Adolescent autonomy to make a ‘catastrophically bad’ decision informed by religious beliefs: To respect or protect?

Abstract

This article examines the Australian case law in which medical treatment has been authorised against the wishes of a Gillick competent adolescent, in circumstances where the adolescent’s decision to refuse treatment has been informed by their religious beliefs. We analyse three factors that have shaped the Australian jurisprudence – medical evidence regarding the adolescent’s prognosis and risk of death; the role of religion in adolescents’ medical treatment decision-making; and the adolescent’s own views and wishes – to expose arguably uncomfortable, value-laden judicial attempts to achieve a ‘nice balance’ between the sanctity of human life and individual autonomy. We argue for greater judicial transparency about when, if ever, an adolescent’s Gillick competence will be respected in refusal of medical treatment cases shaped by the adolescent’s faith.

When should a court override a competent adolescent’s ¹ decision to refuse medical treatment that could save or prolong their life, in circumstances where that decision has been informed by the adolescent’s religious beliefs? Should the court respect the adolescent’s current autonomy, even where their decision is considered to be ‘catastrophically bad’? ² Or should the court seek to protect the adolescent’s future autonomy and ‘right to an open future’? ³ These vexed questions have arisen in only three published Australian judgments to date. In each case, the court has exercised its parens patriae jurisdiction to authorise life-saving medical treatment against a Gillick competent adolescent’s wishes. ⁴ In this article, we analyse these judgments, which span a period of two decades, to expose arguably uncomfortable, value-laden judicial attempts to determine the best interests of the adolescent, in light of two ‘fundamental principles demanding state protection, which must be reconciled in their application to the particular person’: ⁵ individual autonomy and sanctity of life.

We first present an overview of the parens patriae jurisdiction and Gillick competence in the medical treatment context. We then discuss three factors that have shaped the Australian jurisprudence: medical evidence regarding prognosis and risk of death; the role of religion in adolescents’ medical treatment decision-making; and the adolescent’s views and wishes. Our analysis reveals that, while the earlier two cases have focused primarily on the medical evidence and the adolescent’s religious beliefs, the most recent decision expands the evaluation to recognise the impacts on adolescent rights and bodily autonomy. We highlight that the ‘nice balance’ ⁶ the court has sought to achieve in each case, by ‘appropriately’ weighing these various considerations, prompts the need for greater judicial transparency about when, if ever, an adolescent’s Gillick competence will be respected above the state interest in the sanctity of human life.

Parens patriae jurisdiction, Gillick competence and adolescent autonomy

The principle of Gillick competence provides that a child or young person has the capacity to make an informed decision about their medical treatment where they have achieved ‘a sufficient understanding and intelligence to enable [them] to understand fully what is proposed’. ⁷ In practice, a finding of Gillick competence may allow a clinician to rely on the adolescent’s consent to proceed with recommended medical treatment. While Lord Scarman’s oft-cited formulation of the principle, just noted, was expressly approved by the High Court of Australia, ⁸ deeper engagement with the reasoning in Gillick reveals different judicial approaches to the threshold for adolescent decision-making. Notably, as other commentators have observed, ⁹ Lord Scarman agreed with the principle that a child must reach ‘an age of sufficient discretion to enable him or her to exercise a wise choice in his or her own interests’. ¹⁰ The Gillick principle itself therefore arguably imposes a limitation upon adolescent decision-making: namely, that the choice be ‘wise’ or rational. This limitation has manifested in the context of an adolescent refusing medical treatment on religious grounds, with courts requiring that the adolescent’s ‘expressed choices are the product of an independently formed, carefully considered and long-held adherence to the tenets of the … faith’. ¹¹

The parens patriae jurisdiction is ‘essentially protective’: ¹² it serves to protect ‘those (including, but not limited to, minors) who are unable to protect themselves’. ¹³ In the medical treatment context, the court’s role ‘is to exercise an independent and objective judgment and balance the advantage or disadvantage of the medical step under consideration’. ¹⁴ While other interests are ‘not completely disregarded’, the adolescent’s ‘best interests’ or ‘welfare’ are the paramount consideration. ¹⁵ The practical result is that the Gillick competence principle ‘does not diminish the scope of the parens patriae jurisdiction’. ¹⁶

In the next section, we outline the three published Australian judgments in which the courts have overridden a Gillick competent adolescent’s refusal of medical treatment.

When best interests trump Gillick: An overview of the case law

The case of X v The Sydney Children’s Hospitals Network (‘X’) ¹⁷ concerned a 17-year-old, referred to in the judgments as ‘X’, who had been diagnosed with Hodgkin’s disease, an aggressive form of cancer. In Minister for Health v AS (‘AS’), ¹⁸ 15-year-old ‘L’ had been diagnosed with atypical Burkitt’s lymphoma. In each case, the cancer was treatable with chemotherapy; the adolescent’s treating medical team sought to administer a blood transfusion to remedy the side effects of chemotherapy; the adolescent and their parents were Jehovah’s Witnesses; and they withheld consent to treatment on the basis of their faith.

In the X proceedings, although the Supreme Court of New South Wales (Gzell J) ¹⁹ had accepted that X was a ‘mature minor’, the Court authorised the Sydney Children’s Hospital to administer the blood transfusion. The Court of Appeal (Beazley P, Basten JA and Tobias AJA) granted leave to X and his parents to appeal the decision, but ultimately dismissed the appeal. In AS, ²⁰ L was also assessed as Gillick competent. However, the Supreme Court of Western Australia (Pullin J) overrode the wishes of L and his parents and authorised medical staff at Princess Margaret Hospital to administer the blood transfusion.

In the most recent case of H v AC (‘AC’), ²¹ the applicant hospital sought a declaration that 17-year-old ‘AC’ had capacity to refuse ongoing treatment for Ewing sarcoma, a type of bone cancer. Only if AC was found to lack capacity, was the hospital seeking a declaration to lawfully carry out the treatment without the consent of AC or her parents. Despite declaring AC to be Gillick competent, the Supreme Court of New South Wales (Meek J) nonetheless authorised the treatment.

Below, we examine three factors that have shaped the Australian jurisprudence in this area. Our analysis highlights a judicial resolve to ‘save’ adolescents from themselves and their religious beliefs, underpinned by a presumption that they ‘do not choose well or wisely’. ²²

Risk of death and the weight of medical evidence

Medical evidence and opinions of treating practitioners indicating that a Gillick competent adolescent will not survive without the proposed treatment – commonly presented as a percentage prognosis or risk of death – has led Australian courts to conclude that the treatment is in the adolescent’s best interests. The three judgments reinforce the view that even when an adolescent is found to have reached the Gillick threshold, the court will be reluctant to decide contrary to the advice of the medical team. ²³

It is important to highlight the two distinct issues at stake in the judicial deference to medical evidence and advice in refusal of medical treatment cases. The first issue concerns the adolescent’s understanding and assessment of information about the proposed treatment, including the evidence regarding prognosis if treatment is or is not provided. The court may find that an adolescent is not Gillick competent on the basis that they are unable to understand and/or appropriately weigh the risks and benefits of refusing medical treatment, or they ‘lack[] insight into death and end of life issues’. ²⁴ However, it has been argued that such an approach unjustifiably holds adolescents to a higher threshold of capacity and understanding than adults. ²⁵

The second issue, which is the focus of our article, is where the court assesses the adolescent to be Gillick competent, but nonetheless concludes that the adolescent cannot refuse medical treatment that is in their best interests, if that refusal will result in their death. ²⁶ The distinction between these two issues is significant, for the three Australian judgments analysed in this article openly acknowledge that the need to protect the adolescent from death, by overriding the adolescent’s refusal to consent, accords with both the medical evidence and the adolescent’s best interests. ²⁷

Both X and AS concerned an adolescent’s refusal to accept blood (or blood product) transfusions. In AS, L’s treating medical practitioners described the risk of death if a blood transfusion was not performed as ‘considerable’ and ‘likely’, ²⁸ while the prospect of L’s long-term, disease-free survival if he did undergo the blood transfusion was assessed at greater than 70 per cent. ²⁹ In X, the medical evidence was that X had a 40 to 50 per cent chance of being cured if he received chemotherapy treatment. ³⁰ Without the blood transfusion, X’s risk of dying was assessed at approximately 80 per cent. ³¹

In both X and AS, legislation existed to authorise a blood transfusion upon an adolescent without consent in circumstances where, without such treatment, they would likely die or suffer serious damage to their health. ³² In AS, Pullin J considered that s 21 of the Human Tissue and Transplant Act 1982 (WA) would have covered the situation before the Court, such that an application was unnecessary. However, his Honour accepted that the hospital was seeking certainty. ³³ In X, it was argued that s 174 of the Children and Young Persons (Care and Protection) Act 1998 (NSW) provided the only circumstance in which a blood transfusion could be administered without consent. The Court of Appeal disagreed. It held that this provision was silent on the right of an adolescent to refuse life-saving treatment and so it did not, without clear words, abrogate the exercise of the parens patriae jurisdiction. ³⁴

Unlike the earlier two cases, AC did not involve refusal of a blood transfusion. AC had undertaken four cycles of chemotherapy. A subsequent MRI scan revealed a spectacular reduction in the size of her tumour and no visible evidence of cancer. ³⁵ In line with her Christian faith, AC believed that her prayers had been answered and that miraculously, she had been cured. However, the medical evidence was that, while not visible, the cancer would return. AC’s treating practitioners recommended that she undergo a further 10–13 rounds of chemotherapy, in addition to radiotherapy. ³⁶ AC refused this further treatment, preferring to place her trust in God. AC indicated she would consider re-engaging with medical treatment if the cancer returned. ³⁷

If AC undertook the recommended treatment, the medical opinion was that she had a 40 to 45 per cent chance of permanent cure. Without treatment, there was ‘a near 100% chance’ that the cancer would recur. ³⁸ Although limited data existed about what would happen if AC did not begin treatment until such recurrence, the medical opinion was that the cancer would be difficult to cure and would lead to AC’s death. ³⁹ Notably, AC’s oncologist described her decision as ‘one of the most catastrophically bad decisions I’ve ever seen’. ⁴⁰

The medical evidence in X and AS was largely confined to an assessment of the adolescent’s physical health and prospects of survival if a blood transfusion was or was not administered. In AC, Meek J went a step further, meaningfully engaging with the benefits and burdens of the proposed treatment for AC. The evidence attested to AC having tolerated the ‘horrendous ordeal’ of chemotherapy ‘quite well’ and ‘better than many’. AC’s treating practitioner described her as ‘quite resilient … and not overtly significantly distressed by the treatment’. ⁴¹

However, this medical evidence diverged from AC’s own shared subjective experience. AC described various adverse physical, psychological and social impacts of the treatment she had endured to date, including that she felt ‘pretty exhausted both physically and mentally’. ⁴² AC also spoke to the loss of autonomy she would likely experience if she was forced to undergo further treatment. ⁴³ Yet Meek J did not interrogate this divergence between the medical evidence and AC’s evidence.

Judicial deference to evidence that is within the scope of medical expertise – such as the likelihood that treatment will improve a patient’s health, or alleviate harmful side-effects – is unsurprising and uncontroversial. ⁴⁴ However, where medical opinion exceeds the bounds of that expertise, we argue that the court’s unwavering deference must be questioned. ⁴⁵ This is because the determination of whether a Gillick competent adolescent should be required to undergo medical treatment against their will is not simply a matter of scientific fact, but requires an assessment of best interests, which is an inherently subjective, ‘value-laden’ process. ⁴⁶ As Carmody J explained in Dylan v Dylan: ⁴⁷

[B]est interests are really values not facts. They are not susceptible to scientific demonstration or conclusive proof. The same body of evidence may produce opposite but nonetheless reasonable conclusions from different judges. There is not always only one right answer. ⁴⁸

While the three judgments reinforce that a ‘doctors know best’ judicial approach persists, Gillick competent adolescents may have ‘as much, if not more, expertise than the doctors’ in weighing the ‘benefits and burdens’ of treatment. ⁴⁹ However, the influence of religion has been treated by the courts as impairing adolescents’ ability to ‘appropriately weigh’ the considerations for and against treatment. ⁵⁰

Religious beliefs: Court as saviour?

The three judgments bring into sharp focus the complex interplay between the ‘inestimable’ sanctity of life, individual autonomy, and the role of religion in adolescents’ medical treatment decision-making, prompting ‘some of the most palpable forensic debates and challenging legal decisions’. ⁵¹ The judgments also demonstrate that religious beliefs, even when accepted by the courts as conventional, ingrained, and unlikely to change in adulthood, will be construed as motivating the ‘catastrophically bad’ ⁵² decision to refuse medical treatment.

Both L ⁵³ and X ⁵⁴ made their decision to refuse a blood transfusion in line with their faith as Jehovah’s Witnesses. For L, his belief of not accepting blood products ‘had been with him all his life and [was] an integral part of him’. ⁵⁵ In X, Gzell J at first instance considered that X’s ‘life had been cocooned in … [his] faith’, with X’s decision supported by family and friends who were ‘almost exclusively of the Jehovah’s Witness faith’. ⁵⁶ Gzell J highlighted that X’s parent had taught him ‘morals and principles at an early age and introduced him to the scriptures’. ⁵⁷ At the age of 13, X asked elders of his church if he could preach. ⁵⁸

On appeal, Basten JA acknowledged that decisions concerning X’s best interests required more than an evaluation of the medical evidence. Religious motivation was a factor ‘likely to be one which the Court will accord respect and weight, other things being equal’. ⁵⁹ However, his Honour did not extrapolate on what these ‘other things’ were. Basten JA further noted that courts may be reluctant to intervene where the treatment refusal arises after an assessment of the advantages and disadvantages of treatment, and ‘where the choice is one as to which reasonable minds might differ’. ⁶⁰ It might be inferred from this reference to reasonableness that X’s religiously-motivated choice was not considered reasonable. However, absent further engagement with the values, factors and circumstances shaping the determination of X’s best interests, Basten JA’s approach reinforces the imprecision and discretion that underlie the application of the ‘best interests’ principle. ⁶¹ It also illustrates an expansive notion of ‘best interests’ that embraces an adolescent’s spiritual, psychological and social interests, as well as their medical interests. ⁶²

In AC, the evidence confirmed that AC had conferred with her family, friends, school and church. ⁶³ She had weighed the pros and cons of further treatment and refusal in collaboration with her treating team using a ‘decision matrix’ or ‘decision table’. ⁶⁴ Through this process, AC had regard to ‘a number of different considerations, not merely her religious beliefs’. ⁶⁵ However, the evidence of AC’s treating psychologist was that AC could not ‘appropriately weigh the risks and benefits’ due to her staunch belief that ‘God has been active in her plight and that her cancer has been cured by way of a miracle’. ⁶⁶

Recounting passages of the Bible, Meek J analysed the nature of ‘miracles’ and the purpose of prayer. ⁶⁷ His Honour referred three times to evidence that AC regarded Jesus as her ‘saviour’. ⁶⁸ Although Meek J confirmed AC’s beliefs as ‘entirely conventional within the Christian faith’, ⁶⁹ the orders made went beyond those sought by any party – a fact which his Honour expressly acknowledged. ⁷⁰ His Honour became AC’s ‘saviour’. The decision to authorise treatment arguably undermined the beneficence of AC’s God and supplanted it with the beneficence of the Court.

Adolescent views and the rhetoric of rights

Judicial consideration of the views and wishes of Gillick competent adolescents exposes a palpable tension between respecting the evolving capacities and current autonomy of adolescents to make significant life decisions, and the desire to protect adolescents’ ‘right to an open future’ ⁷¹ and future decision-making capacity.

In AS, L had communicated his wishes via a signed form in unambiguous terms: ‘[L] expressly withholds his consent to and forbids under any circumstances the administration of … blood products’. ⁷² L’s resolve was such that, even if ordered by the court, L would ‘still hold his belief that blood products should not be administered’. ⁷³ Pullin J was satisfied that L was of sufficient ‘maturity and intelligence’ to understand the ‘nature and implications of his decision’. ⁷⁴ While his Honour had ‘carefully considered’ L’s wishes, Pullin J was quick to dismiss them as being ‘governed by his religious belief’, which led L ‘to reject the expert medical advice … available to him’. ⁷⁵

X was a mere four months from legal majority when the Court of Appeal’s decision was delivered. At first instance, Gzell J described X as ‘highly intelligent’ and ‘clearly a mature minor’ who had researched possible non-blood treatment options online and expressed a clear view to ‘be responsible for what went into his own body’. ⁷⁶ His Honour described X’s views as ‘strong and genuine’. ⁷⁷ It is clear that Gzell J considered X to be Gillick competent, as noted by Basten JA on appeal. ⁷⁸ Despite these statements, Gzell J concluded that ‘this is not a case of Gillick competency. X is still a child although a mature child of high intelligence’. ⁷⁹ His Honour’s conclusion may be understood as asserting the Court’s power to override a Gillick competent minor’s wishes.

On appeal, Basten JA described the ‘interest of the state in preserving life’ as ‘at its highest with respect to children and young persons’, given they are ‘inherently vulnerable, in varying degrees’. ⁸⁰ X’s vulnerability was framed as deriving from experience as well as age. His Honour was concerned about X’s dependence upon his family and religious community, and his minimal exposure to broader perspectives to enable him to form independent views. ⁸¹ This focus on vulnerability sits uncomfortably with adolescent autonomy and the Gillick competence principle, which respect adolescents’ ‘evolving capacities’ ⁸² for decision-making, and their right to express their views freely in all matters affecting them. ⁸³

The ‘explicit paternalism’ of the parens patriae jurisdiction and its readiness to prioritise certain rights over others raise valid concerns about its ‘potential for overreach’, ⁸⁴ with ‘vulnerability’ used to ‘legitimis[e] adult intervention’ in adolescents’ lives. ⁸⁵ Indeed, the ‘bright line’, ⁸⁶ age-based distinction between minority and adulthood is palpable in refusal of medical treatment cases. However, as the first author has argued elsewhere, ‘[s]omething more than age is used to deprive children of rights-bearer status: … a perceived connection between children’s age and their attainment of a relevant capacity or competence’. ⁸⁷ Cave and Cave consider the existing binary approach towards refusal of medical treatment should be adapted ‘in light of biological, social and psychological evidence that adolescence can extend to the mid-20s’. ⁸⁸ They advocate for the law to further differentiate between groups of adults if they ‘can be shown to be vulnerable and deserving of protection’. ⁸⁹ We maintain that, regardless of where the ‘line’ is drawn, or how ‘bright’ it may be, overriding the decision of a Gillick competent adolescent (or young adult) does not ‘advance the protection of … welfare and autonomy rights’; ⁹⁰ rather, it unjustifiably denies their right to make autonomous choices about their medical treatment.

Also noteworthy are the contrasting judicial approaches to adolescent rights. Pullin J in AS adopted an ‘invisible’ rights approach, characterised by a failure to even identify L’s rights as relevant to the dispute. ⁹¹ By contrast, X and AC referred to the adolescent’s right to refuse medical treatment and their decision-making autonomy. Yet in both cases, the Court’s approach was a ‘rhetorical’ one. ⁹²

In X, X and his parents relied on ‘the right of a mature minor to self-determination and autonomy with respect to medical treatment’ as a factor said to impose a legal limit on the scope of the parens patriae jurisdiction. ⁹³ Basten JA cited various international conventions to confirm that ‘religious beliefs are internationally accepted as an aspect of an individual’s fundamental autonomy with which the state cannot interfere and should not disregard’. ⁹⁴

In AC, Meek J’s assessment of AC’s capacity similarly embraced the language of rights and autonomy. His Honour identified human dignity as ‘a value common to municipal law and to international instruments relating to human rights’, ⁹⁵ and quoted at length from a Supreme Court of Canada decision, which asserted that:

It is not only an option for the court to treat the child’s views as an increasingly determinative factor as [their] maturity increases, it is, by definition, in a child’s best interests to respect and promote [their] autonomy to the extent that [their] maturity dictates. ⁹⁶

The evidence highlighted the ‘loss of autonomy’ that AC would experience if compelled to undergo treatment, ⁹⁷ as well as the ‘centrality’ of AC’s religious beliefs to her right to autonomy. ⁹⁸ The submissions of counsel for AC invoked various provisions of the United Nations Convention on the Rights of the Child ⁹⁹ and proposed that it was ultimately ‘in AC’s best interests to respect and promote her right to autonomy’. ¹⁰⁰

Listing the factors that had informed the decision to order treatment, Meek J noted that ‘sanctity of life’ was ‘an important consideration’ to be ‘appropriately weighed’ against the medical evidence, AC’s religious beliefs, her right to bodily integrity and decision-making autonomy. ¹⁰¹ There was ample evidence of AC’s resilience, ‘maturity and consistency’ in decision-making, ¹⁰² intelligence and reflective approach to her decision, which had been shaped by AC’s ‘conventional’ religious belief in the power of prayer and miracles. ¹⁰³ Yet it is arguably difficult to discern how Meek J’s decision was reached, if not by an unreserved prioritisation of the ‘sanctity of life’ consideration, underpinned by deference to the medical evidence.

Conclusion: Will Gillick ever be enough?

It has been beyond the scope of this article to explore the differences between consent to, and refusal of, medical treatment, ¹⁰⁴ and whether respect for the rights of Gillick competent adolescents demands that their choice to refuse treatment must always be respected. ¹⁰⁵ Rather, we have asked: in what circumstances will the decision of a Gillick competent adolescent to refuse life-saving medical treatment – which has been informed by their religious beliefs – be determinative? Our examination of the Australian case law suggests that currently, the answer is: never.

Auckland’s analysis of UK case law in this area has reached the same conclusion, namely, that ‘judges are reluctant to allow adolescents to “martyr themselves” for their religious beliefs’. ¹⁰⁶ An articulated judicial concern in many of the UK cases was that the adolescent’s decision ‘was not truly their own, or at least not something that, with the benefit of adult reflection, they would necessarily have continued to seek for themselves’. ¹⁰⁷ Auckland has argued that the courts’ ‘reticence to confront’ these ‘authenticity’ concerns has resulted in ambiguity about the extent to which a religiously-motivated foundation for treatment refusal has in fact been the basis for judicial rejection of Gillick competent adolescents’ decisions. ¹⁰⁸

It is equally unclear from the limited Australian case law to date when, if ever, a Gillick competent adolescent’s refusal of life-saving medical treatment will be respected as ‘truly authentic’, ¹⁰⁹ where that decision is motivated by entrenched religious beliefs. The courts’ approach of prioritising medical evidence is not significantly different to that adopted by courts when determining the ‘best interests’ of adults who lack capacity. ¹¹⁰ However, we argue that the views, beliefs and motivations of adolescents must be given significant weight, by virtue of the finding of Gillick competence. If the courts’ continued approach is to unreservedly order treatment in such cases, then we argue that judicial officers must be transparent that the ‘importance of life’ consideration will always prevail in the best interests assessment. ¹¹¹ Express judicial acknowledgement that adolescent autonomy will be limited in these circumstances might offer greater clarity and certainty for adolescents and their parents. ¹¹² It would also support clinicians to feel ‘ethically comfortable in the framework within which they are required to work’. ¹¹³

The Australian jurisprudence suggests that courts are likely to continue to adopt ‘a sceptical approach to autonomy’ in treatment refusal cases involving Gillick competent adolescents. ¹¹⁴ In AS, X and AC, the Court deferred to medical evidence and opinion to prioritise the sanctity of life over the adolescent’s wishes. In each case, the Court intervened to shield the adolescent from a religiously motivated, ‘catastrophically bad’ ¹¹⁵ – and potentially terminal – decision, such that protecting ‘future choice’ ¹¹⁶ prevailed over respecting current autonomy. According to Fortin, this judicial ‘exercise in paternalism’ may be ‘extremely tempting’, as ‘court[s] may find it impossible to conclude that it is in [an adolescent’s] best interests to be allowed to die’. ¹¹⁷ At first blush, references to adolescent ‘rights’ and ‘autonomy’ in the judgments analysed may be interpreted as an evolution in judicial engagement with, and acceptance of, rights-based arguments. However, by overriding the choice of a Gillick competent adolescent – and in AC, by making an order beyond what any party had sought – we suggest that the ‘rhetoric of rights’ ¹¹⁸ continues to conceal paternalistic decision-making that undermines adolescent autonomy in refusal of medical treatment cases.