Children's and Their Parents' Views on Facing Research Risks for the Benefit of Others

Objective  To assess children's and parents' attitudes regarding pediatric research that poses minimal risk or a minor increase over minimal risk and does not offer the potential to benefit the child clinically.

Design  Separate in-person interviews with children and their parents.

Setting  Clinics where the children were receiving clinical care or participating in clinical research for asthma or cancer.

Participants  Children aged 7 to 14 years and their parents or legal guardians.

Intervention  In-person interviews.

Main Outcome Measures  Respondents' willingness to enroll the child in nonbeneficial research and charitable activities that posed the same risks.

Results  Overall, 81 child-parent pairs were interviewed. For a hypothetical study that would not benefit the child and posed a risk of headache, 71% of the children were willing to participate, and 72% of the parents would allow their children to participate. For a hypothetical study that would not benefit the child and posed a very small chance of a broken leg, 43% of the children and 24% of the parents endorsed the child's participation. Overall, respondents were equally willing to have the child help others by participating in nonbeneficial research or by participating in a charitable activity.

Conclusions  Most respondents were willing to have the child participate in some nonbeneficial research, and most did not consider nonbeneficial pediatric research to be more problematic than charitable activities for children. These findings provide empirical data to support the acceptability of exposing children to some research risks for the benefit of others.

To improve pediatric medical care, it is sometimes necessary to perform research studies and interventions with children that do not offer a compensating potential for clinical benefit.^1-3 Guidelines around the world allow children to be involved in such nonbeneficial research when the risks are minimal.^4-6 US regulations also allow children to be enrolled in nonbeneficial research that poses a minor increase over minimal risk, provided that several additional safeguards are satisfied.

Nonbeneficial pediatric research continues to be subject to intense theoretical debate,^6-9 and some commentators argue that it is unethical to expose children to research risks for the benefit of others. Evaluation of this debate should include data on the attitudes of those most affected, namely, children and their parents. One might assume that children and parents accept nonbeneficial pediatric research, as children often participate. Yet, children and parents may agree to the child's participation because they endorse some nonbeneficial pediatric research or because they fail to recognize that it does not offer a compensating potential for clinical benefit. Our study was designed to address this ambiguity by evaluating respondents' attitudes toward research that poses risks and clearly does not offer any potential for clinical benefit.

Data were collected as part of a larger survey of children and their parents or legal guardians. When 2 parents were available, the parent who was more active in the child's medical care was invited to participate. To be eligible, the children had to (1) be aged 7 to 14 years, (2) be receiving clinical care or enrolled during the previous year in an on-going medical intervention study for asthma or cancer, (3) have been seen at least once previously by the treating team, and (4) speak English. There were 7 participating sites: University of Arizona, Case Western Reserve University, Duke University, Georgetown University, Johns Hopkins University, National Institutes of Health Clinical Center, and University of North Carolina, Chapel Hill. Eligible children were invited to enroll based on the availability of the interviewer at the site. The protocol and survey instruments were approved by the institutional review boards at the 7 participating sites and at the Research Triangle Institute.

Parents provided written consent and the children provided written assent. Individuals were informed that declining to participate would not adversely affect them in any way. Those enrolled were informed that they could skip any questions and could stop participating at any time. The surveys were conducted in person by interviewers who were independent of the care team and professionally trained for the study. Children and parents were interviewed separately.

The survey instruments were developed and revised in 8 steps: (1) literature review, (2) draft survey development, (3) review by 3 experts in research ethics, (4) review by 2 pediatricians, (5) review by survey professionals, (6) review by a child psychologist experienced in surveying children, (7) cognitive interviews with 8 children with cancer or asthma, and (8) behavioral pretesting with 4 children with cancer or asthma. Survey questions covered 3 primary domains: (1) understanding of research, (2) attitudes regarding decision making, and (3) willingness to have the child participate in nonbeneficial research and charitable activities. Our article reports findings from the third domain.

The children were asked how willing they would be to participate in a charitable activity that posed specified risks and how willing they would be to participate in a research study that would not benefit them and posed the same risks. For example, one question asked: “How willing would you be to participate in a medical research study if it would not help you but might help other children and there was a one in a million chance of your dying from being in the study?” Parents were asked how willing they would be to allow their child to participate in the same charitable activities and same research studies. Three descriptions were used to represent minimal risk: (1) “chance of getting a headache for a few hours,” (2) “one in a million chance of dying,” and (3) “the same risks as riding in a car.” A “very small chance of breaking a leg” was used to represent a minor increase over minimal risk. The questions and the answer options were read to respondents; answer options were written on cards placed in front of respondents, who could respond verbally or point to an answer.

A binomial test for proportions was used to evaluate whether parents and their children who were participating in clinical research at the time were more likely to prefer that the child help others by participating in research or by participating in a charitable activity. The McNemar statistic was used to compare how willing respondents were to have the child participate in nonbeneficial research studies that posed specified risks with how willing they were to have the child participate in charitable activities that posed the same risk. For this analysis, 95% confidence intervals were calculated for the difference of the correlated proportions.^10,11

Overall, 81 of 117 child-parent pairs invited to participate completed the survey (response rate, 69.2%). Of the children, 28 (35%) were participating in research and 53 (65%) were receiving clinical care; 51.9% were male and 48.1% were female (Table 1). In terms of race, 61.7% of the children considered themselves to be white, 17.3% considered themselves black, and 7.4% considered themselves of mixed race. For ethnicity, 19.8% of the children considered themselves Hispanic.

Of the adults, 87.6% were female (Table 1) and 91% were the child's biological parent (data not shown). Mean age of the adult respondents was 39.8 years, with a range of 25 to 56 years; 55.6% had attended college and 14.8% had attended graduate school. All the adult respondents stated that they felt close to their child. In addition, 89% stated they had an excellent relationship, 9% stated they had a good relationship, and 1% stated they had a fair relationship with their child (data not shown).

When asked whether they would be more willing to help other children by participating in research that would not help them or in a charitable activity, 38% of the children were equally willing to participate in either activity, 33% preferred nonbeneficial research, and 18% preferred a charitable activity (Table 2). Of the parents, 56% were equally willing to have their child participate in either activity, 19% preferred nonbeneficial research for their child, and 19% preferred a charitable activity (Table 2). Children who were participating in research were statistically significantly more likely to prefer to help other children by participating in nonbeneficial research rather than a charitable activity (76.5% vs 23.5%; 95% confidence interval, 0.56-0.97). The parents of these children also preferred that their child help others by participating in research, though this difference was not statistically significant (60.0% vs 40.0%; 95% confidence interval, 0.30-0.90).

When presented with a research study that was described as offering no chance to benefit the child and posing a risk of headache, 82% of the children were willing to participate and 79% of the parents would allow their child to participate (Table 3). For a charitable activity that posed the same risks, 85% of the children were willing to participate and 87% of the parents would allow their child to participate.

In response to a research study that was described as offering no chance to benefit the child and posing a one-in-a-million chance of death, 42% of the children and 18% of the parents were willing to have the child participate. For a charitable activity that posed the same risk, 36% of the children and 18% of the parents were willing to have the child participate (Table 3). When the risks were described as “the same risks as riding in a car,” 89% of the children and 93% of the parents were willing to have the child participate in nonbeneficial research (data not shown). For a research study that was described as one that would not benefit the child and posed a very small chance of a broken leg, 48% of the children were willing to participate and 26% of the parents would allow their child to participate (Table 3). Of the 6 pairwise comparisons of respondents' willingness to have the child participate in nonbeneficial research or a charitable activity that posed the risks described in Table 3, 1 was statistically significant. When there was a risk of a headache, parents were significantly more likely to prefer that their child participate in charity rather than nonbeneficial research (difference, − 9.3%; 95% confidence interval, − 11.9% to − 1.7%) (Table 3).

Overall, 82% of pediatric respondents and 79% of their parents were willing to have the child participate in a research study that was described as offering no chance to help the child and posing a risk of headache. In addition, 48% of the children and 26% of the parents were willing to have the child participate in research that was described as offering no chance to help the child and posing a low risk of a broken leg. These data provide empirical support for nonbeneficial pediatric research that poses low risks.

Whether parents and children will agree to the child's enrollment in a nonbeneficial research study in practice likely depends on the nature of the study and the child's circumstances. Hence, our present data, collected in response to hypothetical studies, do not reveal what percentage of children and parents will actually agree to the child's participation in nonbeneficial research. Nonetheless, the data do suggest that most respondents consider it acceptable to expose children to minimal research risks for the benefit of others. These data also suggest that nonbeneficial pediatric research that poses a minor increase over minimal risk is more controversial but is acceptable to some parents and children.

Empirical data alone cannot establish the ethical acceptability of nonbeneficial pediatric research. At the same time, the finding that parents and children consider some nonbeneficial pediatric research studies to be acceptable undermines the claim that nonbeneficial pediatric research is necessarily unethical. In particular, most of these parents were willing to enroll their child in some nonbeneficial research, suggesting that they do not regard nonbeneficial pediatric research as inconsistent with parents' obligation to protect and nurture their children. Most of the children were also willing to face some research risks for the benefit of others.

The children indicated a greater willingness overall to help others by participating in nonbeneficial research rather than a charitable activity (38% vs 18%), while parents were evenly split on which activity they preferred for their child (Table 2). In addition, only 1 of the 6 pairwise comparisons between willingness to have the child participate in nonbeneficial research vs participating in a charitable activity that posed the same risks was statistically significant (Table 3). Taken together, these data suggest respondents overall were equally willing to have the child help others by participating in research that would not benefit them or by participating in a charitable activity that posed the same risks. This conclusion is consistent with a national survey of 900 Americans that found that respondents regard participation in research to represent a similar “contribution to mankind” as work for a charity.¹² Assuming it can be acceptable for children to face some risks in charitable activities, these findings suggest it can be acceptable to enroll children in nonbeneficial research that poses similar levels of risk, provided that the research also offers at least equivalent potential to benefit others.

Many prior studies have assessed children's and parents' attitudes,^13,14 as well as parent's understanding^15-19 and motivations^20-23 regarding pediatric research. Most of these studies focus on research that offers some potential for clinical benefit,^24,25 including genetic susceptibility studies,²⁶ vaccine trials,²⁷ and randomized treatment trials.^28,29 A few previous studies attempted to gather data on attitudes regarding nonbeneficial pediatric research. One study evaluated respondents' willingness to have the child participate in research to better characterize asthma using spirometry and radiography.³⁰ Another study found that parents were willing to have their children participate in research “that may benefit other children's health, but not necessarily [their] own child's health.”³¹

While these surveys were designed to evaluate children's and parent's attitudes regarding nonbeneficial research, the respondents may have assumed that the studies in question offered some potential for clinical benefit. This possibility is supported by data that individuals frequently assume that pediatric research offers some potential for clinical benefit.^32,33 Surveys of parents and children who participated in nonbeneficial research reveal similar attitudes.^34,35 One survey found that 47% of the children in a nontherapeutic pharmacokinetic study assumed that they would benefit clinically.³⁶ A second study found “poor to no understanding” regarding the benefits and risks in children who had participated in a nontherapeutic blood draw study.³⁷ In a third survey, 98% of Danish children involved in a nontherapeutic study stated that they wanted to help other children.³⁸ This survey does provide data on children's willingness to help others. However, the description of the study apparently did not mention any risks. Thus, it is not clear whether these children were willing to face risks as a part of helping others.

Our survey was designed to address the uncertainty in existing data by evaluating respondents' attitudes toward pediatric research studies that were explicitly described as ones that would not help the child and posed risks intended to represent minimal risk and a minor increase over minimal risk. Respondents overall were equally willing to have the child participate in these research studies and charitable activities that posed the same risks. Presumably, if respondents had thought the described research studies offered the potential for clinical benefit, they would have been more willing to enroll the child in them compared with charitable activities that posed the same risks. The conclusion that respondents understood that the described studies did not offer the potential for clinical benefit is supported by the fact that none of the respondents who preferred that the child participate in research rather than a charitable activity cited the belief that the research might benefit the child as a reason (Table 2).

Children should be enrolled in nonbeneficial clinical research only when there is a compelling justification, the research has important value, and the risks are sufficiently low. Many guidelines define the risk threshold based on the risks children encounter in daily life⁶ or during routine examinations.^4,5,39 While these standards have been widely endorsed, they suffer from a fundamental flaw. Reasonable parents allow their children to participate in the activities of daily life, such as playing sports, because they assume children benefit from them. Hence, the risks posed by these activities do not seem to provide an appropriate standard for evaluating the risks of research studies that do not offer a compensating potential for clinical benefit.

The proposed charitable participation standard would allow children to be enrolled in nonbeneficial research when the risks do not exceed the level of risks to which children may be exposed during appropriate charitable activities. This standard has been defended on the grounds that charitable activities and nonbeneficial research both involve exposing children to risks for the benefit of unrelated others.⁴⁰ Despite these similarities, it has been argued that nonbeneficial pediatric research, in principle, is ethically more worrisome than charitable activities for children.⁴⁰ Our finding that respondents overall were equally willing to have the child help others by participating in nonbeneficial research as they were charitable activities that pose the same risks suggests that these children and parents at least do not regard nonbeneficial pediatric research as more problematic than charitable activities.

Approximately 90% of the children and their parents were willing to have the child participate in a nonbeneficial study that posed risks “the same as the risks of riding in a car.” In contrast, only approximately 30% were willing to have the child participate in a study that posed a one-in-a-million chance of dying. This finding is significant because children face a one-in-a-million chance of dying from many ordinary car trips.⁴¹ Moreover, riding in a car also poses significant risks of morbidity, implying that the difference in the wording of risks resulted in most individuals unknowingly preferring a more risky study to a less risky study.

These data suggest that individuals' willingness to participate in nonbeneficial research may be increased by describing the risks in comparison with the activities of daily life and may be decreased by citing the specific chances of harm. The influence of risk perception on individuals' research enrollment choices is illustrated by the finding that some parents considered clinical research to be more risky than charitable activities and other parents considered it to be less risky, even though the activities were described as posing the same risks (Table 2). Future research should evaluate which descriptions or combinations of descriptions enable children and their parents to understand the risks of clinical research.

These findings are subject to several limitations. First, our study surveyed a relatively small number of respondents. Second, most of the adult respondents were mothers. While these data may reflect which parent tends to be more involved in children's medical care, future research should assess whether the present findings apply to fathers. Third, the present study was limited to children with cancer or asthma aged 7 to 14 years.

In conclusion, our article finds that pediatric respondents and their parents accept pediatric research that poses some risks and offers no chance to benefit the child. In addition, overall respondents did not regard nonbeneficial pediatric research as more problematic than charitable activities for children that posed the same risks. These findings give empirical support for nonbeneficial pediatric research and suggest that exposing children to research risks for the benefit of others can be acceptable, provided that the research has the potential to gather data important for improving health and well-being and includes strict limits on risk.

Correspondence: David Wendler, PhD, Department of Clinical Bioethics, National Institutes of Health Clinical Center, Bldg 10, Room 1C118, Bethesda, MD 20892 (dwendler@nih.gov).

Accepted for Publication: June 6, 2007.

Author Contributions:Study concept and design: Wendler and Jenkins. Acquisition of data: Wendler and Jenkins. Analysis and interpretation of data: Wendler and Jenkins. Drafting of the manuscript: Wendler. Critical revision of the manuscript for important intellectual content: Wendler and Jenkins. Administrative, technical, and material support: Wendler. Study supervision: Wendler.

Financial Disclosure: None reported.

Disclaimer: The opinions expressed are the authors' own. They do not represent any position or policy of the National Institutes of Health, Public Health Service, or Department of Health and Human Services. The work for this manuscript was completed as part of the authors' official duties as employees of the National Institutes of Health.

Additional Contributions: We thank the respondents for their time and effort. Gina Kilpatrick, MEd, and Connie Hobbs, MS, assisted with the study; Ken Tercyak, PhD, assisted with the survey instruments; and Christine Grady, PhD, and Sumeeta Varma, BA, provided helpful suggestions on the text. We also thank the interviewers Shelley Clark, BA, RN, Sowmya Prahlad, MS, Lauren Wine, BA, Marilyn Sampilo, BA, Joanne Lager, MD, Donald Beam, MD, Shelley Ingram, PharmD, Jane Gaillard, BSN, Maria Meyer, BSN, RN, Julie Blatt, MD, Lori Wiener, PhD, Tammy Scott, RN, and Amy Harris Yamokoski, MA.