The Importance of Engaging Children in Research Decision-Making: A Preliminary Mixed-Methods Study

The study sample included 31 parent-child dyads. Interviews were initiated with 33 families; however, responses from two parent-child dyads were not usable because we discovered during the interviews that these families did not meet inclusion criteria due to language limitation in one case and the child’s cognitive limitation in another. The response rate among eligible families approached for participation was 31/76 (41%) (see figure 1 ). The mean age of child participants was 11.7 years (see table 1 ). There were nearly equal numbers of female and male child participants. The majority of children were identified by their parents as white, non-Hispanic. About half of the children had participated in research studies previously. The majority of parent respondents were female, married, and had completed at least four years of college. Of the 31 children in the total study sample, 11 were healthy controls and 20 were enrolled in their research study due to illness. Of children with illness, conditions included spinal muscular atrophy, endometriosis, short stature, asthma, autism spectrum disorder, and sickle cell disease. Forty-two percent of parents preferred equally shared decision-making between parent and child, 35% of parents preferred a greater parental role, and 23% preferred a greater child role. No parents preferred unilateral decision-making by either parent or child. Children reported a greater spread of preferred decision-making roles, with 35% of children preferring equally shared decision-making between parent and child, 35% preferring a greater child than parental role, and 16% preferring a greater parental than child role. Few children preferred unilateral decision-making by the child (3%) or parent (10%). We also asked about actual decision-making in the decision about joining the CTSU study. Most parents reported that they and their child shared decision-making equally or that the decision was mostly or unilaterally made by the parent. No parents reported that a child unilaterally made the decision to join their research study. Similarly, the majority of children reported that the decision was made equally or that the decision was mostly or unilaterally made by the parent. In contrast to parental responses, 10% of children reported that the child unilaterally made the decision to join the research study. Agreement between preferred and actual decision-making roles was lower for children (weighted kappa = 0.2, p = 0.08) than for parents (weighted kappa = 0.39, p = 0.0001). We also assessed concordance between parent and child responses. Parent-child agreement regarding their preferred decision-making role was low (weighted kappa = 0.16, p = 0.16); they more often agreed on who actually made the decision about participation (weighted kappa = 0.29, p = 0.009). Parents and children both reported that they preferred a greater role for child participation than what actually occurred. Parent and child comprehension is reported in figure 2 . The median SSCI score for parents was 12.5 (interquartile range [IQR] = 10.5, 14), and for children it was 9 (IQR = 6, 11) (Wilcoxon signed rank z = −4.2, p < 0.00001). As expected, child comprehension varied significantly by age group (Kruskal-Wallis χ 2 = 9.6, p = 0.008), with children aged 15 to 17 having a higher median SSCI score (median = 12, IQR = 6.5, 14) than children aged 7 to 10 (median = 6, IQR 3, 8.5). The majority of children were able to answer correctly (score = 2 on SSCI) for comprehension elements related to assent: procedures/what will happen (n = 21/31; 68%), choice to participate (n = 28/31; 90%), and choice to withdraw (n = 22/31, 71%). Table 2 summarizes factors related to the child’s actual role in decision-making. Children’s reports of their actual roles in decision-making were correlated with higher comprehension (Spearman’s rho between SSCI and actual decision-making role = 0.35, p = 0.05), being asked about participation by the study doctor (Fisher’s exact p = 0.01, with greater participation when being asked by the study doctor), reading the information that was given to them (Fisher’s exact p = 0.005, with greater participation by reading information), parent marital status (Fisher’s exact p = 0.02, with greater involvement for children of married parents), and child race (Fisher’s exact p = 0.03, with greater involvement by white children compared to African American children), but not with the child’s age (Spearman’s rho = 0.14, p = 0.44) or the amount of information given to the child (Spearman’s rho = −0.07, p = 0.69). Parents’ reports of their child’s actual role in decision-making correlated with the amount of say parents permitted their child (Spearman’s rho = 0.4, p = 0.03, with greater child involvement when parents reported greater input from the child) and the sex of the child (Fisher’s exact p = 0.05, with greater child involvement from girls compared to boys). We also evaluated factors related to higher comprehension for child and parent subjects. Table 3 summarizes factors related to the child’s understanding. Child understanding was positively associated with age (Spearman’s rho = 0.57, p = 0.008); child’s education level (Spearman’s rho = 0.62, 0.003); verbal intelligence, including performance on the vocabulary (Spearman’s rho = 0.73, p = <0.0001) and similarities (Spearman’s rho = 0.61, p = 0.0003) subscales of the WASI; and process variables including reading information (Kruskal Wallis χ 2 = 8.5, p = 0.003). Parents’ understanding was positively correlated with the prospect for direct benefit (Spearman’s rho = 0.4, p = 0.04) and the child’s illness severity as reported on the MISQ (Spearman’s rho = 0.46, p = 0.04), and it was also associated with the child’s sex (Kruskal Wallis χ 2 = 3.8, p = 0.05, understanding greater for parents of girls). In addition to assessing factors related to actual research decision-making, we asked parents and children to provide reasons for their preferred decision-making role, and we asked children to describe their satisfaction with the decision-making process as it occurred. Three themes emerged from parents’ and children’s descriptions of their preferred decision-making roles: reasons for greater parental involvement, reasons for relational involvement, and reasons for greater child involvement. Table 4 shows representative statements by theme. Reasons for greater parental involvement included younger child age, a sense that parents had a primary responsibility for decision-making regarding their child, that the child lacked understanding of the research, that the parent understood more about the research, and that the parent had the primary responsibility for ensuring that the research activity was completed. Reasons for relational involvement included a sense that there was a shared responsibility between parent and child for decision-making, that the parent had a role in helping their child understand the research, and that the parent had a responsibility to inculcate altruism in the child. Reasons for greater child involvement included that the child’s preferences or wishes were important when participating in research, that the child should participate in decision-making because they experience the burden of participation, and that the research concerns the child’s body. Respondents also advocated greater involvement for the child when the child was more mature and when the research did not offer a prospect of direct benefit to the child. All children reported that they were satisfied with the way in which the research decision was made. Two broad themes emerged related to child satisfaction with decision-making: satisfaction with the process of decision-making and satisfaction with participation in the research itself. Reasons for satisfaction related to the process of decision-making included the child’s sense of autonomy in deciding, the parent’s encouraging participation in the research, and the shared nature of the decision-making process with the parent. Reasons for child satisfaction related to the research itself included enjoying the research experience, a sense of altruism, that the research was low burden, that the child received an incentive, and that the child believed he or she would receive benefit from the research. Table 5 shows representative statements related to the child’s satisfaction with the decision-making process.

Ethically engaging children in research requires parents’ and children’s involvement in the decision-making process. Among our study participants, both parents and children desired a greater role for the child in decision-making about research participation than they experienced. Child participation may be improved by increasing parent-child communication about research, altering study team processes to increase engagement of the child, providing literacy level-appropriate reading materials, and assessing comprehension in real time. Respect for traditional parental roles in decision-making, encouragement of relational decision-making between parent and child, and respect for the developing autonomy of the child were important themes that emerged from qualitative analyses. Still, given the limitations of our small sample size and single center design, more research is needed to show whether modifications to the consent and assent process that incorporate these findings will improve children’s participation in decision-making.

We found that, while there was a diversity in preferences for decision-making between parents and children, both parents and children reported that parents actually played a greater role in decisions about whether to participate in the CTSU study. Factors related to greater participation by the child in decision-making included the amount of say they were given, according to the parent; the child’s level of comprehension; whether the child was asked her or his preference for participation by the study doctor; whether the child read materials they were given; the parent’s marital status; and the child’s race. Consistent with prior studies, we found lower comprehension in children when compared to adults, with increasing comprehension among older compared to younger children. Nevertheless, some children as young as seven appeared to understand at least some of the basic elements of assent (understanding of procedures, the ability to choose to participate, and understanding of the right to withdraw). Overall, child comprehension was positively related to age, education level, verbal intelligence, and reading study-related information. Parent understanding was associated with the prospect for benefit and with illness severity. Our study adds important qualitative insights to the literature of decision-making for participation in pediatric research. Both parents and children identified reasons for parent and child involvement in decision-making. Greater parental involvement related to traditional roles of parents making decisions in families, helping younger children, and bearing responsibility for ensuring studies are completed. Parental involvement also related to understanding the research, whereas the child may lack understanding, reinforcing our quantitative findings that decision-making and understanding of research are closely related. Greater child involvement in decision-making related largely to respect for the child, most commonly that the child’s preferences were important, that the child would undergo the burden of the research, and that the child’s body would be physically involved in the research; these factors acknowledge the developing autonomy of the child. Finally, parents and children also described decision-making as relational, with the parent helping the child through the process and the two sharing responsibility. This supports our quantitative findings that encouraging open communication between parent and child improves child engagement. Children described satisfaction with their participation in decision-making both because of their contributions to and participation in research and because of the process of decision-making itself. These findings further support study team attention to involving children in the process of decision-making. The role of children in research decision-making is an evolving area of study. Some studies suggest that parents should remain the primary decision-maker, 18 while others favor joint decision-making based on the child’s age, maturity, open communication between parent and child, and the level of risk of the study. 19 Previous studies have demonstrated variability in whether children understand the elements of consent, including freedom to withdraw, freedom to ask questions, and the potential benefits of participation. 20 Our findings are consistent with prior research suggesting that more open communication between parent and child increases children’s participation in decisions; parents and children should therefore be encouraged to discuss a child’s research participation. The results also have potential implications for study teams, as they identify factors related to children’s participation in research decision-making, including children’s sense of engagement by the study team, the receipt of reading materials, and their understanding of what they were told, as well as sociodemographic factors related to family structure, parents’ marital status, and the child’s race and sex. While assent generally does not aim at the same level of understanding as legally effective consent does, those giving assent should know what procedures are involved in the research, should have the opportunity to freely choose to participate, and should be aware of the opportunity to withdraw. 21 This suggests a need to assess understanding of at least these critical elements prior to obtaining assent. Additionally, knowledge of the potential for sociodemographic factors to impact decision-making may assist study teams in ensuring adequate engagement in the assent process from children who may be less likely to participate in decision-making. We also found that comprehension was higher for parents whose children were enrolled in studies with the prospect of direct benefit and those whose children had higher illness severity. Parents of children with greater illness severity and who are enrolling in studies with the prospect of benefit may have more interactions with the health system and therefore a higher health literacy or may pay closer attention to the details of studies they are considering. These findings support greater attention to parental comprehension, particularly in studies without the prospect of direct benefit and those in healthy children. To improve understanding for both parents and children, study teams could consider real-time assessment of understanding, using teach-back or formal measures such as the SSCI, with the full scale used for parent comprehension and a subset of the three critical elements of assent for younger child subjects. 22 Several limitations of our study should be noted. First, it was conducted in a single center, with a relatively homogenous study sample, potentially limiting its generalizability. Additionally, although the paired data in this study allowed for an assessment of concordance between parent and child perspectives, a larger sample size may have allowed even more precise estimates of correlations between potential correlates and decision-making and comprehension as well as multivariable analyses to identify independent correlates of decision-making preferences and roles and comprehension. Furthermore, analyses were not adjusted for multiple testing. Additional studies should therefore be conducted in larger samples to confirm or refine our findings. Finally, future studies should also evaluate whether using the SSCI or another measure of comprehension in real time during the parental permission and assent process enhances subsequent comprehension of study details.

SUPPORTING INFORMATION The appendix, tables, and figures are available in the “ Supporting Information ” section for the online version of this article and via Ethics & Human Research ’s “ Supporting Information ” page: https://www.thehastingscenter.org/supporting-information-ehr/ .