To the Editors:

This is in response to Richard Horton’s review “In the Danger Zone” [NYR, August 10, 2000]. Mr. Horton’s analysis of John F. Murray’s ethnographic account and meditation on San Francisco General’s intensive care unit is both comprehensive and informative. Intensive Care: A Doctor’s Journal would seem to be a fascinating insider’s perspective on hospital culture and a worthy addition to the genre.

However, Mr. Horton’s piece falters on the facts of the Oregon experience. In his concluding paragraphs, Horton steps outside of his role as reviewer to pontificate. Horton’s statistics (not attributed) conflict with more than two years of data collected and disseminated by the Oregon Health Division’s Center for Disease Prevention and Epidemiology. According to the Health Division, forty-three individuals have used the law since its passage. A patient’s decision to use the law has centered around three core values: autonomy, independence, and quality of life. Women and minorities have not been overrepresented. There have been no reports of coercion, medication failures, or untoward events.

Mr. Horton is of course entitled to his opinions as a physician and citizen but his misinformation does a disservice to the state of Oregon and end-of-life care.

Greg Eddleston
Oregon Death With Dignity Legal Defense and Education Center
Portland, Oregon

To the Editors:

An additional pervasive factor influencing poor outcomes in intensive care units (ICUs) is a perverse set of financial incentives which can result in denial of care or premature termination of care for patients whose best chance of recovery requires intensive care over a prolonged period of time. Such incentives can also increase the chances of medical errors and iatrogenesis due to overworked physicians working in understaffed ICUs. Unfortunately incentives for denial of care and medical errors exist cross-culturally.

In the United States such incentives are provided by managed care, and in nationalized health care systems there is a conflictual dual agency imposed by insinuating that the treating clinician also act as the guardian of the public purse.

Premature denial of care is often rationalized by pseudoscientific guidelines as to “medical futility” or the “hanging of crepe” for the family, by self-fulfilling negative prophecies as to outcomes, which lead to demoralization and a premature “pulling of the plug” in the name of “quality of life.” Medical error is often rationalized by blaming the patient or as being inevitable due to “multi-system problems.”

Harold J. Bursztajn, M.D.
Associate Professor of Psychiatry
Harvard Medical School
Cambridge, Massachusetts

Richard Horton replies:

Dr. Harold Bursztajn’s letter gives an important additional and frequently neglected perspective on the pressures facing those responsible for managing patients in intensive care. Regrettably, I can corroborate his belief that there is a “conflictual dual agency” in nationalized health care settings. Ten years ago in the UK it was common for age to be a crucial determining influence on whether a patient could be admitted to an intensive care or coronary care unit (over seventy, and you were out). Thankfully, those restrictions have relaxed, but difficulties remain. In Scotland, for example, patients are three times more likely to die when levels of workload in intensive care are high. Although I think it is hard to prove cause and effect between workload and death, it seems plausible to assume that poorly staffed and equipped units will have worse outcomes when they are stretched. But how is the public to know which intensive care units are adequately resourced—or not?

Greg Eddleston takes me to task over my interpretation of the Oregon experience. He has a closer understanding of these issues as they affect Oregon residents, and I certainly respect his assurances that all is well with his state’s Death with Dignity Act. However, my reading of the available evidence differs from his in several ways. I hope we can agree that the aim of end-of-life care should be to secure a peaceful and dignified death, ideally with close friends and family in attendance and, when possible, in a patient’s preferred setting. The patient ought to be the central source of authority in decisions about such care. A doctor’s role is to provide advice and support, and to alleviate suffering when all further curative measures have been tried and exhausted. Taken to their logical extreme, these principles seem to make active euthanasia—the direct putting to death of a patient by the doctor, at the explicit request of the patient and with appropriate safeguards—not only justifiable but also the necessary right of every human being. The best-reported experience of active euthanasia comes from the Netherlands. Physician-assisted suicide, in which the doctor simply prescribes a dose of lethal drug, was legalized in Oregon in 1997.

Dutch research, as I noted in my review, raises important concerns about the quality of end-of-life care when active euthanasia is practiced. Events in Oregon should also provoke our anxiety. Mr. Eddleston claims that the figures I give are unattributed. Not so. They are cited (reference 13 in my review) and the medication failures that I discuss are set out in Table 3 of that report.

Mr. Eddleston also argues that there are no “untoward events” in Oregon. In a section in their New England Journal of Medicine paper headed “Problems reported by physicians,” the Oregon researchers raise issues of “unwanted publicity,” “difficulty obtaining the lethal medication or a second opinion,” “difficulty understanding the requirements of the law,” “not knowing the patient,” and difficulties about protecting patient privacy. In recent correspondence to the journal,1 additional concerns have come to light about a probable failure to identify treatable depression among patients seeking physician-assisted suicide. The Oregon investigators agree that “the physicians in our study may not have considered a diagnosis of depression in patients who were dying.” These reported incidents seem like serious “untoward events” to me.

There are two remaining and sometimes contradictory matters that, if I may, require pause for further reflection—namely, patient autonomy and palliative care. Autonomy, although important, is not inviolable. My natural rights are circumscribed by the consequences of my actions for others. Ending a person’s life—by direct lethal injection or indirect lethal prescription—is an extreme interpretation of the doctor’s duty to relieve suffering. One must consider how the benefits of respecting autonomy (satisfying a patient’s stated wishes) might be counterbalanced by potential harm (threatening the trust that the larger community puts in doctors to restore health). If the calculus is that greater societal damage might be inflicted by incorporating active euthanasia or physician-assisted suicide into a physician’s routine practice, thereby diminishing public trust, then patient autonomy might not be the only source of moral authority for a doctor to take into account. The crucial importance of constructing an equation for working out harm and benefit must force us to review every detail of the Oregon experiment—a great deal hangs on its interpretation.

There are further complexities here, largely ignored by advocates of physician-assisted suicide—for example, the wider existential experiences of those with terminal illness. Studies of patients’ understanding of their end-of-life situation reveal that those who request physician-assisted suicide may be responding to a self-perceived loss of personal control.2 In a study of patients with terminal cancer, Ingrid Bolmsjö quotes one individual as follows:

I’m afraid of ending up just lying there, not being able to do anything, and just waiting to die. In that case, I’d rather take an overdose.

Bolmsjö’s conclusion is that “loss of autonomy may also give rise to suicidal thoughts.” The difficulty of ascribing meaning to requests for physician-assisted suicide makes the case for wider availability of palliative care services all the more compelling. I accept that this choice is more difficult: it raises questions about cost, equity of health care provision, and training of many different health care professionals in palliative care techniques (in fact, many of the issues raised by Dr. Bursztajn). But I do not think it can be avoided. In the reported Oregon experience, eighteen of twenty-nine patients who received a prescription for a lethal drug were not offered additional pain relief, treatment for depression, or a hospice referral. Physician-assisted suicide is too often given as a final “cure” for terminal illness without proper exploration of alternative approaches. It allows doctors, perhaps even encourages them through a falsely conceived understanding of autonomy, to unwittingly set aside their caring responsibilities to patients.

This Issue

April 12, 2001