One of the basic standards that must be adhered to in drug research in pediatric populations is that all studies must be carried out according to an appropriate scientific design. A poorly designed experiment or one in which there is no potential for a direct benefit to the participant or to society (indirect benefit) is unethical because there is no chance of obtaining valuable scientific knowledge from it, while at the same time the investigator exposes experimental participants to needless risk or, at the very least, inconvenience.

A second basic standard is that the investigator must be both competent and ethical. The competence and ethical nature of the investigator are the most important safeguards for the protection of the interest and well-being of the child participant. The investigator who wishes to conduct studies utilizing human participants who are in the pediatric age groups must not only be knowledgeable and well trained, but also must understand the feelings and attitudes of the parties involved in the research and be especially sensitive to the special needs and fears of young children.

A third standard requires that the investigator must be fully aware of the need for any research plan to include a consideration of the concept of distributive justice and that
this be adhered to in its execution. Thus, infants and children should not be exposed to unwarranted risks, no matter how minimal, for the convenience of the investigator, and no subgroup of children (e.g., racial, ethnic, and socioeconomic) should bear a greater share of the research burden than others, except as dictated by the clinical and scientific requirements of the investigation itself or of the disease process under study.

A study conducted to evaluate appropriate dosages and/or effects for pediatric patients who require a specific form of therapy should therefore include patients from as wide a segment of society as is practical, whereas a study to evaluate dosages and/or effects of agents used to treat specific diseases should involve individual subjects in reasonable proportion to the risk or incidence of that disease in their subgroup of society.

Legal restraints imposed by the centuries-old common-law standard, based on the Magna Carta, that permission to act on a minor, or his or her property, can only be given when the action can be construed as being to the minor’s own benefit complicated the development of guidelines and standards for many years. However, since the rule was designed in the 13th century to solve a specific political problem of the times, thinking on the issue has undergone significant changes in modern times. Indeed, societal attitudes toward research on children have moved to the current discussion concerning how to define “harm” in a research situation as distinct from “discomfort” or “mere inconvenience” in such a setting. Since social forces lead to changes in public policy, a clearer set of guidelines has been developed to guide investigators, as the legal standards of professional behavior in this area have changed.

We believe that these guidelines, when properly interpreted and modified, can be utilized to carry out necessary studies so that advances can be made for pediatric patients in an ethical manner. Society must continue to afford maximal protection to all individuals (especially to those individuals who are not capable of protecting themselves) while maintaining systems that are seen as equitable by all groups in the society.

The National Commission for the Protection of Human Subjects was mandated by the Congress in 1974. That body debated whether to use the terms “therapeutic” and “nontherapeutic” to describe forms of research in human participants. In the end, the Commission chose not to use these terms but rather to refer to the two classes of research as “research that holds out the prospect of direct benefit for the individual subjects” and its opposite (4). The Canadian Medical Research Council’s Working Group on the matter also decided against the older terminology because “therapeutic” research can be confounded with treatment or care (5). They were rightly concerned that potential participants in experimentation or their parents or guardians might be misled into thinking that there would be a greater probability of benefit in “therapeutic” experiments than would actually be the case.

The AAP guidelines and the DHHS regulations for research on children distinguish between experiments where there is some prospect of direct benefits to the participants themselves (sometimes called therapeutic experiments) and those in which there is not (sometimes called nontherapeutic experiments). They also distinguish between research that does not involve greater than minimal risk to individuals and research that does. We could distinguish, further, research that involves subjects competent to give free and informed consent and research involving subjects not competent to give such consent. Using these three different pairs of distinctions, we can construct the following matrix:

Minimal risk is risk where “the probability and magnitude of harm or discomfort anticipated in the research is not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations” (6). The AAP understands minimal risk to be “a level of risk similar to the risk encountered in the child’s usual daily activity” (3). OHRP defines these risks in 45 CFR 46 in the same way (CFRs 46.102). However, if harm or discomfort is relativized to the individual child as suggested in this definition, it could lead some to conclude that it is permissible to subject children to research which will involve very painful procedures when those children are already undergoing a painful treatment regimen. However, there is a moral basis for rejecting this conclusion and interpreting “minimal risk” in a general way. It is not morally acceptable to put afflicted children at an even greater disadvantage than healthy children in order to benefit other children. That is, it is never morally acceptable to enroll a child in a study that involves significant pain when there is no prospect of direct benefit to that child, even when that child is already subjected to painful treatments/procedures as a part of a therapeutic regimen.

Sometimes it is permissible to expose afflicted children to substantially more harm than most children are exposed to, but this will either be because there is a good prospect that they will directly benefit from being so exposed or there is the potential for obtaining information that cannot be acquired in any other manner and the information is likely to lead to useful information concerning the disorder or condition being investigated. For example, we believe that it can be permissible to take an extra bone marrow aspirate from an adolescent with cancer if the adolescent is capable of assenting, or consenting, to being exposed to more than minimal risk (3). If it were a younger child, one without this capability, we believe that it would not be permissible in an experiment with no prospect of directly benefiting that child, for it would violate the basic moral obligation to protect the defenseless. However, such a procedure could be permissible in a younger child if the research has the prospect of benefiting that child. [We are aware that some institutional review boards (IRBs) now also permit collection of blood samples
from young children for the sole purpose of developing a database.]

Many different grounds have been offered for and against the limit of minimal risk in experiments with no prospect of direct benefit. Ramsey argued that allowing children to be placed at minimal risk where that will not likely benefit them … violates a requirement of respect for persons (7). The Belmont Report says that “respect for persons divides into two separate moral requirements: the requirement to acknowledge autonomy and the requirement to protect those with diminished autonomy.” The Report goes on to say that respect for autonomy requires giving “weight to autonomous persons’ considered opinions,” and this is what founds the requirement of obtaining free and informed consent. According to the Belmont Report, questions about protecting those with diminished autonomy concern “the risks of harm and the likelihood of benefits.” Because very young children are not autonomous individuals, some have reasonably thought that respect for autonomy is not relevant to actions involving them. However, others have thought that what is relevant is hypothetical, or counterfactual, consent, that is, what the young child would want with respect to treatment or becoming a subject of some experiment, if he or she were autonomous (8). For instance, McCormick wrote “… that it is permissible for children to be subjected to minimal risk because they would want to help others, if they could consent, because they ought to aid others when providing that aid requires little of them” (9). In other words, if children could rationally consent, they would want to do what they morally should do. For him, and for others, hypothetical consent (i.e., what the person would want if capable of rationally consenting) is enough; actual consent is not required.

Ackerman argued against both the requirements of actual and hypothetical consent (10). He approached the question about permissible imposition of risk in nontherapeutic experiments (see categories 1c, 1d, 2c, and 2d in the matrix given earlier) by focusing on what sorts of risk or harm parents are permitted to impose on their children to enhance their personal, physical, social, and moral development and also to enhance the interests of other people (10). He argued that the risk of harm that a child encounters in everyday life may be greater than the risk that we can intentionally impose on a child. Children normally engage in certain dangerous recreational activities, for example, skateboarding, swimming in polluted or unsafe lakes, and hopping on and off railroad cars, but it would not be permissible to impose such risks on them if they are not needed to benefit them. He suggested that “minimal risk” be taken to mean “no more than that to which it is appropriate to expose a child for educational purposes in the family situation” (10).

The DHHS regulations (cited in the AAP guidelines of 1995) take the position that all cases of at least no greater than minimal risk involving children who are unable to give their free and informed consent, that is, 2a and 2c, should be treated alike. According to the regulations, appropriate permission and assent is nonetheless required in all such cases. Permission is always required of parents or guardians for all children except for those who are considered to be emancipated or mature minors, as noted later. These regulations also require appropriate assent from children 7 years of age or older, where assent is defined as “active agreement.” The child’s assent must be free and informed, just as consent must be when it is required. The difference between assent and consent in this context is that a lower level of understanding of the risks and benefits involved may suffice for assent than for informed consent.

Ackerman questioned whether assent is required for those between 7 and around 12 to 14 years of age. Starting with the idea that the primary obligation of parents is to the moral, social, personal, and physical development of their child, Ackerman argued that we are not required to go along with a child’s wishes, especially if they are founded on interests that are still developing. However, he did think we should not go against the wishes of children in that age bracket, that is, it is not permissible to include them in nontherapeutic experiments (categories 2c and 2d) against their wishes. This might be called the requirement of no dissent. Ackerman was rightly concerned that involving a child between 7 and 14 years of age in an experiment in one of these categories against the child’s wishes may create so much fear and anxiety (and we would add mistrust) that it could not be justified.

We agree with Ackerman that assent should not be required for those aged between 7 and 14 years, but, instead, that it be required that there is no informed dissent by the child. This clearly deviates from the DHHS regulations, which require assent, where assent is explicitly distinguished from “failure to object” (11). We also agree with his views on subjecting children to more than minimal risk in “nontherapeutic” experiments. He argued that children between 7 and 14 years of age cannot be exposed to more than minimal risk in a research situation, for if they were, given his understanding of minimal risk, they would be exposed to more risk in this situation than it is permissible for parents to expose their children to in nonexperimental situations.

Furthermore, Ackerman argued that assent is required from preadolescent youths aged 12 to 14 years. Their judgment is sufficiently developed that parents and others must pay attention to their wishes. He indicated that they are so developed that they can even assent to a risk that is slightly more than minimal to benefit others by participating in a nontherapeutic experiment. We agree with Ackerman’s views on nontherapeutic experiments, namely that (a) assent is not required for children aged between 7 and 12 years (though lack of dissent is) and no more than minimal risk may be imposed on them and (b) assent is required for preadolescents aged 12 to 14 years and we can permissibly involve them in nontherapeutic experiments involving slightly more than minimal risk.

It cannot be permissible to involve a child in an experiment without her or his free and informed consent if that child is capable of giving such consent. However, such consent is not by itself sufficient to make the inclusion of that child permissible. In some states, emancipated and mature minors are sometimes considered to be capable of giving their free and informed consent to participate in an experiment. But, because there is a high risk that they will be exploited, we think the best policy regarding emancipated or mature minors is that they should not be permitted to be participants in nontherapeutic experiments, that is,
experiments where there is no prospect of direct benefit to them and where they are subject to more than minimal risk, that is, category 1d. Thus, our recommendation is more stringent than the AAP guidelines, which permit emancipated and mature minors to participate in nontherapeutic experiments if the knowledge sought cannot be obtained by using another group of children where parental consent is obtainable (3).

When it comes to research with the prospect of direct benefit to the research participants (categories 1a, 1b, 2a, and 2b), the DHHS requires that “the risk is justified by the anticipated benefits to the subjects” and that there is no other available alternative with a more favorable risk–benefit ratio (3,12). Of course, the problem is that in situations where experiments are in order, the relevant risks and benefits will largely be unknown because one cannot extrapolate directly from results of animal or adult human studies.