to continue with therapy: she was not going to sign out "against medical advice," a very common disposition for intravenous drug users. This scenario was not in any way out of the ordinal. What was troublesome was the patient was threatening at times to spit on people, and she carried the human immunodeficiency virus (HIV).

While at a lunch meeting, and before I had met Mrs. D., I was paged and told that there was trouble, and that I should come to the

patient's room. When I arrived on the floor, I was met by senior nursing administrators who told me the patient had to leave the hospital. They explained to me that the patient had spat at a nurse who was helping her and had threatened others. The house staff agreed that they could not do anything for the patient and that she should leave.

The patient, however, did not want to leave. I felt immediately embattled, and I still had not met Mrs. D. Everyone agreed that the patient was competent to make decisions, and that she wanted to stay in the hospital. No one seemed to think she had done anything criminal, and indeed, it was not clear whether she had spat at the nurse, or simply spat when the nurse was near. I related that I could not really do anything until I saw the patient. What I really wanted to do was to get away from the doctors and nurses, most of whom I knew as extremely cool and competent professionals, but who in this case were extraordinarily agitated.

Going to the patient's room, I had to pass an array of security personnel, all outfitted in gloves and safety, glasses. Entering the room, I expected to find some large and threatening figure. Instead I found a thin and agitated woman, instructing someone on the telephone to hire an attorney. Clothes were scattered about, and it looked as though the other patient in the room had beat a hasty, retreat. Mrs. D. finished her conversation and, fixing me with an intense stare, told me she would not leave the hospital. She explained that she had not spat at anyone, and that she knew she needed care. I agreed she needed antibiotics, but told her she would have to calm down. At the same time I realized privately and sadly that this was really an untenable situation for everyone involved.

I knew that the patient could only lose. It seemed clear from our initial conversation that while competent, she could not control her rage. She said, asserting with great anger, that if she did not get more methadone, she would go out and buy dope. However, she had already received so much methadone that she had fallen asleep in the bathroom, precipitating the violent encounter with the nurse. She knew she was sick and needed antibiotics, and was even willing to back off the central line request, but she only wanted to be cared for by certain nurses and wanted to leave the hospital whenever she wished.

Moreover, it seemed to me very likely that if confronted with authority, she would leave the hospital. Having spent at least ten years abusing intravenous drugs, and ten years needing immediate gratification from them at great risk to her health, did not produce patience and conscientious self-regard. If she was confronted, I was sure she

would leave, even though we warned her that her health was endangered by the infected clot. This is a scenario I had seen played out countless times in the past. Indeed, the drug user's lack of self-regard can be used as a tool for turfing, or getting rid of the patient.

This patient had originally sought care at a community hospital. Even though they knew she was addicted to heroin and would suffer withdrawal without methadone, the health care workers she met in the emergency room of this hospital told her she would not receive any narcotics if admitted there. She left, quite understandably, after signing a form that instructed her she was leaving the hospital against their medical advice. I had little doubt that years of this kind of treatment had left this patient with an ironic understanding that it was always easier to leave than confront hospital authority. Thus while she doggedly stuck to her contention that she wanted to stay, I knew a confrontation would eventually drive her away.

A confrontation was inevitable because the health care workers felt threatened. I am generally impressed by the sense of duty that most hospital workers bring to their job. I am familiar mainly with the house and nursing staffs of large urban medical centers, and so this defines my perspective to a large extent. On the wards of these hospitals, human suffering is close at hand. There is usually a certain serious, often grim atmosphere, which is then lightened only by gallows humor. Nonetheless, there is a feeling of comaraderie, a belief that there is an important job to be done, that makes the work seem special. Moreover, one frequently sees compassion and gentleness in caring for vulnerable sick people. The nurses are of course in the forefront in this regard, as it is they who really take care of the patients. And they perform this task, in my mind, with a sense of tragedy, as their important job often goes unnoticed and their grave responsibilities are attended to with little real control over the situation.

Be this as it may, neither they nor the house staff, nor for that matter the staff physicians, want to feel threatened in their work. When they are threatened, it is usually by some patient who is behaving violently. Typically, security or the police can be called to subdue such an individual, and work goes on. However, health care workers now face another threat—more frightening because it is so ill-defined-from patients who carry the human immunodeficiency virus (HIV). They know that HIV infection often leads to death from complications of the acquired immunodeficiency syndrome (AIDS). They see young people dying of this infectious disease. They read that Hepatitis B is a much greater threat, as is death from an automobile ac-

cident. But most have never seen anyone die from Hepatitis B, and they also know that everyone has a relatively large risk of dying in an automobile, but only they, health care workers, have the risk of contracting HIV at the workplace. It is their own special threat, a risk which they will take seriously no matter what the risk calculations show.

This risk hung in the air on the wards as I left the patient's room to talk to the nursing administrators, now gathered in a nearby conference room. A couple of security officers had put on yellow gowns, along with glasses, hats, and gloves. A small crowd of people had gathered a respectful distance from the patient's room. The situation was charged; in the conference room, even more so. The nurses made it clear. This patient could not stay in the hospital. There had to be rules; nurses could not be threatened. When everyone else went home that night, only nurses would be left. She had spat, and she would probably do so again.

I did not see how we could send the patient away. She had no home, no arrangements to go to a shelter. More to the point, she was sick and wanted to stay. I thought if I could buy time, maybe the tensions would settle down and so would the patient, and then the staff. I demanded that we get a hospital attorney on the ward. This was a stalling action, as I knew the legal situation. The patient denied spirting at a nurse, and even the nurse who had been the target could not say that she had been spat at by the patient. Thus there was no way to make even a weak criminal case (if there had been, the patient would have been put into police custody and shipped to the city, hospital). Moreover, the law regarding abandonment clearly requires one to continue to care for a patient who is seriously ill. Our legal duty was to keep her and treat her.

I felt that the same was true of our ethical duty, as might anyone else not involved in the situation. But I could not directly challenge the nurses, because I had seen the patient's rage and I knew the "threat" before them. The attorney gambit looked like it might work, however. The attorney took his time, and when he arrived, he reiterated the abandonment line. The nurses moderated, a bit. There was an opening.

Unfortunately, it closed quickly. I went back to talk to the patient. It was going well. We would forget about intravenous lines and treat her with intramuscular injections. She would get a single room. She would try to moderate her behavior. But then I used the wrong choice of words to explain something to her. Again she flew into a rage. She

threw trays around the room and screamed. Security bristled. The patient said, all right, she was leaving, that was it. She began to pack. The staff seemed relieved. I could not entreat her to stay, perhaps because I knew she would not. The confrontation was too much for her. She left, with an infected arm, and likely a number of other medical problems.

None of this was too unusual, except for the "threat" of her infection. The risk of contracting HIV had lent an extraordinarily urgent mood to the encounter. The relief when the patient left was palpable; it would not have been so if the patient had been some other troublemaker. There was also a certain embarrassment because many of the people present had wanted a sick person to leave the hospital. I felt, and I think others did, a peculiar uneasiness. I am normally rather unaffected emotionally by my work in the hospital, but I felt drained as I went to the elevator to leave. I walked over to the public health school to a meeting, for which I was an hour late. It was a meeting on the staffing needs caused by the HIV epidemic. I blurted out an account of the incident. People were vaguely sympathetic, but one colleague noted dryly that while these problems are new in Boston, in New York they happen every day over and over again. This put my thoughts about the legal and ethical duties to treat HIV infected patients in a new light.

Indeed, the HIV epidemic goes to the heart of medical ethics in the liberal state. Up to this point we have been concerned about the manner in which physicians must learn to respect the negative freedom of patients. Here we must debate the extent to which physicians must willingly give up some of their own negative freedom in order to demonstrate their concern and respect for all patients, including those with AIDS. We will also discuss the manner in which AIDS highlights the nature of just doctoring, which requires that physicians address important issues of public health. In general, the challenge of AIDS allows us to understand new dimensions of just doctoring.

The (Rather Minimal) Legal Duty to Treat

In previous chapters, we discussed the manner in which legal innovations reminded physicians that they needed to seek informed consent and to provide good quality care, as well as evidence

that the care was good. Some might think, having read the case study I just outlined, that the law should also require doctors and nurses to care for all the sick, including those with HIV. This argument may, however, misconstrue the nature of the law as integrity. The law, especially the common law, protects negative freedom, and thus it provides formally equal concern for all individuals. It also ensures that minorities do not suffer at the hands of majorities, by providing for certain substantive civil liberties. Moreover, the liberal legislature decides democratically the manner in which inequalities can be overcome—the overall balance between negative freedom and inequality. Ideally all citizens receive equal concern and respect, but that is only a goal. In any real liberal state, there will be inequalities, even inequalities that conflict with the ideal of equal concern and respect. Thus some people, because of the nature of their disease, might receive less than equal treatment. The liberal state might be unwilling to constrain the negative freedom of others, for instance, health care workers, so that equal treatment is provided. This appears to be the case with the common law and AIDS: the doctor's freedom to choose is respected, even to the detriment of sick people.

The traditional paradigm for establishment of the doctor-patient relationship is the contract.1 Under common law, physicians have been allowed to refuse to care for patients in nonemergency settings. Physicians are also allowed to limit the care they provide to certain kinds of problems or to certain subspecialities.2 However, once a physician begins to treat a patient, a relationship is assumed and care must be continued until the patient no longer needs treatment for this specific problem.3

This does not mean that a physician must continue to treat a patient until the patient decides to go elsewhere for therapy; if the patient is not acutely ill, the physician may end the relationship after giving enough notice for the patient to find a new physician.4 If a physician ends the doctor-patient relationship while the patient is in need of care, the physician may be found to be in breach of the implied contract, and a patient mas, sue under the doctrine of "abandonment."5 Abandonment doctrine thus prohibits a physician from unilaterally refusing to care for a patient once a relationship has been initiated, unless that patient's health is stable and sufficient notice is given of the physician's intention to withdraw from the relationship.

Abandonment doctrine will have a relatively small role to play in assuring access to health care for patients with AIDS. If a physician is treating a patient who has AIDS, he can withdraw, but he will have

to give ample warning and recommend another physician for the patient. Moreover, the physician cannot withdraw if the AIDS patient is acutely ill. AIDS patients are frequently ill, and many times the illness is critical. Does this mean that once a physician enters a relationship with an AIDS patient, he is committed to treatment of that patient for life? Some commentators have answered this question in the affirmative, but I think this demonstrates an incomplete understanding of the clinical course of AIDS as well as the law of abandonment.6 The disease tends to smolder, then it flares up with opportunistic infections. In the majority of AIDS cases there are times in which patients are quite stable and a new therapeutic relationship could be forged without detriment to the patient. Moreover, the care of patients with AIDS is becoming more sophisticated, and primary, care practitioners, for instance, may cite lack of specialist knowledge as the reason for terminating a relationship with a patient with AIDS. Thus the doctrine of abandonment does not prohibit a physician's withdrawal from a relationship with an AIDS patient.

The same is true for a patient who is HIV seropositive. People who carry the HIV may be completely asymptomatic and may not require any intense therapy. While such patients may potentially become quite sick, they are usually not acutely ill, and proper notice can be given with no problem. Nor will the doctrine of abandonment prevent doctors from requesting antibody screening before they initiate care for individuals. Since there may in the future be new drugs available to treat seropositive patients, a primary care physician can argue that the care of seropositive patients also requires specialist knowledge. Thus the refusal to care can be couched in terms of concern for the patient.7

In the case of Mrs. D. cited above, the patient would have had grounds, had we forced her to leave, for an abandonment suit. We were already providing care for her, and she was acutely ill. Thus the common law would have found us in violation of her negative freedom and the expectation of being treated fairly.

However, hers is not the typical case, although it is the most frightening for many health care workers. Physicians studying this paradigm might well decide that they will limit their care to people who do not have AIDS. They can justify this refusal to care on the grounds that they do not understand all the problems that can develop in people with AIDS, and that AIDS for them is like cancer; it is best handled by specialists. Underlying this seemingly prudent decision, however, will be the realization that they simply do not want to have to face the

incredible social problems presented by many patients with HIV infection, or the fear of contracting the virus from the patient. (The situation is even more difficult for nurses, who have more contact with patients in general, and who might decide that some career other than hospital nursing seems more attractive.)

Fears of contracting the virus may lead to real shortages of practitioners available to care for people with HIV infection. Especially as the epidemic deepens and moves into urban communities which have long been underserved, it may be difficult to find enough physicians willing to undertake the small risk associated with caring for patients with AIDS. In this area, the common law of abandonment tends to emphasize the negative freedom of the physician, and it thus does little to correct the inequalities that might develop around care for those infected with HIV.

Of course, the common law is not the only means the liberal state employs to develop legal duties. The law as integrity is not restricted to traditions of judicial law. In addition to the common law sanction of abandonment suits, there are statutory controls over physicians' practice that could serve as sanctions against physicians who refuse to care for seropositive patients. Physicians are licensed by the state, and the state retains some control over the manner in which physicians practice. This power could be used to prohibit physicians from discriminating against patients who are seropositive.

In New Jersey, for example, the state's licensing authority has stated that physicians cannot discriminate against patients with AIDS or AIDS related complex (ARC).8 The state does not, however, require treatment if the physician states that she does not have the skill or experience to treat the disease. Thus, while state licensing authorities may be able to mandate that if a surgeon is qualified to do an open lung biopsy, he cannot refuse to do one on an HIV seropositive patient, the state will not be able to mandate that all primary care practitioners must care for HIV seropositive patients. This means that even laws emphasizing the civil rights of individuals may prove ineffective in guaranteeing the equal concern and respect that is the ideal of liberalism.

Hospitals and other health care institutions are governed by a set of doctrines different from those covering individual practitioners. (Of course, many physicians are employed directly by hospitals, and the doctrines that apply to institutions will apply to the physicians employed there.)9 Hospitals that have emergency rooms must treat all

patients who arrive in unstable condition.10 Unfortunately, this right to emergency care at a hospital that offers emergency services does little to guarantee that people with AIDS will have unfettered access to health care. If an AIDS patient is suffering from a medical emergency, it is clear that he or she has a legal right to care in a hospital with an emergency service; but the law provides little more than emergency care. Once the acute medical problem stabilizes, the hospital and its employees are able to sever the therapeutic relationship.11 However, the federal government, prompted by tile phenomenon of "dumping" indigent patients in county hospitals, has instituted sharp penalties for hospitals that send unstable patients to other facilities.12 While these penalties will help ensure safe transfer, they do not guarantee access to health care at any particular institution. Indeed, it is likely that private hospitals will increasingly send indigent AIDS patients to county or public hospitals.

These provisions will accommodate at least some individuals with AIDS, but they will not assure the kind of continuing primary care that may be the cornerstone of treatment for HIV in the future, especially as more drugs emerge for the treatment of the asymptomatic carrier.13 Indeed, patients like Mrs. D. may continue to find that legal guarantees of emergency treatment are essentially empty promises. Every emergency room physician realizes that a confrontation with an intravenous drug user over the availability of narcotics will drive the patient away. Thus the law can do relatively little, in real terms, to secure treatment for individuals with HIV infection.

The Ethical Duty to Treat

I have argued that in certain spheres of activity, voluntary moral behavior, involving altruistic actions, contributes to the liberal state's sense of community and helps balance the demands of negative freedom with the ideal of equal concern and respect. The cooperativeness of the liberal state, the sense of the state as a community, is brought about not only by the law as integrity but also by concern for others, both at the individual level and within certain spheres of concerted activities.

The principles of just doctoring define medicine as one of these

spheres and constitute the practice of medicine as a moral activity that contributes to the sense of cooperation within the liberal state. Thus it is clear that medical ethics must naturally address the potential inequalities that would occur if we relied solely on the law to bring about care for those infected with HIV. Just doctoring must include a duty to treat. Before defining that duty, however, we should review the nature of the threat posed by HIV to those who care for patients infected with it.

The magnitude of the problem posed by HIV in this country, alone is well known. Since 1981, there have been nearly a hundred thousand reported cases of AIDS. It is estimated that approximately one million individuals carry HIV, and that many of these individuals will eventually develop AIDS.14 Although there are increased rates of infection in African American and Hispanic communities,15 there is no way to know whether or not a person is an asymptomatic carrier without a blood test.16 Thus many patients may be unrecognized HIV carriers.17

Fears about occupational transmission have developed relatively recently. At the beginning of the epidemic, there was little discussion of this issue, probably because the risk of transmission was thought to be nearly nonexistent.18 However, in the summer of 1987, the Centers for Disease Control (CDC) reported three cases of HIV infection in health care workers who were splashed with HIV seropositive blood, a manner of exposure that was previously thought not to be a hazard.19 Soon thereafter other researchers demonstrated that the HIV infection rate was much higher than expected in patients admitted to emergency rooms.20 These reports demonstrated that occupational transmission of HIV would not be limited to needle injuries involving AIDS patients. At about the same time, the first suit was filed by a physician against a hospital in which he claimed that he was exposed to HIV and developed AIDS after a blood tube accident.21 More suits by health care workers have followed.22

These suits have heightened health care workers' awareness of the dangers of HIV as an occupational disease, and there are signs of changes in professional attitudes. Surprisingly large numbers of surgeons support mandatory testing and refusal of surgery for HIV seropositive individuals.23 Very few dentists accept new patients with AIDS.24 Thus, although the risk of occupational infection is still thought to be very low,25 physicians and other health care workers, as well as the hospitals in which they work, may soon begin to limit care.26 Indeed, new data suggest that the risk of contracting HIV for

a medical house officer is greater than the risks for police officers of being shot, or for asbestos workers of dying from asbestos-related disease.27 A recent debate at San Francisco General Hospital provides a disturbing picture of our current situation. Researchers there have demonstrated that the occupational risk of contracting HIV is quite low; nevertheless, the chief of orthopedic surgery at the hospital has advocated a policy of physician discretion regarding elective operations on HIV seropositive patients.28 Thus a sense of disquiet is spreading, retarded only by rational arguments concerning the minuscule risk of exposure if one adheres to the standards of the CDC and Occupational Safety and Health Administration (OSHA) on safety procedures and precautions.29 The risk is real, if small, and the law can do little to force physicians to care for those infected with HIV.

Even if the risk were more substantial, just doctoring would require that physicians not discriminate on the basis of type of illness; indeed, it demands an assurance of equal care for those infected with HIV. But what are the limits of this ethical duty to treat?

This can be answered in part by addressing the ethical approaches to the duty to treat. Some create open-ended duties. Ezekiel Emmanuel has noted, for example, that "the objective of the medical profession is devotion to a moral ideal—in particular healing the sick."30 In other words, the ethical obligation to heal entails treating all sick people. Abigail Zuger and Steven H. Miles have framed the relationship of principle to obligation in a slightly different manner. They argue that the practice of medicine itself requires the physician to act virtuously, to exemplify honesty, compassion, fidelity, and courage.31 Since refusing to care for HIV seropositive patients is without virtue, physicians have an obligation to treat everyone. John D. Arras elaborates on this principle of virtue, noting that "in refusing to treat, physicians violate their own professional commitment to the end of healing."32

Yet, a theory of medical ethics based on classic liberal principles finds few special ethical obligations attending the occupation.33 Agreeing with Robert Sade that the relationship between doctor and patient is contractual in nature, and that the doctor's rights in such a relationship are symmetric with those of the patient,34 a physician can state, "I practice medicine and I find nothing in the enterprise that creates a special obligation to treat HIV-related illness." More to the point, a physician can say to a colleague, "You recognize an ethical obligation to treat, I do not. I argue that the practice of medicine itself

does not create such an obligation. Just as patients are free to choose doctors, I am free to decide whom to treat."

This is certainly the position taken by many physicians and some medical societies who assert that the practice of medicine does not entail treating all HIV seropositive patients.35 When coupled with a willingness to refer HIV seropositive patients to HIV clinics, this kind of behavior is not on face unvirtuous or unethical. These arguments, then, emphasize the negative freedom of the physician, to the detriment of the patient who is HIV seropositive and cannot find care.

Those ethicists and physicians who base their ethical duty to treat in a beneficence model seem to be unable to counter the negative freedom argument made by other physicians. Instead of pointing out its deficiencies, these ethicists appear only to recognize and lament the problems posed by the changing structure of the practice of medicine and the pluralism this creates. They rue the growth of the metaphor of medicine as business and the influences of pluralism on the liberal state. Indeed, it often sounds as though they would like to be rid of liberalism and instead base the state, as well as professional ethics, on heroic notions of virtue.

Medical ethics as just doctoring does not need to turn its back on the (real) liberal state. To the contrary, just doctoring arises out of and is compatible with modern liberalism, and this compatibility is revealed in its arguments about physicians' duties to treat patients infected with HIV.36

Since the just doctoring model of medical ethics takes as its first principle equal concern and respect for the sick, the "patient comes first" model, and since this principle is universalizable as to the class of all patients, it is unfathomable that physicians would refuse to care for patients simply because the patients are HIV seropositive. The history of medical ethics, and every notion of the physician's commitment to the patient, requires studied ambivalence toward the disease when one is called to care. Since the patient must come first, discrimination on the basis of illness is not a possibility.

Moreover, medicine constitutes a sphere within the liberal state, a sphere defined by physicians' altruism. Liberalism is dependent on this and other spheres to help bring about the sense of cooperation that unifies the state. Liberalism itself is based first and foremost on the notion of equal concern and respect. Therefore, even if physicians did not partake at all in the notion of altruism, the liberal state would probably require that their negative freedom be limited so that all

would receive care. The duty, to care defined by medical ethics obviates the need for the liberal state to take coercive steps to guarantee equal concern and respect for all patients regardless of their disease. It involves self-imposed limits on the physician's negative freedom for the good of patients.

The notion of health care as a sphere of moral activity that creates duties beyond those usually expected in the liberal state provides further grounds for the duty to treat. Since all physicians share the same set of duties to patients, moral imperatives are arguably as universalizable to the class of physicians as they are to the class of patients. Thus physicians owe duties not only to patients, but also to other physicians. In this regard, it would be grossly wrong for some physicians to refuse to undertake the risks associated with caring for HIV seropositive patients, forcing others to assume more risks. Physicians share equally in the requirements of the duty to demonstrate equal concern and respect for patients. This means they must all be willing to act altruistically, and to share in the risks presented by the care of those who are infected. Just doctoring prohibits free-riding by some physicians because health care is a particular good with moral principles that apply equally to all health care providers. In short, the social morality of medicine expressed as just doctoring extends beyond the public morality of the liberal state.

The requirements of just doctoring arc not, however, as open-ended as the duties specified by a beneficence model. In a beneficence theory, it is unclear what sort of risk might be too great to expect that physicians would serve unselfishly. More important, the beneficence approach to a duty to treat does not leave any room for self-regarding or prudent action by physicians. It does not allow physicians to argue that they should be compensated if they contract the virus in the line of duty or to consider mechanisms for dealing with HIV infection as an occupational disease.

Medical ethics as just doctoring remains within the confines of the liberal state. That is, while the social morality of medicine extends beyond the public morality of liberalism, it does not undermine it. Therefore physicians do retain some negative freedom. They voluntarily give up some portion of that freedom when they become health care providers and enter the moral structure of the health care sphere. The negative freedom of physicians is thus not extinguished, but it is diminished somewhat by the need to show the greatest possible re-

spect and concern for all individuals, to certify the equality at the heart of liberalism. However, doctors are still liberal citizens and they are not expected to be saints, to use George Annas's expression. Indeed, just doctoring allows physicians to consider, in a realistic fashion, those incentives that will enable them and their colleagues to undertake their ethical duty with as much support as possible.

What docs this amount to? Just doctoring, while reiterating the duty to treat, allows physicians to advocate measures that will support caregivers who do contract the virus. They can submit reforms to the representative democracy for its consideration. These reforms will probably center on means for compensation for HIV-related illness for health care workers. Thus just doctoring requires us to consider the available means for compensation through the law, again emphasizing the close relationship between medical ethics and the law.

It also allows health care workers to consider the various levels of risk that accompany different jobs. Since they face significantly higher risks, surgeons may expect greater protection from infection than other specialists, and may expect better assurance about compensation.

If a health worker contracts the HIV at his workplace, he will likely be seropositive for life, probably will develop AIDS, and could be disabled for a long period of time before dying,37 There will be tremendous costs associated with these accidents, both in economic and emotional terms.38 These are the costs that society must be prepared to shift from the injured party to other pockets.

The costs of accidents have traditionally been shifted from the injured to other "deep pockets" by insurance and the tort law.39 But for a variety of reasons, tort law will not provide much compensation to health workers infected with HIV at the workplace.40

The relative inapplicability of tort doctrine to HIV transmission accidents docs not foreclose the possibility of compensation for the injured worker. In fact, workplace injuries are typically compensated by an administrative approach called "workers' compensation" in most jurisdictions.41 Workers' compensation also has drawbacks as a means of shifting the cost of occupational HIV infection. One big drawback of workers' compensation is that the benefits are inadequate, especially in occupational disease cases.42 A further problem with death benefits, and indeed with all workers' compensation benefits, is that they are tied to the amount the person is earning at the time of injury. This will affect student nurses and physician members of the house staff, who

could expect higher incomes after completion of training. State legislatures can, however, increase compensation levels and create presumptions to overcome some of these problems.

Since workers' compensation is often inadequate, HIV infection also creates a need for health, disability, and life insurance. While more than 75 percent of Americans have some form of health insurance, and many have life insurance, far fewer have disability insurance.43 Most health care workers would not necessarily be covered for all the economic repercussions of an HIV infection. Hospitals could, however, broaden the coverage they offer as terms of employment, and provide health, disability, and life insurance for employees as a benefit. This would seem a prudent step for hospitals to take in the near future.

In return for providing low-cost insurance for health workers, insurers might require some form of testing for HIV antibody. They would fear, as might hospitals, that HIV seropositive individuals would seek health care employment as a result of attractive insurance policies available to workers. Thus, to qualify for an insurance plan, health care workers might first have to submit to testing. Current employees who tested positive would be removed from work that could infect patients but would suffer no loss of salary or benefits. The employees who tested negative would qualify for insurance, as would any new employees who tested negative. Prospective employees who tested positive would not be given jobs that have a demonstrated risk of infecting others. Those who refuse to test would not be subject to job discrimination, but would not qualify for special insurance benefits.

This kind of testing will probably be required to develop a workable insurance scheme for defraying the costs of occupational HIV infection. A plan along these lines seems appropriate if we are to shift effectively the costs of HIV occupational accident. Moreover, it is essential to a system that respects the negative freedom of physicians without diluting the duty to treat, and without diminishing the equal concern and respect owed every patient. It acknowledges the limits of physician altruism, as we must in the liberal state.

Of course, consideration of means of compensating physicians who contract HIV at the workplace must be part of efforts to reduce the risk of such transmission of the virus. Just doctoring requires that every effort be made to develop new means for avoiding transmission, and for ensuring that existing guidelines are careful followed.44 Prevention is a much better path than compensation for obvious reasons.

However, there will be accidents that cannot be avoided by universal precautions, and thus compensation policies cannot be ignored.45

Now some health care workers might argue that in addition to universal precautions, hospitals and health care workers should be allowed to test patients for HIV antibody, whether or not the patient requests testing.46 They argue that they would be able to protect themselves better if they knew the HIV status of every patient. While just doctoring allows physicians to consider means for maintaining their own negative freedom, it does so only if those means do not diminish the goal and principle of equal respect and concern for each patient. Thus any consideration of mandatory testing requires that physicians look broadly at the potential impact of such testing on patients. Once again, just doctoring requires analysis of concepts and issues at some distance from the doctor-patient relationship.

Screening for HIV Infection

One of the great advances in the fight against HIV has been the development of low-cost and accurate tests for the presence of antibodies that the body makes once it is infected with HIV. These tests can tell us whether asymptomatic individuals are infected with the virus.47

Although quite accurate, the test is not problem free. Diagnostic tests are evaluated according to their specificity and sensitivity. Sensitivity is usually not a large concern with HIV tests; however, specificity is very important. Specificity refers to the probability that the test will be negative, given that the disease stare is absent. A test lacking in specificity has a high false positive rate. Although the Point is somewhat complicated, it is important also to discuss the positive predictive value. This is the probability that a disease is present, given that the test is positive. Positive predictive value incorporates issues of specificity, as well as prevalence of a disease state. A specific test can have poor Positive predictive value in a Population in which the disease is not prevalent.48

The tests for HIV have increasingly good specificity. However, the Positive predictive value for the test is quite low in low-risk populations because the prevalence of the disease is so low.49 This means that

any mass screening will be troubled by numerous false positive test results: many people who are not infected with the virus will test positive. Given that most of us are now familiar with the biological and social repercussions of infection with HIV, it is easy to see that significant numbers of false positive results are an intolerable prospect. For this reason alone, mass testing of those admitted to hospitals is a bad idea.

But what if the tests for HIV continue to improve and the false positive rate continues to drop? And what if prevalence continues to increase, especially in certain emergency rooms,50 and at specific hospitals,51 as it has. And what if the estimates of the risk posed by occupational transmission of HIV continue to increase, as they have over the last four years? Is it then reasonable to argue that health care workers and institutions should be able to test all patients, regardless of whether they would freely consent to testing? Can the negative freedom of physicians (the right to take precautions when a patient is possibly infected) provide a foundation for mandatory testing? This is an issue that medical ethics as just doctoring must address.52

Certainly there are many health care workers who believe that mandatory testing is reasonable and appropriate. In Great Britain, for example, many feel that requiring consent for HIV testing is not a desirable social imperative.53 In this country there also appears to be a great deal of testing, at least in some states, of individuals without their consent.54 Many doctors, it seems, would argue that we often perform a battery of tests on individual patients without seeking their specific consent to the testing. Why is HIV any different, they ask, than a simple CBC (complete blood count)?55

The simple answer has been that the risks of a false positive result outweigh the benefits to the person who is tested. In the past we had very few good interventions for treating HIV infection or AIDS. A false positive result would lead to great and unnecessary personal suffering for a patient. Therefore, a risk-benefit calculus led to the conclusion that mandatory testing could not be justified. In light of this, the World Health Organization, the Centers for Disease Control, the American Medical Association, the American Hospital Association, and the Presidential Commission on the HIV Epidemic have all argued against mandatory testing for patients.56

But the assumptions underlying the risk-benefit analysis are changing. Consider that there are now more and more pharmacological agents that look as though they may prove to be useful in the therapy

for HIV infection.57 For instance, the drug AZT is recommended as therapy for those who are HIV seropositive but who do not yet have AIDS given certain other conditions, such as low T4 lymphocyte cell subsets. More experimental drugs are coming down the line. Thus it may be beneficial to treat HIV infection, and thus beneficial to test and treat people early.

In addition, assume that concerns continue to grow about the safety of those who must, as part of their work, come into close contact with individuals who are infected with HIV. These concerns could lead to mandatory testing. Indeed, they have begun to do so. In Missouri, for instance, a new AIDS statute requires HIV testing for all individuals who enter a correctional institution, presumably at least in part to protect correction officers.58 One should expect that similar statutes will be passed in other states, and the slippery slope leads from prisons to hospitals.

In view of these developments, should health care workers encourage mandatory testing of hospitalized patients? From my viewpoint of medical ethics, it seems the analysis must still center on the encumbrance on the patient represented by such testing. Physicians have negative freedom as well, but just doctoring requires that it must be weighed relatively less than the negative freedom of patients. The altruism that makes health care a sphere of ethical activity requires that the patient come first. This does not mean the physician must totally disregard her own welfare, but it does mean that her autonomy is of less concern in the calculus. Thus while we would grant that knowing the HIV status of a patient might afford marginally greater protection for the health care worker, we must weigh and assess the burden of the test results on the patient.

These burdens can be insurmountable. Consider the following case related by Dr. Renslow Sherer:

In 1985, I was the primary physician for a young man whose life was ruined by the inappropriate disclosure of a positive human immunodeficiency virus (HIV)—antibody  test. A physician ordered the test without consent and no-tiffed the local health department of the positive result. The health department notified the individual's employer and he was promptly fired. These events became common knowledge at his workplace and in his rural Midwestern town and he was shunned. His landlord asked him to move. Ten days after testing, the life he had known for the past ten years was permanently ruined and he left town. With the loss of his job came the loss of health insurance and insurability; he has been unable to obtain health or life insurance since then.

In this case, no purpose was served by obtaining the HIV-antibody test. The patient had been diagnosed with acquired immunodeficiency syndrome (AIDS)-related complex which has a 95 percent correlation with HIV infection six months earlier at Cook County Hospital. He was aware of his diagnosis and its implications. He had been following safe sex guidelines for the preceding 18 months and had never donated blood or semen.59

This passage makes clear the potential devastation of an HIV test. Let us look at the effects in some more detail. First, it is obvious that to be known as an HIV carrier is often to be stigmatized. Since many people are inordinately fearful of infection with HIV, their knowledge of one's status can lead to unreasonable reactions. The patient easily may become isolated from friends and face social exclusion. This can and does lead to depression and increased rates of suicide among individuals who are HIV seropositive.60

Second, the dissemination of the information can cause the loss of essential insurance policies. Most insurance companies want to reduce their expenses from AIDS-related claims. Therefore, they are unwilling to write health policies for individuals who are HIV seropositive. In addition, there are constraints on the availability of life insurance for people who are HIV seropositive.61 Thus an individual who tests positive through mandatory screening at a hospital may find that he loses both life and health insurance, as did the patient described by Sherzer. Those who are HIV seropositive may also face discrimination in their workplace. There have been many examples of people who have been harassed by coworkers, or even fired by employers, because their status as HIV carriers became common knowledge.62 Carriers may also face housing discrimination.

These problems tend to pale, however, compared to the threat of quarantine. There appears to be both public and academic support for coercive public health intervention for those who are HIV seropositive. Some have now recommended quarantine as a solution for "recalcitrant" individuals who carry HIV.63 While these proposals concern only those people who repeatedly endanger others through sexual contact or through sharing needles for intravenous drug abuse, there are no bright line definitions for the term "recalcitrant." Thus proposals regarding limited quarantine would soon lead to broader use.

Of course, quarantine, discrimination in housing and at the workplace, and even, to a certain extent, loss of availability of insurance are unacceptable burdens on the liberty interest of individuals who are HIV seropositive. In the liberal state, which values the concept of

pluralism and the ability, of each citizen to set his or her own agenda within the sphere of negative freedom, such encumbrances are intolerable. Indeed, in our country, we have a series of constitutional protections designed to prohibit such encumbrances.64 The Fourth Amendment of the United States Constitution explicitly protects citizens from unreasonable searches and seizures. This seems to include searches and seizures of one's body and thus, presumably, laboratory testing.65 In other words, our liberal Constitution prohibits mandatory testing against an individual's will. Given that just doctoring must conform with the public morality of the liberal state, it must be unethical for physicians to recommend mandatory testing. Even more important, since just doctoring involves an altruistic commitment to the patient's best interest, and since the individual's best interests may very well not be served by mandatory testing, there seems to be no basis within medical ethics for anything but opposition to mandatory testing of patients.

One may counter that the best interests of the patient might be met by early treatment of HIV infection, and that this early treatment can occur only if we know the patient's HIV status. Of course, the rational approach to accomplishing this goal is not mandatory testing. It is, rather, careful education of the patient about the options available for therapy. Once the physician explains the benefit of early treatment (and thus the benefit of testing) to the patient, and the patient refuses, then the liberal state is served only by respecting the patient's decision. In our discussion of informed consent, we concluded that the informed patient should be able to make decisions regarding medical therapy. An individual who wants therapy for an HIV infection and who is well educated about the benefits of this therapy can undergo voluntary HIV testing. Those who do not wish to know their HIV status because they are unconvinced of the efficacy of therapy, or because they fear the potential constraints on their freedom should others acquire knowledge of their HIV status, should be able to refuse testing. Medical ethics as just doctoring supports this notion and must therefore be opposed to mandatory testing.

One further argument in favor of mandatory testing is that many of the problems associated with testing can be obviated if rest results are kept strictly confidential. Thus no one loses insurance or faces discrimination if their test results are positive. The problem with this is that no one can be assured that results will be kept strictly confidential. Moreover, physicians must be concerned that even though they prom-

ise patients that the test results will be kept confidential, the state might decide otherwise. Indeed, it is conceivable that a majority might some day support the publication of such results in a liberal state. We must, as just doctors, be concerned for the welfare of the minority, especially when this minority consists of all individuals who are HIV seropositive. Therefore, we must be aware that even though confidentiality appears to be assured, there are no airtight guarantees. Confidentiality of results alone does not provide grounds for mandatory testing.

Many of these same arguments apply to converse testing proposals, that is, proposals that all physicians should be screened for HIV. While the chances of an HIV infected health care worker transmitting the virus appears to be low,66 one such case has been reported by the CDC.67 Given patients' informed consent rights discussed in the previous chapter, one could make an argument that the liberal state could require testing, especially of physicians involved in serious invasive procedures.68

Medical ethics as just doctoring does not create an ethical duty to mandate screening. For many of the same reasons cited above regarding testing of patients, mass screening of health care workers appears inappropriate. Physicians retain their liberty rights, and without clear and rather substantive benefit to the patient, these should not be forfeited. Individual physicians may choose to be tested, especially if they are at high risk, and are often involved in potential percutaneous transfer of blood products with patients. They may also choose to disclose the results to patients. But, until there is evidence that there is some measurable risk to patients, any form of screening of physicians and nurses is inappropriate. Of course, this does not mean there is symmetry in the justifications for patient versus health care worker screening. Indeed, the altruistic commitment of physicians to patients may mean that a certain level of risk would justify screening for physicians, but not for patients.

These are issues all health care workers must begin to debate, and for which they must help develop policies. Just doctoring requires public policy development by health care workers and the public. Medical ethics in the liberal state must consider and elaborate propositions on various issues that were previously considered matters for courts or for governments. Health policy, issues are an important part of medical ethics. Thus it is appropriate for physicians to consider the

effects of discrimination on their patients. They should understand that their concern for patient welfare must extend beyond the disease process and into social implications of disease. Moreover, medical care, especially under the conditions of an epidemic, butts up against public health issues. Here again medical ethics moves beyond the individual relationship between doctor and patient and addresses social and political issues. For instance, consider the question of whether the physician has a duty to warn individuals who may be put at risk by contact with the physician's HIV seropositive patient.

Duty to Warn and the Public's Health

Public health is an enterprise based on equality. It considers the individual's negative freedom far less than it does the broader concern of how to procure the best possible public hygiene. It addresses the health needs of citizens as a group. For example, public health officials curb the negative freedom of individuals to dispose of wastes as they see fit in order to ensure that the danger posed by those wastes does not fall selectively on one group of people, that is, those living near the chosen dump. The public health requires use of the police power of the state to protect innocent parties. Public health officials also attempt to minimize public health problems overall through the police power of the state. For instance, they attempt to decrease use of fluorocarbons in order to halt the destruction of the ozone layer, a public health problem that affects all the members of a society.

One can imagine several ways in which public health officials might employ the power of the state to ensure that the burden of the HIV epidemic does not fall selectively or unfairly on some individuals, or to curb the overall spread of the epidemic. Consider, for instance, ways to slow the transmission of the HIV. One approach, long used by public officials, would be to undertake contact tracing.

Contact tracing is the practice of identifying individuals exposed to communicable diseases.69 It can mean assistance for voluntary efforts to trace one's contacts; it can also mean aggressive tracing of infections through health department interviews.70 The public health officials

identify the contacts and warn them that they may have been exposed to the virus. These individuals can then choose whether or not to be tested to ensure that they are not also silently spreading the disease.

While contact tracing has always relied on voluntary disclosure and respect for confidentiality, some argue that AIDS hysteria might create pressures to change the traditional public health approach. They fear that infected patients who refuse to disclose contacts might be subject to penalties and their identity as HIV carriers disclosed. Moreover, they fear that since contact tracing involves public health authorities, it might open the door to further use of the police power of the state and ultimately to quarantine. Thus civil rights organizations and the gay community are profoundly opposed to systematic, aggressive contact tracing by public health authorities.71

Civil libertarians also argue that the efficacy of contact tracing is unknown. While contact tracing appears to have worked well and cost-effectively in a small trial in South Carolina, there are reasons to believe this will not be the case elsewhere.72 Given the long latency between infection with HIV and the onset of symptoms, the difficulties in identifying contacts in urban populations, and the previous failures of contact tracing to curtail sexually transmitted diseases, such as Hepatitis B, which are not readily treated with antibiotics, it appears that the benefits of contact tracing are questionable.73 Given the potential costs to society of a policy that may move us toward quarantine, it seems that medical ethics as just doctoring would take a stance against this kind of public health measure.

Yet the goal of contact tracing is undeniably good, that being to curtail the spread of the virus. In the pursuit of contacts of HIV infected patients, public health authorities demonstrate their concern and respect for those who have been put at risk—a concern the carrier may not have shown. Just doctoring shows regard for seropositive patients by pointing out that they may be subject to discrimination through contact tracing policies. These tensions come about frequently in the liberal state and must be managed through accommodation and compromise.

Physicians, following principles of just doctoring, should be willing to help bring about the values embodied in policies like contact tracing, yet also provide highest regard for the patient's negative freedom. One way to do this would be to have private physicians warn third-party contacts of HIV seropositive patients rather than employing the state to do so. If physicians were to take on this task, there would be

less danger that state power could be used against individual patients who are HIV seropositive, which would help protect the patient's right to privacy. Since the "patient comes first," it makes sense that physicians who are "just doctors" would be willing to warn third parties as an alternative to contact tracing by the state. However, this proposition requires that just doctoring be accommodated with existing laws about doctor-patient confidentiality.

The law regarding confidentiality is quite complicated. The common law did not recognize that doctor-patient communications were privileged.74 Rather, the confidentiality of these communications was created by state legislatures.75 By 1983, all but two states had enacted statutes providing some kind of protection for medical records and communications; this protection was referred to as the doctor-patient privilege. These statutes are usually written with a number of exceptions to the privilege for issues such as reportable venereal diseases, battered children, and workers' compensation claims.76 In some states, courts have argued that these "public policy" limitations create an exception that allows warnings for innocent third parties,77 although physicians can be held liable for groundless breaches of confidentiality.78

With regard to contagious diseases, courts have tended to find a general duty to warn on the part of health care professionals.79 This common law duty to warn in contagious disease cases is buttressed in some states by statutory public health laws that require warnings for contacts of contagious individuals. It would appear, then, that common law and statutory law generally support confidentiality and certainly frown on groundless breaches of confidentiality, but they do require warnings for innocent third parties who may be endangered by the contagious disease of a patient.

The legal duty of physicians to warn those endangered by their patients has expanded generally as a result of cases concerning psychiatric patients, the most important of which is Tarasoff v. Board of Regents of the University of California, in which the California Supreme Court found that a psychotherapist had a duty to warn a third person whom his patient had threatened to kill.80 Tarasoff has had a tremendous impact on the practice of psychiatry as well as the law governing it. The California court premised the duty to warn on a finding that violent behavior toward a particular individual was foreseeable. The emphasis on foreseeability has diminished as courts have found a general duty to warn, even when the victim is a stranger to the patient.

The importance of Tarasoff and its progeny cases in determining whether a health care worker has a duty to warn contacts of HIV seropositive patients cannot be understated. Using an analysis similar to that used in several post-Tarasoff cases, a court could rule that health care workers are liable for failure to warn not only known contacts of seropositive patients, but any possible contacts. Thus Tarasoff and the common law it has engendered create a potentially very. broad duty to warn in HIV cases.

At the present, however, the common law principles that emphasize warning are constrained by statutory law regarding the confidentiality of HIV antibody status. The states in which the prevalence of HIV seropositivity is greatest have also been the states that have recognized the importance of confidentiality.81 In the mid-1980s, these states tended to buttress the privacy of HIV antibody testing results with statutory requirements concerning confidentiality. In California, any disclosure of a patient's HIV antibody test results to a third party, has been punishable by imprisonment for up to a year or a fine up to $10,000.82 In Massachusetts, the state law is clear that no health care worker can disclose the results of the antibody test without the written consent of the person tested.83 Statutory requirements for confidentiality of HIV test results in most states with large numbers of AIDS cases would thus appear to insulate health care workers from a duty, to warn.

Nevertheless, many of these states' legislatures are now considering amending their blanket confidentiality, provisions to allow individual physicians to warn. The Association of State and Territorial Health Officials is considering the same kind of exception to its endorsement of confidentiality of HIV test results.84 California has already amended its strict confidentiality provisions to allow a physician to inform the spouse of an HIV-infected individual. The New York State legislature has passed a bill that allows physicians to warn third-party contacts and also allows doctors to gain the assistance of trained public health counselors.

It would seem that medical ethics as just doctoring would endorse third-party warnings by physicians as an alternative to contact tracing. The duty to warn allows the physician to consider the exigencies of the situation with the particular patient. Moreover, the physician can weigh the potential restrictions on the patient created by breaking confidentiality against the risk to the innocent third party. Perhaps more important, a physician warning a third party forestalls the need for involvement of the state, and the potential for abuse of civil lib-

erties that state involvement may entail. In short, just doctoring requires that physicians play a public health role if that role serves the goals of the liberal state better than does any alternative. In this way, physicians can accomplish the good of protecting innocent third parties without threatening patients' civil rights.

In summary, if medical ethics is to define appropriate public health responses to combat the spread of HIV, physicians must think broadly about their duties to patients and the relations of those duties to other legal and policy matters. Of course, once we begin to consider the relation of the public's health to the principles of just doctoring, we must discuss the issue of public support of initiatives that will bring about better health.

Public Expenditure on AIDS Treatment

As we discussed in the first chapter, the liberal state tends to be based on a market distribution of goods, as the market allows each individual to decide how to use his or her resources. This freedom of choice will in theory bring about the greatest possible equality, of concern and respect for the individual members of society. However, the liberal allows modification of the market in order to overcome the inequalities that develop in any real market economy. If necessary, the liberal will accept a broad welfare state if the market proves to be too inequitable. Public health expenditures are examples of this kind of equality-promoting modification.

Again we return to AIDS, and to the patient I discussed at the beginning of this chapter. The patient was impoverished, having spent most of her money on illicit drugs. She was also HIV seropositive. She was thus a candidate for several types of intervention. First, we would want to begin to measure certain subsets of her T lymphocytes in the blood. We could also measure levels of beta₂ -microglobulin and p24 antigens in order to predict the course of the disease. These tests would also tell us when to initiate therapy.

Therapy for HIV infection is in the early stages of development, but there do appear to be some important interventions. One is AZT, a drug that has been shown to stabilize immunological status in those infected with HIV. The data on zidovudine is so encouraging that the federal government now recommends treatment with this medication

for all individuals infected with HIV who meet certain clinical criteria. Other drugs, for example, imuthiol, are now undergoing trials. We are also trying to treat some of the opportunistic infections associated with AIDS, such as pneumocystis carinii pneumonia (PCP), with early intervention and prophylaxis. All of these therapies might be indicated for my patient.

Yet she has no money to purchase them, and they are expensive. Therapy to treat the HIV state (antiretroviral therapy) might cost up to eight thousand dollars per year. Physician visits and serological tests could amount to over five hundred dollars. Prophylaxis against PCP would be another two thousand dollars per year. Thus the costs of early intervention across the board might range between two and eight billion dollars per year for the United States alone.85 My patient clearly could not have afforded these potentially valuable interventions.

Since health care is a sphere of ethical activity within the liberal state, and since it provides support for the liberal state's commitment to equal concern and respect, I suggest that just doctoring requires public expenditures for the treatment of patients who are penurious. While recognizing that this will decrease the negative freedom of those who will have to pay more taxes, physicians arguably should support these expenditures in light of their commitment to altruistic caring. Denial of treatment because of inability to pay offends any sense of justice or morality that issues from the notion of medical ethics I have described.

Of course, the liberal state is often unwilling to circumscribe further the negative freedom of its citizens in order to fund more public programs. Some would argue that "just doctors" must realize that there are other goods in society that need funding besides medical care. They would assert that it is naive of doctors to expect society to bankroll physicians' sense of justice. There must be some limits on care, in other words, and just doctoring should address these limits. Thus it is appropriate for us now to leave AIDS and to turn to the issue of limits on care and just doctoring.

It is also appropriate at this point to reemphasize the fact that much of my discussion in these chapters is not descriptive but prescriptive. I do not believe that many, let alone all, physicians conceive of medical ethics as just doctoring, at least in the way I define it. My description of just doctoring is meant as an argument to physicians and other interested parties. It is my conception of the best structure for medical ethics.