Introduction
In Dance v Barts Health NHS Trust (Dance), Archie Battersbee, a 12-year-old boy, suffered from a catastrophic hypoxic-ischaemic brain injury and was kept alive only by mechanical ventilation.1 Accordingly, the High Court concluded that Archie had ‘no prospect of meaningful recovery’ and authorised the withdrawal of continued ventilation on the basis that it was not in his best interests.2 Applying the best interests test, the court also refused to allow the parents of Charlie Gard3 and Alfie Evans4 to take their children abroad for continued treatment and ventilation respectively. However, what distinguishes Dance from these cases is that the child’s wish was re-interpreted by the court to mean that he would desire to be kept alive only so long as there was a prospect of some recovery.5 These cases have sparked significant controversy as to whether the court should have the jurisdiction to intervene in such medical decisions made in the private family sphere based on the highly debated best interests test.
The current best interests test stresses that when doctors and parents disagree, they can bring the case to the court, where a judge will overrule either party’s decision when that decision is deemed not to be in the child’s best interests. This article will begin by analysing the current approach adopted by the court, then proceed to argue that the framework may be tainted by subjective value judgements. Given the potential subjectiveness of the court’s judgement, this article will argue that using the best interests test as an intervention threshold is not aligned with the Children Act 1989 and the rights enjoyed by children and their parents. Instead, this article suggests that the court should adopt a ‘significant harm’ threshold to establish its jurisdiction. A distinction is drawn between the intervention threshold of the court and the approach adopted when the court is entitled to intervene. Once the court has established its jurisdiction, it will still be entitled to adopt the best interests test to determine the suitable medical treatment for the child. Lastly, as the significant harm test has been criticised for its indeterminacy and cruelty, these arguments will be considered and rebutted in this article.
The Current Approach
When parents refuse to consent to treatments or withdrawal of treatment that doctors view to be in the child’s best interests, the NHS Trust can apply to the court for a declaration that the treatment, or withdrawal thereof, is lawful. The court may then authorise the treatment through its inherent jurisdiction6 or through a specific issue order under Section 8 of the Children Act 1989,7 but the court must be satisfied that making an order for the child would be better than making no order at all.8 Section 1 of the Children Act 1989 provides that in any question affecting a child’s upbringing, their welfare must be the court’s ‘paramount consideration’,9 which includes the child’s ascertainable wishes, their physical and emotional needs, and the harm that they have suffered.10
Case law has demonstrated that the prospect of recovery is one of the key concerns when evaluating best interests. In Evans, the court considered the risks of taking the child to Italy and held that it was not worth further endangering the child, since there was no prospect of recovery as Alfie was vulnerable to infection.11 According to the judgement in Re Knight (A Child), treatment without the prospect of recovery would be burdensome to the child even though he could not experience pain.12 It should be highlighted that this finding of ‘no prospect of recovery’ does not mean that there is no contrary medical opinion since the court bases the futility analysis on a comprehensive review of all medical evidence and may accord less weight to some of them. For example, in the appeal of Dance, one medical expert referred to the 1% chance of recovery, but the court upheld the finding that there was no prospect of recovery owing to insufficient evidence to challenge the original decision.13 Whilst the court is entitled to determine questions of facts, the family may reasonably disagree with the court that such ‘futile’ treatments are not worth trying.
The concept of futility, or ‘no prospect of recovery’, has been invoked in Yates14 and Evans15 to deny the parents the opportunity to take their children abroad for medical treatment. However, it was not until Dance that futility was invoked to circumvent, if not override, the child’s wishes.16 In Dance, it was recognised that Archie ‘believed in God’, ‘wouldn’t want to leave [his] Mum’, and would ‘try to get out of bed’ if he was on a life support machine.17 Hence, Archie would object to the withdrawal of the treatment based on his religious beliefs and prior wishes. However, given the lack of the prospect of recovery, the court contended that Archie could not have foreseen that his beliefs would be relied on in this situation.18 In doing so, the court interpreted Archie’s words to be based on a premise that there was a prospect of recovery but did not explain convincingly why. The court seemed to be so overly persuaded by the finding of futility that it overlooked Archie’s strong desire to follow his religious beliefs.19 Had the court given more weight to Archie’s desires, it may have found that Archie would like to stay with his mother even though there was no or only minimal prospect of recovery. The possibility of having different interpretations begs the question of whether there is only one objective understanding of Archie’s best interests.
The Interpretation of Best Interests May Not Be Objective
The best interests test was often described in case law as an objective framework, but the relevant analysis is very likely to involve subjective value judgements. Hence, there can be multiple reasonable interpretations of best interests. In Re E (A Child), the court stressed that the best interests of a child can be ‘objectively and independently assessed’, hence the court has the power to intervene by finding this interest.20 However, in Raqeeb, MacDonald J recognised that ‘moral or religious factors extrinsic to the child’ must be considered, which can ‘mean different things to different people in a multicultural society’.21 Although the court in Raqeeb gave weight to the parents’ analysis of their child’s best interests in light of their parental responsibility, it restricted the principle to the rare cases where pain or the awareness of suffering was absent.22
The subjective nature of the best interests test is also evident when the issue is what treatment should be tried and at what costs. In Yates, the parents wished to take their child to the US to receive an experimental treatment of nucleoside therapy. However, the court was doubtful of its effectiveness as the treatment had not been tested on even a mouse model with Gard’s condition; indeed, the US doctors had not even examined Gard’s situation.23 Although the aforementioned evidence has shown that the risk of the treatment may be greater than the benefit, the court nonetheless made a value judgement when they held that the treatment would be ‘of no benefit’ to Gard and hence refused Gard’s parents’ permission to take their child abroad for treatment.24 This stems from the court’s risk-averse attitude, which is itself a subjective value judgement.
It should be noted that the importance attached to the finding of futility also involves a value judgement, particularly because it relates to the sanctity of life. Whilst the court has emphasised in cases like Wyatt v Portsmouth Hospital NHS that the presumption in favour of preserving life can be rebutted when the treatment is futile,25 some may contend that a life of minimal quality should be maintained since life is inherently valuable from some religious or philosophical perspectives. As argued above, the court in Dance seemed to overlook this principle when they held that Archie’s wish to continue the treatment based on his religious belief was contingent upon the assumption that there was a prospect of recovery. This assumption was never explicitly recognised by Archie but was only presumed by the court. As Joseph Goldstein argued, ‘reasonable and responsible persons can and do disagree about whether the “life” after treatment would be “worth living” or “normal”’.26 In assuming the presumption of recovery prospect, the court is arguably relying on its own value judgements, instead of giving effect to the child’s wishes. In fact, given that Archie was brought up by his parents, they were presumably in a better position to infer Archie’s interests based on his wishes and underlying beliefs. In assuming the importance of the recovery prospect in Dance, the court overlooked the value judgements involved and inappropriately disregarded the parents’ observations of Archie’s wishes.
In conclusion, the best interests analysis is likely to be comprised of subjective elements. It is evident in some of the cases, including Yates27 and Evans,28 that the court overlooks the sanctity of life argument when the child in question lacks a prospect of recovery. Dance is particularly striking since the court presumed that Archie’s desire was based on the assumption that he could recover, which he had never explicitly endorsed. Having discussed the subjectiveness of the court’s decisions, the next section will evaluate whether the court is still justified in interfering with children’s medical decisions so readily.
The Best Interests Test Does Not Justify a Court’s Interference
The idea that a child’s best interests cannot be objectively determined is important in light of Section 1(5) of the Children Act, which provides that the court shall not make any order ‘unless it considers that doing so would be better for the child than making no order at all’.29 If there are legitimate differences in what the best interests of the child would truly mean, the court should be mindful about whether making the order would be ‘better for the child’.30 Wilkinson and Nair highlighted that parents tend to make suboptimal decisions about their children, such as their schooling and food, but this position assumes there is an ‘optimal’ decision, which may not exist at all.31 In the context of Dance, it is unclear whether keeping Archie alive would be sub-optimal for him given his religious beliefs and desire to stay with his Mum. Given the parents’ insistence and Archie’s care for his mother’s welfare, it could be argued that allowing him to stay and comfort his mother would be what Archie wants and hence in his best interests, despite the minimal quality of life. Although the ‘no order’ principle is well established in Section 1(5), the court failed to consider, at least in Dance, that authorising the withdrawal of a futile treatment may not always be better for the child, especially when the child has expressed his wish to stay with the family.
Secondly, given that there is a reasonable difference between approaches adopted by different people, it is also unclear if the court can always rely on policy considerations to override the parents’ wishes. In Raqeeb,32 MacDonald J recognised that Raqeeb’s right to free movement and to receive services under Article 56 TFEU33 could be restricted when there was a proportionate public policy justification, namely protecting the child’s best interests.34 He also highlighted a margin of appreciation given to the UK courts.35 However, if the analysis of best interests is inevitably permeated by subjective value judgements, one should re-evaluate whether this analysis could always provide a sufficient policy justification under Article 56 TFEU for the courts to override parents’ decisions. It is not suggested that an assessment with subjective elements can never be used, but that arguably the courts should be less eager to intervene and more hesitant to present themselves as the objective assessor. Given the variety of versions and interpretations of best interests, it is unclear whether such a high standard can be best assessed by the court and safely relied on as the justification to restrict one’s right to free movement. Although this right is not concerned in Dance, it is important when parents seek to bring their children overseas for medical treatments.
Lastly, Auckland and Goold also highlight that such intervention may violate the Article 8 right to ‘private and family life’.36 In MAK and RK v United Kingdom, the European Court of Human Rights held that a medical intervention on a child without parental consent is a violation of Article 8 ECHR of private and family life.37 The court also added that this right can only be restricted in accordance with the law and necessary in a democratic society in pursuit of a legitimate aim such as the protection of health and morals or the protection of the rights and freedoms of others.38 In Dance, the parents argued that a patient had the right to choose the manner of their death under Article 8 ECHR, but the court distinguished the case by holding that this only related to capacitous adults.39 Had the parents argued that an intervention in the medical decision itself was a violation of Article 8 (which is protected by the Human Rights Act 1998),40 it would be unclear whether the intervention was ‘necessary in a democratic society’ and proportional to the aim pursued. This is especially true when a democratic society requires tolerance and broadmindedness, even when one does not agree with others’ decisions.41 As criticised by Helen Reece, the paramountcy principle of children’s welfare has been used to promote the parenting approach that conforms to social norms.42 In adhering strictly to the best interests analysis in this case, the court is also arguably too narrow-minded to refuse all other parenting approaches which deviate from the societal norm. It is unlikely that this dogmatism is inherently necessary in a democratic society.
Time to Turn to a Significant Harm Threshold
In 2020, Baroness Ilora Finlay introduced the Access to Palliative Care and Treatment of Children Bill to the House of Lords, which provided for a presumption that any medical treatment put forward by parents for the child is in the child’s best interests.43 The presumption will be rebutted if ‘the proposed treatment poses a disproportionate risk of significant harm’ to the child.44 The essence of the Bill echoes the proposition made by Douglas Diekema that the threshold of intervention should be replaced by the question of whether the decisions made by the parents ‘significantly increase the likelihood of serious harm as compared to other options’.45 Before discussing the merits of these proposals, a distinction must be drawn between the intervention threshold of the court and the approach adopted when the court is entitled to intervene. If these two suggestions are adopted, the court will be able to establish its jurisdiction when the proposed treatment poses significant harm to the child. This article suggests that once the court has established its jurisdiction, it will still be entitled to adopt the best interests test to determine the suitable medical treatment for the child.
The significant harm threshold may strike a better balance between state intervention and parental freedom. This threshold still entitles the court to intervene when the parents opt for extreme treatment which subjects their child to significant harm or when they refuse a treatment which increases the child’s chances of recovery. For example, in Re JM (A Child), the parents wished to treat their child’s rare and aggressive cancer with Chinese medicine instead of surgery.46 In An NHS Trust v SR, the mother was concerned that radiotherapy may leave her son infertile and hence refused it.47 In refusing treatment which increases their children’s chances of recovery, the parents have arguably posed significant harm to their children, so the court can still intervene in these extreme situations.
It is acknowledged that some families may be more convinced by the efficacy of Chinese medicine or the devastation of being infertile, and their decisions may be overridden even though the significant harm threshold is used. However, the courts in those cases have relied on medical evidence to illustrate the drastic consequences of not receiving Western treatment.48 Arguably, few people would agree that the parents should still be given absolute discretion in such situations. Instead, as this article proposes, the court should only consider the child’s or the parents’ unique beliefs to determine which decision would most align with the child’s best interests after establishing jurisdiction based on significant harm. In these situations, the court’s interference seems more justified since the children will otherwise be subject to serious harm. The court is more likely to be satisfied that the child’s situation will improve when it issues an order to prevent significant harm, which also has a more prevalent public policy consensus within a democratic society. Given that the children’s welfare is in serious jeopardy, the court will be more justified in asserting its jurisdiction and intervening, as compared to a situation like Dance, where the court is simply satisfied that the course of treatment is not in the child’s best interests.
Essentially, the change will reserve a ‘zone of discretion’ for parents to choose the course of treatment. The discretion in this context means an area free from state intervention, which was figuratively described by Dworkin as the hole in a doughnut surrounded by a belt of restriction.49 Had this test been adopted in Dance, the court would not have intervened until the life-sustaining treatment caused significant harm to Archie. In Re King (A Child), the court recognised that the course of treatment proposed by the parents was entirely reasonable, so ‘it is the parents who bear the heavy responsibility of making the decision’.50 Although, in Yates, McFarlane LJ firmly rejected that Re King indicated a shift to the significant harm threshold,51 this significant harm approach should be preferred because only then can the court’s interference be justified given the Children Act 1989 and the rights enjoyed by the children and their parents. Within that zone of discretion where the court intervenes under the current approach, there will lie some solutions or treatments which the parents prefer, owing to their risk-taking attitude or religious beliefs. Unless the treatment poses significant harm to the child, there is no obvious reason why the court should be authorised to invade this zone when they disagree with the parents. Hence, the bar of significant harm should be adopted as a watershed, leaving the rest to the parents’ discretion. The common counterarguments against this threshold are often concerned with the indeterminacy and cruelty of the language of ‘significant harm’, which will be addressed in the next section.
Considering the Counterarguments
Whilst the best interests analysis is likely to involve subjective value judgements, it is equally true that the concept of ‘significant harm’ can be subjective and indeterminate. However, that does not mean that the significant harm test should not be implemented. The subjective element of the best interests analysis was highlighted earlier not to demonstrate that the court must remain entirely objective and refrain from any subjective determinations; instead, it revealed that the court can hardly remain objective in these controversial life-or-death issues that involve the sanctity of life, a cost-benefit analysis, and an attitude towards taking risks. Realising the court’s inability in determining the ‘objective’ best interests of the child, the court should then refrain from exerting the jurisdiction too readily, simply because they perceive that the child’s best interests are not protected. In short, given the potential subjectiveness of the court, it should be more hesitant about whether it has the authority to decide what is best for the child. As this article has argued, a significant harm threshold would be a more modest approach to intervention.
Additionally, although the interpretation of ‘significant’ harm cannot be entirely objective, the evidence below will show that there is a wider consensus regarding its definition. For example, Diekema argued that it may be easier to determine whether something is harmful (a negative outcome) than whether something is the best decision (a positive outcome).52 The fact that a significant harm threshold has been used in care proceedings pursuant to Section 31 of the Children Act 1989 is also proof that a significant harm threshold may be more intelligible to doctors, parents, and judges.53 Due to this wider consensus surrounding the definition of significant harm, the court may be more justified in its intervention. Whilst the exact limit of intervention can be refined by case law, lifting the intervention bar to ‘significant harm’ is a good starting point for reflecting if the court should intervene in these controversial issues that lack common consensus.
Another common objection is that a ‘significant harm’ test would be too harsh for the parents. Rachel Taylor has described it as ‘unnecessarily cruel and combative’ when the parents are required to defend themselves against a test of harm.54 However, when the parents in Yates, Evans, and Dance were denied the opportunity to take their children abroad, they were sorrowful since they could not keep their children alive. Their agony stems from the very result that their children would soon die, and they arguably would not feel significantly different if the language used is ‘causing significant harm’ instead of ‘not being in their best interests’. As Auckland and Goold argue, we should not lightly dismiss the impact of our words, but we should consider whether the potential they have to harm one set of parents is worth the trade-off if it increases the autonomy of other parents.55 If the court refrains from intervening in decisions within the zone of discretion, it will already offer more leniency than the current approach. Hence, although the language of ‘significant harm’ may potentially hurt a few parents, the court should not be dissuaded from switching to this threshold.
In contrast, switching the threshold to significant harm can arguably further facilitate dialogue and consensus-building between doctors and parents. It is appreciated that consensus-building would be the predominant approach before coming to court if the ‘significant harm’ threshold is used. As McFarlane LJ already noted in Yates, ‘very many cases involving children with these tragic conditions never come to court because a way forward is agreed as a result of mutual respect between the family members and the hospital’.56 Using the language of ‘significant harm’ can reiterate the court’s reluctance to intervene unless necessary, thus creating maximum space for doctors to liaise with parents to determine the best course of treatment agreed upon by both. Jo Delahunty argues that the need to prove ‘significant harm’ may cause the position of both parties to harden during their dialogue.57 However, the contrary position seems more convincing. Realising the high bar of ‘significant harm’, doctors will arguably work out and liaise with the parents. Hence, raising the bar to ‘significant harm’ can be less cruel for parents and more conducive to the doctor-patient relationship.
Conclusion
The line of children’s treatment cases, namely Yates, Evans, and Dance, has indicated that under the current law, the court is ready to intervene when it is satisfied that the decision is not in the child’s best interests. However, it is unclear if the court should be entitled to intervene to such an extent, given the implicit value judgements underpinning the best interests test. The judgement of Dance has strikingly interpreted the child’s wishes to be based on an assumption of recovery prospect, whereas this seems to be the assumption of the court itself. When even the child’s wishes can be circumvented or re-interpreted, we should reconsider whether the court is entitled to make these judgements with the best interests threshold. Based on the analysis above, the best interests analysis does not fit well with the ‘no order’ principle in the Children Act 1989 and the rights enjoyed by children and their parents.
A shift to the significant harm threshold would strike a better balance between state intervention and parental freedom. The best interests test should only be used after the court has established its jurisdiction by using the significant harm threshold. Although there has been apprehension about the indeterminacy and cruelty of the language of ‘significant harm’, this article has argued that these concerns are overstated. Therefore, the court should revisit its position and adopt a ‘significant harm’ threshold before intervening in medical decisions for children.
[1] Dance and Others v Barts Health NHS Trust [2022] EWCA Civ 1055.
[2] Barts Health NHS Trust and another v Battersbee and others [2022] EWFC 80 [6].
[3] Great Ormond Street Hospital for Children NHS Foundation Trust v Yates [2017] EWCA Civ 410.
[4] Alder Hey Children’s NHS Foundation Trust v Evans [2018] EWHC 308 (Fam).
[5] Dance (n 1) [49].
[6] Practice Direction 12D 1.1.
[7] Children Act 1989 (Children Act) s 8(1).
[8] Children Act s 1(5).
[9] Children Act s 1(1).
[10] Children Act s 1(3).
[11] Evans (n 4) [63].
[12] Re Knight (A Child) Parfitt v An NHS Trust and Another [2021] EWCA Civ 362 [60].
[13] Dance (n 1) [27].
[14] Yates (n 3) [21].
[15] Evans (n 4) [24].
[16] Dance (n 1).
[17] ibid [6]-[7].
[18] ibid [49].
[19] ibid [6]-[7].
[20] Re E (A Child) [2018] EWCA Civ 550 [127].
[21] Raqeeb v Barts Health NHS Trust [2019] EWHC 2531 (Admin) [191].
[22] ibid [191].
[23] Yates (n 3) [35].
[24] ibid [115].
[25] Wyatt v Portsmouth Hospital NHS [2005] EWCA Civ 1181 [87].
[26] Joseph Goldstein, ‘Medical Care for the Child at Risk on State Supervision of Parental Autonomy’ (1977) 86 Yale Law Journal 645, 654.
[27] Yates (n 3).
[28] Evans (n 4).
[29] Children Act s 1(5).
[30] ibid.
[31] Dominic Wilkinson and Tara Nair, ‘Harm Isn’t All You Need: Parental Discretion and Medical Decisions for a Child’ (2016) 42(2) Journal of Medical Ethics 116-118.
[32] Raqeeb (n 21).
[33] Consolidated Version of the Treaty on the Functioning of the European Union [2008] OJ C115/13 Art 56.
[34] Raqeeb (n 21) [48].
[35] ibid [110].
[36] Cressida Auckland and Imogen Goold, ‘Re-Evaluating “Best Interests” in the Wake of Raqeeb v Barts NHS Foundation Trust & Anors’ (2020) 83(6) Modern Law Review 1, 12.
[37] MAK and RK v The United Kingdom App no 45901/05 and 40146/06 (ECtHR, 23 March 2010).
[38] European Convention on Human Rights 1950 art 8(2).
[39] Dance (n 1) [26].
[40] Human Rights Act 1998 s 6(1).
[41] Dudgeon v The United Kingdom App no 7525/76 (ECtHR, 22 October 2981) [49]-[51].
[42] Helen Reece, ‘The Paramountcy Principle: Consensus or Construct?’ (1996) 49(1) Current Legal Problems 267, 293.
[43] Access to Palliative Care and Treatment of Children HL Bill (2019-21) cl 2(2).
[44] ibid.
[45] Douglas S. Diekema, ‘Parental Refusals of Medical Treatment: The Harm Principle as Threshold for State Intervention’ (2004) 25(4) Theoretical Medicine and Bioethics 243, 252.
[46] Re JM (A Child) [2015] EWHC 2832 (Fam).
[47] An NHS Trust v SR [2012] EWHC 3842 (Fam).
[48] Re JM (n 46) [15]; SR (n 47) [25].
[49] Ronald Dworkin, Taking Rights Seriously (Bloomsbury Publishing Plc 2013) 48.
[50] Re King (A Child) [2014] EWHC 2964 (Fam)[34].
[51] Yates (n 3) [104]-[105].
[52] Diekema (n 45).
[53] Children Act s 31.
[54] Rachel Taylor, ‘Parental Decisions and Court Jurisdiction: Best Interests or Significant Harm?’ (2020) 32(2) Child and Family Law Quarterly 141, 151.
[55] Cressida Auckland and Imogen Goold, ‘Parental Rights, Best Interests and Significant Harms: Who Should Have the Final Say Over a Child’s Medical Care?’ (2019) 78(2) Cambridge Law Journal 287, 318.
[56] Yates (n 3) [112].
[57] Jo Delahunty, ‘The Child and Medical Treatment: The Chance to Live, or to Die with Dignity’ (2019) 29 Family Law 1022, 1030.
Ovi Cheung
LLB LSE ’24 and Notes Editor of the LSE Law Review 2022-23