The Ethics of Net-Risk Pediatric Research: Implications of Valueless and Harmful Studies

Net-risk pediatric research involves interventions and studies that do not offer participants a potential for clinical benefit that justifies the risks they face. These interventions and studies are central to efforts to improve pediatric medical care.¹ To evaluate the brain scans of children with autism, investigators scan healthy children to establish normal values. To test experimental treatments, they administer single doses as part of phase 1 studies. Investigators also randomize pediatric participants to sham vaccines as part of efficacy studies and they perform blood draws that are purely for research as part of potentially beneficial treatment trials.

These interventions and studies are socially valuable. They also expose nonconsenting children to risks for the benefit of unrelated others, thereby raising important ethical concern. Regulations in the United States and many other countries attempt to address this concern by allowing institutional review boards (IRBs) to approve net-risk pediatric research studies and interventions only when they satisfy at least two conditions: the research has the potential to gather socially valuable data, and the net risks are low, often defined as “minimal” or, in some cases, a “minor increase” over minimal.²

One might assume that net-risk pediatric research that satisfies these conditions is ethically acceptable because it has the potential to collect important data and does not expose children to any risk of serious harm. The problem with this proposed justification is that almost all definitions of minimal risk permit interventions that pose some risk of serious harm, provided the chance of serious harm is very low. For example, U.S. federal regulations governing research with human subjects define minimal risks as risks that are “not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.”³ One of the risks that children ordinarily encounter in daily life is the very low, but not zero, chance of dying in a car accident. It follows, according to this definition, that interventions that pose a risk of serious harm, even death, can qualify as minimal risk provided that the chances of serious harm are very low.

Alternatively, one might stipulate that IRBs should approve net-risk pediatric research interventions and studies only when it is known that they pose absolutely no chance of serious harm. Unfortunately, this approach would preclude almost all research interventions. Even brain scans, such as magnetic resonance imaging (MRI), and single blood draws pose some, very low chance of serious harm. A number of commentators conclude that net-risk pediatric research is unethical. Most prominently, Paul Ramsey has argued that “no parent is morally competent to consent that his child shall be submitted to hazardous or other experiments having no diagnostic or therapeutic significance for the child himself.”⁴ In the most prominent U.S. court case to assess net-risk pediatric research, the highest court in Maryland, the Court of Appeals, agreed. The court argued that, even when parents agree, and even when the research is socially valuable, it is unethical to enroll children in research that poses a “negative balance” of risks and clinical benefits.⁵

If these claims are right, net-risk pediatric research represents a failure on the part of parents, investigators, and society to meet their basic obligations to protect children and promote their interests. Hence, social policy should ban such research, and IRBs should decline to approve it. This conclusion raises a pressing need to assess whether it can be ethically acceptable to enroll children in net-risk research that is socially valuable and poses low risks.

The Present Challenge

One option would be to endorse a utilitarian justification. If the research is socially valuable, and the risks are very low, the potential benefits to future patients can justify the risks to participants.⁶ Unfortunately, this approach, like utilitarian justifications in general, has the potential to permit excessive risks. Consider a net-risk study that is part of the development of a vaccine for a disease that kills hundreds of thousands of children every year, such as malaria. Under a utilitarian justification, the potential to ultimately save the lives of millions of children might justify exposing a few pediatric participants to high risks of injury or even death. The fact that these studies would be unethical, no matter how great their social value, refutes a purely utilitarian justification. It also highlights the conditions that a justification for net-risk pediatric research must satisfy. Successful justifications must explain why it can be acceptable to expose children to some net risks for the benefit of unrelated others, while precluding the possibility of exposing children to high net risks, even in the context of extremely valuable studies.

A number of justifications have attempted to satisfy these conditions. Unfortunately, they all have significant shortcomings, leading commentators to conclude that no successful justification for net-risk pediatric research has been identified. Norman Fost has argued that a convincing justification for such research “has never been made,”⁷ while Thomas Murray finds that the arguments in support of it are “notably weaker than the arguments favoring a complete ban.”⁸

To consider one of the more prominent proposals, some commentators argue that pediatric research that is socially valuable and poses low net risks is acceptable because scrupulous or reasonable parents regard it as acceptable.⁹ The first problem with this proposal is that ostensibly reasonable parents have historically accepted practices that we now regard as morally problematic, such as flogging their children.

The second problem is that many reasonable people regard net-risk pediatric research as unethical. A survey of pediatricians and researchers in the United Kingdom and Canada found that approximately half of the respondents believed that it is unethical to enroll children in research that does not offer them the potential for clinical benefit.¹⁰ Similarly, in a survey of the chairpersons of German ethics committees, 41% regarded net-risk pediatric research as unethical.¹¹ These data reveal that, while many reasonable people support such research, many other reasonable people regard it as ethically problematic. Appeal to what the reasonable parent believes will thus not settle the debate. What we need is a positive reason to believe that net-risk pediatric research can be ethically acceptable when it is socially valuable and poses low risks.

The Potential for Nonclinical Benefits

Some commentators argue that net-risk pediatric research is acceptable because it can contribute to the moral development of children and teach them the value of altruism.¹² This proposal offers a critical insight. Participation in research that does not offer the potential for clinical benefit may offer participating children the potential for nonclinical benefits. Indeed, the absence of clinical benefits is critical to the potential for participating children to realize these developmental benefits. It is the experience of facing risks for the benefit of others that offers children the opportunity to learn the value of altruism.

The potential to gain these educational or developmental benefits depends on the specific children and the study in question.¹³ Some children may learn from only certain types of research interactions, and some net-risk studies may not be conducive to learning the value of altruism at all. In addition, the fact that the health benefits of net-risk research may not be realized for years, even decades, after the initial studies are completed suggests that there may be better settings in which to teach children these lessons. Finally, children must have fairly developed cognitive capacities to learn the value of altruism.¹⁴ Hence, this proposal has the potential to justify at most net-risk research with older children.

Recognizing these concerns, a more recent proposal attempts to develop a justification for net-risk research with children of all ages.¹⁵ This proposal points out that young children who lack the cognitive capacities to learn the value of altruism can still make important causal contributions to research. Even one- and two-year-olds can be part of the causal process that identifies normal values for brain scans or evaluates the efficacy of an experimental vaccine. Indeed, clinical research offers the rare context in which even very young children can make valuable contributions to important projects. Making such contributions can benefit children in at least three ways.

First, children who recognize at the time, or learn when they are older about the contributions they made, may benefit psychologically. They may experience pride and increased self-esteem. This possibility is supported by a study of adolescents who had participated in net-risk research. It found that 68.9% felt they were making an important contribution and 80.8% felt proud to be doing so.¹⁶

Second, making these contributions may have a beneficial impact on the children’s lives. They may come to regard these contributions as an important part of their identity and be inspired to make further valuable contributions.

To appreciate the third type of potential benefit, consider why it can make sense for individuals to feel proud of causal contributions they made as young children. Presumably, it makes sense to feel personal pride only with respect to contributions that one has in fact made. This is highlighted by the previously described study; the reason that the respondents felt proud is because of the contributions they made to valuable research. This points to the possibility of a third type of benefit. Specifically, a life that includes contributing causally to valuable projects, even as a young child, is a better life for the individual compared to the same life absent that contribution. Simply put, it can be beneficial for the children themselves to make these contributions.

To see this, consider which of two lives is better for the individual in question: 1) a good and decent life or 2) the same life with the addition of contributing as a child to a research program that leads to the development of an intervention that benefits many other children. I would prefer to have the second life rather than the first. This reaction suggests that the second life is better for the person whose life it is. This suggestion points to the existence of what might be called “contribution” benefits. Briefly, the idea here is that we do not benefit only from having good experiences and avoiding bad ones in our lives. We can also benefit personally by having a better life overall. And whether we have a better or worse life overall depends in part on what we do and the impact that our lives have on others.

Contributing to socially valuable projects as a competent adult can advance our interests significantly; in short, it can even be heroic. This suggests that the possibility of contributing to valuable projects as an adult can justify potentially significant risks. In contrast, contributing to valuable projects before we understand them and willingly decide to participate promotes our interests to a limited extent only. In addition, the possibility of benefiting in the first two ways is uncertain, especially with respect to individuals who participate in net-risk research as very young children. In that case, the children need to be informed about contributions they made years in the past, and, as a result, they need to experience some psychological benefit or embrace the contributions as an important part of their lives.

This analysis suggests that children of all ages can benefit from participating in net-risk research. It also suggests that these benefits will be relatively modest, at least for younger children. While this may be a problem for anyone who wants to justify high-net-risk pediatric research, it represents a virtue for the purposes of satisfying the two conditions on an acceptable justification for net-risk pediatric research. The potential for modest benefits offers the potential to justify exposing children to some risks. And the fact that these benefits are limited suggests that this proposal precludes exposing children to high risks, even when the research is socially very important.

The Significance of Negative and Valueless Studies

Some net-risk pediatric research studies fail to yield valuable data: they fail to recruit enough participants; they fail to identify a safe dose for children; they fail to find benefit from adding an experimental intervention to an existing treatment. Critics have argued that the existence of these studies undermines the present proposal that enrolling children in net-risk research can be justified on the grounds that it offers them the opportunity to contribute to valuable projects.

To evaluate this objection, it is important to distinguish between negative studies and valueless studies. Whether a study is positive or negative depends on the results relative to the goals of the study. If the goal is to evaluate whether an experimental drug is effective for medulloblastoma, a negative study is one that finds that the drug is not effective. In contrast, whether a study is valuable or valueless depends on whether the information collected contributes to the ultimate development of an approach that improves health. Because the number of theoretically possible treatments for any given condition significantly exceeds the number of effective ones, negative studies can be valuable by helping researchers to narrow the field and, ultimately, identify an effective treatment.

For example, a study that finds that an experimental drug is not effective for medulloblastoma when given as a single agent may be the prelude to a study that finds it is effective when given in combination with other drugs. Participating in this study can involve contributing to finding an effective treatment for a devastating condition. This example suggests that children who participate in trials that turn out to be negative can nonetheless make valuable contributions to important projects. And this conclusion suggests that the challenge for the present justification comes, not from negative studies, but from valueless studies. In these cases, the risks to which pediatric participants are exposed cannot be justified by the fact that they contributed to a valuable project. With this in mind, it has been argued that the existence of valueless studies undermines the present justification for net-risk pediatric research: “it is in the nature of research, and the ignorance that prompts its need, that we cannot assign any probability—hope does not count—to whether the particular research project will be successful.”¹⁷

This objection can be understood as follows. Given the existence of some valueless studies, the chances that a given net-risk pediatric study will turn out to be valuable are unknown at the time it is initiated. Hence, estimating the chances that enrollment in the study offers children the opportunity to contribute to a valuable project is not possible. And, if that is right, it seems that the present proposal cannot justify exposing children to even minimal research risks. It seems to follow that the mere possibility that participating children will contribute to a valuable project is “too thin a basis for an ethical justification.”¹⁸ If this conclusion is correct, it suggests that IRBs do not currently have a justification for approving net-risk pediatric research.

Assessing the Concern

Clinical research is very expensive and time consuming. Before committing the resources to conduct it, funders require some reason to think that the intervention in question might be effective. In some cases, funders have positive data that an intervention might be effective in children. Imagine that a medication for type 1 diabetes has been given to thousands of adults with only rare and mild side effects, leading investigators to evaluate it in children. To begin, the investigators propose a phase 1 study in ten children who will receive a single, low dose of the medication. The present criticism points out that there is some chance this line of research will turn out to be valueless. Perhaps this study will recruit only three children and the general line of research will be abandoned before it yields any valuable data.

In that case, the criticism is correct: the risks that the three enrolled children faced turn out not to be justified by their contributing to a valuable project. However, it does not follow that enrolling these children was ethically problematic. That depends on what is known at the time of enrollment about the risks and the potential to contribute to a valuable project. Type 1 diabetes is a significant health problem in children, and identifying better treatments could benefit millions of patients. Hence, if the risks are sufficiently low, the potential to contribute to a valuable project can justify them, even granting that there is some chance the study will turn out to be valueless and the participating children will not realize any contribution benefits.

In other cases, investigators might have only theoretical reason to think that an intervention is effective. For example, there might be anecdotal reports that a treatment helped some patients, or the mechanism of action might provide some reason to think that it will help patients with a given condition. The fact that the support is only theoretical means there is less reason to think the intervention will turn out to be effective compared to cases in which there is empirical data that it is effective. Hence, in this case, there is a lower chance that participating children will contribute to a valuable research program. Yet, if the condition in question is important enough, and the benefits to others would be significant, it seems that even a theoretical chance of benefit can justify some risks.

To see this, consider an analogous clinical case in which a child with a significant illness has attempted all standard treatment options, at which point her doctor proposes to try an innovative therapy based on theoretical evidence that it might be effective. Presumably, even a theoretical possibility of realizing clinical benefit can justify exposing children to some risks. By the same logic, it seems that the theoretical possibility of making a valuable contribution to a very important research program can justify exposing children to very low risks, such as the risks of a few research blood draws or MRIs. To avoid this conclusion, critics will need to show that, although a theoretical chance of realizing clinical benefits can justify exposing children to some risks, a theoretical chance of realizing contribution benefits cannot justify exposing children to even very low risks.

Valuable but Harmful Studies

A different criticism focuses on negative studies in which a potentially promising intervention turns out to be harmful. For example, one commentator, Binik, describes a study of hypothermia therapy.¹⁹ At the time of the study, hypothermia therapy had shown significant improvement in survival and neurological outcomes in rodent models. To assess the safety and efficacy of hypothermia therapy in children with brain injuries, researchers enrolled 255 children in a randomized controlled trial. The trial found that hypothermia therapy does not improve neurological outcomes and may increase mortality.

Binik argues that, even though participation in this trial involved contributing to a valuable study and, thereby, offered some contribution benefits, those benefits are outweighed by the significant risks and, in some cases, serious harms that participants experienced.²⁰ This conclusion seems right and reinforces the point made earlier that contribution benefits cannot justify exposing children to high risks. However, the present justification does not make that claim. Instead, it proposes that the potential for children to realize contribution benefits can justify low risks.

Critics might respond that it is impossible to know at the time of enrollment whether pediatric participants will experience significant harms. This is true, and, on this basis, some parents may prefer not to enroll their children in any net-risk research. At the same time, the possibility that a given net-risk trial may turn out to be excessively risky does not imply that it was unethical to enroll children in the first place. Imagine some parents who enrolled their children in the hypothermia trial because it offered the children a chance for important clinical benefit. The fact that the treatment turned out to be ineffective and possibly harmful may lead the parents to regret this decision. But it does not follow that it was ethically problematic for the parents to enroll their children in the study. With respect to both potential clinical benefits and potential contribution benefits, the ethics of enrolling children depends on whether the potential benefits justify the risks at the time of enrollment.

Implications

The proposed justification for net-risk pediatric research traces to the potential for participating children to contribute to socially valuable studies. This suggests that it can be unethical to enroll children in net-risk research that has no or trivial social value. Hence, it is important to ensure that children are exposed to net-risk research interventions and enrolled in net-risk studies only when they have the potential to collect socially valuable information. Currently, the review of net-risk pediatric research tends to focus on ensuring that the risks are sufficiently low. The present analysis suggests it is also critical for IRBs to ensure that these studies have the potential to collect valuable data.

Next, a study that is terminated early for insufficient recruitment may provide no information on whether the intervention being tested was effective. Such a study may have no value, in which case the exposure of children to net research risks in the study is unjustified. If enough pediatric studies are stopped due to insufficient recruitment, this could undermine the ethical justification for net-risk pediatric research in general. Therefore, to justify pediatric research, it is ethically, as well as scientifically, important to minimize the number of studies that fail due to problems with recruitment. This analysis thus has what might seem a surprising conclusion: to promote the ethics of pediatric research, it is important for IRBs, investigators, and other stakeholders to look beyond risks, potential benefits, informed consent, and subject selection. These stakeholders also need to attend carefully to the plans to ensure that a sufficient number of subjects are enrolled and retained.

Finally, to the extent that the present justification is correct, it supports a modification to current definitions of minimal risk in pediatric research. Most guidelines define minimal risks based on the risks children ordinarily encounter in daily life. As noted previously, the U.S. federal guidelines define minimal risks as “not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.”²¹ Similarly, India’s guidelines categorize research as minimal risk when the “[p]robability of harm or discomfort anticipated in the research is not greater than that ordinarily encountered in routine daily life.”²² The problem with these definitions is that even average, ordinary children encounter significant risks in daily life, such as those associated with snow skiing and playing football.

These risks are considered acceptable because it is assumed that the potential benefits that children realize from participating in these activities justify the risks. Clearly, the same level of risks would not be acceptable in the context of research that does not offer the potential for clinical benefit. This suggests that current definitions of minimal risk in pediatric research are too broad. The present proposal suggests one possible way to address this concern. Rather than define minimal risks in terms of all the risks that children ordinarily encounter in daily life, the definition of minimal risk could be limited to the risks that children encounter in the context of activities that are designed to benefit others, such as charitable activities.²³