The Autonomy Principle

Beauchamp and Childress place the AP with three other principles as the four normative pillars of bioethical discourse: beneficence, non-maleficence and justice.30

Relying on their approach, the patient is granted by the AP the right to an informed choice of medical treatment as long as he is sufficiently competent to make decisions without coercion or under the influence of others. In some cases, as during CPR, the patient is incompetent, and therefore cannot make any decisions. Childress and Beauchamp30 elaborate on this principle providing a two-step framework of standards for surrogate decision making in case the patient has not provided an advanced directive, for example, a living will or a durable power of attorney. They accept two criteria: the Pure Autonomy Standard and the Best Interest Standard. According to the first, designated by some as the Substituted Judgment Standard,1 ,31 family members should interpret the patient's current wishes according to his past actions and decisions, as if they were asked ‘if the patient could have made a decision, what would it have been?’ This judgment is usually relayed to professionals through relatives, perceived to be the most familiar with the patient's past decisions. The second standard would be used in case the patient's relevant preferences cannot be known due to lack of past relevant decisions or of relatives who possess that knowledge. According to this standard ‘…a surrogate decision maker must determine the highest net benefit among the available options’ (p.138).30

The two-step framework for surrogate decision making should be considered as part of a Three-Tiered Decision Making Process. First, the patient should be treated according to his advance directives. Second, in case the latter does not exist or relevant information could not be found, professionals should take recourse to the Pure Autonomy Standard. Third, in case the latter cannot be exercised, the Best Interest Standard should be implemented.

Applying the Three-Tiered Decision Making Process, we need to affirm that any patient undergoing CPR is incompetent. First, professionals should follow specific instructions given in advance directives, which possess the ultimate authority to determine who will be present during CPR. In case such instructions exist, there is no need for professionals to establish who is a ‘select family member.’ Second, if there are no relevant instructions, professionals need to ask relatives, ‘would the patient have wanted to allow you (his relatives) to witness his CPR?’ and accommodate them accordingly. Third, in case there is a lack of relevant past decisions or a lack of capable relatives who could relate that information, a surrogate decision maker should decide according to the perceived best interests of the patient.

From our discussion thus far, a strong case could be made in favouring FPDR according to the decision making process outlined above. However, as Sulmasy and colleagues32 and others31 ,33 ,34 demonstrate, empirical data counteract both the morality and practicability of this model. Family members often lament of emotional stress when expected to function as surrogates for their loved ones, begging the question whether it is morally justified to expect them to do so. From a practical perspective, many difficulties arise: First, only few patients have advance directives, and even when they do, these are commonly ignored by professionals. If patients do have advance directives, and these include information that may be slightly relevant to FPDR, that information tends to be too vague or too specific; it might be difficult for patients to dictate specific wishes regarding end-of-life care in advance and they may change their minds. Also, patients may prefer to provide their surrogates some flexibility in following their advance directives. Second, family members can predict their loved one's wishes at the end-of-life with only a relatively low accuracy. Third, in assessing the patient's best interests, family members may confound the perceived values of the patient with their own worldviews.

Thus, Sulmasy and Snyder deem the Autonomy Model of Decision Making inadequate and instead offer one that considers the specific clinical circumstances and the patient's unique characteristics.1 Before turning to the new model, we examine the applicability of the BP.

The Beneficence Principle

Beauchamp and Childress30 define beneficence as including three norms: the obligation to prevent harm, the obligation to remove harm and the obligation to promote good. They acknowledge that in certain cases the patient's best interests could only be guaranteed if we temporarily suspend the patient's autonomy and act solely according to the BP. This assertion leads to a discussion of paternalism, which is relevant here since many opponents of FPDR purport to promote the patient's care, dignity and privacy by excluding the family. They recognise two forms of paternalism. Soft paternalism constitutes overriding another's autonomy for the sake of achieving a greater beneficial value identified as such by all parties involved. Hard paternalism constitutes overriding another's autonomy for the sake of achieving a greater beneficial value identified as such only by those who perform the act of suppressing another's autonomy.

Accordingly, opponents of FPDR argue that in some cases, such as CPR, the AP conflicts with the BP. It is soft paternalism if the outcome valued the most by both the patient and professionals is the survival of the patient. The two parties might disagree on the best means to achieve that outcome; the patient might believe that having relatives present will increase his chances of survival, while professionals might believe that having the family present could potentially jeopardise his chances of survival. It is hard paternalism if the patient wishes relatives to be present because he believes it will benefit them, perceiving their well-being as the most important outcome, while professionals maintain that the most important outcome is the survival of the patient, which they believe will be maximised if relatives are excluded.

There are good reasons to be suspicious of medical paternalism, but occasionally in medical practice the circumstances justify its use. One justification is that while medical paternalism implies the superiority of medical judgement over that of the patient, it undermines only temporarily the will of the patient until his true autonomy is restored. A second justification goes back to the difficulties that arise from implementing the Three-Tiered Decision Making Process.

Do these two justifications provide professionals the moral legitimacy to deny FPDR in the name of the BP? The first one, if acceptable, is certainly a required condition, but not a sufficient one, mainly because the patient may never regain his autonomy and so he may never be able to reverse the medical decision. The second justification is also legitimate, since it is based on empirical data.

Thus, on one hand, it seems that respecting the AP would mean placing the decision whether to permit FPDR in the hands of the patient or his family. On the other, opponents of FPDR claim that a consideration of the patient's best interests overrides his autonomy, framing their argument as soft or hard paternalism. Unfortunately, as discussed by Daniel Callahan,35 medicine cannot be considered to be value-free; its applications, priorities and taboos are deeply embedded in the governing power. Accordingly, in some cases, the invocation of the BP should not be taken at face-value; frequently we find that professionals respect a patient's decision only insofar as it correlates with their own judgment. Therefore, professionals might oppose FPDR because it clashes with their own worldviews, rather than their concern for the patient's best interests.

Additionally, medical science may not necessarily be the best tool for discerning the patient's best interests. Considering the grim CPR survival rates, and acknowledging that professionals performing CPR commonly know their patients only superficially, one should wonder to what extent should decision making rely on the prospective of the patient's survival.

Thus, we have identified several potential difficulties in applying either the AP or BP in the context of FPDR. In the next section, we discuss another model which seems to be more satisfactory in the case of FPDR.

Part III: the Authenticity Principle

Ronald Dworkin distinguishes two kinds of criteria used in decision making: Experiential Interests and Critical Interests. Decisions based on experiential interests account for the patient's present preferences and future expectations in light of future quality of life. Decisions founded upon critical interests entail a more holistic approach, considering personal values and past and future experiences. Whenever decisions are made based on the AP, they should be consistent with a conception of the integrity of a person's life.36

Daniel Sulmasy shares Dworkin's assertion and emphasises the importance of a holistic, integrated view of one's life when addressing end-of-life issues.37 In light of the limitation of the Three-Tiered Model for Decision Making, Sulmasy and Lois Snyder propose a model based on an ‘Authenticity Principle’, which could be treated as a subset of the AP as it places the final decision in the hands of the patient and relatives. According to this model, any medical decision making process should not be solely that of the patient, the patient's family or the physician. Rather, it should be a shared, dialectical process between the different agents involved, as exemplified by the following: ‘[o]ur physicians involve the family in the decision making process and elicit the wishes of the patient from them. If resuscitation efforts are deemed futile, the family members are told so and they are given an opportunity to be part of the decision to cease efforts.’1

Sulmasy and Snyder introduce the standard of Substituted Interests and propose an Integrated Model of Surrogate Decision Making. While substituted judgment requires us to ask the family what would the patient have wanted, substituted interests requires asking the family ‘tell me about the patient.’ Family members should be encouraged to provide professionals relevant information regarding the patient's beliefs, moral and religious values, and so on, with the aim of uncovering his ‘authentic’ convictions and preferences. Importantly, family members would not be the sole decision maker; they will provide valuable information to the professionals who in turn would provide their recommendations to the particular case.1

Sulmasy's integrated model seems to address the limitations of the Autonomy-based Three-Tiered Process and the BP in the context of FPDR. First, it lessens the possibility of soft or hard paternalism which might occasionally be used by professionals, thus also reducing the risk of confounding the patient's best interests with their own worldviews. Second, inquiring about the patient's character as a whole, rather than presenting only one question which requires a cleat-cut reply, might prevent cases in which the patient's true wishes may be misrepresented by his family, and reduce the emotional stress experienced by family members who act as surrogates. Clearly, it could be argued that within this new model family might provide false information in replying to the request ‘tell me about the patient’. However, since this request is more general than the question ‘what would the patient have wanted,’ we believe it is less likely to misrepresent the patient.

Even in the case of CPR, when rapid and efficient action arguably determines life and death, we believe that it is reasonable for professionals to approach the family and briefly ask them about the patient. Both open-ended questions, such as ‘tell me about the patient,’ or narrow questions, such as ‘did the patient involve his family in important decisions during his lifetime?’, could be used. Arguably, one might posit that an active resuscitation is no time for any discussion that is not essential, however brief it may be. We have three replies. First, given the poor chances of surviving CPR, we argue that a great deal of attention should be devoted both to the patient's well-being and to the patient's family who would eventually suffer most over his loss. Second, even if there is no time to approach the family members during CPR, there is no reason why excluding them should be the default practice rather than allowing them into the room. Once we find agreement on this, professionals might be encouraged to include time during protocol to discuss the matter with family members and ascertain whether the patient would have wanted them present. Third, it is the experience of one of the authors to suggest that, often, there is enough time during CPR to conduct a brief discussion with the family as described.