Distancing sedation in end-of-life care from physician-assisted suicide and euthanasia

Adoption of higher brain definitions of death

Lipuma argues that the presence of higher brain function bestows moral worth and meaning to a human life.(2) According to him, the induction of deep, irreversible unconsciousness, which leads to “dying an experiential death, that is, chemically and purposefully simulating the condition of one who is dead, based on higher brain functioning”, denies the patient of a meaningful existence, particularly when there is little chance that cognitive function can be regained.(2) He opines that it is higher brain function that gives a person a meaningful life, while lesser brain function merely represents a physiological function that keeps the body alive, bereft of meaningful existence. To 'correct' this implausibility and to accurately classify such patients as 'dead' despite still possessing functioning cardiovascular and respiratory systems, Lipuma proffers the “higher brain (neocortical) definitions of death”. As a result, patients on CSD would effectively be classified as 'dead', given that they had no chance of recovering their higher brain function.(2)

Argument on reversibility criterion

At the crux of Lipuma's earlier point is the issue of the irreversibility of iatrogenic-induced deep sedation in CSD. He explained that the cessation of sedation would simply result in the resumption of suffering ostensibly till death, which would be antithetical to good end-of-life care and the goals of care. As a result, Lipuma believed that CSD could not be reversed, thus leaving the 'reversibility criterion' without merit.(2)

Argument on double effect

A common defence of CSD has come from the invocation of the doctrine of double effect (DDE).(9) The DDE is seen to contain the following four principles: (1) the nature of the act must be 'good' or at least 'neutral'; (2) the 'good' effect cannot be brought about by the 'bad' effect; (3) the act must be proportional to the exigencies of the situation; and (4) the intention behind the act must be morally 'good' or at least 'neutral'.(10,11) Lipuma, however, contends that the induction of CSD would be in breach of three of the four central principles of the DDE.(12-14) The first is that the 'nature of the act' is neither 'good' nor 'neutral' given its associations with bringing about death.(12-14) The second condition within the DDE is also violated, as success of CSD can only come about through the 'bad' outcome; in other words, relief of suffering is only brought about by the induction of deep and irreversible unconsciousness, which is tantamount to 'death'.(12-14) The final condition of the DDE is infringed, as the good effect of symptom amelioration is realised only by bringing about the foreseen ill effects of deep irreversible unconsciousness.(12-14) Lipuma argued that the application of CSD in the knowledge that it is irreversible and that “eliminating consciousness is in effect the same as inducing a state that is functionally identical with neocortical death” crystallises the wish to hasten death.(2)

Condensing Lipuma's position into the practical triumvirate and personhood triad

Beyond the three arguments raised, a closer scrutiny of Lipuma's position reveals the presence of six considerations – three interconnected factors that relate to medical practice issues, referred to as the “triumvirate of interlinked practical considerations (or practical triumvirate)”, and three interrelated considerations that pertain to personhood, referred to as the “triad of personhood-related considerations (or personhood triad)”. For Lipuma's position to be accepted, both the personhood triad and practical triumvirate require universal acceptance in their own right.

The personhood triad arises from the belief that conscious function defines “a meaningful way for a human being to exist or to be considered alive”.(2) The personhood triad is composed of the beliefs that: (a) personhood pivots upon the presence of conscious function; (b) an irreversible loss of consciousness leads to an irretrievable loss of personhood; and (c) a loss of personhood negates 'what is meaningful to human life' and renders a patient dead.(2)

The practical triumvirate requires a homogenous practice of CSD in order to highlight the similarity of this practice with PAS/E.(2) The practical triumvirate is made up of the requirements that: (a) there is a common understanding of the practice, goals and indications of CSD; (b) the practice of CSD is applied in a consistent manner; and (c) when applied, CSD leaves little chance for the re-establishment of personhood.(2)

Primary sedative agents in CSD

On the issue of drug selection, Lipuma's attention is focused on the employment of barbiturates as the primary agent in inducing CSD, given that its use for induction of deep sedation is both risky and disproportional to the exigencies of the situation. Lipuma's motive for doing so is to crystallise the intention to hasten death. To lend credence to these efforts, he turns to Lo and Rubenfeld's account of the terminal sedation of Mrs B,(7) a 49-year-old woman with metastatic breast cancer who was admitted to the hospital for pain control of bone metastasis.(7) To ameliorate her intractable suffering, Lipuma highlights the use of barbiturates to induce CSD in this patient.

A closer scrutiny of Lo and Rubenfeld's article reveals that Mrs B received barbiturates to induce deep sedation only after the failure of treatments such as hydromorphone, fentanyl, morphine, celecoxib, amitriptyline, lorazepam and a course of radiotherapy to alleviate her symptoms, and as a result of dose-limiting side effects.(7) Furthermore, the authors added that phenobarbital was only employed as “midazolam was not permitted on the unit where the patient was receiving care”.(7) Our scrutiny of Mrs B's case further reaffirms the following cardinal features of acceptable sedative use in end-of-life care:

1. Last resort: CSD is a treatment of last resort when symptoms remain intractable and alternative treatment options are exhausted. 'Risky' options are only administered when 'safer' means have been exhausted or unavailable, and suffering persists.

2. Safety: CSD is employed only when there are effective means of ensuring safety and titration of these drugs. Lo and Rubenfeld state that a significant consideration for the employment of CSD is “that the site of care provides an appropriate level of nursing care and monitoring”.(7) In Mrs B's case, treatment was titrated under monitoring to ensure her life was not jeopardised. Data would suggest that Lo and Rubenfeld's use of focused and proportional application of sedatives may even prolong life in a minority of cases.(7,20-22)

3. Choice of agent: contrary to Lipuma's assertions, a review of regnant palliative care formularies and PS guidelines reveals that barbiturates are rarely employed in the induction of deep and continuous sedation in end-of-life care. In all recent PS guidelines, barbiturates do not feature either as first- or second-line agents in the application of CSD. Benzodiazepines are preferred to barbiturates primarily as result of their well-studied effects, rapid onset of action, short half-life and high therapeutic index, as well as greater physician experience in their employ, and the presence of an antidote, should the need arise.(7) The overall intent of safeguarding the patient is exemplified by the choice of a medication with an easily accessible antidote. Thus, should adverse reactions (such as respiratory depression) develop, prompt and appropriate remedial action can be taken to reverse the effects of the benzodiazepine and maintain the safety of the patient. This safety-first approach serves to refute Lipuma's objection to the reversibility criterion, i.e. that the effects of sedation cannot be reversed.(7,23,24) Indeed, a practical interpretation of the reversibility criterion may be that the effects of sedation can be reversed in light of potential risks to the patient's life, with the emphasis on the goals of CSD to ameliorate suffering safely and without the abbreviation of life.(7,23,24)

4. Impact of sedation: the impact of sedation on the cognitive abilities and conscious levels of most patients on which CSD is applied appears to be overstated. Claessens et al and Hui et al reveal that most patients tend to experience varying levels of consciousness as part of the natural dying process, which tends to occur 7–10 days prior to death.(25,26) This point is important given that Kohara et al(27) and Mercadante et al(28) both showed that CSD tends to be applied 2–3.4 days before death, suggesting that most patients would already be experiencing varying levels of sedation and cognitive impairment before CSD is applied.(25-28)

5. Variations in level of consciousness to sedation: Lipuma's key argument is that CSD results in the negation of higher brain function, which he believes is integral to a meaningful existence.(2) To advance such a stance, sedation applied in CSD must be shown to consistently suppress consciousness. Porzio et al report that midazolam-induced CSD fails to maintain deep continuous sedation in about a third of patients.(29) Barbato's use of bispectral index monitors also found variations in levels of consciousness in terminally ill, sedated patients.(30) In many cases, patients revealed some awareness or showed only minimal sedation. Before we highlight the stark implications of these findings upon Lipuma's higher brain function definition of death, we acknowledge that it is not midazolam that Lipuma discusses, but barbiturates. However, given that such practice is largely defunct, it is more clinically relevant to comment upon the effects of midazolam upon sedation, and the study by Barbato has shown that higher brain function is not entirely suppressed as a result of this practice.(30)

Clinically assisted nutrition and hydration

Mrs B's case(7) serves as a platform to discuss another of Lipuma's assertions – the routine cessation of clinically assisted nutrition and hydration (CANH) upon application of CSD. Lipuma believes that ceasing CANH in an unconscious patient leaves them susceptible to death by dehydration, which he asserts, uncovers the true intentions of the physician in applying CSD.(2) Such assertions highlight a number of misconceptions regarding the prevailing practices of sedation in end-of-life care. A review of the prevailing frameworks of PS shows that the provision of CANH is assessed on a case-by-case basis, thereby ensuring that clinical and psychosocial factors are given adequate consideration in each case.(31) Such heterogeneity in the employment of CANH in CSD, which highlights the variance in CSD practices, as well as the disproving of the belief that CANH is routinely ceased, further erodes Lipuma's position.(31)

In truth, Lipuma's conjecture on the underlying intention of physicians in 'routinely' ceasing CANH highlights further misreading of clinical practices, as well as Lo and Rubenfeld's case; the authors clearly stated that the patient had “already stopped eating and would have died shortly even had they [CANH] been administered.”(7)

Authorisation of treatment

Lipuma's comparison between CSD and PAS/E also extends to the manner that CSD is authorised. His concerns pivot on two key issues: first, he likens the authorisation process of CSD to that of PAS/E, and second, he asserts that authorisation of CSD is a clinical decision bereft of psychosocial input and oversight, and hence, paves the way for a slippery slope to PAS/E. Lipuma posits that the adoption of either practice crystallises and equates the intentions of the clinicians to those underpinning PAS/E.(2) Lipuma also suggests that authorisation for the application of CSD in Mrs B's case was done upon request, in a manner similar to PAS/E.(2) However, a review of Lo and Rubenfeld's original case presentation reveals that CSD was applied only when Mrs B's condition met the prerequisites for CSD set out within prevailing guidelines, rather than simply upon the request of the patient.(7)

Lipuma also uses the case of Mr Leo Oltzik, a mentally incompetent patient suffering from terminal agitation and dementia, congestive cardiac failure and kidney problems, as an example of CSD being applied surreptitiously with little involvement of the patient's family.(2,8) Lipuma suggests that circumnavigation of familial involvement in this case, drawn from Hartocollis' 2009 New York Times article, ushers in concerns of a 'slippery slope' leading to PAS/E. Lipuma further asserts that lax oversight of the application of CSD compounds these concerns.(2,7,8)

A closer scrutiny of Hartocollis' account, however, reveals that Lipuma's conclusions were misguided.(2,8) There are also a number of concerns with regard to the way Lipuma has employed Hartocollis' account of Mr Oltzik's care.(2,8) First, this was a newspaper article charting a medical encounter that reported the observations of a journalist who was neither part of the medical team nor likely to be privy to the circumstances and finer details of the case. Furthermore, the report did not meet the rigorous standards required for a case report published in most medical journals, and lacks critical clinical information and medical analysis. These concerns are amplified by Lipuma's paper, which omitted some critical information from the original newspaper report.(2,8) Contrary to Lipuma's assertions that the decisions for CSD and cessation of CANH were made unilaterally and without the input of the family, Hartocollis' report clearly states that both Mrs Oltzik and her daughter were aware of and party to the decision to sedate the patient.(8) Similar to Mrs B's case, CSD was applied to Mr Oltzik in keeping with the prevailing clinical guidelines for PS.(2) These guidelines included a holistic review by a palliative care team and/or an independent review by a professional, a prognosis of death in less than two weeks, the presence of intractable suffering, exhaustion of all viable treatment options and the likelihood of continued suffering until death without the provision of CSD.(32,33)

It is also clear that simplistic inferences as to the intentions behind the application of CSD cannot be defined solely upon a review of the actions surrounding the employment of CSD or indeed, the cessation of CANH. Krishna and Capps argued that intentions must be seen upon a more holistic plane and within the specific confines of the particular case setting.(4) The actions themselves must be taken in totality, and ought to take into account discussions and the decision-making process that preceded the action rather than be judged solely upon the 'final act'. Further, Krishna and Capps, and Krishna and Chin argue that delineation of the intentions that underpin clinical undertakings must be considered upon an evidence-based platform and in keeping with prevailing practice guidelines, legal frameworks and social expectations.(4,11,34) Any digression from accepted practice must be clearly justified, documented and ideally carried out with the oversight of a multidisciplinary team, if concerns about the intentions of the clinicians are to remain above reproach.(4)

Use of the term 'CSD'

Any comment about the intentions of a clinician must be made upon a clear understanding of the practice being discussed. This is particularly evident in Lipuma's misreading of Rietjens et al's study concerning the opinions of nurses who had experience employing PAS/E and CSD.(18) Lipuma suggests that nurses with experience employing PAS/E and CSD opined that there was little difference between the two interventions.(2,18) Scrutiny of Rietjens et al's original paper reveals that only four of the 16 nurses interviewed reported concerns with the employment of CSD.(18) Critically, the authors noted that these four nurses “had little experience with palliative sedation”.(18) This serves the further purpose of drawing attention to the poorly defined term employed by Lipuma to delineate the practice of sedation in end-of-life care.(2,18) Later studies by Rietjens and colleagues reaffirm that underpinning the fears voiced by the four nurses in the 2007 study by Rietjens et al was a lack of clarity on the practice of CSD.(35,36)

Proportionality and the overall goals of care

A further abortive effort to aggregate CSD under the umbrella of PAS/E sees Lipuma suggesting that “[b]oth procedures are only successfully completed upon Mrs B's complete biological death”.(2) Such a posit runs contrary to PS guidelines.(19) Prevailing guidelines on PS pivot upon the principle of proportionality and the temperate and monitored employment of sedation.(4,34) The endpoint of treatment is in fact the effective amelioration of diagnosed intractable suffering in the safest, most proportionate, effective and efficient means, using the lowest possible doses to do so.(4,34) There is no desire to hasten death.(8,4,34)

The higher brain function definition of death

The higher brain function definition of death, which Lipuma employs to proffer a new definition of death, straddles the practical triumvirate and personhood triad. According to this definition of death, any irreversible suppression of conscious function would be regarded as being tantamount to death.(2) There are two foci to Lipuma's efforts in advancing this definition of death. The first is to show that an irreversible loss of conscious function is tantamount to a loss of personhood and thus, equivalent to death. The second is to show that the application of CSD, which would in fact sedate a patient to unconsciousness, is an irreversible process. We have already refuted Lipuma's assertions against the reversibility criterion and will now address the issues pertaining to defining death on the basis of an irreversible loss of consciousness and the wider practical considerations of this framework.

There is significant data to suggest that contrary to Lipuma's belief that, once applied, sedation levels will be stable. Barbato revealed that there is a great variation in the levels of sedation and levels of consciousness among iatrogenic, deeply sedated patients.(30) These findings are further validated by data suggesting that even patients in persistent vegetative states and minimally conscious states manifest variations in consciousness levels when studied using positron emission tomography-computed tomography and electroencephalograms.(30,37-40) Attempting to establish an irreversible loss of consciousness, much less a stable level of unconsciousness required for the determination of brain death, becomes technically difficult.(24,34-43)

This is particularly so when changes in pharmacokinetics and efficacy of medications used in the treatment of conditions such as delirium, epilepsy, schizophrenia and pain can affect levels of sedations. Therefore, there is no reason to overturn the Academy of Medical Royal Colleges' 2008 Code of Practice for the Diagnosis and Confirmation of Death, which states that “the definition of death should be regarded as the irreversible loss of the capacity for consciousness, combined with irreversible loss of the capacity to breathe.”(44)