The results indicate incongruent approaches to addressing emerging issues, if indeed they are addressed at all. There are several explanations for this, and each issue may manifest different rationales for variability. With respect to consent, for example, it is generally accepted that the failure to address in some capacity a child’s assent and then later consent as they mature (if the research project continues over a prolonged period) can undermine the integrity of the project and create schisms in the current and future protection of a child’s wellbeing and developing autonomy. These issues are of particular importance in longitudinal and biobanking studies that intend to use samples and data indefinitely. But re-contact, perhaps in studies not specifically longitudinal in nature, may be regarded as unforeseen, unfeasible, or unrealistic. Hence, an explanation for the common implementation of ‘broad consent’ in the consent forms we reviewed may be that it is viewed as the most practical, efficient, and appropriate scope of consent, provided the samples and data are coded rather than anonymised so that re-contact is possible. It should be noted that broad consent was largely seen, as expected, in the longitudinal and biobanking research studies.
That the majority of the consent forms we analyzed did not address cumulative or non-physical risks may speak to the definitional ambiguity of ‘risk’ and the tendency to focus on physical risks. Risk is defined in Canada’s 2010 Tri-Council Policy Statement (TCPS) c as ‘a function of the magnitude or seriousness of the harm, and the probability that it will occur’ (TCPS, Ch. 2B), but this does not address the ambit of harm. In paediatric research, harm may encompass psychological, social, financial, and community risks, particularly in genetic, biobanking, and longitudinal research. This is more than an ethical concern. In Canada, full disclosure of risks in research is legally required [28, 29]. Along the same lines, given that most forms stated that indirect benefits (e.g. a societal benefit from biomedical advancement) would be achieved, one must conclude that the research projects entailed no more than ‘minimal risk’ in order to receive REB approval, as required by the TCPS (Art. 4.6). Yet, it remains an open question whether these projects truly pose a minimal risk if they do not disclose non-physical risks, consider cumulative risks, or consider risks from the perspective of the child whose perspective may differ drastically depending on age .
The lack of specificity in some of the forms that the child could withdraw rather than the parent providing the authorisation, and the lack of a resolution procedure for withdrawal conflicts between children and parents, may be due to normative guidelines that generally encourage, rather than require, researchers to respect a child’s decision to withdraw from research if the child has the capacity and maturity to make an independent choice [19, 31]. The consent forms that failed to clearly state to the parents that withdrawal may not in fact be absolute could be due to instances where data and samples are irretrievably de-linked to an identifiable person (i.e. they are anonymised). Yet, almost no forms indicated anonymisation was involved, so this is only a partial explanation. This also leaves open the question as to whether parents or children are aware that should they later change their mind about participating in a study that has anonymised their data or samples, destruction of their data or samples is no longer possible as they cannot be identified.
The return of research results has garnered significant discussion, especially with the advent of whole genome and exome sequencing [20, 32, 33]. Incidental findings, defined as ‘unanticipated discoveries made in the course of research but that are outside the scope of the research’ (TCPS Article 3.4), are becoming increasingly important as data-intensive science expands and next generation sequencing technologies are employed [34, 35]. Currently, there is neither national nor international consensus on the treatment of incidental findings in paediatrics, and the consent forms we reviewed reflect this variability and continuing lack of consensus, with respect to both return of research results and incidental findings. In cases where results reveal a clinically significant condition for which there is current treatment or prevention, the parents cannot refuse to know and the child’s right to medical care prevails [14, 21, 36]. In other situations, researchers – and parents – may opt to wait until the child is mature enough or reaches majority before disclosing research results if the results are not materially relevant until the child reaches adulthood. Either way, the potential for these situations, especially if their possibility is known at the time of consent, should be explained in the consent form and further elaborated upon by the individual obtaining consent. Similarly, consent forms should be clear if the child’s consent or assent to receiving the information would be sought so as to not compromise the child’s ‘right to an open future’ , and if the ‘right not to know’ necessarily includes the right not to have information included in the medical record if it entails an actionable result.
It is a positive sign that most of the forms we analysed provided standardised nomenclature for sample or data identifiability (e.g. ‘coded’), since terminological confusion has long been an issue . However, standardised nomenclature may be seen as only the first step to addressing privacy and confidentiality concerns. Challenges to privacy and confidentiality are amplified by genomics and other biomedical research projects , which are often internationally collaborative and engage in perpetual data linkage across jurisdictions. Biobanking or genetic research consent forms may need to declare that neither anonymised nor coded data and samples can guarantee privacy, as knowledge of even a small number of genetic variants can lead to matching of samples to individuals with a high level of confidence (we did not observe consent forms that disclosed this potential privacy risk).
All of the consent forms assumed parents could access research-related information about their child. And yet, multiple tensions can arise between the child’s privacy interests and the parents’ general legal right to their child’s health information [40–42]. For example, a child may not want her parents to know about a pregnancy test result or habitual drug use, but this desire could conflict with legal duties of parental access to health information that compel a researcher to disclose such information. At a minimum, consent forms should disclose the kind or extent of information communicated to the parent and information which shall require the child’s consent.
Lastly, while ethical norms generally support transfer of data and samples with certain safeguards, there is continuing debate about the parameters of that transfer and the various organisational safeguards, technological measures, and physical measures that should be adopted and updated – and disclosed – in the consent form. Research participants remain woefully uninformed of the transfer of their samples and data, particularly when it may carry a commercial use . Devising and disclosing a method for listing all approved projects that are accessing the data and/or samples could alleviate this. While it may be the case that all of the consent forms we analysed that did not address transfer of data and samples simply did not envision transfer, in the absence of explicit disclosure that ‘no transfer will occur’, uncertainty remains and creates the risk of future ethical concerns with maturing children.