Update on Ethical Issues in Pediatric Dialysis: Has Pediatric Dialysis Become Morally Obligatory?

Abstract

Improvements in pediatric dialysis over the past 50 years have made the decision to proceed with dialysis straightforward for the majority of pediatric patients. For certain groups, however, such as children with multiple comorbid conditions, children and families with few social and economic resources, and neonates and infants, the decision of whether to proceed with dialysis remains much more controversial. In this review, we will examine the best available data regarding the outcomes of dialysis in these populations and analyze the important ethical considerations that should guide decisions regarding dialysis for these patients. We conclude that providers must continue to follow a nuanced and individualized approach in decision making for each child and to recognize that, regardless of the decision reached about dialysis, there is a continued duty to care for patients and families to maximize the remaining quality of their lives.

In his 1964 Presidential address to the American Society of Artificial Organs, Belding Scribner predicted that with improving outcomes and increased availability of treatment, the center of ethical focus on dialysis would move from “who should receive the artificial kidney” to “who shouldn’t.” (1) Fifty years later, improvements in pediatric dialysis and renal transplantation have changed the decision to pursue dialysis as a bridge to transplant in countries where it is available from “yes/no” to “yes” for the vast majority of children with ESRD (2). However, the answer remains less clear for certain populations. Although guidelines, such as those published by the Renal Physicians Association, provide useful advice for clinicians caring for these children and their families, a comprehensive analysis of the underlying data and ethical questions are beyond their scope (3). The purpose of this review is to perform a focused analysis of the most current data and ethics of pursuing, withholding, or withdrawing dialysis in these special populations, including children with multiple comorbid conditions, children with limited social resources, and neonates. It is generally accepted that there is no ethical distinction between withholding and withdrawing medical treatments, even if there may be an emotional distinction for the patient, family, and medical team, and the subsequent analysis therefore applies equally to both withholding and withdrawing dialysis (3,4). However, as the decision to withhold, rather than withdraw, dialysis is more common in the pediatric setting, we will primarily focus on withholding dialysis in this article. Finally, we would note that while the data presented predominantly discuss peritoneal dialysis, the core ethical principles addressed can be applied to all forms of mechanical RRT, including hemodialysis and continuous RRT.

Children with Multiple Comorbid Conditions

Advances in neonatal and pediatric intensive care, nursing, nutrition, and medical technologies have resulted in an increasing number of children surviving with multiple comorbid conditions. Some of these children may develop ESRD as part of their underlying condition or as a consequence of medical therapies received.

The decision of whether to initiate or forgo dialysis for patients with severe comorbidities is a complex one, comprised of at least two distinct questions. The first is whether dialysis in these patients results in outcomes comparable to other populations. The second is whether dialysis is a meaningful medical intervention for a given patient. Each of these questions requires a distinct analysis and consideration.

In regards to the first question, limited data are available. At least one study suggests no difference in outcome on peritoneal dialysis for children with special needs or social disadvantage (5). In contrast, Neu et al., reporting data from the International Pediatric Peritoneal Dialysis Network, demonstrated a higher hospitalization rate and lower 4-year survival rate in patients on peritoneal dialysis with at least one comorbidity (73% versus 90%, respectively) (6). Other studies have suggested renal transplantation is a therapeutic option in this population, with excellent graft survival and reported improvement in quality of life (7,8).

The second question is more complex. With the exception of a few extreme cases (e.g., severe sepsis precluding extracorporeal therapies or abdominal ailments limiting peritoneal dialysis) it can be presumed that the technical goals of dialysis, fluid and solute removal, will likely be achievable. However, whether this intervention is clinically meaningful depends on two factors: the impact that dialysis will have on the length and quality of the patient’s life. If a patient is extremely ill, it is possible that providing dialysis will have a negligible benefit in extending the patient’s life, due to their concurrent medical conditions. In that case, the benefits and burdens of the procedure must be weighed accordingly. The impact on the patient’s quality of life can be more difficult to quantify. Although initiation of dialysis can have significant burdens resulting from both initial and continuing provision of associated care, this must be balanced against the burdens of the patient’s acute or chronic disease and likely death resulting from forgoing dialysis.

In a recent article, Dionne and d’Agincourt-Canning offered an ethical decision-making framework based on the four-box method developed by Jonsen et al. (9), which centered on an analysis of benefits and burdens to the child in order to aid in the decision of whether to pursue dialysis in children with ESRD (2). The authors suggest that in situations where dialysis is clearly unlikely to be beneficial, it may be the duty of the medical team to be directive and take responsibility for a decision not to pursue dialysis (2). Similarly, Goldberg et al. (10) recently proposed a framework for the evaluation of kidney transplant candidacy for children with multiple comorbidities. The authors recommend individualized assessments and therapeutic plans based on an analysis of the benefits and burdens of therapy for the particular child, as opposed to criteria based on a certain level of intelligence.

While both frameworks do include other factors as considerations, both are clear that the focus should be on benefits and burdens directly to the child at present and in the future. Such an assessment can be challenging in any individual with a limited ability to communicate, including all young children and children with significant neurodevelopmental disability. In these cases, the calculus will almost always favor initiation/continuation of RRT, except in circumstances where the provision of dialysis results in additional burdens (essentially pain related to dialysis or transplantation) so great that any benefit of continued life would not be worthwhile.

Wilfond has offered an alternative framework for these considerations, arguing that clinicians have a special responsibility to provide services available only with the active participation of health care providers (such as dialysis or mechanical ventilation). Given that families of these children must live with the experience of these decisions for many years, their preferences should be given special privilege within the decision-making process (11). As a result, in situations of disagreement between providers and families, a family’s wish to pursue life-sustaining therapy should be accepted (11). This argument is based on the concept of relational potential developed by Arras, who argued that rather than focusing on interests of the child alone, we ought to consider whether a child has the capacity to form relationships with others and whether there are people who value those relationships (12). This means that depending on the family, it may be appropriate to initiate or discontinue life-sustaining dialysis for a child with a severe neurodevelopmental disability. This framework can be broadly applied to each of the special populations we discuss in this article, and may account for the numerous studies that suggest nephrologists do consider impact on families when making treatment recommendations and why the European Pediatric Dialysis Working Group Guideline (EPDWG) for the care of infants with stage 5 CKD recommends consideration of quality of life for both the child and the family (13–16).

Some readers may argue that dialysis should only be pursued in children as a bridge to renal transplantation. It has been argued, citing the principle of nonmaleficence, that when a child is not a candidate for transplantation, dialysis may only prolong the dying process and so should be withheld or withdrawn (2). Renal transplantation is generally accepted as a superior modality of time-limited RRT, allowing children the best opportunity for growth and development. However the exclusion of children who are not transplant candidates from dialysis bears further scrutiny. This sentiment is based on two different arguments. First, that some children should not be candidates for transplantation, and second, those children who are not candidates for transplantation will not benefit from dialysis therapy.

A complete discussion of inclusion criteria for transplantation is beyond the scope of this article, however, traditional reasons to exclude children from transplantation have included intellectual or physical disability, other comorbid medical conditions (such as untreated cancer or sepsis), and those imminently dying. Recent controversies have suggested that traditional exclusion criteria may lack a strong ethical support and instead have reflected societal values of certain traits (such as intelligence) (10,17). Furthermore, newer research has demonstrated excellent outcomes in this population, including perceived improvement in quality of life (7,8). Practically, the number of children who could be candidates for dialysis but could not be future candidates for transplantation is very small.

The second argument fairs worse. It is unclear why a child who would not be a transplant candidate, such as a child with a recurrent form of FSGS, should not be offered dialysis therapy for that reason alone. In adult nephrology, patients are offered dialysis with or without the option of transplantation and it is unclear why children should be denied this opportunity (18). Dialysis, while clearly burdensome, does offer children the benefit of continued life even if it is at a reduced quality relative to transplantation. Finally, it is important to note that the ethical distinctions between dialysis and renal transplant as forms of RRT may be less substantive than they appear. Although few would argue that transplantation is not the superior modality, it is also a temporary measure and many children receiving a transplant will require dialysis as either a transient or permanent form of RRT later in life. It is not clear why these transitions in modality would change the ethics of providing RRT.

Some children who develop ESRD may be dying from a separate chronic or acute condition. In a situation where dialysis provides meaningful benefit to a dying child, it would not be reasonable to refuse to provide the therapy simply because the child will die from a separate condition. As in all other situations of considering initiation, withholding, or withdrawal of dialysis, such a decision requires a more complete and nuanced assessment of the family’s (and, when applicable, the patient’s) goals of care (2).

We feel that current research does not support a broad approach of limiting pediatric dialysis in the United States among children with multiple medical comorbidities even if the patient may not be considered a future transplant candidate. Future research is needed to increase our understanding of the prognosis of more common comorbid conditions, as well as to better understand the expected burdens faced by patients and families in those circumstances. Additionally, it would be useful to examine any potential benefits experienced by these families resulting from the initiation of dialysis and also renal transplantation, in as much as the decision of whether to proceed to transplantation in patients with severe comorbidities remains quite controversial (10). This will allow us to better counsel patients and their families during the decision-making process.

Children and Families with Few Social Resources

Pediatric dialysis has long been recognized as a significant burden for families, particularly in the case of the smallest children and those with multiple conditions (19,20). It could be argued that these burdens are especially great on those families with few social resources. The challenges faced by the families of patients on pediatric dialysis are especially important to understand, as the quality of care provided by the family is a major factor in successful dialysis (21).

The limited research available shows that pediatric dialysis impacts an entire family. In a series of in-depth interviews with parents of children with CKD, Tong et al. identified four major themes: absorbing the clinical environment, medicalizing parents, disruption of family norms, and the need to develop new coping strategies and support structures (22). In the study, parents consistently reported having a child with CKD as “a pervasive and profoundly negative experience.” (22) Other studies have noted pediatric dialysis disrupts family life (23,24), affects parental marriages (23,25,26), parental health-related quality of life (HRQOL) (24), parental stress (26–29), and results in less parental time spent with healthy siblings (30) and more schooling problems in siblings (23). While all of these studies are limited by small sample sizes, future research may provide better guidance. In 2014, Parham et al. published the Pediatric Renal Caregiver Burden Scale, a new, validated survey of dialysis caregiver burden (31). The K-CAD study, a prospective cohort study of children with CKD and caregivers in Australia, began in 2014 and planned to enroll 756 families (32). These studies, as well as others, may allow us to better quantify the burdens experienced by families.

In addition, many families caring for children on pediatric dialysis will face substantial financial costs incurred via missed work, transportation expenses, and the like. Prior research has demonstrated that having a child on pediatric dialysis negatively affects family’s finances, a burden that will be experienced by the family as a whole (29,30,33). These expenses reinforce the importance of continued advocacy by the nephrology community for the provision of resources to our patients and their families, but also raise questions over how much expense is reasonable for a family to take on.

Parents possess finite psychologic, physical, and financial resources. It is possible that exceptionally needy children, such as a child with ESRD and multiple comorbidities, may take up so much of the parents’ psychologic and physical energy and financial resources so as to make them unable to meet needs of other children or themselves. Murray has argued that if the worth of the parent–child relationship is judged by how successfully it avoids unpleasantness and increases pleasures and that the burdens of raising a child with chronic illness are most likely to fall on the parents, then it would be reasonable to defer decision-making authority to parents in cases of poor prognosis (34). Such an approach suggests that parents should be able to refuse dialysis for children for whom their care will be particularly burdensome.

There are clear limitations to this argument. This approach overlooks the possibility for positive effects for siblings and parents, which may be difficult to perceive or forecast when initially considering the future treatment course. Furthermore, it is important to note that burdensomeness of any child is also a reflection of the kinds and intensity of available community support. Considering burdensomeness in this way favors pursuing life-sustaining treatment for children with families with greater social resources. The exclusion of these children based on consideration of their family’s resources would strike many individuals as inherently unfair, as the children bear no responsibility for those resource limitations, and such exclusion is explicitly contrary to the goal of providing pediatric medical care independent of their resources, as illustrated by programs such as the Children’s Health Insurance Program in the United States and similar approaches worldwide. A similar sentiment led to the inclusion of all dialysis patients within the Medicare entitlement.

Finally, it is not clear whether, if the burdens experienced by a particular family are overwhelming, that forgoing dialysis for the child would be an ethically appropriate remedy. While in many nations, the availability of resources necessary to provide care for a patient or support for their family may be an absolute limitation, this is rarely the case within the United States, Canada, and other developed nations. In these circumstances we instead experience a relative limitation, in that we lack the mechanisms to provide the necessary support or have not yet developed a social consensus on the appropriateness of providing the necessary degree or type of support. Even with these limitations, it does not mean that no interventions are available. The more modest form of these interventions include things like financial support offered by charities or through resources directed by medical centers themselves. More intensive forms of support include interventions such as home-based nursing support, or placement in a foster home or communal setting for ongoing therapies. These more intensive interventions can also result in burden or harms for the patients and their families, and as a result may not be appropriate in some or even many circumstances, but cannot be dismissed from consideration out of hand.

We feel that current research does not support a broad approach of limiting pediatric dialysis in the United States due to limits on families’ social resources as an isolated factor. Additional research is needed to characterize the costs of dialysis, both social and financial as well as both direct and indirect, for children and their families. This will allow health care providers to better assist patients and families in decision making as well as inform both institutional and societal approaches to this at-risk population.

Neonates and Infants

Early reports of neonatal and infant dialysis from the 1980s showed high mortality and morbidity, including poor neurodevelopmental outcomes. Responding to these reports, Cohen (19) concluded in 1987 that dialysis for very young infants was not morally obligatory and the therapy was disproportionally burdensome, did not have a predictably good success rate, was technically difficult to provide, could create pain and suffering, and had uncertain short- and long-term risks. Since that time, multiple factors including advances in dialysis technology (such as continuous ambulatory peritoneal dialysis and continuous cyclic peritoneal dialysis), improvements in nutrition and nursing care, and the recognition of the deleterious effects of aluminum (35) have led to dramatic improvements in the outcomes of neonates with ESRD (Table 1) (35–45). A 2007 North American Pediatric Renal Trials and Collaborative Studies special analysis demonstrated no difference in survival between children initiating dialysis at <1 month of age compared with those aged 1–24 months at initiation (46). An updated analysis in 2015 demonstrated further improved survival for neonates and infants initiating dialysis (47). While these studies must be interpreted with some caution due to selection and publication bias, they clearly show a dramatic improvement in outcomes over time. With these improvements, individual centers have become more willing to provide dialysis in increasingly complex clinical circumstances. Recently, case reports and the popular press have highlighted successful dialysis performed in an extremely low birth weight (<1500 g birth wt) newborn and preterm infant with bilateral renal agenesis with survival and apparently intact neurodevelopmental function (48,49). In addition to improvements in mortality, data suggest that the risk of morbidity in this population may be less severe than previously thought. A recent prospective study of infant dialysis found that infants without preexisting neurologic abnormalities did not develop severe additional neurodevelopmental conditions after initiation of peritoneal dialysis (45).

Table 1.

Survival after initiation of RRT in neonates and infants

Publication	Dates	Age at Onset of RRT	Outcome
Rheault et al. (36)	1995–2004	<28 d	52% 1 yr survival; 48% 5 yr survival
Hijazi et al. (37)	1983–2008	<1 mo	46% survival
Pollack et al. (38)	1997–2013	<1 mo	86% survival (HD only)
Vidal et al. (39)	1995–2007	<1 mo	80% survival
Wood et al. (40)	1992–1996	<3 mo	83% 1 yr survival
Alexander et al. (41)	1992–2007	<3 mo	63% survival
van Stralen et al. (42)	2000–2011	<1 mo	81% 2 yr survival; 76% 5 yr survival
Wedekin et al. (43)	1997–2008	<1 yr	79% survival
Ledermann et al. (20)	1986–1998	<1 yr	80% survival
Mekahli et al. (44)	1986–1997	<2 yr	87% 1 yr survival; 77% 5 yr survival; 75% 15 yr survival; 72% 20 yr survival
Laakkonen et al. (45)	2001–2005	<2 yr	95% survival
Carey et al. (46)	1992–1999	<31 d	68.7% survival
Carey et al. (47)	2000–2012	<31 d	78.6% survival

These studies report outcomes for all children for whom RRT was initiated rather than all infants with ESRD. Reasons for exclusion could include small size, prematurity, compounding medical conditions, or ethical views of parents/providers. As such, these studies may skew to more favorable outcomes. HD, hemodialysis.

Despite these improved outcomes, the beliefs of pediatric nephrologists have remained relatively static. Two international surveys performed in 1998 and 2008 show an overwhelming majority of nephrologists (93% and 98%, respectively) will offer dialysis to some infants with ESRD, but not to all (41% and 30%, respectively) and the majority feel that it is acceptable for parents to refuse RRT in children <1 month (81% and 73%, respectively) (Table 2) (15,16). In 1996, Bunchman advocated for an approach of shared decision making to choose between courses of aggressive dialysis therapy, “wait and see,” and comfort care only (20). This view was echoed more recently in the 2012 EPDWG Guidelines recommendation that “consensus decision-making based on guidelines is recommended for both commencement and withholding of RRT in the neonate or infant.” (14) The guideline authors recommended consideration of short- and long-term prognosis, medical care issues, and the predicted quality of life for the child and the family (14). An informal survey of members of the EPDWG found that 55% of respondents felt it is acceptable to respect a parent’s wish to refuse dialysis for an otherwise normal neonate, whereas only 26% felt parental refusal was acceptable for an infant 1–12 months of age (50). In 2013, Lantos and Warady highlighted the improved outcomes in this population, but again concluded that comfort care only should continue to be offered as a treatment option in neonatal dialysis, citing concerns of burdens placed on the family, expense, and a different societal perception of resuscitation in neonates compared with others (51).

Table 2.

Pediatric nephrologists’ beliefs about dialysis for infants

Series	Geary (15)	Teh et al. (14)
No. of responders (response rate)	361 (43%)	180 (21%)
Offer RRT to any infant <1 mo	93%	98%
Offer RRT to all infants <1 mo	41%	30%
Believe it is never acceptable for parents to refuse RRT for infants <1 mo	19%	27%
Believe it is usually acceptable for parents to refuse RRT for infants <1 mo	50%	NA

NA, not applicable.

Janvier et al. have demonstrated that pediatricians and the educated general public consider interventions differently in the very young versus older children and adults (52,53). As a result, we may be less supportive of aggressive interventions in this population. This should be coupled with awareness of the fact that we are poor at prognosticating renal outcomes in antenatal and neonatal patients with structural renal abnormalities (54). Hogan et al. demonstrated that among ten infants for whom a prenatal termination of pregnancy was recommended due to oligohydramnios and structural renal abnormalities, 50% survived and demonstrated normal renal function following the neonatal period (54). Additionally, data suggests that neurologic and respiratory outcomes for neonates with antenatal diagnosis of oligohydramnios may be better than previously thought (55). It is important for providers to be aware of these considerations when making treatment recommendations, particularly for newborns.

In light of the improvements in outcomes, newborns with ESRD may now be expected to have comparable survival to newborns with Down syndrome and intestinal atresia (56). We recently explored why intervention may be considered mandatory for neonates with Down syndrome and intestinal atresia and not neonates with ESRD (56). Children with both conditions will require greater familial, medical, and societal resources than an otherwise healthy child. Possible differences could be the greater certainty that newborns with ESRD will continue to require intensive, ongoing medical interventions than a newborn with Down syndrome, the greater availability of resources to support children with Down syndrome, and the limited ability of some centers to perform infant dialysis (56,57). Another possible explanation could be perceptions by providers of lower quality of life among children on dialysis and high burdens placed on parents and caregivers. Although children with ESRD do report diminished HRQOL scores when compared with healthy controls, HRQOL scores for children after renal transplantation appear to be similar to children with other types of chronic diseases (58–62). Furthermore, it is not clear why future quality of life would result in withholding treatment for a newborn, but not a toddler or older child.

We feel that current research does not support a broad approach of limiting pediatric dialysis in the United States based on the age of a child as an isolated factor if it is technically possible to perform the dialysis. Rather, neonates should be considered in the same manner as other children, including considerations of comorbity and family social resources, as discussed above. Future research efforts will need to better describe outcomes of all newborns with ESRD, including survival, quality of life, school participation, and expense. This information will aid physicians in appropriately counseling parents regarding the decision to pursue RRT. Given the surprising findings of the study by Hogan et al. (54), it is clear that particular focus should be paid to potential markers of poor outcome that can be detected in utero, to best aid nephrologists in providing appropriate prenatal consultation.

Conflicts in Decision Making

It is possible, despite best intentions, that these cases may result in intractable disagreement or significant moral distress for the patient, family, or health care team. In such a case it is reasonable to consider consultation with a hospital ethics committee. Request for state intervention to enforce a treatment decision over parental objection represents a serious challenge to parental authority and autonomy and may permanently alter a family’s future interactions with medical providers. Pursuit of state intervention should be considered only after the medical team has clarified the reasons for the family’s reluctance to give permission and only after any misconceptions or possible alternatives have been addressed.

Conclusions

The decision to initiate or forgo RRT is a difficult one and requires a careful examination of a child’s best interests and the benefits and burdens of therapy for both the child and his or her family. Although many patient populations face particular challenges, it is important to realize that our presumptions about the extent of those challenges may not be supported by existing data. Additionally, while mortality and clearly defined morbidity are relatively easy to measure, these measures may not capture the most meaningful aspects of a given patient’s care. Providers must continue to follow a nuanced and individualized approach for each child and their family and recognize that, regardless of the decision pertaining to the use of RRT, there is a continued duty to care for patients and their families so as to maximize the quality of their lives. This may mean that RRT should be pursued for most children with ESRD.

Disclosures

None.