A Randomized Study of a Method for Optimizing Adolescent Assent to Biomedical Research

Abstract

Introduction

Ethicists and biomedical researchers have made significant progress in our understanding of child and adolescent assent to biomedical research, although these efforts have also revealed the intricacy and complexity of the assent process (Scherer et al. 2013). Important questions remain unanswered, particularly those regarding child and adolescent competence to assent and how to best respect the autonomy of children and adolescents to assent (Baines 2011, Hein et al. 2015, Kodish 2003).

Ethical standards require that research participants with uncertain competence be granted a maximal level of decision-making autonomy consistent with their decision-making capacities, which is contingent in large part on their developmental maturity (Field and Behrman 2004, Santelli et al. 2003, Rossi, Reynolds, and Nelson 2003, Ross 2006). Early studies of assent suggested that adolescents had decision-making competencies that were similar to those of adults (Weithorn and Campbell 1982, Weithorn and Scherer 1994). Subsequent studies, however, have emphasized the variability in cognitive, socio-emotional, and neurological development in children and adolescents (Kon 2006, Masty and Fisher 2008, Spears 2010) and the differences in how adults, children, and adolescents process risk and benefit information (Abramovitch et al. 1995, Annett et al. 2004, Brody et al. 2005, Brody et al. 2006, Brody, Dalen, et al. 2012, Burke, Abramovitch, and Zlotkin 2005, Cohn et al. 1995, Lidz et al. 2004, Reynolds and Nelson 2007). These studies have brought the recognition that it may be implausible to establish a single age-based standard for when to defer to the judgments of participants.

Contemporary research on adolescent assent has emphasized the highly contextualized nature of the assent process, with thorough discussions of volition in research decision-making in the extant literature (Nelson et al. 2011, Weithorn and Scherer 1994). Consequently, studies of assent have focused on adolescent autonomy, particularly how adolescents interact with parents and with biomedical researchers to arrive at a mutual research participation decision (Miller, Reynolds, and Nelson 2009, Swartling et al. 2011). These studies have raised consistent concerns about minors’ capacities to assent or, more importantly, dissent under social pressure (Abramovitch et al. 1995, Baumrind 2005, Brody et al. 2003, Brody et al. 2005, Brody et al. 2006, Brody, Turner, et al. 2012, Miller, Reynolds, and Nelson 2009, Swartling et al. 2011, Varma, Jenkins, and Wendler 2008).

Current recruitment procedures in biomedical research complicate and likely attenuate adolescent autonomy. In an effort to facilitate recruitment efforts, many biomedical facilities employ recruiters who establish contact with potential participants through their health care providers and make them aware of research opportunities. This approach may establish a “foot-in-the-door” tactic for adolescents whose parents have been contacted and who agree to seek further information (Burger 1999). Moreover, adolescents ordinarily play a subordinate role within families and experience significant power differentials in their relationships with adults. Families employ different degrees of adolescent involvement in research participation decision-making and finding an effective balance of parental control and adolescent inclusion in the process can be complicated (Swartling et al. 2011). Parents tend to expect their children and adolescents to acquiesce, and children and younger adolescents often do. Lastly, the validity or quality of assent and parental permission to participate in biomedical research is contingent on the parents and adolescents having sufficient understanding to make a knowledgeable decision. The relevant knowledge varies from one study to the next, but routinely includes matters such as medications, research procedures, the rights and privileges that research participants have, and the risk and benefits of the study.

These difficulties arise from the larger issue of how best to obtain assent from adolescents. The challenge is to define an assent process that can be adapted for general use in pediatric biomedical research and that takes into account how social context affects autonomy, particularly with adolescents with varied competencies (Gibson et al. 2011, Kumpunen et al. 2011, Masty and Fisher 2008). Brody and colleagues (Brody et al. 2003, Brody et al. 2005) have examined parent and adolescent participation decision-making in simulated biomedical research participation decisions. They found that both parents and adolescents claim decision-making responsibility for biomedical research participation decisions, with parents expecting adolescents to acquiesce, and adolescents indicating they would not have to comply with their parents’ wishes. Studies of parental influence and family dynamics in the assent process suggest there are different strategies employed in making a research participation decision. For example, one study has determined that family communication and decision-making strategies vary, from parents assuming complete control to family decision-making processes that are open and inclusive (Snethen et al. 2006). Others have observed that the presence or absence of parents during an assent discussion has a profound effect upon how researchers, parents, and adolescents behave. These investigators have recommended that adolescents and parents be given explanations of a study separately and that adolescents be given an opportunity to ask questions in private (Broome et al. 2003). One alternative that has been suggested is to customize assent conferences to meet the specific needs of families and children (Masty and Fisher 2008, Salas et al. 2008). These considerations suggest the need for additional studies of various techniques for obtaining assent are needed, preferably including actual studies of the “real world” assent process (Campbell 2005).

To address these issues within the context of a randomized trial, we evaluated the effect of adolescents and parents being together or separated for the consent/assent procedures. We conceptualized the separation of adolescents and parents as an autonomy-enhanced assent/consent procedure. Our main hypothesis was that separating adolescents from their parents would improve the quality of adolescent assent by increasing their knowledge and appreciation of the clinical trial and willingness to participate. We further hypothesized that the impact of the intervention would be stronger for older adolescents since older adolescents are often presumed to have greater maturity (Coleman and Rosoff 2013). The procedures provided opportunities to explore participant satisfaction with the consent procedures and coordinator observation of engagement, understanding, and enthusiasm for trial participation. We hypothesized that there would be no consent satisfaction difference between adolescents and parents, while there would be intervention-related differences in parent and adolescent engagement, understanding, and enthusiasm.

Method

Results

A total of 64 adolescent-parent dyads completed procedures for this study. Only one family refused participation in the assent study. Mean age of the sample was 14.16 (SD=1.69) (range: 12–17) and sex included 52% male and 48% female. In regards to asthma severity, 34% reported mild symptoms, 36% reported moderate symptoms, and 5% reported symptoms in the severe range (Table 1). In order to examine the impact of adolescent age upon willingness to participate, study knowledge, and satisfaction, a median split created two groups (ages 12–14; 15–17). The consent intervention was comprised of two groups; usual assent (n=34) and separated assent (n=30). No differences were observed between demographic characteristics and the consent intervention condition (usual versus separated). Adolescents were most commonly accompanied by their mothers and there were no differences in parent demographic characteristics between the two groups (Table 2).

Knowledge and appreciation (KAS)

Our main hypothesis examined the impact of the consent intervention (usual consent versus separated consent) and adolescent age group (12–14; 15–17) upon parent and adolescent KAS domain percent correct scores. The proportion of participants providing a correct response to individual items is presented in Table 3. Parents and adolescents demonstrated a similar response pattern with several important differences. Both parents and adolescents achieved the highest percent correct scores in the research process (parent 86.87% correct; adolescent 82.26% correct) and rights/privileges domains (parent 85.69% correct; adolescent 82.06% correct).

Table 4 demonstrates the impact of adolescent age on KAS domain scores. Adolescent age operated in two important ways, influencing knowledge about asthma medicines and risks/benefits. For the asthma medicines knowledge scale, older adolescents (M=84.6%, SD=16.9%) had higher scores than younger adolescents (M=74.1, SD=15.6), [F(1,62)=6.41; p=.01, eta²=.09]. More nuanced was the finding that the separate assent procedure intervention increased knowledge for risks/benefits with older adolescents [F(1,23)=7.68, p=.011, eta²=.25].

Table 4

Percent correct on knowledge survey domains

Domain	Parent of Adolescent <15 (N=38)	Parent of Adolescent ≥15 (N=26)	Adolescent <15 (N=38)	≥15 (N=26)
Asthma Trial Medicines* T A	85.3 (17.4)	90.1 (12.0)	74.1 (16.9)	84.6 (15.6)
Research Process*	86.9 (13.4)	86.8 (11.4)	81.3 (13.6)	83.5 (15.5)
Rights and Privileges*	84.1 (10.9)	88.0 (10.5)	79.7 (11.2)	85.5 (10.0)
Risk and Benefits* TT	76.0 (15.8)	75.0 (17.3)	61.5 (22.2)	68.5 (15.8)

Open in a separate window

^*Mean (SD)

^TParent > Adolescent F(1,60)=11.12; p=.001; eta² .16

^TTParent > Adolescent F(1,59)=10.19; p=.002; eta² .15

^AAdolescent age ≥15 > Adolescent age <15; F(1,63)=6.41; p=.01

Further detail on the interaction between the separate assent intervention and younger (<15 years) versus older (>15 years) adolescents on knowledge of risks and benefits is provided in Figure 1. Parents scored higher than adolescents regardless of the assent intervention [F(1,36)=9.59, p=.004, eta²=.21]. For younger adolescents, the intervention did not impact knowledge of risks/benefits. The separate procedure resulted in greater appreciation of study risks and benefits only for older adolescents and their parents [F(1,23)=7.68, p=.011, eta²=.25]. Parents who received study information separately (i.e., autonomy enhanced) demonstrated increased knowledge over those who received the information in the usual consent condition [F(1,23)=4.73, p=.04, eta²=.17]. Likewise, the separate intervention also resulted in significantly greater percentage of correct responses for adolescents [F(1,23)=4.49, p=.04, eta²=.16].

Open in a separate window

Figure 1

Relationship of consent intervention on KAS knowledge of risk and benefit, The group X A age interaction F=3.097; p=.05; eta2=.096

Discussion

This is an in vivo study that experimentally manipulated autonomy in relation to assent/consent processes within a randomized trial. The type of assent/consent procedure did not affect the participation decision or willingness to participate in the asthma study; almost every family agreed to enroll in the asthma trial. This highlights one challenge in studying the impact of different consent processes on the willingness to participate in research; the difficulty in isolating the consent decision so that it can be experimentally manipulated. The reality of many consent/assent processes is that study information may come from a variety of informants (e.g. advertisements, recruitment brochures, physician referrals, research recruitment staff) long before the assent/consent conference, and, as a consequence, some filtering and predisposition to enroll occurs over the period of time prior to the formal assent/consent conference.

Our study revealed that, depending on adolescent age, a separate informed consent procedure (which we conceptualize as an autonomy-enhanced consent procedure) significantly improved the quality of information processing on the knowledge and appreciation of study risks and benefits. Improved knowledge of study risks and benefits was observed for adolescents 15 years of age and older, and for their parents, when the trial was explained separately. Unfortunately, this autonomy-enhanced procedure did not result in significantly better knowledge of study risks and benefits for younger adolescents (age 12–14) or for their parents.

One possibility is that we observed a “diffusion of responsibility” in the usual assent condition. Our coordinator ratings provide tentative support for this finding, indicating that the usual intervention resulted in decreased coordinator observed engagement with parents and adolescents. Parents of older adolescents may be expecting their adolescent to take responsibility for the study participation decision, and consequently listen with less attention when the adolescent is present. This may be particularly relevant for a non-therapeutic research study where the risks of participation may be perceived as low. Similarly, an older adolescent may expect their parent to attend to details of risks and benefits since they have historically been less involved in and have less experience making medical-related decisions. However, when separated, both parents and adolescents may accept greater responsibility for the decision and attend more closely, which has been suggested by others (Broome et al. 2003). We did not find the same pattern of responses for parents of younger adolescents or for younger adolescents. Their knowledge of the risks and benefits of the study were the same regardless of the intervention. It may be that parents of younger adolescents intended to make the participation decision for their youth and so were unaffected by the intervention, and similarly, younger adolescents intended to acquiesce to their parents’ views.

One remaining question is why this pattern of findings was seen for the risks and benefits knowledge category but not for asthma trial medicine, research process, or research rights and privileges categories. These domains may have a different level of salience for adolescents and parents that consequently affects their attentiveness. For example, in regard to the medication domain, it may be that adolescents and parents of adolescents are more familiar with issues surrounding medication, or view them as more important, and are therefore more likely to attend to this information in all circumstances.

Finally, we observed similar patterns of responding and fairly dramatic differences in the level of knowledge for all age levels across the four knowledge and appreciation domains. Of particular interest was the finding that younger adolescents were much better able to comprehend information on their rights and privileges and idiosyncratic research processes of the trial than the other knowledge categories. These research processes represent the most concrete aspects of what the adolescent will be required to do in the study. This suggests that, when assenting younger adolescents, specific attention to explaining procedures and rights will offer the greatest opportunity to include them meaningfully in the assent process. However, younger adolescents in our study appeared to have difficulty comprehending issues related to study medication and risks and benefits. The language used to explain these concepts may have been too complex for younger adolescents to fully comprehend. This suggests that when seeking assent from younger adolescents, researchers may need to pay careful attention to explaining medication and study risk and benefit issues.

Our findings point towards future research challenges, including processes for valid informed consent with adolescents who are compromised by disease or neurodevelopmental related competencies (e.g., brain tumor; intellectual disability), let alone disease progression that can, over time, adversely impact competency. Further complicating these issues is an adolescent’s motivation towards participation and pre-existing experience of empowerment in complex decision making. These challenging problems warrant research designs that are both focused and sensitive to these delicate situations.

There are limitations to our study and to the generalizability of our findings. Asthma research provides unique opportunities to study adolescent autonomous assent, since research participation is almost always optional, in contrast to biomedical treatment for other childhood conditions (e.g., cancer). It is possible that our findings were due to unique features of asthma research or this specific research protocol and may not generalize well to other biomedical research. In addition, differences observed in satisfaction with the consent may have an association with reading level of the consent utilized. Consequently, the findings from this study need to be replicated and verified with other disease research.

Research on adolescent assent is complicated by difficulties with realistically extending findings from a single decision-making encounter to other biomedical research decisions. Scherer and colleagues have previously highlighted these challenges, particularly as they pertain to research decision making across pediatric diseases (Scherer et al. 2013). The body of research in this area is still quite small, and caution is appropriate when considering generalization of decision-making findings to other pediatric diseases. As Scherer acknowledges, disease-specific and trial-specific factors may limit generalizability yet call for identifying common study elements that may promote improvements in obtaining valid adolescent assent.

The results of this study provide valuable guidance for conscientious researchers seeking an optimal method for assuring quality assent/consent for adolescents and their parents making biomedical research participation decisions. Holding separate assent/consent conferences for adolescent and parents may be a better method of providing research information and seeking assent from older adolescents, at least as far as appreciating research risks and benefits are concerned. This approach has the added benefit of optimizing adolescent autonomy in research participation decision-making. Moreover, our results suggest this approach can be utilized without impacting the probability of an affirmative research participation decision. Giving older adolescents decision-making responsibility and choice may further improve their willingness for participation and the quality of data they provide. The more traditional method of seeking assent/consent by addressing parents and adolescents together may be a better option for younger, less mature adolescents who have a lesser expectation for decision-making autonomy and who may be more vulnerable and in need of parental support for biomedical research participation decision-making.

Acknowledgments

Abbreviations

Footnotes

References