Bioethics for Neonatal Cardiac Care

A 28-year-old female at 20 weeks gestation presents following an ultrasound general anatomy screen. Results are concerning for presence of congenital heart disease.

Pregnant women who screen positive on their anatomy scans often present to a pediatric cardiologist for identification and counseling on a wide range of heart disease and defects.⁴  The pregnant woman is still under the care of her own doctor and the dichotomy between obstetrics and pediatrics are highlighted in their different approaches to clinical, legal, and ethical aspects of prenatal care.^5–7  Although obstetricians are centered on 2 patients, their time is largely spent with 1 adult who is the decision-maker with the ethical principles of autonomy and beneficence employed for the woman and fetus. However, beneficence has traditionally been based on meeting the woman’s goals and, thus, leans largely on respect for the woman’s personal autonomy. Whereas obstetricians are experts in maternal health, they lack expertise in counseling on the full spectrum of variability and implications of pediatric conditions. Pediatric specialists are “future child” focused and see the pregnant woman as a parent and surrogate decision-maker.⁷  They are less likely than obstetricians to think of termination as an option, less likely to weigh maternal health and risks, and are more likely to consider court intervention to protect fetal welfare.⁶  The American Academy of Pediatrics and the American College of Obstetricians and Gynecologists jointly acknowledge these different views and the difficulties that can arise in the care of the pregnant woman and fetus, and therefore recommend a cooperative, multidisciplinary approach.^8,9  The benefit of including both cultures is that the pregnant woman can be better informed to make decisions. Depending on the condition, prenatal diagnosis may provide information for the families to weigh all treatment options, including termination, fetal therapy, expectant management followed by postnatal life-sustaining treatments and palliative care.¹⁰  The prenatal family may be more ready to hear and understand things than the shocked family whose baby is sick and diagnosed postnatally.¹¹  The antenatal diagnosis is considered valuable in providing better perinatal planning: including prenatal monitoring, alternate delivery hospital to avoid mother and infant separation, and mode of delivery,^11,12  all of which might improve infant outcomes.^13,14  The more diagnostic and prognostic uncertainty, the more difficult it is to declare a “fetal best interest,” allowing more weight to the maternal interests, and less moral obligation to the fetus.¹⁵  Professionals should allow the mother, father, and others they may choose to help make decisions, weighing values and competing interests while informing the family that even if the diagnosis may have a good level of certainty, aspects of the future for that fetus and family are still largely unforeseeable. The discussions and planning, when operating with these unknowns, will attempt to balance the well-being of the fetus and the pregnant mother. Other stakeholders may include: high risk obstetrics, maternal fetal medicine and pediatric subspecialists in: prenatal and transitional pediatrics, cardiothoracic surgery, cardiology, neurology, radiology, neonatology, critical care, palliative care and ethics. Discussions should be documented in the medical record.

Organizational interests, such as a call to become a center of excellence, and financial ramifications impacted by these place-of-delivery decisions should be acknowledged, yet not be primary components of decision-making. These cases affect health care resource utilization and expenditure. Added fetal echocardiograms, neonatal transports, NICU admissions, and echocardiograms will be done that will escalate costs for some otherwise healthy children in the name of safety and better outcomes for those with a defect.

Some centers offer fetal cardiac interventions, including pharmaceutical treatments for fetal arrhythmias, fetoscopic procedures, and specialized delivery management for severe CHD.¹⁶  For the pregnant woman there are procedural risks for the current pregnancy, risks of maternal illness, considerations of future reproduction, and death with outcomes that are far from guaranteed. When approaching the case of a possible fetal intervention, it is again paramount that the fetal care center self-reflect and acknowledge what is standard of care, medical necessity versus convenience in care, innovative clinical medicine, or research.¹⁷ 

Good quality SDM and communication is required to navigate the above decisions and considerations.^9,18  In the next section, we explore best practices and evidence of SDM. This foundation begins in the prenatal or perinatal period and continues throughout each interaction with the medical system throughout the patient’s life.

• Multidisciplinary approach to antenatal counseling is reasonable to aid in good quality SDM. Consider including the following specialists in discussions with the parents: high risk obstetrics and pediatric subspecialists in: prenatal and transitional pediatrics, cardiothoracic surgery, cardiology, neurology, radiology, neonatology, critical care, palliative care and ethics (Class IIa, LOE C-EO)

The infant is born with evidence of severe congenital heart disease. A difficult choice must be made between at least 3 less-than-desirable alternatives.

Patient autonomy is not possible in the neonate, rather, when outcomes are uncertain and decisions between ethically acceptable treatment courses are required, clinicians and parents must work together to make plans that are in the baby’s best interests, or as Ross and Frader present, a best-interest standard that “does not require what is ideal, but what is reasonable.”¹⁹ 

The goal of SDM in this population is to engage parents and clinicians in partnerships to make medical decisions, especially when the best options are not clear.^20,21  In today’s health care environment, SDM has widespread implementation and is supported by many medical professional organizations as well as by policy initiatives.²² 

Debate may exist when more than 1 ethically-reasonable choice is available for some perinatal and neonatal cardiac conditions. To better recognize when SDM is appropriate, it is helpful to delineate when it is not appropriate. SDM is not presenting the appearance of a choice between treatments when only 1 choice is ethically acceptable.

SDM is not a discrete event related to a single decision. Instead, it is an evolving process or conversation between clinicians and parents.^23–25  This process begins with prenatal diagnosis, decisions about genetic testing or surgical intervention, and in the case of severe illness, optimal timing for inclusion of palliative care involvement. These decisions may become more complex and value laden over time, especially when complications occur.

We will consider a few key concepts and questions clinicians should ask of parents and themselves when engaging in SDM for care of neonatal cardiac disease include:

1. What is your understanding of your baby’s condition? Authentic and effective SDM is more than information disclosure and exchange. It requires assessment of the parents’ understanding of the decision and its implications. As we continue to strive to improve the process of shared decision-making, we should aim for a caring relationship as proposed by Noddings where “the care-er puts aside her own values and projects and tries to understand the expressed needs of the cared-for.”²⁶  In addition, a true dialogic conversation with parents, as described by Bakhtin, is where “our words are always half ours and half someone else’s.”²⁷  Families present with different decision-making frameworks based on medical experience, spirituality, education, race, and ethnicity, which make up their values. Because of this, caring parents may make different decisions on behalf of their neonates.²⁸  Only when we develop this type of relationship with the parents of neonates with congenital heart disease can we understand which outcomes are most important to discuss with them.

2. How do you want to be involved in making decisions? Clinicians should understand that different families have varying degrees of desire for involvement in complex decisions.²⁹  A one-size-fits-all model to include families in decisions ignores the tremendous variability among them. True partnership in decision-making requires recognition and support of these differences. As in the PICU literature, clinicians should be aware that the desired role of a parent in decision-making might change with the gravity of the decision or the lack of clarity of outcomes.³⁰  Families may also differ on who they rely on for support to make their decisions.³¹  Clinicians should ensure that all families feel supported and included in decisions at the level they choose.

3. How can clinicians communicate facts with transparency in the setting of uncertainty? SDM becomes more challenging with lack of standardization across centers and in the face of uncertainty. Clinicians must become well-versed and be able to discuss with families not only mortality data, but also neurodevelopmental outcomes, long-term effects on family members and dynamics and reported quality of life outcomes for both the groups that chose surgical interventions, and those who did not. Families facing difficult decisions should be informed if there is no agreement in the larger field or among centers about the recommended path for management. Clinicians must feel comfortable disclosing that oftentimes there is a choice beyond surgical repair and alternative treatment offered by centers other than their own. They should not solely recommend surgery if palliative care is a reasonable option.

4. How to acknowledge clinicians’ bias in practice? Everyone, including physicians and health care professions, has implicit or unconscious biases that affect how they interact with the world and others around them.³²  Effective decision-making requires an unbiased approach,³³  and yet we find this barrier difficult to overcome. Though outcomes continue to improve for hypoplastic left heart syndrome, the example often presented shows that only 44% of clinicians would hypothetically choose surgery for their own child.^34–36  And yet, we see personal experiences change attitudes, as NICU nurses who delivered premature infants themselves report making different choices than what they would have made prior.³⁷  Awareness of how language choices can introduce bias will also allow for a more balanced discussion. How a decision is framed, which words are chosen, and the tone in which they are stated may all influence decision-making.³⁸  Implicit bias training and policy insights have increasingly been recognized as important steps in mitigation of bias. The use of language interpreters and cultural navigators, when appropriate, is key to engaging families in shared decision making as is allowing sufficient time to process information in the iterative process.

5. How to give recommendations? Clinicians have an ethical obligation to provide guidance and adapt their recommendations to help families achieve their goals. Parents want honest and clear communication and clinicians should therefore minimize scientific language and attempt to connect and synchronize with the family.^39–41  A range of ethically justifiable approaches in clinician directiveness is appropriate in helping families navigate decisions.⁴²  Failure to provide these recommendations might paradoxically represent inadequate shared decision-making.⁴³ 

1. Discerning when SDM is appropriate is a reasonable first step (Class IIa, LOE C-EO)

2. Listening and acknowledging parents’ goals and values is recommended (Class I, LOE B-NR [Non-Randomized])

3. Assessing parents desire to be involved in the decision can be useful (Class IIa, LOE C-LD [Limited Data])

4. Being clear, consistent, and honest is indicated (Class I, LOE B-NR)

5. Bias should be acknowledged and mitigated (Class I, LOE B-NR)

6. Helping families achieve their goal by adapting recommendations and titrate directiveness is useful (Class I, LOE C-EO)

7. Recognizing that diagnosis informs but does not fully determine prognosis is reasonable (Class IIa, LOE C-EO)

Though the parents of the neonate had declined genetic testing during the fetal period, anatomic features and other patterns strongly suggests a presence of a genetic disorder. The team wishes to complete genetic testing to better understand prognosis.

When a genetic diagnosis is suspected based on a combination of cardiac or other abnormalities, it can be important to the clinical management of the patient to diagnose the genetic disorder more specifically. Some families may refuse testing, whereas others will be eager for extensive testing and the value of a contribution to disease based registries. The team should not underestimate the stress of a genetic diagnosis on a family, as they may blame themselves for causing the condition or worry about other family members being at risk for the diagnosis. Although the best interests of the patient remain of a primary concern,⁴⁴  it is important to both be aware of the potential need for additional family support and to examine the strength of a recommendation for testing. Ongoing discussions with the family and a model of titrated shared decision-making can be helpful in the approach to genetic testing, with consideration placed on the direct benefits to the child. For example, testing that will affect further management in the pediatric period should be explained to parents as beneficial and valuable to the care plan, whereas information gleaned for the child’s adult life or reproductive future may be better delayed for when autonomy is present.

Some lessons could be learned from the history of cardiac repairs in children with Down syndrome. As public opinion gradually became more accepting of persons with Down syndrome and data suggested they had similar postoperative outcomes, cardiac repairs in these children became more common, though not fully standard of care until 2 decades later. This population highlights the degree to which social, ethical, and legal factors define the context of clinical decision-making.⁴⁵ 

Over the last several years, there has been a small yet increasing interest in corrective and palliative surgery for CHD in infants with Trisomy 13 or Trisomy 18. Some centers now offer surgical interventions for neonates with Trisomy 13 and Trisomy 18, typically in the range of bands and shunts and ventricular septal defect closures with some degrees of success. However, the overall early mortality rate remains high in this select neonates cohort^46–48  and many parents desire expectant management with palliative care.⁴⁹  Given the significant morbidities accompanying these diagnoses, shared decision-making is appropriate to decide what intervention, if any, is appropriate.

With advances in technology for chromosomal testing, clinicians must educate themselves on the sensitivity and specificity of genetic screening versus diagnostic tests. A negative screening test does not rule out a genetic problem. In-depth genetic testing (eg, chromosomal microarray, expanded genome testing) for cardio-genetic conditions may shed light on clinical conditions when pathogenic variants are identified, but variants of uncertain significance lead to difficulty in interpretation. A variants of uncertain significance classification as outlined by the American College of Genetics and Genomics signifies that there is insufficient or conflicting evidence regarding a molecular alteration’s role in disease.⁵⁰ 

Clinical geneticists and genetic counselors can help parents and physicians process the uncertainties introduced by cardio-genetic evaluations.⁵⁰  Attitudes toward prenatal testing have largely paralleled the rise of genetic counseling as a medical profession and the commercialization of many tests. The demand for genetic counselors far outweighs current numbers in the workforce, which may result in suboptimal counseling and disclosure, adding to the situation’s complexity.⁵¹  Hence, it is important to determine whether a recognizable genetic pattern exists. Cardio-genetic conditions may involve other important organ systems, impact clinical and outcome prognostication, inform genetic reproductive risks for the particular family, and suggest genetic testing for other family members.⁵² 

• Involving clinical geneticists and genetic counselors to assist in the complex decision-making and interpretations of genetic testing results can be useful (Class IIa, LOE C-EO)

The team discovers that the infant qualifies for participation in a research study examining a new medication. Parents are approached for consent.

Respect for persons (ie, autonomy), justice, and beneficence are the principles to guide ethical research as delineated by the Belmont Report in 1979⁵³  and served as the basis for creating Institutional Review Boards to ensure ethical conduct in research. The current Code of Federal Regulations Title 45, part 46 includes additional protections for neonates and unborn fetuses in subparts B and D, respectively as vulnerable populations who lack autonomy.⁵⁴  Excluding these populations from research development has contributed to the need for off-label use of devices and medications and potentially infringes upon justice, which allows equal access to the risks and the benefits of research. To counteract this effect, more recent legislation incentivizes pharmaceutical and medical device companies to study medications in pediatric populations.⁵⁵  As Sable and colleagues explained in a 2017 multiorganizational Health Policy, “an understanding of the importance of research to advance childhood therapies has led to a new perspective that neonates need appropriate protection of rights and safety during research rather than protection from research.”⁵⁶ 

Several excellent collaborative papers by professional organizations have addressed some of the more difficult challenges cardiology studies present. First, outcomes of clinical care and investigative work may be misrepresented (highlighting successes over failures, configuring favorable denominators) through a desire to innovate with a new procedure, attract referrals, or bring a new medical device to market. To this end, in 2011, 45 North American pediatric cardiology centers released a consensus statement of adherence to ethical marketing.⁵⁷  Pressure to expedite approval for a new device or medication is offset by statements of acceptability of using off-label treatments by an American Academy of Pediatrics joint Sections of Cardiology and Cardiothoracic Surgery and Orthopedics statement in 2017.⁵⁸ 

Equipoise, the agreement of clinical equivalence of risks and harms by a medical community, provides ethical justification for exposing neonates to different treatments in a randomized controlled trial. However, as we have learned from retrospective studies of surgical procedures for cardiac anomalies, such as transposition of the great arteries, which outcomes must be equivalent, are not always clear. Increased mortality in the introductory years of the arterial switch for transposition, for example, occurred before an improvement in long-term outcomes could be measured only years later.^59,60  How does one weigh a suboptimal procedure in which most surgeons are proficient against a new technique that promises physiologic superiority? How should a data safety monitoring board respond to reports of early deaths when a retrospective study years later may show a marked improvement in overall mortality with the same investigative procedure?

We suggest optimizing 4 concepts to help address these quandaries: (1) maximizing data gathering while minimizing risk; (2) attention to therapeutic misconception; (3) rigorous standardization in a trial’s design, and (4) transparency in informed consent. Clinical registries can teach us about long-term outcomes, such as neurodevelopment and later morbidities, and allow longitudinal study of patients receiving off-label treatments.⁶¹  Comparative effectiveness and outcomes research are newer methods of gathering evidence. Multicenter and even international trials can improve generalizability of studies with larger sample sizes. As clinicians evaluating the ethical acceptability of a trial, we must be cognizant of the powerful effects of therapeutic misconception in investigators and parents which may enhance the allure to “do something” over more conservative plans.¹⁷  Potential harms introduced by learning curves for procedures treating relatively rare conditions can be minimized through confining study to a few research centers by select trained investigators.⁶²  Through standardization and ample enrollment, trials should be designed to detect differences with an adequate power so that risks taken are minimized and results are dependable.^63,64  An expert consensus paper addressing cardiac catheterization standards advises conflicts of interest may also be minimized when regimens are strictly followed.⁶⁵  Finally, clinicians and their research colleagues must be transparent in the information they supply parents considering research enrollment for their neonates. This must include explaining potential conflicts of interest, unknown aspects of the study treatments, and alternative options, even those outside of their center.

At some point, clinicians and professional organizations must decide whether there is enough evidence to promote a successful treatment from “innovative” to “standard” treatment. Carefully designed, ethically-grounded controlled studies will provide quality evidence to facilitate such considerations.

1. Maximizing data gathering while minimizing risk is recommended (Class I, B-NR)

2. Minimizing therapeutic misconception is recommended (Class I, LOE B-NR)

3. Rigorous standardization in trial design is recommended (Class I, LOE B-NR)

4. Optimizing transparency in informed consent is recommended (Class I, LOE B-NR)

The infant has survived a series of medical and surgical treatments, however, has had multiple complications and now several cardiac arrests. She is currently dependent on life-sustaining technology, including ventilator and renal replacement therapy.

Parents of neonates born with CHD hope that their child will have a complete repair with a normal lifespan⁶⁶  and that their child will live as long and as comfortably as possible.⁶⁷  When the clinical team anticipates disability or death, it has an obligation to speak “clearly, compassionately, and collaboratively striving toward improving the quality of life for children, including when necessary, that part of life that is dying.”⁶⁸ 

The focus in this population for cure and intervention can have several unintended consequences. First, it leaves parents feeling unprepared when their child experiences clinical declines, inhibiting conversations discussing the quality of end of life⁶⁹  care, including discussions about their goals.^66,70,71  Second, this focus may result in moral distress for the medical team⁷²  and may make team members feel complicit in needless suffering or misappropriation of resources. They may even make a claim that continued interventions and the pursuit of a cure are futile.

Though not unique to neonates with CHD, early involvement of a pediatric palliative care (PPC) team can help guide these conversations and may ameliorate suffering at the end of life when appropriate. Earlier PPC involvement, including during the prenatal period, allows families more time to understand their child’s diagnosis and engage in complex decision making, with a deeper sense of support from the medical team.^67,73,74  PPC can reinforce good communication among the health care team, and between the team and family, help with anticipation all while accompanying the family longitudinally through prenatal and postnatal periods.⁷⁵  How parents define their role as “good parents” will influence their decisions at the end of their child’s life.^71,76,77  Exploring this concept with families in addition to remaining engaged with empathic curiosity is key to building trust and to understanding how to guide conversations and what recommendations to give. Ethics consultations can be useful as well and can help support families and teams using ethical frameworks and by sharing concepts based on contemporary and empirical ethics literature. Ethics consults should be documented in the medical record.

Palliative care can be provided concurrently with interventional and surgical pathways and need not be a dichotomy. Components of palliative care, such as advanced care planning,⁷⁸  pain and symptom management, and psychological and spiritual support can be provided alongside disease-directed care in children with CHD. Similarly, clinicians should not see PPC involvement as a sign of “giving up.” Earlier and longitudinal integration of PPC improves shared decision-making, communication, and coping in a population with high levels of parental depression and anxiety.⁷⁹  And evidence is growing that children with CHD and PPC consultation experienced less invasive interventions at end of life.^80,81 

Some centers may find automatic triggers for PCC useful in the perinatal CHD population. Given that a genetic diagnosis typically reduces surgical success rates,⁸²  diagnosis of overlapping genetic and cardiac conditions would be a useful trigger for pediatric palliative care involvement.^83,84  Other examples include those with no further surgical management options, antenatal diagnosis of single ventricle, ventricular device in situ, symptomatic heart failure, those awaiting heart transplantation,⁷⁸  and patients on extracorporeal membrane oxygenation. More needs to be understood on how to integrate palliative care principles into the care of children and neonates with CHD.⁸⁰ 

Ethical obligations at the end of life in the neonatal cardiac population thus rely on clear, consistent, and honest communication with families and require that teams adapt to their family’s individual needs, goals, and preferences, centering on what is in the best interest of the neonate.⁷¹  Maintaining strong relationships within the medical team and with families may help mitigate concerns for suffering and allow instead a focus on quality EOL care, with fewer interventions for the child and more preparation for the family. Similar to parents of children who die of cancer, the communication between parents and their medical team can reduce the perception of suffering and can help redirect focus on quality end of life care.^71,85 

1. Early palliative care involvement even in the prenatal period can be useful (Class IIa, LOE C-LD)

2. Using a genetic diagnosis with a cardiac condition as a trigger for palliative care involvement is reasonable (Class IIa, LOE C-LD)

As we look forward, reevaluation of outcome reporting through an ethics lens should place greater emphases on benchmarks of health equity and overall health and quality of life for patients and families over mortality statistics.

Recent studies analyzing CHD outcomes with respect to social determinants of health demonstrate a complicated interplay of factors. Disparities are evident, beginning with a higher incidence of congenital heart disease in newborns with maternal comorbidities more prevalent in women of color⁸⁶  and maternal socioeconomic hardship markers.⁸⁷  Health inequities continue to manifest throughout congenital heart disease diagnosis, treatment decisions, and long-term outcomes. A commitment to justice will require that outcomes incorporate addressing these health disparities with interventions targeting maternal, newborn, community, and regional levels.⁸⁸  Centers may fare worse when judged by typical benchmarks, such as length of stay while caring for higher-risk patients (eg, coming from difficult socioeconomic situations), thus requiring modified measures of quality health delivery.⁸⁹  State-wide registries that incorporate socioeconomic factors can help study causes and interventions and adjust such measures.⁹⁰ 

Cardiac centers should also be judged on their ability to prioritize ethical decisions shared with families. Outcome reporting on neurodevelopmental (motor, communication and cognition), functional capacity, and long-term quality of life assessments has remained suboptimal and inconsistent, despite advocacy for the use of neonates-reported health status as a key measure of cardiovascular health.⁹¹  Clinical ethics could inform and perhaps redefine which outcomes are most important in neonatal cardiology. Such determinations result in a change of resource allocation to often neglected aspects of holistic care of the CHD neonates, such as social support and a sense of coherence, which may have a stronger impact in quality of life than clinical outcomes.⁹²  Additionally, we must acknowledge the conflict of interest that these reporting systems have created in some instances, limiting the therapeutic options that a center is willing to offer. A modification of our current reporting system could support families who make difficult ethical decisions – surgery with survival, trials of surgery and withdrawal if factors other than survival have been impacted, and quality comfort care.

Furthermore, we believe that our day to day clinical practice should change according to those practices that have an impact, not only upon mortality, but also on neurodevelopmental outcomes and patient and parent-determined quality of life. Research should be geared toward reducing those modifiable risk factors associated with the worst neurodevelopmental outcomes, such as hospital length of stay,⁹³  duration of cardiopulmonary bypass,⁹⁴  use of volatile anesthetics,⁹⁵  and birth weight or prematurity.^93,96  Lastly, we advocate for neurodevelopmental screening and referral to outpatient neurodevelopmental clinic in moderate to severe CHD populations to become standard practice following algorithms such as those described in the scientific statement by the American Heart Association.⁹⁷  Examples of children with moderate to severe CHD include: coarctation of the aorta, atrioventricular septal defect, ventricular septal defect with comorbidities, and tetralogy of Fallot. Those with severe or CHD of great complexity include: single ventricle, truncus arteriosus, and complex transposition of the great arteries.⁹⁷  This will allow timely detection and treatment when neurocognitive deficits are encountered, and thereby improve the quality of life in this population.⁹⁸ 

1. Detecting and mitigating health disparities in cardiology outcomes is indicated. (Class I, LOE B-NR)

2. Not hindering individual EOL and palliative decision-making is reasonable despite affecting mortality outcomes (Class IIa, LOE C-EO)

3. Standardizing neurodevelopmental screening and reporting in high-risk populations can be beneficial (Class IIa, LOE C-EO)

4. Standardizing reporting of functional capacity and long-term quality of life assessments is probably recommended (Class IIa, LOE C-EO)

5. Critical assessment of resource allocation that could impact quality of life is reasonable (eg, social support) (Class IIa, LOE C-EO)

Clinicians face a myriad of ethical considerations when taking care of the neonate with CHD. Ethical treatment of this vulnerable population and their families is grounded in respect for persons and honest communication. It also evolves over time as clinicians learn about the values of parents. As our therapies and treatment options continue to increase, we encourage continued reexamination on ethical best practice approaches with continued inquiry and debate. Please refer to Table 1 for a summary of vital points.

Future targets with gene therapies and precision medicine may again alter the landscape and may further change many of the ethical considerations presented herein.

CHD

congenital heart disease

LOE

Level of Evidence

SDM

shared decision-making