The invisible child of personalized medicine

ABSTRACT This article seeks to bring the invisible child of personalized medicine out from the shadows through legal analysis and empirical data. It uses Denmark as a case to argue that existing policies, laws and practices on personalized medicine neglect the legal and ethical issues specific to children. The article investigates Danish laws and practices in clinical genetics and describes how the law neglects children’s right to self-determination in three ways. Firstly, while child participation is provided for by law, no guidelines have been created to operationalize this norm. Secondly, children’s right not to know is inadequately reflected in current policies. Thirdly, the storage of information from prenatal genetic sequencing raises important issues that are in need of reflection. Several recommendations are made, including for strengthening children’s participation and limiting parents’ access to secondary findings where they relate to untreatable or unpreventable conditions. It furthermore recognizes, however, that children’s self-determination in some circumstances should be viewed relationally due to the interconnected nature of genetics.


I. INTRODUCTION
Governments, researchers, and other stakeholders increasingly promote personalized medicine as cutting-edge health technology that will improve patient and societal † Katharina Ó Cathaoir is Associate Professor at the Faculty of Law, University of Copenhagen. Katharina researches in Danish and global health law, with a focus on the implications of Big Data for the human rights of patients. Katharina is affiliated with several interdisciplinary research projects, including Personalized Medicine in the Welfare State (MeinWe) based at the Department of Public Health, University of Copenhagen. In this project, Katharina explores the Danish legal landscape governing use of genetic data in healthcare and research, focusing on patients' rights and vulnerable groups, including children and minorities.
secondary findings related to children's future health mandate adequate reflection on how best to preserve this 'special category' of data. This article therefore seeks to bring the invisible child of personalized medicine out from the shadows through legal analysis and empirical data. It examines the Danish approach as a case study of how the law governs (or fails to govern) children's rights and interests within genetics. The article focuses on the Danish legal context: Denmark has established a National Genome Centre (2018), launched a policy strategy on personalized medicine (2016) and amended the Health Act (2018) to facilitate personalized medicine. 10 Furthermore, the National Genome Centre has selected 12 patient groups to be offered whole genome sequencing as a priority, including children and young persons with cancer and psychiatric conditions, as well as fetal medicine. 11 While the neglect of the ethical and legal issues specific to children in government strategies and law making does not necessarily mean that children are not being given any consideration, it does signal that children's rights and interests are not center stage. It also suggests that Denmark is not applying long-standing ethical, legal, and social science (ELSI) research on genetics and children to personalized medicine, which this article seeks to remedy.
As self-determination is a central value of the Danish health system and the Health Act, the article explores to what extent the law guarantees the latter value for child genetics patients. In the article, law and practice is presented with reference to international norms on children's rights, primarily Articles 3 and 12 of the CRC and Council of Europe biomedicine treaties, as well as academic literature and recommendations. Furthermore, ELSI research relating to minors and personalized medicine was reviewed to contextualize and assess the legislation.
The article finds, firstly, that the law seeks to provide a framework for children's right to information but in line with previous research, practical barriers remain. Secondly, children under 15 do not have legal capacity, which means that they do not control return of secondary findings and thereby, their right not to know is not given legal protection. This is especially clear in prenatal sequencing as, because the fetus lacks legal personhood, secondary findings are not saved in the future child's health records.
In response, section 4 makes recommendations for how these concerns could be mitigated in a manner that upholds children's best interests. The article recognizes the complex ethical issues raised by advanced genetic sequencing and acknowledges that self-determination is rarely absolute. It suggests that the law on genetics should be interpreted through the prism of relational autonomy, provided adequate safeguards are ensured. This includes strengthening children's participation and limiting the longevity 4 • The invisible child of personalized medicine of parents' decisions. Technical solutions should be developed to ensure that data and results are later accessible.
At policy level, the author hopes to contribute to important questions, such as, should children's genomes be sequenced before they can provide a valid consent, and should advanced sequencing be permitted on healthy children? When is it legitimate for a tax-funded health system to use genetic sequencing as a diagnostic test for children? Which findings should parents and children receive? How are the role and responsibilities of geneticists challenged by personalized medicine?
The article begins by introducing personalized medicine and its implications for children, followed by outlining the methods used in this piece. Thereafter an analysis of Danish law and practice is presented with a focus on children's self-determination. Finally, recommendations are made for how the concerns explored in section 4 may be addressed.

II. WHAT IS PERSONALIZED MEDICINE?
Personalized medicine marks a shift from a 'one size fits all', population-based approach to healthcare to individualized treatment and prevention. In the twentieth century, as medicine became increasingly evidence-based, randomized clinical trials became the gold standard, with clinical application based on patients' symptoms or diagnosis. 12 However, for many patients, the prescribed treatment still does not produce the desired clinical effect or no satisfactory diagnosis is reached. Personalized medicine seeks to address this societal challenge. Big data is central, spanning omics, in particular genetic data, but also patient records and registries. Algorithms analyze these data to identify patterns in datasets and stratify patients, with a view to developing more effective, targeted treatment and prevention.
There are multiple definitions and descriptions of personalized medicine. At a European level, the following definition was relied upon by the Council of the European Union: [personalized medicine is] a medical model using characteristics of individuals' phenotypes and genotypes (eg molecular profiling, medical imaging, lifestyle data) for tailoring the right therapeutic strategy for the right person at the right time, and/or to determine the predisposition to disease and/or to deliver timely and targeted prevention. 13 The Danish strategy emphasizes genetic data as a means of achieving better diagnosis and treatment options, while recognizing molecular profiling, imaging, and usergenerated data as increasingly relevant in the coming years. 14 The benefits of personalized medicine are purported as earlier diagnosis, prevention, and more effective treatments. 15 Faster diagnosis can spare families the strain of a 'diagnostic odyssey' and facilitate access to social supports, as well as inform reproductive decision-making. 16 While a diagnosis is not a legal requirement for social benefits, Danish parents interviewed in connection with this research project spoke of how a diagnosis was often requested by social welfare authorities and made access easier. 17 Prospectively, identifying an increased risk of developing certain diseases, notably breast, ovarian, and colon cancers, could allow for increased screening and early intervention. 18 Pharmacogenetics, including through genome wide association studies identifying biomarkers that signal an adverse reaction to certain drugs, can avoid unsuitable treatments that produce harmful side effects.
Personalized medicine for children is promising for several reasons. There are relatively few clinical trials on children, meaning that childhood diseases remain illunderstood. 19 Children have instead been treated as 'small men', 20 with off label drug use common. Studies have however shown that some children experience different reactions to certain drugs than adults, notably Warfarin. 21 In cancer care, personalized medicine is encouraging as childhood cancers carry relatively few mutations (compared to in adults). 22 Personalized medicine could also have significant implications for rare diseases, which overwhelmingly affect children, and for which clinical trials are often viewed as too expensive. 23 Several studies point toward the potential of personalized medicine to, for example, treat cystic fibrosis. 24 Furthermore, while the focus of personalized medicine until now has been on patients with cancer and rare diseases, most strategies, including the current Danish strategies, discuss expanding sequencing to healthy individuals who would act as controls. Several proposals have been made for whole genome sequencing of newborns as part of postnatal screening. For example, in November 2019, the then British Health Secretary announced his ambition for WGS be carried out on all newborns in Britain. 25 This would allow diseases to be detected in children before they have decision-making capacity and potentially result in their biological material being stored and used as a research 'resource' . If such strategies were realized, the laws relating to genetics would have relevance for all newborns.
Personalized medicine thereby raises legal and ethical questions, including concerns regarding informed consent, data protection, autonomy, and privacy. The expansion of personalized medicine will require gathering, sequencing, storing, and sharing of genetic and health data. Routine genetic sequencing of children raises distinct concerns, as minors cannot consent. It is therefore timely to analyze how children and genetics are currently regulated.

III. LOOKING TO CLINICAL GENETICS: DANISH LAW AND PRACTICE
Having introduced personalized medicine as an emerging healthcare technology that (in the Danish Strategy) relies on advanced genetic sequencing, this section outlines current law and practice as a case study on children's self-determination within clinical genomics. The ensuing analysis relies on a legal doctrinal review of the legislation governing clinical genetics (the Health Act and related secondary legislation and guidance, as well as intersecting legislation governing children in health care). In the below analysis, I evaluate domestic provisions in light of the CRC. The CRC is an internationally binding treaty, which Denmark has ratified but not incorporated into national law. 26 Therefore, as a dualist legal system, the Convention does not have direct effect in Danish law and a domestic law cannot be set aside based on the CRC alone. 27 Thus, my criticism of Danish law in relation to the CRC relates to the state's obligations.
In Denmark, court cases are rare in the field of health law and there have been no judicial pronouncements on how the sections discussed in this article should be or are interpreted in practice. For this reason, there is a gap between the law 'on the books' and how the law is applied in the clinic. Those interested in ascertaining the content of the law on children and genetics therefore need to look beyond the legislative text, which this article aims to do.
In order to do so, I conducted semi-structured interviews with 11 clinical geneticists to gain insight into their daily experiences with interpreting and applying the legislation in question. The central focus of each interview was an assessment of whether the law is adequately equipped for the introduction of personalized medicine in Denmark and what barriers, if any, they considered the law posed for their ability to diagnose and treat children.
Clinical geneticists were chosen as they play an important role in interpreting and implementing law on a daily basis. 28 Previous research has highlighted the 'significant degree of responsibility' given to professionals who make key decisions in the local pediatric genetics context. 29 Clinicians from each of the national clinical genetics departments were interviewed with a view to gaining a nationwide perspective. Danish healthcare is divided into five regions within which there are seven departments of clinical genomics spread throughout Denmark. 30 Interviewees were selected through contact to the head of the clinic and for their experience with pediatric patients. The interviews were undertaken from September 2018 to May 2019 and therefore reflect the law in force at that time. An interview guide was prepared with a list of interview questions which all interviewees were asked, however, the interviewees were also free to stray if they wanted to raise other issues. The interviews (besides one group interview) were transcribed and subsequently coded with a focus on self-determination.
In the subsequent analysis, I shine a light on the limits to children's selfdetermination in the current governance regime. Firstly, despite legislation requiring child participation in medical decision making, practical barriers remain. Secondly, children's status under law means that their right not to know is not given legal protection, in contrast to the position of adults. Thirdly, the challenge of safeguarding the right to information in prenatal genetic sequencing is introduced. In section 4, the implications of these findings are discussed and suggestions are made for how children's self-determination could be encouraged through relational autonomy.

III.A. Children's Participation in Consent to Genetic Sequencing
In Denmark, the central piece of legislation governing genetics in healthcare is the Health Act (Sundhedsloven). The law, inter alia, 'sets out the requirements for the health care system in order to ensure respect for the individual, their integrity, and selfdetermination' . 31 The Act regulates a number of important areas, such as the duties of the healthcare system, municipalities and hospitals, patient safety and the rights of patients, including, self-determination, freedom of information, and confidentiality. It entered into force in 2006 and has been amended numerous times since. In July 2018, amendments to the Health Act were adopted to create the national infrastructure deemed necessary to facilitate the introduction of personalized medicine in the clinic, including establishing the National Genome Centre. 32 The Minister for Health issued secondary legislation in February 2019. 33 While the Health Act was amended to facilitate personalized medicine, the legal and ethical position of children was not addressed. Thus, as before the amendments, no specialized legislation governs genetic sequencing of children in Denmark. 34 Instead, general legislation applies, whereby minors can consent to treatment, including genetic analysis from age 15. 35 In contrast to other countries, 'mature minors' do not have capacity to consent to treatment before age 15. 36 Furthermore, until age 18, the child's guardian should be informed and consulted in healthcare decisions. This provision reflects a compromise as children aged 15-17 years remain minors, to whom their guardians owe duties of care. 37 Alternatively, it can be criticized for encroaching on children's privacy and autonomy, especially in intimate areas such as access to contraception. However, the Agency for Patient Complaints and the Disciplinary Board have found that parents' entitlement to information can sometimes be limited. 38 For example, the Disciplinary Board has stated that parents' access to information should be viewed in light of the treatment's character, the seriousness of the illness, the type of information, the young person's age and maturity, as well as the need for follow up at home in light of parents' caring responsibilities. 39 Still, more clarity as to the circumstances under which physicians may legitimately withhold information would create greater certainty for all parties. Furthermore, although minors under 15 do not hold decision-making capacity, the law requires that they be facilitated in participating in healthcare decisions. Healthcare workers must inform the child to the extent they understand the treatment and unless informing would cause harm; the child's views should be given weight to the extent they are current and relevant. 40 Ultimately, however, decision-making lies with the guardian, who must make medical determinations based on the child's interests and needs. 41 Danish law thereby recognizes elements of article 3 CRC (best interests) and article 12 CRC (participation). However, whereas the CRC requires that all children capable of forming views have a right to express them, the Danish provision states that children should be informed unless it would cause harm. The Committee's interpretation of the Convention, while not binding, does not use harm as a reason for limiting participation (but does specify that states should ensure that the right is not misused in a manner that is harmful to the child). 42 In fact, the Committee has stated that 'there is no tension between articles 3 and 12' . 43  'adolescents need to have a chance to express their views freely and their views should be given due weight'; again without limiting the provision through the prism of harm. 44 Commentaries on Article 12 likewise do not mention that participation is subject to the harm principle, instead focusing on the benefits of participation to the health of the child. 45 Specifically, children should be presumed competent, which I consider rendering involvement subject to harm to go against. 46 It has been underscored that 'children's capacity to express views must not be assessed through the experiences and expectations of adults as to what constitutes a valid or appropriate view deserving of recognition or attention' . 47 In contrast, Caroline Adolphsen concludes that while Danish legislation does not provide as much empowerment as corresponding Norwegian legislation, it complies with the CRC because no age limit is specified for when children should be involved in medical decision-making. 48 However, in my view, this interpretation is not in keeping with the CRC nor prominent academic literature, given that harm is not mentioned as a criterion for limiting children's right to be heard. 49 Focusing on harm risks rendering participation vulnerable to paternalism, instead of recognizing children as rights holders, which is the aim of the CRC.
The prepatory works of the Health Act furthermore suggest that children should normally be informed, stating that children under 15 are often capable of understanding, at least important elements, depending on their development and maturity. 50 At the same time, the prepatory works also make clear that the parents are expected to attribute weight to their children's views but the provision does not give physicians the competency to overrule parents' decisions if they are in conflict with the child's wishes. 51 During my interviews with geneticists, it became clear that despite the legislative intent, there are practical barriers to operationalizing participation in genetics. 52 Central challenges include the complexity of genetics, as well as the uncertainty of potential findings. With advanced sequencing, geneticists no longer determine whether a child suffers from a monogenetic disease but instead try to identify mutations that explain a phenotype, and thereby could lead to a diagnosis. A further challenge is that many genetics patients are younger children with cognitive impairments, who often have difficulty concentrating, processing information, and conveying their needs. Finally, for families that do not (fluently) speak Danish, an interpreter is required to ensure that valid consent is given, which can further limit the child's participation, depending on the quality of interpretation. 53 Many geneticists, when asked, relayed that they perceived it almost impossible to include children in the consent process. Cassandra, an experienced clinical geneticist with a strong interest in children's rights, thoughtfully relayed: In rare cases I do [include children], but many of the children [I see] are mentally disabled, which makes it even more difficult. Besides, these are some complicated issues, so I think that in many cases it would largely be impossible to involve the child. Many times, if they are smaller children, [the parents] do not have them with them because it often just becomes a bit distracting.
This observation aligns with previous research, which suggests that while international and domestic law support child involvement in healthcare decision-making, operationalization remains challenging. 54 Barriers include protective attitudes of families and physicians, and the sociocultural context. 55 Yet, interviews with children show that most would like to be informed and allowed to voice preferences, 56 even where children agree that decision-making be delegated to their parents. 57 Furthermore, failure to respect children's right to participation based on lack of resources (such as time) may amount to a state violation of the CRC. 58 Under the CRC, a child 'capable of forming his or her own views' has a right to express such views, which should be 'given due weight in accordance with the age and maturity of the child' . The formulation of the text suggests that in some circumstances it may not be practicable to involve children in decision-making. However, the CRC Committee considers that Article 12 must be interpreted in light of children's right to freedom of expression under Article 13 and has taken a more radical approach, stating that all children, regardless of age, should be actively listened to, including through nonverbal forms of communication. 59 Furthermore, the CRC must be interpreted in light of the Convention on the Rights of Persons with Disabilities (CRPD) and it should not be presumed that children with disabilities are incapable of forming views. Instead, information should be tailored to their level of understanding in light of age and maturity. While legislation requires children to be informed and their views solicited in decision-making on genetic sequencing, clinicians appear to experience barriers to realizing this obligation. This is particularly problematic for self-determination as the results of advanced genomic sequencing can relate to the child's future health, as explored in the next section. Furthermore, the CRC Committee has held that providing training on Article 12 to professionals working with children is a core obligation. 60 Notably, in 2011, the CRC Committee recommended that Denmark 'ensure, through appropriate training, that all professionals and staff dealing with children's issues are informed and competent to support the expression of children's views' . 61 Yet, child participation is never mentioned in Danish political or legal documents on personalized medicine.

III.B. The Right Not to Know
Under Danish law, patients have the right to receive adequate healthcare but are generally not entitled to demand specific treatments. Likewise, in genetics, legislation does not specify for which genetic conditions children can be tested and this instead lies within the medical professional's discretion. In the absence of law, ethical and professional norms have filled the vacuum and traditionally protected the child's right to know. However, advanced genetic sequencing challenges these (largely) unwritten rules as these techniques can unintentionally reveal secondary findings. This can lead one to question whether children under 15 in fact have a right not to know in genetics.
Geneticists usually refuse to test minors for conditions that debut in adulthood as this is widely regarded as unethical. 62 The traditional position is that out of respect for children's autonomy and right to an open future, their opportunity to decide whether to be informed of genetic predispositions should be safeguarded until they reach adulthood; parents should not make this decision. The classic example is Huntington's disease-an illness that cannot be prevented and will usually manifest in one's thirties. This position is empirically supported by data that show that adults at risk of Huntington's disease often decide against testing. 63 Yet, for some parents, it is difficult to accept this principle. I have had parents who became enraged with me because I did not want to examine their 5-year-old child for the mother's mutation, which would first affect the child as a 50-yearold. . . . Well, I said, we don't [screen children]. It is child abuse to make presymptomatic diagnostics on minors.
Suspecting the doctor was being slightly ironic in his remarks, I asked again whether he really thought testing is such cases amounts to abuse: Yes, if the child is [legally] incompetent, there are some who believe that it is child abuse, because the child must be allowed to decide . . . Pausing, he stated 'it has no consequence for their treatment' . The doctor's experience reflects the pressures that geneticists face from concerned parents, keen to avail of technological advancements. It illustrates geneticists' gatekeeper role, whereby they must negotiate expectations and sometimes limit access to technology based on unwritten ethical norms. As the doctor highlights, testing for such conditions has no consequences for the child's current health.
However, personalized medicine challenges this prevailing approach as it relies on next generation sequencing, which carries a risk (1-10 per cent) of secondary findings, ie findings that go beyond the original aim of the testing or the patient's phenotype. 64 While analyzing a patient's genome, variants relating to adult onset diseases may be 'accidentally' revealed. To safeguard patients' right to self-determination, the Danish National Genome Centre has developed a national consent form, which requires patients to decide whether they wish to be informed of findings 'of importance to health' (1) where there is an option of treatment or prevention, (2) where there is no treatment or prevention (3) or not to be informed of any secondary findings. 65 In the case of children aged under 15, their guardians make decisions on secondary findings. On the one hand, this reflects the standard legal position for children. Hens et al. point out that parents already obtain all sorts of information related to their children's future, such as IQ. 66 Two geneticists also presented this view; Ingrid stated: The parents are the guardian of the child and [their decision] is what I have to follow. Just as parents are also responsible if their [child] doesn't put socks on that day and gets sick or doesn't put their glasses on and gets run over . . . I think that is the responsibility of the parents. And if I did it differently, I would be a guardian of the child. I would not think that was right . . .

Another concurred separately:
The child did not ask to have a genetic test, and of course it is true, but the child also did not ask to come for an eye examination . . .
The child thereby in a limited way ceases to be the only patient. This echoes the recommendations of the European Society on Human Genetics (ESHG) that if an incidental finding suggests 'serious health problems' in either the person being tested or close relatives, where treatment or prevention is possible, the healthcare professional should report the findings. 70 However, the ESHG also recommends that in the case of children, guidelines should be established to 'balance the autonomy and interests of the child and the parental rights and needs (not) to receive information that may be in the interest of their (future) family' . 71 Furthermore, while the law does not draw any red lines in terms of which secondary findings are not in minors' interests, geneticists appeared to. Several declared with conviction that they would not inform a family if they found a mutation for early onset dementia: 'that I would struggle to put on a child'! This consensus seems to be based on the absence of prevention and treatment options currently available for dementia, coupled with the heavy burden of the information. It suggests that in spite of viewing parents as decision makers, geneticists are generally willing to limit parents' access to information, where no health intervention is possible. This is in line with doctors' professional responsibilities to safeguard children's interests.
Finally, a further unresolved issue raised by Adolphsen is whether parents can receive information about their child aged 15-17 where the child has exercised their right not to know. In genetics, hypothetically this could arise where the child does not want, for example, information on secondary findings but the parent does. Adolphsen suggests that parents should normally be considered to have an independent right of access to information although this could be limited based on the child's best interests. 72 The Danish strategies and political process never discuss whether these current practices are in children's best interests or should be reconsidered. The current consent forms and policies on return of secondary findings appear adult focused, lacking reflection on child patients. This is of concern given that the National Genome Centre has prioritized children with certain diseases.

III.C. A Right to Information in Prenatal Sequencing?
Finally, an interesting issue that emerged during interviews was the status of the right to information in prenatal genomic sequencing. The latter is offered to couples with a history of fatal fetal abnormality. 73 A goal of the analysis is to inform the woman's decision on whether to continue the pregnancy. In Denmark, abortion is a right until 12 weeks gestation. 74 After this, the woman has to apply to a regional administrative body to be granted permission to have an abortion. 75 Abortion can be granted, inter alia, based on hereditary disease or disease that will cause serious mental or physical illness. 76 The further question can arise as to whether and which secondary findings geneticists should return in these cases. Previous research suggests that there is no agreement among health professionals as to which findings should be returned. 77 Interviews conducted with 27 laboratory personnel revealed that they did not routinely report variants of unknown significance (VUS) to clinicians, especially if unconnected to the phenotype. 78 Uncertainty has been identified as a central element of fetal genome testing due to VUS and secondary findings. 79 The return of VUS can have unforeseen consequences for the future child, as explained by Irmgard: . . . the things I have said [to mothers] about a potential risk continues to be a blemish on the children involved . . . the mothers are nervous about risks . . . even if they have given birth to a perfectly healthy child . . . . how do I keep my information from following the child, even if the child is completely healthy?
In a recent article, Lou, Jensen and Vogel explore mothers' perspectives on return of VUS in Denmark. While the women's opinions on receiving uncertain results varied, some struggled with the decision of deciding which results are 'important' . 80 Danish law does not help to clarify this uncertainty. The consent form developed by the National Genome Centre uses the term 'important health related secondary findings' . 81 While the consent form must be used whenever advanced genetic analysis is performed, it does not have a legal status. As the phrase 'important' is not used in the right to information under the Health Act, a conflict could arise if a higher level of information is found to be required by law. This leads to uncertainty for the health professional as to which information they must return and which they can deem too uncertain or remote to share. There is a risk that the Agency for Patient Complaints or the Disciplinary Board will interpret 'important' differently to geneticists.
Finally, under Danish law, health professionals are required to record results of procedures in the patient's health record. 82 As the fetus lacks legal personhood (and thereby an independent health record) and the mother is the patient, any findings will belong to her health record. If the mother continues the pregnancy, the resulting data may have implications for the future child's health but not be easily accessible to them. This can have implications for the child's self-determination, health, and privacy. In contrast, if the woman terminates the pregnancy, the information remains only retrievable by her, although it may also have relevance to the father. Yet, none of these potential concerns are found in the Danish strategies or legislation, despite fetal medicine now forming a central part of the Ministry of Health's strategy for personalized medicine.

IV. IMPLICATIONS FOR CHILDREN AND THE FUTURE OF PERSONALIZED MEDICINE
This article has noted several aspects of Danish law on genetics that have neglected children's rights, focused on the rights to participation and information. These issues should have been addressed or at least discussed in the strategies and political debates on personalized medicine, instead no mention of them is found in publicly available documents.
The first issue relates to children's legal status; minors under 15 years cannot make decisions relating to their treatment. The strict age limit can be criticized for failing to take children's evolving capacities and circumstances into account. Research highlights, for example, that children with disabilities and histories of illness may mature earlier and have insights and experiences that their guardians lack. 83 While legislation seeks to give children a voice in healthcare, operationalizing this norm is challenging. Literature highlights that children should not be assumed incapable of participation; all age groups can struggle to understand the implications of genomics. 84 It is submitted that guidelines on how to pursue children's participation in personalized medicine would be beneficial. It is a lost opportunity that the Danish personalized medicine strategies and law reform process have not reflected upon how to address the vulnerabilities of those who lack decision-making capacity. In contrast to the British 100,000 genomes project, the Danish National Genome Centre has not created publicly available, child-centered information. Previous research has identified, for example, training health professionals as an important aspect of improving child participation. 85 Under Danish law, it is the clinician's responsibility to determine how a child should be included in a manner that includes their views, while not proving overly burdensome for the child. 86 Therefore, it is recommended that information directed at children be made available to ensure they are informed of genetic sequencing and the potential results, even when they do not have legal capacity to consent. Furthermore, professionals should be assisted in fulfilling their obligation to inform all patients regardless of legal capacity through for example, training. This would furthermore be in keeping with the recommendations of the CRC Committee, as introduced above.
Secondly, in light of children's best interests and autonomy, I conclude that the current options for return of secondary findings are unsuitable for children. Instead, it is usually in the best interests of the child that treatable/preventable findings with relevance to the child's current health be communicated, regardless of parents' preference. For preventable/treatable findings that relate to the child's future health, an individual assessment should be carried out as to whether the child is sufficiently mature and early information would thereby be in the child's best interests. Through participation, children must have a role to play in determining their best interests. 87 This is in keeping with the interpretations of best interests established by the CRC Committee. As the CRC Committee views it, 'a larger weight must be attached to what serves the child best.' 88 For example, Helen Stalford argues that 'it is generally accepted (in principle at least) that it is right and fair that children's interests should carry more weight in such decisions because their outcome is likely to have much more profound effects on children in the immediate and longer term' . 89 The CRC Committee suggests that 'if a legal provision is open to more than one interpretation, the interpretation which most effectively serves the child's best interests should be chosen' . 90 Best interests can also be understood as a form of child autonomy: as a means of ensuring children's right to an open future, ie ensuring that children enter adulthood with the maximum available options. 91 However, child autonomy is not absolute, and should in some circumstances be viewed in relation to the family. Maya Sabatello et al. posit that reliance on individualized autonomy may be unsuitable for children and instead favor a relational approachpersonalized autonomy. 92 This approach favors viewing children and guardians' interests as intertwined, given that child health is highly dependent on the parents. Notably, child-adult dependency may be heightened in a family with a sick child where a parent plays a significant caregiving role.
Still, there should be limits to the level of information that could be provided purely by reference to relational autonomy. To inform parents of a disease that cannot be prevented or treated risks instrumentalizing the child. In her relational account of autonomy and genetics, Anne Donchin reflects: Children need protection lest they be treated, Cinderella-like, as mere conveniences to advance others' ends . . . justice would not be served unless they set constraints on moral sacrifices family members can be expected to make for one another. 93 Therefore, I recommend that variants related to adult-onset disease with serious implications for close family members could exceptionally be communicated with reference to relational autonomy. Under this view, information regarding, for example, BRCA (increased risk of breast and ovarian cancer) or lynch syndrome (increased risk of colon cancer) could be returned given the risk of a parent developing a lifethreatening form of cancer that may require invasive treatment. For preventable or treatable health information that does not need to be acted on immediately, guardians and children should be given the choice to reject receiving such information. Again, depending on the child's age and maturity, she should play a role in this determination.
In my discussions with geneticists, it was clear that a form of relational autonomy was being practiced. Geneticists reflected on the child's interpersonal context: did the parents have other children/plan to have other children, were there tensions in the family, reasons in the family history that suggest against informing, such as a history of abuse? The doctors highlighted that the patient-not the family-was their primary concern but were keenly aware that the former could not be completely isolated from the latter.
Paradoxically, geneticists were concerned that by moving genetics into mainstream healthcare, the degree of personalization currently provided would not translate. This is because doctors without specialization in genetics will increasingly order sequencing and return findings to patients. These doctors may not have adequate time to get to know the person behind the patient, nor detailed knowledge of the potential variants that could be returned. While today clinical geneticists with a research background are often involved in interpreting variants, the planned expansion of genetic sequencing would render this unfeasible.
Furthermore, non-treatable/preventable findings should not be communicated in childhood as neither the child's-nor relational-autonomy provides an adequate justification. While literature suggests that most adults would find this information useful for life planning and personal empowerment, the nature of the information strays far from the phenotype that triggered the sequencing. Furthermore, sharing this information exposes children to privacy risks that are in need to greater consideration, such as 'sharenting' on social media or uploading children's data to nonsecure websites. 94 Children may also be discriminated against based on genetic diagnoses; in February 2021, the Danish Supreme Court held that an insurance company may ask whether a child had been referred to a clinical genetics department and that the presence of an inherited mutation (without symptoms) amounted to a current condition and was thereby not protected by law. 95 Thus, revealing that a child carries a BRCA or other mutation could have long-term consequences for the child, such as exclusion from certain insurances.
In the case of prenatal sequencing, the absence of guidelines leaves professionals in a difficult position: do they inform parents of variants of unclear significance based on the patient's right to information, or withhold data due to their uncertainty and potential to stigmatize a future child? In the Danish context, it could be argued that only information that helps to inform reproductive decision-making, ie indicates serious illness, should be given. If a mother receives VUS that do not relate to serious illness, she is unlikely to be granted an abortion on this basis. 96 This question is perhaps best resolved with reference to the obligations of the professional, upon which one geneticist reflected: I work with sick children. I am not interested in whether someone is disposed to dementia. That is not realistic as there is neither time nor money for that. Doctors should treat sick patients-not do consumer genetics.
Similarly, Clarke and Thirlaway reflect that, with the expansion of the field, genetics professionals may become 'lifestyle coaches, requiring a very different set of skills' . 97 In the Danish, tax-payer funded healthcare system, this would seem an inappropriate use of resources. Similarly, the British 100,000 Genomes project does not transmit secondary findings where the clinician is in doubt of the health consequences of the result or the information is of nonmedical character. 98 Therefore, while patients have a right to information, it should not be viewed as absolute and should focus on the health of the child.
Finally, the question should be asked as to what becomes of future information revealed about a child or fetus. Given the highly digitized nature of Danish healthcare, 99 it has been suggested that such information should be stored and made available to the child later in life. For children under 15 this could be achieved, for example, through making such findings inaccessible in the child's electronic health record until they reach 15 and thereafter informing the child of the possibility of accessing the results. In the case of prenatal sequencing, it has been suggested that a separate fetal health record could be created and should the fetus be born, the information be transferred to the child's health record. 100 In conclusion, three years after the entry into force of the legislation establishing the Danish National Genome Centre, uncertainty remains in several areas related to children and prenatal sequencing. These issues should be addressed both in light of children's rights and to create clarity for clinicians as to their duties. 101