Вы находитесь на странице: 1из 16

Journal of Medicine and Philosophy, 33: 321336, 2008 doi:10.

1093/jmp/jhn018

Kant and Therapeutic Privilege


Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

CHRIS BROWN
National University of Singapore, Singapore

Given Kants exceptionless moral prohibition on lying, one might suspect that he is committed to a similar prohibition on withholding diagnostic and prognostic information from patients. I conrm this suspicion by adapting arguments against therapeutic privilege from his arguments against lying. However, I show that all these arguments are importantly awed and submit that they should be rejected. A more compelling Kantian take on informed consent and therapeutic privilege is achievable, I argue, by focusing on Kants duty of benecence, which requires us to aim at furthering others ends. But I show that there are some cases in which furthering a patients ends requires withholding material medical information from her. Although I concede that these cases are probably quite rare, I conclude that the best Kantian thinking agrees with that of therapeutic privileges advocates. Keywords: informed consent, Kant, lying, therapeutic privilege I. INTRODUCTION In the United States, physicians are expected not only to seek consent from their patients before treating them but also to ensure that this consent is informed by an adequate understanding of the diagnosis, the various treatment options, and their respective risks and advantages. Medical law and codes of medical ethics emphasize patient autonomy, and doctors can be censured, decertied, and even prosecuted for failures to obtain informed consent.1 Whereas the disclosure of diagnostic and prognostic information was once regarded as optional for physicians, it is now a matter of legal and ethical debate whether nondisclosure is ever permissible.2 Arguably, there

Address correspondence to: Chris Brown, National University of Singapore, Department of Philosophy, Faculty of Arts and Sciences, 3 Arts Link, Singapore 117570. E-mail: cabrown@ u.arizona.edu.
The Author 2008. Published by Oxford University Press, on behalf of the Journal of Medicine and Philosophy Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org

322

Chris Brown

would be no such debate were it not for the fact that people tend to be made anxious by bad news about their health. In some cases, however rare, it can be foreseen that the psychological distress brought on by full disclosure would worsen the patients condition and/or severely undermine her ability to deliberate about treatment options. And many consider it permissible for physicians to withhold information from their patients in cases of this sort, even though these patients consent then could not be called informed consent. This alleged exception is called therapeutic privilege and will be the focus of this paper. Opponents of therapeutic privilege might reasonably presume that they have a strong philosophical ally in Kant. For one thing, respect for autonomy is central to Kantian morality, and withholding material information from a patient might seem incompatible with respecting her as an autonomous agent. Moreover, given Kants insistence that lying is never morally permissible, one might suspect that he is committed to an absolute moral prohibition on all alternatives to straightforward truth-telling. In fact, however, these aspects of Kants moral theory do not help the case against therapeutic privilege. As Barbara Secker (1999) has shown, Kants notion of autonomy is importantly different from the notion of autonomy standardly presupposed in bioethics. Kantian autonomy is just the capacity to freely choose to live according to moral principles, whereas the bioethical notion of patient autonomy concerns the agents capacity/authority to decide which treatments will be provided for her. The connections between these two kinds of autonomy are by no means straightforward. In any case, withholding diagnostic or prognostic information from a patient for the sake of therapeutic ends is compatible with respecting her capacity for free moral choice. Objections to therapeutic privilege therefore cannot be grounded on a moral requirement to respect patients Kantian autonomy, so I will ignore this approach. Appeals to Kants philosophical thought on dishonesty, on the other hand, are not so easily dismissed. For, as I will show, his arguments against lying can be adapted into arguments against medical nondisclosure. So, there is a genuinely Kantian case to be made against therapeutic privilege. But I argue that it is not a good one. The adapted arguments inherit some important defects from the originals, and all of them should be rejected. Finally, I turn to the Kantian duty of benecence, which requires us to aim at furthering others morally permissible ends. I argue that, in the vast majority of cases, this duty by itself requires physicians to fully inform their patients and then to seek their consent, since this is most often the best way to ensure that the treatments administered are consistent with the patients ends. This, I suggest, provides a compelling account of the moral importance of seeking informed consent. But I also demonstrate that, in special cases, fullling this duty requires the withholding of diagnostic or prognostic information.

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

323

II. KANTS ARGUMENTS AGAINST LYING Kant presents two distinct arguments against lying, understood as intentionally making statements one believes to be untrue. On the one hand, he argues that, since a maxim prescribing the intentional statement of a falsity cannot hold as a universal law of nature, such statements are not morally permissible:
Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

When the maxim on which I intend to give testimony is tested by practical reason, I always consider what it would be if it were to hold as a universal law of nature. It is obvious that in this way everyone would be necessitated to truthfulness. For it cannot hold with the universality of a law of nature that statements should be allowed as proof and yet be intentionally untrue. (CPrR, 5:44)3

I will refer to this as the universality argument. A maxim, for Kant, is a principle that prescribes an action, and his famous categorical imperative (in its rst formulation) commands that we should act only on those maxims that we can also will as universal laws (GMM, 4:1921).4 Kant holds that we can apply the categorical imperative to determine all of our moral duties. From the fact that some maxim cannot hold as a universal law, it follows (according to Kant) that it is impermissible to act on it. It is in this way that Kant purports to establish certain moral prohibitionsin this case, a prohibition on lying.5 On the other hand, Kant also argues that telling lies harms humanity generally because lying undermines the bases of all rights. In his famous On a Supposed Right to Lie from Philanthropy, Kant argues that even when one cannot evade answering a question, one still is not morally permitted to be untruthful:
Truthfulness in statements that one cannot avoid is a human beings duty to everyone, however great the disadvantage to him or to another that may result from it; and although I indeed do no wrong to him that unjustly compels me to make the statement if I falsify it, I nevertheless do wrong in the most essential part of duty in general by such falsication, which can therefore be called a lie (though not in a jurists sense); that is, I bring it about, as far as I can, that statements (declarations) in general are not believed, and so too that all rights which are based on contracts come to nothing and lose their force; and this is a wrong inicted upon humanity generally. (Kant, On a Supposed Right to Lie from Philanthropy, 8:426; Gregor [1996], 612.)

This argument, which I will call the contract argument, makes no appeals to the nonuniversalizability of any maxim prescribing a lie. Rather, what makes lies impermissible, according to this argument, is the adverse effect they have on the believability of statements in general, and hence on contractually based rights. The idea is that if contracts are to be binding, the statements on which they are based must be believed. But if it is not the case that statements generally are believed, contracts will not bind, and so the rights based on contracts will have no force.

324

Chris Brown

It is worth noting the implications of Kants view. In his contract argument, Kant asserts that truthfulness is a duty no matter how disadvantageous it may be to oneself or others. He famously uses this to argue that one is not morally permitted to lie even to a would-be murderer who inquires about the whereabouts of his intended victim (Kant, On a Supposed Right to Lie from Philanthropy, 8:427). Kant, of course, has not explicitly addressed the question of whether a physician may lie in response to a patients questions about her diagnosis or prognosis. But if he is to be consistent, he would have to regard this as impermissible, no matter how severe the consequences for the patient are expected to be. For, neither the universality argument nor the contract argument leaves room for any exceptions to the prohibition on lying.6 III. KANTS ARGUMENTS AND THERAPEUTIC PRIVILEGE There is, of course, a difference between lying and withholding information. What I will show, however, is that the universality and contract arguments commit Kant to a prohibition on withholding information under certain conditions and that these conditions are satised by the doctor-patient relationship. To get at the conditions I have in mind, suppose you have agreed to notify me if and when a certain event, E, occurs, and we both know that you would be able to notify me promptly. But suppose also that you nd out that E has occurred, and yet choose not to notify me. Let us say that you know I will assume that E has not occurred unless I hear otherwise from you since I have no way of nding out for myself. But, because you do not want me to know about E, you remain silent on the matter. For a straightforward example, suppose you and I are both expecting the arrival of a mutual friend from out of town, with whom we are both eager to spend time. You have agreed to call me as soon as our friend arrives at your house, but when he arrives you change your mind, and decide to wait a while before calling. What I am describing is a species of what are commonly called lies of omission. What distinguishes these lies of omission from others is that these occur in the context of an agreement to provide a specied kind of information. This distinction will play an important role in the remaining discussion, so it would be helpful to have a label that emphasizes it. So I will call such an omission a defection on an agreement to notify (DAN). In this section, I will argue rst that, in applying the universality and contract arguments to DAN, we get the conclusion that DAN is never morally permissible. I then argue that, in virtue of the doctor-patient relationship, withholding diagnostic or prognostic information from a patient amounts to an instance of DAN. The point of this is to show that the universality and contract arguments commit Kant to the view that invoking therapeutic privilege is never morally permissible. Recall, the key premise in the universality argument is that it cannot hold with the universality of a law of nature that statements should be allowed as

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

325

proof and yet be intentionally untrue. The thing about agreements to notify, though, is that they create a situation in which both statements and omissions of statements are allowed as proof. If and when you notify me that E has occurred, I take your statement as proof that E has in fact occurred. And, given my knowledge that you can always notify me promptly, I would take your silence as proof that E has not occurred. Our agreement therefore makes it the case that your silence is to count as proof that a certain proposition is true (namely, the proposition that E has not occurred). But if you defect, you intentionally allow your silence to express a falsehood. And if it cannot hold with the universality of a law of nature that statements should be allowed as proof and yet be intentionally untrue, it also cannot hold that silence be allowed as proof and yet intentionally express a falsehood. Thus, DAN is prohibited by the universality argument. The key idea in the contract argument, on the other hand, is that if statements in general are not believed, then rights that are based on contracts lose their force. Obviously, Kant presupposes that we should have contractbased rights; otherwise there could be no problem with their loss of force. But no such rights could be established if people did not talk about matters relevant to their interactions in the rst place. Remaining silent on issues that need to be discussed for the sake of establishing such rights would appear to be no less problematic than lying about them, on Kants view. For, in either case, there would be no binding contractually based rights. In short, remaining silent on matters that are relevant to our interactions with each other is problematic for contractually based rights in much the same way that lying about such matters is. Can we conclude from this that the contract argument says anything against DAN? It might seem not, since relatively few instances of DAN involve remaining silent on rights-relevant matters. To be sure, if people came together to discuss putative contracts as possible bases of rights, they might agree to try and bring up all relevant issues and concerns. And remaining silent on some of these would then be an instance of DAN that might properly be regarded as impermissible. For, one would thereby intentionally bring it about that contractually based rights would be less than fully informed or that some important rights would not get established at all. But imposing a prohibition on all instances of DAN on these grounds seems odd. (From the fact that some types of DAN would interfere with establishing contractually based rights, how could it follow that all instances of DAN are morally impermissible?) Notice, however, that Kants original contract argument is odd in precisely this way. Clearly not all lies are relevant to contractually based rights; the overwhelming majority of lies do not concern rights issues at all. (Certainly, statements made to a would-be murderer about the whereabouts of her intended victim do not inform rights!) And, given that people can tell whether they are discussing rights-relevant matters, the forcefulness of contractually

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

326

Chris Brown

based rights would be preserved if only lying about rights-relevant matters were prohibited. But Kant asserts that, by lying, I bring it about, as far as I can, that statements (declarations) in general are not believed, which leads him to the more general conclusion that all lies are impermissible. But if this is so, then the following also holds: By defecting on an agreement to notify, I bring it about, as far as I can, that silence in the context of such agreements generally is not trusted. Thus, if Kant were to apply his contract argument consistently, he would have to also conclude that all instances of DAN are impermissible. So, Kants contract and universality arguments yield the same result for DAN as they yield for lying: both are absolutely prohibited. We are now in a position to see why Kant must deny that invoking therapeutic privilege is ever morally permissible, if he is to be consistent. As we have seen, agreements to notify create situations in which remaining silent (when there is something to report) is, according to the universality and contract arguments, impermissible. But explicit agreements to notify are not the only things that create such situations. For, there are certain relationships where remaining silent on certain matters amounts to expressing a falsehood. And the doctor-patient relationship is one of these. Patients expect their doctors to give them bad diagnostic or prognostic news when they have it. Thus, when a doctor has nished telling a patient about her condition, it rarely enters the patients mind that material diagnostic or prognostic information may have been withheld. To be sure, doctors do not explicitly tell each patient that they will notify them of all the important medical facts pertaining to their case. (Any doctor who said this knowing that she might nd reasons to invoke therapeutic privilege would, of course, be lying.) This is merely assumed by the patient. But doctors also allow their patients to maintain this assumption since they do not make it explicit that they may invoke therapeutic privilege. In effect, then, doctors maintain a tacit agreement that they will notify their (competent) patients of bad diagnostic or prognostic news, an agreement from which they defect when they invoke therapeutic privilege. And, since the universality and contract arguments render all instances of DAN morally impermissible, Kant must deny that invoking therapeutic privilege is ever morally permitted. IV. AN ASSESSMENT OF KANTS ARGUMENTS Now that we have seen the implications the universality and contract arguments have for therapeutic privilege, it is time to give those arguments a closer, critical examination. In this section, I argue that neither of these arguments is sound. Recall, the key premise in Kants contract argument is his claim that, in lying, [one brings] it about, as far as [one] can, that statements (declarations) in general are not believed, and so too that all rights which are based on contracts come to nothing and lose their force. Again, the contract argument

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

327

makes no appeals to ideas about universality but is instead concerned with the consequences of lying. Now, surely Kant does not mean to suggest that telling any kind of lie is sufcient to completely undermine the believability of statements and the forcefulness of contractually based rights. For, obviously, we have at least some tendency to believe others statements and to expect others to honor their contracts, even though we are all aware that people sometimes lie. Perhaps Kant is trying to say that, in lying, one lowers the extent to which statements are believed and, as a result, the extent to which contracts are expected to be honored. But this is not always the case. If, for example, one makes sure that the lie one is about to tell will not be exposed as a lie, the general believability of statements, etc, may go completely unscathed. So, Kant would still need to explain what is wrong with lying in cases where there is no threat to the general believability of statements, etc. But there is another important sense in which the contract argument is incomplete. It presupposes that there are no duties that could, in some situations, require one to tell a lie despite the risk of diminishing the general trust in statements and contracts. In effect, Kant is saying that maintaining the general trust always has priority over all other moral ends, including that of providing needed assistance for other rational beings. (Let us not forget, Kant explicitly denies the permissibility of lying for the sake of preventing murder.) But this is not obviously correct, so it should not simply be stipulated. Kant needs to explain why preserving/promoting the general trust has the priority he alleges. To be sure, doing what one can to preserve and promote the general believability of statements and the forcefulness of contractually based rights seems to be an important moral aim. Our abilities to communicate, to propose contracts, and to bind ourselves through them signicantly improve our potential for learning, cooperation, and mutual advantage. We can safely suppose that maintaining a general trust in statements and contracts serves many moral ends. The question Kantians should be asking, however, is whether the general trust is an end in itself, that is, something that is worth preserving and promoting for its own sake (and not merely a means for achieving moral ends). For, if it is not, undermining it can be permissible and, in some cases, required. But it seems Kant is in no position to claim the general trust as an end in itself since he is explicit that there is only one end in itself: rational nature (GMM, 4:428). And if the general trust is not an end in itself, treating rational beings as ends in themselves would appear to have priority over preserving and/or promoting the general trust. In fact, Kant is explicit that all of our duties are derivable from the requirement to treat rational nature as an end in itself (GMM, 4:429). So, it seems he would have to admit that in those rare cases where lying, or withholding information, is the only available way to treat rational nature as an end in itself, this is what one ought to do. To summarize, the contract argument seems to exaggerate both the frailty and the importance of both the general believability of statements and

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

328

Chris Brown

the forcefulness of contractually based rights. For, it is possible to lie while preserving the general trust. And there appear to be duties which Kant himself recognizes and which could, in some cases, require an agent to lie despite the damage this would do to the general trust. Furthermore, it is not as if the contract argument succeeds when applied to other forms of deception (e.g., defections on agreements to notify). For, even if we set aside questions about the importance of the general trust, the fact remains that we often nd ourselves in a position to deceive others while preserving the general trust. And, for all the contract argument says, any kind of deception that preserves the general trust could be morally permissible. But, supposing there are moral ends that are more important than preserving/promoting the general trust, there may be cases in which one is morally required to deceive another even if one knows this would guarantee a decline in the general trust. Now, one might nevertheless suppose that the contract argument (or something much like it) still provides a good Kantian case against therapeutic privilege. For, one might think that, even if Kant does exaggerate the frailty and importance of the general trust, patients trust in their doctors is both precarious enough and important enough to make it morally impermissible for a doctor to withhold material medical information from a patient. But this, too, seems to overstate things. A doctor can withhold material information from a patient without diminishing patients trust. And, among the various moral ends that doctors can pursue, there are some (e.g., providing needed assistance) which appear to have priority over preserving patients trust. So, though it seems safe to assume that patients trust is exceptionally precarious and important, we need not conclude from this that doctors are never morally permitted to withhold material medical information from their patients. We have seen that the contract argument provides no sound basis for thinking that dishonest acts are generally impermissible since it exaggerates the importance and frailty of the general trust in statements and contracts. Nor is it conclusive when applied to special relationships where trust is exceptionally important and precarious. I now aim to show that the universality argument does no better. The problem with the universality argument is a little more difcult to illuminate, so it will be worthwhile to present the argument again in its entirety:
When the maxim on which I intend to give testimony is tested by practical reason, I always consider what it would be if it were to hold as a universal law of nature. It is obvious that in this way everyone would be necessitated to truthfulness. For it cannot hold with the universality of a law of nature that statements should be allowed as proof and yet be intentionally untrue. (CPrR, 5:44)

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

As noted earlier, Kant has it that it is morally impermissible to act on any maxim that cannot hold as a universal law. The key premise in the above,

Kant and Therapeutic Privilege

329

expressed by the last sentence, is that there could be no universal law of nature that lies be allowed as proof. This is supposed to explain why acting on a corresponding maxim is morally impermissible, and thus why the only permissible kind of testimony is truthful testimony. Unfortunately, Kant has omitted a few crucial details. It is unclear what the precise content of the maxim in question is, and it is therefore also unclear why Kant thinks it cannot hold as a universal law. So we need to do our best to ll these gaps in order to evaluate the argument. Kants specication of a maxim always includes an action or a deliberate omission. And it is clear enough from the above that the action in this case is making an untrue statement, telling a lie. Usually, however, he also includes a description of the agents situation and/or the aim she intends her action/omission to serve. The universality argument omits a description of the situation, presumably because Kant is after the general result that lying is in all situations morally impermissible. And we can only guess that he takes the aim of the lie to be: getting someone else to accept ones statement as proof. It is not clear how strong a notion of proof Kant presupposes here. He would obviously be wrong to say that, in lying, we generally intend our claims to stand as conclusive evidence for the truth of a false proposition. In order for an agent to see a point in lying, it is sufcient that she expects or hopes that her false statement will incline her audience to give her the benet of the doubt. More plausibly, then, Kant means to say that, in lying, one intends ones statement to be taken as evidence (simpliciter) for the truth of a false proposition. The maxim, then, must be: I will lie for the sake of evidencing to others a false proposition. Now, why Kant thinks this maxim cannot hold with the universality of a law of nature is far from obvious. The only explanation that readily suggests itself is that Kant here presupposes a line of reasoning he has employed elsewhere, in an argument against the moral permissibility of making a false promise for the sake of acquiring a loan one has no intention of repaying.7 The line, suitably adjusted, would have it that a law of nature that people tell lies whenever they aim to evidence false propositions could not sustain itself. For, if people always lied when they wanted to evidence false propositions, it would become common knowledge that people do this, which in turn would only ensure that people would no longer believe any statements at all. And then one could no longer evidence a false proposition by telling a lie since no one would take statements as evidence anymore. Thus, telling a lie would no longer be a way of evidencing a false proposition. In effect, this law of nature would eliminate itself since it would make such deception impossible. Assuming this is how Kant intended to be understood, the universality argument, like the contract argument, appeals to the general believability of statements. Unlike the contract argument, however, it does not rely on the idea that maintaining the believability of statements is morally important in

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

330

Chris Brown

its own right. Rather, Kants point here is that because the above lying maxim would, as a universal law, completely undermine the believability of statements, lying would then no longer be a way of evidencing false propositions. Thus, the lying maxim cannot hold as a universal law. According to the universality argument, then, the impermissibility of lying follows from the impermissibility of acting on maxims that cannot hold as universal laws. So, obviously, one cannot object to the universality argument on the grounds that it exaggerates the importance of maintaining the general believability of statements. For, it makes no claims about this at all, nor does it rely on any such claim. It does, however, rely on a claim about the frailty of the general believability of statements. For, it presupposes that no statements would be believed if the above lying maxim were to become a universal law of nature. And this, as I will show, is precisely where the argument goes wrong. The maxim in question prescribes lying for the sake of evidencing false propositions. The corresponding universal law of nature, then, would make it the case that, whenever a person has decided to evidence a false proposition, she will lie. To be sure, if this became a law of nature, and if everyone knew it, the general believability of statements would probably suffer. But just how much it would suffer is not clear. Notice, even if everyone knew about this law of nature, this does not mean that they would also know when a lie was being told. And so, even if the frequency of lies increased signicantly as a result of this law of nature, the frequency of exposed lies might not increase at all. But it is the frequency of exposed lies that affects the general believability of statements, not the frequency of lies. So, there is no reason to think that, under this law of nature, people would no longer believe anything they were told. Nor, then, is there any reason to think that it would be impossible to evidence a false proposition by lying. In other words, it seems that, contrary to the universality argument, the above lying maxim can hold as a universal law of nature. The problem with the universality argument, then, is that it (like the contract argument) exaggerates the frailty of the general believability of statements. It incorrectly assumes that the lying maxim would, as a universal law of nature, have the result that people would no longer believe anything they were told. Similarly, in order to apply the universality argument to some other form of dishonesty (e.g., defecting on an agreement to notify), one would have to assert that the relevant (dishonest) maxim would, as a universal law of nature, completely undermine some kind of believability or trust. (For, as we have seen, one of the key moves in the universality argument involves the idea that, when the relevant kind of trust is completely undermined by the new law of nature, the relevant form of dishonesty can no longer serve its purpose.) But it is not clear that any such assertion is correct. I submit, therefore, that both the contract argument and the universality argument are unsound. They do commit Kant to the view that physicians should never lie to their patients and should never defect on their tacit or

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

331

explicit agreements to notify their patients of material medical information. But those who share this view should not think that Kants take on dishonesty provides solid philosophical support for their position. And, at the outset of this discussion, we noted that the notion of autonomy central to Kants moral theory shares too little with the bioethical notion of patient autonomy to be of any use in the debate on therapeutic privilege. So, if we are looking to Kant for sound philosophical guidance on this issue, we must look beyond his notion of autonomy and his arguments concerning dishonesty.8 V. INFORMED CONSENT AND THE KANTIAN DUTY OF BENEFICENCE I believe it is most fruitful to focus on Kants duty of benecence. It not only provides the foundation for a compelling account of the moral importance of seeking informed consent but also gives physicians a useful and instructive way of thinking about nondisclosure. The duty requires each of us to be willing to make sacrices for the sake of furthering others morally permissible ends.9 There is no privileged way of fullling this duty, according to Kant, and he is explicit that there is no principle determining how much any agent is supposed to sacrice for its sake (MM, 6:393). What is required, however, is the adoption of an aim to do what one tolerably can to assist others with their ends. One appealing feature of this duty rests in the fact that it does not presuppose any one-size-ts-all conception of well-being, utility, or the good as the thing to be promoted on others behalf. Rather, it is the putative beneciarys morally permissible endswhatever these arethat matter. And I aim to show that this makes the Kantian duty of benecence especially well suited as a physicians moral duty, one which provides a compelling account of the moral importance of seeking informed consent. But we will also see that furthering a patients morally permissible ends sometimes requires her physician to withhold material medical information from her, albeit only in very special cases. Thus, in the absence of exceptionless prohibitions on lying and DAN, Kantianism provides sound philosophical support for the view that therapeutic privilege is sometimes morally legitimate. Let me begin with the Kantian take on the moral importance of seeking informed consent. A doctor can reasonably assume that her patients have an interest in their own health. In most cases, one genuinely benets another person by promoting her health. Why, then, is it not generally permissible for a doctor to initiate or halt medical treatments based solely on her best estimates on which of the available treatments is/are most conducive to a patients health? One answer (perhaps the answer) begins with the fact that, often enough, the best way of promoting a patients health is not the best way of furthering her ends. A patient may nd it important to forego medical treatments that she recognizes as most conducive to her health, if she also recognizes that those treatments would compromise her pursuit of other ends she has.10 For example, a blood transfusion may be most conducive to a

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

332

Chris Brown

patients health and yet be incompatible with her spiritual or religious ends. But, the answer continues, patients do not wear their respective systems of ends on their sleeves, and doctors are not mind-readers. So, doctors face certain epistemic difculties when it comes to determining which treatments are the best means for furthering their patients ends. But by disclosing diagnostic and prognostic information, they can give their patients the opportunity to determine this for themselves. Given the opacity of others ends, this is most often the only way for a physician to ensure that, in administering or prescribing some treatment, she will further her patients ends rather than compromise them. Thus, most often, the Kantian duty of benecence requires doctors to explain material medical information to their patients and then to allow them to make treatment decisions. However, full disclosure and best serving ones patients permissible ends do not always go hand in hand. Anxiety brought on by bad medical news can undermine a patients capacity to deliberate about her treatment options. It can also induce any number of adverse physiological responses. In some cases, the net result would be that the patient is no longer competent to make treatment decisions. So, full disclosure does not always help physicians ascertain which treatments are most consistent with their patients ends since it does not always facilitate informed consent. In other cases, the patients competence would survive, or at least would not be permanently undermined by the trauma from bad medical news, but the physiological harms brought on by that trauma would be debilitating, leaving her much less able to pursue her ends. In these cases, the possibility of obtaining informed consent would not be forestalled, but the patients prospects for achieving what she values would be signicantly diminished by full disclosure. In short, although full disclosure is most often conducive to furthering a competent patients ends, it can be ineffective or counterproductive. I grant that, presumably, it is very rarely the case that a physician reasonably expects that full disclosure will bring about either of the above kinds of adversities.11 But the fact that this can happen paves the way for a defense of therapeutic privilege, assuming doctors have a fundamental moral duty to do what they can to further their patients permissible ends. For, if they do, there is a strong reason to withhold material information when one reasonably expects that full disclosure would seriously compromise ones patients ends. And, as I will argue here, this reason is not always overridden by opposing moral considerations. There is a lesson to be gleaned from the previous sections analysis of Kants contract and universality arguments. Kant seems perfectly right to think of the general trust (the extent to which statements are believed) as morally important. Recall, my complaint was just that he exaggerates its frailty and its importance. One can lie without undermining the general trust, and lying sometimes serves moral ends that have priority over preserving or promoting the general trust. I made similar points about patients trust and

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

333

nondisclosure: One can withhold material information from a patient without undermining patients trust, and nondisclosure sometimes serves moral ends that have priority over preserving or promoting patients trust. Now, of course, patients trust and the general trust are not the only morally important interests that can be undermined by treating a patient in the absence of informed consent. For example, if policy or law absolutely prohibits nondisclosure, withholding information from a competent patient could result in sanctions that radically diminish ones opportunities for assisting others. Also, when practices and hospitals can be sued because of an instance of nondisclosure, a single omission by one physician can adversely affect other physicians opportunities to provide assistance. But here, too, we should not exaggerate the moral importance or the frailty of the interests in question. To be sure, the importance of maintaining or improving ones own and others prospects for providing assistance is hard to exaggerate. But I suspect that very many experienced physicians are perfectly capable of withholding material information from a competent patient without compromising these prospects at all. These considerations therefore are not prohibitive for doctors who are so capable. Arguably, full disclosure tends to serve a number of other morally important interests. My suggestion, however, is that very many physicians know how to avoid compromising those interests as well, should they choose to withhold material information from a patient. So, when such a physician reasonably expects that full disclosure would seriously compromise the ends of one of her patients, concerns about the other morally important interests at stake do not render nondisclosure impermissible for her. In order to show that nondisclosure is impermissible even in such cases, one would have to show that (medical) nondisclosure is in and of itself morally wrong. Now, I concede that nondisclosure can be regarded as a form of dishonesty or bad faith. In Section III, I argued that doctors in effect maintain a tacit agreement that they will notify their competent patients of all material medical information. By administering treatments in the absence of informed consent, one defects on an agreement one has deliberately kept in place. However, we have already seen problems with the Kantian reasons for thinking that such defections are in and of themselves morally wrong. In addition, it would be implausible to say that one should never defect on an agreement. It is not difcult to imagine both medical and nonmedical cases in which adherence to an agreement would defeat the purpose of the agreement and in which defection is the only way of preventing serious harms to one or more of the agreements intended beneciaries. When reasons such as these are precisely what motivate a doctor to withhold information from her patient, characterizing this as an instance of dishonesty or bad faith would be misleading at best. And those who consider it morally impermissible would need an extremely compelling case for the absolute moral priority of honoring ones agreements. On the view proposed here, a doctor should withhold material medical information from her patient when she reasonably expects both that (a) full

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

334

Chris Brown

disclosure would seriously compromise her patients permissible ends and that (b) her nondisclosure will not compromise other morally important interests. I cannot say how often it is that a doctor reasonably believes that both of these conditions are satised. Again, I do suspect that very many experienced doctors are capable of withholding material information from competent patients without compromising other important interests. But I presume that even an experienced doctor rarely has a reasonable expectation that the permissible ends of the patient in front of her would be seriously compromised by full disclosure. It might be thought that this is so rarely the case that even experienced doctors would be well advised to adopt an exceptionless policy not to withhold material information from competent patients. However, this presupposes that the advantages of adherence to such a policy outweigh the disadvantages to those patients who would be seriously harmed by it. And, as rare as the exceptional cases may be, I contend that the balance of advantages and disadvantages works in favor of adopting a more sophisticated policy. Precisely because full disclosure so often is most conducive to serving various morally important interests, including the permissible ends of competent patients, it is rightly regarded as a sound default approach. But when a physician is presented with compelling evidence that full disclosure would seriously compromise her patients permissible ends, the fact that this rarely happens is, of course, irrelevant. She should do what she can to ascertain whether the other moral reasons that usually prohibit nondisclosure do so in her present case. If they do not, invoking therapeutic privilege is, all things considered, morally required.12 NOTES
1. For accounts of the history and legality of informed consent, see Faden and Beauchamp (1986) and Berg, Appelbaum, Lidz, et al. (2001, 4193). 2. See Tickner (1995), Stauch (2000), and Berg et al. (2001, 7593, 14666), for more on the legal debate. As for the ethical debate, Katz (2002, in particular, 20729) and Pirakitikulr and Bursztajn (2006). 3. I use the translations of Kant included in Gregor (1996). Subsequent references to Kants writings will be cited using the following abbreviations: GMM (Groundwork of the Metaphysics of Morals); CPrR (Critique of Practical Reason); MM (The Metaphysics of Morals). This will be followed by the volume and page numbers from Kants gesammelte Schriften, edited by the Royal Prussian Academy of Sciences (Berlin: Georg Reimer, later Walter de Gruyter & Co., 1990). 4. For some recent commentaries on the categorical imperative, see Timmons (1997), Guyer (2000, 172206), Kerstein (2000), Millgram (2003), and Kitcher (2004). 5. Kant also employs the categorical imperative to argue for positive duties (as opposed to prohibitions), such as benecence and self-perfection (GMM, 4:4223). 6. Few contemporary Kantians agree that lying is generally morally prohibited. See Paton (1953 1954), Sedgwick (1991), Koorsgaard (1996, 13358), and Shapiro (2006). 7. In Groundwork of the Metaphysics of Morals (4:422), Kant argues:
Another nds himself urged by need to borrow money. He well knows that he will not be able to repay it but sees also that nothing will be lent him unless he promises rmly to repay it within a determinate time. He would like to make such a promise, but he still has enough conscience to ask himself: is it not forbidden and contrary to duty to help oneself out of need in such a way? Supposing that he still decided to do so, his maxim of action would go as follows: when I believe myself to be in need of money I shall

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Kant and Therapeutic Privilege

335

borrow money and promise to repay it, even though I know that this will never happen. Now this principle of self-love or personal advantage is perhaps quite consistent with my whole future welfare, but the question now is whether it is right. I therefore turn the demand of self-love into a universal law and put the question as follows: how would it be if my maxim became a universal law? I then see at once that it could never hold as a universal law of nature and be consistent with itself, but must necessarily contradict itself. For, the universality of a law that everyone, when he believes himself to be in need, could promise whatever he pleases with the intention of not keeping it would make the promise and the end one might have in it itself impossible, since no one would believe what was promised him but would laugh at all such expressions as vain pretenses.

8. An alternative defense of Kants position might be constructed on his behalf by invoking his idea of a Kingdom of Ends, an ideal state of mutual respect between all rational beings (GMM, 4:43337). Kant is explicit that we are to act only on principles that could serve as laws in a Kingdom of Ends (GMM, 4:436). And one might suppose that, at least for Kant, there can be no deception in a Kingdom of Ends, and therefore that it is wrong to act on any principle that prescribes deception. The problem with this defense, however, is that there can be deception in a Kingdom of Ends. For, Kant is explicit that the laws of a Kingdom of Ends are grounded entirely on the fundamental requirement to treat rational nature as an end in itself (GMM, 4:433). On his view, it is precisely when all rational beings respect each other as ends in themselves that a Kingdom of Ends is achieved. But, as I will illustrate in Section V, it is possible to deceive for the sake of treating a rational being as an end in itself. So, deception is possible even in the ideal state where all rational beings always treat each other as ends in themselves. 9. Kant argues that the ends of a subject who is an end in itself must as far as possible be also my ends, if that representation is to have its full effect in me (GMM, 4:430). I have concerns with this argument, but there are other arguments available to Kant in support of the duty in question. I favor one which begins with the rst formulation of the categorical imperative (act only in accordance with that maxim through which you can at the same time will that it become a universal law [GMM, 4:421]) and argues that one cannot will as a universal law a maxim not to further others morally permissible ends, since, as a human rational being, one wills that others assist one with ones own morally permissible ends. Kant also describes the duty in terms of promoting others happiness; in fact, this is his favored description. And he seems to presuppose at least two distinct conceptions of happiness: (a) the satisfaction of habitual desires (GMM, 4:399, 405; CPrR, 5:73; MM, 6:212) and (b) agreeableness and the absence of disagreeableness (CPrR, 5:22, 60). One could, therefore, plausibly suppose that Kant is in fact committed to three distinct duties of benecence. But I will ignore this here, since I believe it will be most instructive to focus on the duty to further others morally permissible ends. For conveniences sake, I will call this the Kantian duty of benecence. 10. Johnston and Holt (2006) emphasize a similar point. 11. It can hardly be said that we have much in the way of scientic knowledge of the ways in which bad medical news can predictably harm patients. See Buckman (1992, 53). 12. I would like to thank David Schmidtz, Mark Timmons, and Tom Christiano for their comments throughout the development of this paper and for many stimulating conversations. Sincerest gratitude also goes to two anonymous referees, especially for their role in shaping the nal section of this manuscript, and to Houston Smit, whose exegetical expertise and guidance have proven invaluable. Finally, I thank the Earhart Foundation for its support during several key phases of my research.

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

REFERENCES
Berg, J., P. Appelbaum, C. Lidz and L. Parker. 2001. Informed consent: Legal theory and clinical practice. Oxford: Oxford University Press. Buckman, R. 1992. How to break bad news: A guide for health care professionals. Baltimore, MD: The Johns Hopkins University Press. Faden, R. and T. Beauchamp. 1986. A history and theory of informed consent. New York: Oxford University Press. Gregor, M. (ed. and trans.) 1996. Practical philosophy. Cambridge: Cambridge University Press. Guyer, P. 2000. Kant on freedom, law, and happiness. Cambridge: Cambridge University Press.

336

Chris Brown

Johnston, C. and G. Holt. 2006. The legal and ethical implications of therapeutic privilegeIs it ever justied to withhold treatment information from a competent patient? Clinical Ethics 1:14651. Katz, J. 2002. The silent world of doctor and patient. Baltimore, MD: The Johns Hopkins University Press. Kerstein, S. 2000. Kants search for the supreme principle of morality. Cambridge: Cambridge University Press. Kitcher, P. 2004. Kants argument for the Categorical Imperative. Nos 38:55584. Koorsgaard, C. 1996. Creating the kingdom of ends. New York: Cambridge University Press. Millgram, E. 2003. Does the Categorical Imperative give rise to a contradiction in the will? The Philosophical Review 112:52560. Paton, H. 19531954. An alleged right to lie: A problem in Kantian ethics. Kant-Studien 45:190203. Pirakitikulr, D. and H. Bursztajn. 2006. The Grand Inquisitors choice: Comment on the CEJA report on withholding information from patients. The Journal of Clinical Ethics 17: 30711. Secker, B. 1999. The appearance of Kants deontology in contemporary Kantianism: Concepts of patient autonomy in bioethics. Journal of Medicine and Philosophy 24:4366. Sedgwick, S. 1991. On lying and the role of content in Kants ethics. Kant-Studien 82:4262. Shapiro, T. 2006. Kantian rigorism and mitigating circumstances, Ethics 117:3257. Stauch, M. 2000. Taking the consequences for failure to warn of medical risks. The Modern Law Review 63:2618. Tickner, K. 1995. Rogers v WhitakerGiving patients a meaningful choice. Oxford Journal of Legal Studies 15:10918. Timmons, M. 1997. Decision procedures, moral criteria, and the problem of relevant descriptions in Kants ethics. Jahrbuch fr Recht und Ethik 5:389417.

Downloaded from http://jmp.oxfordjournals.org at UNIVERSITAT POMPEU FABRA on May 28, 2010

Оценить