Medical Treatment at the End of Life—A British Doctor's Perspective
Medical Treatment at the End of Life—A British Doctor's Perspective
Abstract and Keywords
The rising number of prosecutions for medical manslaughter understandably is a matter of concern within the medical profession, and the complex ethical issues with which a medical professional must contend when confronted by a patient requesting assisted dying can only be exacerbated by the knowledge that his consequent actions may educe criminal liability. This chapter provides a UK clinician's perspective on these issues, and identifies three key issues in regard of the way ethical principles impact upon clinical practice at the end of life: greater demand, and respect for ‘patients' rights’; advances in medical practice that prolong life; and the Human Rights Act 1998. Together, these elements can lead to tensions for a doctor trying, at one and the same time, to respect his patient's vulnerability, assist the patient in articulating his or her authentic wishes, and to follow these wishes while remaining within the law.
Ethical medical practice is based on a code that expresses what is ‘right’ in a particular circumstance, a position reached through consideration of moral, ethical, and cultural norms, filtered through clinical decision-making. These considerations need to embrace what is legal, but are not concerned that the ‘right’ decision is necessarily consistent with the law. It is also likely to be a more nuanced approach than that based on a legal approach, where the basis of a decision is anchored in both primary legislation and case law.
Reaching a solution by reference to case law has considerable drawbacks, and may not always provide a good and consistent reference in dealing with similar cases. Courts have reached what appear to health care professionals to be confusing and conflicting conclusions in deliberations about what are inaccurately portrayed as ‘right to life’ cases.1 These judgments are often applicable to the individual case, rather than forging a new legal framework for similar dilemmas.
Arguably, ethical consideration is a more flexible instrument than legal argument played out in the confines of a court, contained as it is by legal process and procedural argument. Ethicists have the freedom to debate, while the courts have the duty to decide, but clinicians charged with a responsibility to resolve ethical issues often bear a heavy responsibility. This burden is both lightened and more focused because the patient, being at the centre of all decisions, is either present to play this crucial role through the expression of his or her autonomy, or others are charged with fulfilling this function as a representative of the patient's wishes.
While there are many influences on the way that ethical principles in clinical practice have been shaped in relatively recent times, the most important in terms of treatment at the end of life are:
• greater demand, and respect for ‘patients’ rights’;
• advances in medical practice that prolong life; and
• the Human Rights Act 1998.
• greater patient autonomy in decisions about the timing and mode of death;
• an increasing perception that what patients want in this regard is being denied by doctors; and
• the view that laws on assisted dying are restrictive of autonomous patient choice.
There is a paradox in the ethics of end-of-life treatment. The time at which, by virtue of the terminal, or life-threatening nature of the disease, patient autonomy should be at greatest influence, it may be at a low ebb. It is the duty of health care professionals caring for the patient to do their best to respect this vulnerability, to help the patient express his or her wishes, and, as far as possible, follow these wishes in treatment, or non-treatment decisions.
Recent cases relating to patient refusal of treatment have commonly been taken under the umbrella of the Human Rights Act 1998. Many will see the application of the Act to medical decision-making as inappropriate, given its origins in the European Convention on Human Rights of 1950, which was itself a response to gross abuses in pre-war and wartime Germany and elsewhere. The two central Articles are the ‘right to life’ (Article 2), and the right not to be subjected to inhumane or degrading treatment (Article 3). Article 2, in fact, provides no such thing as a right to life in all circumstances. At the very least, to do so would imply that the law delivers protection from the inevitability of death. The medical profession's settled view is that this clause should be interpreted as placing a responsibility on the doctor to respect human life, and to seek to preserve it, except when to do so comes at the price of a level of suffering that is unacceptable to the patients. This suffering may be physical, mental, or both.
‘Treatment’, in the general context of Article 3, is not a direct reference to medical care, and the attempted application of the Act to health care issues (such as rationing) dilutes the concept of ‘degrading’ or ‘inhumane’ treatment to a questionable and unhelpful degree.
When applied to decisions at the end of life, a tension is immediately created between the concept of a right to continuing life, and the right not to be treated inhumanely. In medical terms, inhumane or degrading treatment is that which has no possibility of achieving a net benefit, and whose continuation therefore represents a net harm. Cases in which autonomy has been expressed in terms of the right to refuse treatment (which is a different context than the right to remain alive) have almost always been decided in favour of the right of a patient to expect an autonomously expressed refusal of treatment to be followed and respected, even if that will lead to the likely or certain death of the patient.2
(p.167) It is important to stress that these judgments have been delivered in spite of, rather than because of the Human Rights Act, since they are based on the common law. Any doctor who overrules a competent refusal of treatment (see below) commits a trespass or assault.3 The Human Rights Act does not need to be employed. Where it does have (some) relevance is in its application to end-of-life decisions by the patient in which current laws preventing the doctor acting upon patient requests for assisted dying are restrictive of patient rights.4
Where courts have been asked to arbitrate between Articles 2 and 3, judgment has usually been given in support of a patient's right to refuse treatment. If pressed, one would have to argue that, in these cases, the respect for patient autonomy reflects a superiority of Article 3 over Article 2, since autonomy is reflected in support for a decision not to be subjected to inhumane treatment. The only circumstances in which courts have effectively found in favour of a clinical decision to continue life-supporting treatment have been in cases where uncertainty exists regarding the capacity of a patient to make decisions, or where there is evidence of clinical uncertainty.
At a simple level, respect for the principles of beneficence and non-maleficence reflects the idea of a health care professional seeking to do good. In terms of clinical outcome, the linked concepts of net benefit and best interests come into focus. All medical procedures carry risk, and many have negative, as well as positive outcomes. The short term outcome of the risk, and the pain of a Caesarean section carries an extremely positive net benefit in the delivery of a healthy baby. Similarly, the brief pain of an injection or venepuncture is justified by the associated immunization, medication, or blood test.
In all therapeutic encounters, it is a fundamental principle that the doctor approaches these from a clinical perspective. That is both the primary duty of a doctor, and the job for which he or she has been trained. Every encounter starts from this point, and from it flow the completion of a diagnosis, a decision about management, and the weighing up of alternative therapeutic plans. This last stage also includes a choice between treatment and non-treatment, although a decision not to treat does not automatically terminate the doctor/patient relationship, but transfers its content to one of palliative, or supportive care.
The weighing up of different therapeutic decisions under the rubric of net benefit is therefore a clinical process that is familiar to the doctor at every step. What has changed, and arguably only in the last thirty years, has been the additional concept, the required next step, in framing these decisions in a new context, that of the best interest of the patient; it is only relatively recently that both the legal and ethical duties to view the clinical decision through the eyes of the patient have become a feature of good practice.5 In many cases, this is a decision inseparable from that (p.168) of net clinical benefit. One fits the other like a glove. The careful control of the blood sugar in a diabetic patient reduces the risk of both short-term and long-term problems, and compliance with treatment is clearly in the patient's best interest.
However, at the end of a life characterized by disability, physical restriction or pain due to multiple pathology, a doctor should pause before assuming that a chest infection should be treated, even if the outcome will be the (probable) control of the illness, and the preservation of life. In this type of scenario, how the patient views the probable outcome of a successful resolution of bronchitis or pneumonia, beyond the purely clinical, is central to the decision-making process. Doctors have had difficulty in seeing their responsibility as being to respect this aspect as much as the need to achieve net benefit, and the influence of patient demand and expectation has been a strong driver in changing attitudes.
How can best interests be assessed? Often, with great difficulty. Even if a patient is fully competent, informed, and engaged, there may be communication problems. These will not just relate to language, but also to the circumstances in which the patient is communicating with doctors. A feeling of vulnerability may exacerbate a reluctance to openly express wishes and fears, for example. The doctor needs to understand this, and be open to the variations of expression that often reveal underlying fears and anxieties.
In the case of an incapacitated patient, the problem of identifying best interests assumes a new difficulty. The opinions and wishes of others close to the patient may be extremely helpful in understanding a patient's wishes, expressed prior to the onset of incapacity, but they are not determinative.6 Many family members find this confusing and difficult, feeling that they should have more control over the outcome, but an explanation of the legal limits of their involvement, combined with an approach that stresses the importance of their contribution, will usually improve their feeling of involvement.
How the autonomy of an incapacitated patient can be respected is obviously particular to the case, and absolute certainty that treatment will follow a patient's wishes is never complete, but a careful assessment of evidence from family, friends, and other health care staff involved in their care, and an approach based on teamwork and consensus, will almost always produce a decision that feels appropriate and ‘right’. It will, at least often, be the case that such an approach will also provide the health care team with protection from the attentions of the criminal law.
(p.169) The advent of the Mental Capacity Act 2005 offers a more precise instrument by which autonomy can be respected, providing, as it does, the legal protection of the advance directive.7 Although uncertainty as to the purpose of these instruments caused much parliamentary confusion during the passage of the then Mental Incapacity Bill, their importance in end-of-life treatment lies in the fact that they are now legally binding.
A contemporaneous refusal of treatment, given by an informed and competent patient, must be respected. As already mentioned, cases have confirmed that a refusal to honour such a refusal places the doctor in a position of assault or trespass. The force of a relevant advance refusal is no different, but an assessment of its relevance to a particular clinical situation demands some work.
In general, an advance refusal should determine medical treatment if:
• it was completed with a full understanding of the context in which it could be applied; and
• it is relevant to the situation the patient is now in.
It must be the case that patients who complete an advanced directive can never fully anticipate the circumstances in which it might be operative. A common experience is that the factors we might believe would influence choices of care or treatment become less certain once that situation is reached. Many of us, seeing a dependent, elderly relative suffering under the weight of ill-health, deteriorating capacity, and limited opportunities, are quick to make the assumption that we would reject living a life so constrained, on the basis that it would have little value or meaning. It is the experience of many who work in such clinical environments that different perspectives come to bear when we reach such situations, and absolute certainties become more nuanced.
These considerations are of less importance when applied to an advance statement that outlines general requirements and expectations of care. It should be emphasized, that, in spite of the judgment in the case of R (on the application of Burke) v GMC, there is no specific legal requirement for a doctor to provide treatment that is demanded in an advance statement, but good advice, based on ensuring patient trust, must be to take such a requirement into account, even if there is a potential ‘waste’ of resources in so doing.
Where the advance statement requires more careful interpretation is when it becomes an advance refusal of treatment. This will usually refer to a refusal of life-supporting treatment in which the outcome may (refusal of antibiotics, refusal of dialysis), or almost certainly will (refusal of CPR, ventilation, or artificial nutrition and hydration (ANH)), lead to death.
Here, the imperative is to ensure, as far as possible, that the patient, when completing an advance refusal, understood and foresaw the circumstances in which it might apply. As we have seen, neither this, nor the absolute certainty that a (p.170) patient had full information and capacity at the time of completion, can be achieved, and the advance refusal, while a significantly useful new instrument, needs to be applied with care, and with a proportionate approach. This approach demands that some weight is given to the patient's previously expressed wishes, as evidenced by family and friends, as well as an assessment of the relevance of the document to the clinical situation now presenting.
Debates around end-of-life treatment are always passionate, and frequently polarized. They also tend to be further muddled by a number of confounding factors. These include a confusion between what is assisted dying, and what is palliation, a problem with the ethical and legal distinctions between euthanasia and withdrawal of treatment, a poor understanding of the ‘double-effect’ doctrine, and a substantial level of ignorance about what actually occurs when treatment, particularly ANH, is withdrawn.
In the course of a final illness, there is usually clarity for doctors, and for many of our patients, about when it is no longer appropriate to persist with treatment, and when that treatment itself has no further purpose, and therefore becomes a burden. The doctor's duty in these cases is the relief of suffering, both physical and mental. This will nearly always bring into consideration whether ‘active’ treatment—treatment to preserve life and to act to treat a condition—should be withheld or withdrawn. For some, often driven by strong religious belief, there is an acceptance that while palliation is a central obligation, treatment should never be refocused on assisting the patient to die. This is one area of continuing difficulty since, for such people, a decision to withdraw treatment is tantamount to assisting dying, because the doctor cannot separate foreseeing the likely consequence of the decision from an intention to assist the patient's death. This was at the heart of opposition at the final stages of the passage of the Mental Capacity Bill, when a passionate debate ensued on whether patients who took advantage of an advance refusal of life-supporting treatment were effectively creating a legally enforceable suicide request.
This leads us into considering the ethical and legal distinctions between euthanasia and withdrawal. It is both an ethical and legal dilemma because, although many doctors would claim to be able, in all cases, to separate an intention to relieve suffering from an intention to end life, these thoughts cannot in reality be distinguished in the doctor's mind. Physicians are driven by a fundamental imperative to reduce pain and suffering. When faced with a patient who is suffering both physically and mentally, the primary effort will be to give whatever treatment is appropriate to achieve this aim, providing net benefit in terms of best interest. It is unrealistic to think that the doctor at all times keeps from his mind the thought that part of that benefit includes death. Obviously, whether the doctor acts on this thought will define whether he or she remains within accepted ethical and legal practice, but that these thoughts will arise must be simply a natural part of a doctor's professionalism. They are particularly (p.171) difficult to deal with when the patient also articulates a wish to discuss assisted dying.
The principle, or doctrine of double effect has long been a convenient protection, used (often with some sleight of hand) to avoid conviction for murder. It is a relic of a time before doctors developed palliative skills, and when the relief of pain depended on the use of powerful drugs given in sufficiently high doses to threaten the patient's life. With the development of good, although far from universal palliative care, in which careful attention is paid to dosage and outcome, the double effect concept is not only irrelevant, but a positive barrier to informed and honest debate on end-of-life care. Nowadays, a major ethical requirement of doctors is to provide evidence-based care, and practice, and evidence confirms that, for the majority of patients, the appropriate use of pain-relieving drugs involves prescribing them at a level that does not, in itself, threaten life, although, of course, the primary threat to life remains the terminal nature of the illness.
We are therefore now at a point at which this outdated ‘principle’ has become harmful to rational debate. It also acts as a protection for an outdated paternalistic attitude to terminal care, in which the ‘doctor knows best’ approach prevents further involvement, both of the patient, in expressing his or her wishes about treatment, and of those close to a patient who has lost capacity, in advising the doctor on what the patient's known wishes might be.
Finally, there is the wide level of ignorance amongst patients, politicians, and, unfortunately, many health care professionals, about the processes involved in the withdrawal of ANH. Ending treatment that involves the provision of intravenous fluid and nutrition via a nasogastric tube carries profound implications, since the consequence of so doing will be the certain death of the patient, over a number of days. The period between the withdrawal and death is often uncertain, but the relationship between the two is absolute. There is a less strong link between the withdrawal of, say, antibiotics for a chest infection in a patient with advanced dementia and the patient's death, as a number of other factors will influence the course of the illness.
There are, of course, as many scenarios as there are patients. Let us consider the case of a patient with a terminal malignancy. In this case there is a reasonable predictability about the course of the final illness. Experience suggests that when there is a strong likelihood that death will occur within one month, then artificial feeding makes no difference to the course of the illness. If that is the case, it follows that withdrawal of ANH will also not influence the outcome, and therefore that doing so cannot reliably be predicted to accelerate the dying process, and therefore be the primary cause of death.
In cases which are very familiar to doctors, where the course of the disease, although inevitably terminal, is less predictable, the place of ANH is also unclear in terms of whether the outcome (‘outcome’ here being delaying dying) is helped (p.172) by instituting feeding, or harmed by withdrawing it. If the final course of the illness is likely to last less than a month, again the provision, or non-provision of ANH appears to have little effect in terms of predictable outcome. If the terminal phase extends beyond a month, the removal of ANH may accelerate the dying process, but a direct causal relationship is hard to identify.
In an increasingly common group of patients, those suffering from long-term, debilitating, but not immediately life-threatening illness, the contribution of ANH to the final outcome is also debatable, but for different reasons. A patient suffering from advanced dementia will be prey to a variety of potentially life-threatening conditions, each of which may individually cause death (whether treated or not) in an already vulnerable, and weakened patient. It is, therefore, impossible to say with any certainty what contribution the withdrawal of feeding makes to the final outcome, although, of course, it may feel to the health care professionals involved that a decision to withdraw ANH, after a long period of its provision to the patient, will create a strong association between the act and the outcome, even if the causal relationship is absent or weak.
These difficult and unpredictable scenarios cause discomfort to those who want more certainty in medical treatment and outcome, where it often does not exist. The debate is further polarized by the habit of those opposed to the withdrawal of ANH as describing its effect as death by starvation. As well as being wrong, this view also overlooks the fact that good palliative care must be provided to all patients in which a non-treatment decision has been reached.
What has been missing from the debate is an ethical approach that provides the same rights to patients, and which respects their autonomy in the way the law already allows. There is a strong, and well-tested legal protection for the principle that a competent patient can decide to refuse life-supporting treatment, even if the likely consequence of that decision will be death. There is a clear disjunction between the law and ethics in these cases, and ethics needs to catch up with what the law already permits.
Euthanasia is a minority issue, in the sense that the vast majority of decisions about end-of-life treatment will be ones involving withdrawing and withholding treatment, rather than whether to actively end life. As more of us approach old age, with a greater certainty that our life-ending cancers and vascular diseases will be successfully treated, we—or, very often, those close to us—will be making decisions about the circumstances in which we would wish to be treated, or not treated.
In this context, the opposition of many doctors, and some politicians and pressure groups to opening up legal opportunities for patients to be provided with assisted dying, has an impact beyond the particular issue. The debate around euthanasia may not appear very important in terms of the small number of patients who would wish to benefit from a relaxation of the law of murder, but as (p.173) a symbol of the current gap between what our patients want, and what we doctors think they want, it is extremely potent.
Whether the medical profession maintains its curiously paternalistic need to control patient wishes on the most important area of care we will all eventually face is not that significant when applied to assisted dying. But it carries strong implications for the way in which we approach the care of our patients at the end of their lives. Doctors do need to be more sensitive to the autonomy of their patients if they are to be effective partners for them and their families when making decisions about withdrawal and withholding life-supporting care. (p.174)
(1) For example, R (on the application of Burke) v GMC  2 FLR 1223.
(2) For example, Ms B v An NHS Hospital Trust  EWHC 429 (Fam).
(3) F v West Berkshire AHA  2 AC 1.
(4) R (on the application of Pretty) v DPP  1 All ER 1.
(5) Here, I am arguing that it is in the best interests of a competent patient to act in accordance with her wishes, even if the result will be less ‘good’ in clinical terms than the doctor intended/expected/wished. This was exactly the situation in the case of Ms B, where the doctors behaved both unethically and illegally in denying her wish to cease ventilation. The overriding condition, of course, is that the patient makes that decision with full knowledge of the (likely) outcome of her decision, as well as understanding all the available other options.
(6) Although the Mental Capacity Act 2005 enables Lasting Powers of Attorney (LPA) to be conferred by a person. Under an LPA, the person confers authority upon another to make decisions that can include decisions about his or her personal welfare relating to the carrying out or continuing medical treatment when he or she lacks capacity (s 9–11).
(7) See ss 24–6 of the Mental Capacity Act 2005.