Defining Neonatal Serious Illness

Abstract

Background:

One major challenge to the conduct of rigorous neonatal palliative care research is the lack of robust universally agreed upon definitions of key concepts central to pediatric and neonatal palliative care.

Objective:

We sought to define neonatal serious illness as a foundational concept for neonatal palliative care.

Design:

Survey study.

Setting/Subjects:

Practitioners in the United States with expertise in neonatal serious illness.

Measurements:

Participants ranked 15 components according to how important each would be to include in a conceptual definition of neonatal serious illness. Based on rankings and free text responses, a working definition was created and a follow-up survey was circulated. Participants then ranked the extent to which the proposed definition comprehensively defines neonatal serious illness. The definition was further refined based on responses to the second survey.

Results:

Eighty experts responded to our first survey. Definition components ranked as most important included “high risk of short term mortality” and “results in shortened lifespan.” Analysis of free text responses revealed additional components viewed as important. We developed the following conceptual definition: “Neonatal serious illness 1) carries a high risk of short term mortality OR lifelong medical complexity with probable shortened lifespan, 2) may involve substantial prognostic uncertainty (especially in regard to neurodevelopment) that complicates medical decision-making, and 3) significantly impacts the patient and family's life now or in the future with strain related to treatments and care.”

Conclusion:

We believe our definition of neonatal serious illness will facilitate future study essential to the advancement of care for this population.

Introduction

Because pediatric palliative care remains a relatively new field, palliative care research within the field of neonatology is similarly nascent.¹ One major challenge to the conduct of such research is the lack of robust universally agreed upon definitions of key concepts central to pediatric and neonatal palliative care. Such definitions exist in adult palliative care and have been critical to the advancement of rigorous research in that field.² Prior work has defined “serious illness” in adult populations as a condition that carries a high risk of mortality and either negatively impacts a person's daily function or quality of life or excessively strains their caregiver.^2–4

There is not yet a robust conceptual definition of neonatal serious illness. There are three broad categories of Neonatal Intensive Care Unit (NICU) patients who may require or benefit from palliative care: (1) extremely preterm infants, (2) those with life-limiting anatomic or developmental differences or anomalies, and (3) those with genetic abnormalities.⁵ Many of these patients are now diagnosed prenatally, which may allow for early palliative care engagement.⁶

Prior literature has sought to define what constitutes palliative care in the neonatal population, and to begin to establish which babies in the NICU might benefit from palliative care consultation.^7,8 Authors point to a need for provision of palliative care services for seriously ill neonates, but it is not well established which neonates qualify as having serious illness.^9,10 We, therefore, sought to define neonatal serious illness as a foundational concept for neonatal palliative care, which will be critical to the conduct of future research involving seriously ill neonates and to the provision of high-quality care.

Methods

Overview

To begin developing a definition of neonatal serious illness, we created and circulated a survey to neonatologists and those with expertise in pediatric palliative care. This study was deemed exempt human subjects research by the Icahn School of Medicine at Mount Sinai Institutional Review Board.

Sample

To identify potential participants, e-mails were sent using snowball sampling.¹¹ This was achieved by the authors sending e-mails to individuals with relevant expertise and then asking that each recipient forward the survey to five other individuals who would then forward to five others and so on. To ensure a diversity of responses, our survey was then circulated using e-mail lists of neonatologists and experts in pediatric palliative care.

Measures

Our survey was developed by experts in neonatology, palliative care, and defining serious illness. Survey respondents were asked to appraise the suitability of the adult definition of serious illness for neonatal populations and to suggest modifications, if indicated. Participants were also asked to rank a list of 15 definition components (such as “carries poor prognosis” or “creates caregiver stress”) according to how important each would be to include in a definition of neonatal serious illness (with 1 being most important and 15 being least important). Components were generated from the adult definition or were produced by the authors.

Participants were also asked to discuss how they would define neonatal serious illness in free text responses. Finally, we asked respondents whether they believed the definition of neonatal serious illness for the purposes of research would be different from the definition for the purposes of clinical care. Once the proposed definition was developed (see Results section), it was sent with a two-question follow-up survey to a subset of participants who had expressed willingness to comment on additional iterations of the definition.

The first question asked participants to rank, using a 5-point Likert scale, the extent to which they felt the proposed definition comprehensively defines neonatal serious illness, with an optional free text response to suggest changes. The second question asked whether participants felt parents would define neonatal serious illness differently than clinicians and to explain any potential differences expected. Because this study did not include parents' perspectives, we asked this question to begin to explore whether providers anticipate that parents may define neonatal serious illness differently. This question does not substitute for parent perspectives but instead may serve as a bridge for future research in which such perspectives can be more fully explored.

Analysis

Using responses to the first survey, we reviewed component rankings and free text responses to generate a working definition of neonatal serious illness. Highly ranked components were prioritized for inclusion and free text responses were analyzed for common themes that were also included. The authors revised, refined, and then circulated the proposed definition as part of the second survey. The definition was then further refined based on responses to the second survey.

Results

Eighty physicians responded to our first survey, representing a range of disciplines, including neonatology (65%) and pediatric palliative care (41.2%) (Table 1). Respondents were primarily from academic institutions (93.7%) and represented a broad range of ages, professional experience, and geographic locations across the United States. The majority of participants (76%) reported spending <25% of their time conducting research. Definition components with the lowest median scores (meaning participants found them to be most important) included “high risk of short term mortality,” “results in shortened lifespan,” and “negatively impacts quality of life now or in the future” (Table 2).

Table 1.

Demographics and Participant Characteristics

	n (%)
Sex, female (N = 80)	46 (57.5)
Age, years (N = 80)
25–34	4 (5)
36–44	25 (31.2)
45–54	23 (28.8)
55–64	11 (13.8)
>65	17 (21.2)
Geographic location in United States (N = 79)
Northeast	27 (34.2)
Southeast	8 (10.1)
Midwest	24 (30.2)
West	15 (19)
Southwest	5 (6.3)
Primary specialty (N = 80)
NICU	52 (65)
Pediatric palliative care	33 (41.2)
General pediatrics	8 (10)
Pediatric critical care	6 (7.5)
Pediatric hematology oncology	2 (2.5)
Pediatric emergency medicine	1 (1.25)
Adult medicine	6 (7.5)
Setting of primary institution (N = 80)
Urban	58 (78.5)
Suburban	18 (22.5)
Rural	2 (2.5)
Other	2 (2.5)
Practice type (N = 79)
Academic	74 (93.7)
Community hospital/practice	6.3 (5)
Years in primary subspecialty (N = 80)
0–10	28 (35)
11–20	12 (26.3)
21–30	14 (17.5)
>30	17 (21.2)
Percent time providing clinical care (N = 80)
None	1 (1.2)
1–25%	9 (11.2)
26–50%	20 (25)
51–75%	27 (33.7)
76–100%	23 (28.7)
Percent time doing research (N = 80)
None	15 (18.7)
1–25%	46 (57.5)
26–50%	13 (16.2)
51–75%	6 (7.5)
76–100%	0 (0)

NICU, Neonatal Intensive Care Unit.

Table 2.

Participant Ranking of Importance of Potential Definition Components

Component	Median	IQR	Min	Max	% responded
High risk of short-term mortality	1	1	1	14	92.5% (74/80)
Results in shortened lifespan	3	3	1	15	83.8% (67/80)
Negatively impacts QOL now or in future	4	2	1	13	95.0% (76/80)
Negatively impacts daily function now or in future	5	3	1	13	88.8% (71/80)
Associated with poor neurodevelopmental outcome	5	6	1	13	90.0% (72/80)
Cannot be “cured”	6	11	1	14	71.3% (57/80)
Has/will have burdensome symptoms	6	3	1	14	95.0% (76/80)
Has/will have burdensome treatments	7	3	1	13	90.0% (72/80)
Associated with future lack of independence	9	5	2	15	82.5% (66/80)
Associated with future need of long-term care	9	4	2	15	88.8% (71/80)
Associated with future technology dependence	9	8	3	14	85.0% (68/80)
Creates caregiver stress	10	4	3	14	83.8% (67/80)
Negatively impacts family QOL	10	4	3	15	82.5% (66/80)
Creates stress for other family members	11	3	4	14	81.3% (65/80)
Other	15	1	1	15	52.5% (42/80)

1 = most important, 15 = least important.

IQR, interquartile range; QOL, quality of life.

Analysis of free text responses revealed that, although not captured in component rankings, “prognostic uncertainty,” “strain related to treatments,” and “impact on family” were frequently viewed as important (Table 3). Based on the responses to both surveys, we developed the following conceptual definition: “Neonatal serious illness 1) carries a high risk of short term mortality OR lifelong medical complexity with probable shortened lifespan, 2) may involve substantial prognostic uncertainty (especially in regard to neurodevelopment) that complicates medical decision-making, and 3) significantly impacts the patient and family's life now or in the future with strain related to treatments and care.”

Table 3.

Themes Present in Free Text Responses

Theme	Example
High risk of death, shortened life expectancy	A neonate with a “serious illness” has a condition where there is a high risk of death or severe disability
High risk morbidity	Carries a high risk of … future morbidity secondary to an underlying medical cause
High risk disability	High risk of chronic physical disability
Impacts development, poor neurodevelopmental outcomes	An illness that threatens the life or developmental future of the child
Impacts QOL	Could negatively impact QOL and daily function
Complex medical decision making	Need for complex medical decision making
Likely result in chronic illness, medical complexity, and high utilization of medical services	High risk of … lifelong medical complexity and/or need for technologies to survive
Technology dependence	Developmental outcomes that would render the infant dependent on technology for essential life functions for potentially the entirety of their life
Future lack of independence	Limiting the child's potential to gain functional independence as they age
Prognostic uncertainty	Given an infant's early stage in development, it may be accompanied by poor neurological outcomes and with high prognostic uncertainty
Burdensome symptoms and treatments for patient	Burdensome lifelong symptoms and treatments
Impact on family	Whose conditions may create overwhelming burdens, stress, and distress on caregivers and family units
Strain related to treatments for caregivers	Significant care burden for families and caretakers
Results in lack of independence	Limiting the child's potential to gain functional independence as they age

Half of the respondents to our first survey (43/80) felt that the definition of neonatal serious illness should be the same in both clinical and research contexts. Most commonly, this was for reasons such as “Research exists to advance clinical care; they are inherently intertwined. Therefore, I think the definitions should be the same, and encompass both quantitative and qualitative descriptions” and “should be the same to be meaningful clinically.” Twenty-five percent of participants felt that the definitions should be different.

One participant stated “For research purposes, the definition of neonatal serious illness will likely include binary conditions/outcomes (i.e., will this child be blind? yes/no) … families are often more concerned about their child's capacity to feel and experience life …. These questions are not currently covered in most research but are often very important when discussing serious illness in the clinical setting.” Another said, “I think for the purposes of research it would be helpful to define serious illness using a criteria …. For clinical care, the definition is a little more broad and includes situations where there is significant family distress about the illness even when perhaps the patient would not meet the [proposed definition].” A substantial minority (21%) were unsure or did not answer the question.

The overwhelming majority of respondents to the second survey (87%) agreed or strongly agreed that the proposed definition comprehensively defines neonatal serious illness. Free text responses collected along with our second survey suggested minimal edits to our proposed definition. In the few comments provided, respondents suggested edits to wording surrounding quality of life, with a focus on avoiding language that might be perceived as judgmental or subjective (such as “negatively impacts quality of life” instead of “impacts quality of life.”)

A large proportion of respondents to our second survey (70%) felt that the definition of neonatal serious illness would be different or somewhat different for parents when compared with clinicians and researchers. There was not a uniform perception of how parents might define neonatal serious illness and some participants expressed uncertainty as to how parents might define it. One respondent stated “The definition may be similar, but the interpretation of the terms may differ. How doctors perceive quality of life, in particular, may differ depending on family values and beliefs.”

Discussion

Based on an iterative process involving content experts in neonatology and pediatric palliative care, we propose a novel conceptual definition of neonatal serious illness. Our findings echo previous literature that serious illness differs in important ways for neonates when compared to adults. Notably, pediatric patients, and especially neonates, may survive for months to decades with serious illness, whereas their adult counterparts often have shorter lifespans once serious illness develops.¹ Our definition, therefore, departs from the adult definition by adding “lifelong medical complexity” as a marker of serious illness in neonates. In addition, prognostic uncertainty is extremely common in seriously ill neonates and was commonly expressed in free text responses.¹² Our definition, therefore, includes such uncertainty to capture all potentially seriously ill newborns and to highlight that a patient does not have to be dying imminently to qualify as seriously ill. Instead, there is often uncertainty about likely outcomes that can be especially challenging for both parents and the medical team. Finally, parents and family play a unique role in the context of neonatal illness. Although family is also unquestionably impacted by serious illness in adults, the presence of a child who is sick or who lives with medical complexity has meaningfully different implications for family. Although the importance of impact on family was not reflected in participants' ranking of components, this theme came up frequently in free text responses. In addition, the literature supports a unique and profound impact of serious illness on the families of pediatric patients.¹³ Our definition, therefore, includes a focus on family to account for this key component.

Because family is so important in the context of pediatric serious illness, it must be acknowledged that the views of parents are missing from this study. This is especially notable given that a large majority of respondents believe parents will likely define serious illness differently from clinicians, and many endorsed uncertainty surrounding the definition parents might utilize. In addition, parents are eager to be included in research designed to advance the field. Involving parents results in more useful, more comprehensive, and more inclusive research studies.^14,15 It will, therefore, be critical to conduct additional studies that both integrate their voices and compare parent and provider perspectives to better understand neonatal serious illness.

We sought to create a conceptual definition that can be operationalized with future study. Our goal was to produce a definition that was “sensitive” to make it useful for research and clinical applications. As a result, we may have sacrificed a degree of “specificity.” We accept this trade-off to produce a definition less likely to omit neonates who should be reflected in future research and who might benefit from services offered to those with serious illness, such as palliative care consultation or follow-up with complex care. Prior work defining serious illness in adult populations has similarly focused on sensitivity in the early conceptual phases of development. Based on such studies, we anticipate being able to work toward improved specificity in future studies aimed at operationalizing this definition.¹⁶

There is a growing need for palliative care services in the NICU. With more infants surviving critical illness, conversations often require nuanced discussion of prognosis and quality of life.¹⁰ Babies who would previously have died early in their NICU course may now experience extended hospitalization before death.¹⁷ Such scenarios require strong primary palliative care skills. However, to effectively provide palliative care in the NICU, it is essential to first identify which infants are seriously ill. Provision of palliative care services alone is not enough to ensure that seriously ill infants receive compassionate care that aligns with their families' goals. Such care must be provided early enough to have impact. Successfully identifying which infants are seriously ill is necessary for timely provision of palliative care. Prior work has demonstrated substantial benefit of early palliative care in non-NICU populations, yet much of neonatal palliative care to date has focused only on care for infants near death.⁹ This has led to significant underutilization of palliative care services in the NICU. Similar trends have previously been observed in adult surgical populations with the advent of clear population definitions offered as a potential solution.¹⁶

Historically, there has been some resistance to engagement of palliative care among intensivists, including in the NICU.¹⁸ Such resistance needs to be further characterized and acknowledged to work toward improved utilization of palliative care in the NICU. This challenge may, in part, explain our relatively low response rate. However, our small sample size may also relate to the relative paucity of physicians with expertise in neonatal palliative care. Along with successfully identifying patients in need of palliative care, addressing this resistance will be critical to improving care for seriously ill neonates. In addition, to successfully integrate palliative care in the NICU, clinicians should become familiar with the resources and guidelines which currently exist, upon which we hope to build with this study.^4,19–21

Given that the Institute of Medicine has recommended that all providers who care for patients with advanced serious illness maintain at least basic palliative care skills, additional training and resources to work toward that goal would undoubtedly be justified.^10,22 Defining serious illness in this and other pediatric populations will be essential to answering this call. We hope that this study presents a first step toward identifying this population to provide goal-concordant care and optimize outcomes for parents and babies.

Our study has important limitations. First, snowball sampling is an imperfect method that can result in a nonrepresentative sample. We utilized snowball sampling in the hope of accessing individuals with knowledge of and/or interest in neonatal palliative care. With this access came potential limitations. We attempted to mitigate sample-related challenges by ensuring that e-mails were sent to a geographically and institutionally diverse cohort. Based on the composition of the study cohort (Table 1), these efforts were relatively successful. In addition, to further ensure a representative cohort, we circulated our survey through e-mail listservs to further enrich our cohort. Although we are confident in the methods used in this study to produce a definition through an iterative process, our results reflect the views and attitudes of 80 individuals. Additional perspectives may be useful, including from those practicing outside of academic institutions. Our findings come exclusively from physicians with content expertise and are not representative of parental perspectives. Similarly, the views of nurses, social workers, and other members of the interdisciplinary team are not reflected in this study and may differ in important ways from those of physicians. The responses in this study also reflect the challenge of definitions designed for research versus clinical care. We hope that our definition may be useful in both contexts, given that a research definition that lacks relevance in a clinical context is likely of very limited utility. With heavy representation from nonresearchers, it will be important to conduct additional studies to operationalize this definition for the purposes of research. Although we are optimistic about the utility of this definition, we do not yet know if it can be successfully operationalized.

We hope that the definition of neonatal serious illness will facilitate future study essential to the advancement of care for this population. We intend to incorporate parental perspectives as we work toward operationalizing this definition for future study to better serve this understudied population.

Authors' Contributions

Dr. K.G. conceived of the study, acquired and interpreted data, drafted the original article, and approved the final version of the article. A.W. and R.B. made substantial contributions to the conception of this study, analyzed the data, critically revised the article, and approved the final version of the article. A.K. made substantial contributions to the conception of this study, critically revised the article, and approved the final version of the article. All authors agree to be accountable for all aspects of this study in ensuring that questions related to the accuracy or integrity of any part of the study are appropriately investigated and resolved.

Funding Information

A.K.'s work is funded by NIA K24AG062785.

Author Disclosure Statement

No competing financial interests exist.