The Effects of Introducing a Harm Threshold for Medical Treatment Decisions for Children in the Courts of England & Wales: An (Inter)National Case Law Analysis

In S31 orders it is the task of the applicant, in all identified cases the Local Authority, to prove, on the balance of probability [59], that the harm threshold is crossed.

To do so the Local Authority submits a ‘threshold document’ to the court in which they set out their evidence. For an applicant it is advantageous to present as much evidence as possible in order to maximise the chance that the harm threshold is considered crossed. Due to the objective standard against which parental care is measured there is a focus on parental behaviour and characteristics that can be presented as parental failings. In the included cases Local Authorities have presented relatively trivial evidence i.e. missed medical appointments [52, 57] and previous occasions when parents did not follow medical advice [3, 56, 57] as well as more serious concerns such as ‘inappropriate’ behaviour towards healthcare professionals [56], evidence about parental abilities and/or mental health [14, 42, 46, 53, 57] and parental personal history and relationships [14, 53, 57]. This scrutiny of parental failings is lacking in court decisions about medical treatment using a best interests test; whilst parental reasons are scrutinised, parental failings are not. Due to the emphasis on parental failings the content of the threshold document can be experienced as both intrusive and adversarial which may negatively impact future relationships between the family and the clinical team when the applicant is a NHS Trust.

In order to decide whether the harm threshold is crossed judges must first establish the facts. In decisions about medical treatment the medical evidence about the health condition of the child and the benefits/harms of proposed treatment is crucial in the establishment of those facts. Whilst the traditional deference of the court to medical experts may have abated to some extent [15, 62] it is still undeniably true that doctors are in a much better position to provide medical evidence than parents. More so as the decision that the harm threshold is crossed is accepted to be a value judgement [61]. Indeed, in all identified cases the medical evidence was accepted and thus the harm threshold considered crossed. That includes decisions in which the medical evidence is an opinion rather than fact-based. In Re R[57] for examplethe parents preferred to tube-feed their child by blending regular food (blended diet) rather than using commercially prepared feeds. Despite a lack of evidence that a blended diet is inferior, the judge decided the child should be fed with commercially prepared feeds given the preference of the treating doctor. The same importance to medical opinion is attached when establishing future harm. To establish future harm the applicant has to show that future harm is a ‘real possibility’ [61]. In M-W (a child) [46] an appeal court directed a case to be re-listed in High Court, to enable a medical opinion regarding the child’s future psychological and emotional development on the basis of maternal characteristics to be taken into account when no harm was demonstrable at the time of the ruling. The approach was confirmed by the Supreme Court a few years later when, in the absence of demonstrable harm, a care and adoption order was approved against parental wishes based on parental characteristics only [55].

In summary, in applying the harm threshold for medical treatment decisions in S31 orders factors come into play that are not present in decisions using the best interests test. In the context of S31 orders the threshold decision also takes into account parental characteristics and behaviour interpretable as parental failings resulting in a more adversarial procedure. In addition, medical evidence plays a crucial role in the determination whether the harm threshold is crossed also when it is based on opinion rather than scientific facts, putting parents at a considerable disadvantage.

In the first case the proposed treatment was experimental and had an estimated chance of curing the child’s leukaemia of 10% [29]. The case was brought to court because the parents disagreed; the mother favoured treatment given its life-saving potential, the father declined because of its burdens. In view of the low likelihood of cure and reasonable arguments of both parents in future similar cases might not cross the harm threshold and thus denied a court decision. However, when parents cannot agree and the issue is a serious one, there is an expectation that clinicians apply to the court for a decision [5]. Should the introduction of the harm threshold for triage deny cases with limited harm and reasonable but disagreeing parents a court hearing, those parents might be forced to litigate against each other instead.

In the second case the NHS Trust requested the court’s consent for surgery to enable continuation of haemodialysis for a child with kidney failure. The clinicians agreed that the choice between active treatment and palliative care for this particular child was evenly balanced [44]. A parental decision to refuse consent would thus seem reasonable. However, the parents based their refusal exclusively on their religious views, favouring prayer therapy instead. Likely, the clinicians’ unease about parental decision-making led to the court application. For now it remains an open question whether a parental preference for unproven treatment would fall below the standard of ‘reasonable parent’ when proven treatment has limited chance of success. Should in future such cases be denied access to the court an application contesting parental capacity might be made in some cases.

In 16 cases across the jurisdictions and 4 in England & Wales consent was sought from the court to allow or prohibit non-life saving treatment.

In all three jurisdictions cases in which treatment sought to prevent the loss of vision or hearing have been litigated, including in a S31 order [14]. Similarly, cases about proposed medical treatment for children with severe mental health issues have been heard in Germany, the Netherlands and in a S31 order [62]. Such cases will thus continue to be heard in England & Wales should the harm threshold for triage be introduced.

However, harm to the child is not always imminent. In 6 cases litigated in the Netherlands and Germany a variety of medical treatment was proposed aiming at preventing harm (mostly) experienced in adulthood such as short stature and cardio-vascular disease. Should similar cases arise in England & Wales the harm threshold in itself should not deny such cases a court hearing.

In one of the remaining two cases about non-life saving treatment decided in England & Wales, the parents did not accept the diagnosis of incurable cancer and therefore palliative treatment [7]. Given that the diagnosis was well established the parental reasons for their refusal to consent might fall below the standard of reasonable parents. Moreover, the child was in considerable pain. Taken together, this case would have crossed the harm threshold and similar cases can be expected to be heard in court.

The second case is more controversial. The parents refused to consent to a brainstem test, to determine whether the child was legally dead, as they feared that the test might further injure their child [11]. The child was thought by clinicians to have lost the capacity to suffer and thus would not suffer himself from continued intensive care treatment. Under those circumstances it seems possible and perhaps even likely that the harm to the child posed by continued medical treatment will not be significant enough to cross the harm threshold and similar cases might not be heard in court.

In summary, after an introduction of the harm threshold for triage most cases about life-saving treatment will continue to be heard in the courts of England & Wales. A likely exception are cases in which the chance of successful treatment is low. So far such cases are rare and only lead to court applications when there are concerns about parental decision-making. Most cases litigating non-life saving treatment would also still be heard given the extent of potential harm in cases litigated so far. However, cases in which the child is no longer capable of suffering may be denied a court hearing.

In 30/32 cases the courts provided substituted consent for proposed medical treatment intended to be life-saving. In more than 30 years of litigation across the three jurisdictions only twice did courts approve parental refusal of life-saving treatment.

In Re T[54], the appeal court in England & Wales, decided it was not in the child’s best interests to have a life-saving liver transplant. In the specific circumstances the transplant would involve the mother being forced to stay in England to care for a child with substantial medical needs due to a procedure she opposed given the burdens of treatment, whilst the father lived and worked abroad. Re T does however seem to be an outlier as equally invasive treatment also requiring considerable familial input, e.g. a bone marrow transplant, has since been allowed against the wishes of the parents [49]. Similarly, courts in Germany and the Netherlands have authorised intensive treatment, i.e. chemotherapy demonstrating that a harm threshold does not prohibit intensive medical treatment against parental wishes when potentially life-saving.

In the second case a 4-year old German child in ‘Wachkoma’² but suffering from severe, painful muscle spasms following a hypoxic incident resided in a rehabilitation centre. Given that the child was not expected to regain awareness and the painful symptoms could not be relieved except by deep sedation the parents decided stop her clinically-assisted feed and allow her to die at home under the supervision of a palliative care specialist. Her doctors disagreed. The lower court declared the harm threshold crossed [4]. However, the appeal court found that given the prognosis and the inability to treat her pain other than by sedation the foreseeable death of this particular child did not cross the harm threshold [50].

The above demonstrates that, regardless of the standard used, for parents to obtain the court’s approval of their refusal of life-saving medical treatment is a very high bar indeed. In the two cases in which the death of the child was allowed, the exceptional circumstances determined the court’s ultimate decision. The second case also demonstrates that health care professionals are uncomfortable when parents decide, ahead of them, that their child should be allowed to die. Such cases have been discussed in the literature [16, 64] and in England & Wales has been the subject of a S31 order [57].

In most cases, the courts also allow medical treatment that is not intended to be life saving. In 12/14 cases (85%) in which parents refused proposed medical treatment not intended to be life-saving, the courts provided substituted consent. Both cases in which the court declined to do so were decided using a harm threshold.

In Germany deaf parents refused the implantation of a cochleair device in their youngest child [6]. In a reasoning very similar to that in Re T[54] the German court balanced the impact on family life, the child spending many hours outside the family home in order to acquire spoken language after implantation of the device, against the harm of not being able to hear and concluded that the harm threshold was not crossed. The case demonstrates that a court decision using a harm threshold can be equally holistic as when using a best interests test. However, whether a holistic approach is taken does depend on how the harm threshold is defined. Should the German case described above have been heard in the Netherlands the court might well have provided substituted consent for cochleair implantation because the Dutch harm threshold only allows consideration of the health of the child. Moreover, that factors can be taken into account does not mean that they will be taken into account. For example, in circumstances in which parents were thought not to be able to afford the implantation of a cochleair device a court in England allowed a care and adoption order [14] against their wishes. The judgment takes into account medical evidence but not the potential harm caused by the child losing contact with her birth family.

In the second case, a Dutch court was asked to allow a renal transplant for a child suffering from chronic renal failure [23]. Whilst the child’s health was stable on dialysis, the main aim of the transplant was to prevent serious cardio-vascular problems in adulthood caused by metabolic dysregulation inherent to renal failure. Because the harm would be experienced in adulthood and treatment could be deferred to when the child could decide the court considered the harm threshold not crossed. Given that part of the harm does occur in childhood it is likely that the best interests test would have allowed the renal transplant to go ahead in childhood. The argument that a decision can be deferred until children can decide for themselves has also been used in Germany to prohibit cross-sex hormone treatment for gender dysphoria. However, deferring treatment until the child can decide may be less convincing to the courts in England & Wales. In 2013 a court approved booster immunisation against mumps, measles and rubella (MMR) for a healthy 15-year old against her wishes [24]; a decision that could have been deferred until she reached adulthood with minimal or no harm.

In summary, in the vast majority of cases in which parents and clinicians disagree about proposed treatment the outcome is similar regardless of the standard used in court. In most cases the courts provide substituted consent. The similar outcomes are likely due to the severity of preventable harm in the cases heard in the courts across the three jurisdictions.

Taken together, the results indicate that introducing the harm threshold either for triage or as standard for decision-making will not substantially limit the number of future court applications. Nevertheless, the investigation found three characteristics of cases for which litigation might change. When introduced for triage two types of case could be denied a court hearing; 1) cases in which harm is limited and there are concerns about parental decision-making and 2) when parents refuse treatment for a child not capable of suffering. Only under these specific circumstances will the introduction of the harm threshold likely increase parental discretion. An introduction of the harm threshold as standard might change the outcome of cases when treatment can be deferred to adulthood increasing the autonomy of the child rather than their parents.

Parental discretion might be further increased should a harm threshold be introduced that does not contain an objective standard of parental care. An introduction of a harm threshold for medical treatment decisions that directs the courts to focus on the significance of harm only might change the outcome of cases in which the harm is limited. For example, in the Netherlands a mother based the refusal of a renal transplant on her belief that her child would be miraculously cured [23]. In England & Wales it is not uncommon for parents to place their faith in a ‘miracle cure’ [48] or have other faith-based reasons [45] which currently are unlikely to take precedence over medical evidence. A harm threshold that only takes into account the harm and not parental reasoning may go some way to broaden parental discretion when harm is limited.