The Physician’s Duty to Treat During Pandemics

Abstract

Physicians assume a primary ethical duty to place the welfare of their patients above their own interests. Thus, for example, physicians must not exploit the patient–physician relationship for personal financial gain through the practice of self-referral. But how far does the duty to patient welfare extend? Must physicians assume a serious risk to their own health to ensure that patients receive needed care?

In the past, physicians were expected to provide care during pandemics without regard to the risk to their own health. In recent decades, however, the duty to treat during pandemics has suffered from erosion even while the risks to physicians from meeting the duty has gone down.

After exploring the historical evolution of the duty to treat and the reasons for the duty, I conclude that restoring a strong duty to treat would protect patient welfare without subjecting physicians to undue health risks.

Physicians assume a primary ethical duty to place the welfare of their patients first, above their own interests. Thus, for example, physicians must not exploit the patient–physician relationship for personal financial gain through the practice of self-referral.¹ But how far does the duty to patient welfare extend? Must physicians assume a serious risk to their own health to ensure that patients receive needed care?

The Spanish flu pandemic presented that question in stark terms 100 years ago, and the question has continued to come up in the years since, resurfacing during the HIV pandemic in the 1980s, the severe acute respiratory syndrome (SARS) epidemic in 2003, and most recently during the Ebola outbreak that began in 2014 in West Africa. According to one estimate, more than 600 civilian physicians in the United States died from the Spanish flu pandemic; nearly two percent of South African physicians reportedly died.² During the Ebola outbreak, hundreds of health care workers in West Africa died.³

On one hand, we have to worry that society might impose too strict a duty to treat in the face of a risk to physician health. The public’s welfare could be compromised if health care providers succumb to disease and can no longer attend to their patients. On the other hand, we have to worry that society might impose too lenient a duty to treat. Physicians, like people generally, may have an exaggerated perception of their personal risk during a pandemic, or they might share a prevalent invidious bias toward a class of persons at high risk for infection. Physicians might be too quick to abandon patients in need. And these are not hypothetical risks. During the HIV pandemic in the 1980s, some physicians refused to treat HIV-infected patients because of their hostility toward homosexuality.

In exploring this issue, I will cite primarily the American Medical Association (AMA) Code of Medical Ethics because it is the most prominent professional code of medical ethics in the United States, it has had the most to say on the duty to patients in the face of personal risk, and it has been adopted as an authoritative source of a physician’s professional responsibilities by many state licensing boards and courts.⁴ The AMA code is important both because it is grounded in fundamental principles of medical ethics and because it can be enforced as a matter of law.

EVOLUTION OF ETHICAL STANDARDS

In 1918, the ethical standard was quite clear—patient welfare should come first even in the face of a serious risk to physician health. According to the AMA’s code, physicians were expected during epidemics to continue their provision of care to patients “without regard to the risk to [their] own health.”^5(pW7) This strong duty dated back to the AMA’s original code, which was adopted in 1847 and which stated that physicians were duty-bound to provide care during epidemics “even at the jeopardy of their own lives.”^5(pW6)

Although some physicians fled to safety in 1918, many physicians remained at their stations. Indeed, whereas the historical record indicates that physicians generally fled during medieval outbreaks of the plague, the Spanish flu pandemic and other 19th- and 20th-century pandemics were generally characterized by “medical heroism.”⁶ Similarly, during the SARS outbreak in 2004, physicians in Canada acted in an “exemplary” fashion.⁷ For the past century and longer, physicians have acted on an ethical commitment to patient welfare in the face of health risks to themselves.

Yet during that same time, ethics standards have qualified their characterization of the physician’s duty to patients during a serious epidemic. By 1957, for example, the AMA’s code simply stated that during emergencies, physicians “should render service to the best of [their] ability.”^5(pW7) And even that limited duty was eliminated from the AMA code as a historical anachronism in 1977.⁵ More recently, especially in the wake of the HIV pandemic in the 1980s, ethics standards have restored much of the duty to treat. According to Opinion 8.3 in today’s AMA code, for example, physicians’ “commitment to care for the sick and injured” requires doctors “to provide urgent medical care during disasters” and this “obligation holds even in the face of greater than usual risks to physicians’ own safety, health, or life.”^8(p127)

But the duty to treat still falls short of the 1918 code’s admonition for physicians to provide care without regard to the risk to their own health. Some consideration may be given to the risk to physician welfare. According to AMA Opinion 8.3,

Physicians also have an obligation to evaluate the risks of providing care to individual patients versus the need to be available to provide care in the future.^8(p127)

The American Nursing Association also counsels care providers to take into account the risk to their own health. Its ethics code states,

Nurses are morally obligated to care for all patients. However, in certain situations the risks of harm may outweigh a nurse’s moral obligation or duty to care for a given patient. . . . Accepting personal risk exceeding the limits of duty is not morally obligatory; it is a moral option.⁹

As a general matter, it seems reasonable to qualify the duty to treat in terms of risk to the care provider. No personal or professional duties are absolute. Thus, for example, the duty not to kill has exceptions for self-defense, and the duty to preserve patient confidences has exceptions for reporting child abuse and other threats to public safety. With regard to the duty to provide care, it would not make sense to expect physicians to treat patients who cannot be saved but would readily transmit their deadly infection to their care providers.¹⁰

But in the face of a serious pandemic, physicians could easily use the exceptions in the AMA code to swallow up the rule, especially because the code does not explain how to weigh the needs of current patients against those of future patients.

RESTORING A STRONG DUTY TO TREAT

Moreover, it is counterintuitive to see a weakening of the duty to treat in an era when advances in medicine make it much more likely that physicians can provide effective care to affected patients and much less likely that physicians will themselves succumb to a new public health pandemic. Although the Spanish flu of 1918 was unusually lethal among influenza viruses, it also is true that the virus spread at a time when there were no flu vaccines, antiviral drugs, or intensive care units. Physicians could do little for either their patients or their colleagues who were infected, other than try to limit the further spread of the virus by isolating the infected and quarantining the possibly infected.

Today, on the other hand, physicians and nonphysicians alike who are infected with a potentially lethal virus can be treated with medication or at least supported with artificial ventilation and other intensive care until the infection runs its course, as happened with many individuals infected with Ebola or the SARS coronavirus. Thus, the SARS death rate for physicians was far lower than was the Spanish flu death rate; in Canada, for example, only one physician died during North America’s most serious SARS outbreak, with two nurses also suffering a fatal infection.¹¹ Laboratory studies indicate that today’s antiviral drugs would have been as effective against the Spanish flu as they have been against more recent influenza strains.¹²

Just as the risks to physicians from viral pandemics have declined substantially, so have the more routine risks of medical practice. Between 1920 and 1940, for example, 10% or more of a medical school class would go on to develop tuberculosis.⁵ With the decrease in occupational risk to physicians both in general and during pandemics, one would expect the duty to treat to strengthen rather than weaken.

Of course, even advanced medical care cannot subdue all lethal viruses, as in the early days of the HIV pandemic. HIV is not a readily transmissible virus, however, so the risks to health care providers were not particularly high even before the development of anti-HIV drugs. In November 2016, the Centers for Disease Control and Prevention reported a cumulative total of only 58 cases of confirmed occupational transmission of HIV to health care workers in the United States and a risk of infection of less than 1% after a needlestick injury involving HIV-infected blood.¹³ One can hypothesize an easily transmitted, highly lethal virus that withstands all that modern medicine can offer, but until such an event actually occurs, it is difficult to see how the risk to physicians can justify limits on the duty to care for patients during a pandemic. The benefit to risk ratio seems to favor a strong obligation to treat.

There are other important reasons to restore a more robust duty to provide care. As we saw with the HIV pandemic during the 1980s, refusals to treat may be rooted in invidious bias rather than legitimate concerns about the risks to physician health.¹⁴ Marginalized groups often are at greater risk for illness; some physicians may be influenced more by the patient’s disfavored status than the risk to the physician’s health. Allowing physicians to take into account the risk to their own health would open the door to pretextual denials of care to unpopular patients.

Moreover, duties to provide care for those in need are difficult to meet when the burden falls on a small number of physicians. For example, all physicians can afford to commit some of their time to unreimbursed care for the poor, but few physicians could afford to commit most of their time that way. Similarly, it is much more feasible for any one physician to assume the risks of providing care during a pandemic when all physicians are sharing the risks. Universal participation minimizes the magnitude of the risk per physician. By contrast, permitting physicians to opt out of their duty to treat and thereby increase the risks to colleagues would encourage more physicians to opt out, increasing the risks even further and leading even more physicians to opt out. Exceptions to the duty to treat can trigger a self-reinforcing cycle of physician withdrawal that ultimately would defeat the duty.

Rather than trying to identify limits on the duty to treat, it makes more sense to ensure that all steps are taken to minimize the risks to physicians and other health care providers from treating patients during pandemics.¹⁵ During the Ebola outbreak, for example, there were times when health care providers did not have access to appropriate protective gear when attending to infected patients. We cannot ask physicians to assume risks to their health unless we do all we can to reduce the risk. Physicians should have access to gowns, gloves, and other necessary personal protective equipment, as well as sterilizers and isolation rooms with negative pressure. We also must take steps to improve the public health infrastructure in less-developed countries, where viral outbreaks spread more widely and cause greater morbidity and mortality to both patients and physicians.¹⁶ It is important as well to ensure that solid financial benefits are available to providers and their families in the event of disability or death.¹⁵

Is a duty to provide care during pandemics consistent with other physician duties to provide care? Of course, doctors do not owe a general duty to provide care to all in need, but we do expect each physician to provide some care to those in need. For example, we expect doctors to provide a meaningful amount of charity care to the uninsured poor, just as we expect all persons to help meet the needs of the indigent. We also expect physicians to provide care in emergencies. When a patient has a dire need for treatment, and the patient cannot shop around for a doctor, we expect the available physician to provide care.

Similarly, for a number of reasons, we expect physicians to provide care even at risk to their own health. When medical students embark on their careers, they understand the risks that they will face. Just as police officers and firefighters accept the risks of their jobs by virtue of their choice of profession, physicians accept the risks of their jobs.⁷ The occupational risks for physicians are by no means exceptional. The occupational death rate for doctors is below the average for US workers and well below that of firefighters, police officers, truck drivers, and construction workers.¹⁷

In addition, the duty to treat is an important element of the social contract between physicians and the public. Society funds much of a physician’s education and grants the medical profession many privileges, including the authority of self-regulation and the right to control the use of medical treatments and technology. In return, physicians assume a responsibility to provide care when needed by the public.⁷

APPLYING A DUTY TO TREAT

Although a duty to treat in the face of risk to self seems clearly justified, how far does it extend? One can suggest a couple of alternatives. At the least, a duty to treat would mean that when a physician ordinarily would provide care to a patient, the duty to treat would preclude a physician from refusing to treat on account of the risk to self. For example, an emergency department physician could not refuse to evaluate a patient who might be suffering from a pandemic flu, and an internist or infectious disease specialist could not refuse to treat the patient. Under this view of the duty, the risk to the physician would not be a factor for the physician in deciding whether to provide care (as opposed to whether another physician would have more appropriate professional expertise for the patient’s needs).

In another, broader view of the duty, it would create more affirmative obligations on physicians. In the case of a major outbreak that taxed a hospital’s regular staff, other physicians would have a duty to join their colleagues in providing needed care. This broader view makes good sense. We might analogize to firefighters who come from other communities to help douse a serious wildfire. As I have said, physicians, like people generally, have a duty to help out in exigent circumstances, and in the case of a health exigency, physicians have a special duty because of their greater expertise.⁷

Even the broader view has limits. Some physicians serve in administrative roles and may not be able to draw on their medical expertise to provide any better care than a nonphysician.¹⁸ For these physicians, it would not make sense to impose a professional duty to provide care, but the general duty of any person to help out in exigent circumstances would remain.

CONCLUSIONS

The duty to treat during pandemics has suffered from erosion in recent decades even while the risks to physicians from meeting the duty has gone down. A strong duty to treat ensures that patient needs will be met, and such a duty would not subject physicians to undue health risks.