American Journal of Law & Medicine

Some realism about end of life: the current prohibition and the euthanasia underground.

It is revolting to have no better reason for a rule of law than that so it was laid down in the time of Henry W. It is still more revolting if the grounds upon which it was laid down have vanished long since, and the rule simply persists from blind imitation of the past.

--Oliver Wendell Holmes (1)



This well-known quotation from one of America's foremost judges provides an important admonishment about the role of history within the law. (2) Holmes' admonishment is that, even in common law systems, we should not allow ourselves to become too dependent on legal rules laid down in the past. Legal rules laid down long ago are as likely as newly constructed ones to have flaws in their reasoning or be difficult to implement in real-life situations. Furthermore, even in cases where the rule may have originally served some purpose, it may nevertheless be out of date and therefore have little continuing benefit to a particular legal system. Holmes thought the reason behind the particular rule ought to be of paramount importance. For these reasons, if a legal rule's purpose has continuing validity and the rule continues to provide a reasonable and practicable way to accomplish that goal, then there may be sufficient reason to continue to use the rule. If there is not, then the mere fact that a legal rule has existed for a long period is an insufficient reason for keeping it in force.

Thus, Holmes' point is that historical longevity is not, in and of itself, a sufficient reason for following a particular legal rule. Instead, we should follow legal rules because they serve a certain function. If a long-standing legal rule serves that function, then we ought to keep it. Otherwise, a different legal rule, even if new, is the better option. Even in cases in which an old legal rule continues to satisfy its specified aim, a newer legal rule might be preferable if it can do so in some way that is better (e.g. it is more efficient or does a better job of satisfying the purpose of the rules in question). Of course, historical longevity itself may provide evidence that a specific legal rule sufficiently serves its aim. It may also provide other benefits that a new rule could not. For example, the costs of educating citizens about a new legal rule are either reduced or avoided entirely for old legal rules. Nevertheless, the benefits of historical longevity do not, standing alone, counsel in favor of retaining an old legal rule when other options are available.

This means at least two important things when considering whether an existing legal rule ought to continue to be in force. First, when exploring the option of a new legal rule, we must remain cognizant of the fact that we are comparing two distinct rules--the new one and the old one. Thus, we cannot merely examine whether the new legal rule is likely to be as effective as we hope it will be; we must also consider its effectiveness in light of the current legal rule. In doing this, we may find that the new rule, although not ideal, provides substantial benefits over the existing rule. Consequently, we may decide that it is the better alternative.

Second, it is also critical to any evaluation of competing legal rules to apply the same metric to each when evaluating them. In other words, whatever standard used to explore whether the new regulation is beneficial should also be applied to the existing rule. Otherwise, any resulting comparison is likely to be unhelpful, or, at worst, misleading. For example, if we were to examine a new rule on the basis that it does not result in cases of unwanted action, then we must apply the same standard to the existing rule. If instead we impose differing standards, we may find that one rule unjustifiably fares better. It is thus imperative that when comparing two different legal approaches to an issue, that they be compared on the same basis.


1. The Arguments

With these lessons in mind, I want to consider a particular legal problem. That problem asks whether we should legalize assisted dying in the form of either physician-assisted suicide ("PAS") or active voluntary euthanasia ("AVE"). (3) Of course, many scholarly works examine whether current attempts to legalize assisted dying are likely to be effective. (4) The most common conclusions of such examinations suggest that legalizing PAS and AVE will not prevent cases where patients are coerced or pressured into 'choosing' assisted dying; where patients choose assisted dying without proper understanding and information; or where patients are killed without or against their consent. (5) I do not want to take issue with those conclusions. In fact, I think it extremely likely that if we were to legalize assisted dying either in the form of PAS or AVE or both, such cases would inevitably arise. This hypothesis derives from the fact that there appears to be no rule in any legal system throughout history that has had a total success rate. Even the most stringent regulations coupled with the most extreme punishments are not able to reduce the attempts to break the law to zero. I therefore think that no matter how narrowly tailored or well-crafted a law legalizing PAS or AVE is, it will not eliminate all instances of abuse.

I differ, however, with those who suggest that these findings necessarily counsel against legalizing PAS or AVE. The predominate reason that I disagree with this conclusion is because the argument that legalization will not reduce all instances of abuse to zero is only a useful argument if the current legal rule (that of prohibiting PAS and AVE) can reduce instances of abuse to zero. (6) If the current prohibition cannot do this, and it appears that it cannot, then the question is not whether any particular regulatory scheme can reduce instances of abuse to zero, but which scheme is better able at reducing the number of those instances. In other words, which system of regulating doctor's actions at the end of life reduces the number of abuses--a system which allows PAS and/or AVE, or a system which prohibits both?

2. The Issues

As the cases against legalizing PAS and/or AVE have been made in a number of circumstances, this article will attempt to address, in some sort of systematic way, how effective the current prohibition is at removing or limiting instances of abuse. To do this, it is first necessary to determine what counts as an instance of abuse. Instances of abuse, at least for this article, will focus on two of the main players in cases of assisted dying: the patient and the health care practitioner (who is usually, but not always a doctor). There are other players involved in cases of assisted dying; including family, friends, partners, other members of the health care team and even the community at large. It will be more efficient at this stage, however, to narrow our focus to the vital issues related to the patient and the health care practitioner. Moreover, I will focus on those examples of abuse traditionally leveled against the legalization of assisted dying. These include, in relation to the patient, issues concerning (1) increases in the overall rate of suicide within a community; (2) protection from discrimination; (3) the incidence of non-voluntary and involuntary euthanasia; and (4) whether PAS and/or AVE places the individual over the community. In relation to the health care practitioner, this article will address issues concerning (1) the role of medical ethics; (2) the psychological and emotional effects of participating in an illegal practice; (3) the amount of control exercised by doctors; and (4) the workings of the 'euthanasia underground.'

3. The Data

It is also important to mention the type of evidence available to examine these possible instances of abuse. There is, rather unfortunately, a very limited amount of research available about the actual practices of doctors at the end of life. This is especially true in cases, such as in PAS or AVE, where the conduct is illegal. The evidence that is available for evaluation consists of surveys and interviews done in jurisdictions, such as the Netherlands or the U.S. state of Oregon, where either PAS and/or AVE has been legalized. (7) There have also been a few studies based on the Dutch survey models, but even these only provide limited information--and some of that information is in dispute. (8)

Another source of available information is an interview study done by Roger Magnusson, an Associate Professor of Law at the University of Sydney. Magnusson documented that study in his book, Angels of Death, at which time he detailed a series of forty-nine interviews of professionals in health care and related professions. (9) The interviewees were predominately from Sydney, Melbourne and San Francisco and primarily (but not exclusively) treated or interacted with AIDS/HIV patients. (10) Interviewees were guaranteed anonymity and confidentiality. (11) For that reason, it is not possible to verify Magnusson's data. Nevertheless, the guarantee of anonymity and confidentiality did provide the interviewees with a shield that Magnusson believed allowed them to be more honest about their practice than they otherwise might have been. (12) Additionally, the study comprises a small number of interviews with interviewees that self-selected for participation. (13) For these reasons, any attempt to generalize from the conclusions drawn from Magnusson's study should be subject to some skepticism. This is most important when considering some of the incidence rates, such as the number of interviewees participating in these practices. Since the interviewees were self-selected, it is likely that this rate is not in line with the general population. Even so, the results of the study, including the anecdotal evidence presented to Magnusson, provide us with valuable data about practices by doctors involved in end-of-life decision-making. More importantly, it appears to be the only study of its kind completed in jurisdictions that currently outlaw assisted dying.

As Magnusson's book is not as well known as the Dutch surveys, it is worthwhile to summarize the basic results of the study here. Of the 49 interviews, researchers recorded 203 anecdotes, 105 of which dealt with PAS and/or AVE, while another 34 dealt with suicide. (14) The remaining anecdotes related to patient care or other topics. (15) Of the 105 anecdotes that specifically mentioned PAS and/or AVE, 88 involved the interviewee directly. (16) Although the remaining anecdotes did not involve the interviewee directly, researchers were able to distinguish them from mere rumor or speculation. (17) Thus, even these other anecdotes provided some information that made them more substantial, at least to the interviewer, than the general rumors that may exist about end-of-life practices. (18)

Even more surprising than the number of anecdotes was the number of interviewees directly involved in end-of-life practices. Of the 88 first-hand accounts, 56.8% of the interviewees took a 'hands-on' role in the death. (19) Another 6.8% of the interviewees were active at the scene (i.e. they were not merely present at the scene yet they were not as involved as those taking a hands-on approach). (20) Finally, 21.6% of these interviewees were indirectly involved in facilitating death (e.g. prescribing drugs later used at the scene). (21) To look at the information in another way, of the forty-nine interviewees in the study, twenty-six (53%) took a hands-on role at some time; four (8.4%) were active at the scene; seven (14.3%) participated in an indirect facilitation; while only eleven (22.5%) had no involvement at all. (22)

Additionally, of the number of interviewees who had participated in end-of-life activities, quite a few of them were involved on numerous occasions. For example, the table on the following page, (Table 1), relays the experiences of twelve different interviewees, illustrating their repeated participation.

These individuals included not only general practitioners and hospital physicians, but hospice nurses, community nurses and therapists as well. (23) It is also important to note that the table does not reflect all of those interviewees who told multiple stories. Of course, as mentioned previously, one should not read too much into the figures. It is impossible to determine if these figures are representative of the larger populations of health care providers even within a particular specialty in a particular city. Thus, general conclusions drawn from the figures are likely to be inaccurate. Even considering that point, however, the fact that over half of the interviewees participated in a hands-on role during PAS and/or AVE indicates that the problem of illegal end-of-life involvement is larger than previously thought. (24) With this information in mind, it is now time to turn to the specific issues related to end-of-life practices. As indicated, this discussion will begin by considering those issues that are related to the patient.



One of the primary concerns about any legislation permitting assisted dying relates to whether that legislation would be able to prevent those suffering from depression or people with suicidal tendencies from choosing PAS and/or AVE prematurely.

The first point that bears mentioning is that this particular concern assumes that suicide is a bad option at all times for all people. However, this is far from a universally accepted idea. For example, Margaret Battin, Distinguished Professor of Philosophy and Adjunct Professor of Internal Medicine, Division of Medical Ethics, at the University of Utah, quotes Seneca as saying:

   Living is not the good, but living well. The wise man therefore 
   lives as long as he should, not as long as he can. He will observe 
   where he is to live, with whom, how and what he is to do. He will 
   always think of life in terms of quality, not quantity.... Dying 
   early or late is of no relevance, dying well or ill is. To die 
   well is to escape the danger of living ill ... 
   Just as I choose a ship to sail in or a house to live in, so I 
   choose a death for my passage from life. Moreover, whereas a 
   prolonged life is not necessarily better, a prolonged death is 
   necessarily worse. (26) 

As Battin notes, the point of this particular passage is that the length of one's life is not the issue. (27) The question of whether a particular life has been good is not based on how long that life existed. In fact, it would be strange if we did make decisions about a life's worth on that basis. We do not judge books or movies based on their length. Seneca's statement suggests that the same is true of a person's life. (28) As Battin argues:

   A self-embraced death, a suicide, Seneca argues, is not a 
   premature end to a complete life. It isn't like a journey cut 
   short, which is incomplete because you don't get there; rather, 
   Seneca maintains, a life cut short can still be complete if it 
   has been lived well--you do get there, so to speak; you've 
   actually lived your whole life. (29) 

Thus, suicidal thoughts or even suicide may not, standing alone, always be irrational. (30)

Even assuming suicide, whether rational or otherwise, should not be encouraged, that argument alone provides an insufficient reason to warrant outlawing assisted dying. In fact, it may be that legalizing assisted death would see the overall rate of suicide go down. Additionally, it may be the case that even if the numbers do not decrease, treatment for those seeking suicide would increase (or at least not decline). It is therefore worth considering what effect PAS and/or AVE would have on suicide rates before deciding whether these practices should be legal.

All of this may sound counter-intuitive. How is the addition of PAS and/or AVE, themselves types of suicide, going to reduce the overall rate of suicide? The intuitive response would be that adding an additional form of 'legal suicide' would increase the number of suicides. Thus, is there any reasonable argument to suggest otherwise?

There are several reasons why legalized assisted dying may correspond with a decrease in the rate of suicide. Richard Posner conceived at least one. In Aging and Old Age, Posner presents the hypothesis that allowing PAS and/or AVE may reduce the rate of suicide generally because it affects the cost of committing suicide and the cost of staying alive. (31) Assuming people are acting rationally, they will only choose suicide when the costs of committing suicide are less than the burdens of staying alive. (32) Thus, when making the decision, an individual will consider a number of factors. The concerns in question are broadly based that may include a number of personal factors outside of quantifiable reasons, such as financial costs. In cases of terminal illness, however, the balance may be skewed. For example, if an individual faces a terminal illness, it may be that, at present, the benefits of continuing to live outweigh the burdens; while in the future, there may be a time when the burdens will become too onerous. If the individual recognizes the possibility that when this happens, they will be unable to take their own life, for example, because of physical limitations, this may affect their present perception of the benefit/burden balance of continuing to live. (33) They may therefore decide that future incapacity tips the balance in favor of taking their life now. (34) If, however, it were possible to defer suicide until that later point, the individual would likely choose that option instead. (35) This latter possibility explains why legalizing assisted death may affect suicide rates. When that future time comes, the individual may change his or her mind and forego suicide. Thus, the argument runs that the legalization of assisted suicide could potentially decrease the overall suicide rate because people might wait longer to commit suicide, which increases the possibility that people might change their minds.

At the time of his writing, Posner was unable to provide any substantial support for this argument. (36) He did compare suicide rates from the Netherlands with other European countries, but was unable to reach any firm conclusions. (37) New evidence unavailable to Posner may provide a clearer picture about his hypothesis, although much of it is anecdotal. First, there is evidence from the annual reports prepared by the Oregon Department of Human Services indicating that not all patients prescribed lethal medication actually take the medication. In fact, this phenomenon has been reported in every year of the annual reports. (38) While some incidents may be due to a time lag between the prescription of the drugs and their ingestion by the patient or due to patients dying of other causes, at least some are not. (39) In those latter cases, one possibility is that the patient obtained the drugs in the event that they elect PAS in the future. There is further evidence from Angels of Death that supports this possibility. In that study, there were many instances of people smuggling drugs to a patient for the patient to take at a later date. (40) We thus have some evidence, albeit limited, which has some people acting in ways that appear to be consistent with Posner's hypothesis.

Another way in which legalizing assisted suicide may contribute to an overall decrease in the rate of suicide is the life-affirming value that actively refusing lethal drugs can give to an ill patient. In some cases, patients actively seek lethal prescriptions without necessarily intending to use the medication. Instead, they seek the prescriptions to maintain some aspect of control over their lives. …

Log in to your account to read this article – and millions more.