NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Committee on Clinical Research Involving Children; Field MJ, Behrman RE, editors. Ethical Conduct of Clinical Research Involving Children. Washington (DC): National Academies Press (US); 2004.

Cover of Ethical Conduct of Clinical Research Involving Children

Ethical Conduct of Clinical Research Involving Children.

Show details

5Understanding and Agreeing to Children's Participation in Clinical Research

My daughter will be nine years old and she needs some kind of input in what's going on with her… . She's presently in a study and I need for her to be able to understand what she's getting herself into… . She's at the point where she asks a lot of questions … which is good.

Andrell Vaughn, parent, 2003

The quote above shows a mother's concise, personal appreciation of the values underlying requirements that children be involved, when appropriate, in discussions and decisions about their participation in research. This mother recognizes her daughter's growing maturity, increasing curiosity, and developing moral right to be involved in choices about “what's going on” in the context of a close parent-child relationship.

When research involves children such as this 9-year-old, investigators and institutional review boards (IRBs) cannot rely on the conventional concept of informed consent, which applies to decisions about research participation made by those with the legal and intellectual capacity to make such choices in their own right. Children usually lack such capacity. Instead, legal authority to allow a child's participation in research rests with parents or guardians, who must provide their permission. In addition, with respect for children's emerging maturity and independence and consistent with federal regulations, investigators—when appropriate—seek to involve children in discussions about research and obtain their assent to participation.

Although debate continues about the ethical dimensions and boundaries of parental permission and child assent, especially when adolescents are involved, much attention today focuses on more practical questions about how to interpret and implement these concepts and honor their underlying ethical principles in research practice. For example, at what age should investigators, as a general rule, begin seeking a child's assent? What information and what role in decision making are appropriate for children at different stages of development? How should relationships between children and parents be approached?

This chapter considers two elements in the charge to the committee. The first relates to the written and oral process for obtaining assent and permission for research participation from children and their parents, guardians, or other legally authorized representatives.1 The second involves an examination of children's and parent's expectations and comprehension of the direct benefits and risks associated with a child's participation in research and, in particular, their understanding of the distinction between research and treatment. The discussion begins with an overview of the ethical principles and legal requirements for obtaining parents' and children's agreement to a child's participation in research, including circumstances in which minors can make decisions on their own behalf. Next is a review of the literature relevant to parents' and children's understanding of research participation, which also covers the general research literature on the development of children's cognitive and decisional capacities. The last sections of the chapter offer recommendations.

ETHICAL PRINCIPLES AND LEGAL REQUIREMENTS

The Ethics of Informed Consent, Permission, and Assent

For pediatric ethics, informed consent is more properly understood as a combination of informed parental permission and (when appropriate) the assent of the child.

Kodish, 2003b, p. 90

As described in Chapter 1, the ethical principle of respect for persons underlies the obligation of investigators to treat individuals as autonomous actors who must provide their informed and voluntary consent to participate in research. Requirements for parental permission serve the ethical obligation of investigators to respect and protect vulnerable individuals. Provisions for involving children in discussions about research participation and seeking their assent, when appropriate, attest to respect for children's developing autonomy. Permission and assent are thus the intertwined foundations of ethical research involving children.

Much of the ethical analysis that underlies the principles and processes of informed consent in research derives from or is similar to the analysis of informed consent to receive medical treatment. Nonetheless, agreement to participation in research differs from agreement to receive clinical care. As discussed in this chapter, patients—and investigators—may sometimes not clearly understand that research has purposes distinct from clinical care. Such a lack of understanding can compromise the objectives of informed agreement to participation in research.

General Conditions for Informed Consent

In the classic analysis by Faden and Beauchamp (1986), the conditions for autonomous decisions or actions include intentional action, understanding of the action, and voluntary or uncontrolled action. Informed consent may be viewed as one type of autonomous choice, but a choice that must also be considered within “the web of cultural and policy rules and requirements … that collectively form the social practice of informed consent in institutional contexts” (Faden and Beauchamp, 1986, p. 277).

The conditions of understanding and voluntariness are central to requirements that investigators provide relevant information to prospective research participants, evaluate their understanding of this information, and assess situations for possible coercion or undue influence on the decision to participate in research. Voluntary choice extends beyond the initial, explicit agreement to participate in research to the sometimes implicit and sometimes explicit agreement to continue participation on the basis of an understanding of one's right to withdraw.

Parental permission is ethically distinct from consent, but the conditions of understanding and voluntariness and freedom from coercion or undue influence still apply to assessments of the procedures for obtaining permission. With appropriate regard for a child's or an adolescent's maturity, they also apply to procedures for seeking assent.

Although parental permission is sometimes described as “consent” or “proxy consent” and children's agreement is also sometimes labeled as “consent,” such labels are misleading. They do not properly reflect the ethical differences between permission or assent and informed consent. Again, as the concept has developed, only those who are held competent to make autonomous decisions on their own behalf can provide informed consent. Thus, parents can provide informed consent only for themselves. Minors cannot provide informed consent unless they have been judged to be able to act as adults for that purpose (see the discussion below of emancipated and mature minors).

Within the limits of parental discretion described in Chapters 3 and 4, the ethical and legal provisions for parental permission assume that parents will make decisions about research participation that protect their child's interests.2 In exceptional circumstances, for example, in certain situations involving child abuse and neglect, that assumption fails. Alternative mechanisms (e.g., appointment of a guardian) may be necessary to protect the child. In less exceptional situations, investigators who engage parents in careful discussions of research participation may still occasionally conclude that parents are not adequately considering their child's well-being. They may then judge that the ethical standards for the child's participation are not met, even though the parents are willing to give permission.

Consent, Permission, and Assent as Processes

As familiarity with the practical realities of obtaining informed consent to medical care or research participation developed, ethicists, investigators, and policymakers recognized that creating understanding is more than a simple matter of providing information or preparing clear consent forms. Rather, a careful process of communication is necessary, one that includes the opportunity for parents and children to ask questions and investigators to make assessments of the extent to which a decision about participation in research (and about continued participation, once it has started) is made freely and with understanding (Faden and Beauchamp, 1986).

The inclination to concentrate on physical consent forms is, however, strong. Forms are fairly easy to review, and their use is relatively straightforward to document. Two recent reports on research ethics still found it necessary to stress in their recommendations that informed consent should be understood as a process and not merely a form (NBAC, 2001b; IOM, 2003a). A third report, which focused on social and behavioral research, concluded that IRBs focus too much on the consent form and also fail to calibrate their attention to the level of risk posed by the research (NRC, 2003). That is, they do not match the intensity and specificity of their review of consent documents and processes to the risk presented by the research. The recommendations at the end of this chapter suggest some ways to direct investigator and IRB attention to the process of seeking parents' and children's agreement to research participation.

The focus on forms also reflects, in part, the concerns of research institutions and sponsors about litigation that might cite omitted information or other alleged deficiencies in consent forms. This concern is evident in the legal tone of many consent, permission, and assent forms, which include terminology that is unfamiliar and potentially intimidating to many. Such language may satisfy the “informed” aspect of consent but limit a prospective research participant's true understanding of the proposed research. Again, the focus should be on the process, and the forms should support that process.

Although failures of comprehension about any major element of research participation raise ethical concerns, a particular concern is whether prospective research participants understand that they will be participating in research and that the purpose of research differs from the purpose of normal clinical care. The purpose of research is to generate knowledge, usually for the benefit of patients or individuals in the future. The belief that the purpose of research is treatment is termed the therapeutic misconception (Appelbaum et al., 1982).3

In some cases, the description of an area of research may encourage this misconception. For example, Churchill and colleagues (1998) and Lysaught (1998) have made this argument about the labeling of gene transfer research as “gene therapy.” As discussed in Chapter 1, references to “therapeutic” and “nontherapeutic” research may likewise contribute to misperceptions that the purpose of clinical trials is individual clinical care. Referring instead to research interventions or procedures as having or not having the prospect of direct benefit may help reduce misunderstandings.

King (2000) has argued that the threat that participants will misunderstand the purpose of research is large enough that IRBs should devote more attention to the description of potential benefits in the consent process. King argues, for example, that IRBs should require a statement in consent forms that benefit is not expected in early-stage studies, unless the investigators can make a good case for a reasonable prospect of direct benefit. King also argues that in consent forms the introductory discussion of what a study hopes to prove may mislead participants who are eager to receive treatment that will improve their lives. In addition, one of the anonymous reviewers of this report observed that in his long experience, protocols for clinical trials do not always clearly distinguish between the potential benefits or harms generally associated for treatment of a disease and the potential benefits and harms specific to the intervention being tested.

In an analysis of the therapeutic misconception, Dresser observed that patient advocacy groups “often portray study participation as the way to obtain cutting-edge therapy” (Dresser, 2003, p. 240; see also, Dresser, 2001). Dresser suggested that exaggerated expectations about research benefits not only can undermine informed decision making by individuals but also can diminish social activism that promotes better access to existing beneficial care (also see, more generally, Callahan, 2003).

When research includes medical procedures or interventions, clinical investigators have the obligation to provide them as carefully and as competently as they would in usual care situations. Nonetheless, to protect the validity of the research and its potential to create knowledge that will benefit future patients, certain elements in a research protocol (e.g., random assignment to an experimental or control group and blinding of investigators to the experimental or control group) typically limit how much a particular patient's care may be individualized to better meet that patient's best interest. “The opportunities for [clinical] choice disappear when neither the health care professional nor the subject plays a role in deciding on treatment” (Lidz and Appelbaum, 2002, p. V-57). This underscores the importance of helping prospective research participants understand what participation in the research will and will not involve and what limitations the research protocol will impose on practitioner and participant decisions about treatment.

Legal Requirements for Permission and Assent

General

The concept and practice of informed consent have evolved as much through the judicial route as through the ethical route. For the most part, however, courts have focused on informed consent to clinical care, not research participation. The term informed consent itself gained currency following a 1957 court case in which the court held that physicians had the duty to disclose “any facts which are necessary to form the basis of an intelligent consent by the patient to proposed treatment” (Salgo v. Stanford Jr., 1957).

Aside from the direction provided by the Nuremberg tribunal, legislators and regulators, rather than judges, have been the major source of requirements for consent to research participation. As described in Chapter 1, the 1962 amendments to the Federal Food, Drug, and Cosmetic Act (P.L. 87-781) included requirements for informed consent from those participating in research involving investigational new drugs “except where it is not feasible or it is contrary to the best interests of such human beings” (21 USC 355(i)(4)). In 1966, the U.S. Surgeon General significantly expanded requirements for informed consent in clinical research funded by U.S. Public Health Service grants. Government officials provided more explicit guidance and direction in subsequent years, notably, in 1981 U.S. Department of Health and Human Services (DHHS) regulations that remain in place today.

The regulatory language concerning informed consent is fairly detailed. Box 5.1 includes the main provisions that apply by reference to parental permission for a child's participation in research. Statements from the regulations often appear verbatim in instructions to clinical investigators and in consent forms (see, e.g., NIH, 2000a). Some studies have, however, documented discrepancies between the elements required and the elements actually found in consent forms (White et al., 1996; Silverman et al., 2001).

Box Icon

BOX 5.1

Selected Regulatory Provisions for Informed Consent That Also Apply Generally to Parental Permission for a Child's Participation in Research. Except as provided elsewhere in this policy, no investigator may involve a human being as a subject in research (more...)

Parental Permission

Federal regulations usually require that parents provide permission for a child's participation in research (45 CFR 46.408; 21 CFR 50.55). (Consistent with earlier chapters, the text refers to the regulations that apply to federally conducted or supported research as DHHS regulations to distinguish them from the similar but not identical FDA regulations.) As discussed further below, the regulations define children with reference to state laws establishing the legal age for consent to the treatments or procedures involved in the research. In general, the rules for informed consent listed in Box 5.1 also apply to permission. That is, parents or guardians are to be provided the same kinds of information and disclosures that would be provided to adults consenting to research in their own right. The regulations define parents as biological or adoptive. A guardian is someone who is appointed by a court and who can authorize medical care for a child.

Both parents must provide permission for the child to be included in research when research involves greater than minimal risk and does not hold out the prospect of direct benefit to the child. The regulations allow exceptions when one parent is dead, incompetent, not reasonably available, or not legally responsible for the child's care. In these situations, permission from one parent will suffice.

Waiver of Parental Permission

Under certain circumstances when parental permission is not a reasonable requirement to protect a child, the DHHS regulations allow an IRB to waive parental permission—provided that an appropriate mechanism for protecting the child is substituted and the action is consistent with federal, state, and local laws (45 CFR 46.408(c)). As an example of a situation in which a waiver might be appropriate, the regulations cite only situations involving neglected or abused children. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (hereafter, the National Commission) mentioned several other circumstances in which parental or guardian permission might be waived, including when parents are incompetent to provide permission, when the research involves conditions for which state laws allow adolescents to be treated without parental consent, or when the research involves mature adolescents and only minimal risk (National Commission, 1977).

The DHHS regulations do not describe what additional protective procedures might be appropriate if parental permission is waived but only state that the choice will “depend upon the nature and purpose of the activities described in the protocol, the risk and anticipated benefit to the research subjects, and their age, maturity, status, and condition” (45 CFR 26.408(c)). One option is to appoint an independent research monitor or participant advocate. His or her role would be to verify the adolescent's understanding of the research and the assent process, determine that the agreement to participate is voluntary, and document the basis for the waiver of parental permission. If another adult (e.g., a grandparent) is knowledgeable and involved with the minor, the advocate might also seek that person's views. Other protective strategies are described later in this chapter.

In a departure from DHHS regulations, the regulations of the Food and Drug Administration (FDA) do not allow the waiver of parental permission, even when it is not a reasonable requirement to protect the child (FDA, 2001b). FDA has argued that its statute does not permit such a waiver. As noted earlier, the statute governing FDA requires the consent of those to whom investigational drugs are being administered (or their representatives) except when it is not feasible or is contrary an individual's best interests. In the committee's view, this language allows for waivers consistent with DHHS regulations.

A recommendation presented later in this chapter urges that FDA bring its policies on waiver of parental permission into agreement with those governing research conducted or supported by DHHS. It also urges FDA, the National Institutes of Health (NIH), and the Office for Human Research Protections (OHRP) to cooperate to define more explicitly the factors that should be considered in waiver decisions and the safeguards that are appropriate for different situations when parental permission is waived.

Although not limited to research involving children, OHRP (then called the Office for Protection from Research Risks) and FDA have since 1996 provided for waivers of consent or permission when research involves certain emergency situations (Ellis and Lin, 1996; 45 CFR 46.116(c) and (d); FDA, 1996b; 21 CFR 50.24). A detailed set of requirements must be met and documented before a waiver can be approved. For example, “the subjects are in a life-threatening situation, available treatments are unproven or unsatisfactory, and the collection of valid scientific evidence, which may include evidence obtained through randomized placebo-controlled investigations, is necessary to determine the safety and effectiveness of particular interventions” (21 CFR 50.24(a)(1)); “[t]here is no reasonable way to identify prospectively the individuals likely to become eligible for participation in the clinical investigation” (21 CFR 50.24(a)(2)(iii)); and “participation in the research holds out the prospect of direct benefit to the subjects” (21 CFR 50.24(a)(3)). In addition to IRB review, draft regulatory guidance from FDA specifies a process of community consultation and public disclosure (FDA, 2000a). The committee learned that at least one proposal for FDA approval of research under these regulations was being prepared under the auspices of the Pediatric Emergency Care Applied Research Network (James Chamberlain, M.D., Children's National Medical Center, personal communication, October 13, 2003).

Children's Assent

Although decisional capacity develops through childhood and even into adulthood, practical and policy considerations have led policymakers to require, in most cases, that individuals achieve a specified age (the “age of majority”) before they can enter into contracts, consent to medical care, and make other crucial decisions in their own right. State policies are important for clinical research because federal regulations on human research protections define children as “persons who have not attained the legal age for consent to treatments or procedures involved in the research, under the applicable law of the jurisdiction in which the research will be conducted” (45 CFR 46.402; 21 CFR 50.3(o)). As described below, minors (those who have not achieved the age of majority) can still be held competent to consent to medical treatments under certain circumstances.

Consistent with recommendations in the National Commission's 1977 report, federal regulations generally require assent to research participation from children judged capable of providing it. They define assent as “a child's affirmative agreement to participate in research” and also state that “[m]ere failure to object should not, absent affirmative agreement, be construed as assent” (45 CFR 46.402(b); 21 CFR 50.3(n)). Assent is not, however, required if an IRB concludes that the research has the prospect of directly benefiting the child and that potential benefit is available only in the research context (45 CFR 46.405; 21 CFR 50.52). Assent may also be waived under the same conditions in which adult's informed consent may be waived.

The federal regulations do not specify an age at which IRBs should expect investigators to begin to seek assent from children. Rather, IRBs are to consider the ages, maturity, and psychological state of the prospective research participants. In the committee's experience, most IRBs require investigators to seek assent from children over the age of 6 or 7 years. The judgment about assent may apply to all the children who are to participate in a study or to each child individually. As discussed later in this chapter, the assent process for older adolescents may differ little from the process for seeking informed consent if that consent process is suitably structured. Thus, investigators would provide the same kinds of information as is required for adults.

The regulations permit oral presentation of assent information and allow discretion on the part of the IRB about the way assent is documented (45 CFR 46.117 (c)(2) and 45 CFR 46.408 (e); 21 CFR 50.55(g)). FDA guidance on pediatric drug research (developed by the International Conference on Harmonisation [ICH]) provides that “[p]articipants of appropriate intellectual maturity should personally sign and date either a separately designed, written assent form or the written informed consent” (FDA, 2000b, p. 19781, emphasis added). Although the ICH guidance puts more stress on getting a signature than do the federal regulations, the guidance also allows discretion.

The regulations do not describe what information is to be presented as part of the assent process. In general, the type and amount of information presented should be adapted to the child's cognitive and emotional status and experiences.

The process of involving children in research typically involves a progression of assessments that combine acknowledgement of the child's vulnerability with respect for the child's developing maturity (Fisher, 2003; Kodish, 2003a). IRBs must first decide whether the research is approvable under federal regulations and that it includes appropriate processes for requesting parental permission and child assent. For approved protocols, parents usually then decide whether participation in the research is appropriate given their child's characteristics and experiences. If parents agree, the child decides whether the research, as he or she understands it, is an activity in which he or she wishes to participate. Particularly with young children, whose decisional capacities may fluctuate with fatigue and irritability, it may be reasonable—as part of a respectful and noncoercive assent process and with parental agreement—for investigators to come back to a child who says “no” to see whether he or she may feel differently later. Depending on the child's age, the nature of the research, and the investigator's perspectives on involving children, the discussion about research participation may start with the parents or may include the child from the outset.

Emancipated and Mature Minors

Even when individuals have not reached the age of majority, states provide, under certain circumstances, that they may consent to medical treatment without parental agreement. Appendix B describes state policies (including court decisions) in more detail. It makes clear that most states have no explicit provisions about a minor's agreement to participation in research without parental consent. (See also, English and Kenney, 2003.)

In most states, minors can become emancipated and attain legal majority through certain actions, which usually include marriage, enlistment in the military, or being self-supporting and living independently. Although the terminology or label used in different states varies, emancipated minors can make decisions as if they had reached the age of majority. They thus can provide informed consent to medical treatment in their own right. Emancipation sometimes requires a court order.

In addition, states permit minors to consent to treatment for certain problems (e.g., drug or substance abuse or sexually transmitted disease) without parental permission. In many states, the statutes allowing such treatment also forbid the health care provider from billing the parent, lest confidentiality be breached. A number of states have general consent statutes that permit minors to consent to medical or surgical treatment at a specified age; usually it is age 16 years but in some states it is as young as 14 years. Furthermore, the jurisprudence in many other states has accepted the “mature minor” rule, in which a minor who is subjectively assessed as capable of giving the same degree of informed consent as an adult may be treated without a parent's involvement. In some states, minors who are parents may be in the paradoxical position of being able to permit treatment on behalf of their child but not on their own behalf.

Again, as Appendix B explains, statutes and case law rarely refer specifically to decisions by minors about participation in research. Because state laws are so variable and often vague, investigators and IRBs must be knowledgeable about federal, state, and local laws related to consent, confidentiality, and related matters when they consider research that involves waiver of parental permission. Differences among state policies could complicate certain studies involving sites in more than one state and might preclude the inclusion of certain sites. In addition to seeking legal counsel, investigators and IRBs may find it prudent to consult representatives from communities that might be affected by the proposed research and to consider the climate in the state regarding parental prerogatives. As emphasized earlier, federal regulations require that appropriate protective procedures be in place if parental permission for research participation is to be waived.

Wards of the State

Responding in part to some of the abuses and controversies described in Chapter 1, federal regulations impose special conditions for research participation by children who are wards of the state. Although the DHHS regulations do not define the term ward, the interim regulations issued by FDA define a ward as “a child who is placed in the legal custody of the State or other agency, institution, or entity, consistent with applicable, Federal, State, or local law” (21 CFR 50.3(q)). When the state assumes the role of parent for these children, it may still defer to the child's biological parent(s) if the custody is temporary.

For research that involves more than minimal risk and that does not hold the prospect of direct benefit to the child participant, IRBs can approve research involving a ward of the state only under specific circumstances. These are when the proposed study (1) is related to the child's status as a ward or (2) is conducted in settings such as schools or hospitals where the majority of children participating in the study are not wards. When such research is approved, each child must have an independent, competent appointed advocate (in addition to any other appointed guardian). If a minor is under detention by the state, the provisions of Subpart C of the DHHS regulations, which govern research involving prisoners, may also apply.

As discussed in Appendix B, state policies on research participation by wards of the state rarely appear to be codified in statute or clear case law. Rather, state agencies responsible for these children often appear to have little formal, written guidance for making decisions. As a consequence, investigators likewise may have little guidance beyond the federal regulations. Some states appear to be much more restrictive than others in permitting wards to be included in research and may thereby discourage research that might increase understanding of this population of vulnerable children.

RESEARCH RELEVANT TO PARENTS' COMPREHENSION OF CHILDREN'S PARTICIPATION IN RESEARCH

It's a lot easier to make a decision for yourself rather than for somebody else … you just don't know if you've done the right thing.

Parent (quoted in Caldwell et al., 2003, p. 557)

The literature on adults' understanding of and decisions about their own participation in research is considerably more extensive than the corresponding literature on parents making decisions about their child's participation in research. The committee believes that findings about the reasoning that adults use and the decisions that adults make on their own behalf are generally relevant to an understanding of decisions that adults make on behalf of children. It recognizes, however, that the anxieties associated with making decisions on behalf of one's child, especially one's sick child, may put particular stress on an adult's comprehension, reasoning, and decisionmaking capacities (Ruccione et al., 1991; Levi et al., 2000; McGrath, 2002). Some research findings about informed consent for clinical care are also relevant to the research context, although the ethical and informational requirements are usually more stringent for informed consent in the context of research.

What should parents be able to understand to make an informed decision about their child's participation in research? In general, parents of prospective research participants should have a basic understanding of

  • the purposes of the research and what procedures (medical and otherwise) that the child will undergo (or may undergo, if random assignment to treatment and control groups is planned);
  • the potential consequences of the child's research participation, including the likelihood, significance, and duration of possible harms and benefits;
  • their right to accept or refuse research participation and to withdraw the child from a study once it has started;
  • their responsibilities as the parents of a child participating in research and the child's responsibilities (if any); and
  • the responsibilities of investigators to parents and children, including answering questions, maintaining the confidentiality of data, minimizing the risks of participation, and providing information about study findings.

If the child has a medical condition, certain additional areas of understanding are also important when participation in a clinical trial is proposed. Parents considering an ill or “at-risk” child's participation in a clinical trial should understand

  • their child's medical condition and prognosis;
  • the difference between receiving usual clinical care and participating in a trial;
  • their child's options for care (e.g., standard treatment, monitoring, or hospice care) outside the trial;
  • the research methods, including, if applicable, methods for assigning research participants to intervention or control groups; and
  • the potential harms and benefits for the child of participating or not participating in the trial, given the child's medical status and prognosis.

The following discussion looks first at the general literature relevant to adult's comprehension of research participation. It then considers some research specific to parents' decision making on behalf of their child. The research literature supports increased attention to the process of seeking parental permission. It also supports a recognition that decision making about research participation—whether on one's own behalf or that of another—is an imperfect process.

General

Reading Levels, Readability, and Informed Consent and Permission Forms

Consent and permission forms are only one element of a meaningful process of parental decision making about a child's participation in research. Nonetheless, to the extent that the forms fail to assist parents in making decisions because they are not easily comprehended, this failure constitutes an important deficit in the informed consent and permission process. In principle, investigators can compensate for some shortcomings in consent forms as they talk with people and explore their understanding of research participation. In reality, little is known about these conversations, including whether they can compensate for deficient consent forms. Although forms should not be the main focus of the consent and permission process, the exercise of developing a clear, readable, and informative consent document can contribute to the design and implementation of a clear, understandable, and constructive process for seeking informed consent or permission.

Research institutions and IRBs typically provide the same forms and related guidance for studies that involve adults making decisions about their own participation in research and for studies that have parents making decisions about their child's participation. Study after study has found that the consent forms for medical treatment or research participation typically require high levels of education or, at least, levels of education higher than the average for key patient or study populations (e.g., low-income patients) (see, e.g., Ogloff and Otto, 1991; Rivera et al., 1992: Grossman et al., 1994; Hopper et al, 1995, 1998; Goldstein et al., 1996; and Paasche-Orlow et al., 2003). This is a serious problem given, that surveys suggest that more than one-fifth of adults in the United States have very limited literacy skills and that another quarter have seriously limited skills (Kirsch et al., 1993).

Consistent with other studies of consent forms for adults, one analysis of forms for the parents of prospective child participants in an array of biomedical research studies at a single hospital found that forms were written at the college-graduate level (Tarnowski et al., 1990). The researchers also found that forms had gotten longer but not more readable over a 10-year period.

Another study of institutional consent forms (for clinical care) and vaccine information pamphlets intended for parents reported that the forms generally required a reading level of 12th grade or above and the pamphlets required an 11th-grade reading level or higher (Davis et al., 1990). In reality, the inner-city parents at the investigators' institution had a median reading level just below the seventh grade. A second study with the same lead author assessed educational materials from the American Academy of Pediatrics, the Centers for Disease Control and Prevention, and other sources. It found that parents at the outpatient pediatric clinic tested at a seventh- to eighth-grade reading level but that 80 percent of the educational materials required at least a 10th-grade reading level (Davis et al., 1994).

Unfortunately, efforts to make consent forms more readable have their limitations. The authors of one recent analysis of consent forms concluded, “IRBs commonly provide text for informed-consent forms that falls short of their own readability standards” (Passche-Orlow et al., 2003, p. 721). Nonetheless, the authors concluded that the sample text provided by some IRBs indicated that forms could be written clearly and simply at a fourth-grade level and still provide essential information.

An earlier analysis by Hammerschmidt and Keane (1992) reported that IRB review of consent forms usually did not improve their readability levels. Furthermore, after a simplified consent form was compared with the standard consent form for a clinical oncology trials group, researchers found that participant comprehension was essentially the same for the two forms, although nearly all participants found the simplified form more readable (Davis et al., 1990; 1994; 1998; see also Duffy and Kabance, 1982; Charrow, 1988; and Taylor et al., 1998).

A recent randomized clinical trial of an easy-to-read consent statement for several cancer treatment trials concluded that documents could be simplified without omitting important information (Coyne et al., 2003). Such documents could increase patient satisfaction and reduce the anxiety associated with the consent process. (The authors noted that participants in their study tested at an average literacy level at or above ninth grade, higher than might be found in many research contexts.)

Evaluations of consent forms often focus on quantitative “grade-level” measures of readability, but other aspects of written forms should also be considered.4 These include the logical flow of ideas, the organization of headings and subheadings, type style and size, page layout, the helpful use of graphics, and cultural and linguistic appropriateness or adaptations. Actual comprehension of the text may depend not only on the characteristics of the text and the level of literacy of the individual, but also on other individual characteristics, such as motivation to learn, interest in the material, and experience with the subject matter.

The National Cancer Institute has developed guidelines for researcher design and use of consent forms and has also provided a template to serve as the basis for devising forms for specific protocols (NCI, 2001). In general, the development of templates (informed by evidence about adult learning and comprehension of information about research participation) is a useful step, especially for studies conducted at multiple sites. As discussed elsewhere in this report, local IRBs may modify consent forms in ways that compromise the research and the practical purposes of having a common form for multicenter studies. Local modifications may sometimes be appropriate for local study populations, however.

Presentation of Quantitative Information

Just as the readability of research consent and permission forms may affect parent's comprehension, so may the way in which information about the probability of potential harms and potential benefits of research participation is presented. The committee found no studies that assessed the quantitative information provided to prospective research participants or to the parents of prospective research participants. Some studies in other areas, including medical decision making, are generally relevant.

A number of studies indicate that adults often have considerable difficulties accurately understanding and using quantitative information (see, e.g., the review by Stanovich and West, 2000). Several studies also indicate that the manner in which quantitative and other information is presented can affect accurate understanding of the likelihood of harms and can also influence an individual's choices (Edwards et al., 2001). For example, studies of people's understanding of quantitative information about the outcomes of medical screening (e.g., mammograms) suggest that they find frequency information (e.g., a 4 in 1,000 chance of an event) more understandable than probabilities (e.g., a 0.04 percent chance) (Gigerenzer, 1996). Information presented in relative terms (e.g., a 50 percent increase or decrease in some outcome) tends to be more “persuasive” than information presented in absolute terms (e.g., decrease from 2 in 10,000 to 1 in 10,000) (Forrow et al., 1992; Malenka et al., 1993; Ransohoff and Harris, 1997).

Other research suggests that framing quantitative information about possible outcomes in positive terms (e.g., a 3-in-4 chance of improvement or survival) rather than negative terms (e.g., a 1-in-4 chance of deterioration or death) may encourage individuals to choose less risky options (see, e.g., Rothman et al., 1993, Rothman and Salovey, 1997; see also, Llewellyn-Thomas et al., 1995; McGettigan et al., 1999). Increasing the number of options or the number of negative outcomes discussed can produce unexpected shifts in people's choices, at least, in hypothetical situations (Ubel, 2002). Finally, although it is tempting and common to substitute verbal descriptions for numerical descriptions, research indicates that people vary in their interpretation of qualitative expressions such as “rare” or “most” (see, e.g., Sutherland et al., 1991; Mazur and Merz, 1994; Cohn et al., 1995; Man-Son-Hing et al., 2002; Mazur and Hickam, 1991). Providing quantitative descriptions of probabilities (e.g., a 1-in-4 chance) rather than or in addition to verbal descriptions (e.g., a moderately low chance) may reduce inconsistency in the interpretation of risk information and encourage more deliberative thinking.

A considerable body of research considers people's reasoning and decision-making capacities more generally. Much of this research describes deficiencies in people's reasoning capacities, for example, selectively focusing on information that supports one's views (Stanovich and West, 2000). This underscores the importance of a careful process for helping prospective research participants reach decisions.

Comprehension of Research Purposes

In the end, it is only the benefit of furthering knowledge that can be honestly guaranteed to a potential research subject.

Advisory Committee on Human Radiation Experiments, 1995, p. 476

Most research evaluating people's understanding of the difference between research and usual clinical care has involved adults consenting to research participation in their own right. It generally indicates that avoiding or overcoming the therapeutic misconception can be a formidable challenge. As Appelbaum and colleagues (1982) observed many years ago about participants in psychiatric research, “subjects' ability to distort small aspects of the study design often had the effect of maintaining their therapeutic misconceptions, while they gave the appearance of having a good general understanding of the study” (p. 328).

Other studies also suggest that research participants may have difficulty understanding the purpose of research (see, e.g., Yuval et al., 2000 and Daugherty et al., 1995). When Edwards and colleagues (1998) reviewed 61 studies about attitudes toward clinical trials, they found that people mentioned self-interest more often than altruism as the reason for participating in trials. (Often, however, the specific questions and methods were not fully enough described to provide a clear picture of the results.) They also cited three studies from the 1990s that reported a near majority or majority of physicians believed that research participants did not understand the information given them or realize that they were participating in research.

Several studies suggest that research participants frequently have expectations of benefit, even in clinical trials that test safety but not efficacy (see, e.g., Daugherty et al., 1995; Schutta and Burnett, 2000; Meropol et al., 2003; and Weinfurt et al., 2003). Likewise, although they may understand and approve the knowledge-generating purpose of research in general, participants may view their own participation in research primarily in terms of benefit to themselves (Cassileth et al., 1982; Bevan et al., 1993; Wilcox and Schroer, 1994; Aby et al., 1996; Hutchison, 1998; Yoder et al., 1997; Cheng et al., 2000; Madsen et al., 2000). Individual reactions to randomization and placebo-controlled trials suggest a tendency to view the intervention arm of a trial as desirable or beneficial (see, e.g., Snowdon et al., 1997; Fallowfield et al., 1998; and Welton et al., 1999).

A survey and interviews conducted by the Advisory Committee on Human Radiation Experiments (ACHRE) produced a number of interesting findings about people's understanding of research participation (ACHRE, 1995).5 The survey involved more than 1800 patients at 19 health care institutions across the country. A third of those surveyed believed that patients who participated in medical research usually or always benefited medically compared with those who did not. About two-thirds of the patients who had participated in research indicated that they had done so to obtain better treatment; a similar percentage reported that being in research gave them hope. Results from 103 in-depth interviews were generally consistent with the survey findings. The researchers, however, reported that patients who had been in diagnostic, epidemiologic, or survey research were more likely to differentiate between research and treatment than those who had participated in studies testing a potentially therapeutic intervention.

About three-quarters of those surveyed for ACHRE indicated a desire to help others or advance scientific knowledge. Suggesting that this altruistic motivation should be “tapped explicitly” when investigators are recruiting research participants, the ACHRE report explains that such an emphasis might reduce the potential for research to be misunderstood by underscoring “for patients that the primary objective of research is to create generalizable scientific knowledge” (ACHRE, 1995, p. 476).

Patients considering research participation may not necessarily anticipate improvement in their condition from the intervention being tested, but they may expect to receive better diagnostic evaluation, closer medical monitoring and follow-up, and more information about their condition (Mattson et al., 1985; ACHRE, 1995; Yoder et al., 1997; Madsen et al., 2000). For example, for Danish patients in a trial of interventions for inflammatory bowel disease, an important reason for participating in the research was “the expectation of being ‘a special patient' during the trial” (Madsen et al., 2000, p. 463).

Implications

Although more studies of ways to present qualitative and quantitative information about research are needed, the research summarized above provides some guidance to investigators on how they can prepare to discuss a child's research participation with parents. It suggests that documents to support the permission process can often be written at a simpler level than is customary, although investigators and IRBs should still be aware that more readable information and forms do not guarantee better understanding of research participation. A number of research institutions and IRBs set an eighth-grade reading level as the target for written materials, but as discussed earlier, a target of a sixth-grade reading level may benefit many parents and other adults as long as essential information can still be accurately presented.

Research about the comprehension of quantitative information suggests that investigators should present estimates of potential harms and benefits using frequencies rather than (or in addition to) probabilities and using absolute terms rather than (or in addition to) relative terms. They should not rely on qualitative descriptions of potential harms and benefits but should include quantitative descriptions or estimates to the extent possible. Strategies for communicating about potential harms and benefits of research should consider characteristics of study populations as well as individual variability in education levels, experiences, and other characteristics.

The general findings summarized above underscore, again, the importance of focusing on consent and permission as a process. That process should include discussion and time for questions. It should also, depending on the study, provide for a tactful assessment of the prospective participant's understanding of research purposes, potential harms and benefits, voluntary participation, and other key information.

Research on Parents' Decisions About a Child's Participation in Research

We didn't have time to think.

Mother of child with cancer (quoted by Pletsch and Stevens, 2001, p. 57)

Most parents … don't know anything beyond their current crisis, and they are trusting in the physician to carry them through.

Nancy Sander, family advocate, 2003

Several studies have explored factors associated with parents' decisions to involve their child in research. Many involved small numbers of parents of limited sociodemographic diversity. Some studies involved validated survey or interview instruments; others apparently did not. Publications did not always report quantitative information on parents' responses, and some relied on quotes from parents that conveyed views vividly without providing a clear sense of how representative those views were.

A few studies included both parents who permitted their child's enrollment in research and parents who did not, but most included only parents who had agreed to participation. A number of studies included at least some parents of children with serious illnesses or at risk of such illnesses. Several studies were conducted in other countries, mainly in Canada, Australia, and Western European nations. Differences in cultures, research practices, and health care norms could limit the relevance of these studies to the United States.

Reasons for Permitting or Not Permitting a Child's Participation in Research

Some studies suggest that many parents (like many adults making decisions in their own right) view research as an opportunity to gain additional access to care and information about their child's condition and treatment. In a Washington State study, 44 parents who had agreed to their child's participation in an asthma study completed a questionnaire about their motives (Rothmier et al., 2003). The authors concluded that “[a]lthough altruistic motives are present in pediatric asthma research, most parents/ guardians gave consent for their child to learn more about their child's asthma” (p. 1037). Access to the newest drugs and relationships with staff were also positive factors. Parents with lower family incomes were more likely to respond that access to free medications was a factor in their decision.

Of 181 Dutch parents who agreed to their child's participation in a study of ibuprofen for the prevention of seizures, approximately one-third reported an expected benefit to their child as the reason for agreeing (van Stuijvenberg et al., 1998). Slightly more than half cited a desire to contribute to science as a motivation for agreeing to their child's participation. Two-thirds of the parents reported that they saw no disadvantages to their child's participation in the research.

Hayman and colleagues (2001) surveyed 94 New Zealand parents who agreed (69 percent response rate) and 103 who declined (47 percent response rate) to enroll their child in a physiological study (without a prospect of direct benefit) related to sudden infant death syndrome. All those who participated reported doing so for altruistic reasons. Just over a quarter of these parents were initially concerned about safety issues but no longer expressed concern after the study. Of the parents who declined to participate, about half cited inconvenience and a quarter cited safety concerns.

In a study that included both parents who agreed (n = 221) and parents who did not agree (n = 208) to enter their infant in a randomized clinical trial of pertussis vaccine, Canadian researchers concluded that altruistic motivations (a desire to contribute to medical knowledge and to help others) were major factors contributing to parents' willingness to enter their infants into the trial (Langley et al., 1998). The involvement of the family physician was also important to a majority of those who agreed. Concerns about painful procedures and blood drawing were important to parents who did not agree to their child's participation.

Another group of Canadian investigators questioned parents who had agreed (n = 103) or declined (n = 37) to have their newborn infant included in one of three randomized, controlled clinical trials in a neonatal intensive care unit (Zupancic et al., 1997). (The response rates were between 80 and 85 percent for the two groups.) The researchers reported that parents giving permission for the infant's enrollment were more likely to see the research as probably benefiting their child and less likely to perceive it as risky than were parents not giving permission. Parents who allowed their infants to enter the study were also less likely to view the consent process as too complex, although only a minority in each group gave this response—10 percent of those who gave permission versus 22 percent of those who did not. Nearly all parents who gave permission believed that it was important for children to take part in research because it would help other children. A large majority (84 percent) of the parents who did not give permission also responded positively about the importance of research participation. The two groups did not differ significantly in socioeconomic characteristics.

In a study that included parents permitting and parents not permitting a child's participation in anesthesia research (168 and 78 parents, respectively), Tait and colleagues (1998) reported that parents who did not give permission cited fear for the child's safety and the potential risk to the child as the most important factors in their decision. Very few reported that a lack of understanding of the research was a factor, but 15 percent were concerned about having insufficient time to decide. None reported feeling pressured to agree. Parents who gave permission cited the importance of the research and the low risk to the child as key factors in their decision. The parents who allowed their children to enter the study were more likely to have a child who had participated in a previous study and to have read the consent form completely.

In a statement presented to the committee, a group representing parents of children with asthma noted that parents may simultaneously understand the importance of allowing children to participate in research and be reluctant to allow their own children to serve, in their words, as “guinea pigs” (AANMA, 2003, p. 2). The statement cited a number of reasons for such reluctance, including concerns about safety, pain, and inconvenience (e.g., disruption of the children's routines and missed work by the parent).

In sum, research suggests that parents have various reasons for permitting their children to participate in research. Some may involve expectations of a direct treatment benefit or perceptions that research participation increases access to care and information. Altruistic motives figure in some parents' agreement to their children's participation in research, usually in studies involving no prospect of benefit. Parents not permitting a child's participation are more likely to cite concerns about safety and pain or discomfort than parents agreeing to participation.

Experience of Illness and Expectation of Benefit

There is never enough time … in those kinds of situations to make an “informed decision” [about research]. I think that to say … that either the parent or the patient is making an “informed decision” is not correct…. And, of course, we signed [the form] because that's what you do…. It's always at the worst time to be reading this type of material…. But parents aren't in control when they're doing this, nor is the child.

Joseph Lilly, parent, 2003

The circumstances under which parents face making a decision about a child's participation in research may have a profound effect on their ability to evaluate information, ask questions, and make reasoned assessment of the potential harms and potential benefits presented by the proposed research. In particular, the conditions for informed and reasoned choice are threatened when parents are confronting a new diagnosis of a life-threatening medical condition and a crisis situation in which immediate decisions are sought.

Pletsch and Stevens (2001) investigated the factors affecting initial decisions about research participation by mothers of children who had leukemia or diabetes. (The investigation was part of a larger study that also studied children's comprehension and perceptions of research participation, as discussed later in this chapter.) The authors concluded that “the boundaries between research and treatment were unclear or irrelevant” for many of the 24 mothers whose children had cancer (p. 59). The mothers described being in shock and not having time to think. The urgency of the threat to their child dominated their decision making. In contrast, the nine mothers of children with diabetes faced circumstances more supportive of informed decision making. They were more concerned about the risks of research for children whose condition had, for the time being at least, been stabilized. These mothers were able to clearly differentiate between their child's usual care and the research activities. Their families had had time to adjust, to some extent, to the diagnosis and did not confront the need for crisis-driven decision making about the participation of the child in research. They also had had time to develop some expertise in negotiating and working within the health care system.

Kodish and colleagues (1998) reported that pediatric cancer researchers believed that parents' “state of shock” was the major factor interfering with understanding of the participation decision (see also, Simon et al., 2001). These findings are consistent with those of other studies about parents' decision making about clinical care in the face of a child's cancer diagnosis. These studies point to a high state of parental anxiety and the frequent sense of not having a real choice (Ruccione et al., 1991; Hinds et al., 1996; Levi et al., 2000). Levi and colleagues (2000) reported that these stressed parents “did not verbalize distinctions between their understanding of their child's medical treatment, [and] research participation” (p. 3).

Deatrick and colleagues (2002) undertook a retrospective analysis of transcripts of interviews and other information related to an earlier study of decision making by parents (n = 21) of children with advanced cancer about the child's participation in phase 1 trials. All the families were facing situations in which all other treatments with curative intent had failed. More than half the parents (n = 13) reported that they had no choice; as one said, “you have to try everything … keep fighting to keep her here with me” (p. 118). As described by the researchers, the parents' expectations included “providing treatment, buying time for another therapy, working a miracle, being altruistic, and delaying death” (p. 117). Some parents were clear that they were hoping for a miracle. At least one parent spoke of the doctor's optimism about the drug as the deciding factor.

Using a combination of interviews and focus groups, Caldwell and colleagues (2003) studied 33 Australian parents whose children's health status ranged from healthy to seriously ill. The investigators reported that parents of ill children, some of whom were or had been involved in clinical trials, saw the greatest potential for direct or personal benefit from the trial. Not surprisingly, these parents were also more knowledgeable about trials. As viewed by one parent whose child had participated in a trial, children not participating were “missing out,” whereas participating children were “getting the benefit” (p. 556). Parents of ill children were concerned about “blinded” studies that withheld information about the treatment that their child was receiving. Parents of children with cancer mentioned that they wanted to monitor medications because of past experience with medical errors. (The majority of parents approached by Caldwell and colleagues either declined outright to participate in the interviews or did not participate after indicating that they would.)

Differences between families with different experiences of illness were also reported by Geller and colleagues (2003). They interviewed 37 pairs of parents and children or adolescents about participation in genetic susceptibility research. During interviews that asked questions about research that might offer benefit, the parents from families with a history of breast cancer indicated that they would not allow their daughters to refuse participation in potentially beneficial research. In contrast, the parents with a family history of heart disease were more likely to talk about “overriding” the wishes of a child who wanted to participate in a study (p. 266). Parents in both groups strongly agreed that children should not be enrolled against their will in research intended only to gain knowledge.

In the study by Zupancic and colleagues (1997) cited earlier, the investigators reported no differences in perceptions of severity of illness between parents who agreed and those who did not agree to their neonate's participation in clinical trials in the intensive care unit. As described above, parents who agreed were more likely than parents who did not agree to believe that the trials would probably benefit their child and were less likely to believe that the trials were risky.

Overall, these studies point to the need for particular care in the design and review of processes for seeking parental permission for research that will involve seriously ill children and high-stress situations. Concerns about the inadequacy of current procedures have led to suggestions that the consent process be designed (and tested) as a “continuous” one that involves giving and then repeating information and explanations as research proceeds (see, e.g., Allmark et al., 2003). When a study involves a discrete, one-time intervention and an urgent situation, this approach likely will not be relevant.

Involvement of Children in Discussions and Decisions

The committee found limited information on parent's views about the involvement of their child in discussions and decisions about the child's participation in clinical research. It likewise located little research on the dynamics of family relationships in decisions about research participation or on the consequences for family relationships of different strategies for conducting the permission and assent process.

Based on a questionnaire study of 100 pairs of parents and adolescents at children's hospitals, Sikand and colleagues (1997) reported that 57 percent of the adolescent-parent pairs agreed that parents' permission for an adolescent's participation in research was unnecessary for anonymous surveys. In other situations, the parents were consistently more likely to favor permission requirements. For example, 78 percent of the parents but only 59 percent of the adolescents favored parents' permission for blood testing for human immunodeficiency virus (HIV) infection, and 62 percent of parents but only 48 percent of teens supported parents' permission for face-to-face interviews. The researchers suggested further research with larger groups to identify factors associated with parent and adolescent views and differences.

In the Australian study by Caldwell and colleagues (2003) cited earlier in this chapter, the investigators reported that parents thought that children's preferences should be considered in decision making about participation in clinical trials. These parents said that they were prepared to let adolescents make decisions about trials involving quality-of-life interventions. They were not, however, prepared to let children or adolescents make the final decision about participating in clinical trials involving life-threatening conditions.

In a small study involving six parent and child pairs interviewed about hypothetical research scenarios, Rossi and colleagues (2003) have reported that parents sometimes but not always considered their child's wishes. No parent wanted the child to be approached for assent before the parent.

The study by Geller and colleagues (2003) cited earlier also found that most parents wanted to make the initial decision about their child's participation in research. When asked about a hypothetical research scenario, parents who were uncomfortable with the research varied in whether they would let the child know about the study. The parents and children differed in their views about how the consent process (i.e., the permission and assent process) should be structured, including whether the parents should be approached first and whether a facilitator should be involved in explaining the study to the family and answering questions. Parents' preferences were influenced by the child's age and the family's style of communication.

In a presentation to the committee, Kodish (2003a) reported that the parents he and his colleagues had interviewed were very clear that the parents and the clinicians should negotiate in advance a child's involvement in discussions and decisions about participation in cancer clinical trials. He also reported that parents were offended by the idea of having their comprehension tested after the consent conference, suggesting that an assessment of parent's comprehension needs to be tactful and discreet.

Independent of parent's attitudes about their children's involvement in discussions and decisions about research participation, it is appropriate to consider how children's involvement might affect the flow of information and the interaction between parents and investigators. Olechnowicz and colleagues (2002) evaluated 85 “informed consent” conferences, comparing those with the child present versus those with the child not present. Of the 34 conferences that concerned children over age 7 years, the child was present during 14. The investigators reported that parents asked fewer questions when the child was included in the conference. In addition, when the child was not present, clinicians were more likely to use quantitative descriptions of the child's prognosis (Kodish, 2003a). To the extent that asking questions is a key element in informative communication, this finding raises some concerns and points to an issue for further investigation.

Based on their findings, Geller and colleagues (2003) concluded that the challenges for researchers are both to assess family communication (especially to identify families that do not communicate well) and to “harness a spirit of connectedness while minimizing undue influences on children by parents or the researchers” (p. 269). The investigators argued for a model of shared decision making that, first, gave parents and children adequate time to ask questions, discuss views, and clarify preferences and, second, gave “children the opportunity to exercise their right to refuse participation without parental influence” (p. 270). The researchers also recommended further research to assess strategies to accommodate differences in parent and child characteristics and respect family relationships while still providing safeguards that give children a voice.

The committee agrees that more research would be helpful to understand parent and child views about children's participation in discussions and decisions about participation in research. It also encourages more research on the effects on communication of including children in initial discussions of research participation and, more generally, the effects of different ways of structuring the permission and assent process. Such research should help guide investigators and IRBs in developing procedures that both support parental roles and family relationships and provide for the developmentally appropriate and respectful involvement of children in discussions and decisions about research participation.

The Role of Physicians

Research findings are limited and somewhat mixed on parents' views of the physician's role in decision making about a child's participation in research. In the study by Zupancic and colleagues (1997) cited above, the investigators reported that one-third of parents (including those who did not agree to participation) agreed with the statement, “I would prefer to have the doctors advise me whether my baby should be in the study, rather than asking me to decide” (p. 3). Investigators in another Canadian study found that parents believed in the necessity of research but often believed that their own knowledge was limited, so they needed to depend on their baby's doctor (Singhal et al., 2002). A subsequent study, which involved a different group of parents from the same Canadian research site, found that when parents were explicitly asked who should make the decision about enrolling the newborn, nearly all parents rejected the option of having the physician decide (Burgess et al., 2003).

In his presentation to the committee, Kodish (2003a) reported that trust was an important theme in interviews with parents about their decisions on their child's participation in research. As expressed by one parent, “if I have to put my faith in a human, it has to be the doc. So, I have to trust her unconditionally.” Nonetheless, although parents saw a clinician's presentation of participation in a clinical trial as an implicit recommendation, they wanted an explicit recommendation only upon their own request.

Reporting on a study of 64 Australian parents questioned after completion of a clinical trial, Harth and Thong (1995) found that 15 percent of parents viewed strict informed-consent procedures as unnecessary because they would follow the doctor's advice. A majority thought being in a hospital-based trial involved little or no risk, and few realized that the study was to assess safety as well as efficacy. Only one-third of the parents were aware that they could withdraw their child from the trial. These results suggest deficits in the consent and permission process, even if the parental bias is to rely on the doctor's advice.

In a study of decisions about cancer treatment (not research), Pyke-Grimm and colleagues (1999) studied the preferences of 58 Canadian parents about the role of physicians in decision making. They found that parents most preferred collaborative decision making between families and physicians, with a minority of participants preferring either a passive role (of parents) or active involvement. The investigators also reported that parents placed more emphasis on their needs for “concrete” information than on information about possible emotional or family consequences. Parents' views about their information needs were highly variable, which led the investigators to stress that the provision of information needed to be tailored to the individual families. This is possible in a well-designed consent and permission process.

In another study that examined parents' preferences for clinician involvement in decisions about clinical care (not research) during elective surgical procedures, Tait and colleagues (2001) showed parents generally preferred a process of shared decision making with the anesthetist. A significant fraction, however, wanted a more active role; those parents were less satisfied with their child's care.

Although limited, these findings generally support the movement in recent decades away from medical paternalism toward a collaborative model of decision making about medical care and research participation. They also suggest that investigators should be sensitive to and respectful of differences in parents' decision-making styles and preferences. At the same time, investigators must be mindful of the potential for undue influence based on power and information imbalances between investigators and parents. They also must still fulfill their ethical obligations to offer the information and explanations that parents need to provide informed, voluntary permission. As Kodish observed (2003a), all this can make for a “very delicate pas de deux” between investigator and parent.

Views and Results of the Consent and Permission Process

Several studies have sought parents' views on the importance of different aspects of the consent and permission process. In another article from their study of parents' agreement to children's participation in a clinical anesthesia or surgery study, Tait and colleagues (2002) asked 184 parents and 38 investigators to rank elements of informed consent by importance. Both groups considered information about risk to be the most important element. Parents, however, put more emphasis than investigators on information about potential benefits (to their child or other children) and less emphasis on voluntary agreement.

Tait and colleagues (2003a)—in another study involving permission for clinical anesthesia or surgery research—found that although parents perceived their own understanding of elements of research “consent” as high, independent assessors were less positive. The investigators reported that consenting parents showed more understanding than those who did not consent. Factors related to parents' understanding included education level, the clarity of the disclosure information, whether their child had previously been involved in a research study, and parental attention to disclosure and consent forms.

Kodish and colleagues (1998) reported that parents found discussion with staff more helpful than the consent and form. Although they found the whole process somewhat confusing, most parents believed that the amount of information provided was appropriate. In contrast, about half of the investigators surveyed thought that too much information was presented.

In the study by Hayman and colleagues (2001) cited earlier in this chapter, the investigators reported that 85 percent of the parents who agreed to enroll their child in research believed that the verbal explanation was the most useful source of information about the trial. All the parents said they understood the study's purpose and procedures. Only 6 percent wanted more information about the study.

A small pilot study of parental decision making in cancer clinical trials found that parents were generally satisfied with the “informed-consent” process, despite their distress, the time pressures, and the amount of information presented to them (Kupst et al., 2003). During interviews after the consent discussions, parents recalled information about the diagnosis, intervention, and survival statistics but were less clear about the research procedures, including randomization. The parents expressed a wish for more information about alternatives to the trial intervention and about research aspects of the study.

Kodish and colleagues (2004) recently reported that 50 percent of parents who had participated in a discussion of their child's participation in a cancer clinical trial did not understand randomization. The lack of understanding was higher among minority parents and parents of lower socioeconomic status. Physicians had explained randomization in 83 percent of the discussions. As the researchers observed, “to make informed consent more effective, future research must seek to improve communication” (Kodish et al., 2004, p. 470).

European research also raises concerns about parental understanding. In an analysis of data from nine European countries on 200 parents giving permission for their neonates' participation in clinical trials, Mason and Allmark (2000) reported that three-quarters of the permissions provided by parents were deficient. The deficiencies were most common for emergency situations. The most frequent problem areas involved the information provided and the parents' understanding of the research. These findings were one impetus for the suggestion for a continuous consent process as mentioned above.

In a study designed to learn more about groups at risk of poorly informed decision making, Simon and colleagues (2003) compared discussions of research participation involving non-English-speaking Latino parents (n = 21), English-speaking minority parents (n = 27), and English-speaking majority (white) parents (n = 60). The research involved cancer clinical trials. The investigators reported that the non-English-speaking parents were significantly less likely than the English-speaking majority group to receive explanations of the consent form, to be provided with as much detailed information or explanation about the trial, and to ask questions about the trial. For example, 48 percent of the non-English-speaking group received explanations of the consent document, whereas 92 percent of the English-speaking group did. In addition, independent raters assessed the explanations and found that they were less successfully provided to the former group than to the latter group. Not surprisingly, the investigators reported that problems with understanding were most frequent among the non-English-speaking parents. Only 14 percent of these parents understood randomization, whereas 60 percent of the English-speaking majority group did; 60 percent of the non-English-speaking group understood that research participation was voluntary, whereas 90 percent of the English-speaking majority group did. In their discussion of the findings, the authors encouraged investigation of the “role of interpreter accuracy and communication-fatigue, which commonly affects both interpreters and the original speakers” (Simon et al., 2003, p. 217).

Implications

Studies of adults asked to provide consent for research participation in their own right and studies of parents asked to provide permission for their child's research participation are generally consistent in a number of respects. They suggest that parents' understanding of research purposes is often incomplete and that the process for seeking parental permission should be improved. Particular care needs to be taken when seeking permission from parents who are just beginning to cope with a child's serious illness or injury and its treatment. In these circumstances, investigators and IRBs should consider structuring parental permission as a continuing process that is sensitive to parents' developing knowledge of their child's diagnosis and prognosis, as well as the treatment options. The process should also take into account how parents' emotional states may affect their understanding and consideration of options.

Given parents' usual reliance on and trust in their child's physician, research pointing to parents' willingness to defer to the physician's judgment about research participation is not surprising. It reinforces the need for investigators and IRBs to be particularly attentive to the potential for undue influence or conflict of interest when the child's physician is also the investigator responsible for seeking informed consent.

The needs of non-English-speaking parents warrant further attention. Qualified translators or interpreters should be provided, as discussed later in this chapter. Both the discussion and written materials should be adapted to the expected language spoken by the parents to be approached for particular studies as well as to other capacities of the parents.

For most research situations, the presentation of a consent and permission form is ethically secondary to the discussion about research participation. This discussion should include opportunities for the parents to ask questions and—when appropriate—probing by investigators so that they feel comfortable with the parents' understanding of the proposed research, its requirements, and the rights and protections available to the child and family (e.g., the right to withdraw from the research). The research reviewed by the committee does not explicitly identify situations in which the parents fail to protect their child adequately, but investigators must also be alert to this possibility.

In addition to materials describing an individual research protocol, research institutions, collaborative research groups, and research sponsors should consider developing more general guidance to help parents think about their child's participation in research. Box 5.2 lists some questions that might be included in such guidance. Parents may also be referred to more detailed guidance that has been developed for adults considering participation in clinical trials (see, e.g., ECRI, 2002).

Box Icon

BOX 5.2

Questions Parents May Want to Ask When Considering Their Child's Participation in Clinical Research. What is the purpose of the research? Who is paying for it? Where will the research be done? How long will it last?

CHILDREN'S AND ADOLESCENTS' COMREHENSION OF RESEARCH PARTICIPATION

At age 14, my parents brought me to the National Institutes of Health for care … I was encouraged to participate in clinical research…. I welcomed the chance to interact with CF [cystic fibrosis] physicians and researchers on a more professional level. … My initiation into clinical research as a teenager was one of deliberation and independence—the first time I became a partner in my health … It was a wonderful learning experience.

Suzanne Pattee, Cystic Fibrosis Foundation, 2003

Children and adolescents who participate in clinical research may have vastly different experiences and perceptions of their experiences depending on the nature of the research and their own medical, psychological, and social circumstances. Some may not really understand that they are participating in research or may be reluctant participants. Others—like Suzanne Pattee—may find it to be a positive experience that gives them a greater sense of control over their situation.

Although the literature is not extensive, the committee identified some research that is relevant to assessments of what children and adolescents understand about research participation and when and how their agreement to participate in research should be sought. A more extensive literature on children's general cognitive or decision-making capacities also has implications for these assessments.

Development of Cognitive Capacities in General

Studies continue to expand understanding of the patterns of cognitive development and variation in children and adults. Many psychologists agree, in general, that children's cognitive development follows a fairly regular course, although they disagree about many specifics. In a highly simplified account, as described by followers of Piaget,6 children first develop sensory and motor skills (from approximately birth to age 2 years) and then begin to use language and images (to approximately ages 6 or 7 years), develop certain concrete reasoning and problem-solving abilities (to approximately age 11 or 12 years), and, lastly, acquire more advanced capacities to think and reason about complex, abstract concepts, relationships, and processes (adolescence). Other psychologists put more emphasis than Piaget and his followers on the role of cultural and social factors in children's development, in particular, children's interactions with parents, other adults, and peers (see, e.g., Vygotsky, 1978; and Doolittle, 1997). Still others take an information-processing approach that focuses on maturation of the sensory capacities, encoding abilities (attentiveness to and retention of information relevant to a task), and information retrieval skills (Siegler, 1991).

Particularly for more abstract concepts and reasoning capacities, research suggests that both adolescents and adults vary considerably in their capacities. It also suggests that culture, context, experience and training affect both the pace and the eventual adult level of cognitive development and performance (see, e.g., Capon and Kuhn, 1982; Chi and Ceci, 1987; Nisbett et al., 1987; Chien et al., 1996; Morris and Sloutsky, 1998; Segall et al., 1999; and Klaczynski and Robinson, 2000; etc.). These findings reinforce the argument for viewing informed consent (as well as permission and, as appropriate, assent) as a process that should stress careful explanation; provide opportunities for questions; and take into account the language, cultural, and other characteristics of prospective research participants. Although IRB policies and other sources often require assent starting at age 7 years, investigators should treat children as individuals and consider when it may be reasonable and respectful to provide children under that age with simple information about what will happen.

Table 5.1 depicts Broome's summary of how the general theoretical and research literature on children's development relates to the more specific question of children's developing capacity to understand and make decisions about research participation. It refers to social and emotional as well as cognitive factors.

TABLE 5.1. Developmental Differences Influencing a Child's Participation in Research.

TABLE 5.1

Developmental Differences Influencing a Child's Participation in Research.

Understanding of Research Participation

[P]ediatric researchers should not underestimate the awareness and maturity that some children possess when addressing issues of concern to them.

Geller et al., 2003, p. 269

I would like to see the age limits completely scrapped and maturity brought in. As you grow up your age has a stereotype. I'm trying to escape from that stereotype.

Robin, age 13, (quoted in Alderson, 1993, p. 9)

Recent years have seen an increasing interest in determining what children and adolescents comprehend about research participation and using that knowledge to guide decisions about when and how to seek children's agreement to participate in research. The findings are not entirely consistent, and each study has limitations in terms of the methods and the populations studied that affect generalizations. For example, most of the studies involved healthy middle-class children, although some included children with different kinds of medical problems. The focus was usually on children's intellectual or cognitive capacities rather than their emotional state or development. One exception involves studies of gravely ill children (Bluebond-Langner, 1978; Bluebond-Langner et al., in press). The possible effects of mental disorders (e.g., depression) on children's involvement in discussions and decisions about research participation have received little attention.

Some studies that the committee reviewed presented children with hypothetical research situations; others took place in the context of actual clinical research or other psychosocial research (i.e., research not involving the study of research understanding as such). Studies varied in the complexity of the hypothetical or real situations facing children and also in the way in which information was presented (e.g., in writing or orally). Few studies have included children as young as age 7 years, the age at which many IRBs and others advise that assent first be sought (see, e.g., AAP, 1995).

Some of the research reviewed below examined children's understanding of or agreement to medical care rather than research participation. Findings from these studies are relevant to the extent that they relate to children's and adolescent's knowledge of their medical condition (if any) and their understanding of medical procedures or other dimensions of clinical care (e.g., the potential harms and benefits of different treatments) that are identical or similar to those used for clinical research.

Very little research has investigated the actual process of obtaining children's assent to research participation. One study, already discussed above in the section on parents, analyzed 85 informed-consent conferences that concerned research participation by children with cancer (Olechnowicz et al., 2002). (Under federal regulations, assent was not a necessary condition for this research to proceed.) Of the 34 conferences that concerned children over age 7 years, the child was present during 14. The mean age of these children was 14 years (range, 10 to 18 years); the 20 children not included were younger (but their average age was not reported). The study authors reported that the investigators conducting the conferences differed in how they approached the discussion; for example, whether they indicated that the decision should be a family decision, the parents' decision, or “up to” the child (p. 808). During the conferences, older children asked more questions than younger children. Most children, however, talked very little and real discussion was uncommon. The majority of the questions from children focused on the disease and the treatment.

Although it did not assess child or adolescent understanding directly and did not observe the process of assent, another study examined 70 informed-consent documents that were used with adolescents being recruited for genetic research (Weir and Horton, 1995). The investigators concluded that the documents were often confusing and unclear. This is consistent with the findings from evaluations of informed-consent documents for adults. The investigators also concluded that the documents did not reflect “developing ethical and legal standards” for research involving adolescents (p. 347).

Comprehension of Research Purposes, Characteristics, and Possible Consequences

I would like to see consent forms be presented in [a clearer] manner. Some of the terms are a little bit ambiguous, especially for the patient, and if you're going to be the one getting treatment, it would seem to me that you would like those things to be clarified.

Carolyn Brokowski, research participant (at age 16), 2003

As reviewed below, most studies of children's understanding of the purposes and procedures of research suggest that children under age 9 or 10 years have a limited ability to understand the purposes, risks, and potential benefits of research, especially more complex research. These younger children are better able to grasp the more practical aspects of research (e.g., what they are expected to do) than they are to understand the more abstract dimensions (e.g., expectations of confidentiality). Most studies and summaries conclude that by the age of 14 or 15 adolescents differ little from adults on measures of research comprehension, but some suggest caution about other dimensions of adolescents' understanding and judgment.7

Children Compared to Adolescents A few studies suggest that children as young as age 6 or 7 years can comprehend some aspects of research. One study that included 6- to 9-year-olds (studied in groups not individually) concluded that although older children comprehended more about a vaccine study, children as young as 6 years could ask what the investigators viewed as reasonable questions about the research (Lewis et al., 1978). They could comprehend certain basic information about what might happen to them in medical treatment or research and what they were expected to do. A study by Susman and colleagues (1992) of hospitalized young people ages 7 to 20 years did not find important differences between children ages 7 to 13 years and young people ages 14 to 20 years. Both groups did better in understanding concrete features of research (e.g., what they were to do) than in comprehending more abstract features (e.g., alternatives and the right to withdraw). The investigators concluded that emotional factors (e.g., anxiety and a sense of control) were more important than age or cognitive development in affecting the participants' understanding of research participation (Dorn et al., 1995).

Most studies, however, have reported differences in comprehension between younger and older children. For example, in assessing the results of a study of 102 children 7 to 18 years of age, Tait and colleagues (2003b) concluded that children under age 11 years had particularly limited understanding of the study protocol (which involved an anesthesia or a surgery study), the potential benefits, and their right to withdraw from the study. Another study (Nannis, 1991) reported that more than 84 percent of the fifth graders (approximately 11 years old) in the study correctly understood what research was, whereas only 35 percent of the third graders (approximately 9 years old) answered the question correctly. In a study that included only second, fourth, and sixth graders, Hurley and Underwood (2002) reported that few of the participants in their psychological study understood its primary goals, even after a debriefing. They speculate that their results may have reflected the complexity of both the study (which investigated how children coped with peer provocation) and the debriefing.

Broome and colleagues (2001) found that children younger than age 10 years could describe overall goals and potential benefits but did less well in describing research risks. (This study, which also involved a parent component, as discussed earlier, included 24 children and young people with leukemia and 10 children and young people with diabetes; participants were 8 to 22 years of age and were involved in various studies related to their condition.) The investigators also reported that both children and adolescents who could discuss the purpose of the research could not always “translate [that understanding] into an acknowledgement that they themselves were involved in a clinical trial” (Broome et al., 2001, p. 41).

In a study of 18 children ages 5 to 18 years who were involved in nutrition research that did not offer the prospect of direct benefit, Ondrusek and colleagues (1998) concluded that compared with older children and adolescents, children under age 9 years showed deficient understanding of the purpose of the study, its potential harms and benefits, and their right to withdraw. (Children under age 13 years were read information about the study; older participants received a written form with the same information. This difference in presentation might have affected the results.) Younger children were, for example, more likely than older children to relate the research to checking their own health rather than to gaining information that would help sick children. The authors concluded that “the current age of 7 years for initiating assent (in addition to parental consent) is possibly not appropriate and should be reconsidered” (Ondrusek et al., 1998, p. 158).

Based on a study of student understanding of research and research rights (which included 4th, 7th, and 10th graders and college students), Bruzzese and Fisher (2003) reported that a majority of students at each level answered the questions correctly. A sizeable minority of the fourth graders, however, appeared to be confused about the purpose of the research (which was to compare the ability of children in different age groups for their abilities to comprehend the meaning of research). Overall, the study reported a steady increase by age in student's comprehension of the purpose and nature of the study as well as the risks and potential benefits of the study.

In sum, most research indicates that children between ages 6 or 7 years and ages 9 or 10 years can understand the more practical features of research but show a lesser appreciation than older children of the more abstract features of research (e.g., understanding of risks). The committee notes the shortfalls in younger children's understanding of research but believes that evidence does, on balance, support their involvement in age-appropriate discussions of participation in research and their ability to dissent (as provided for by federal regulations) from participation in research that does not present the prospect of a direct benefit. Such participation respects children's developing maturity and emerging autonomy. This is consistent with the moral purpose of assent.

Younger and Older Adolescents Consistent with a continuing maturation of cognitive capacities, most research shows differences between younger and older adolescents in their comprehension of providing informed consent to participate in research. Research on differences between adolescents and adults is more mixed.

A 1981 study by Lewis (involving 108 adolescents in the 7th, 8th, 10th, and 12th grades) looked at adolescents' recognition of risks and future consequences in a hypothetical situation involving peer counseling about cosmetic surgery or a trial of acne medication. She found sizeable differences by age, with the older adolescents showing greater awareness of risk and future consequences. Older adolescents were also more likely to suggest the need for independent professional opinion about the situations presented to them. When Halpern-Felsher and Cauffman (2001) replicated the study, however, they did not find important differences by age in the ability to recognize of risks and future consequences. Their findings are consistent with the findings of studies by Kaser-Boyd and colleagues (1985) and Ambuel and Rappaport (1992).

In their study of competency to consent to medical treatment, which involved 98 participants at ages 9, 14, 18, and 21 years, Weithorn and Campbell (1982, p. 1589) reported that the 14-year-olds in their study did not differ from adults on four standards of competency: understanding, choice, reasoned outcome, and rational reasons. A number of summaries of the literature also suggest that by the age of 14 years, most adolescents possess the “psychological elements of ‘intelligent' consent” (Grisso and Vierling, 1978, p. 420; see also Weithorn and Campbell, 1982; Leiken, 1993; Weir and Peters, 1997; Thompson, 2000b).

Others are more cautious about the capacity of adolescents to evaluate participation in research. For example, after reviewing the few studies of age-related differences in judgments of risk, Millstein and Halpern-Felsher (2001) state that, depending on one's perspective, these studies could point either to a “heightened sense of vulnerability [among young people] compared to that among adults” or to a picture of adolescents being less accurate in their judgments than adults (p. 30). After noting the limitations of the available research (e.g., a lack of longitudinal studies that differentiate cohort from developmental influences), they conclude that “the finding of age-related increases in risk identification does call into question the degree to which we should consider adolescents, particularly younger adolescents, competent” (p. 24). Based on their review of the literature on the development of judgmental maturity, Steinberg and Cauffman (1996) concluded that evidence points to greater differences between early adolescents and mid- and later adolescents than between the two older groups. The authors suggested that more research is needed on the cognitive and noncognitive factors associated with mature decision making as needed.

In presenting initial results from an Australian study of 234 participants ages 12, 15, 18, and 21 years who were asked about several medical treatment vignettes, Bartholomew (1996) reported that 12-year-olds showed less evidence of decisional capacity than older participants. In contrast, 15-year-olds did not show significant differences from 18-year olds on measures related to comprehension and the amount of information sought. On other measures that related to reasoning capacities, however, 15-year-olds were more like 12-year-olds than they were like older participants.

A study by Bernhardt and colleagues (2003) of children's views of participation in a genetic susceptibility study suggested that the children (ages 10 to 17 years) did not initially appreciate the study's risks and potential benefits. When conversation encouraged them to “personalize” the possible consequences of genetic testing, they developed a fuller appreciation of the possible risks. This supports the emphasis on discussion in the design of the assent process.

Other research suggests that the development of decision-making capacities continues into adulthood, although “not all of these developmental changes point to increased rationality” (Millstein and Halpern-Felsher, 2001, p. 37). For example, certain biases in decision making may increase with age, including the tendency to reject evidence inconsistent with one's views and accept evidence that supports them (Stanovich and West, 2000; Millstein and Halpern-Felsher, 2001; but see also, Klaczynski and Gordon, 1996 and Klaczynski and Narasimham, 1998).

Although a number of studies indicate that adolescents differ little from adults in certain elements of competent decision making, research overall suggests that abilities to comprehend important dimensions of research participation continue to evolve during adolescence. As the research reviewed earlier in this chapter made clear, adults are not always well informed and often show deficits in understanding related to research participation, especially when they are under stress. Overall, the committee concludes that it is usually advisable for parents to be involved in decision making about research participation for adolescents. Adolescents' capacity to make informed decisions should not, however, be dismissed out of hand, especially when requiring parental permission would endanger adolescents or preclude their participation in research with important potential to benefit them or other adolescents in the future.

Comprehension of Rights in Research

Several studies suggest that younger children have less understanding than adolescents of their rights in research, including rights related to confidentiality and their choice to participate in a study (or withdraw from a study once it has started). For example, Bruzzese and Fisher (2003) asked children in the 4th, 7th and 10th grades to identify violations of research rights in several hypothetical vignettes. The youngest children were least able to identify such situations. In a study of 9-, 15-, and 21-year-old males, Belter and Grisso (1984) concluded that the 9-year-olds were less able to recognize violations of patient rights and less able to protect against such violations than the 15- and 21-year-olds.

With respect to confidentiality specifically, Hurley and Underwood (2002) found that sizeable minorities of second graders in their study did not understand the “fine points of confidentiality,” despite repeated explanations in different contexts that information from their study would not be shared with family or school personnel (p. 140). Fourth and sixth graders did better. In their study of children ages 5 to 12 years, Abramovitch and colleagues (1991) likewise found younger children to be more uncertain about confidentiality.

In contrast, Bruzzese and Fisher (2003) reported that a large majority of the students at each grade level (4th, 7th, and 10th grades) in their study understood confidentiality. In discussing the difference between their findings and those of Abramovitch and colleagues (1991), Bruzzese and Fisher (2003) suggested that their repeated explanations of rights—before students took the permission form home to parents and again before the testing began—may have contributed to their better results among the younger children. Hurley and Underwood (2002) also provided repeated explanations, without the expected effects, but the youngest children in the study by Bruzzese and Fisher were older than the youngest children in the two other studies.

On the question of what would happen if they wanted to stop participating in the study, the children in the different age groups studied by Hurley and Underwood (2002) had similar, high levels of accurate responses. In contrast, the study by Ondrusek and colleagues (1998) cited earlier found that children under age 10 years had a low level of understanding that they could withdraw from the study. Even the older children believed that their parents would not like them to withdraw. Abramovitch and colleagues (1991) also found that few children ages 5 to 12 years believed that they had the right to withdraw from the study. In the study by Brusseze and Fisher (2003), fourth and seventh graders were less likely than older adolescents to understand that they had the right to withdraw from the study, although the majority in both groups did indicate accurately that they had that right. The investigators noted that they had, on the basis of findings from previous research, made particular efforts to provide explicit explanations of voluntary participation and how to withdraw from participation.

Bruzzese and Fisher (2003) concluded that providing a brief lesson in the rights of research participants improved understanding of these rights for individuals in each of the grade levels studied as well as for college students. (The investigators compared the responses of the students receiving the research rights lesson with students receiving education about an unrelated topic.) Their findings suggest, however, that even young adults may not be aware of their right to ask questions of investigators. They may also be uncomfortable asking questions in some research situations.

Overall, Hurley and Underwood (2002) also concluded that processes for obtaining assent and debriefing children after research participation were “moderately effective” in helping children understand their rights as research participants (p. 139). Younger children understood less than older children. The authors cautioned that debriefing might “do more harm than good” for younger children who do not understand the information provided (p. 139).

Although her opinions were focused on medical care rather than research, the views of a 7-year-old on how hospitals could make children more comfortable suggest that some very young children do have concepts of responsible action by health professionals. As expressed to her mother (a newspaper columnist), the child set forth several rules (McIntyre, 2002; used with permission; excerpts not in original order):

  • “Don't surprise me. Tell me what you are going to do before you do it.”
  • “Ask my permission before you put any part of your body on mine.”
  • “Get down on my level. If I'm in the bed sit down.”
  • “Be honest. I get really upset … when they say something isn't going to hurt [and it does].”
  • “Always think of a less painful way of doing things.”
  • “Try to keep the doctors and nurses who come into my room the same.”
  • “Stop saying it's no ‘big deal.' It might not be a big deal to you because it's not happening to you.”

To summarize, findings about children's and adolescents' understanding of their rights in research are generally consistent with findings about their comprehension of research purposes and characteristics. Older children and adolescents understand more than younger children. Some studies indicate that careful explanations of research rights can help both younger and older children have an improved understanding of their rights as research participants.

Influence of Experience with Illness

Few studies have directly examined the influence of children's or adolescents' experience with illness on their understanding of research participation. The literature cited above in the general discussion of children's cognitive development suggests that culture, context, and experience as well as age can influence children's cognitive skills and understanding of the world.

In their research on children's experience of grave illness (e.g., cancer recurrence), Bluebond-Langner and colleagues (in press) found “evidence for children's ability to hold several, sometimes contradictory views of their illness and what a particular treatment, drug or trial would offer.” The children—and their parents—could recognize that their condition was not curable but still express belief or hope in the possibility of a cure. In the context of grave illness, the views of many children about participation in a clinical trial may be captured in one child's statement: “What choice is there? It is this or nothing.” The investigators reported that children's experiences with grave illnesses are more predictive of their understanding of their illness and prognosis than is their age. Earlier qualitative studies suggest that children who have serious chronic illnesses may develop an earlier conceptual understanding of death than healthy children (see, e.g., Bluebond-Langner, 1978; and Sourkes, 1995).

In the study by Broome and colleagues (2001) cited earlier, the investigators found that half of the 24 children with cancer indicated awareness that they were in research, whereas all 10 diabetic children knew they were involved in research and could more clearly differentiate between clinical care for their condition and the treatment-related aspects of the clinical trials in which they were participating. The authors link the differences in understanding to differences in experiences. The children with diabetes had lived with their condition for a longer period, were on a stable treatment regimen, and had been involved in multiple studies since the diagnosis of their condition. As an incentive for their participation in research, these children also received a gift certificate or payment at the end of the studies. In contrast, the children with cancer had more recently received the diagnosis, had little or no previous experience with research, were presented with the option to participate in research soon after receiving the diagnosis under highly stressful circumstances, and were never offered financial incentives to participate in the research. These children were often involved in several studies, and the studies were typically simultaneous and related to each other and their clinical care. Some of the children with cancer were in data collection studies that did not involve an experimental intervention, and some of them may not have understood that this data collection was research. In addition, several of the children with cancer “were removed from the discussions” surrounding their enrollment in research because of pain or the severity of their illness (p. 44).

Parental Influence

As children develop, one mark of increasing maturity is an increasing desire for autonomy in decision making. Voluntary choice and independence in thinking are important dimensions of informed consent. The influence of parents appears to be strongest for younger children but parents influence adolescents as well.

In a study with individuals ages 9 to 10, 14 to 15, and 21 to 25 years that used hypothetical vignettes about medical treatment, investigators reported that considerable deference to parents' preferences was evident at all ages (Scherer, 1991). The youngest children, however, were the most deferential, particularly for vignettes involving more severe and complicated situations. The children gave various reasons for their deference, including respect for parental judgment and knowledge, need for parental support, desire to avoid conflict, and a belief that they had no choice. Based on responses of just the 14- and 15-year-olds in this study, Scherer and Reppucci (1988) reported that deference varied depending on the degree of parental pressure depicted. Abramovitch and colleagues (1991), in their study of participation in psychological research by children ages 5 to 12 years, found children across this age spectrum were concerned about withdrawing from research for fear of disappointing their parents and health care professionals.

Based on their interviews with children with cancer and children with diabetes, Broome and Richards (2003) reported that nearly all the children had faith that their parents would protect them. Responses were more mixed on whether parents would support children's choices about research participation, although most thought their parents would take their views into account.

In their study cited earlier, Geller and colleagues (2003) suggested that the older and more mature the child the greater the likelihood that decisions would be made jointly rather than by the parents alone. The children and adolescents that they interviewed generally believed that the decision to participate should be the child's, “unless they're like babies” (p. 264). All, however, wanted to know what their parents thought. A minority (all younger girls) said that they would defer to their parents.

Reporting on a study with 37 adolescents and their parents, Brody and colleagues (2003) found that the two groups agreed approximately three-quarters of the time on decisions involving an asthma research vignette. Both, however, claimed that they had “ultimate responsibility for the participation decision” (p. 79).

Bluebond-Langner and colleagues (in press) have stressed that one complication in determining what gravely ill children understand or want with respect to treatment or participation in research is that these children, in particular, may not express their concerns, wishes, or fears to their parents, clinicians, or investigators. This reticence may reflect not only the children's deference to their parents but also the children's desire to protect their parents from the additional distress of knowing their child's fears or facing the mutual acknowledgment of the prospect of death. The investigators refer to a process of “mutual pretense” that constrains free discussion and questions about treatment or research in the context of grave illness.

Children's dissent or unwillingness to agree to research participation is not well studied. Citing the work of Abramovitch and colleagues (1991), Thompson (2000b) notes that children may find meaningful dissent difficult “not only because of limitations in judgment, but also because their invitation to participate typically occurs in a context of prior parental permission, institutional support (whether the institution is a school, childcare center, hospital, or other setting), and adults' interests in furthering the research enterprise” (pp. 164-165).

When a parent believes that the child should participate in research for altruistic or other reasons and the child does not want to participate, the actual decision about research may involve sensitive discussions with the parent and child. As long as the child is not coerced into assenting, it is reasonable for parents to engage in persuasive discussion, for example, about the importance of helping others. Even young children know what it means to “help” someone, although the ability of children to reason abstractly about altruism develops with age. Children also show empathy, sympathy, and other traits associated with the moral motives for adult altruism (Hoffman, 1990; Eisenberg et al., 2002; Bernhardt et al., 2003). Nonetheless, consistent with the regulations, the child's dissent should override parental assent when the research does not promise direct benefit to the child.

Implications

I think that it has to be for children a very interactive process in terms of consent. After I was asked this entire list of questions, I began to question my own first response. I think that just giving a kid a piece of paper, no matter how comprehensible the piece of paper is, is not going to be effective. I think the kid needs to be prompted with questions because I would never have thought of those issues, even as a 13-year-old… . When I was asked by the researcher did you consider this, what about these consequences, I began to change my idea of whether I would really want to participate in that study.

Sarah Lippincott, research participant, 2003

Although research is limited and findings are not entirely consistent, they support a gradual expansion of the involvement of children in discussions and decisions about research participation. For younger children, the emphasis should be on providing basic information about what will happen, responding to their questions and concerns (including those not explicitly expressed), and—particularly when the research does not offer the prospect of direct benefit—recognizing when children do not want to participate. Even when federal guidelines do not require the child's assent, investigators should still inform children (as appropriate given the circumstances) about what will happen and answer any questions that the child may have.

As children mature, they can participate more fully in discussions and decisions about their participation in research, although the involvement of parents is still required and prudent in most situations. Older adolescents may not have the legal capacity to make decisions in their own right, but the research reviewed here generally suggests that their level of comprehension of research approaches that of adults. Many aspects of the assent process can be similar to the consent process for adults if that process has been designed to accommodate people of various educational, social, and cultural backgrounds. For children and adolescents as well as adults, physical incapacity or distress related to serious illness may limit the extent to which they are involved.

Fisher (2003) has described a general approach to assent that aims for a “goodness-of-fit” between children's maturing skills and the assent processes. Because children have limited experience exercising their rights in response to requests from adults (especially in educational, medical, or unfamiliar settings), the assent process should be designed to demonstrate that participation is voluntary and dissent will not be penalized. In some cases, tutorials on research procedures and protections may be an appropriate element of the assent process. Opportunities for supported decision making that involves a parent-child discussion can also be created; a facilitator may be helpful in such a discussion. The broad objective is to create assent contexts and processes that minimize stress, encourage children's involvement in decisions about their participation in research, and ensure that their wishes and concerns are adequately communicated and considered.

Based on their analysis of findings related to both parents and children, Broome and colleagues (2003) similarly argue for variations in the permission and assent process to fit the circumstances, particularly in studies involving adolescents. They conclude that the explanations of research should be provided separately to adolescents and their parents and that parents and adolescents should, when feasible, be provided more time (hours to days) to consider participation in clinical trials related to serious illness. For adolescents and their parents, agreement to research participated should be regarded as a “shared endeavor … based on implicit, developmentally based negotiations” (Broome et al., 2003, p. S23).

Among the sensitive issues related to research participation is when adolescents should be free to consent to research participation without parental permission. As noted above, although the DHHS regulations make provisions for the waiver of parental consent when appropriate, the FDA regulations do not. The research reviewed here suggests that the DHHS regulations reasonably provide for waivers, as long as safeguards are provided that are appropriate for the individual adolescent being considered for research participation.

IRB AND INVESTIGATOR POLICIES AND PRACTICES

Information about IRB and investigator policies and practices related to parental permission and child assent to research participation is limited. McWilliams and colleagues (2003) analyzed information from 31 (of a total of 42) cystic fibrosis care centers related to IRB review of a genetic epidemiology study that included children. The genetic study involved the collection of data from medical records and the analysis of a blood sample. Twenty-one of the centers reported that investigators were required to obtain assent from children, but 10 centers reported that assent was not required. Of the 21 centers obtaining assent, 10 used a separate assent form with an explanation of the study, whereas the rest provided for initials or a signature on an adult consent form. Some centers used distinct parental permission forms. Local templates often did not include information appropriate for genetic studies (e.g., assurances of confidentiality for family members). Few of the approved forms included all the information that has been recommended for inclusion by guidelines on the banking of genetic information. The researchers noted criticisms of IRB preparation to deal with genetic research, quoting Frances Collins's statement that the “IRB Guidebook is dusty and out of date for genetics research” (McWilliams et al., 2003, p. 364, citing Collins, 2001, online, unpaged).

Mammel and Kaplan (1995) undertook a national survey of IRB chairs to describe policies for obtaining the agreement of adolescents to participate in research; they received usable responses from 30 percent of the 600 IRB chairs contacted. Of those responding, 70 percent said that the IRB required parental permission for all minors involved in studies; that is, they did not allow waivers of parental permission for certain kinds of research involving adolescents. IRBs that reviewed more than 10 protocols involving “adolescents” a year were less likely to have this blanket requirement (p. 323). Approximately one-third of the respondents said that their IRB would waive parental permission for a study that involved anonymous testing of adolescents for HIV infection.

A statement to the committee by the Allergy and Asthma Network included observations on the experiences of one office-based researcher, who reported that sponsors often did not provided assent materials and that protocols were “often approved without an assent document being reviewed by the IRB” (White, 2003, p. 7). In this physician's view, adolescents generally needed a simplified version of the consent document to help them understand the research.

Several studies have documented considerable variability in the practices of IRBs related to informed consent. In a study supported by NIH, researchers surveyed the chairs of nearly 500 IRBs that had conducted at least 10 protocol reviews in 1995 (Bell et al., 1998). More than a third of respondents reported that their IRBs had suspended or terminated a research project; of this group, just over 20 percent of the actions had been based on investigator failure to obtain informed consent. When questioned about their priorities for the initial review of protocols, about 20 percent of IRB chairs indicated that consent forms were a priority and another 11 percent mentioned informed-consent procedures. When questioned about deficiencies in protocols during the initial review, 60 percent of IRB chairs said that excessively technical language in the consent form was often a problem. Less than 10 percent said that they often saw problems with other aspects of consent forms (e.g., understatement of risks or benefits) or with the consent process. More than half, however, said that they sometimes found problems with consent processes that did not “promote comprehension” (p. 62). For multisite studies that originated at another institution, more than half the chairs reported that their IRBs had required modifications in the (model) consent form. Almost 90 percent of the chairs and administrators said that the failure of investigators to provide an acceptable consent form with their initial submission was a problem. “Issues relative to assent procedures” had been encountered by almost 90 percent of the IRB chairs (p. 22).

For the survey just cited, Bell and colleagues (1998) also questioned investigators and administrators associated with 300 IRBs. Among the investigators who reported that they had modified a research protocol in response to IRB concerns, more than three-quarters reported that they had to modify the consent form and one-fifth mentioned that they had to modify the consent process. Responses from investigators questioned about informed-consent practices indicated that the median time spent in discussion related to consent was 20 minutes (mean of 30 minutes). Nearly one-third of the investigators reported that they spent 10 minutes or less on consent. Investigators from institutions with high volumes of research spent more time, on average, in these discussions (means, 34 versus 25 minutes). About half the investigators indicated that they shared responsibility for discussions of consent with other members of the research team, whereas a third took sole responsibility for that activity.

In a study of 16 institutions involved in a single multicenter trial of treatment for patients with acute lung injury, Silverman and colleagues (2001) found that only 3 institutions had prepared forms for the study that included all the information that is specified in the federal regulations. The reading levels for the forms varied from the 8th to the 13th (college freshmen) grade. One of the 16 IRBs waived the requirement for informed consent.

Burman and colleagues (2003) used independent reviewers to evaluate changes made in consent forms for two multicenter tuberculosis studies. The researchers concluded that 85 percent of the changes altered “wording without affecting meaning,” that 11 percent of the changes introduced errors, and that two-thirds of the locally approved forms had at least one error related either to the protocol or to federal regulations (Burman et al., 2003). Another study that involved 44 institutions reported that only eight applications for IRB approval were approved without changes and that 91 percent of the changes requested in the other applications involved consent forms (Stair et al., 2001).

In its summary of common findings of noncompliance after investigation of complaints, OHRP listed several topics under the heading of informed consent, including failure to obtain legally effective consent and various deficiencies in consent documents (OHRP, 2002d). None of the listed findings explicitly mentioned children's assent or parent's permission.

Overall, these surveys and studies suggest room for improvement in the processes for obtaining informed consent, permission, and assent. The following discussion focuses on steps to strengthen the processes related to permission and assent.

IMPROVING PROCESSES FOR REQUESTING PERMISSION AND ASSENT

Accomplishing the goal of real understanding as a precondition to a meaningful decision to participate [in research] will require a sea change in [review board] and investigator perception and practice.

Institute of Medicine, 2003a

Informed consent is widely regarded as a cornerstone of ethical research. Because children usually do not have the legal capacity or maturity to provide informed consent, the concepts of parental permission and child assent have been developed as a foundation for ethical research involving children. (Again, in this report, the term parent covers guardians as well.)

Taken together, a considerable body of research—albeit imperfect— points to important shortfalls in the current processes for seeking permission and assent. It also points to some directions for improvements, while underscoring the challenges of securing informed agreement, especially under conditions of stress, desperation, and crisis. On the basis of its experience and judgment as well as its review of the research literature, the committee developed the conclusions and recommendations presented below.

Focus on Permission and Assent as Processes

We have to make sure that in our attempts to protect children we aren't making the consent process so onerous and paper heavy that we're actually prohibiting or inhibiting new studies from opening, or conversely scaring families away.

Lise Yasui, parent, 2003

As discussed earlier, informed consent—and, by extension, permission and assent—should be viewed as a process, not a form. Committee experience and the limited evidence about IRB practices suggests reason for continued concern that investigators and IRBs devote disproportionate attention to consent forms, while paying less attention to the overall process for seeking consent and ensuring, insofar as possible, that consent is informed, reasoned, and voluntary.

This committee reiterates the importance of placing more emphasis on the design and implementation of ethical processes for seeking parents' permission and children's assent to a child's participation in research. Discussions about research participation should allow sufficient time for questions and, if necessary, further explanations. Such discussions should always precede the presentation of forms. Before any forms are presented for signature, investigators and others involved in the process of requesting permission and assent should feel comfortable that the parents and, when appropriate, the children have an adequate understanding of the course to which they are agreeing.

Recommendation 5.1: To focus attention on the process of requesting parents' permission and children's assent to research participation, investigators should provide and IRBs should review protocol descriptions of

  • who will request permission and assent;
  • how and when permission and assent will be requested;
  • who should be contacted if parents have questions or concerns about the research; and
  • for studies that extend over considerable periods of time, when and how permission and assent may be requested again, for example, as children reach important developmental milestones.

A one-time process for requesting permission and assent may be all that is needed for many studies, for example, those involving a single encounter or a brief hospital stay and routine follow-up. Other studies should provide for periodic revisiting of the parents' and children's agreement to the child's participation in research. This may occur yearly or more frequently, depending on the nature of the research and, possibly, changes in the child's health status. For example, if children come in for testing every 6 months, particularly for research that does not have the prospect of benefiting them directly, it usually will be appropriate to check that they want to continue to participate. (Some children will signal their unwillingness to continue by persuading their parents not to bring them in for the testing.) In addition, if the research changes in ways that could affect the parents' and children's assessment of the risks in relation to the potential benefits, permission and assent should also be sought again.

For children with serious medical conditions, the initiation of permission and assent procedures for clinical research is often inextricably intertwined with decisions about basic clinical care. Discussions of research participation may occur close in time to a family's first being told of the child's diagnosis or learning about a setback in the child's response to treatment. These discussions may also cover acute care that is not part of the research protocol. In these circumstances, permission and assent should be conceptualized as continuing processes that are sensitive to parents' developing knowledge of their child's condition and their shifting emotional reactions to the diagnosis, prognosis, and treatment options. The processes also should be sensitive to changes in the child's cognition, emotions, and physical health that may alter the balance between research risks and burdens in relation to potential benefits. Such changes may also affect whether it is possible or appropriate to involve the child in discussions and decisions about research participation.

Recommendation 5.2: When appropriate for research involving children with acute illnesses or injuries, investigators and institutional review boards should provide for ongoing processes for permission and assent that will accommodate a family's evolving understanding of the child's condition, the child's emotional state and decision-making capacity, and the child's changing medical and psychological status. These processes are not matters of signing or updating forms but, rather, of continuing communication based on appreciation of the difficult and even overwhelming circumstances in which parents may be asked to make grave decisions about their child's future.

Informed Parental Permission

Parents will inevitably vary in their levels of understanding of their child's health status, the difference between clinical care and clinical research, and other matters critical to informed decision making. To some extent, but not entirely, this variation may reflect differences in educational and reading levels and, possibly, proficiency in English. Sociocultural differences may also shape how parents understand and engage in the process of deciding about their child's participation in research. Regardless of the study population, consent and permission forms and processes should avoid language that creates or reinforces confusion about the difference between research and clinical care.

In designing and reviewing procedures and related written materials for seeking parental permission, investigators and IRBs should consider what is known or expected about the educational, language, cultural, and other relevant characteristics of the populations to be involved in particular studies. Studies consistently indicate that informed-consent materials are geared to individuals with reading skills at levels higher than those that many prospective research participants or the parents of prospective child participants possess. Some research also suggests that most information in consent documents can be presented in much simpler language that will still convey information essential for decision making. For groups with low levels of literacy, investigators and IRBs should consider the use of information tools (e.g., simple graphics and oral explanations or special videos) that do not depend on written documents. Again, investigators should view the presentation of consent, permission, or assent forms as the end of the process, not the beginning.

When language barriers exist, investigators should see that the consent forms translated from English to other languages are adequate and that the content is equivalent to that in the original. As necessary for a meaningful process of seeking parental permission, investigators should use interpreters who have the knowledge, language skills, and awareness of cultural factors that are necessary both to translate information about clinical research and to assess the level of understanding of that information by the participant of the participant's parents. In support of effective communication with patients, accreditation standards and various federal and state policies now expect hospitals to have available interpreters of the languages frequently used by their constituent groups of patients (see, e.g., Perkins et al., 1998; OCR, 2000; and Somers, 2003). Except in a clear emergency when no acceptable alternatives exist, children should not serve as interpreters for their parents. The practice may compromise parent-child relations and increase the risk of misinformation.

Investigators using translators in hospital and other settings should also be sensitive to issues of confidentiality. Particularly in small ethnic communities, those who can appropriately act as interpreters may have business or social relationships with prospective research participants and their families (Fisher et al., 2002).

The committee recognizes the cost implications for clinical studies of provisions for adequate translation and interpretation services and materials otherwise appropriate to the language skills of target research populations. To support investigators in meeting their ethical obligations as part of the permission and assent process, research sponsors should recognize these costs.

Recommendation 5.3: Investigators—with assistance and oversight from institutional review boards, research institutions, and research sponsors—should design procedures for seeking parental permission for a child's participation in research that are sensitive to educational, cultural, and other differences among families and include provisions for

  • educating—not merely presenting information to—parents about issues critical to informed decision making and, as appropriate, assessing the degree to which these critical issues are understood;
  • writing consent and permission materials in the simplest language that still conveys essential information about the study; and
  • providing competent, trained translators and interpreters, when needed, and otherwise assisting parents with limited English-language proficiency with making informed decisions.

Particularly when a child has a serious medical condition, parents may be concerned that their child's access to care will be negatively affected if they do not agree to their child's participation in a clinical study. Even though it is not the intent of the investigator, the result may be permission that is less than fully voluntary. Investigators should work with IRBs and research institutions to develop safeguards to ensure that parents understand that their child's normal access to clinical care will not be disrupted if they choose not to permit their child's participation in research or if they choose to withdraw the child after the start of a study.

When no effective standard therapy exists or standard therapies have failed, parents may see participation in a clinical trial as the only option offering any prospect of extending or improving their child's life. In these situations, investigators still have an obligation to discuss the expected or possible burdens that research participation may impose on the child, to offer their best assessment of the prospect of benefit, and to explain what relevant options for palliative care may be available within or outside the research setting.

Waiver of Parental Permission

As described earlier and consistent with the recommendation of the National Commission, DHHS regulations permit IRBs to waive parental or guardian permission for a child's participation in research when such permission (1) is not a reasonable requirement to protect the child and (2) is not inconsistent with federal or state laws. The regulations also require the substitution of an appropriate protective mechanism when a parental waiver is approved. The FDA regulations do not allow such waivers, even when the research involves, for example, abused or neglected children who would thereby be denied the potential benefits of research participation.

For a variety of reasons, it is generally desirable to seek parental involvement in decisions about children's and adolescents' participation in research, especially when the research involves a clinical trial or presents more than minimal risk. In some situations, however, requiring parental permission could put a child or an adolescent at risk of violence, expulsion from the home, or other harm. This could impede some important research. For example, researchers trying to identify risk factors for HIV infection among different groups of adolescents might be successful in encouraging some teens to involve their parents but be unable to recruit those teens who are most fearful of violent parental reactions. Such reactions may be part of a family dynamic that is, itself, a risk factor.

In addition, some adolescents have no parents or guardians and are not wards of the state, and others are runaways or “throw-aways” (adolescents who are evicted from or asked to leave the home by their parents) (Blustein et al., 1999). Research to understand the circumstances of these adolescents—including their health status—is unlikely to be feasible if parental permission is required. Likewise, it may not be possible to enroll these adolescents in clinical studies that have the prospect of direct benefit.

In some cases, an adolescent for whom parental permission is not feasible may be able to receive an investigational product outside a trial if the product (e.g., an antidepressant) is approved for other uses or age groups. (Such use will still require either parental consent or a judgment that such consent is not required under applicable laws.) Although this approach may, on balance, be prudent and may potentially benefit the individual adolescent, it will not advance knowledge to benefit future adolescents.

Committee experience and some research suggest that many IRBs are reluctant to consider waivers of parental permission for adolescent's participation in research even though the proposed waivers are consistent with DHHS regulations and relevant state laws. Restrictive IRB policies and practices can be just as constraining as current FDA policies, thereby limiting research that can benefit adolescents.

Again, parental involvement and permission is usually advisable, especially for research involving more than minimal risk. IRBs should not, however, employ a blanket policy of denying protocols that provide for waivers of parental permission. Rather, they should consider protocols on their merits taking into account their purposes and rationales (e.g., parental neglect, fear of parental violence), the basis for assessing the adolescent's capabilities to make decisions, proposed safeguards, and applicable federal and state policies.

Recommendation 5.4: Institutional review boards should consider granting waivers of parental permission for adolescent participation in research when

  • the research is important to the health and well-being of adolescents and it cannot reasonably or practically be carried out without the waiver (consistent with 45 CFR 46.116(d) and 45 CFR 408(c)) or
  • the research involves treatments that state laws permit adolescents to receive without parental permission (consistent with the definition of children at 46 CFR 402(a)) and when
  • the investigator has presented evidence that the adolescents are capable of understanding the research and their rights as research participants and
  • the research protocol includes appropriate safeguards to protect the interests of the adolescent consistent with the risk presented by the research.

The committee agrees with the National Human Research Protections Advisory Committee (NHRPAC) (Marshall, 2001) and the Society for Adolescent Medicine (SAM, 1995) that FDA's restrictive policy hampers important research involving adolescents. As discussed earlier in this chapter, the committee believes that FDA's statute provides discretion when parental permission is not feasible or it is contrary to the child's best interests. Furthermore, as NHRPAC has noted, the FDA regulations define children with reference to state laws that apply in the locale where the research takes place, which reasonably means that the age of consent (i.e., when parental permission is not required) depends on the law of that state. States also permit minors to consent to treatment in their own right under a range of circumstances, although these provisions do not explicitly mention research.

Recommendation 5.5: The Food and Drug Administration should adopt policies consistent with federal regulations at 45 CFR 46.408(c) that allow institutional review boards with appropriate expertise to waive requirements for parental permission in research, provided that additional, appropriate safeguards are in place to protect the child's or the adolescent's welfare.

OHRP, FDA, and NIH should cooperate to develop guidance for investigators and IRBs on the additional safeguards that should be in place when a waiver of parental permission is sought and obtained. The safeguards should fit the risks associated with specific studies. In making recommendations about health research with adolescents that involves only minimal risk, the Society for Adolescent Medicine advises that IRBs should find that the research meets several conditions (SAM, 1995). Box 5.3 provides an example of questions devised by one IRB based on this advice. Beyond these questions, IRBs and investigators should be familiar with state policies relevant to the medical treatment of minors without parental permission.

Box Icon

BOX 5.3

Questions and Information for Investigators Seeking Waiver of Parental Permission for Adolescent's Participation in Research. Please specify why the research could not be practically conducted without a waiver and why parental permission is not a reasonable (more...)

In some cases, safeguards might involve a research ombudsman or independent advocate who would review the adolescent's situation, identify relevant state laws, observe or participate in the consent discussion (including, unless inappropriate in the circumstances, discussion of the desirability of parental involvement), perhaps hear the views of a nonparent relative or other adult who is providing support and guidance, and document the reason for waiving parental permission. IRBs should consider the guidelines from the Society for Adolescent Medicine. For research involving more than minimal risk, they advise (among other protections) that IRBs determine that someone other than the investigator will assist the adolescent by identifying an adult familiar with his or her situation, committed to his or her well-being, and willing to offer appropriate emotional support. Further, the Society recommends that an appropriately trained professional (e.g., psychologist or masters level social worker), who is not involved in the study should confirm

the capacity of the adolescent as a mature minor to give informed consent by finding evidence of: (a) cognitive ability to comprehend the objectives and requirements of the research and other important considerations (e.g., the voluntary nature of participation, the potential of risks and benefits) as would be required for a competent adult; (b) reasonable judgment as evidenced by the ability to address problems, to foresee the long-term consequences of action or inaction, and to evaluate the validity of information; and (c) personal responsibility to be able to comply with the requirements of the research protocol, especially those designed to ensure individual safety. (SAM, 1995, p. 266.)

Finally, it is important for regulators and institutions to develop appropriate policies and procedures for waiving parental permission and child assent to permit research involving children who have acute, unanticipated health problems (e.g., extreme prematurity or head trauma) when parents are not immediately available or not medically competent to provide permission and when the child is incapacitated. Consistent with OHRP and FDA policies and regulations on emergency research, IRBs should develop policies and procedures for simultaneously protecting the well-being of the affected children and the rights of the parents while still permitting the acquisition of new knowledge concerning the most frequent causes of child morbidity and mortality in the United States, which are—for children beyond infancy—intentional and unintentional injuries. The committee recognizes that many institutions have been unwilling to consider such research out of concern about adverse outcomes and the concurrent legal liability associated with these outcomes.

Children's Assent or Dissent in the Context of the Family

Rather than viewing the process of requesting a child's agreement to participation in clinical research as an independent event, investigators and IRB members should adopt a family systems perspective. In the committee's experience and as indicated in some of the research reviewed above, parents often want to decide whether their child will be approached about research participation. Children are usually not approached unless parents are willing to consider the child's participation.

In some cases, parents may agree to a child's participation in research with the expectation that the child, once approached, may not want to participate and that the child's wishes will be accepted. In other situations, parents and children may negotiate decisions about research participation, with the views of adolescents tending to have more weight. Sometimes, parents may expect their wishes for the child to participate in research to prevail over a child's wish not to participate in research, especially when the research has the prospect of benefiting an ill child. Federal regulations governing research involving children appropriately do not require the child's assent when research offers the prospect of an important direct benefit that is available only in the context of the research” (45 CFR 46.408(a); 21 CFR 50.55 (c)(2)). Investigators should avoid suggesting to children that they can make the participation decision when their parents can override their wishes.

Nonetheless, even when it can be overridden, a child's dissent should not be regarded as simply inferior to the parents' views and decisions but, rather, should be viewed as reflecting a different and still important point of view. Investigators or others involved in the assent process should work with parents to help them understand children's views and treat them respectfully. It will usually be better for all involved if agreement between parent and child—especially an older child or adolescent—can be achieved. This may involve a developmentally variable process of negotiation. As discussed earlier, it may sometimes be reasonable—as part of a respectful and noncoercive assent process with parental agreement—for investigators to come back to a child who says “no” to see whether he or she may feel differently later.

In rare instances, understanding and respect may lead an investigator or clinician involved with the child's care to advocate—with great sensitivity to the family's wishes and values—on behalf of a dissenting older child or adolescent when the assent process reveals compelling reasons for encouraging parents to consider accepting rather than overriding the dissent. One such situation may arise when a mature adolescent who has a grave medical problem that has been unresponsive to treatment asks not to be subjected to a burdensome investigative procedure that offers at most a remote prospect of meaningful benefit. When parents are desperate to do something or fear feeling guilty in the future if they refuse proposed interventions, they may find it difficult to fully consider the burdens that research participation is likely to impose on their child.

When research does not have the prospect of direct benefit to the child, the committee interprets the regulatory requirements that “adequate provisions are made to solicit the child's assent” to mean that the child's dissent overrides parental permission. Investigators should be alert to nonverbal indications of children's wishes about research participation, particularly when children may be reluctant to distress their parents by voicing disagreement. The federal regulations specify that failure to object to participation should not, by itself, be construed as assent. Again, in the committee's view, a child's dissent does not preclude investigators seeking the child's assent at another time, as long as such practices are respectful and not coercive and parents are involved and supportive.

Some research has the potential to reveal sensitive information about a child or, more often, an adolescent that might provoke negative, even punitive, parental reactions. Investigators should be sensitive to this potential and should determine when adolescents and possibly younger children should be told about the kinds of information that may be shared with their parents if they agree to participate in research. Researchers studying adolescents have been concerned that the federal privacy regulations (see Appendix C) may allow parents to have access to information about adolescents that would otherwise be held confidential. The regulations defer to state laws on parental access to the health records of minors, which means that research-related personal information may be protected in some states and not in others.

Recommendation 5.6: In designing and reviewing procedures for seeking a child's assent to participation in research, investigators and institutional review boards should aim to create assent processes that consider and respect the child and the family as a unit as well as individually. The process for requesting assent should

  • be developmentally appropriate given the ages and other characteristics of the children to be approached;
  • provide opportunities for children to express and discuss their willingness or unwillingness to participate;
  • clarify for parents and children (as appropriate) the degree of control that each will have over the participation decision; and
  • when appropriate, describe to children and parents the kinds of information about the child that will or will not be shared with the parents.

Age-Appropriate Assent Processes

The construction of age-appropriate assent materials and procedures should be informed by the literature on cognitive development in general and on child and adolescent capacities for assent in particular. For children and adolescents at all ages, the assent process should be designed to be an empowering and respectful experience.

The research on children's cognitive, intellectual, social, and emotional development and their understanding of research participation provides somewhat mixed directions on the question of a minimum age for seeking affirmative assent. As long as investigators and IRB members recognize that the assent process for young children should be age appropriate and quite simple, it is not unreasonable for investigators to be asked to seek assent for children as young as age 7 years. (Such an age-defined policy may be useful for practical reasons, but investigators should also be encouraged to provide simple explanations for children younger than that age who seem to be capable of understanding basic information.)

For younger children (to age 9 or 10), assent information and procedures can be limited and can focus on a subset of the information required for adults. For example, the discussion and, if appropriate, written material or assent form for younger children should explain in simple language

  • what the study is about and whether it might help (e.g., “We want to see whether a new medicine will or won't help children like you who have earaches.” or “We want to understand your illness better.”);
  • what will happen and when (e.g., “You will have to come to the clinic once a week for 8 weeks to have a shot, you will not be allowed to eat for 12 hours, and then the following things will happen: …”);
  • what discomfort there might be and what will be done to minimize it (e.g., “Your head may hurt after the test and you may need to stay in bed for a day.” “You will get a cream on your skin that will keep you from feeling the shot.” or “You will have someone to whom you can talk about your feelings.”);
  • who will answer the child's questions during the study; and
  • whether an option to say “no” exists (e.g., “You do not have to be in this study and no one will be mad at you.” or “If you say “yes” and then change your mind, that is okay.”).

With the increasing age of the child, assent information should become correspondingly more substantive and specific and should use language appropriate to the cognitive and emotional maturity of the child or adolescent. The nearer an adolescent is to the age of majority, the more the assent process should resemble a well-designed informed-consent process for adults. Most research suggests that by age 14 or 15 years adolescents are similar to adults in their ability to comprehend the meaning of participation in research, although they may still be maturing in other areas of understanding and decision making. If consent forms for adults are written to be easily read by those with reading skills at less than a high school level, they may be appropriate to use as the assent form for middle to older adolescents. As for adults, in designing assent processes and written information for children and adolescents, investigators should take into account the expected language competencies and preferences of the groups to be included in the research. For emancipated and mature minors who are making decisions in their own right, the process for seeking assent should still be sensitive to the level of maturity of these adolescents (e.g., their attitudes about risk taking) and appropriate safeguards should be provided.

Discretion in Documenting Assent

As described above, federal regulations permit oral presentation of assent information and allow the IRB to use discretion about the way in which assent is documented. Although the guidelines on pediatric drug research adopted by ICH (and issued as FDA guidance) are more specific (requiring signatures on assent forms for participants of “appropriate intellectual maturity” [FDA, 2000b, p. 13]), they, too, allow some discretion and should not be rigidly interpreted.

On the basis of the literature reviewed in this chapter, the committee suggests that requiring a child's signature on an assent form may be developmentally inappropriate for younger children (e.g., those under age 9 or 10 years). A signature requirement may confuse rather than inform a child, and a child may even experience this situation as intimidating or coercive.

Recommendation 5.7: Guidance and education for investigators and members of institutional review boards should make clear that federal regulations allow discretion—based on children's developmental maturity—about the way in which information is presented to children and the manner in which assent is documented. Investigators and institutional review board members should apply that knowledge in determining what procedures will best serve the goals of assent for particular research protocols and populations.

While allowing for discretion with individual children or adolescents, IRBs and institutions can provide guidance about age-appropriate options for documenting assent. For example, in addition to developing sample age-appropriate information and assent materials for younger children, institutions can suggest that investigators use their judgment in determining when it is appropriate and respectful to ask younger children to sign simple assent forms. If investigators have developed an adult consent or permission form that is suitable for use with adolescents, the form could be used to document adolescent assent. Alternatively, then adolescents could sign a simpler assent form that includes age-appropriate information beyond that included on the form for younger children.

Investigator Knowledge and Skills

Developmentally appropriate and family-sensitive procedures for obtaining children's assent and parents' permission for a child's participation in research require communication skills and knowledge of child psychology and family dynamics. Education in these skills may not be part of the standard training for clinical researchers, including some investigators who conduct studies that include children. Investigators and other team members who are experienced in communicating with children and families will usually be better prepared to undertake these procedures than those who work entirely or mainly with adults. Even experienced pediatric researchers may, however, benefit from formal training in both the regulations related to permission and assent and the means of implementing permission and assent as “a process, not a form.”

Recommendation 5.8: To increase investigator competence in communicating with children and parents about research participation, educational programs for investigators and research staff who expect to do research involving children should include training and evaluation in developmentally appropriate and family-sensitive processes for seeking permission and assent.

Directions for Research

Although the committee was encouraged to find some research on parental permission and child assent, much of that research involves very small numbers of participants and other limitations. The range of pediatric conditions and research experiences covered by studies of permission and assent is also fairly limited.

Better evidence about how parents and children of different ages comprehend various dimensions of research and research participation is needed to help investigators and IRBs devise ethical and effective processes for explaining research and seeking permission and assent. To the same end, more research is needed to describe how permission and assent are actually sought in different contexts, how processes vary, and what processes appear to better serve the goals underlying the requirements for permission and assent, given the variations in research purposes and populations.

Recommendation 5.9: Federal agencies, private foundations, and advocacy groups should encourage and support research on existing and innovative permission and assent processes and information materials to support improvements in these processes and guide the education of investigators and institutional review board members.

One template or model for permission and assent clearly will not fit all parents and children, and certain research contexts (e.g., acute serious illness) will make reasoned reflection and decision making more difficult. Some topics that should be considered for investigation include:

  • current practices for seeking permission and assent for different types of research and for children of different ages and with different conditions (e.g., mild chronic conditions or life-threatening diseases);
  • the effect on the flow of information of different processes for seeking permission and assent (e.g., approaching the parent first, involving both the parent and the child in the first meeting, or including a facilitator or research advocate for certain kinds of research);
  • the consequences for communication, comprehension, and decision making of parents and children with different characteristics (e.g., education level, language, and cultural background) and different research purposes or contexts; and
  • the opportunities that children actually are given to dissent, the way in which they express a reluctance or an unwillingness to participate in research, and the reactions of parents and investigators.

In addition, the committee encourages public and private support for research on innovative ways to improve permission and assent processes generally. One starting point would be research on shared decision making between clinicians and patients and methods for improving patient understanding of and decision making about clinical services (see, e.g., O'Connor et al., 1999; Deyo et al., 2000; and Volk et al., 2003). Depending on the research and the circumstances, strategies might include interactive videos, computer programs, and Internet resources.

As described in Appendix B, policies on research participation by wards of the state are poorly documented. The committee encourages a systematic effort to document state policies, their rationales, and their application as a basis for further discussion of the appropriateness of these policies and practices. Some state policies and practices that were an understandable reaction to historical abuses and controversies may now go beyond the provisions and protection of federal regulations in ways that unduly restrict research participation that could benefit the population of children and adolescents who are wards of the state.

CONCLUSION

Both ethical principles and legal requirements place strong emphasis on having competent adults provide informed consent for their participation in research. Because children are usually not legally or intellectually competent to provide consent, an alternative approach has been devised that relies on parental permission and, when appropriate, children's assent to participation in research.

For adults agreeing to clinical research in their own right and for parents agreeing to a child's participation in clinical research, studies suggest that truly informed consent or permission is unlikely to be completely achieved for all individuals in all research situations. Parents who are asked to provide permission for their child's participation in clinical trials are, in particular, often making decisions under great stress and time pressure. Some prefer to trust the physician's assessment rather than make their own, and investigators must be acutely sensitive to the influence that they wield in discussions with parents of ill or injured children. A significant minority of parents may misunderstand the purpose of research, especially when the research tests an intervention for a medical condition. Nonetheless, the goal of having parents provide informed permission remains an important protection for children, both when participation in research is initially sought and throughout the course of a study.

The capacity to make voluntary, informed decisions clearly evolves from birth through adolescence and into adulthood. It also clearly varies among individuals of the same age. The goal should be to involve children in discussions and decisions about research participation as appropriate, given their cognitive and emotional maturity and psychological state. Involving children in discussions and decision making respects their emerging maturity, helps them prepare for participation in research, and gives them an opportunity to express their concerns and objections—and, possibly, influence what happens to them.

This chapter has discussed some circumstances that may raise concerns about undue or even coercive influences on family decision making, for example, when a child's physician is also the investigator seeking agreement to the child's participation in research. Payments to parents or children related to research participation have also raised questions about the potential for undue influence. The next chapter examines these questions.

For simplicity, the rest of this chapter will refer only to parents.

Although parents are often expected to make decisions in their child's best interests given the available alternatives, some arguments allow a role for other considerations or, at least, reasonable freedom from intrusive second-guessing by others (see, e.g., Brock, 1994 and Ross, 1998). One argument is that having assigned parents the main responsibility for raising children, society should grant them reasonable discretion in doing so according to their own values. For example, under this argument, parents may consider peripheral benefits of research participation (e.g., socialization of the child in the values of altruism).

Research participants may also expect that being in a clinical trial may provide them benefits in the form of more information about their condition and closer monitoring than they would otherwise receive. A recent analysis of studies that compared the outcomes of patients participating or not participating in cancer clinical trials reported that “there are insufficient data to conclude” that enrollment in clinical trials leads to improved outcomes (a “trial effect”) (Peppercorn et al., 2004, p. 263). Although the results were more favorable for pediatric cancer trials than for adult cancer trials, the article emphasized the significant limitations in the quality of the data; it did not single out these trials as an exception to the general conclusion. That is, until better data are available to support a trial effect, “patients with cancer should be encouraged to enroll in clinical trials on the basis of trials' unquestioned role in improving treatment of future patients” (p. 263).

Assessments of readability typically rely on formulas that are convenient but that may have a number of limitations (see, e.g., Duffy, 1985; and Baker et al., 1988). The simplest formulas usually involve counts of the numbers of letters or syllables used in words and the number of words used in sentences, on the basis of the assumption that longer means more complex. For texts about medical topics, however, even short words (e.g., shunt and lesion) may be unfamiliar. Some formulas take familiarity into account.

The survey also had some interesting findings on patient interpretations of terms used to describe research, specifically, medical research, clinical trial, clinical investigation, medical study, and medical experiment. The last term, medical experiment, produced the strongest negative associations. Respondents thought patients in medical experiments were at higher risk and were likely to fare worse than patients in medical research. Although patients in medical research were viewed as being at greater risk and more likely to get unproven treatments than those in clinical investigations or clinical trials, they were thought to be more likely to benefit medically. The term medical study was viewed more positively than medical research, perhaps because it seemed to suggest research involving records and not patients.

The basics of Piaget's conceptualization of development are covered in Inhelder and Piaget, 1958; Piaget and Inhelder, 1969; and Grisso and Vierling, 1978. For those citing Piagetian concepts in the context of children's comprehension of and decisions about medical care and similar important matters, see, for example, Weithorn and Campbell, 1982; Garrison, 1991; Ambuel and Rappaport, 1992; Leiken, 1993; Wertz et al., 1994; Britner et al., 1998; and Broome, 1999.

Some studies of decision making in a legal context have similar conclusions. For example, Bartholomew (1996) concluded that research shows similar decision-making capacities for adults and adolescents over the age of 14, but performance is more variable for younger individuals (see also Ambuel and Rappaport, 1992 and Scott et al., 1995).

Footnotes

1

For simplicity, the rest of this chapter will refer only to parents.

2

Although parents are often expected to make decisions in their child's best interests given the available alternatives, some arguments allow a role for other considerations or, at least, reasonable freedom from intrusive second-guessing by others (see, e.g., Brock, 1994 and Ross, 1998). One argument is that having assigned parents the main responsibility for raising children, society should grant them reasonable discretion in doing so according to their own values. For example, under this argument, parents may consider peripheral benefits of research participation (e.g., socialization of the child in the values of altruism).

3

Research participants may also expect that being in a clinical trial may provide them benefits in the form of more information about their condition and closer monitoring than they would otherwise receive. A recent analysis of studies that compared the outcomes of patients participating or not participating in cancer clinical trials reported that “there are insufficient data to conclude” that enrollment in clinical trials leads to improved outcomes (a “trial effect”) (Peppercorn et al., 2004, p. 263). Although the results were more favorable for pediatric cancer trials than for adult cancer trials, the article emphasized the significant limitations in the quality of the data; it did not single out these trials as an exception to the general conclusion. That is, until better data are available to support a trial effect, “patients with cancer should be encouraged to enroll in clinical trials on the basis of trials' unquestioned role in improving treatment of future patients” (p. 263).

4

Assessments of readability typically rely on formulas that are convenient but that may have a number of limitations (see, e.g., Duffy, 1985; and Baker et al., 1988). The simplest formulas usually involve counts of the numbers of letters or syllables used in words and the number of words used in sentences, on the basis of the assumption that longer means more complex. For texts about medical topics, however, even short words (e.g., shunt and lesion) may be unfamiliar. Some formulas take familiarity into account.

5

The survey also had some interesting findings on patient interpretations of terms used to describe research, specifically, medical research, clinical trial, clinical investigation, medical study, and medical experiment. The last term, medical experiment, produced the strongest negative associations. Respondents thought patients in medical experiments were at higher risk and were likely to fare worse than patients in medical research. Although patients in medical research were viewed as being at greater risk and more likely to get unproven treatments than those in clinical investigations or clinical trials, they were thought to be more likely to benefit medically. The term medical study was viewed more positively than medical research, perhaps because it seemed to suggest research involving records and not patients.

6

The basics of Piaget's conceptualization of development are covered in Inhelder and Piaget, 1958; Piaget and Inhelder, 1969; and Grisso and Vierling, 1978. For those citing Piagetian concepts in the context of children's comprehension of and decisions about medical care and similar important matters, see, for example, Weithorn and Campbell, 1982; Garrison, 1991; Ambuel and Rappaport, 1992; Leiken, 1993; Wertz et al., 1994; Britner et al., 1998; and Broome, 1999.

7

Some studies of decision making in a legal context have similar conclusions. For example, Bartholomew (1996) concluded that research shows similar decision-making capacities for adults and adolescents over the age of 14, but performance is more variable for younger individuals (see also Ambuel and Rappaport, 1992 and Scott et al., 1995).

Copyright © 2004, National Academy of Sciences.
Bookshelf ID: NBK25560
PubReader format: click here to try

Views

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...