Some patients have no chance of surviving if not treated, but very little chance if treated. A number of medical ethicists and physicians have argued that treatment in such cases is medically futile and a matter of physician discretion. This paper critically examines that position.
According to Howard Brody and others, a judgment of medical futility is a purely technical matter, which physicians are uniquely qualified to make. Although Brody later retracted these claims, he held to the view that physicians need not consult the patient or his family to determine their values before deciding not to treat. This is because professional integrity dictates that treatment should not be undertaken. The argument for this claim is that medicine is a profession and a social practice, and thus capable of breaches of professional integrity.
Underlying professional integrity are two moral principles, one concerning harm, the other fraud. According to Brody both point to the fact that when the odds of survival are very low treatment is a violation of professional integrity.
The details of this skeletal argument are exposed and explained, and the full argument is criticised. On a number of counts, it is found wanting. If anything, professional integrity points to the opposite conclusion.
Statistics from Altmetric.com
Is it ever acceptable not to treat a patient who wants treatment and who will certainly die if not treated? The answer may not be immediately obvious, but upon reflection most people would say, yes. In extreme circumstances, those in which the need for treatment far outstrips its availability, and patients are divided as (synchronically or diachronically imposed) triage would dictate, a patient whose chances for survival are very slim if treated but none if not, might be justifiably denied treatment in favour of those whose chances are better. Other things being equal, both justice and utility point to such a conclusion. A second case in which it has been claimed that it is permissible not to treat is much more controversial. If treatment involves procedures or an outcome that a physician has strong ethical objections to, then, some have argued, it is permissible not to treat, even if death is the result. If saving the life of a woman involves aborting her pregnancy, for instance, and a physician thinks that abortion is very wrong, he need not operate, according to the view in question. Finally, it is arguable that the same conclusion follows if there is no triage and no life threatening pregnancy, but a painless (or painful) death is inevitable in the near future whether or not treatment is administered. Like the second case, however, this third one is not clear. Many bioethicists think that treatment should be administered if the patient or surrogate requests it, even if the chances of it ameliorating the patient’s condition, or saving his life, are zero. For the sake of argument, though, I shall assume that treatment is a matter of physician discretion in cases of this sort. The main argument of this paper will be all the stronger if that is not so.
Some people think there are other reasons, besides those just mentioned, for a yes answer to the question. They think that sometimes the physician may, in fact should, decide not to treat a patient, even if there’s no triage, no problematic pregnancy, no question of not being compensated for services rendered, some possibility that death can be averted if treatment is administered, and certainty that death will occur if it is not.1–3 Similar views can be found in papers by Schniederman, et al, and Miles.4,5 (I shall be referring exclusively to the Brody articles reprinted in Mappes and DeGrazia throughout, however since they contain the clearest and most comprehensive defence of the position.1,3) If a treatment is “futile”, according to Howard Brody, physicians “have no obligation to provide, and patients and families have no right to demand, medical treatment” (Tomlinson et al,1 p 335). It is this view that I will be examining at length here.
“When resuscitation offers no medical benefit,” Brody says:
The physician can make a reasoned determination that a DNR order should be written without any knowledge of the patient’s values in the matter. The decision that CPR is unjustified because it is futile is a judgment that falls entirely within the physician’s technical expertise (Tomlinson et al,1 p 336).
…communication with the patient or family should aim at securing an understanding of the decision the physician has already made [in such cases]. Eliciting the patient’s values or involving the family in the decision in not required. Rather, the discussion should inform them of the medical realities and attempt to persuade them of the reasonableness of the DNR order ([though] this is not to say that the physician should callously override or ignore the wishes of a patient or family that insists on resuscitation) (Tomlinson et al,1 p 337.)
In fact, “the physician has no duty to ascertain the patient’s preferences,” and “there [is] no need for discussion, since the justification for the [DNR] order would not rest on information about the patient’s values or preferences” (Tomlinson et al,1 p 338).
Since Brody is not just talking about cases in which CPR has no chance of saving a patient’s life, this is pretty strong meat. To some people, it is simply physicians usurping life and death decisional power when such power properly resides elsewhere, and doing so with little more in the way of justification than presumptuously waving a banner that says “technical expertise”.
To his credit, Brody later realised that his views are problematic in certain respects. In subsequent papers, he admitted that (a) it’s very difficult to define “medically futile” in any very precise way, that (b) judgments of futility “contain an irreducible value component”, and that (c) “the sorts of values that go into futility judgments are not within the exclusive expertise of physicians” (Brody,3 p 345).
(a)–(c) are certainly plausible, but while (a) is not an especially damaging admission, both (b) and (c) are, at least prima facie.
If judgments of futility are value laden, as (b) alleges, it at least seems that (i) patient or family values, and therefore (ii) patient or family input, are relevant to a DNR decision. After all, it’s the patient who’s going to die if not treated, and one of the primary values in question has to be her life. But if patient or family values are relevant, then judgments of medical futility can’t be made “without any knowledge of the patient’s values”, or without “involving the family in the decision”. It would thus follow that the physician has a “duty to ascertain patient preferences” in such cases. Also, if patient or family input is relevant, as would seem to follow from the fact that patient and family values are relevant, then it’s false that “there is no need for discussion” in such cases. Finally, given these two points, it would be difficult to maintain that a judgment of “no medical benefit” falls “entirely within the physician’s technical expertise”.
These points are reinforced by (c). Even in isolation, (c) suggests that patient and family values should be factored into a judgment of futility, that input, from the patient and or from his family, is relevant to a decision to treat or not to treat, and that the technical expertise of physicians isn’t sufficient justification or proper authority for a judgment of futility.
Brody himself doesn’t draw out the implications of (b) and (c) noted above, much less respond to the objections implicit in them. Still, he could reply that the arguments that (b) and (c) make for the claim that treatment in cases of medical futility isn’t a matter of physician discretion aren’t definitive. This can be seen if the chances of patient survival are zero whether or not CPR is administered. Treatment in such cases is truly medically futile. And since the patient is going to die no matter what, the value that fuelled the prima facie arguments, based on (b) and (c), for the importance of patient and family input—the value of the life—is of little or no moment. Treatment would do no good, even if it would be in keeping with the patient’s or family’s wishes, and even if he or they would feel good as a result of knowing that their wishes would be honoured. In those circumstances, the physician wouldn’t be just wasting but misapplying his skills and time, for he would be deliberately disregarding the professional norm to use his abilities in an effort to do good.
Two explanatory points need to be made in relation to this last remark. First, the goods that physicians pursue can be of many kinds, though the primary ones relate to proper biological function or condition (which is itself responsive to widespread social and personal norms to some extent—for example, the ability to stand, bend over, or coordinate certain kinds of eye and hand movements). Other goods include the aesthetic (for example, cosmetic surgery); the axiological and hedonic (for example, analgesics for the relief of pain); the personal (for example, guided nutrition and exercise for the achievement of an athletic goal, a prescription for birth control, or tubal ligation in order to avoid conception, transsexual procedures), and the psychological (for example, anti-depressants for the lifting of depression). On occasion, these goods can come into conflict with each other. A lesser good would then have to be sacrificed to achieve a greater good, as when, for instance, a limb is amputated to halt the spread of gangrene. More pointedly, a biological good relating to normal function or condition can come into conflict with a personal good, such as the avoidance of conception. When that occurs a trade off of values is inevitable.
Second, to say that the physician pursues the good doesn’t mean that he must always make an effort to secure a good. A competent patient may not want treatment, or prolonging life may just mean more misery. It is to say, though, in a phrase that echoes Aristotle’s general view of human action, that all efforts should aim at the good.
In any case, the above is very much in keeping with (b) and (c), for a judgment of medical futility would then contain an irreducible value component, and the value in question, relating to professional practice as it does, would be accessible to, and capable of being comprehended and appreciated by, those outside the medical profession.
Perhaps that is why, despite the strong prima facie arguments entailed by the admission of (b) and (c), Brody continues to hold what I shall call the discretionary thesis (DT):
In cases of medical futility, it is a matter of physician discretion whether to treat a patient.
In Brody’s own words, physicians “have no obligation … to provide medical treatment” in such cases (Tomlinson et al,1 p 335). So far, though, all that’s been done is that (1) certain arguments against (DT), those based on (b) and (c), have been shown to be less than definitive, and (2) a positive argument, supporting (DT) in a limited range of cases, namely, those in which the patient is going to die no matter what, has been provided. Proponents of (DT) evidently want it to have far wider application than that. They therefore have to provide more in the way of an argument for it.
Brody’s argument is built on the notion of professional integrity—which is not surprising, considering that, as argued above, the best way to respond to the objections based on (b) and (c) is in terms of professional integrity. This is certainly a legitimate and important concept, and Brody is right that “there is a set of treatment decisions over which professional integrity … hold[s] sway” (Brody,3 p 346). To explain the idea of professional integrity, he uses a via negativa; he explicates the concept negatively, by presenting examples of breaches of it. Such breaches occur when a physician:
is required to perform CPR for a patient whose empirical likelihood of regaining consciousness or of being discharged from the hospital alive is less than 1%;
is required to prescribe an antibiotic for a patient who, based on all appropriate diagnostic criteria, has a viral infection;
is required, or offers, to perform a cholecystectomy upon a patient with no detectable disease of the gallbladder;
offers to prescribe laetrile for a cancer patient;
offers to prescribe anabolic steroids for a weightlifter;
offers, or is required, to inject a lethal does of a drug as part of a state execution;
engages in a sexual relationship with a patient (Brody,3 pp 346–7).
“Violation of professional integrity” does not mean “behaviour that violates whatever rules the AMA or a professional medical organisation has on the books, or that is not in keeping with what the majority of practitioners think acceptable”. Such rules or common practice could be arbitrary, on the order of: “Do not practise medicine on Saturdays”. A rule, principle, or ideal has to be a valid, or an ethically acceptable, part of professional practice in order to exemplify professional integrity. Professional integrity refers to de jure rules, principles, and ideals, not de facto ones (though, of course, in most cases the two will coincide). Professional integrity is thus a normative notion, and concerns justifiable standards of conduct as defined by the very nature of medicine itself, considered as a professional practice.
A practice involves a unified body of goal oriented practitioners and their activities, rules, roles, and relations as practitioners. It is the goals of a practice, more than anything else, that help to define the legitimacy and desirability of various relations, positions, actions, and inactions—in other words, that help to determine professional standards of conduct. Medicine, in one of its acceptances, is a practice centred on the human being, and takes as its principal goal the promotion of human welfare in as much as humans are biological entities. But as an institution within larger society, it is also responsive to extramedical norms and pressures, as well as itself having, as a self-constituted and evolving subsociety, its own legitimate internal norms and rules.
Such, at least in brief outline, is my basic understanding of a practice and professional integrity. In Brody’s view, the treatment or conduct mentioned in cases (1)–(7) violates standards of professional integrity, even if it’s requested and desired by a competent, informed patient.
I take it there would be little disagreement about the majority of (1)–(7). In fact, it might even be argued that professional integrity requires more, at least with some of them, than Brody indicates. Case (2), for example, says it is a violation of professional integrity to require a physician to prescribe an antibiotic for a viral infection. That, however, leaves open the possibility that it is permissible for a physician to offer to do so, and permissible for him to do so. “Wrong to be required to do” does not entail “wrong to offer to do” or “wrong to do”. It might be wrong (for me) to be required to teach an overload, but it would not be wrong (for me) to offer to teach an overload, or wrong (for me) to teach an overload. Some might argue that in case (2) professional integrity is violated by any of the three, that is, to be required to write, to offer to write, or actually to write a prescription for antibiotics. Similarly, in cases (4) and (5) it could be, and I think would be, a violation of professional integrity not just to offer to prescribe laetrile or anabolic steroids, but actually to prescribe either, or to be required to prescribe either; and in case (3), a violation of professional integrity to perform a cholecystectomy, as well as to be required to or to offer to perform one. In all such cases, treatment and, by extension, its conceptual kin—requiring treatment and offering treatment—violate standards of professional integrity because treatment would do no good for a patient, and/or disturb, upset, or destroy normal function or condition, and/or promote abnormal or harmful function or condition, and/or aid and abet ethically wrong or illegal activity—for example, the use of anabolic steroids in athletic competition. Any such treatment, or requirement to, or offer to perform such treatment, violates a constitutive norm of medical practice, dictated by the goals that define it. (An additional concern of some importance in relation to these cases is also discussed below.)
Cases (6) and (7) are a little different. In at least the great majority of cases, sexual relations with a patient are a violation of the norms that have to be maintained between physician and patient in order for a properly therapeutic relationship to exist. Therapy isn’t incompatible with romance or excitement—some physicians treat their spouses or lovers—but absent such a prior disqualifying non-therapeutic relationship, the possibility of endangering therapy is all too real. Human beings being what they are, a sexual relationship is not a handshake, or even a friendship. Even more importantly, sexual relations endanger the institution of medicine itself, by sending mixed, confused signals about the nature and bounds of medical practice. The roles and relations of participants have to be clearly defined and delimited, in order for a profession to establish and maintain itself as professional and not personal, and in order for it not to be seen in society’s eyes as even possibly exploitative. Some things are not done, and foremost among them are the sexual.
(6) concerns the proper role of the physician in larger societal affairs, and alleges that the role of physician precludes certain extraphysician activities, namely, those involving the deliberate taking of human life in a non-life threatening situation, as a form of punishment. The idea is that, given the norm of doing good in relation to humans as biological beings, or at least not harming them, activities such as assisting at or performing an execution violate the constitutive norms of medicine, or emblematically project a wrong, harmful, or distorted image of the physician and medical practice.
This view can be questioned on a number of grounds. The demands of justice, based in overarching and general societal concerns, may more than counterbalance an internally generated case against physician assisted execution, and tip the scales in favour of professional permission to perform or assist at an execution under clearly defined conditions. As mentioned above, the internal goals, goods, roles, and relations of a practice are not the only factors that go into a determination of appropriate professional norms. Apart from that, the rationale (or at least the rationale that I have provided) for thinking that performing or assisting at an execution is verboten, is too broad as it stands, as it would also entail that murder (by a physician) is a violation of professional integrity. Murder may be wrong, but it is not a matter of violating the code of a professional practice. Finally, a prohibition on performing or assisting at an execution, almost regardless of rationale, very much suggests that performing or assisting at an abortion would also be a violation of professional integrity, something that many physicians would be loath to agree to.
Brody’s explanation of why cases (2)–(7) violate professional integrity is rather different from mine. His explanation relies on two principles, quoted in the next paragraph. One principle speaks of not causing harm disproportionate to foreseeable benefit, the other of not fraudulently misrepresenting medical knowledge or skill. Despite his remarks on the cases, (Brody,3 p 347) however, it is difficult to see how such principles rule out any of the six as violations of professional integrity, except possibly (2). Case (2) aside, neither disproportionate harm nor fraudulent misrepresentation (at least in any simple or straightforward way) needs to be present in any of them.
The most important case is certainly (1). The argument for (DT) is essentially the argument for (1). Why does Brody think it violates professional integrity to require a physician to administer CPR if the likelihood of a patient regaining consciousness or being discharged from the hospital is less than 1%? Essentially, his argument relies on “two basic ideas”:
Physicians ought not [to] administer treatments that cause harm disproportionate to any foreseeable benefit.
Physicians ought not [to] fraudulently misrepresent the knowledge or skill of medical practice (Brody,3 p 347).
The argument follows directly from this.
The duty (if any) not to perform CPR … has something to do with harm, though the disproportion between harm and benefit might be disputed by the informed patient. But it seems to have more to do with the appearance of fraud. When physicians administer a treatment in appropriate circumstances, they do more than treat the individual patient; they proclaim, in effect, that the treatment is medically indicated for that condition under those circumstances …. [P]hysicians agree that there is a line which cannot be crossed, lest the physicians be unable to distinguish themselves from the proverbial snake oil salesmen …. [There are the twin dangers of the physician being] railroaded into being an agent of harm … [and of the physician feeling] resentment at being forced to perform in a charade, as the phrase “show code” makes clear (Brody,3 pp 347–8).
The first thing to note is that this is really an argument for more than (DT), physician discretion. That is, the argument is not supposed to show merely that the “physicians have no obligation to provide … medical treatment,” but that there is a “duty not to perform CPR”. A duty not to do x means that doing x isn’t permissible, while not having a duty (obligation) to do x leaves open the possibility that doing x is permissible.
In any case, there are really three arguments to consider here, not two. As Brody himself admits, the first, the harm argument, is very weak in the case of a competent patient who knows of the possible harm—presumably, broken ribs, internal bleeding, minor burns, and temporary pain—he is threatened with, and yet judges that he wants CPR even so. He values his life more than he disvalues any such harm, even with the odds of survival stacked against him, and even though the probability that he will suffer harm is enormously greater than the probability that he will enjoy benefit. That is not an irrational decision. Much the same holds if the patient is incompetent and the decision is made for him. As the President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research has recommended, if CPR would benefit an incompetent patient, or even if the benefit of CPR is unclear, and a surrogate favours treatment, then treatment should be administered.6 The first argument, then, is very weak.
The second argument is that performing CPR, or perhaps agreeing to perform CPR, involves fraud, and thus violates professional integrity. It is hard to see how this could be so. Fraud involves deliberate misrepresentation—in effect, lying—or deliberately withholding vital information, and if a patient is truthfully told that the chances that he will survive are less than 1%, fraud has not been committed. In fact, if physicians are truthful, then fraud in any larger sense, fraud in relation to larger society, is also out of the picture. No misrepresentation or deception, no fraud.
The third argument picks up on the phrase “the appearance of fraud”. This is the argument that Brody is really banking on. Unlike the first two arguments, it doesn’t make direct contact with principles 1 and 2. The central ideas are two. The first is the principle that no matter what patients may have heard or thought—that is, even if they have been told the truth—and no matter what patients want, physicians have a duty not to prescribe something of no value, not to engage in worthless treatment, not to raise false hopes, and, perhaps, to positively lead, educate, or train the public not to crave such treatment. This is connected to the second idea, which is that because medicine is a public institution, to treat a patient is, by that very fact, to make a public statement. Treatment is thus automatically of symbolic significance. It proclaims: “This is medically appropriate behaviour in the context”. When the odds of survival are less than 1%, however, that is simply not true. To treat under those conditions, then, is to make a fraudulent misrepresentation, regardless of what the patient has been told. True professional integrity is thus served by not prescribing something of no value, not engaging in worthless treatment, not raising false hopes, and not making fraudulent misrepresentations.
The basic principle here is correct, but although it has something to do with why the performance of a pointless cholecystectomy, the needless prescription of antibiotics, and the administration of a worthless drug like laetrile are violations of professional integrity, I do not see that it has anything to do with the performance of CPR when the likelihood of recovery is less than 1%. The crucial difference is in the numbers. Antibiotics, a cholecystectomy, and laetrile are truly “of no medical benefit”, for the odds of their helping are zero. Not so with CPR. Granted, the chances are it will not help; still, the chances are not zero. Everyone would agree that if the odds of patient survival were 50%, CPR should be administered, so in essence Brody’s claiming that the number, being near zero as it is, is too small to bother with, and that the medical profession has the right to consider that number to be zero, and thus to conflate the case of CPR with that of laetrile. Supposedly, this is based on a professional duty not to do something of no worth, not to raise false hopes, perhaps to educate or lead the public, and ultimately not to make false misrepresentations. But the premises here are unwarranted. Cardiopulmonary resuscitation might, just might, be of some worth; the hope raised is not necessarily false; and the education or training of the public is better done through truth telling on the physician’s part and bitter experience on the patient’s, family’s, and general public’s part. And as for fraudulent misrepresentation: even if to treat a patient is always to make a public statement—a remark I find overly strong, because it exaggerates the prominence and importance, and thus the symbolic significance, of medicine in society—still, the claim that the representation made in the low-odds CPR case is fraudulent is question begging. It is question begging because to say that it is false that “treatment in this case is medically appropriate” is to say that such treatment is medically inappropriate. But whether such treatment is medically inappropriate is precisely what is at issue. It is what is to be proved, not assumed.
There is also something more positive to say than that Brody’s argument fails. It is that more than anything else, it is the patient’s life to lead, and death to die. A concern with professional integrity cannot trump that when there is some possibility, however small, of survival, and the patient wants to take it. It is not for the physician to say that a patient—one out of one hundred, but still a patient—who would have lived will instead die, because of his, and not the patient’s, decision. Physicians should not decide that for them—or decide for the other 99 that they have no chance. In fact, for physicians to do so, for them to arrogate such decisions to themselves just because of their medical expertise, is itself a violation of professional integrity.
My thanks to two anonymous referees for their useful comments on an earlier draft of this paper. Special thanks, however, must go to Pete Alelyunas for hounding me to write up my ideas in the first place.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.