Issues of consent are at the heart of many of our ethical concerns about individual autonomy and treatment decisions. Proxy consent presents an interesting extension of the consent process to cover those incapacitated or otherwise incapable of providing consent for themselves. The Mental Capacity Act 2005 (MCA) has recently formalised proxy consent across England and Wales by the creation of a “lasting power of attorney” (LPA), which is conferred on a person designated by another individual to assume decision making power with regard to their personal welfare, property and affairs, or other specified matters should they become incompetent or lack capacity to decide.¹ The Act replaces an earlier “enduring power of attorney” (conferred by the Enduring Powers of Attorney Act 1985 and covering only matters relating to finance and property) by extending the authority of LPA to cover matters of personal welfare such as decisions about medical treatment.

The theoretical basis underpinning the proxy’s role can differ depending upon which principle of surrogate decision making is adopted:

• substituted judgement, where the proxy uses their special knowledge of the patient’s preferences to make the decision that the patient would have made were he competent;

• best interests, where the proxy makes an assessment of the patient’s best interests and makes a decision based on that assessment.

The important difference between best interests and substituted judgement is that in the best-interests role, although the proxy may use her knowledge of the patient’s preferences, her decision should be, not a direct reflection or second-guessing of the patient’s wishes, but only what the proxy would consider best for the patient in a particular set of circumstances.

Although best-interests judgements can, and often do, involve taking into account the wishes, feelings and so forth of the person who lacks capacity, they do not have to. This can be seen as both an advantage and a disadvantage of the approach. As the proxy would be making decisions on behalf of the person lacking capacity, it is advantageous, because it does not require the proxy to get inside the complex psychology of another person. However, this is also seen as a disadvantage, because it would allow the proxy to make decisions based on what she herself (or an independent, rational person) would view as best in those circumstances rather than on any particular, subjective views a particular patient might have.

Opinion is divided as to which principle should be followed.² Some, such as Brock,³ favour substituted judgement, a principle that has been adopted in US states such as Pennsylvania and California.⁴ Other writers, such as Harris,⁵ have argued for a best-interests test.

The MCA¹ fashions the role of the proxy with LPA along the lines of the best-interests view, declaring (section 1(5)) that the acts or decisions for the person lacking capacity must be “done, or made, in his best interests” but that this “does not authorise the giving or refusing of consent to the carrying out or continuation of life-sustaining treatment” (section 11 (8a)) unless there has been express provision in the form of an advance directive detailing this as something the proxy may consider.

However, the MCA contains certain elements more associated with the substituted-judgement approach that are to play a role in determining what constitutes an action or decision in the person’s best interests, notably considering “the person’s past and present wishes and feelings” (4.6a), “the beliefs and values that would be likely to influence his decision if he had capacity” (4.6b) and “the other factors that he would be likely to consider if he were able to do so” (4.6c). Although these elements can play a role in determining what is in someone’s best interests if a proxy sees fit, if they are recommended or required by the Act in the process of making a best-interests judgement, they will be open to the same problems that befall the pure substituted-judgement approach. Moreover, the moral authority of the proxy to make treatment decisions for the patients is limited by sections 24–26 of the MCA, which make clear that a valid advance directive takes precedence over the decisions a proxy with LPA may make with regard to any treatments.

The provisions of the Act send out a rather confused and, ultimately, ineffectual account of the role of the proxy with LPA, to the point where, I shall argue, the proxy has no genuine moral authority to make treatment decisions. Ultimately, we are better viewing her as nothing more than an advisor who can help the treatment team in establishing what is in a patient’s best interests.

THREE CONCERNS OVER PROXY CONSENT

There is a substantially important ethical question surrounding proxy consent, as to whether the proxy should have the moral authority to make treatment decisions for those lacking capacity of the same scope as the patients themselves would if they had the capacity to decide. There are substantive arguments as to why we should never see proxy consenters as having such treatment-determining authority.

Three major issues are directly relevant to the critical analysis of the current provision for proxy consent: pragmatic judgements, philosophical underpinnings and legal constraints affecting the moral authority of proxies. I shall deal with each in turn.

Practical problems over accuracy of judgement

There is evidence to indicate that we are simply not very good at making substituted judgements for other people, not even for close relatives, as revealed in a study by Seckler et al.⁶ They concluded that surrogates’ decisions using substituted judgement “may be seriously flawed” (p 95) and that these standards for decision making are “compromised by their inability to truly approximate the patient’s wishes” (p 97). This empirical evidence suggests that as a practical means of extending the autonomous wishes and desires of a patient who now lacks capacity, substituted judgement is an extremely poor method, as it is likely to be unrepresentative and could lead to errors.

A supporter of substitute decision making might be tempted to try to explain away the practical problems by suggesting that perhaps we are poor at making such decisions because we are unused to carrying them out and that greater awareness of the issue and communication of desires and wishes would resolve the problem. Therefore, what is needed is for proxy consent to be more widely adopted among patients, after better discussion and awareness of this as a means of consent.

However, it is unlikely that such changes in our attitudes towards proxy consent will ever be sufficient to avoid our poor ability to make judgements that accurately represent the views or desires of others, because of the second major concern over proxy consent—philosophical problems.

Philosophical problems with the issue of representation

The central question here is whether one person can ever simulate the decision making process of another person. This is a major issue in the philosophy of mind, where there are two leading theories that attempt to capture our psychology of mental representation of other persons: Simulation Theory and Theory Theory. Much of this debate has been gathered in collected works by Davies and Stone (see box). Although there is substantial disagreement between these two theories as to how we replicate or predict the thinking, decision making, emotional responses, etc, of other people, if either theory is correct, the outlook as to the possibility of successfully making something as complex as a proxy consent substituted judgement is remarkably negative. This seems to be borne out by the findings of empirical studies, such as that of Pearlman and colleagues ⁷, which showed that intervention to improve the accuracy of substituted judgements produced little improvement over chance.

Simulation Theory has its roots in an important aspect of our scientific methodology. Simulations are used widely in empirical science to make predictions of the behaviour of objects under various circumstances. For example, the effect of a new drug on people can be simulated in clinical trials by having another person or persons take the same drug under the same conditions. The conditions can then be varied relatively easily to make further predictions about the behaviour of the drug. Without this empirical process of simulation, all predictions would have to be based on some previous body of knowledge and extrapolated in the imagination, quite markedly minimising progress and our ability to gain new knowledge.

Although the simulation of processes and conditions can be repeatedly tested in the empirical domain, the realm of the mental is significantly different because of the difficulty of simulating the mental life of another person. As the memories, emotions, beliefs, desires, experiences and so on of another person—even a person one knows well—can differ radically from one’s own, placing oneself in the mental and physical situation of another in order to simulate that person’s decision making process is close to unachievable. Predicting the decision that person would make therefore pushes the boundaries of what is achievable by attempting to simulate another’s thought processes.

This is not to say that we do not attempt to simulate the thought processes of others, with varying degrees of success, but that we lack any systematic and accurate means of doing so, which are crucial in the distinctive and complex setting of decisions that relate to their medical treatment.

So while we possess an adequate folk psychology that allows individuals to reasonably determine the tastes, likes, dislikes and attitudes of another person in a range of everyday settings, such as buying a present for someone we know, the epistemology changes in the case of medical decisions. Decisions regarding our welfare and medical treatment involve such a wide and complex arrangement of values, emotions and attitudes that simple folk psychological predictions are unable to capture this.

Simulation Theory relies upon the assumption that the mental resources we possess are sufficient to simulate in our imagination the thoughts and decisions of other people. The simulation of another person’s thoughts is viewed as being “process-driven” in a way similar to empirical simulations: the same processes are occurring in the simulation as in the object which is being tested. Hence the position of the simulation theorist is that some mental processes operate in just the same way when we simulate in our imaginations being in a particular situation as they would were we actually in that situation. We can thereby explain substituted judgement as a means of proxy consent by making a series of considerations: first, by reflecting upon what I would decide to do if I were in that incapacitated situation; I can then predict what someone else would do by simulating in my imagination what their attitudes and wishes are and then utilising my own reasoning to come up with an answer—an answer I can rely on if I take my mental thought processes to be suitably similar to the other person.

This theory relies heavily on two extremely contentious assumptions: that mental processes behave in the same way in reality as in the imagination in which we simulate them; and that the person whose thoughts we are simulating is similar to us in the relevant ways. These assumptions, however, are even more dubious under the particular conditions of substituted proxy consent in a medical setting.i

In particular, the most positive affirmation one could give them would be that if these assumptions hold, they could only do so for our purely rational mental processes. Other mental processes—attitudes, beliefs, wishes, emotional responses and so on—differ so markedly from person to person that they are extremely unlikely to be accurately simulated. This is borne out by the high rate of prediction failures: if processes were so similar, then simulating other people’s thoughts would be more successful than it actually is. When this is combined with the complex and demanding situation of treatment preferences in a medical setting, where a whole host of attitudes about values, pain, life-and-death decisions, stress and various non-rational factors enter in, the chances of creating a genuine prediction are extremely remote and certainly not due to a systematic representation, as the simulator would be unable to take on the non-rational aspects. The Theory Theory explanation of how we might simulate the mental life of another person takes the differences in mental states and attitudes to be so insurmountable in trying to make an accurate imaginary simulation of another’s thought processes as to make mental simulation of others necessarily reliant on some pre-established theory as to how people normally think and respond to situations (a “folk psychology”). This view is also prominent in Dennett.⁸

It is precisely this adherence to a tacitly acquired folk psychology that would lead to problems if one were to adopt the Theory Theory approach. The failure to accurately predict other people’s behaviour and thoughts is explained in Theory Theory by the use of an incorrect psychological theory to explain the various third-person psychological processes. However, this explanation of prediction failure also highlights the general inadequacy of Theory Theory itself as a reliable means of prediction: any person attempting to predict the choices, etc, of another person must have acquired a correct psychological theory of mental processes. Even if there were a reasonably accurate and sufficiently broad psychological theory available, it would be entirely unreasonable to expect anything other than an approximation of thought processes that is unlikely to capture the complexity of the thought processes involved in medical decisions.

This leaves the substituted judgement view of proxy consent in disarray: we have neither the ability to mentally simulate another person’s thought processes nor an adequate psychological theory to represent and predict their thoughts and wishes. Any attempt to give proxy consent on the basis of substituted judgement has such a dubious basis as to make accuracy in determining another person’s thoughts unachievable.

These criticisms of substituted judgement do not, however, affect a purely best-interests view. Whereas the substituted-judgement account requires the proxy making that judgement to “get inside the mind” of the person she is proxy for, this is not a requirement of a best-interests judgement. There is no doubt that we are, to some extent, interested in what the beliefs and values of a patient are in making a best-interests judgement for them. Yet it is the very fact that such judgements are made on behalf of an individual that means that such judgements are not required to be based on accurate representations of their mental life.

This leads to the third and final concern over proxy consent.

Legal constraint on the moral authority of the proxy

This further problem is that the construction of the MCA provides legal constraints on proxy decisions, which again leads the whole process to the problem of professional, paternalistic, best-interests judgements.

The MCA places fairly obvious restrictions on what a proxy can and can’t consent to. Notably, sections 1 and 9 of the MCA set out that actions and decisions by the proxy on behalf of the person lacking capacity must be done in their best interests. Furthermore, section 9(4) also states that a proxy’s authority is subject to any conditions or restrictions that the incapacitated party may choose when creating the LPA document as well. This has particular significance, because it means that a proxy cannot refuse life-sustaining treatment on behalf of the incapacitated patient unless this condition is contained specifically as an advance statement forming part of the Lasting Power of Attorney.ii

Such legal restrictions on decision making powers substantially curtail the moral authority of the proxy. The crucial and testing question when it comes to determining the moral authority of any consenter is to ask what happens if she makes a decision that others don’t agree with (on the grounds that it is not in the patient’s best interests). This question has its foundations in Mill, for whom the right of a competent adult to make decisions that are not in their own best interests is at the very roots of the concept of liberty (and the association with our modern understanding of autonomy) as opposed to paternalism.⁹ This leads directly to a question about the ethical status of the proxy in terms of whether she has the moral authority to make those decisions that autonomous agents would make for themselves—that is, does she have the same right to make supposedly “irrational” decisions as the patient would if they were actually competent with regard to their personal medical decisions?iii As the MCA requires either that decisions be in the patient’s best interests or that there be pre-established consent in an advance statement, the proxy has little of the moral authority granted to a fully autonomous agent and hence their role as a consenter is markedly restricted.

It is not even clear that the proxy either does or should have the ability to determine what is in a patient’s best interests. Special attention is paid to what constitutes “best interests” in section 4 of the MCA. This acknowledges both multiple factors about the patient’s past and present wishes and feelings, their beliefs and values that would influence their decisions if they had capacity; and that multiple parties are to be consulted where feasible to determine what actions are in the best interests of the incapacitated patient, of which a proxy consenter within the scope of an LPA is only one. Others include the patient’s care team and anyone named by the patient as someone to be consulted on relevant matters. This approach acknowledges a long-standing view (see, for example, Buchanan and Brock¹⁰) of what constitutes best interests as involving pain and suffering and also prognosis, previously held beliefs and values, and to some extent the institutional options and other people’s interests.

The proxy’s powers are further limited by section 6(7) of the MCA, which still allows continuation of life-sustaining treatment or treatment to prevent serious deterioration of a condition until a decision is made by a court even if this conflicts with the decision made by a proxy. Moreover, there is an expectation that what she “personally feels about, or wants for, the person concerned”¹ (c. 9, section 4, subsection 5, note 31) does not motivate the proxy or decision maker. Assessment of what constitutes the best interests of the patient is still therefore very much a clinical matter, with the medical team in a position to decide whether treatment is in a patient’s best interests. This reflects a previously understood aspect of interpreting best interests under common law, where there is no obligation on doctors to continue to provide treatment, even life-sustaining treatment, that is not deemed in a patient’s best interests.

Proponents of the substituted-judgement approach readily acknowledge the need for the medical team to be more than just passive observers of the process, on the grounds that choice of medical treatment must fall within the range of medically sound options (see Buchanan and Brock,¹⁰ p143, note 17). However, if there is little possibility of making an accurate substituted judgement that can reproduce the various attitudes and specific values that an individual holds, then the role of the proxy cannot be anything more than an advisor as to a likely component of a best-interests calculation. The epistemic standards of decision making in this context strongly tip the balance towards a professional, paternalistic judgement, on the grounds that in most cases the professionals possess the specialist knowledge that determines the initial choice of treatment options and also that the presentation of various medically sound options cannot be readily divorced from the ethical components of what constitutes “best”. As “all sound medical options” is not the same as simply “all medical options”, any professional medical team will have already made numerous ethical decisions about that patient’s welfare and interests. With so much of the determination of treatment taken up by the medical team, the role of the proxy is markedly reduced in scope to work well within the decisions that have already been made on behalf of the patient.

The detailed and specialist knowledge of the treatment team undoubtedly can—indeed should—be supplemented by additional information about patient preferences as part of a best-interests judgement. However, that there might be additional values, beliefs and wishes that an individual has cannot be absolutely settled if there is no substituted judgement, only advice given as to what these may be. Even with the proxy in the role of such an advisor, without the moral force attached to the consenting process that a fully autonomous agent would have, the relating of such values cannot be a deciding factor in determining treatment. A proxy’s view of what a patient’s wishes might be is not sufficient to allow her to make a whole range of decisions that would run counter to those the professional care team have already deemed to be the best course of action. Therefore, such factors should be taken into account only if they do not diverge significantly from the factors determining the course of treatment to be taken. (The term “diverge significantly” is deliberately vague to allow scope for debate. However, I have in mind cases where there is no clearly preferred path between various treatment options, so that the additional information as to what the patient’s values, wishes and so forth are likely to be would be enough to tip the balance in favour of one particular option.)

These limitations upon the role of the proxy clearly indicate how anyone acting under an LPA should be seen, not as a genuine “proxy consenter” in terms of someone with the moral authority to make treatment decisions on behalf of the patient, but rather as a “special-interests advisor”, namely, someone who has been designated by the patient (while still competent) as someone who has particular insight into some of the factors that may be relevant. The MCA does, in one sense, succeed in creating such an advisor. However, the very fact that this individual’s authority to make treatment decisions is so heavily curtailed by pragmatic matters, by philosophical reflection and by the multiple limiting clauses in the Act would indicate that the proxy should not be seen as anything more than an advisor, with little or no moral authority to make treatment decisions.

CONCLUSION

Upon analysis, the MCA has created a situation where lip service is paid to the notion of a proxy consenter, but when the matter is pursued, the ethical and, ultimately, legal status of such a proxy seems diminished to that of an advisor. This should not be taken in a negative light, however, because this role of advisor to a professional medical team is the most useful and morally authoritative role a proxy can take. When faced with treatment decisions for an incapacitated person, attempting to provide a substituted judgement of what that person would want is difficult, bordering on the impossible. The only acceptable alternative approach is to act in our best interests. It is clear that the professional medical team will have much of the specialised knowledge that is required to make such a decision. Nominating a proxy who is able to use some special knowledge of the incapacitated party’s general values and interests will add to the body of knowledge that goes to determine what these best interests are. However, to see the proxy as anything other than a useful advisor to the professional team would be (1) to misjudge what she is capable of doing; (2) to mislead oneself into what role she will be undertaking for the patient when the patient becomes incapacitated, and hence positively inhibit, not enhance, the patient’s autonomous decision making power; and (3) to project onto her a legal and ethical status that she does not possess. Ultimately, “proxy consent” should not be seen as consent at all, but rather “assistance” to those best placed to judge the patient’s best interests.