Answering the challenge: moral philosophy and the cognitively disabled Dominic Wilkinson 29/9/08 Does cognitive disability present distinct problems for philosophy and moral philosophy? Until fairly recently philosophers have given precious little attention to the question of the care, the internal experience and the moral status of those human beings whose cognitive capacities are significantly less than the rest of the species. But in the last 2 decades or so there has been increasing focus on the ethical and metaethical questions that relate to the cognitively disabled, particularly those who are most profoundly disabled. These questions were the focus of a conference in New York in late 2008 – “Cognitive Disability: a challenge to moral philosophy”.1 The conference raised a number of interesting questions, and the presentations will be available online as podcasts2, in a special issue of the journal Meta-philosophy, and in a published collection by Blackwell. Here I will briefly outline and draw links between two questions that underpinned a number of the presentations. Firstly, what is the appropriate conception of the good for individuals with severe cognitive disability3? Second, how should we think about and treat individuals with severe cognitive disability (SCD) compared to non-human animals with similar (or greater) cognitive capacities? Is it even appropriate to make such comparisons? Finally I will point to some important distinctions that are relevant to attempts to answer the question of how we should treat humans with SCD. What makes for a good life? This question is one of the most fundamental in philosophy, and great minds have been attempting to answer it for thousands of years without reaching a consensus. One broad way of classifying the answers that are given to this question is into subjective and objective.(DeGrazia 1995) Subjective theories point to the ultimate value of pleasure (hedonistic theories) or the fulfilment of desires (preference-based theories). On such theories things can be good for us insofar as we are aware of them and they lead to positive experiences, or if they satisfy our preferences. These things can be said to be in our interests. On the other hand objective value theories hold that certain activities or features make life good regardless of whether or not they are consciously perceived or preferred. These might include for example the development of deep reciprocating relationships, the pursuit of knowledge, the development and attainment of goals. There can be hybrid theories that include both subjective and objective elements (for example thinking about ‘rational preferences’ as well as actual preferences, or incorporating preference satisfaction into lists of objective goods), and to many, such combined theories are the most plausible accounts of what makes our lives valuable. SCD can be relevant to either of those two broad types of theory.4 The problem facing subjective theorists is firstly epistemic - individuals with SCD are frequently unable to communicate to us the nature of their experiences. We may not be able to tell whether or not an adult or a 1 18-20th September 2008, Stony Brook University, New York 2 http://www.stonybrook.edu/sb/cdconference/ 3 In this paper I will use the term ‘severely cognitively disabled’ to refer to… 4 Note that these issues are quite distinctive for cognitive disability compared to physical disability, and in particular for severe or profound cognitive disability. It is less clear, for example that moderate or mild cognitive disability affects wellbeing in such profound ways. child with SCD is feeling pleasure or pain. We may have little or no idea of their preferences, and indeed it may not be clear to what extent they are capable of forming preferences. Secondly, if SCD limits awareness and the development of desires or preferences, it may fundamentally change the range of things that are of value to an individual with SCD. For example, it has been argued that individuals who are not self-conscious cannot have a desire or preference for the future, and that consequently it may be much less of a harm to them to die (if it is a harm at all). (Singer 1993) Objective theories of value are also challenged by SCD, since many of those features that are typically listed as being valuable in human life are dependent upon rationality. Since cognitive capacity and reason are the features that are often noted as being distinctive of human life (as opposed to the life of non-human animals), this is not surprising.(Wilkinson 2006) Yet the implication of such theories is that the lives of those with SCD may contain less that it is of value (to them) than the lives of individuals without cognitive disability. This may give rise to controversial questions about priority when the interests of individuals with SCD conflict with the interests of the cognitively normal. Less controversially it may also have implications for consideration of the interests of patients with SCD who are faced with disorders or medical procedures that are likely to cause the patient to suffer. (Wilkinson 2006) Only one paper in the New York conference explicitly raised the question of value theory in relation to individuals with SCD, though the problem lurked beneath the surface of many points of contention. Leslie Francis and Anita Silvers argued that the aforementioned epistemic problem for subjective theories of the good could be overcome by using family, friends and guardians to help assess what is of value for the individual.5 They saw the trustee acting not by stepping in to the individual’s role, but by serving in a manner akin to a prosthetic limb. The trustee carries out a task, rational reflection, that the disabled individual is not able to perform unaided. Francis and Silvers’ paper was one of several at the conference that attempted to find a way to incorporate individuals with SCD into a liberal theory of justice. They, (as well as several other speakers at the conference), drew attention to the problem that Rawl’s contractualism appears to exclude the cognitively disabled. For Rawls the good was defined by a “rational plan of life”.(Rawls 1999, 380) He argued that individuals should be free to decide upon their own conception of the good, though when deliberating on the form that society should take (in the ‘original position’) they should be ignorant of their own individual values. (Rawls 1999, 347-9) However Rawls’ approach seems to have little to say directly about justice for those with cognitive disabilities; worse, some have argued that it may aggravate the disadvantages of those individuals compared to the rest of society. Martha Nussbaum presented her alternative approach to justice at the conference,6 and several speakers commented on Nussbaum’s book Frontiers of Justice, and the degree to which her 5 Francis L, and Silvers A. (Conference paper) Rethinking “Conceptions of the Good” in Light of Intellectual Disabilities. Cognitive Disability: a challenge to moral philosophy 2008 6 Nussbaum M. (Conference paper) Capabilities and Disability. Cognitive Disability: a challenge to moral philosophy 2008 favoured ‘Capabilities approach’ succeeded in overcoming the limitations of Rawls’ theory.7 Yet Nussbaum and the other speakers remained largely silent on the question of what is of value to individuals with cognitive disability and how that might be different from cognitively normal individuals. Nussbaum’s list of capabilities to be promoted appear to be based upon an Aristotelian idea of human flourishing. During her presentation Nussbaum mentioned that they had emerged from international discussion and consensus. They include being able to “live to the end of a life of a human life of normal length”; being able to use the senses “to imagine, think and reason – and to do these things in a “truly human” way”; “being able to form a conception of the good and to engage in critical reflection about the planning of one’s life”; “being able to participate effectively in political choices that govern one’s life”. (Nussbaum 2006, 76-7) However it was not clear from the discussion whether Nussbaum thinks of these capabilities as being objectively valuable, or whether they are instrumentally valuable in promoting the subjective wellbeing of individuals with SCD. If they are objectively valuable, then in the absence of ways to enhance the cognitive abilities of individuals with SCD their lives appear to be stripped of many of the distinctive values of human life. If they are supposed to be subjectively valuable then it is not clear how some of the capabilities promote the wellbeing of the severely cognitively disabled. One of the implications of Nussbaum’s capabilities approach that she freely acknowledged, was that individuals with cognitive disability or their guardians should have a right to participation in the political and judicial process in terms of voting and involvement on juries.8 There are some reasons why political participation may be of instrumental benefit to adults with SCD.9 Politicians have traditionally neglected the welfare of the disabled. Extending voting rights to individuals with SCD may force political parties to think seriously about the ways in which their policies will affect, and be perceived by, disabled adults and their guardians. This step might also reduce other forms of discrimination against disabled individuals because of its explicit recognition of their place in society. Yet such instrumental justification might also justify granting political representation to others who are traditionally discriminated against in the political process such as future generations or non-human animals.10 Nussbaum explicitly rejected this latter possibility on the grounds that certain goods are valuable only for members of particular species; she claimed that voting would not be of interest or benefit to an elephant for example.11 But why does species membership matter when we are considering how we should behave towards an individual? The question of how we should treat non-human 7 Berube M. 2008 (Conference paper) Response to Nussbaum; C. Stark. 2008 (Conference paper) Reasons, persons and capabilities; Wong S. 2008 (Conference paper) Duties of justice to citizens with cognitive disabilities. 8 Nussbaum M. 2008 (Conference paper) 9 Nussbaum rejected the idea that her capabilities approach would also require the extension of voting rights to children or their guardians. It was not clear however how this distinction was justified. 10 Singer P. 2008 (Conference paper) Speciesism and moral status. 11 Nussbaum M. 2008 (Conference paper) response to questions animals compared to humans with SCD came up repeatedly during the conference. Some speakers questioned whether it is appropriate to even ask this sort of question, and the debates about the relevance of species membership for moral status yielded some of the most heated discussions of the conference. Yet in itself the question of whether species membership is relevant to our moral obligations seems not to be one that should generate offence. It is no different to asking whether height, or religion or social class are relevant factors that we should consider. I suspect that the reason that some individuals have such visceral responses to making comparisons between species relates to two factors. Firstly individuals with SCD are related to other humans in ways that non-human animals are not. They are someone’s daughter, someone’s son, someone’s brother or nephew or cousin. We feel strong bonds of kinship with our family in ways that we do not with non-human animals. A number of the speakers at the conference had children or close relatives with disability. Such experience helps to enrich philosophical discussions about disability, but it also makes it hard for those who have opinions that diverge from those of disability advocates to express their arguments without risking causing offence. Eva Kittay, whose daughter has multiple disabilities expressed her palpable “nausea” at being asked to justify a distinction between what we owe to animals such as a chimpanzee, a dog or a pig and cognitively disabled humans. I will return to the question of the ‘relatedness’ of individuals with SCD and whether it justifies treating them differently from non-human animals with similar cognitive capacity. However if relatedness matters, it does not seem to invalidate the question of comparisons between species. The second reason that some may object to comparisons between species is that they lead to conclusions that conflict fairly strongly with commonly held beliefs. Philosophy may offend by taking us to places where we would rather not go. One way to avoid such unpalatable conclusions is to avoid the comparison in the first place – however it is not clear that there is a non-question-begging way of justifying this. Peter Singer presented a familiar argument about moral status and ‘speciesism’.12 Both he and Jeff McMahan13 drew attention to a commonly cited defense of differential treatment of non-human animals ie that they differ in cognitive capacity, rationality, and use of language from humans. Yet some animals at least have cognitive capacities that are on a par with many humans (Singer cited a gorilla able to use and understand 1000 signs, and a parrot able to understand abstract concepts like ‘shape’ and ‘colour’). And some humans with SCD have cognitive capacities that are lower than many non-human animals. Singer argued that cognitive capacity is relevant to interests, and that consequently we should treat individuals with similar cognitive capacities similarly.14 12 13 Singer P. 2008 (Conference paper) Speciesism and Moral Status McMahan J. 2008 Cognitive disability, cognitive enhancement and moral status. 14 Interestingly Singer was attacked after his presentation for caring more about nonhuman animals than about humans with disabilities, and for supporting better treatment of animals than of individuals with SCD. His reply was he did not advocate, and had never advocated such a difference, and seemed puzzled that his interlocutors had misunderstood his position. One possible explanation for the misunderstanding is a sort of parallax error in interpretation of Singer’s views. He argues for animals to be treated better than they currently are (and than many people think they should be), while also arguing that our moral obligations to individuals with SCD are less than currently held. This conjunction of views might give the impression that he is Singer did not discuss this, but it should be noted that this does not require equal treatment, since the desiderative interests of individuals may diverge. However it seems to require that the minimum that is morally required of us should be broadly similar. But what does this mean in practice for our treatment of humans with SCD and non-human animals. McMahan has suggested in his book ‘The Ethics of Killing’, that we have three options. (McMahan 2002, 206) We could treat humans with SCD on a par with our current treatment of non-human animals (an option he calls Consistent Elitism); we could improve our treatment of non-human animals to the level that we currently treat humans with SCD (Radical Egalitarianism); or alternatively we could improve the treatment of non-human animals, but at the same time lower the standard for treatment of humans with SCD so that they come together (Convergent Assimilation). McMahan rejects the first two as implausible, and argues that the third option is the only defensible alternative. Yet even if some form of convergent assimilation is accepted, this does not require us to treat identically humans and non-humans with similar cognitive capacity. There are two reasons for this. (McMahan 2002, 231) Firstly, while the interests of an individual may determine what we are morally required to take into account, it is always possible for us to do more than this. There is nothing to say that we cannot treat humans better than non-humans, as long as this does not prevent us from treating non-humans to the minimal standard that their interests require. Such preferential treatment might be thought of as superogatory. However it may also reflect the difference that I noted earlier, that humans with SCD are related to other (cognitively normal) members of the community in ways that non-human animals are not. These relationships may create agent-relative obligations for family members of an individual with SCD. They also mean that there are other significant interests that may be affected by our treatment of humans with SCD that do not need to be taken into account (or are less strong) in the treatment of non-human animals.15 ‘Relatedness’ may justify different treatment of humans with SCD than cognitively similar non-humans. However there are also reasons why some may have qualms about relying on it to justify a kind of speciesism. Firstly, it may imply that where such relations do not hold, it may be justifiable to treat a human with SCD quite differently from the way that they are at present treated. For example, a child who was orphaned, without any available adoptive parents. It may also mean that the level of treatment of individuals with SCD would be contingent upon the degree to which their families care for their wellbeing. Would society be justified in treating less stringently a child or adult whose family had lost interest in them? Secondly it may mean that if there were relationships of a certain strength holding between humans and non-human animals, that such animals should be treated relevantly similarly to advocating preferential treatment of animals, yet in fact there is such a wide gulf between our treatment of humans and non-humans, that his arguments merely represent some narrowing of that gap. 15 Of course most non-human animals are also ‘related’, in the sense that they are the son, daughter, brother etc of other non-human animals. These relationships are important to them, and may historically have been underestimated by humans, however the strength of the future-directed preferences of such relatives are likely to be less (as a rule) than their human counterparts, and this may lead us to think that the corresponding interests are less strong. humans with SCD. In fact some existing non-humans have very strong relationships with humans, and where those humans are able to afford it, that does lead them to be treated in a way that is substantially greater than other non-humans, and not that differently from humans. Again such comparisons risk causing offence, but there may be good reasons for making them.16 In this paper I have outlined some of the challenges that SCD presents to moral philosophy. I have explored two issues that underpin philosophical debate in this area. The relevance of SCD to wellbeing depends upon our conception of the good. Some disagreements about SCD, and how we should treat disabled individuals turn on this question of what makes a life valuable to that individual. Unless our answer to this question is dependent upon species membership, the question of how we should treat individuals with SCD compared to non-human animals is impossible to avoid. I have argued that even if speciesism is rejected, it may still be possible to justify preferential treatment of humans with SCD over similarly cognitively endowed non-humans. It is beyond the scope of this paper to outline what should be the level of care of humans with SCD. However I will end by listing a number of key distinctions that I believe should be made when attempting to answer this question. When considering the treatment of individuals with SCD we should distinguish 1. Decisions about the lives of individuals with SCD, from decisions about whether they live 2. What we are morally required to do (the minimal standard), from what we are permitted to do, or what may be superogatory to perform 3. What we are required to do by virtue of the intrinsic morally relevant properties of an individual from what we owe to them by virtue of any special relationships that we may have with them, or they may have with other individuals 4. What we actually care about from what we have reason to care about 5. What we should in general do for individuals, from what may be, in certain circumstances, justifiable. These distinctions are sometimes blurred in debates, and were at times during the New York conference. Being clear about these may help to resolve disagreement and answer the challenge that cognitive disability presents to moral philosophy. REFERENCES DeGrazia, D. 1995. Value theory and the best interests standard. Bioethics 9(1): 5061. McMahan, J. 2002. The ethics of killing: problems at the margins of life. New York: Oxford University Press. 16 There may be some residual asymmetry between human-human and human-animal relations. It is not clear whether the strength of feeling of a human to a dog, or a horse, or other animal can ever reach the same level as the strength of feeling of a parent for their child. It also does not seem that an animal-human relationship would generate the same sort of agent-relative reasons as a parental or filial relationship. Nussbaum, M. C. 2006. Frontiers of justice : disability, nationality, species membership. Cambridge, Mass. ; London: The Belknap Press of Harvard University Press. Rawls, J. 1999. A theory of justice. Oxford: Oxford University Press. Singer, P. 1993. Practical Ethics. Cambridge: Cambridge University Press. Wilkinson, D. 2006. Is it in the best interests of an intellectually disabled infant to die? J Med Ethics 32(8): 454-9.

‘Neglected Personhood’ and Neglected Questions: Remarks on the Moral Significance of Consciousness Open peer commentary on Fins. et al “Neuroimaging and disorders of consciousness: envisioning an ethical research agenda” Dominic Wilkinson1,2, Guy Kahane1, Julian Savulescu1 1. Oxford Uehiro Centre for Practical Ethics, University of Oxford 2. The Ethox Centre, University of Oxford Corresponding Address: Dr Dominic Wilkinson, Uehiro Centre for Practical Ethics, Littlegate House, St Ebbes St, Oxford, OX1 1PT, United Kingdom dominic.wilkinson@green.ox.ac.uk Word Count: 1483 words Page 1 of 10 Introduction Fins et al. (2008) outline a number of ethical issues associated with neuroimaging in conditions where consciousness is absent or impaired, and make a number of useful suggestions. However they are virtually silent on what we take to be crucial questions Here we briefly discuss two key issues. Firstly, talk about consciousness is notoriously contentious and can refer to a range of potentially distinct phenomena, not all of which might be of moral significance. Secondly, if some of these patients are indeed conscious in a morally significant sense, what would be the implications for treatment? The significance of consciousness Our existing everyday and clinical criteria for ascribing consciousness are in the main behavioural. This is reflected in current diagnostic criteria for VS, where absence of consciousness is defined in behavioural terms. However we should not confuse a phenomenon with our means of finding out about it. The great interest in neuroimaging severely brain damaged patients is precisely that it offers an entirely new means of detecting consciousness even in the absence of behavioural response. (It is of course a separate and controversial matter whether the brain activation detected by neuroimaging amounts to more than a further source of evidence for the presence of consciousness.) Page 2 of 10 Consciousness is central to the ethical issues raised by patients in VS. Controversies about withdrawing life sustaining treatment (LST) from such patients have revolved around the question of whether they have irreversibly lost the capacity for consciousness, a matter taken to have momentous moral significance. By contrast, patients in the minimally conscious state (MCS) are commonly assumed to have a measure of consciousness, although accompanied only by minimal cognitive capacities. Given that current neuroimaging studies appear to suggest that some brain-damaged patients diagnosed as in VS might in fact be in MCS, a crucial ethical question must concern the moral significance of the difference between the absence of consciousness and its minimal presence. The moral problem of consciousness Consciousness is a contentious concept. Fins, Illes et al. (2008) write that consciousness encompasses “subjectivity, sentience, self-awareness and an ability to appreciate the relationship between the self and the environment,” and recognize that it has been philosophically identified both with access to information and with phenomenality. Indeed philosophers of mind distinguish these and further senses of consciousness. As Ned Block recently remarked, “the word ‘consciousness’ connotes a number of different concepts and denotes a number of different phenomena” (Block 2002; see also Levy 2008). When Fins et al. (2008) write that consciousness is required for intentional behaviour, this may be correct only for certain senses of consciousness. The problem that this diversity raises for medical practice cannot be side stepped, Page 3 of 10 as the authors suggest, by adopting a neurobiological approach that focuses on the different neural processes that might underlie these phenomena. After all, the difficulty raised by disorders of consciousness in brain damaged patients is precisely that these different aspects of consciousness seem to come apart, and it cannot be simply assumed that all of them have equal (or any) moral significance. For example, even if Owen et al (2006) are correct to claim that their fMRI study showed their patient to be conscious in some sense (a matter still under scientific and philosophical dispute) this finding might make no moral difference if she is not conscious in a sense that is significant. Here we can only make brief remarks about the moral significance of one central notion, that of phenomenal consciousness or qualia. This notion is notoriously difficult to define: a state is phenomenally conscious if there is something ‘it is like’ to be in it. We assume that genuine VS patients lack subjectivity whereas patients in MCS possess it at least intermittently. From a philosophical standpoint, however, talk of minimal consciousness is misleading, as what MCS patients have only in minimal form are rather cognitive and motivational capacities. Arguably, the value of phenomenal consciousness is grounded in the moral significance of interests (Kahane & Savulescu 2008). The interests of persons include hedonic, desiderative and objective elements. Hedonic interests relate to states of suffering or enjoyment. Desiderative interests relate to the satisfaction of a person’s desires. Objective interests relate, more Page 4 of 10 contentiously, to having a meaningful existence by possessing goods such as relations with other people, achievement, or knowledge. It is clear enough why phenomenal consciousness is needed if one is to have hedonic interests. Pleasure and pain matter only because of what it is like to experience them. Arguably, phenomenal consciousness is also needed if one is to have desiderative and objective interests: without subjectivity there is no point of view to which such interests can be ascribed. Treatment decisions and consciousness One of the most ethically contentious decisions concerning patients with severe disorders of consciousness is the withdrawal of LST. Ethical and legal justification for withdrawal of LST in such patients is based upon patients’ prior wishes as well as upon considerations of futility or the patient’s best interests. Some of the scientists involved in fMRI research in VS patients have suggested that neuroimaging evidence for consciousness in such patients would have profound implications for decision-making about withdrawal of LST, (Owen 2008) and something of this nature appears to be implied by Fins et al. (2008) concern about “neglected personhood”. There have already been calls for the use of fMRI in this setting in a legal case in the UK.(Lewis 2007) There appears to be a prevailing assumption that such evidence for consciousness would provide a strong reason not to withdraw LST from a patient. However this common assumption is highly problematic, and there are grounds for thinking that the opposite might in fact be the case. Life with minimal consciousness may be worse than the VS.(Wilkinson et al 2008) Page 5 of 10 Neuroimaging findings might be relevant to such decisions in several ways. They may be relevant to the patient’s previously-expressed wishes and their desiderative interests; this will depend upon how their desires were expressed. For example a patient may have expressed a desire for their life not to continue if they were permanently unconscious or they may have explicitly referred to the possibility of minimal consciousness. Fins et al. (2008) discuss the potential incorporation of our changing understanding of VS/MCS into advanced directives. Neuroimaging may be relevant for the prognosis of patients with disorders of consciousness, and thus to considerations concerning futility. Patterns of preserved higher cortical function may place patients into a subgroup of vegetative patients with a better prognosis, or may portend partial recovery as reported for a small number of cases,(Owen et al 2006, Di et al 2007). Further evidence is needed before such claims could be drawn with confidence. And on the basis of the patients reported to date, the extent of the recovery appears to be limited. The three patients who recovered were reported as reaching the MCS,(Owen et al 2006, Di et al 2007) and one has apparently subsequently relapsed.(Owen, personal communication) Whether such findings portend meaningful recovery is another question. Indeed minor degrees of recovery may be worse than none at all. If some patients diagnosed on behavioural grounds as being in VS are actually in a state closer to the MCS, this may be worse for them in one Page 6 of 10 important way. If patients in MCS possess phenomenal consciousness then, unlike patients in VS, they are able to feel pain, although they are largely unable to communicate their distress and discomfort.(Nelson et al 1999) We may do them significant harm by keeping them alive: it is doubtful that a life that contains suffering while containing little or no desiderative or objective goods is a meaningful life. Judicial and ethical argument in VS has sometimes relied upon an inversion of the best interests standard, to suggest that since such patients are completely unaware, they do not have interests; consequently continuing LST cannot be in their interests.(Fenwick 1998) However if such patients are partly aware and suffering we may have a stronger reason to let them die – it may be against their interests to continue LST. Beyond considerations of pain and suffering, the compromise of their objective and desiderative interests may make their lives worse than nonexistence. Conclusion Fins et al. (2008) make a number of recommendations for a framework for ethical research in this field. They have however remained silent on some of the most important questions about the clinical translation of neuroimaging findings. Difficult philosophical questions about consciousness and its moral significance need to be addressed before the scientific and medical community can adequately interpret the results of research into consciousness in brain damaged patients. Most importantly, the question of what makes life worth living will need to be addressed if we are to decide Page 7 of 10 whether we are benefiting or harming this group of patients by keeping them alive. Greater clarity on these issues is needed to avoid fostering false hope in families of affected patients. Page 8 of 10 Block, Ned, Some Concepts of Consciousness, in Philosophy of Mind: Classical and Contemporary Readings, David Chalmers (ed.) Oxford University Press, 2002. Di, H.B., Yu, S.M., Weng, X.C., Laureys, S., Yu, D., Li, J.Q., Qin, P.M., Zhu, Y.H., Zhang, S.Z., and Chen, Y.Z. Cerebral response to patient’s own name in the vegetative and minimally conscious states, Neurology 2007; 68; 895-899. Fenwick A. J. Applying best interests to persistent vegetative state--a principled distortion? J Med Ethics 1998;24(2):86-92. Fins J and Illes J, Bernat J, Hirsch J, Laureys S, Murphy E. Neuroimaging and disorders of consciousness: envisioning an ethical research agenda. American Journal of Bioethics 2008. Kahane, G. and J. Savulescu. Brain-damaged Patients and the Moral Significance of Consciousness. Under review. 2008 Lewis P. Withdrawal of treatment from a patient in a permanent vegetative state: judicial involvement and innovative 'treatment': an NHS Trust v. J. Medical law review 2007;15(3):392-9. Levy N. Going beyond the Evidence. American Journal of Bioethics 2008 Nelson L. J., Cranford R. E.. Michael Martin and Robert Wendland: beyond the vegetative state. The Journal of contemporary health law and policy 1999;15(2):427-53. Owen, Adrian M., Coleman, Martin R., Boly, Melanie, Davis, Matthew H., Laureys, Steven, Pickard, John D. Detecting Awareness in the Vegetative State, Science, 313, 8 September 2006. Owen, Adrian M. and Coleman, Martin R. Functional neuroimaging of the vegetative state. Nature Reviews: Neuroscience. 9, 2008. 235-243. Page 9 of 10 Wilkinson, D., G. Kahane and J. Savulescu. Letting Vegetative Patients Die: What are the Implications of Functional Neuroimaging? Under review. 2008 Page 10 of 10

Journal of Perinatology (2006) 26, 772–776 r 2006 Nature Publishing Group All rights reserved. 0743-8346/06 $30 www.nature.com/jp STATE-OF-THE-ART Influenza in the neonatal intensive care unit DJ Wilkinson1, JP Buttery2,3,4,5 and CC Andersen1 1 Neonatal Unit, Mercy Hospital for Women, Melbourne, Australia; 2NHMRC Centre for Clinical Research Excellence in Child and Adolescent Immunisation, Melbourne, Australia; 3Murdoch Children’s Research Institute, Melbourne, Australia; 4Department of Paediatrics, University of Melbourne, Melbourne, Australia and 5Department of General Medicine, Royal Children’s Hospital, Melbourne, Australia Biology of influenza Influenza is an RNA virus of the orthomyxoviridae family, classified into types A, B or C. Type A affects many animal species, and is responsible for most clinical infection in humans. Type B is responsible for about 11% of non-pandemic influenza in humans, whereas Type C causes only a mild coryzal illness.4 Influenza A nomenclature is further subdivided on the basis of the surface glycoproteins hemagglutinin (H) and neuraminidase (N), for example, ‘H1N1’. At least 15 hemagglutinins and nine neuraminidases have been described and all are antigenically distinct. These glycoproteins are major virulence factors. Hemagglutinin mediates viral binding to cell receptors, whereas neuraminidase has a crucial role in the release of virus from the cell following viral replication.5 Minor changes in surface antigens (antigenic drift) contribute to seasonal epidemics of influenza A, whereas major changes (antigenic shift) have been associated with pandemics. Influenza has historically been an uncommon illness in the newborn period, although epidemic outbreaks in neonatal intensive care units have been described. There is currently significant concern about the possibility of a new pandemic of influenza in the near future. During a pandemic neonates are likely to be exposed, with significant illness more likely in preterm newborns due to reduced levels of passively transferred protective maternal antibodies. While newer therapies have been shown to be effective in reducing the severity of illness in adults and children, such therapies are untried in neonates. Supportive care and measures to contain and prevent spread of infection may well be the most important measures in the event of a neonate acquiring influenza, including the avian variety. Journal of Perinatology (2006) 26, 772–776. doi:10.1038/sj.jp.7211625 Keywords: influenza; infant, premature; infant, newborn; intensive care, neonatal; neuraminidase/antagonists and inhibitors Introduction There is widespread concern about a potential new pandemic of influenza A of avian origin with increasing reports of H5N1 avian influenza affecting birds and humans across multiple continents.1,2 Pandemic planning has focused for the most part on adults and older children.3 While influenza is an uncommon illness in newborn infants, it would seem timely to review the clinical features and management of influenza in the newborn. This paper will also review the epidemiology of influenza in newborn infants, anti-viral therapy, vaccination and control measures in a neonatal intensive care unit in the event of an outbreak. The rarity of influenza in this population, and the lack of controlled trials of therapy, of necessity means that recommendations are based largely upon level 4 and 5 evidence. Correspondence: Dr DJ Wilkinson, Neonatal Unit, Mercy Hospital for Women, 163 Studley Road, Heidelberg, Victoria 3084, Australia. E-mail: dwilkinson@mercy.com.au Received 31 May 2006; revised 1 September 2006; accepted 19 September 2006 Influenza pandemics and newborn infants There were three pandemics of influenza in the 20th century, in 1918, 1957 and 1968,3 but the most recent of these occurred before the development of modern neonatal intensive care. There are few descriptions of the effect of those pandemics on newborn infants. It is clear however that the 1918 to 1919 pandemic was associated with an increase in neonatal and post-neonatal infant mortality (as well as an increase in pre-term delivery).6 In more recent times, outbreaks of influenza have been described in neonatal intensive care units,7–10 and these have sometimes coincided with community-wide epidemics. Furthermore, recent avian influenza (H5N1) disease has been particularly severe in infants and young children.2 Transmission of H5N1 influenza to date has been almost entirely owing to direct acquisition from birds. The fear is that change in the viral genome, potentially by genetic mixing with coinfecting human influenza strains, could make human to human transmission possible, thereby setting the scene for rapid spread within the population.1 Influenza in the NICU DJ Wilkinson et al 773 When the next pandemic arrives, it is likely that newborn infants will be exposed and develop clinical illness. Clinical features of influenza in newborn infants Influenza is a significant cause of pediatric hospital admission,11 with its highest attack rate in pre-school and school age children.4,12 The classic clinical features include high fever, myalgia, headache and malaise, whereas a small proportion have signs of pulmonary involvement.11,13 In comparison, influenza is an uncommon infection in the first 6 months of life with generally mild symptoms.12 Community studies and studies of infants during epidemics have found that a high proportion of infections are asymptomatic.8,9,12 Where infants do manifest symptoms those most commonly described are abrupt onset of high fever and symptoms of upper respiratory tract infection.7–10,13 Such clinical features may be indistinguishable from bacterial sepsis.11,13,14 The milder illness in infants and newborns has been attributed to transplacental acquisition of protective antibodies (which can provide protection for 3 to 6 months after birth),9,11,15 as well as protection from breast milk.11 However, epidemic and pandemic influenza is associated with significant mortality in infants.6 There were 153 influenza-related pediatric deaths reported during the US influenza season in 2003 to 2004, with the highest pediatric mortality rate in infants less than 6 months old.16 Severe influenza has been described in a small number of newborn infants (Figure 1).9,12,17,18 Glezen et al.12 described an 11 day-old-term infant who died suddenly from pneumonitis presumed related to influenza A. That infant’s mother had developed influenza A (H3N2) 6 days post-partum, and the infant had no detectable antibody to the epidemic virus in her cord serum. A 27-week gestation twin died during an epidemic described by Cunney et al.9 of influenza A virus-associated hemophagocytic syndrome. Van den Dungen et al.18 described a 29-week gestation infant who became unwell with a respiratory illness on day 17 of life (viral cultures grew influenza B), developed significant lung injury and neurological complications and subsequently died. Epidemic influenza in the neonatal intensive care unit Part of the reason that influenza is so uncommon in the newborn period is presumably related to the reduced contact neonates generally have with unwell adults or children. However, there have been several descriptions of nosocomial spread of influenza A within neonatal units.7–10 These outbreaks usually coincided with epidemics of influenza within the broader community, although in each case it was difficult to pinpoint the source. In the largest outbreak, 30 of 95 infants admitted to two neonatal units in Spain were positive for influenza A virus.8 Cunney et al.9 described a similar outbreak involving 19 out of 54 infants in a level III nursery in Canada. In both of these instances, the majority of newborns were asymptomatic. Furthermore, with the exception of one pre-term infant, all of the symptomatic newborns recovered completely. Mechanical ventilation, twin pregnancy, gestational age and birth weight were identified as risk factors for infection.9 Anti-viral agents in newborn infants A number of different anti-viral therapies have been used for the treatment of influenza in children. However, in infants there is very limited evidence for either safety or efficacy of these agents. None of the M2 or neuraminidase inhibitors are licensed for use in young children either in Europe4 or in USA.19 Similarly, there have been no studies demonstrating safety of anti-viral therapies for influenza during pregnancy or breast-feeding, and they should be used only if the potential benefit to the mother outweighs the possible risks to the fetus or infant.20 Ribavirin Ribavirin is a nucleoside analog with some in vitro activity against influenza A and B. Its main use has been in the treatment of respiratory syncytial virus (RSV) infection.21 Ribavirin is usually administered by inhalation, but can be given by intravenous infusion, the latter having been described in adults with severe influenza,22 and a young child with influenza A and cardiomyopathy.23 Ribavirin is teratogenic, making inhalational administration problematical owing to potential staff exposure.24 M2 inhibitors Amantadine and rimantadine inhibit the replication of influenza A by acting on the viral protein M2 (they have no activity against influenza B).25 They have been used for some years for prophylaxis and treatment of influenza infection,19 but resistance may develop rapidly, and in addition they are associated with reversible central Journal of Perinatology Figure 1 Influenza A pneumonitis in a 27-week gestation pre-term infant who became unwell on day 54 of life. Chest X-ray on day 3 of admission. Influenza in the NICU DJ Wilkinson et al 774 nervous system side effects.10,19 There are no randomized placebocontrolled trials of either agent for the treatment of influenza in children.19 They are not effective against some of the possible pandemic strains (e.g. H5N1).3 Amantadine and rimantadine have been shown to be teratogenic at high dose in animal studies and should not be used in pregnant women or while breast-feeding.20 Neuraminidase inhibitors The most recently developed agents are the neuraminidase inhibitors oseltamivir and zanamavir. They are effective against both influenza A and B, and have been shown to reduce duration of illness in previously healthy children26 and adults.27 They are also associated with less adverse events than the M2 inhibitors.4 Treatment (as with any of the above agents) must be started within 48 h of the onset of symptoms to be effective.11 For prophylaxis, neuraminidase inhibitors and M2 inhibitors are effective in adults in preventing influenza when started within 48 h of exposure, but there have been few studies in children.25,26 Zanamavir and oseltamivir cross the placenta and are excreted in breast milk in animals, but have not been found to be teratogenic.5 Their safety in humans during pregnancy or breast-feeding has not been established. Specific safety concerns about the use of neuraminidase inhibitors in infants were raised after the release of a US Food and Drug Administration animal trial of oseltamivir in 2003.28 Seven-day-old rats were fed 1000 mg/kg of oseltamivir (250 times the dose recommended in children). The treatment led to the death of many of the rats, with brain levels 1500 times higher than adult animals exposed to the same dose.28 The clinical relevance of this is unclear, however it has led to recommendations that neuraminidase inhibitors not be used in infants.28 Nevertheless, in Japan, where the rate of encephalopathy after influenza in childhood is high, many children and infants have been treated with oseltamivir.29 Two case series of a total of 150 infants younger than 1 year treated with oseltamivir found no association between oseltamivir use and mortality or encephalopathy.29,30 The other limiting factor for the use of anti-viral agents in critically ill neonates relates to the method of administration. Zanamavir and ribavirin are usually administered by inhalation.19 Oseltamavir, amantadine and rimantadine are only administered Table 1 orally, with suspensions available for each.19 The only intravenous preparation is that of ribavirin. Influenza vaccination Influenza vaccination is the primary means of prevention of influenza in the community, and has been shown to be safe and effective in preventing significant illness in adults and children.20 Inactivated and live attenuated forms are available. Inactivated influenza vaccine has been approved for use in infants older than 6 months, and is routinely recommended in the US (although not currently in Europe or Australia) in infants aged 6 to 23 months.4,20,31 The vaccine is not thought to be immunogenic in infants less than 6 months.20,32 Live attenuated vaccines are not recommended for children younger than 5 years.20 Vaccination is safe throughout pregnancy and while breastfeeding.20 Influenza infection during pregnancy is associated with increased morbidity, and consequently influenza vaccine is recommended for women who will be pregnant during an influenza season.20 Additionally, vaccination may provide protection for the neonate during the first 6 months of life.33 All current vaccines are grown initially in embryonated hens’ eggs, and consequently there is a risk of hypersensitivity reactions in individuals with severe egg allergy.20 There are candidate pandemic vaccines in development, some of which have been shown safe and immunogenic; however, there are no efficacy data to date.2 Prevention of nosocomial influenza Given the lack of proven and safe treatment for influenza in neonates, prevention of spread of infection may be the most effective measure in the context of an epidemic or pandemic (Table 1). Reducing transmission of virus Influenza virus is primarily spread from person to person by respiratory droplets.34,35 Consequently cohorting and isolation of affected infants is recommended by most authorities. The relative importance of airborne (droplet nuclei) spread is unclear, and in general negative pressure isolation is not required.34,35 In a nosocomial RSV epidemic in a pediatric intensive care unit droplet Key measures in the event of an outbreak of influenza:37,43 Screen patients using a rapid diagnostic test to enable cohorting of pre-symptomatic or asymptomatic infants. Affected infants should be cohorted and droplet precautions taken to minimize spread. Infants admitted from home and other units should be isolated until screening test results are available. All staff and parents should be immunized. Staff or parents who are unwell should not enter the unit, or should be sent home. Anti-viral prophylaxis should be provided for staff and parents if unvaccinated, or if infection has occurred with a strain not covered by the vaccine. Journal of Perinatology Influenza in the NICU DJ Wilkinson et al 775 precautions were more important than a physical barrier.36 Influenza can survive on a variety of fomites and in the environment, and although the clinical significance of direct contact spread is unclear, good hand-washing is likely to reduce transmission.35 Timely diagnosis is essential for effective cohorting. Traditional methods of detection have included immunofluorescence or rapid diagnostic tests (using immunoassay or viral neuraminidase detection).19 The use of rapid diagnostic tests for influenza enabling detection and cohorting of infected infants has been associated with reduced nosocomial transmission in pediatric hospitals.37 Detection of viral RNA by polymerase chain reaction (PCR) is the most sensitive technique, and the recent development of real-time PCR38 has led to this becoming the method of choice for rapid, reliable diagnosis of influenza. To avoid exposure of infants to influenza, staff are recommended not to work when unwell; however, compliance with this by medical and nursing staff has historically been poor.9,34,39 Other measures to reduce spread are of unknown benefit. For example, the application of viral filters to the expiratory arm of ventilator circuits has been described in the neonatal setting (in the context of an outbreak of severe acute respiratory syndrome (SARS)),40 and may be useful for ventilated infants with influenza. During one of the aforementioned neonatal intensive care outbreaks procedures for changing ventilator tubing were altered in order to reduce aerosol spread.9 Immunization To prevent transmission of influenza to vulnerable populations, it is recommended that health-care workers should be immunized annually against influenza.20 Again compliance rates amongst staff in neonatal units (and elsewhere in hospitals) have been low.9,10 Routine immunization of parents of high-risk infants in the neonatal unit in addition to that of pregnant mothers has also been advocated.41 It is unclear how much protection vaccines will provide in the event of a novel strain. Chemoprophylaxis During epidemics, prophylactic treatment with anti-viral agents for unimmunized parents and staff has been suggested by some authorities.20 They are an adjunct to immunization and should not replace it.20,42 Most reported experience has been with M2 inhibitors. Amantadine use has been described in the two larger neonatal intensive care outbreaks.8,9 It remains to be seen whether neuraminidase inhibitors will be more effective and better tolerated.34 outbreaks within neonatal units have been surprisingly low. However, pandemic influenza may well be more severe in this population because of the lack of antibody protection to a novel strain. Although anti-viral agents have been shown to be effective in children and adults, there are no agents with proven efficacy or safety in infants, and the available formulations limit their use in pre-term infants. This may lead to difficult decisions about whether to use unproven therapy in the context of life-threatening illness. If such agents are used, parents should be informed of the limited evidence and provide informed consent. Careful follow-up may provide useful information for future infants. The most important measures in a neonatal intensive care unit in the event of a pandemic are likely to be preventive ones (Table 1). If an effective vaccine is available, staff and parents should be immunized. Anti-viral prophylaxis of staff, parents and even newborns may have a role, although despite stockpiling, availability of these agents during a pandemic may well be limited. Isolation of affected infants and strict adherence to infectioncontrol precautions are critical to control spread within a unit. Acknowledgments We are grateful to Jacqui Keene for help with image preparation. This study was undertaken at the Neonatal Unit, Mercy Hospital for Women, Melbourne, Australia. Duality of Interests: JPB has performed clinical vaccine trials with vaccine manufacturers including Sanofi-Pasteur, Wyeth, GlaxoSmithKline, CSL and MedImmune while working for the Universities of Melbourne and Oxford and the Murdoch Childrens Research Institute. He has been supported by vaccine companies for travel to scientific meetings. All honoraria for consultancies are paid to an educational fund held by the Murdoch Childrens Research Institute. The remaining authors do not have any duality of interests. Funding support: No financial support was used in the preparation of this paper. References 1 Monto AS. The threat of an avian influenza pandemic. N Engl J Med 2005; 352(4): 323–325. 2 The writing committee of the World Health Organization (WHO). Consultation on Human Influenza A/H5. Avian Influenza A (H5N1) Infection in Humans. N Engl J Med 2005; 353: 1374–1385. 3 Woods CR, Abramson JS. The next influenza pandemic: will we be ready to care for our children? J Pediatr 2005; 147(2): 147–155. 4 Teo SS, Nguyen-Van-Tam JS, Booy R. Influenza burden of illness, diagnosis, treatment, and prevention: what is the evidence in children and where are the gaps? Arch Dis Child 2005; 90(5): 532–536. 5 Gubareva LV, Kaiser L, Hayden FG. Influenza virus neuraminidase inhibitors. Lancet 2000; 355(9206): 827–835. 6 Reid A. The effects of the 1918–1919 influenza pandemic on infant and child health in Derbyshire. Med Hist 2005; 49(1): 29–54. Journal of Perinatology Conclusions It is likely that some newborn infants will be affected in the setting of a new pandemic of influenza. Mortality rates in previous Influenza in the NICU DJ Wilkinson et al 776 7 Bauer CR, Elie K, Spence L, Stern L. Hong Kong influenza in a neonatal unit. JAMA 1973; 223(11): 1233–1235. 8 Sagrera X, Ginovart G, Raspall F, Rabella N, Sala P, Sierra M et al. Outbreaks of influenza A virus infection in neonatal intensive care units. Pediatr Infect Dis J 2002; 21(3): 196–200. 9 Cunney RJ, Bialachowski A, Thornley D, Smaill FM, Pennie RA. An outbreak of influenza A in a neonatal intensive care unit. Infect Control Hosp Epidemiol 2000; 21(7): 449–454. 10 Munoz FM, Campbell JR, Atmar RL, Garcia-Prats J, Baxter BD, Johnson LE et al. Influenza A virus outbreak in a neonatal intensive care unit. Pediatr Infect Dis J 1999; 18(9): 811–815. 11 Munoz FM. Influenza virus infection in infancy and early childhood. Paediatr Respir Rev 2003; 4(2): 99–104. 12 Glezen WP, Taber LH, Frank AL, Gruber WC, Piedra PA. Influenza virus infections in infants. Pediatr Infect Dis J 1997; 16(11): 1065–1068. 13 Kao HT, Huang YC, Lin TY. Influenza A virus infection in infants. J Microbiol Immunol Infect 2000; 33(2): 105–108. 14 Dagan R, Hall CB. Influenza A virus infection imitating bacterial sepsis in early infancy. Pediatr Infect Dis 1984; 3(3): 218–221. 15 Puck JM, Glezen WP, Frank AL, Six HR. Protection of infants from infection with influenza A virus by transplacentally acquired antibody. J Infect Dis 1980; 142(6): 844–849. 16 Bhat N, Wright JG, Broder KR, Murray EL, Greenberg ME, Glover MJ et al. Influenza-associated deaths among children in the United States, 2003–2004. N Engl J Med 2005; 353(24): 2559–2567. 17 Joshi VV, Escobar MR, Stewart L, Bates RD. Fatal influenza A2 viral pneumonia in a newborn infant. Am J Dis Child 1973; 126(6): 839–840. 18 van den Dungen FA, van Furth AM, Fetter WP, Zaaijer HL, van Elburg RM. Fatal case of influenza B virus pneumonia in a preterm neonate. Pediatr Infect Dis J 2001; 20(1): 82–84. 19 Uyeki TM. Influenza diagnosis and treatment in children: a review of studies on clinically useful tests and antiviral treatment for influenza. Pediatr Infect Dis J 2003; 22(2): 164–177. 20 Harper SA, Fukuda K, Uyeki TM, Cox NJ, Bridges CB. Prevention and control of influenza. Recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2005; 54(RR-8): 1–40. 21 Ventre K, Randolph A. Ribavirin for respiratory syncytial virus infection of the lower respiratory tract in infants and young children. Cochrane Database Syst Rev 2004; (4): CD000181. 22 Hayden FG, Sable CA, Connor JD, Lane J. Intravenous ribavirin by constant infusion for serious influenza and parainfluenzavirus infection. Antivir Ther 1996; 1(1): 51–56. 23 Ray CG, Icenogle TB, Minnich LL, Copeland JG, Grogan TM. The use of intravenous ribavirin to treat influenza virus-associated acute myocarditis. J Infect Dis 1989; 159(5): 829–836. 24 Gladu JM, Ecobichon DJ. Evaluation of exposure of health care personnel to ribavirin. J Toxicol Environ Health 1989; 28(1): 1–12. 25 Meissner HC. Antiviral drugs for prophylaxis and treatment of influenza. Pediatr Infect Dis J 2001; 20(12): 1165–1167. 26 Matheson NJ, Symmonds-Abrahams M, Sheikh A, Shepperd S, Harnden A. Neuraminidase inhibitors for preventing and treating influenza in children. Cochrane Database Syst Rev 2003; (3): CD002744. Cooper NJ, Sutton AJ, Abrams KR, Wailoo A, Turner D, Nicholson KG. Effectiveness of neuraminidase inhibitors in treatment and prevention of influenza A and B: systematic review and meta-analyses of randomised controlled trials. BMJ 2003; 326(7401): 1235. Wooltorton E. Oseltamivir (Tamiflu) unsafe in infants under 1 year old. CMAJ 2004; 170(3): 336. Okamoto S, Kamiya I, Kishida K, Shimakawa T, Fukui T, Morimoto T. Experience with oseltamivir for infants younger than 1 year old in Japan. Pediatr Infect Dis J 2005; 24(6): 575–576. Tamura D, Miura T, Kikuchi Y. Oseltamivir phosphate in infants under 1 year of age with influenza infection. Pediatr Int 2005; 47(4): 484. Recommendations for influenza immunization of children. Pediatrics 2004; 113(5): 1441–1447. Groothuis JR, Levin MJ, Rabalais GP, Meiklejohn G, Lauer BA. Immunization of high-risk infants younger than 18 months of age with split-product influenza vaccine. Pediatrics 1991; 87(6): 823–828. Gall SA. Maternal immunization. Obstet Gynecol Clin N Am 2003; 30(4): 623–636. Salgado CD, Farr BM, Hall KK, Hayden FG. Influenza in the acute hospital setting. Lancet Infect Dis 2002; 2(3): 145–155. Bridges CB, Kuehnert MJ, Hall CB. Transmission of influenza: implications for control in health care settings. Clin Infect Dis 2003; 37(8): 1094–1101. Thorburn K, Kerr S, Taylor N, van Saene HK. RSV outbreak in a paediatric intensive care unit. J Hosp Infect 2004; 57(3): 194–201. Maltezou HC, Drancourt M. Nosocomial influenza in children. J Hosp Infect 2003; 55(2): 83–91. Kuypers J, Wright N, Ferrenberg J, Huang ML, Cent A, Corey L et al. Comparison of real-time PCR assays with fluorescent-antibody assays for diagnosis of respiratory virus infections in children. J Clin Microbiol 2006; 44(7): 2382–2388. Adal KA, Flowers RH, Anglim AM, Hayden FG, Titus MG, Coyner BJ et al. Prevention of nosocomial influenza. Infect Control Hosp Epidemiol 1996; 17(10): 641–648. Ng PC, So KW, Leung TF, Cheng FW, Lyon DJ, Wong W et al. Infection control for SARS in a tertiary neonatal centre. Arch Dis Child Fetal Neonatal Ed 2003; 88(5): F405–F409. Shah S, Caprio M. Optimizing long-term care by administration of influenza vaccine to parents of NICU patients. J Perinatol 2004; 24(5): 273–274. Statement on influenza vaccination for the 2006–2007 season. An Advisory Committee Statement (ACS). Can Commun Dis Rep 2006; 32(ACS-7): 1–27. Centers for Disease Control and Prevention. Interim Recommendations for Infection Control in Health-Care Facilities Caring for Patients with Known or Suspected Avian Influenza 2004, available online: www.cdc.gov/flu/ avian/professional/infect-control.htm. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Journal of Perinatology