Loretta M. Kopelman, PhD

A difficult ethical and social problem exists about how to promote the best interests of children as a group through research, while protecting the rights and welfare of individual research subjects (Kopelman, 2000). While children and adults often have the same diseases and are treated with the same interventions, few modalities have been well tested on pediatric populations and a majority of marketed drugs are unlabeled for children's use (AAP, 1995). Given this lack of information about safety and efficacy, clinicians face a dilemma. If they use untested interventions, they may endanger their patients. Many neonates had died before careful studies showed that their immature livers could not tolerate the commonly used antibiotic Chloramphenical. Yet if clinicians only use tested interventions, they are severely limited in their treatment options (FDA, 1998). Children with schizophrenia, for example, are routinely undertreated because many interventions used for adults have not been tested on children (Findling et al., 1996 and Quintana and Keshavan, 1995).

The difficulty is that until research is conducted, investigators do not know if treatments will be safe or effective, and the testing may harm some subjects. Unlike adults, or even mature or emancipated minors, younger children lack the capacity to decide whether they want to run risks of harm for uncertain benefits or the common good (Buchanan and Brock, 1989; Kopelman, 1995). Thus, someone else must decide for them. As children get older, they generally have increasingly greater capacity to participate in such decisions. This paper, however, focuses upon those who lack capacity to give consent, which is where the problem arises in its starkest form.

In what follows, I will consider two kinds of regulatory solutions to this problem. One is a non-consequentialist response and seeks to protect children’s rights and welfare by drawing the line at certain sorts of studies no matter what the anticipated social benefit. A second response draws no such line, but requires members of review boards such as Institutional Review Boards (IRBs) or Research Ethics Committees (RECs) to determine if proposed research has an acceptable balance of likely benefits and harms to subjects and to society. This“ balancing” or consequentialist strategy may be found in The Belmont Report (1979), the U.S. Federal Regulations (45 CFR 46), The Council for the International Organizations of Medical Science (CIOMS, 1993), and many other national and international guidelines. Unfortunately, this popular approach offers too little guidance about how to do this balancing, what constitutes risks of harm, or what benefits to the person or society are salient. Consequently, people sometimes reach radically different judgments. Some suggestions for improvement are made in the final section.

NON-CONSEQUENTIALIST APPROACHES
One way to deal with the problem is to protect children’s rights and welfare by establishing thresholds, drawing a line at certain sorts of pediatric studies. Certain studies must not be authorized, according to such non-consequentialist standards, no matter what the anticipated social benefits. Two very influential codes, The Nuremberg Code and pre-2000 versions of The Declaration of Helsinki, take this approach.

(1) The Nuremberg Code
The Nuremberg Code (1949), the first international research code, states that the “ voluntary consent of the human subject is absolutely essential.” Drawing the line here would halt research with children or others lacking capacity to give informed consent (Kopelman, 1995). Yet members of these groups have unique medical problems and drug reactions, so studies done with competent adult subjects may not be applicable to them. Competent normal adults cannot test drugs intended for use by premature infants or serve as subjects in studies about the efficacy of treatments for childhood schizophrenia. To test the safety and efficacy of many standard, innovative, or investigational treatments for distinctive groups, some members of the groups must be subjects. Participation in controlled research, moreover, may not always violate the rights and welfare of persons who lack capacity to give informed consent, if it offers opportunities to get innovative treatments or if the risks are low. Excluding all persons who are unable to give consent means that they cannot participate in important and potentially beneficial or safe studies that advance our knowledge for them. This neither honors nor promotes their welfare. Rather, adopting such a restrictive research policy for children would ignore their needs; it would also fail to honor their rights to equal consideration by limiting their access to safe and efficacious treatment modalities.

(2) The pre-2000 versions of the Declaration of Helsinki
The pre-2000 Declaration of Helsinki (1964 to 1996) distinguishes between acceptable and unacceptable research differently. It holds that persons who lack the capacity to volunteer or give informed consent may be enrolled only in clinical or therapeutic research studies (seeking generalizable knowledge but intending to provide medically beneficial and acceptable therapy for the individual). They cannot be enrolled in non-therapeutic biomedical research (studies seeking generalizable knowledge but not intended as therapy to benefit the individual directly). Therapy is designed to benefit the person, so this policy permits only therapeutic biomedical research for people who lack the capacity to give informed consent. This policy has several problems.

First, classifying studies as either therapeutic or non-therapeutic may be misleading or arbitrary, or create false expectations (Kopelman, 1995). Therapeutic studies may have burdensome nontherapeutic features such as additional costs, tests, hospitalizations, or visits to clinicians. Second, some nontherapeutic studies provide important information involving little risk, so it seems unjustifiable to forbid them. Epidemiological investigations, for example, are needed to learn about children's normal growth and development and what diseases are most common in certain age groups. Prohibiting such studies, rather than protecting them, may deny them the benefits of evidence-based medicine for arbitrary and problematic reasons.

Deontological or line drawing approaches set some non-negotiable threshold' to limit regarding when nonconsenting subjects can participate in studies. Such inflexibility, however, can be a weakness if it prohibits studies that reasonable and informed people of good will agree ought to be done. As we saw, the two most prominent line drawing approaches, the Nuremberg Code and Pre-2000 Declaration of Helsinki, have this flaw. It is possible that other deontological responses to this problem might draw a more defensible upper limit of risks of harm, but most rules now take a different approach.

CONSEQUENTIALIST APPROACHES
Many countries such as the U.S., Canada, the United Kingdom, Norway, and South Africa, as well as international organizations, such as CIOMS (1993), employ another policy option in their codes for research involving children (Freedman et al., 1993; Kopelman, 1995). They require studies to be reviewed for many things including proper ranking of potential benefits and harms by IRBs and RECs. Approval is based on finding that subjects have been selected fairly and that the risks to subjects are minimized and reasonable in relation to anticipated benefits of the study to the subject or to society (45 CFR 46). Moreover, adequate provisions must also be made for the safety and confidentiality of the subjects and for parental permission and, when appropriate, the child's assent. Children's refusals are not binding where their parents and doctors judge it is in the best interests of the children to participate, as in studies where children may obtain a scarce resource to treat a deadly disease. The thresholds that exist are vague and negotiable, or so I argue.

I will focus on the U.S. Federal Regulations, although analogous issues arise in the other national and international guidelines taking this consequentialist approach. There are, however some differences worth noting. Other regulations lack the 407 approval mechanism, permitting federal approval for higher risk studies in some circumstances. No upper level of risk is given. The 2000 version of the Declaration of Helsinki gives no upper threshold either, but lacks the risk levels provided in the U.S., leaving it up to review boards to rank potential benefits and harms.

In using the likely harms to benefit calculation, the U.S. regulations specify four categories of research with children (45 CRF 46.404-7). As risks increase, the regulations require increasingly more rigorous documentation of appropriate parental consent, children's assent, direct benefits to each child, or benefits to other children with similar conditions. If children are healthy or the research does not hold out direct benefit to them, at least as great as available alternatives, then the risks must be no more than minimal. If children have a disorder or condition then, even if it does not benefit them, research may be approved having no more than a minor increase over minimal risk. Studies having more risk cannot be approved by IRBs unless they hold out the prospect of direct benefit to the children at least as great as available alternatives, as in therapy. In the U.S., IRBs are permitted to seek federal approval for higher risk studies.

(1) Vague thresholds
Unfortunately, no definition is given of the crucial upper limit of risk that local IRBs can approve for children's research, that of “a minor increase over minimal risk.” Also the pivotal definition of "minimal risk" is unsatisfactory (Kopelman, 1981, 1995). According to the U.S. regulations:

“Minimal risk means that the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical and psychological examinations or tests” (45 CFR 46.102i).

According to this definition, something has the probability and magnitude of a minimal research risk of harm if it is comparable to the risks in either or both parts of this disjunctive phrase. This definition is very common in international and national codes (Kopelman, 2003).

(a) The first part of the definition of“ minimal risk” might mean we decide what is a minimal risk based upon the ordinary risks encountered in the particular person’s daily life or by any of us; this is called the relative standard in the literature.

The relative interpretation of “everyday risk” standard judges research risk relative to the individual’s daily life. It is widely regarded to be untenable. Some of us, such as combatants, police, firefighters or children in war zones or in institutions, face high risks every day. A policy that sets minimal risk thresholds based on the risks encountered in the everyday lives of different populations is socially destabilizing and unfair. Children in the widely criticized Willowbrook Studies were given hepatitis on the grounds that this was a minimal risk for them, since hepatitis was endemic to the institution (Kopelman, 1981). This reasoning is flawed. It is morally unacceptable to locate children in places with increased risks of harm, and then use that increased risk to redefine "minimal risk" or "justify" higher research risks for them. One standard prevents singling out a few people with higher daily risks in order to "excuse" higher research risks as minimal.

The absolute interpretation of “everyday risk” standard judges research relative to the sorts of risks all of us encounter. It has epistemological and moral problems. How do we calculate our “everyday” background risks as varied as car accidents, terrorist attacks, plane crashes, floods, and threats of nuclear and conventional wars? Moreover, risks of harm even within one nation are not homogeneous, varying from community to community. This interpretation assumes that first, we know the probability and magnitude of harm or discomfort normally encountered in daily life, and second, this information should serve as a morally justified norm to estimate research risks as minimal. Each assumption is challenged.

(b) The “routine-examination” standard, or the second part of the definition of “minimal risk,” says that the minimal risk of harm or discomfort should be comparable to the probability and magnitude of those ordinarily encountered in routine clinical examinations. Here too, a relative interpretation seems untenable (Kopelman, 2003). Children facing death or disability may have painful experiences far removed from those normally encountered by most people in their daily lives. Their daily experiences should not be used to "justify" higher risks as minimal for them, or else the door is open to treating some groups in a discriminatory way and exploiting them.

An absolute interpretation of the “routine-examination” standard gives clear direction about what physical risks are minimal, since it means research risks are minimal when they are like those in routine examinations we all encounter; and those can only be routine examinations for healthy persons. There is a problem, however, for the absolute interpretation of the “ routine-examination” standard. Clinicians " ordinarily encounter" highly sensitive information in routine examinations and testing, including data about family assaults, substance abuse, sexual preference. Moreover, they make sensitive diagnoses such as psychiatric disorders or late-onset genetic diseases. Such information can affect how people are viewed, including whether they will be able to gain jobs or
buy insurance. Thus, this part of the definition lacks guidance about how to assess psychosocial risks. Some genetic research involving testing has low physical risks, but high psychosocial risks. Clinicians are bound by the primary duty to act in patients’ best interest, all things considered. But investigators also have duties to gather the best information.

(c) The pivotal threshold of "a minor increase over minimal risk” is undefined; it is the upper limit of risks IRBs can approve for pediatric populations that do not hold out the prospects of direct benefits for the subjects as in therapy at least as great as available options. Whatever uncertainties exist about what constitutes a minimal risk of harm are compounded when reviewers are asked to determine what constitutes a minor increase over it. Not surprisingly, since they are given a free range of interpretation, they disagree and have done so for two decades (Kopelman, 2000; 2002). Studies show considerable differences among pediatric experts, in both treatment and research settings, about how to assess the risk of such procedures as venipuncture, arterial puncture, and gastric and intestinal intubation (Janofsky and Starfield, 1981). Investigators and others have concluded that better standards of risk assessment in children's research need to be formulated (Lascari, 1981). More recently, this lack of consensus and regulatory guidance within the research community was apparent in testimony before the National Bioethics Advisory Board (NBAC, 1998). They all agreed, however, that the definition of “minimal risk” was vague, and that “a minor increase over minimal risk” was vaguer.

(2) Four moral problems
There are four unresolved moral issues with the definition and threshold of “minimal risk” and “a minor increase over minimal risk” defined in terms of “everyday” risks. First, since probability ranges from none to certainty, and magnitude of harm from trivial to horrific, how should we establish and use thresholds regarding both the probability and magnitude of harm to identify everyday risks? The assessment is complicated since some extremely low risks of substantial harms might be approved in some circumstances, while high risks of moderate harm might not.

Second, why should normally encountered, everyday risks be taken as a morally justifiable guide for estimating that comparable research risk as minimal? Some everyday hazards are trivial, and some catastrophic, some rare and some common, so what does knowing whether a hazard is normally encountered on a daily basis have to do with estimating minimal research risk? Clearly, some people face death and disabilities in the normal routine of their daily lives, but this is not minimal.

Third, given the different hazards in different countries and communities, what locale(s) should be regarded as appropriate to assess everyday risks? In some war zones, death and starvation are every day hazards, especially for young children. Some diseases, such as AIDS threaten entire populations.

Fourth, if this is a useful and clear standard why has there been sustained disagreement over whether common procedures should be viewed as minimal risk, a minor increase over minimal risk, or greater? (Janofsky and Starfield, 1981; Lascari, 1981, NBAC, 1998) Asking whether the potential harms are like those in the child's everyday life is pertinent, but not decisive.

(3) Ranking potential harms and benefits to society

Another problem is that RECs and IRBs are given little direction about how to understand social utility in evaluating and balancing burdens and benefits. Something is not useful per se, but useful to some end. Research may help the participating subjects gain services, aid patients collectively by improving their therapies, build careers, or advance commercial interests. Reviewers, however, are instructed to balance the social utility of research against the potential benefits, inconveniences, or harms of the study for the subjects. How should they do this ranking or balancing? Some studies are more important for helping many people get well who are not very sick, others for helping a few who are very sick get well, still others for providing long term care rather than cure, and so on. Different views or theories of justice balance potential harms and benefits differently. With little guidance about how to rank or balance them fairly, IRBs and RECs must rely on their own judgments.

The consequentialist or “balancing” approaches taken by the U.S. Federal Regulations and other codes, then, are vague. Of course, many laws are unclear initially, but become clarified in use. These ambiguities about how to assess potential harms and benefits for pediatric and other populations, however, remain after several decades. There is an increasingly greater need to address the issues of risk assessment. For example, reviewers need direction about how to evaluate psychosocial risks associated with leftover tissue samples, the genome project, electronic chart reviews, and other recent innovations. With key terms undefined or defined poorly, investigators and reviewers must follow their intuitions and consciences, which may be arbitrary, inconsistent, or not good enough for some members of the public.

(4) Improving the consequentialist or balancing approaches

The consequentialist approaches could be improved in at least three ways: a) offer better standards for risk assessment, b) reject crude utilitarianism, c) identify and justify normative or moral judgments and d) acknowledge extra-regulatory thresholds and deontological duties.

a) Offer better standards for risk assessment. As noted, the regulations offer poor guidance about risk assessment, and this was apparent in the disagreement among knowledgeable people of good will about assessing the risk to children of undergoing multiple placebo injections or lumbar punctures. Better definitions or examples of the risks of harm associated with many common medical procedures could help improve consequentialist approaches and assist review board members in making risk assessments. If we had clear and openly adopted examples of minimal risk, minor increase over minimal risk, and so on, such examples could serve as paradigms to make other decisions, and help answer the four moral questions posed above. Methodologically this approach presupposes that we can match disputed cases to morally unambiguous situations where we agree about what we ought to do (core cases, exemplars, or paradigms). That is, it assumes that we generally agree about what cases are morally unambiguous and when disputed situations are adequately similar to the core cases.

Adopting exemplars in an open and public way would bring differences to light, and resolving these disputes would help forge socially sanctioned paradigms. Investigations may assess potential harms to subjects and potential benefits differently from civil rights attorneys. Public discussions, however, could help address such different points of view. Is there a consensus about whether a hundred placebo injections or one or more lumbar punctures on a child is a minimal risk, a minor increase over a minimal risk, or higher? Clarifying publicly such paradigms or definitions shows the public that the research community wishes to specify these notions in an open way. Without public trust and a sense that those applying the regulations are competent and informed people of good will, however, the system, I believe, cannot work.

b) Reject crude utilitarianism. An underlying philosophical problem with the consequentialist methodology is: to what extent should people's rights and welfare be "balanced" with social utility? Crude utilitarian calculations permit decision-makers to assess the rightness or wrongness of particular acts simply in terms of what is best able to promote the general good. If notions about maximizing the good are not restricted, however, they could swamp individual rights or welfare. Suppose it maximizes the over all good to clandestinely and without subjects’ consent conduct studies that expose people to radiation or leave untreated their syphilis; or imagine it would maximize the good to learn about hepatitis by pressuring parents to "consent" to expose their children to this disease. On a crudely utilitarian view, authorizing such secret and coerced studies would be the right thing to do if it would maximize the net good, even marginally. Yet similar studies resulted in public and professional condemnation, and increased public scrutiny of investigators and their institutions; in addition there were doubts that such studies would maximize utility. This entails that the public and professional organizations should reject a crudely utilitarian approach to assessing utilities and determining the common good in research, indicating agreement with standard philosophical objections to this stance (Sober, 1990). (More sophisticated forms of utilitarianism would not compromise basic liberties.) The consequentialist methods of risk assessment in the U.S. Federal Regulations, then, should be sharply distinguished from a crudely utilitarian position. These criticisms may be avoided by a rule utilitarian moral view (assessing rightness or wrongness in terms of the rules best able to promote the good).

No upper level of risk exists in the U.S. Federal Regulations, but they are not freestanding sets of rules. Rather they exist in a web of other social and legal duties or expectations. Clinicians and investigators must adhere to legal and professional codes in taking care of patient-subjects. Parents have duties not to abuse or neglect their children. All are private citizens who must obey laws on assault and battery, and so on. Moreover, there are penalties if the responsible parties do not adhere to these rules or other relevant laws.

Investigators are subject to a variety of sanctions including loss of funding, peer pressure, or denial of publication if they do not adhere to social and legal rules of conduct. If members of IRBs or RECs fail to live up to their responsibilities there are consequences to them as well, including possible lawsuits for themselves or their institutions and loss of federal funding. In making their moral judgments about whether studies should be permitted, the needs and rights of subjects, as well as the public interest and common good, should be considered by members of IRBs or RECs in authorizing studies. Salient considerations include not just research rules, therefore, but also wider
social, legal, and professional norms.

c) Identify and justify normative or moral judgments. A consequentialist approach depends on many people making sound moral judgments, especially parents, investigators, and members of IRBs and RECs. People who are informed and reasonable must act with a good will to identify and weigh salient features to reach decisions about which studies ought to be authorized. They must decide how to protect subjects' rights and welfare, what utilities are important, what weights to give them, and what
constitutes the common good.

Although this approach requires weighing and balancing of values, we do this differently. A civil rights attorney who is a member of an IRB may see more dangers than other members about waiving subjects' rights to consent. The deeper philosophical problem is that it is extremely difficult to clarify and rank utilities in a way that does not rig or beg the question about the value of different alternatives. Acknowledging this problem has implications. First, to combat bias, wide representation on IRBs and RECs is important. For example, many such committees have community representatives, who should play an important role. Multi-disciplinary representation on review boards and panels also helps insure that, in addition to the research regulations, many professional codes of responsibility in many fields help guide what is in the best interest of the subjects and the public in authorizing studies.

Second, openly acknowledging that judgments made by IRB and REC members are moral judgments has important consequences. It means that their assessments can be tested against what is required of good moral reasoning (such as whether the problem has been framed reasonably, whether the best information is being used, whether consideration has been given to all available options; whether the appropriate weight has been given to relevant policies, rights, duties or values; and whether due consideration has been given to the weaknesses of the decision). Two other features of moral judgments also help to focus issues. Decision makers should consider how they would want to be treated in similar circumstances. This is a variation of the "golden rule" or test of universality; would you want your children to be in the study? A second test concerns whether the decision-makers would be sanguine if their decision were publicized widely. These two tests are not sufficient, but help determine if the judgment is morally justifiable

Third, members of IRBs and RECs ought to represent the public and act for the common good. The concept of acting in the public's interest requires that, in addition to notions about maximizing good, that virtuous people try to uphold institutions and a variety of policies relating to the common good (Weale, 1998). Good research policy and oversight must rely upon investigators, parents, and members of IRBs and RECs generally making good decisions about how to pursue the common good, while protecting subjects' rights and welfare.

d) Acknowledging existing thresholds and non-negotiable duties. The research regulations are not meant to override state or federal laws, and so exist within a framework of policies protecting children from abuse and neglect. Morally and legally, it is well established that parents and the state have duties to children that set non-negotiable thresholds (Kopelman, 2002). These help determine the appropriate level of research risk, even using a balancing approach. Parents who make imprudent or neglectful decisions may lose custody temporarily or permanently, and there are well-established social and legal thresholds for state intervention when children are placed at risk. There are also possible sanctions against institutions or investigators who cross those thresholds. Consequently, the consequentialist approach for assessment of research exists in a context of deontological or non-negotiable duties to children.

CONCLUSION
Moral and social responses to the problem of how to promote the best interests of children as a group through research, while protecting the best interests of individual subjects, take two general forms. First, some non-consequentialist approaches set thresholds, forbidding certain kinds of studies with children as subjects, whatever the social benefit. The Nuremberg Code and the pre-2000 Declaration of Helsinki draw different lines, but both do so in a way that is too restrictive for the best interest of children individually and collectively.

Second, The U.S. Federal Regulations have vague thresholds and set no crucial upper limit of risks of harm for pediatric studies. Rather, they favor consequentialist approaches, allowing studies when anticipated risks of harm to potential subjects are offset by likely benefits to the subjects or society The U.S. Federal Research Regulations, as well as a variety of national and international policies, permit RECs and IRBs to balance potential harms, inconveniences, and benefits, and authorize studies if the risks of harm to potential subjects are low enough and the social utility of the study appropriately high. Unfortunately, they offer no good criteria for identifying or measuring either risks of harm to individual subjects or benefits to society. In short, little direction is offered on how to “balance” potential harms or benefits to individual subjects or society. As a result, there is considerable disagreement over how to do this and growing concern about the quality of the oversight and monitoring function of local IRBs and RECs (Moreno et al., 1999; NBAC, 1999). Several ways to improve consequentialist or balancing approaches are suggested, including clarifying assessments of harms and benefits, rejecting crude utilitarian calculations, and identifying and justifying normative and moral judgments.

Balancing approaches, as represented in the U.S. Federal Regulations and other codes, should not be understood as reflecting a crude utilitarian moral theory, since they exist within a context of extra-regulatory thresholds and duties. Members of IRBs or RECs should assess whether studies should be approved in terms of the rules in the research regulations, as well as other legal and moral expectations. Clarifying these rules and goals, and wider representation on these committees, should help promote the public interest, the common good, and trust.

The quality of health care for children varies with the caliber of information and therapies available, which in turn varies with the quantity and quality of the research conducted on issues involving children. This is a source of the enduring moral and social problem: What sort of research should be permitted with children as subjects, since research promotes the welfare of children as a group, but may harm or not benefit particular children who are subjects?

Unfortunately, we currently lack a satisfactory solution to this difficult question.

BIBLIOGRAPHY

American Academy of Pediatrics. 1995. Committee on Drugs, Guidelines for the Ethical Conduct of Studies to Evaluate Drugs and Pediatric Populations. Pediatrics 95:2, 286-294.

Buchanan, A., and D. Brock. 1989. Deciding for Others: The Ethics of Surrogate Decision Making. New York: Cambridge University Press.

Council for International Organizations of Medical Science (CIOMS). 1993. International Ethical Guidelines for Biomedical Research Involving Human Subjects. Geneva, Switzerland: CIOMS.

Findling, R.L., Groevich S.J., Lopez, I. and Schulz, S.C. 1996. Antipsychotic Medications in Children and adolescents. Journal of Clinical Psychiatry 57 Supp 9, 9-23.

Freedman, B., Fuks, A., and Weijer, C. 1993. In Loco Parentis: Minimal Risk as an Ethical Threshold for Research Upon Children. Hastings Center Report 23 no. 2, 13-19.

Janofsky, J. and B. Starfield: 1981. Assessment of Risk and Research on Children. Journal of Pediatrics 98, 842-846. Kopelman, L.M. 1981. Estimating Risk in Human Research. Clinical Research 29, 1-8.

Kopelman, L.M. 1995. Research Policy/II. Risk and Vulnerable Groups. In W.T. Reich (ed.) Encyclopedia of Bioethics, Revised Edition, and New York: Simon & Schuster MacMillan, 4: 2291-2296.

Kopelman, L.M. 2000. Children as Research Subjects: A Dilemma. The Journal of Medicine and Philosophy, 25: 745-764. (This article is based upon this paper)

Kopelman, L.M. Pediatric Research Regulations Under Legal Scrutiny: Grimes Narrows Their Interpretation. Journal of Law, Medicine & Ethics, 30, 2002, pp. 38-49.

Kopelman, L.M. 2003. Minimal Risk as an International Ethical Standard in Research. The Journal of Medicine and Philosophy.

Lascari, A.D. 1981. Editorial: Risk of Research in Children. Journal of Pediatrics 98, 759-760.

Moreno, J. et al. 1999. Updating Protections for Human Subjects Involved in Research. Journal of the American Medical Association, 280 (22), 1951- 1958.

National Bioethics Advisory Board. 1998. Regulatory Understanding of Minimal Risk. Testimony before the Human Subject Subcommittee of the National Bioethics Advisory Commission Meetings, January 8, 1998 in Arlington, Virginia.

National Commission for the protection of Human Subjects. 1979. The Belmont Report. Washington: Department of Health, Education, and Welfare. Nuremberg Code. 1949. Nuremberg Military Tribunals, 181-182.

Quintana, H. and Keshavan, M. 1995. Case Study: Risperidone in Children and Adolescents with Schizophrenia. Journal of the American Academy of Child & Adolescent Psychiatry 34 (10), 1292-1296.

Sober, E. 1990. Core Questions in Philosophy. New York: MacMillan Publishing Company. U.S. Code of Federal Regulations. 1991. Title 45 CFR Part 46.

U.S. Department of Health and Human Services. 1983. Additional Protection for Children Involved as Subjects in Research. Federal Register 48 (46), 9814-9820.

U.S. Department of Health and Human Services Food and Drug Administration. 1998. Regulations Requiring Manufacturers to Assess the Safety and Effectiveness of New Drugs and Biological Products in Pediatric Patients. Federal Register, December 2, 1998. Vol. 68, number 321, pp. 66631-66672.

U.S. National Institutes of Health. 1992. Human Growth Hormone Protocol Review Committee. Report of the NIH Human Growth Hormone Protocol Review Committee. Bethesda, MD.

U.S. National Institutes of Health. 1998. NIH Policy and Guidelines on the Inclusion of Children as Participants in Research involving Human Subjects.

Weale, Albert. 1998. Public Interest. Routledge Encyclopedia of Philosophy, London and New York: Routledge, Vol. 7, pp.832-835.
World Medical Association. 1964. Declaration of Helsinki: Recommendations Guiding Medical Doctors and Biomedical Research involving Human Subjects. Adopted by the 18th World Medical Assembly, Helsinki, Finland and amended in 1975, 1983, 1989, 1996, 2000.