Although parents are typically the most appropriate decision-makers for their children, there are limits to this authority. Medical providers may be ethically obligated to seek state intervention against a parental decision if the parent places a child at significant and imminent risk of serious harm. When parents make medical decisions for their children, they assess both the projected benefits and risks of their choices for their family. These assessments are impacted by uncertainty, which is a common feature of neonatal intensive care. The relative presence or absence of uncertainty may impact perceptions of parental decisions and a medical provider’s decision to seek state intervention to overrule parents. In this article, we propose a model integrating prognostic uncertainty into pediatric decision-making that may aid providers in such assessments. We will demonstrate how to apply this model to 3 neonatal cases and propose that the presence of greater uncertainty ought to permit parents greater latitude to incorporate family values into their decision-making even if these decisions are contradictory to the recommendations of the medical team.

Medicine’s ground state is uncertainty. And wisdom- for both patients and doctors- is defined by how one copes with it.

Atul Gawande1 

One of the basic tenets of biomedical ethics is that good ethics begins with good facts. Although this statement is undoubtedly true, the intricate complexities inherent in some medical cases often are without a full complement of the “facts” essential for rendering a clear, ethically appropriate medical decision. Nowhere is this more on display than in the NICU, where predictions about the prognosis of an infant are at times imperfect and almost always fraught with uncertainty. Providers and parents alike must grapple with the challenge of deciding what is “right” for an infant when making difficult health care decisions, all of which is magnified when there is uncertainty.

Uncertainty exists in many forms, including uncertainty about a specific medical diagnosis, prognostic uncertainty, and uncertainty about a parent’s role in the medical decision-making team.2,3  Prognostic uncertainty is especially prevalent in the NICU given the wide range of outcomes for any particular neonatal disease process. Although outcome data exist for many of these diseases, applying these statistics to individual cases can be extremely difficult, in particular when predicting long-term outcomes for individual neonatal patients. This is further compounded by the challenge of explaining the possible range of outcomes in a meaningful way to parents, often in the midst of a high-acuity situation.4 

When there is significant uncertainty, parents must rely on family values that they find most important when making important and complex medical decisions. Sometimes, their choices may go against medical providers’ recommendations or differ from what providers would choose for their own child. Such cases can often lead to significant distress, especially when it is unclear whether the providers should challenge the parents’ decisions by seeking state intervention.

Infants lack the capacity to make medical decisions for themselves. Instead, surrogate decision-makers (parents in most cases) choose for them. Parents are typically best situated to weigh individual family values and relationships in health care decisions, all vital for determining what is in an infant’s best overall interest.57  This best interest standard asserts that parents must choose an option that maximizes the benefit to their infant in the context of their family. Additionally, parents share a special relationship with their children in which they have inextricably linked interests and will bear unique responsibility for the consequences of any such decisions made for their children.8,9 

There are limits to parental decision-making authority. When a parent’s decision places an infant at significant risk of serious, imminent harm compared to a recommended alternative, action must be taken against that decision.10,11  Many in pediatrics have adapted this harm principle as a threshold that, when crossed, ethically compels providers to seek this state intervention to provide care that is in opposition to parents’ requests.

Although the harm principle serves as a threshold for intervening against parental decisions, the best interest standard is often used as a way to conceptualize an optimal decision for an infant.12,13  Between the 2 extremes of an optimal decision and a decision that would cause clear harm exists a morally significant gap, sometimes referred to the zone of parental discretion (ZPD) or gray zone of decision-making (Fig 1).14,15  Within the ZPD, parents have the latitude to make difficult health care decisions on the basis of the values they deem most important, regardless of whether the medical team would make the same decision for the infant.

FIGURE 1

The ZPD.

To help address the struggle that providers encounter when viewing parental decision-making in the setting of prognostic uncertainty, we propose a new model that incorporates the concept of the ZPD with varying degrees of this uncertainty (Fig 2). The model reveals the ZPD as varying in size dependent on the level of prognostic uncertainty. On the left side of the model, there is less prognostic uncertainty and the ZPD is narrow, with the harm threshold existing close to the level of the best interest standard. As prognostic uncertainty increases toward the right side of the model, the ZPD widens, with the harm threshold existing further and further away from the level of the best interest standard.

FIGURE 2

Uncertainty and the ZPD: an integrated model.

FIGURE 2

Uncertainty and the ZPD: an integrated model.

Close modal

Cases exist within neonatal medicine that possess different levels of prognostic uncertainty, ranging from low to high. In cases in which greater uncertainty exists, we propose that the ZPD should be bigger, specifically impacting when the harm threshold is crossed, and permitting more latitude for parental decision-making. In these cases, parental discretion and assessment of their own values ought to play a more significant role in deciding the best course of action for their infant. Conversely, when relatively little prognostic uncertainty exists, the ZPD ought to be narrower, with a higher harm threshold. This model allows providers to more easily determine when the harm threshold is crossed whereby treatment becomes either mandatory or medically unreasonable.

We present 3 neonatal cases that have varying degrees of uncertainty, along with brief discussions of how to apply our model in each of them. Each case is real, with details changed to protect anonymity.

A female infant was prenatally diagnosed with a critical pulmonary valve stenosis by fetal echocardiography. Her parents were informed that given the degree of pulmonary stenosis and relative right ventricular hypoplasia, their infant would likely have a single functioning left ventricle and require several cardiac surgeries within the first several years of life. After many conversations and consideration of the medical information they received, the family chose to pursue palliative measures at the time of birth, forgoing all medical and/or surgical interventions. When the infant was born, however, she was noted to be clinically well appearing. A postnatal echocardiogram revealed isolated pulmonary valve stenosis and a normal-appearing and functioning right ventricle. A single catheter-based procedure could correct this condition. The medical team described this to the parents, along with the expectation of an excellent long-term outcome for their daughter. Despite this information, the family continued to desire palliative measures only and declined the proposed procedure.

The long-term outcomes for infants with isolated pulmonary valve stenosis are well known and more predictable, with low expected mortality and an 80% probability of not requiring any reintervention by 20 years of age.16  This case would fall to the left side of the model where uncertainty is low. As such, the ZPD should be narrowed and parental refusal of the recommended catheter-based intervention crosses the harm threshold, placing the infant at risk for serious, preventable, and imminent harm.

A term male infant was born after an uncomplicated pregnancy. During the delivery, the fetal heart rate dropped precipitously and, at birth, the infant had no detectable heart rate. He underwent extensive delivery room resuscitation, including endotracheal intubation, chest compressions, and administration of epinephrine. He recovered a heart rate by 8 minutes of life and was transferred to the NICU, where he underwent 3 days of therapeutic hypothermia, along with other standard treatments.

After the first few days of life, he was extubated and weaned off respiratory support and supplemental oxygen. An MRI revealed global hypoxic-ischemic injury to the brain. Although these MRI findings are associated with severe neurodevelopmental impairment or mortality at 18 months of age, his clinical findings suggested a much milder impairment at the same age.17  The medical team met with the infant’s parents and communicated that given this disconnect, there was a wide range of possible long-term outcomes, ranging from severe intellectual and physical disabilities to more mild impairments. After careful consideration, the family chose to withdraw hydration and nutrition.

In this case, there is a great deal of uncertainty about the future of this child. It is impossible to accurately predict what his life will look like in the long-term or how that will affect his family. His uncertain neurodevelopmental prognosis includes a wide range of possibilities, from mild cognitive impairment to profound intellectual disability and cerebral palsy. Different families will assess his quality of life across this wide range as more or less acceptable to them. Therefore, this child falls on the right side of the model where there is the greatest degree of prognostic uncertainty. Here, parents should be given wider latitude to make decisions for their child on the basis of the values they deem most important, regardless of whether they are in agreement with what the medical team would choose.

A mother presented to labor and delivery triage at 26 weeks’ gestation in preterm labor. The obstetric team predicted that she would deliver in the next day and asked the neonatologist to provide a prenatal consult. The neonatologist reviewed the expected course for an infant born at this gestational age, including an expected survival of >80% and survival without major morbidity of ∼75% with active treatment.18  They also reviewed the possible complications for an infant born at this age to include necrotizing enterocolitis, bronchopulmonary dysplasia, retinopathy of prematurity, intraventricular hemorrhage, and a range of long-term neurodevelopmental impairments. After a period of discussion, the couple requested that the medical team provide comfort measures only, citing concern for their infant’s potential for long-term suffering.

Although some facts are known in this last case, there still exists a moderate amount of prognostic uncertainty regarding the expected outcome for this infant. An infant at this gestational age does have a high chance of survival, but it is far from certain and does not account for other complications that may develop after birth.19  Given the intermediate level of uncertainty, this case would fall in the middle of the model where the ZPD ought to be slightly restricted, although parents should still be afforded a degree of latitude to incorporate their values into decision-making for their infant. The medical team deemed that the parental decision to forgo any resuscitation crossed the threshold for harm and felt it would be ethically unacceptable to follow their wishes. The medical team did, however, feel justified in restricting therapies within the NICU that had less clear benefit and cause pain, such as pursuing a surgical intervention if the infant developed necrotizing enterocolitis, because this would be consistent with the parental desire to limit their child’s suffering.

Although medical teams should always strive to understand parental values that are important in driving the medical decisions they make for their infants, they also bear the responsibility of recognizing when parents’ decisions would place their infant at significant risk of serious and imminent harm when compared to an alternative. The harm threshold coupled with the concept of the ZPD gives guidance to those making these assessments but falls short in cases in which the prognosis is unknown or uncertain. By adjusting the limits of the ZPD and the level of the harm threshold on the basis of the degree of prognostic uncertainty, our proposed model acknowledges this uncertainty and incorporates it into a more robust adaptation of the ZPD.

A significant limitation in our model is that it is often challenging to truly assess how much uncertainty exists in a particular case. We recognize that determination can be highly subjective. Providers using the model may have difficulty determining how best to apply limits to parental decision-making if they are unclear about the level of prognostic uncertainty. In future work, researchers should look at how to better assess the degree of uncertainty and how that assessment fits within our model.

By adjusting the limits of the ZPD on the basis of the degree of prognostic uncertainty, our proposed model acknowledges this uncertainty and incorporates it into a more robust adaptation of the ZPD. With a more nuanced understanding of how uncertainty affects this concept, those caring for infants may be better equipped to seek intervention against a decision when necessary but also feel more comfortable accepting parents’ decisions that differ from what they would professionally recommend.

We thank Dr Jennifer Kett for her review of the article.

Dr Krick conceptualized the manuscript, drafted the initial manuscript, and reviewed and revised the manuscript; Drs Hogue, Reese, and Studer reviewed and revised the original conference abstract and presentation and reviewed and revised the final manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

The views expressed in this presentation are those of the authors and do not reflect the official policy or position of Madigan Army Medical Center, the US Department of the Army, the US Department of Defense, or the US Government.

FUNDING: No external funding.

ZPD

zone of parental discretion

1
Gawande
A
.
Complications: A Surgeon’s Notes on an Imperfect Science
, 1st ed.
New York, NY
:
Metropolitan Books
;
2002
2
Han
PKJ
,
Klein
WMP
,
Arora
NK
.
Varieties of uncertainty in health care: a conceptual taxonomy
.
Med Decis Making
.
2011
;
31
(
6
):
828
838
3
Mishel
MH
.
The measurement of uncertainty in illness
.
Nurs Res
.
1981
;
30
(
5
):
258
263
4
Green
J
,
Darbyshire
P
,
Adams
A
,
Jackson
D
.
Balancing hope with reality: how neonatal nurses manage the uncertainty of caring for extremely premature babies
.
J Clin Nurs
.
2015
;
24
(
17–18
):
2410
2418
5
Kopelman
LM
.
The best-interests standard as threshold, ideal, and standard of reasonableness
.
J Med Philos
.
1997
;
22
(
3
):
271
289
6
Kopelman
LM
.
The best interests standard for incompetent or incapacitated persons of all ages
.
J Law Med Ethics
.
2007
;
35
(
1
):
187
196
7
American Academy of Pediatrics Committee on Bioethics
.
Ethics and the care of critically ill infants and children
.
Pediatrics
.
1996
;
98
(
1
):
149
152
8
Walter
JK
,
Ross
LF
.
Relational autonomy: moving beyond the limits of isolated individualism
.
Pediatrics
.
2014
;
133
(
suppl 1
):
S16
S23
9
Wightman
A
,
Kett
J
,
Campelia
G
,
Wilfond
BS
.
The relational potential standard: rethinking the ethical justification for life-sustaining treatment for children with profound cognitive disabilities
.
Hastings Cent Rep
.
2019
;
49
(
3
):
18
25
10
Diekema
DS
.
Parental refusals of medical treatment: the harm principle as threshold for state intervention
.
Theor Med Bioeth
.
2004
;
25
(
4
):
243
264
11
Diekema
DS
.
Revisiting the best interest standard: uses and misuses
.
J Clin Ethics
.
2011
;
22
(
2
):
128
133
12
Kopelman
LM
. Using the best interests standard in treatment decisions for young children. In:
Miller
G
, ed.
Pediatric Bioethics
.
Cambridge, MA
:
Cambridge University Press
;
2010
:
22
37
13
Salter
EK
.
Deciding for a child: a comprehensive analysis of the best interest standard
.
Theor Med Bioeth
.
2012
;
33
(
3
):
179
198
14
Gillam
L
.
The zone of parental discretion: an ethical tool for dealing with disagreement between parents and doctors about medical treatment for a child
.
Clin Ethics
.
2016
;
11
(
1
):
1
8
15
McDougall
R
,
Gillam
L
,
Gold
H
. The zone of parental discretion. In:
Rosalind McDougall
R
,
Delany
C
,
Gillam
L
, eds.
When Doctors and Parents Disagree: Ethics, Paediatrics and the Zone of Parental Discretion
.
Sydney, Australia
:
Federation Press
;
2016
16
Devanagondi
R
,
Peck
D
,
Sagi
J
, et al
.
Long-term outcomes of balloon valvuloplasty for isolated pulmonary valve stenosis
.
Pediatr Cardiol
.
2017
;
38
(
2
):
247
254
17
Sabir
H
,
Cowan
FM
.
Prediction of outcome methods assessing short- and long-term outcome after therapeutic hypothermia
.
Semin Fetal Neonatal Med
.
2015
;
20
(
2
):
115
121
18
Rysavy
MA
,
Li
L
,
Bell
EF
, et al;
Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network
.
Between-hospital variation in treatment and outcomes in extremely preterm infants
.
N Engl J Med
.
2015
;
372
(
19
):
1801
1811
19
Glass
HC
,
Costarino
AT
,
Stayer
SA
,
Brett
CM
,
Cladis
F
,
Davis
PJ
.
Outcomes for extremely premature infants
.
Anesth Analg
.
2015
;
120
(
6
):
1337
1351

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.