John Harris 1
|The principles of medical ethics and medical research |
Os princípios da ética médica e da pesquisa médica
|1 Centre for Social Ethics & Policy, University of Manchester. Humanities Building Oxford Road M13 9PL Manchester, England. email@example.com||Abstract In this paper I discuss the application of the principles of medical ethics and of medical research to the case of children and others whose consent to treatment and to research is problematic. Public health depends substantially on the possibility of ongoing research into all conditions which affect the health of the people. Constraints on this research are therefore a public health issue. Moreover and more importantly the possibility of predictive testing and indeed of screening for health-relevant conditions is an important public health tool, and limitations on the use of this tool are of great significance to public health medicine. Having considered the particular problems created by research and predictive testing on children for late-onset conditions I go on to discuss research on those whose consent is problematic more generally. I conclude with radical recommendations for the reform of The Declaration of Helsinki and of the International Ethics Guidelines for Biomedical Research Involving Human Subjects, prepared by the Council for International Organizations of Medical Sciences (CIOMS). |
Key words Bioethics; Research; Public Health
Resumo Nesse artigo, discuto a aplicação dos princípios de ética médica e da pesquisa médica ao caso de crianças e outros cujo consentimento a tratamento e pesquisa é problemático. A saúde pública depende substancialmente da possibilidade da continuidade de pesquisas dirigidas a todas as condições que afetam a saúde da população. Restrições nessas pesquisas são, portanto, uma questão de saúde pública. Além disso, e mais importante, testagens preditivas e, especialmente, screening para condições relevantes (para) à saúde constituem-se em valiosos instrumentos de saúde pública. Limitações no uso desses instrumentos são da maior significância à medicina ligada à saúde pública. Levando em consideração os problemas particulares criados pela pesquisa e testagem preditiva para crianças no que se refere a condições que se manifestam tardiamente, prossigo discutindo investigações naqueles cujo consentimento é problemático em termos mais gerais. Concluo com recomendações radicais para a reforma da Declaração de Helsinki e das Orientações Éticas Internacionais para Pesquisa Biomédica Envolvendo Sujeitos Humanos, preparadas pelo Council for International Organizations of Medical Sciences (CIOMS).
One of the problems that has preoccupied me recently is that of formulating ethical principles appropriate to testing and screening for late-onset disease. As a member of the recently established United Kingdom Advisory Committee on Genetic Testing (ACGT) and in particular of its working party on Genetic Testing for Late-Onset Disorders, I have been particular interested in the question of the appropriateness of testing children, or testing prenatally for late-onset conditions.
There seems to be an orthodoxy against testing children which perhaps needs some further reflection.
In this paper I will attempt two tasks. The first is to say something very brief about the dilemmas of testing and screening as they apply to testing children for late or 'later' onset conditions. I shall then, in the main part of what I have to say, try to address the very general issue of the ethics of research on children and others who either cannot give authentic consent or whose consent to research may be problematic.
Should Children be tested for late-onset conditions?
If we ask whether or not it is legitimate to test children for late-onset conditions, there seem to be powerful arguments on both sides.
1) Many adults refuse testing and this is some evidence as to the likely choices of children when they become competent.
2) Testing children preempts their autonomous decision.
3) There are important insurance and employment consequences of testing which are adverse to the individual tested. These are not of course an argument against testing per se but so long as they remain true they must be taken into account.
4) Children may grow up with a sense of a blighted existence and deprivation relative to their friends and peers.
1) A clear result enables rational planning for a probably shorter life. Deferred gratification may be less appealing as may decisions which presuppose a normal life span.
2) Preventive strategies may be possible which would not be thought of without clear knowledge of the condition (possible preventive mastectomy, for example).
3) Reproductive planning becomes possible, either to have an early family with early prenatal testing for genetic diseases or to decide not to have a family.
A policy proposal
The United Kingdom Government Advisory Committee on Genetic Testing has produced a consultation document which addresses the policy questions raised by such dilemmas. I must stress that these are suggestions for consultation purposes and are by no means settled policy proposals. However, they are, I believe, of considerable interest as one possible way forward. Below I reproduce pages 22 and 23 of this document. I must emphasize that this document is for consultation purposes only and has no official standing.
"Presymptomatic Genetic Testing of young children for late-onset disorders:
a) 'Over the counter' genetic testing of children is not appropriate.
This was the conclusion of the committee's previous report and we continue to endorse this approach. The issues involved may be complex and they require professional involvement.
b) Any request from a person or persons with parental responsibility to professionals, for presymptomatic genetic testing of a young child unable to give consent, should be fully discussed in context of the particular family situation, the nature of the disorder, the possible medical or other benefits, and the implications of testing for the child and family.
Requests for such testing are frequently part of a more general need for information and require full and sensitive discussion with a professional who can provide this; they should not be simply accepted or declined on ground of general policy. If a child is to be tested then the consent to be given for the test by the person or persons with parental responsibility should be on the basis that it is in the child's best interest.
c) "Presymptomatic testing of young children for disorders not currently influenced by therapy, and where onset is normally in adult life, is not recommended, though there may be special situations that are exceptions to this".
Since young children are usually unable to give consent, it is preferable for important decisions relating to their adult life to be left to a time when they can give consent themselves, especially since it is known that a considerable proportion of adults may decline testing in such a situation. This has also been the conclusion of other bodies, such as the Clinical Genetics working group report, the Genetics Interest Group and the US Task force on Genetic Testing. The situation for a particular disorder will require reassessment in the light of advances in therapy.
d) Where genetic tests are being used diagnostically in the context of a child with a family history of a late-onset disorder showing clinical features, it is essential that these clinical features are relevant to the nature of the family disorder.
Genetic testing is increasingly part of diagnosis in pediatric practice. Where a late-onset disorder may occasionally occur in childhood, there is the risk that in a child with symptoms, these might be unrelated to the disorder in the family; this would give a real risk of inadvertent detection of the gene being present, with serious implications for later life. Many clinicians do not at present recognize this potentially dangerous situation.
Genetic testing for late-onset disorders in adolescents:
Requests from adolescents themselves regarding presymptomatic genetic testing for late-onset disorders deserve full and sensitive discussion, taking into account the individual and their family. If the adolescent is competent to give consent but there are not clear benefits of testing, we would recommend that ideally testing be deferred until the age of majority is reached. However, the competent adolescent is entitled to make a personal decision on this matter after a full discussion and exploration of the issues.
Adolescents will vary greatly as to their emotional and mental maturity in relation to complex issues for later life. Current experience is that such requests are few but that they need full discussion with an experienced person. Pregnancy could be a special situation where testing needs to be considered.
In some late-onset disorders, e.g. familial adenomatous polyposis, for which treatment might be possible or for which screening might detect the condition at an early enough stage for treatment to be effective, there is good reason to consider early testing of adolescents. In conditions such as familial cardiomyopathy, where there may be a family history of sudden death in young family members, there may be benefit in carrying out predictive testing in children and adolescents. The situation for each disorder requires careful consideration (Advisory Commitee on Genetic Testing, 1997)
We have reviewed some of the particular problems raised by testing children for later onset diseases and examined one possible set of policy recommendations to cope with these difficulties. For the remainder of this paper I will now turn more generally to the ethics of treatment and research on those whose consent is problematic and I will, I hope, challenge what I have come to regard as a pernicious and unsustainable orthodoxy (Harris, 1997).
Problematic consent to treatment
We will begin by considering the case of medical treatment of those whose consent is problematic and then move on to the issue of research. There is a continuum between these dimensions of consent which will, I hope, become obvious.
Because there are so many cases in health care practice which necessitate touching patients in circumstances where their consent cannot be obtained and where knowledge of their wishes is absent, the law has contrived various fictional consents to protect well-intentioned practitioners from the guilt of unlawful conduct. The moral necessity of obtaining a valid consent where this can be obtained does not require further discussion. To violate the bodily integrity of persons who reject such violation is a form of tyranny and should be accepted and treated as such. We must however look more closely at those cases where consent or its refusal is problematic, and at the fictionalized consents that are often manufactured in these circumstances.
There are a number of instances in health care where the patient's consent is appealed to and used, where actual consent is unobtainable. These are circumstances in which the patient is either unconscious or unable to process the information required to give a valid consent, or is temporarily or permanently lacking the relevant capacity to consent. In such cases terms like 'proxy consent', 'substituted judgment', 'presumed consent' or even 'retrospective consent' are used to justify treating a patient.
Provision for these sorts of 'consents' is endorsed by most of the leading international protocols on research. For example, the Declaration of Helsinki provides that "Where physical or mental incapacity makes it impossible to obtain informed consent, or when the subject is a minor, permission from the responsible relative replaces that of the subject..." (World Medical Association, 1996). The other leading source of guidelines in this field are the International Ethics Guidelines for Biomedical Research Involving Human Subjects, prepared by the Council for International Organizations of Medical Sciences (CIOMS) in collaboration with the World Health Organization (WHO). Their Guideline 1 states: "...in the case of an individual who is not capable of giving informed consent, the proxy consent of a properly authorized representative" must be obtained" (CIOMS/WHO, 1993).
However, not only are these all fictions, but they totally fail to be justifications for treating the patient in particular ways.
Here of course we shall be advancing a thesis that runs counter to much contemporary thinking on consent which seems at home with attributing consent to individuals who are totally unaware that they are supposed to be consenting or were unaware at the time the consent is operative (as in the case of retrospective consent).
The reason why it is right to do what presumed consent or substituted judgment seems to suggest in these cases is simply because treating the patient in the proposed ways is in his best interests and to fail to treat him would be deliberately to harm him. It is the principle that we should do no harm that justifies treating the patient in particular ways. The justification for treatment is not that the patient consented, nor that he would have, nor that it is safe to presume that he would have, nor that he will when he regains consciousness or competence, but simply that it is the right thing to do, and it is right precisely because it is in his best interests. That it is the 'best interests' test that is operative is shown by the fact that we do not presume consent to things that are not in the patient's best interests, even where it is clear that he would have consented. We do not infuse known heavy smokers with cigarette smoke while they are unconscious even where it is reasonable to suppose they would have consented, and patients are often denied access to alcoholic beverages or cigarettes, even when they specifically request them.
It is widely held that not only should we not harm people who do not want to be harmed, we also should not harm even those who do want to be harmed, and that this is sufficient reason not to withhold treatment the absence of which would harm. This raises the question of the right to harm oneself, which I have no space to discuss further here.
Not only do we not need the concept of implied or assumed or proxy consent, because it literally does not work; we do not need it because it misleads us as to the character and meaning of our actions. The nineteenth century English philosopher Jeremy Benthan was rightly scathing of fictional consents when he remarked: "In English law, fiction is a syphilis, which runs in every vein, and carries into every part of the system the principle of rottenness... Fiction of use to justice? Exactly as swindling is to trade... It affords presumptive and conclusive evidence of moral turpitude in those by whom it was invented and first employed." (Benthan, apud Steiner, 1994: 258).
So where, in medical contexts, we act in the best interests of patients who cannot consent, we do so, I suggest, because we rightly believe we should not harm those in our care and not because some irrelevant person or the law has constructed a consent. This does not of course help with the vexed problem of who is and who is not competent to consent, but it does explain the justification for intervening in the lives of those we are satisfied are not able to give the consents that would otherwise be required.
If the treatment of those whose consent is impaired is problematic, and decisions to discharge them from care equally so, what of research?
Research on children or on cognitively impaired subjects
There is of course nothing special about research on cognitively impaired elderly subjects. Cognitive impairment is only significant in this context insofar as it also involves impairment of autonomy and the elderly are not importantly different from other age groups from the point of view of consent (Hirsch & Harris, 1988).
First we should note the difficulty of distinguishing between research and therapy. Many definitions have been provided and most rely at some point on making the distinction by reference to the intentions of those carrying out the work (Ciba Foundation Study Group, 1980; British Medical Association, 1993). However, any distinctions based on intent are at best suspect and at worst susceptible to manipulation. Firstly the intentions of people are usually mixed and confused. Medical staff may, and doubtless often do, simultaneously intend to do many things when they treat their patients. They may intend to care for patients, offer relief of symptoms, identify and eradicate causes of disease, hone their medical skills, further their careers, satisfy their curiosity, generate research data, carry out instructions, and so on. Identifying primary intent is hardly more helpful because it either relies on a degree of self-awareness that few can achieve, or, it simply stipulates one of many possible or actual intentions as paramount for the purposes of bringing the activity into an appropriate or permitted category.
This does not mean that there is no distinction to be drawn between research and therapy, but rather that the distinction cannot be drawn sufficiently clearly to sustain claims that therapy is justified whereas research is not. This is perhaps less worrying than might at first appear. Firstly the work supposedly done by the distinction between research and therapy might as easily be accomplished by concentrating on the degree of advantage to the patient. We could say that interventions, whether motivated by research or therapeutic imperatives, are permissible if either the patient accepts and consents to the interventions, or, where consent is unobtainable, that the interventions are the best available treatment for the patient, whether or not they also constitute research.
However, it is worth pursuing the distinction between therapeutic and non-therapeutic research a bit further, because, insofar as it is coherent, it relies on a suspiciously narrow view about the definition of 'therapeutic' and also about what is, in fact, in every patient's best interests.
The following two claims are often made in this context:
1) The patient's interests are paramount.
2) That research on patients who cannot consent for themselves must be either:
i) Therapeutic or
ii) In the patient's own interests or of potential benefit for the patient herself or
iii) If not directly for the benefit of the particular patient in question, at least for the benefit of the category of patients to which the subject belongs, so that for example, if the patient is suffering from Li-Fraumeni disease, then research will only be justified if the research will be of benefit if not to the patient himself, at least to other Li-Fraumeni victims.
Neither claim seems sustainable and since the claims are often repeated, it is worth taking a moment to see why this is so.
The patient's interests cannot be paramount for the simple and sufficient reason that being or becoming a patient is not the sort of thing that could conceivably increase either your rights or your moral claims. All people are morally important and with respect to one another each has a claim to equal consideration. No one has a claim to overriding consideration. To say that the patient's interests are paramount, if it means anything, must be seen as a way of reasserting that health professionals are concerned primarily for the patients in their care and may have special contractual duties to them. However as a general remark about the obligations of the health care system or of society it is not sustainable.
The second claim is equally problematic. To assert that research must be therapeutic or in the patient's own interests is not plausible when interpreted as confining research to work that will benefit particular patients directly, in the sense that they will themselves, so to speak, 'feel the benefit', or that it will serve their continuing interests.
Let us look at the idea of what is or is not in someone's interests first. We must be wary of being too conservative about what does or does not benefit someone or of defining someone's interests too narrowly. We all benefit from living in a society in which medical research is carried out and which utilizes the benefits of past research. It is both of benefit to patients and it is in their interests to be patients in a society which pursues and actively accepts the benefits of research and where research and its fruits are given a high priority. We all also benefit from the knowledge that research is ongoing, into diseases or conditions from which we do not currently suffer but to which we may succumb. It makes us feel more secure and gives us hope for the future, for ourselves, our descendants, and others for whom we care. If this is right, then I have a strong general interest that there be research, and in all well-founded research; not excluding but not exclusively, research on me and on my condition. All such research is also of clear benefit to me. A narrow interpretation of the requirement that research be of benefit to the subject of the research is, I believe, perverse. If the claim that research on subjects who cannot consent for themselves is justified only if narrowly targeted on conditions from which those subjects currently suffer, and which will likely bear fruit in time to be of direct benefit to them is to be sustained, then it requires separate argument and support. There is no reason however to take this narrow interpretation as the standard one.
The third suggestion, that research which is not directly beneficial to the patient be confined to research that will benefit the category of patients to which the subject belongs is also untenable. What arguments sustain the idea that the most appropriate reference group is that of fellow sufferers from a particular disease, Li-Fraumeni for example? Surely any moral obligation I have to accept risk or harm for the benefit of others is not plausibly confined to those others who are narrowly like me. This is surely close to claiming that research should be confined to others who are 'black like me' or 'English like me' or 'God-fearing like me'? The most appropriate category is surely 'human like me'.
Where I benefit from research but refuse to participate in it I am clearly acting unfairly in some sense. I am free-riding on the back of the contribution of others. While we may conclude that people are ultimately entitled to act unfairly in this way if they choose, that they should not normally be compelled to contribute or participate, there is no reason to presume that those who cannot consent would have wished to be free riders. Indeed, as we have argued, there is no justification for any presumptions about their willingness to consent or otherwise, in the absence of clear indications about their preferences. What is clear, I believe, is that there is no basis for any presumption that those who cannot consent should be excluded from research, whether that research is targeted on their condition or not. Nor, of course, is there any basis for the presumption that those who cannot consent should be, so to say, 'professional research subjects'. They should neither be automatically included in research, perhaps because they happen to be readily available in institutions or under continuing care, nor should they be automatically excluded.
We have been talking very generally of course and assuming a favorable risk/benefit balance for the research subjects. Namely that the risks and pain, discomfort, or inconvenience of the research are minimal, and the projected benefits clear. There is clearly a trade-off between risks and benefits and a fairly steep upward curve where we demand that the benefits be clear and urgent before we will accept significant risks or pain, inconvenience, and so on for ourselves. We should surely apply the same standards to those who cannot consent. I cannot say anything useful or original about this balance. Clearly I will only accept significant risk of death for myself if without accepting such risk there is an almost certainly and substantially worse outcome for myself or those I care about. However even here the risks are not plausibly undertaken only to benefit oneself. Live kidney donors for example clearly accept significant risks for others and we usually applaud their decision so to do.
There clearly is an obligation (sometimes) to make sacrifices for the community or an entitlement of the community to goes so far as to deny autonomy and violate bodily integrity in the public interest and this obligation is recognized in a number of ways. The following areas in which this is already recognized and accepted to some extent will serve as a reminder: control of dangerous drugs, control of road traffic, vaccination, screening tests, blood donation, quarantine for communicable disease, compulsory military service, detention under mental health acts, restriction on sexual activities and professional activities of HIV positive people. All these involve some denial of autonomy, some imposition of public standards even where competent individuals do not consent to that imposition. However there are clearly some exceptional cases where overriding moral considerations take precedence over autonomy.
An example which seems to me to illustrate, and to an extent explain, our attitude to the imposition of risk in the public interest involves the following story. Imagine an ocean liner on a cruise. The captain receives a radio message that there is another ship in distress some miles to the north. There are two hundred people aboard this other ship and his liner has one thousand aboard. His is the only ship that can effect the rescue before the stricken ship will founder. He knows that if he diverts into the storm he will impose some risk on his passengers and crew. There will be a small but significant risk of death for all. The storm is a bad one but the modern liner should be able to cope. There is a greater, but still small, risk of death for a few of his passengers and crew in the rough and tumble of the rescue. Finally, because the storm is severe, he will almost certainly be subjecting his many elderly passengers to risk of minor injuries in the rough seas and certainly to discomfort, fear, and inconvenience. Significantly, we don't have to ask what he should do. The captain knows he must attempt the rescue and subject his passengers and crew to the attendant risks and few would disagree. He also knows that he can and must do so without asking for the consent of his passengers and crew, for they would be wrong to withhold their consent and the captain would be wrong to act on it.
The list of areas in which, I believe, similar decisions are made in our society shows that the principles involved are not unfamiliar or indeed unacceptable. Why then should we assume, when considering the ethics of research involving say, children or cognitively impaired elderly subjects, that different principles should apply?
Firstly, let's just remind ourselves what these principles are, and then go on to see if they are constrained in any way by the situation of patients whose consent is problematic. The principles involved are all dimensions of the principle of equality from which our first principle of research ethics was derived. That principle is, it will be remembered, that each person is entitled to the same concern, respect, and protection of society as is accorded to any other person in the community. This principle reminds us that the passengers on the stricken liner have as good a claim to our protection as any other persons, (although they may not be fellow citizens) and that while we are not obliged to afford that protection at all costs, we are obliged to act morally when the costs of so doing are reasonable given the importance of what is at stake. I hope it is obvious that if acting morally were only obligatory when doing so was cost-free to the agent, morality would not exist. It is not plausible to believe that the costs of acting morally fall only on those competent to consent. So long as we ensure that such costs do not fall more heavily on those not competent to consent than on others I can see no sound argument for exempting them from the demands of morality. They may not be accountable in law, if they do wrong, but there is no reason to ensure that they do wrong, by exempting them from their moral obligations.
The fact that the moral obligations in question are not mandatory, in that competent individuals may not, usually, be coerced into fulfilling them, does not seem to be a reason for exempting those not competent to consent. We do not allow children, (or we should not) to do wrong because some adults may freely do so. However we constrain children not only because we are, in part, morally responsible for their actions, but also because this is part of an educational process. The right parallel with adults who lack competence might be to include them in research when, although not competent to consent or refuse, they make no overt objection to inclusion or complaint about it. Where they do object or complain, we should perhaps respect that, albeit incompetent, rejection of this particular moral obligation.
Doubtless this defense of the idea that there can be an obligation to participate in research in certain circumstances, whether those who participate are capable of consenting or not, will strike some as controversial. The prohibition of research on those incapable of consent except where the research is in their own therapeutic interest is a principle founded on the highest of motives, that of the need to protect the vulnerable. However, the same motive animates the position defended here. It is not only the incompetent who are vulnerable in the requisite sense. We are all vulnerable unless research is pursued. The issue is one of balance. If research can be pursued without recourse to those whose consent is dubious, then so much the better. This should be our first choice. However, if such a prohibition jeopardizes our capacity to pursue well-founded research then perhaps we should remember that free-riding is not an attractive principle; nor is it a moral principle. We should not, as I have indicated, assume that those incompetent to consent would wish to be free riders, nor that they be excluded from discharging an obligation of good citizenship which we all share.
ADVISORY COMMITEE ON GENETIC TESTING, 1997. Report on Genetic Testing for Late Onset Disorders. London: Health Department of United Kingdom.
BRITISH MEDICAL ASSOCIATION, 1993. Medical Ethics Today: Its Practice and Philosophy. London: British Medical Association.
CIBA FOUNDATION STUDY GROUP, 1980. Medical research: civil liability and compensation for personal injury a discussion paper. British Medical Journal, 280:1172-1175.
CIOMS (Council for International Organizations of Medical Sciences)/WHO (World Health Organization), 1993. International Ethics Guidelines for Biomedical Research Involving Human Subjects. Geneva: CIOMS/WHO.
HARRIS, J., 1997. The ethics of clinical research with cognitively impaired subjects. Italian Journal of Neurological Sciences, 18:9-15.
HIRSCH, S. & HARRIS, J., 1988. Consent and the Incompetent Patient: The Royal College of Psychiatrists.
STEINER, H., 1994. A Theory of Rights. Oxford: Blackwell.
WORLD MEDICAL ASSOCIATION, 1996. World Medical Association Declaration of Helsinki. 48th General Assembly of World Medical Association. Helsinki: World Medical Association.