Abstract and Keywords
Chapter 4 focuses on evaluating arguments which grant the concerns raised by nonbeneficial pediatric research, and then try to meet Ramsey's challenge. The chapter first considers Utilitarian justifications and why they are rightly rejected. It then considers the argument that nonbeneficial pediatric research can be justified when it is consistent with the parents’ rights to make decisions for their children. Next is the argument that participation in nonbeneficial research can be in children's interests by teaching them to be moral. While this argument does not provide a general justification it does reveal that a response to Ramsey's challenge need not be based on the potential for participating children to realize clinical benefits in particular. This suggestion will be pursued in subsequent chapters. Chapter 4 ends by considering whether it is possible to meet Ramsey's challenge by appeal to a variation on the Rawlsian approach of evaluating what policies are appropriate based on which policies a reasonable person would endorse from behind a veil of ignorance.
The Argument Thus Far
The previous chapter considered arguments that attempt to determine whether nonbeneficial pediatric research is acceptable, in effect, by analyzing the nature of the research itself and the standard guidelines under which it is conducted. We considered arguments that such research is necessarily unethical because it violates children's autonomy, exposing those who cannot consent to risky procedures that are not in their medical interests, and also involves forcing them to contribute to the goals of projects they do not understand, much less embrace. The problem with this approach is that it ignores the possibility that children's inability to give consent does not merely alter their ability to decide the course of their lives; it transforms what constitutes appropriate treatment of them. Although this negative conclusion removes certain arguments against nonbeneficial pediatric research, it does not provide a positive argument regarding whether it can be acceptable to expose children to research risks for the benefit of others.
McCormick's attempt to justify nonbeneficial pediatric research on the grounds that children would consent to it if they could consent fails to take seriously the fact that children are not autonomous. Next, we bracketed for consideration in the present chapter McCormick's claim that children are obligated to participate in nonbeneficial research when the risks are sufficiently low. Finally, we considered the claim that nonbeneficial pediatric (p.78) research does not raise serious ethical concern when the risks are negligible or low. Understood as allowing research that poses no risk of serious harm, this approach fails to address the majority of nonbeneficial pediatric research or, alternatively, implies that the majority of nonbeneficial pediatric research that is being conducted, and has been conducted under existing regulations is unethical.
Proposed justifications of the first type are unsatisfactory because they do not take seriously the fact that children are children. Those in the second set do not take seriously the fact that nonbeneficial pediatric research often poses some risk of serious harm to participating children. This is not to say that such research is necessarily unethical, only that proposed justifications should be consistent with this fact. One way to think about this point is in terms of the function of the existing risk standards. These standards allow nonbeneficial pediatric research when the level of risks does not exceed some threshold, most notably, the level of risks average children ordinarily face in daily life. The present claim is that research which satisfies this standard may well be ethically acceptable. That standard may provide approximately the correct threshold for the level of risks to which children may be exposed in this context. The problem comes when we regard this standard as providing not simply a metric for which risks are acceptable, but also a justification for studies that satisfy the endorsed metric: nonbeneficial pediatric research that meets this standard is acceptable because the risks are no greater than the risks average children ordinarily encounter in daily life.
As we have seen, the risks of daily life tend to be accepted in most cases because they attach to activities that are thought to offer the participating children a compensating potential for personal benefit. Because nonbeneficial pediatric research does not offer participating children a compensating potential for clinical benefit that justification appears lost when we apply the standard to nonbeneficial pediatric research. An alternative justification cites the fact that we tend to ignore the (familiar) risks of daily life when going about our lives. The problem with this justification is that this attitude toward the risks of daily life is either itself unjustified, or it depends on a kind of rough cost/benefit calculation regarding which risks are worth paying attention to.
The fact that it might not be worth parents’ time to pay attention to very low risks to their children does not provide a justification for society to support and clinical investigators to engage in an activity which exposes children to those risks for the benefit of others. We saw this point with respect to the comparison between the risks of nonbeneficial pediatric research and the risks of charitable activities for children. The fact that it is widely regarded as appropriate to expose children to some risks for the benefit of others in the (p.79) context of charitable activities provides reason to think that exposing (nonconsenting) children to risks for the benefit of others is not necessarily unethical, as Ramsey claims. The problem, of course, is that this comparison begs the question of whether in fact it is acceptable to expose children to risks in the context of charitable activities, and why. A further problem arises from the fact that clinical investigators actively expose children to risks, giving them medications, inserting needles into their veins.
The thought that charitable activities for children can be appropriate because it is good for the children to participate in them begins to suggest one way to try to justify exposing children to risks for the benefit of others. This suggestion brings us to the work of Ramsey, Bartholome, and others who argue that it might be possible to justify nonbeneficial pediatric research by appeal to the fact that it can promote the nonmedical interests of pediatric participants. We will pursue this suggestion further in the following chapter. For the present, it is sufficient to note that this approach also may offer a justification for the active role that investigators play in the context of pediatric research. It is one thing to insert a needle in a nonconsenting child's arm for the benefit of others. This is the perspective from which Ramsey and other critics argue that nonbeneficial pediatric research is necessarily unethical. It is a very different thing to insert a needle in a child's arm for the benefit of others, in a context in which the child stands to benefit in some way. To set the stage for that consideration, the present chapter first considers prominent attempts to justify nonbeneficial pediatric research which take seriously the fact that it exposes children to risks, including some risks of serious harm, but do not appeal to the possibility that the participating children themselves stand to benefit.
Utilitarianism as an Epithet
The puzzle offered by (the paradigm instances of) nonbeneficial pediatric research is to justify directly exposing some children to risks for the benefit of unrelated others, and thereby having them contribute to ends they do not understand and do not positively endorse. Utilitarianism provides a possible solution to this puzzle by largely ignoring the distribution of benefits and burdens that arise from a given policy or practice. A Utilitarian justification would focus on whether a given nonbeneficial intervention or study is expected to produce on balance positive benefits to the extent that the net benefits are greater than what could be realized by pursuing available alternatives. (p.80) On this argument, it is acceptable to expose children to research risks when the study offers the potential for benefit that outweighs the risks and burdens the participants face (and the burdens and costs of conducting the research), independent of the extent to which the participants themselves will benefit.
Although Utilitarianism offers a straightforward argument, this view, at least in the simplistic form assayed here, is rarely endorsed by supporters of nonbeneficial pediatric research. It is more often cited by opponents who oppose current regulations on the grounds that they are thought to justify nonbeneficial pediatric research on purely Utilitarian grounds. These arguments are regarded as criticisms because it is assumed that purely Utilitarian justifications are inadequate. Nonetheless, evaluation of the possibility of developing a straightforward Utilitarian argument will be useful because it reveals the extent to which current regulations for nonbeneficial pediatric research are not consistent with a Utilitarian calculus and also provides important insights into what a successful justification might look like.
One virtue of the Utilitarian approach is that it connects any possible justification of nonbeneficial pediatric research and its value in what seems like the right way. Intuitively, one assumes that any justification for nonbeneficial pediatric research will cite in some way its social value. It is after all the potential for benefit that is the reason for doing the research in the first place. And, yet, since the benefits do not accrue to those who face the risks, it is hard to see how to connect the justification for exposing subjects to risks to the benefits it offers to others. Indeed, while the value seems central to the justification of nonbeneficial pediatric research, it similarly raises concerns about exploitation or unfair use of children for the benefit of others. The accounts we have considered thus far, and several we will consider in the present chapter, address this dilemma by providing a justification for doing the research in the first place, and then a distinct justification for enrolling children in the research. The research is justified by its potential to gather information of significant social value. Enrolling specific children is justified on independent grounds; for example, on the grounds that the children would consent to it if they could.
This is not to say that these justifications ignore the social value of the research. The claim that individuals would consent is based, at least in part, on the assumption that the research has sufficient value. It is assumed that reasonable individuals would consent because the research is valuable (and presumably not excessively risky or burdensome). Although such justifications cite the value of the research, the fact that the studies have value is not, in (p.81) theory, necessary for them. One could argue that individuals would consent to participate on other grounds, perhaps to satisfy a desire to confront risks or to assist one's colleagues in an endeavor that is important to them. In this sense, the value of the research is not necessary for the proposed ethical justification. It is, instead, just one reason to draw a conclusion that one might arrive at for other reasons.
I will argue that none of the existing justifications for enrolling children succeeds in the sense of providing a justification for the range of research that seems intuitively acceptable. This conclusion will provide impetus to consider attempts that justify the enrollment of children by citing more directly the social value of the research. This approach will have a second virtue. By focusing the justification for enrolling children on the social value of the research, it will emphasize the importance of ensuring the research is valuable. Accounts that attempt to justify nonbeneficial pediatric research by citing the low risks involved tend to focus almost exclusively on the risk level, losing sight of the importance of ensuring the research is valuable.
Similarly, nonbeneficial pediatric research is thought to provide a paradigm case of exploitation because it involves the use of some (without their consent) for the benefit of others. If the research offered no social value, enrolling children might be wrong on other grounds, but it would not involve possible exploitation of the enrolled children (since it would not benefit others). The social value of the research thus seems to represent at best an embarrassment, at worst an ethical affront and the source of the primary ethical concern posed by the research.
Many of those who are involved in nonbeneficial pediatric research tend to ignore the fact that the interventions are designed to benefit others when considering the acceptability of enrolling specific children. They do not emphasize to themselves, the children, or their parents the fact that the interventions are designed to benefit others because doing so raises the concern that the participating children are thereby being exploited. They instead focus on the requirement for low risks and try to convince themselves that this precludes the possibility of serious harm. Or, they focus on the possibility, sometimes extremely remote, that the child might derive some kind of clinical benefit, perhaps as the result of an incidental finding of the type that we considered in the first chapter, or in the long term as the result of new medications identified by the series of studies of which the present study is a part. An approach that connects the justification for enrolling specific children to the social value of the research might offer the pragmatic benefit of reassuring those involved in nonbeneficial pediatric research and possibly making more transparent the nature of the research itself. That is, it might clarify the ways in which the (p.82) social value (i.e., the potential to benefit others), far from being an affront and the possible basis for a charge of exploitation, provides the basis for a justification for enrolling some children in some nonbeneficial research.
What we might think of as a simple-minded Utilitarian justification connects the social value of nonbeneficial pediatric research and the justification for enrolling children by dropping the requirement that those who face the risks must stand to benefit, and instead evaluating the overall balance of risks and benefits independent of the actors to whom they attach. A Utilitarian account can explain why it is acceptable to expose children to risks, provided the expected risk–benefit profile of enrolling them is more favorable overall than other available options. Indeed, when these condition are satisfied, one might argue that doing otherwise than enrolling children in the research would be unethical since the other options, by hypothesis, are expected ex ante to yield lower benefits overall. It follows that it would be acceptable for parents to enroll their children, and it would be acceptable for investigators and society at large to involve themselves in such efforts. Doing so is the ethical thing to do.
This same aspect of Utilitarianism famously explains why almost no one regards it as an acceptable ethical justification. By splitting apart the burdened from the benefited, it allows for the possibility of increasing, essentially without limit, the risks and burdens to which some are exposed provided doing so brings with it a compensating potential for benefit somewhere in the system. The implication for nonbeneficial pediatric research would be that it would be acceptable to expose children to essentially any level of risks provided the study offers sufficient potential for social benefit.
The costs of exposing children to risks in a given nonbeneficial research study is, at least on this first pass, a relatively fixed sum: the risks or the harms and burdens caused by the study to the children who are enrolled in it. Of course, most interventions require a series of studies, not a single one. Thus, we need not assume that just one cohort of children will face the risks of research participation. Nonetheless, the harm is fixed to a relatively small set of children. The potential benefit, in contrast, can be thought of as an ever-increasing sum in real time. Improved medical treatments benefit all children who are helped by them. If the disease in question affects many children, the number benefited could be enormous. And, if the disease remains active for generation after generation, each treated cohort would add to the sum of total benefits realized. Or, if the research produces a cure, the beneficiaries would be all future children who thereby avoid the disease. The rotavirus vaccine studies considered in the first chapter have the potential (again, assuming funding is made available to provide it to children in lower-income (p.83) countries) to benefit hundreds of thousands of children each year, possibly for decades to come.
This potential benefit underscores another common concern with straightforward Utilitarian approaches, namely: a relatively minor benefit to the members of a sufficiently large group can justify very serious harms to a small number. At some point, we could justify serious harms, even deaths, to children in a study of a new treatment for acne, since the new treatment could benefit so many children. To take a frightening, but potentially actual example, testing of HIV vaccines takes many years, not to mention the fact that the studies cost a great deal. Essentially, investigators must give the putative vaccine to one group and a placebo vaccine to another group and then wait many years to see whether there is a significant difference in the rate at which individuals in the two groups acquire HIV. Because the risk of acquiring HIV for a given individual tends to be low, these studies must enroll very large numbers of individuals in both groups, typically in the thousands, and must wait a long time to get a result, up to 10 years. This approach to testing vaccines is extremely inefficient. In addition, while we wait 10, 20, perhaps 50 years for an effective vaccine to be identified, millions more will become HIV infected.
To address this concern, HIV vaccine studies often are conducted in groups at higher risk for acquiring HIV, such as intravenous drug users and prostitutes. Although possibly decreasing the time to a significant finding, this approach raises a host of additional ethical concerns linked to the focus on vulnerable populations. From the perspective of overall benefit, independent of who benefits and who is harmed, a more efficient approach would be to do injection studies in which you give a few individuals the vaccine and then inject them with HIV virus. Clearly problematic studies of this type could be justified on a strictly Utilitarian approach. The fact that existing regulations would not allow these studies in children, or even allow an injection study of a vaccine for a disease that was significantly less risky, say hepatitis A, reveals that these regulations are not, as critics have argued, based on a Utilitarian justification alone.1
There are various ways in which one might try to develop a more nuanced and plausible Utilitarian justification for nonbeneficial pediatric research. One might argue on rule Utilitarian grounds that we need to consider not only the consequences of a given study, but also the consequences of adopting a policy which permits that type of study. It might turn out that a policy of allowing what we regard as excessively risky nonbeneficial pediatric research studies would have worse consequences in the long run. This possibility could be made even more plausible by citing our intuitions that such studies are wrong. (p.84) The fact that this is a fairly deeply ingrained intuition—assuming that it is—would suggest that the implications of adopting a policy that contradicts the intuition need be considered as well, and might tip the balance of risks and benefits against a policy of allowing such studies.
One likely could come up with a Utilitarian story that roughly matched our intuitions. The final account might not match our existing intuitions exactly. But, that result in itself is not a knock-down objection. Our intuitions in this regard are relatively rough and ready, and reasonable people can disagree about where to draw the line on acceptable risks. However, as a justification, the present approach would remain problematic because it explains why some nonbeneficial pediatric research is acceptable based on the fact that we think it is acceptable, and explains why riskier research is unacceptable based on the fact that we think it is unacceptable. But, the intuitions regarding where to draw the risk threshold for nonbeneficial pediatric research are not supposed to be the explanation for why the line should be drawn at that point. Rather, the explanation is supposed to provide some story that is independent of and makes sense of our intuitions. Moreover, if this account were correct, it would provide what seems a very unstable defense of present regulations.
The claim here is that extremely risky pediatric research studies should be prohibited because a policy that allows them would be traumatic, an offense to our feelings on the matter. This argument provides a defense of current regulations only to the extent that we continue to find extremely risky studies troubling. As such, this line of reasoning provides a reason to change our views on these studies, if that can be done in a relatively cost effective way. It suggests that a better world would be one in which we conduct such research and find it unproblematic. The intuition that this would not be a better world ethically undermines this approach.
To consider one final version of the Utilitarian argument, one might attempt to develop a justification that allows nonbeneficial pediatric research on consequentialist grounds, subject to plausible side constraints. One version of this approach would argue that nonbeneficial pediatric research is acceptable provided the research is valuable, the risks are low, and the child's parents give their permission. Although this framework seems reasonable and mirrors many existing regulations on pediatric research, it does not provide a justification for allowing children to be exposed to some risks, nor does it offer a justification for the side constraints on the pursuit of social benefit. With respect to the former aspect, this account does not offer a response to Ramsey's claim that exposing children to risks for the benefit of others is unethical. One might respond that the side constraints minimize the extent to which children (p.85) are exposed to risks for the benefit of others. That response, however, effectively grants Ramsey his central claim and then tries to minimize the ethical damage.
Second, this approach provides no justification for placing a threshold on nonbeneficial pediatric research. If some risks are allowed for an important study, significantly greater risks seem acceptable for even more important studies. Arbitrary side constraints on the level of allowable risks do not provide a very reassuring safeguard against incremental increases in the risks to which children may be exposed in the context of nonbeneficial research. A more effective response to this concern would provide a framework that not only insists on a particular risk threshold, but justifies the choice and explains why greater risks in nonbeneficial pediatric research are unethical, even if the research thus precluded offers the potential for enormous social benefit.
The Scope of Parental Authority
Most attempts to justify nonbeneficial pediatric research focus first on the children and the impact the research has on them. The hope is to show that, on the proposed analysis, this impact is of such a nature that parents can enroll their children in nonbeneficial pediatric research and investigators can conduct it. A different approach effectively reverses the priority of these considerations, arguing that a proper analysis of parents’ authority to make decisions for their children implies that they may enroll their children in at least some nonbeneficial clinical research. The level of risk to which children may be exposed in nonbeneficial clinical research is determined, using this approach, not directly based on what it is acceptable for investigators to do in research, but on the scope of parental authority, on the level of risks to which parents may expose their children.2
One of the most prominent proponents of this view is Lainie Ross. She begins with a general analysis of the scope of parental decision making in a liberal society, and then systematically applies the resulting account to decision making for clinical care and clinical research.3 Ross points out that the two primary standards for health care decision making for individuals who are unable to make their own decisions are the best interests and substituted judgment standards.4 The latter directs surrogate decision makers to try to make the decision the patient would have made in the circumstances, given the patient's competent preferences and values. Although important to decision making for adults who have lost the ability to make their own decisions, (p.86) for instance, those who fall into a coma as the result of head injury, this standard does not apply to children, at least not to younger children who have yet to develop competent preferences and values.
This conclusion seems to imply, and many commentators and courts agree, that decisions for children should be governed by the best interests standard. Ross points out that endorsement of a strict best interests standard ignores the fact that children typically are reared in the context of a family, often with other children who have needs, preferences, and values of their own. The fact that siblings’ concerns and interests can conflict implies that parents cannot adopt the best interests standard all the time, at least not with respect to all their children. Attending the school play of one child can necessitate leaving a younger child with a baby sitter. This can be an acceptable choice to make even though it often is not in the best interests of the child left temporarily behind. Strict adoption of a best interests standard also would require parents, implausibly enough, to “disregard all personal interests in order to fulfill the children's needs, interests, and wants. That parents are individuals with needs, interests, and rights is ignored.”5
Ross argues that, in liberal societies, parents should have wide latitude to make decisions for their children, what she calls a model of “constrained parental autonomy.” In this model, parental authority to make decisions for their children is constrained only by a negative and positive obligation. The negative obligation specifies that parents may not abuse, neglect, or exploit their children, whereas the positive obligation requires parents to respect their children as developing persons and provide them with the means to live an autonomous life.6 Some have argued that parents’ obligation to provide their children with the means to live an autonomous life places very strict demands on parental decision making. Specifically, it has been argued that the ability to lead an autonomous life requires that individuals have a sufficiently rich range of life choices from which to choose.
Joseph Raz argues that realization of personal autonomy presupposes that individuals “have available to them many forms and styles of life” thus providing them with an “adequate range of options” for their lives from which to choose (not to mention the mental capacities to make this choice and lead the life thus chosen).7 Similarly, Joel Feinberg argues that, in the absence of wide agreement on what constitutes the best life, parents have the obligation to provide their children with an “open future,” one sufficiently rich in possible life plans from which the future adult can choose.8
Ross does not see the proper relationship between parents and their children in a liberal society in the same way. To a certain extent, she prioritizes parental autonomy over the developing autonomy of the child. She regards the (p.87) absence of any “consensus on what the good life entails” as implying that “parents in a liberal society must be free to choose whether to expose their child to a wide or narrow array of coherent life plans.”9 On this view, parents have the right to mold their children as they see fit, provided the children are given the means to live a viable adult life. Ross gives the example of parents who may choose to not motivate, or even to discourage their children from developing an interest or talent in sports that would conflict with the religious life the parents envision for their children.
Parents, of course, may make mistakes in deciding which type of life to prepare their children for. But as long as they provide for the child's basic needs and prepare them for some viable form of life, the state must respect the parents’ decisions.10 Applying this view of the family and parental decision making authority to clinical research, Ross argues that as long as the risks of the research are sufficiently low, enrolling children in research is one of the things that parents can do free of state intervention: “Parental authorization or prohibition of their child's participation in minimal-risk research is not abusive or neglectful, even if the child is forced to participate against his will.”11 She also allows minor increase over minimal risk pediatric research, provided the investigators obtain parental permission and the assent of children who are capable of providing it. The child's assent in this case “signifies that his participation promotes goals that he identifies as consonant with his ends (without seriously threatening his future well-being).”12 Ross argues that nonbeneficial pediatric research that poses more than a minor increase over minimal risk “is always immoral and must be prohibited” on the grounds that enrolling children in such research conflicts with parents’ obligation to protect the child's “self-regarding interests and developing personhood.”13
Ross supports the claim that parents have the authority to enroll their children in nonbeneficial research that poses minimal risk or a minor increase over minimal risk on the grounds that parents have the right to help shape their children's values and the type of adults they will become. Within the scope of appropriate parental authority is the right for parents “to steer their child's development into a socially responsible adult.” The parents may not succeed, but “it is reasonable for them to try to guide his development in this way.”14 Ross repeats this stance in her more recent book. Referring back to her earlier work she writes:
Two slightly different ways to read this view differ in terms of how strict one sees the constraints on parents’ authority to make decisions for their children. On the first reading, parents have the right to enroll their children in research that does not offer a prospect of direct benefit, provided the risks do not exceed a minor increase over minimal. This simply is within the purview of parents’ decision making authority, and the state has no claim to intervene. This argument captures one version of the parental authority assumption; namely, that parents are allowed, that it is within the scope of parental authority, to make this decision for their children.
In brief, I have argued that parental autonomy should be respected unless their decision is disrespectful of the child's developing personhood. Parents who value participation in social projects like advancing science may try to inculcate similar values into their child. (p.88) Even if the child never shares in these goals, they are goals which responsible parents may try to inculcate.15
With respect to the example of choosing between different life plans (or more accurately deciding what range of life plans their children will be prepared to pursue), it seems plausible to argue that parents can choose in a way that is not in their children's best interests, and yet still in a way that is consistent with the children's basic needs and developing autonomy. Of course, there may be disagreement regarding the extent of the range of life plans that should be open to one's children. Some argue for a wider range, others might argue for the acceptability of a narrower range, perhaps even preparing one's child for only one viable form of life. Bracketing that debate, Ross's view seems plausible in the case of life plans because it seems clear that children's basic needs can be satisfied within a wide range of more and less optimal ways of life, or ranges of life options.
Parents faced with the possibility of enrolling a child in nonbeneficial research face a different sort of decision in an important respect. Here, the question is what decisions are consistent with the parents’ obligation to protect the child's basic needs. Under this authority, parents may enroll their children in research that offers a compensating potential for clinical benefit. Parents also, presumably, may enroll their children in nonbeneficial research when it poses absolutely no risk of serious harm, and there may be, as we have seen previously, research like this. Survey research may provide some examples. But, most minimal risk research poses some risk of serious harm. The research qualifies as minimal risk because the chances of those harms occurring are very low. It is not clear how enrolling children in such research can be consistent with the obligation to protect their basic interests. In the absence of any chance for clinical benefit, how does one determine how great the likelihood of harm has to be before the decision to enroll one's child conflicts with one's obligation to protect her basic interests?
(p.89) In theory, it seems plausible to assume that a complete analysis would show that the costs and burdens of state intervention, not simply on the state, but on the children themselves, would imply that they are better off without it. Here, in theory, the threshold of risks to which parents may expose their children free from state intervention would depend largely on the costs of state intervention. There are costs to allowing the state to prohibit parents from making these decisions, and that level of costs could be used to set the level of risks to which parents may expose their children in the context of nonbeneficial research.
In some cases, state intervention can have enormous costs on the children it is trying to protect. It is not in children's interests for their parents to fail to compliment them, and one might think that the state should force parents to do so. In addition to dramatically curtailing the ways in which parents may raise their children, enforcement of this requirement would involve enormous intrusion into family life that likely would be contrary to the interests of many children. Prohibition of nonbeneficial pediatric research, in contrast, likely would not place a high burden on the children themselves. It would not require children to be taken from their parents, or even require the state to enter the home and chastise the parents. Instead, the state could prevent parents from making these decisions by directly prohibiting such research. On the assumption that few parents are likely to conduct research on their children, this approach would protect children from research risks without requiring direct intervention into the home or family life.
More importantly, Ross focuses on the decisions that parents can make free of state intervention. She concludes that we allow parents to make certain decisions. Although this argument is important for present purposes, it is not sufficient to justify pediatric research. To that end we need to answer the further question of whether it is ethically appropriate for parents to make these decisions for their children. Specifically, what level of risks, if any, is it ethically appropriate for parents to expose their children to in the context of nonbeneficial research? This question is not directly germane to Ross's task of developing an account of the proper scope of family decision making, but it is crucial for determining whether and when it is acceptable for parents to enroll children in nonbeneficial research, and when it is acceptable for investigators to conduct and society to sponsor and benefit from it.
To consider a specific instance of this difficulty, it clearly is within parental authority to at least occasionally curse at their children free of state interference. Unless the behavior rises to the level of abuse, the state does not have sufficient reason to intervene, not because we think that cursing at one's children is ethically acceptable, but because we place great value on a protected family sphere in which parents get to decide, free from state and other intervention. The serious (p.90) costs attendant on state intervention imply that states should intervene only at the point at which the cursing becomes abusive.16
Nonbeneficial pediatric research poses a different dilemma. Here, it is not simply a question of whether the state should decline to prohibit parents from making certain decisions for their children or treating their children in specific ways. Rather, the question is whether society may establish the system in the first place, encourage, and even participate in it. It would be as if someone concluded from the fact that parents have the authority to mistreat their children slightly that it would be acceptable for society to establish programs, staffed by physicians and nurses, which provide parents with this opportunity free from state intrusion. In this case, it would be feckless to enquire of such programs precisely how much the parents could mistreat their children. It is wrong to set up programs for this purpose, even though parents may do it in their own homes.
Investigators and society do not, unlike parents, have the authority to simply expose children to risks. For these individuals to be engaged in the activity, as is often true for third-party actions in general, there needs to be an independent reason to think that the practice is appropriate. As mentioned previously, a number of other authors, including Ross, offer an independent argument by citing the possibility that enrolling children in research offers one possible way to teach them to become responsible adults. This line of argument offers a possible response to Ramsey's challenge that pediatric research can be acceptable only when it is in the interests of the participating children, and the next section turns to it.
Teaching Children to Be Moral
The process of raising a child to become a mature, responsible, autonomous adult is a long and arduous one. To be successful, parents cannot simply protect their children and provide for their basic needs. The best way of doing that presumably would be to keep the children warm, well fed, and locked in a protective bubble in the attic until age 18, to be unleashed and sent forth. This overprotective approach to raising children, taking the view that one's duty as a parent is to protect one's children from harm, while psychologically understandable, is counterproductive. It has negative effects on the child and also denies him valuable experiences that he can have only when he is a child, and many of the experiences he needs to become a mature and responsible adult. Protecting children from too many risks poses its own serious risks.
(p.91) This point is crucial for the present analysis. To promote the interests of one's child, parents cannot merely adopt an approach of protecting the child from risks. The parent must be willing to take risks with and for the child. Mature and responsible adults are able to interact with and care about others, and to navigate through the environment. They must have some confidence in themselves, and enjoy some activities, projects, or hobbies. Children are not born with these abilities; they need to learn them. Many things cannot be learned second-hand or by example. Children must do to learn. Many lessons should be learned during childhood, and others must be learned during childhood, if they are to be learned at all.
There are several reasons for this. One is the nature of human development. Parents cannot wait until children reach age 18 to expose them to language for the first time and expect them to become effective users of language. The nature of brain development is such that children need to begin this learning process while relatively young. A second reason is timing. Arguably, it is in children's interests to meet Gandhi, but this opportunity might arise only when the child is young. Third, some experiences are important ones only if they occur when individuals are children. Meeting Mickey Mouse may be an experience like this. A fourth reason has to do with socialization. To lead a flourishing adult life, one must learn to live with and interact with others, to develop and maintain deep and important relationships. And to have a flourishing adult life, one must have a reasonably flourishing childhood, which itself requires the right kinds of interactions with others outside of a protective bubble.
By choosing particular experiences for one's children as a means of helping them to develop the skills necessary for a successful adult life, parents inevitably influence the kind of person the child is and becomes, her character, personality, values, and preferences. As a result, the parents’ choices will necessarily slant their children toward certain lives and away from others. To be in a position to earn a music scholarship for college, children must learn to play an instrument and practice a great deal while they are children. A girl who does not hold a tennis racket until age 18 has almost no chance of becoming a world-class tennis player. A child who is never exposed to others will have less chance of hearing mean things said to and of them, but also will have significantly less chance of developing good relationships as an adult.
In many cases, parents cannot determine which experiences to provide their children and which paths to steer them toward by appealing to the values and preferences the child has or will ultimately develop. Often parents need to make these decisions before children have developed the relevant preferences and values, and the decisions the parents make will influence the preferences (p.92) and values the children ultimately develop.17 This fact reinforces the point made earlier that parents get considerable latitude in deciding how to raise their children. It is not just that parents are allowed to influence their children based on the parents’ own preferences and values. In many cases, these are the only preferences and values extant in the family upon which the parents might base their child-rearing decisions.
In the process of rearing their children, parents certainly have the latitude and, arguably, have an obligation to teach their children to be altruistic, moral, and decent members of a community. A number of commentators have argued that one way of imparting this lesson is to enroll children in nonbeneficial research.18 Bartholome argues with respect to children that “What warrants our intervention into their lives is not our thesis about what they would or should want. What warrants our involvement is that they stand in need of our help, our care.”19 One way to respond to a child's need is to help her learn to be a responsible adult through participation in nonbeneficial research, when it poses no discernible risk. He argues that this is not obligatory, but the “child and parents might select this activity as one that could enhance the moral development of the child”20. Dan Brock offers a nice summary of this argument: “when parents teach their children that they should be willing to participate in medical research designed to benefit others, this is a small, though not inconsequential, respect in which they are helped to become moral beings, with a concern for the well-being of others besides themselves and their close relations and friends.”21
This argument provides an important lesson for evaluating the different ways in which one might justify nonbeneficial pediatric research. The fact that nonbeneficial research offers no compensating potential for clinical benefit does not necessarily imply that there is no chance for children to benefit in other ways that might justify the risks to which they are exposed. Educational benefits, especially in learning to be moral individuals, but also perhaps in learning about science, medicine, and the research enterprise, are some of the possible benefits children may realize from participating in nonbeneficial research. Notice that the ethical concern posed by nonbeneficial pediatric research, the fact of exposing some children to research risks for the benefit of others, in effect becomes a virtue on this argument. The fact that the research is designed to help others presumably becomes central to its potential to teach children lessons regarding altruism and the importance of helping others.
Learning to be altruistic is an important and valuable lesson. However, to use this as a justification for enrolling children in nonbeneficial pediatric research, one must be clear on why it is in children's interests to learn these lessons. Presumably, we all think that it is a good thing for people in (p.93) general to be willing to help others. This makes for a better society overall. But this is not sufficient in the present context to justify nonbeneficial pediatric research. We have acknowledged that nonbeneficial pediatric research has the potential to benefit society by improving overall health and well-being. Justifying the enrollment of children in nonbeneficial research on the grounds that it makes the participants better members of society only increases the puzzle, it does not answer it.
What we need is an argument that becoming an adult who is (more) likely to help others is good for the children themselves in some way or other that justifies the risks to which they are exposed.22 Other kinds of education are more clearly in the child's benefit, such as learning how to read and write. These are to the advantage of the child on most life plans in most (current) cultures. However, without further argument, a skeptic might claim that helping others is a benefit for the child himself only if the child values doing so. Therefore, guiding the child to help others cannot be justified on the grounds that it will benefit the child.
The empirical possibility that enrolling children in nonbeneficial pediatric research may teach them to be moral, and further teach them to value being moral, is both the strength of this view and also its downfall to a certain extent. Whether it provides a justification for nonbeneficial pediatric research depends on the empirical facts. Thus, pending further empirical research, it is unclear whether and to what extent it justifies nonbeneficial pediatric research. The facts may turn out in the right way to allow only the research and the levels of risks that we do allow. That is possible. And some data suggest that participation in charitable activities may provide children indirect psychological benefits.23 , 24 , 25 , 26 However, the existing data are not that strong, and the facts may turn out in ways less congenial to this justification.
Bracketing the question of how these data might turn out with respect to older children, it seems extremely unlikely that enrolling very young children in research before they can understand it has any chance of teaching them to be moral. In this way, the present argument will not offer a justification for nonbeneficial research using infants and very young children. Furthermore, the existing data on pro-social behavior, while relatively scant, suggest that children do not develop the concepts necessary to understand that there are moral reasons to help others until approximately age 10, or older.27 Before that age, they help others when they think that they have to, or they assume that they will be punished if they do not. Relying on those notions to teach children to help others might be an effective way to get them to act in accord with morality, but it does not seem an effective mechanism for teaching children to be moral; that is, getting them to do the right thing for the right reason.
(p.94) Next, many nonbeneficial research studies seem particularly ill suited for teaching children to be moral. The principal problem is that the children often experience only the burdens and pain of the procedures required, for instance, undergoing a needle stick. Undergoing this experience seems an unlikely way to teach children to be moral. Moreover, the benefits of any given research study often are not realized for many years afterward and require a good deal of sophisticated understanding to appreciate. The process by which someone else eventually benefits as the result of an investigator inserting a needle into a child's arm, taking blood, looking at that blood in a laboratory, comparing it to the blood taken from other children, and developing, testing, and marketing a new treatment is less than a perspicuous one. One who hopes to teach children a sense of altruism likely would do better to focus on contexts in which the connection between the effort and the benefit is clearer and closer.
Learning to be moral is certainly a benefit of participation in nonbeneficial pediatric research to the extent that it occurs. However, this justification applies to only some research for only some children. It follows that if educational benefit is necessary to justify nonbeneficial pediatric research, a good deal of it could not be justified. In addition, pending further empirical study it remains unclear whether any research participation is justified on this view. Given the alternative of trying to teach children these lessons in safer contexts, reliance on this approach would put nonbeneficial pediatric research in jeopardy. This conclusion is problematic if one relies solely on the educational benefit argument. Yet, there are good reasons to think that this argument does not offer the only justification for nonbeneficial pediatric research.
Imagine if future research on neurological development reveals that children are unable to learn to be altruistic between the ages of, say, 12 and 14. What would we conclude from this research? It seems unlikely that we would conclude that it is always unethical to enroll 13-year-olds in nonbeneficial research. We might conclude that it is better to engage older children who can learn from the experience. But, when the risks are low, the benefit important, and particularly when children of this age are the only suitable participants, it seems we would be willing to expose them to some research risks for the benefit of others. This suggests that the potential for educational benefits is supplementary in some cases, not the primary justification for enrolling children in nonbeneficial research. One possibility is that children should sometimes at least act in ways that help others, even at some risk to themselves. It is hoped that they will learn from these experiences, but that is not the justification for having the children act in this way. This suggests a slightly different argument, according to which we allow children to be enrolled, not because it necessarily will teach them to be moral, but in some cases at least (p.95) because we think that children should do the moral thing. This line of reasoning recalls a widely cited passage in the literature on nonbeneficial pediatric research.
Willard Gaylin quotes a father who orders his son to give a small amount of blood for research purposes after the son has refused to do so on the grounds that the needle stick will hurt. In the father's words: “I’ll be damned if I was going to allow my child, because of some idiotic concept of children's rights, to assume that he was entitled to be a selfish, narcissistic little bastard.”28 This passage is quoted frequently by individuals who endorse the argument that enrolling children in nonbeneficial research can teach them to become altruistic and unselfish members of the community. However, from the passage, it is not clear that this is what the father is saying. The father is concerned that the child not act selfishly. That is, the father wants the child to do the right thing, now. One assumes that even if this father learned that the child was not of the proper age to learn to be altruistic he would still want his child to do the right thing.
The difficulties with the educational benefit argument are underscored by the competing analyses offered of this example. Bartholome argues that the father was teaching the child that physical force wins.29 Diekema states that the father was “teaching the child that if you want something from someone else, it is okay to try to get it without asking permission first.”30 In Ross's view, the father was teaching the child that others may veto the child's participation, despite the fact that the father endorsed it. This, she argues, “teaches the child that strangers can restrain his parents’ authority, even when it is not abusive or neglectful.”31 Which of the interpretations is correct depends on what, in fact, children learn in these contexts. And answering that question will require future research into the very complex process of learning. Future research may turn out to support one view or another, but that is not certain at the beginning and, thus, does not provide much support for the claim that nonbeneficial pediatric research is acceptable.
Granting the limitations of this argument for the purposes of providing a justification for nonbeneficial pediatric research, recognition that children can benefit from participation in clinical research in nonclinical ways suggests the possibility of considering (which we will pursue) whether children might benefit in other ways from participation in nonbeneficial research.
Children's Moral Obligations
A number of commentators argue that nonbeneficial pediatric research can be acceptable, not because it teaches children valuable lessons, but because (p.96) children have an obligation to contribute to the common good, to helping others. Robert Veatch argues that “if the individual is seen as a member of a social community, then certain obligations to the common welfare may be presupposed even in cases where consent is not obtained.”32 Veatch's argument suggests that children have this obligation as the result of enjoying the benefits of living in a community. Or, more specific to clinical research, children have gained tremendous benefits as a result of the conduct and completion of previous clinical research, and from those individuals who faced the attendant research risks for the benefit of others, including, in effect, if not specifically in intent, the present child.
The enjoyment of benefits typically does not impose obligations on recipients who have no choice but to accept them. You are not obligated to reciprocate if I offer you a gift you literally cannot refuse. Organizations often mail random individuals small gifts, calendars, and greeting cards. These organizations rely on some individuals wanting or perhaps even feeling obligated to reciprocate by donating money to the organization. While this may be a laudable response, individuals who receive these gifts unsolicited are not obligated to donate to the contributing organizations, even if they retain the gifts and benefit from them.
Children likewise do not have the opportunity, at least at a young age, to decline the benefits of past research. Future children who receive the rotavirus vaccine will have benefited greatly from the efforts of those children who participated in the studies to develop the vaccines in the 1980s and 1990s. However, in many cases, the children have this benefit forced on them. Most children, at least at the time they receive the vaccines, try their best to avoid the benefit. The children do not reject the vaccine under the description of the benefit it conveys, because they cannot understand that description at the time they receive the vaccine. But they often do their best to avoid the injection that delivers the vaccine, and certainly are not accepting the benefit in a context in which they meaningfully could have declined it. Similarly, younger children cannot give informed consent to being enrolled in clinical research because they cannot understand sufficiently the potential benefits and risks of the research, including the potential benefits and risks of the studied treatments. It follows that they cannot understand information that would be essential to their being able to agree to participate in future research in exchange for present medical benefits.
One might argue that at least one instance exists in which enjoying a benefit does confer obligations, even though the recipients did not understand the nature of the benefit and did not voluntarily accept it in a context in which they could have declined. Specifically, children might have an obligation once (p.97) they become adults to care for their parents in the parents’ old age. The children benefited enormously from their parents efforts in rearing them. The parents, let us assume, worked hard and sacrificed for 18 years, attending to the children's needs and providing them with the environment to develop the skills and capacities for a valuable adult life. This effort represents enormous expenditure on the parents’ part and the receipt of enormous benefits on the children's side, benefits that the children did not understand for many years and may well have never voluntarily accepted.
It does not seem completely implausible to argue that, in these cases, children have a moral obligation to care for their parents. For present purposes, the important point is not to settle this issue, but to note a key aspect of it. Whether this view ultimately succeeds, it is based on the plausible assumption that, if nonvoluntary and even unknowing receipt of a benefit ever can obligate one to help others, the obligation is owed to those who bestowed the benefit. The plausibility of the example of children having an obligation to assist their parents traces to the extent of the parents’ efforts for their children. One wants to say that an individual has to help in a time of need or incapacity someone who did so much for them when they were in a time of need and incapacity. The obligation is owed to those who bestowed the benefit, and denying it would be ungrateful, possibly to the point of being unethical.
This line of argument is not being made in the case of nonbeneficial pediatric research and, in effect, this argument cannot be made. Research intended to develop medical treatments for adults must be conducted with adults because they differ from children in terms of their metabolism, physiology, immune systems, and the like. For the same reasons, medical interventions for children must be tested on children. Moreover, as we have seen, the process of testing medical interventions typically takes many years, yielding a generational gap between participants and beneficiaries. This gap is especially extended for pediatric research that does not offer a compensating potential for clinical benefit because such research typically involves the earlier phases of a research project, the phases of identifying possible medications and determining appropriate dose levels and toxicities. These phases occur even before the tests to evaluate the possible efficacy of the treatment in question, and long before possible approval and marketing of the new treatment.
Children who participate in early-phase nonbeneficial clinical trials tend to benefit other children 10-15 years after their participation. At that point, the children who were enrolled in the trials are now teenagers or more typically adults, assuming their diseases were not fatal. As such, the participants no longer need medical interventions designed for young children. In this way, (p.98) those children who will benefit from the research involvement of this generation are not in a position to repay these same individuals, at least not by participating in clinical research themselves. The research efforts of one generation benefit subsequent, not previous generations. If current children owe a debt that goes beyond gratitude to those who participated in previous research, they will have to discharge it in some way other than participating themselves in nonbeneficial research.
Finally, if the justification for enrolling children in nonbeneficial pediatric research is that these children benefited, however unwittingly and possibly unwillingly, from the research efforts of previous children, it is unclear how one establishes the system in the first place. If the justification for enrolling children in nonbeneficial pediatric research is based on the benefits they have enjoyed from previous research, how does one justify the original nonbeneficial pediatric research studies? By assumption, the children who participated in the earliest clinical trials did not benefit from the research efforts of previous generations of children because there were no previous clinical trials. Thus, either the enrollment of children in the initial studies was unethical or there exists some other justification for enrolling children in nonbeneficial research that does not depend upon their having enjoyed its benefits.
Two authors have developed accounts of an obligation to participate in clinical research that avoid the present concerns. Arthur Caplan considers individuals who receive care in teaching hospitals that pursue research and, it is assumed, provide better care in the process.33 He argues that at least some of the individuals who repeatedly return to teaching hospitals for their medical care could have instead gone to nonteaching hospitals. These individuals are repeatedly accepting the benefits of past clinical research in a context in which they could have declined them. Caplan argues that in this context: “Fair play requires that those who knowingly and willingly seek out and accept the benefits of better care, closer attention, and the higher levels of medical skill that are often available in a teaching hospital incur a general obligation to serve” as subjects of medical teaching and as subjects of clinical research.34
By focusing on individuals who continue to enjoy the benefits of past research over time, and who could have declined to enjoy those benefits, Caplan tries to develop an argument for an obligation to participate in future research in contexts that fall short of one formally agreeing to do so. Whether this argument works for competent adults who have the option of seeking care elsewhere, it does not apply to children who are brought to the hospital by their parents. As with the proposed justifications based on consent, the present argument succeeds only to the extent that one ignores the premise on which the dilemma is based; namely, that children cannot understand and (p.99) consent, whether it is to receiving the benefits of health care or being enrolled in clinical research.
John Harris claims that the argument from fair play does apply to children.35 He argues that our duty to help others and the notion of fair play yield an obligation on the part of those who have enjoyed the benefits of previous clinical research studies to participate in future clinical research studies which have the potential to improve health and well-being for others. As he points out, at least everyone living in higher-income countries has benefited from the results of clinical research. Harris is not arguing for an enforceable obligation. He is not arguing that individuals who have realized these benefits can be forced to participate in future research studies. Rather, he is arguing that our duty to help others and the notion of fair play provide strong reasons for each of us to participate in important research. And this reason, he claims, applies to children as well:
This passage suggests three related but distinct arguments. First, children may have an obligation to participate in research simply by virtue of having received and benefited from previous research efforts. This argument we have considered and rejected. A second argument is that in deciding whether to enroll children in research parents should ask what it would be reasonable to assume the child would decide, if the child were able to make her own autonomous decisions. On this view, declining to enroll the child in research is “to impute moral turpitude as a default.”37
If children are moral agents, and most of them, except very young infants are, then they have both obligations and rights; and it will be difficult to find any obligations that are more basic than the obligation to help others in need. There is therefore little doubt that children share the obligation argued for in this paper, to participate in medical research. A parent or guardian is accordingly obliged to take this obligation into account when deciding on behalf of her child and is justified in assuming that the person they are making decisions for is or would wish to be, a moral person who wants to or is in any event obliged to discharge his or her moral duties. If anything is presumed about what children would have wished to do in such circumstances the presumption should surely be that they would have wished to behave decently and would not have wished to be free riders.36
It is not clear when making other decisions for their children that parents ask or should ask what decision the child would make in the circumstances. (p.100) This certainly is not the common approach used to determine whether to enroll children in school or music lessons. For younger children, it is not even clear, as we saw previously, that it makes sense to ask what decision they would make if they were able to make a decision. To answer that question, one must first imagine that the children are competent adults. And that supposition eliminates from the children the very characteristics which press the problem in the first place.
A third approach would be to argue that the obligation to participate in nonbeneficial research traces to an obligation to support and participate in the social structure or system of which nonbeneficial pediatric research is a part. On this view, the obligation does not trace to benefits the individuals in fact received from the efforts of previous research participants. Rather, the obligation is to the overall social system of which clinical research is a part. This argument has the potential to avoid the present concerns. In particular, it may offer the possibility of justifying the first nonbeneficial pediatric research interventions and studies, and it does not necessarily imply that those who enjoy the benefits of the system are obligated to help those specific individuals who faced burdens and risks in the studies from which these benefits were derived. The next section considers perhaps the most prominent way to argue for such an obligation.
The View from Behind the Veil of Ignorance
In the previous section we considered the possibility that children are obligated to participate in nonbeneficial research given the medical benefits they enjoy as the result of the research-specific efforts of previous children. Here it is believed that the obligation traces to the fact that a particular child enjoyed specific benefits that were the result of trials conducted previously. The failure of this argument suggests the potential for a more general one. Rather than the obligation following from the receipt of particular benefits, one might argue that the obligation follows from the more general benefits children receive as the result of being raised in the context of a cooperative scheme or society. Children are obligated to do their part for the scheme because of the many benefits they enjoy as a result of their good fortune of being born within it.
Initially, the same challenge arises here of explaining why children have certain obligations as the result of receiving benefits unknowingly and often involuntarily. One way to justify the obligations that arise out of a cooperative social scheme is on contractualist grounds. However, contract theories have (p.101) difficulties with precisely those groups, such as children, who do not accept in any meaningful way the benefits of the social system under which they are living. It has difficulty with those who are not able to meaningfully enter into a contract.
One might try, in a Rawlsian vein, to establish children's obligation to participate in nonbeneficial pediatric research based on the choices individuals would make regarding the structure of society from a position of ignorance regarding their own place within the society, from behind a veil of ignorance. To make this argument, one will have to modify the Rawlsian argument in several respects. The knowledge that one is an adult could well bias one's decision regarding nonbeneficial pediatric research from behind a veil of ignorance in favor of more, and perhaps more risky, pediatric research. After all, those who know they are adults at the time the decision is being made are no longer in danger of being exposed to the risks of nonbeneficial pediatric research. They may thus be inclined to regard the overall benefits to be worth serious risks to a few children. Conversely, the knowledge that one exists without knowing one's age may well bias one in favor of very limited or even no pediatric research. This information implies that one is in a position to benefit from the research efforts of previous generations. Any further benefits to be realized from the conduct of more nonbeneficial pediatric research are unlikely to be realized during one's childhood, when they would be of interest to oneself.38
Dan Brock argues that, to avoid these biases, the “veil of ignorance has to be stretched so that one does not know to which generation one belongs—past, present, or future—when one decides whether to accept and to participate in a practice in which children, though unable to give or withhold consent, participate in research potentially beneficial to other children.”39 He argues that individuals under this stretched veil of ignorance presumably would agree to pediatric research as long as the benefits of the practice exceed its overall burdens. On this basis, one could argue that justice as fairness gives all individuals, even children, an obligation to participate in clinical research when their turn comes.
This approach seems to have the advantage of explaining why we can expose children to some risks for the benefit of others, and why parents can give permission for their children to participate in such research.40 This approach has the added virtue of avoiding concerns raised by the substituted judgment version considered previously. It does not require us to imagine what decisions a particular child would make, if he were competent. The individuals behind the veil of ignorance are not the specific individuals being enrolled in a given study. Instead, we are to imagine what choices reasonable and competent persons would make in that situation.
(p.102) Rawls argued from the original position to the basic structure of society; that is, to a fair arrangement of the basic institutions in the society. If the structure of society meets these basic conditions, members of the society cannot argue that the resulting distribution of benefits and burdens is unfair. Thus, in considering what the argument would imply for more specific questions, such as our policy regarding nonbeneficial pediatric research, we are taking the argument beyond the areas in which Rawls used it. This is not to claim that such uses are inappropriate, only that they should be recognized as extensions of the Rawlsian approach.
The conclusion, on Rawlsian grounds, that the structure of society meets the conditions for fairness, does not imply that individuals are obligated to participate in the society so structured. Competent adults can decide to leave a society that meets these conditions (whether they have any better places to go is another question). The right of exit suggests that the fairness of the system does not generate an obligation to participate in it, but instead defends the system against those who would argue that it is unfair to some of the participants over others. At most, then, the present argument can show that it is not unfair to enroll a given child in a research study, that this is a reasonable thing for all children. But the fact that the system is fair in this sense does not provide a compelling reason for enrolling children in nonbeneficial research. Put differently, a system may expose individuals to risks for the benefit of others in a fair way without implying that it is acceptable to expose individuals to those risks in the absence of their consent.
A second concern relates to whether hypothetical contract arguments provide an independent or noneliminable reason for a particular program or policy. To see this, consider why rational individuals behind the veil of ignorance would select one policy over another. The answer presumably is that there are good reasons to select that option, that on balance the relevant reasons support its selection. But, on this explanation, it appears that one does not need to appeal to the hypothetical contract at all. One could simply ask which option the relevant reasons on balance support.41 To put the same point in a negative frame numerous commentators who have systematically analyzed the ethical appropriateness of nonbeneficial pediatric research have been unable to agree on the appropriate policy governing it. In this context, it is not clear how we might make progress by imagining, say, these same individuals being challenged to select the most reasonable approach from behind a veil of ignorance. The difficulty we have had developing a consensus does not trace to the kind of biases that Rawls attempts to avoid. Rather, the lack of consensus traces to the fact that it is not clear what the right approach or policy is.
(p.103) Third, it is important to consider the factors to which individuals behind the veil of ignorance can appeal in making their choices. In particular, are these choices constrained or guided by moral considerations? An obvious response is to think that the decisions would be so constrained. After all, we are asking what is the ethical approach or policy with regard to nonbeneficial pediatric research. The problem, then, is that what answer we get in this case may depend significantly on which ethical constraints we build into the system, making the whole approach question-begging. If we include Ramsey's constraint that it is unethical to expose those who cannot consent to research risks for the benefit of others, the reasonable policy to choose from behind the veil of ignorance would be one that prohibits nonbeneficial pediatric research.
Proponents might avoid this dilemma by assuming that individuals behind the veil of ignorance will make decisions based purely on self-interest, unconstrained by moral limits or considerations. Brock considers this approach and points out that, under those conditions, rational and self-interested individuals presumably would agree to any system that on balance produces more benefits than harms overall. Presumably, many different systems would satisfy this requirement. In particular, the system that produces the greatest amount of benefits overall may well be one that we regard as unethical. This possibility raises concern that the present version of the Rawlsian approach is ill-suited to identifying the morally appropriate approach, leaving aside the extent to which it might identify which approach would be reasonable for such individuals or which approach they have reason to endorse.
Consider the example of the HIV injection studies mentioned previously as a (presumably unethical) means to developing a vaccine for HIV. Here, the risks are a high chance of acquiring HIV for a few individuals in the context of research studies until an effective vaccine is developed. Assuming an effective vaccine is identified, the ultimate benefit of this approach will be that of saving tens of millions and, if one stretches the benefit into future generations, hundreds of millions of people from becoming infected with HIV. Redescribing this choice set as one that faces a single individual who makes the decision on self-interested cost–benefit grounds yields the question of whether such an individual would accept the risk of being one of the individuals enrolled in these studies for the benefit of otherwise avoiding HIV infection. If we assume the studies can be completed in a generation or two (assume for this purpose the existence of a sufficient number of potential vaccines), several hundred of the approximately 10 or so billion (assuming current world population figures) people who are alive in the world during that period of time will need to be enrolled. Because the vaccine protects future generations as well, the number (p.104) of those who benefit is a multiple of those alive during the time the vaccine studies are conducted. Assuming this approach would speed identification of an effective vaccine by even a few generations suggests that enrolling 500 people in vaccine challenge studies might benefit 50 billion people. The choice facing individuals behind the veil of ignorance, then, is: Would a reasonable person accept the risk of approximately a 1 in 100 million chance of being injected with the HIV virus after being given a potential vaccine in order to gain the benefit of otherwise being protected from the virus?42
The veil of ignorance filters out facts that would allow individuals to determine their personal chances of acquiring HIV. They do not know if they will be rich or poor, male or female, live in Sweden or Zimbabwe, a priest or addicted to intravenous drugs. Presumably, on the present application of the original position, all they know is the chances that they will become HIV infected based on the probabilities that an average person will become so infected. Under these conditions, it seems likely that the self-interested individual would choose to allow the HIV injection studies for the simple reason that the gamble will likely be a positive one. Specifically, the background average chance of acquiring HIV infection is likely to be significantly greater than the risk of acquiring HIV as a result of being enrolled in one of these studies. It follows that a self-interested individual would allow the studies, a result which we have agreed is unethical.43
The problem is that a purely self-interested individual trying to maximize expected benefits from behind a veil of ignorance makes a decision based on overall benefits and burdens, independent of where those benefits and burdens happen to fall. More precisely, the individual wants the benefits to be sufficiently distributed to provide a sufficient chance that he will enjoy them. The problem is that the individual is likely to prefer the option that places all of the burdens on a few individuals so that the chances of his experiencing them approaches zero. For this reason, the individual ends up making a decision on grounds similar to a Utilitarian approach in the sense of not caring about a fair distribution of risks and benefits, nor caring about the way in which the risks and burdens are realized (with the difference that the individual makes the decision based on an averaging of the risks and benefits rather than on the overall risks and benefits).
Rawls famously avoids this concern when he uses the original position to determine the basic structure of society by arguing that individuals behind the veil of ignorance will choose the greatest overall benefits tempered by the so-called maximin strategy. That is, Rawls assumes that individuals behind the veil of ignorance will be conservative in the choices they make in a particular way. They will choose the option that works to the benefit of those who are (p.105) least well off in the society. For example, Rawls claims that reasonable individuals would not select a world in which the bottom tier of the population lives in misery, even if this allows everyone else to live in opulence. To avoid the chance of having a life of misery, reasonable individuals would instead choose a world in which everyone lives a comfortable middle-class existence.
In the Theory of Justice Rawls acknowledges that the maximin principle “is not, in general, a suitable guide for choices under uncertainty.”44 He argues that it becomes plausible in situations, like the original position, that have certain special features.45 It becomes less clear whether these features are part of the original position when one applies it beyond the question of the basic structure of society, as is being considered here. Leaving aside the question of the extent to which the present argument retains key features of the original position as Rawls envisioned it, it is not clear that appeal to the maximin strategy will save the present version of the Rawlsian argument. The question here is not whether a reasonable person would choose to make the poor even worse off in order to elevate the status of those more privileged. Rather, both options involve some individuals being in very unfortunate circumstances (i.e., infected with HIV). The difference is that the one option involves many more individuals becoming infected over time, whereas the other option involves significantly fewer individuals being infected, but some as the result of being injected with HIV in the process of identifying an effective vaccine. Since the less-preferable option (being infected with HIV) is the same in both cases, the reasonable choice, even if one accepts the maximin strategy, seems to be whichever option reduces the number of individuals who occupy that position. As we have seen, this may well be the world in which the injection studies are allowed.
Finally, the present defense of nonbeneficial research seems to misconstrue the relevance of consent to exposing someone to risks for the benefit of others. The problem comes when one tries to apply the hypothetical contract model not to the basic social structure of society as Rawls did, but to the specifics of the way in which individuals will be treated. It appears that a rational person behind the veil of ignorance, unencumbered by ethical principles, would accept the HIV injection studies. And that conclusion can be avoided only if the background or average rate of acquiring HIV is low enough that the chances of acquiring it are less than the chances of acquiring it in a world in which the injection studies are allowed. This suggests that the current argument will be relatively unstable. Whether it justifies nonbeneficial research studies of different kinds will depend upon empirical facts that may well change over time. That seems problematic in itself, but hides a deeper concern.
(p.106) If the probabilities turned out that the reasonable person would not agree to allow the studies, we would conclude, as seems correct, that these studies are unethical. But, notice what the argument is for their being unethical. It is that allowing the studies is not something to which a reasonable person would agree behind a veil of ignorance stretched across generations. That explanation seems to misconstrue the primary ethical issue raised by nonbeneficial pediatric research. The concern with enrolling children in such studies is not, at least in the first place, a matter of whether competent and reasonable persons would agree to allow these studies and risk being enrolled in them. The concern is that children are not competent and reasonable, hence, cannot consent to their being so treated. The present approach misses the ethical work done by actual consent, as opposed to hypothetical consent. The very act of consent changes the normative conditions relevant to a situation. In the present case, it seems that very high risks might be acceptable, if they can be acceptable at all, only if the individuals so exposed in fact give their full and voluntary consent. Exposing children to such risks is worrisome because they cannot consent. Their inability in this regard is not remedied by imagining what choices a competent and reasonable individual would make from behind the veil of ignorance. In effect, then, the present concerns echo the problems we have seen with Ramsey's and McCormick's arguments, namely, a failure to take seriously the fact that children are unable to consent.
(1) . Critics of the U.S. regulations might point out that the 407 category does not have a strict limit on risk level, but requires only that the “research will be conducted in accordance with sound ethical principles.” Although I will not argue it here, this requirement does not seem a Utilitarian one but includes an implicit risk limit. Even if one does not accept this argument, the fact remains that the other categories are not strictly Utilitarian, nor are the many non-U.S. regulations that do not include an analogous category.
(2) . This account may have difficulty explaining why it can be acceptable for individuals other than parents to enroll children in research.
(3) . Ross LF. Children, Families and Health Care Decision Making. Oxford: Clarendon Press, 1998.
(7) . Raz J. The Morality of Freedom. Oxford: Clarendon Press, 1986:425.
(8) . See, for instance, Feinberg J. The child's right to an open future. In: Aiken W, LaFollette H, eds. Whose Child? Children's Rights, Parental Authority, and State Power. Totowa, NH: Littlefield, Adams, 1980:124–153.
(9) . Ross LF. Children, Families and Health Care Decision Making, p. 49.
(10) . Ibid., p. 23. It is worth noting that the disagreement may not be as stark as all this. Commentators often focus on what parents should do, how ideally they should raise their children. Ross may have no objection to the claim that parents should raise their children to have a wide range of life plans. Her focus instead is on the limits of state intervention in the family context. What obligations do parents have to their children such that failure justifies state intervention?
(11) . Ibid., pp. 92–93. One noteworthy implication of Ross's view is that she would not require, contrary to the view of many regulations and commentators, pediatric assent for nonbeneficial pediatric research that poses minimal risk. In her view, the independent determination of the extent to which the child has a say in whether (p.305) he participates in minimal risk research represents an improper infringement on parental decision making for their children.
(15) . Ross LF. Children in Medical Research: Access Versus Protection. Oxford: Clarendon Press, 2006:22.
(16) . This discussion recalls the distinction between what Buchanan and Brock call guidance principles and intervention principles. See Buchanan AE, Brock DW. Deciding for Others: The Ethics of Surrogate Decisionmaking. New York: Cambridge University Press, 1989:10–11. The thought that best interests is the correct guidance principle does not imply that any deviation from this principle justifies intrusion by the state or others to block parental authority. Several reasons for this exist, most notably, the fact that state intrusion poses many costs.
(17) . For discussion of this point, see Ackerman, TF. Fooling ourselves with child autonomy and assent in nontherapeutic clinical research. Clin Res, 27 (5), 345-8, 1979.
(18) . See Curran WJ, Beecher HK. Experimentation in children a reexamination of legal ethical principles. JAMA 1969;210:77–83.
(19) . Bartholome WG. Central themes in the debate over involvement of infants and children in Biomedical Research: A critical examination. In: Van Eys J, ed. Research on Children: Medical Imperatives, Ethical Quandaries, and Legal Constraints. Baltimore: University Park Press, 1978:73–74.
(21) . Grodin MA, Glantz LH, eds. Children as Research Subjects: Science, Ethics and Law. New York: Oxford University Press, 1994, p. 89.
(22) . Alternatively, one might argue that parental obligations go beyond the child's own best interests to include raising the child to be a better person. Here one would need to explain the basis for this obligation. If the obligation is a social one, one wonders whether the same account might imply that children have an obligation to participate in some research, thus obviating the need for the intermediate step of the child's learning to be moral.
(27) . For argument and citations, see Wendler D, Shah S. Should children decide whether they are enrolled in non-beneficial research. Am J Bioethics 2003;3:1–7.
(28) . Gaylin W. Competence: No longer all or none. In: Gaylin W, Macklin R, eds. Who Speaks for the Child: The Problems of Proxy Consent. New York: Plenum Press:27–54.
(29) . Bartholome WG. Ethical issues in pediatric research. In: Vanderpool HY, ed. The Ethics of Research Involving Human Subjects: Facing the 21st Century. Frederick MD: University Publishing Group, 1996:356–361.
(30) . See Diekema DS. Conducting ethical research in pediatrics: A brief historical overview and review of pediatric regulations. J Pediatr 2006:149:S3–S11.
(31) . Ross LF. Children in Medical Research: Access Versus Protection, p. 97.
(32) . Veatch RM. The Patient As Partner: A Theory of Human-Experimentation Ethics. Bloomington: Indiana University Press, 1987:57.
(33) . Caplan AL. Is there a duty to serve as a subject in biomedical research? IRB: Ethics and Human Research 1984 Sep–Oct;6:1–5.
(35) . Harris J. Scientific research is a moral duty. J Med Ethics 2005;31:242–248.
(38) . For a discussion of the application of contract theory to participation in clinical research and the resulting generational problem, see Heyd D. Experimentation on trial: Why should one take part in medical research? Jarhrbuch for Recht und Ethik (Annual Review of Law and Ethics) 1996;4:189–204.
(39) . Brock DW. Ethical issues in exposing children to risks in research. In: Grodin MA, Glantz LH, eds. Children as Research Subjects, p. 95.
(40) . This argument, as Brock acknowledges, implies not simply that clinical research is acceptable, but that individuals have an obligation to participate in it. That is, it implies that adults are obligated to participate in clinical research, although for practical reasons we might refrain from forcing them to do so.
(41) . The classic formulation of this concern was voiced by Ronald Dworkin (Dworkin R. The original position. In: Daniels N, ed. Reading Rawls. Stanford, CA: Stanford University Press, 1989:16–53, especially pages 17–21). Cynthia Stark tries to fend off this “standard indictment” of hypothetical consent by arguing that hypothetical consent does not yield obligations but rather provides one with a moral reason for acting in accord with the decision that one would have made in the hypothetical context (Stark C. Hypothetical consent and justification. J Philosophy June 2000;97(6):313–334).
(42) . One could, if one is uncomfortable with the implicit assumption of 100% efficacy, adjust this figure. The conclusion here would be unaffected as long as the vaccine is at least somewhat effective and would be widely available. Rawls claims that individuals in the original position do not have access to probabilities: “the veil of ignorance excludes all knowledge of likelihoods. The parties have no basis for determining the probable nature of their society, or their place in it. They have no basis for probability calculations” (Rawls J. A Theory of Justice, revised edition. Cambridge, MA: Belknap Press, 1999:134). Leaving aside the question of whether individuals in the original (p.307) position have any basis for determining probabilities, it is unclear how one could insist on this limitation in the present case and yet still use the original position to determine the appropriate policy with respect to nonbeneficial pediatric research. It seems unclear, for example, how one would go about deciding whether to allow nonbeneficial pediatric research without knowing anything about the probabilities of its leading to harm and benefit, for example.
(43) . A similar problem faces a recent attempt to justify nonbeneficial research with adults who are not able to provide informed consent (e.g., adults with severe Alzheimer disease). The argument is that nonbeneficial research using incompetent adults is ethical because such individuals are better off in a world that allows such research to proceed. The overall benefits to individuals of allowing such research justify the risks they face of being enrolled in specific nonbeneficial studies. The author responds to the present concern by arguing that high-risk research would subject individuals to significant burdens that could well outweigh the benefits from research conducted with those who cannot consent. The HIV challenge studies illustrate the problem with this response: it depends on the details of the case and the specific number of individuals who will benefit. Yet, the intuition that the challenge studies are unethical is not a function of the number of people who will benefit. See Coleman CH. Research with decisionally incapacitated human subjects” an argument for a systematic approach to risk-benefit assessment. Indiana Law Journal Summer 2008; 83:743.
(44) . Rawls J. A Theory of Justice, p. 133.
(45) . The three special features are that a) the veil of ignorance excludes all knowledge of likelihoods (see footnote 41); b) the choosers have a conception of the good, leading to their caring “very little, if anything, for what he might gain above the minimum stipend” (p. 134); and c) the alternatives have outcomes that “one can hardly accept” (Rawls J. A Theory of Justice, pp. 133–135).