Conditions and diseases

Not all children who die receive palliative care services, but national profiles of children who die provide the strongest epidemiological evidence to characterize the conditions and diseases of children receiving pediatric palliative care. Examining mortality data in the United States between 1999 and 2006 (Fig. 2-2), one notices immediately that infant deaths are due mostly to specific perinatal conditions, such as premature birth and congenital syndromes or chromosomal disorders, whereas older children are most likely to die from external causes, such as traumatic injury.

Fig. 2-2 Major causes of pediatric death in the United States, 1999-2006.

(Data from National Center for Health Statistics.)

We can examine the number of children who die from complex chronic conditions (CCCs) as one statistic that can inform the field of pediatric palliative and hospice care by defining it as:

For children with CCCs, congenital and chromosomal abnormalities were associated with 5819 (20.3% of 28,527 total) infant and 1087 (4.4%) child deaths, diseases of the nervous system caused 373 (1.3%) infant and 1173 (4.8%) child deaths. Deaths from cancer, which is perhaps the condition most strongly associated with the imagery of palliative care in mind of the public and much of the medical profession, accounted for only 141 (0.5%) infant deaths and 2150 (8.9%) deaths past infancy. A few studies of patients enrolled in pediatric palliative care programs support the evidence of diverse conditions in this population. In a recent study of 8 pediatric palliative care programs in Canada, patients had as their primary diagnosis nervous system disorders (39.1%), cancer (22.1%), or perinatal-onset conditions or congenital anomalies (22.1%).¹

Trends and age-specific patterns of mortality

Over the past century, child mortality has been declining for both children suffering from trauma and from complex chronic conditions. Examining U.S. mortality data from 1979 to 2006, the rates of death attributed to CCCs as well as injury, sudden infant death syndrome, and other causes declined substantially (Fig. 2-3), with the corresponding numbers of deaths also declining but less so since 2000 (Fig. 2-4).

Fig. 2-3 Trends in the rates of pediatric death by causes, 1979-2006.

(Data from National Center for Health Statistics.)

Fig. 2-4 Trends in the number of pediatric death by causes, 1979-2006.

(Data from National Center for Health Statistics.)

One published study focused on the interval between 1979 and 1997, when the annual death rate due to non-cancer and cancer-related CCCs declined for almost every age group (from a 7.1% decline for infants with cancer to a 49.9% decline for 1 to 4 year olds dying from non-cancer CCCs). The one exception to this decline was the mortality attributed to non-cancer CCCs among 20 to 24 year olds for whom the rate of death actually increased by 11.6% over this 18-year interval, which may be due to the longer lifespan of younger children with CCCs dying at later ages.² For this reason, studies of pediatric deaths and palliative care need to include the experiences of young adults, perhaps into their 20s or 30s, who die from conditions with congenital or childhood-onset.

Symptoms and quality of life

The management of physical and psychological symptoms is of utmost importance in palliative care. Most studies published are on bereaved parents of children with cancer. In one such study, 96 bereaved Australian parents of children with cancer reported that, in the last month of life, 84% of children suffered from at least one physical symptom (pain, fatigue, poor appetite, constipation, dyspnea, nausea/vomiting, diarrhea, and seizures), with pain, fatigue, and poor appetite the most frequent.³ Nearly half (43%) of children suffered from three or more physical symptoms. Parents also reported that 42% of children had been “more than a little sad,” 38% experienced “little or no fun,” and 21% were “often afraid.” In a Swedish study of 449 bereaved parents of children with cancer, physical fatigue was the most frequently reported symptom (86%) to have a higher or moderate impact on the child’s well-being, with reduced mobility (76%), pain (73%), and decreased appetite (71%) also major concerns.⁴ In this study, parents were more likely to report anxiety in children older than 9 years of age than in younger children (relative risk [RR] 1.8, 95% confidence interval [CI]: 1.2-2.6) and 16 years of age (RR 2.0, 1.3-2.9). Children also suffered from difficulties in swallowing, depression, reduced mobility, impaired speech, swelling, disturbed sleep due to anxiety, and urinary problems. Problems with breathing during the last day of life, and with loss of motor function in the last week of life, were also reported by 28 bereaved parents whose children had been enrolled in a pediatric hospice program in St. Louis.⁵ Another study of 65 parents of 52 children reported similar symptoms, which did not differ according to patient gender, disease, or location of death, such as the ICU, elsewhere in hospital, or home.⁶ A study of bereaved parents of 48 children who died from cancer also cited that their children suffered from anxiety, which was not treated effectively.⁷ Given the prevalence of symptoms, and the nature of the child’s condition, it is challenging to measure quality of life for children receiving palliative care services, especially using generalized instruments that may not be suitable for this population of children.⁸ Nevertheless, much remains to be done in improving the treatment of symptoms in this population.

Pharmacologic treatments and other interventions

Medications are one of the mainstays of modern medicine. Polypharmacy is usually considered a problematic phenomenon among geriatric patients, but children admitted to children’s hospitals are typically (median count) exposed to five different medications during their hospitalization, with 10% of the patients exposed to 14 or more medications. By the time pediatric patients have been hospitalized for a month or more, the 10% receiving the most medication have been exposed to 40 or more medications (unpublished data).

The drugs used to treat children who are receiving palliative or hospice care have not yet been well described. Among pediatric patients who died while hospitalized, the most common exposures on the last day of life are to epinephrine (45%), dopamine (45%), fentanyl (42%), midazolam (33%), morphine (30%), vancomycin (29%), and ranitidine (26%); examining only oncology patients, the list and proportions are essentially the same (unpublished data). In a cohort of 834 pediatric patients enrolled in hospices across the United States, the most commonly used class of drug was opioids (72% of all patients), followed by hypnotics (60%), laxatives and other medications for the bowels (50%), antacids and other medications for the stomach (41%), and anti-epileptic drugs (38%) (unpublished data).

The use of specific medications appears to vary substantially across hospitals and hospices. For instance, in a study of 1466 pediatric oncology patients who died in children’s hospitals, the proportion of patients who received opioids daily during their last week ranged across the hospitals from zero to 90.5%, a variation in practices that did not diminish substantially even after adjustment for patient characteristics (Fig. 2-5).

Fig. 2-5 Wide variation in opioid daily receipt by pediatric oncology patients who died in children’s hospitals.

(Data abstracted from Orsey AD, Belasco JB, Ellenberg JH, Schmitz KH, Feudtner C. Variation in receipt of opioids by pediatric oncology patients who died in children’s hospitals. Pediatr Blood Cancer. 2009; 52(7): 749-50.)

Medications are only one of many possible interventions used in palliative or hospice care or end of life situations to ameliorate symptoms or improve the quality of life. Much less is known about the epidemiology of other interventions, such as a child’s exposure to complementary and alternative medical (CAM) practices. We do know that in one study of 92 parents of children with cancer in the United States, 49% reported that their children used at least one type of CAM and 20% used herbal remedies, homeopathy, or vitamins in the two months preceding the survey,⁹ which suggests if any of these children died, some likely continued to receive CAM treatments at the end of life.

Family structure

Families can have many different structures, and families of children with complex chronic conditions are no different. In two-parent married households, research has reported mixed results as to the effects on the marriage. Some studies, such as a case-control study of marital quality in mothers and fathers parenting children with spina bifida, report no difference in marital distress than in parents of healthy children.¹¹ Another study on the same disease reports marital breakdown.¹² Other studies, such as one on functioning in parents of children with cancer, suggest that while marital difficulties may occur, positive benefits such as feeling closer as a family and supporting each other may offset the stresses.¹³ When parents are divorced, communication difficulties may exist, especially when the nonresidential parent has a different opinion about the child’s care than that of the custodial parent.¹⁴ Single-parent households have been less frequently studied,¹⁵ but some evidence suggests that single mothers of children with cystic fibrosis may have higher stress-related symptoms and that their children may have higher hospitalization rates;¹⁶ another study on handicapped children suggests that single mothers have greater difficulty with finances, but also were more flexible in adapting to new family roles and arrangements.¹⁷ Step-parents in the home also alter the family structure.¹⁸

Siblings add another dimension. While the national percentage of children with complex chronic conditions who have siblings is unknown, meta-analyses investigating the effects of childhood chronic illness on siblings suggests a modest, negative effect on psychological functioning, peer activities, and cognitive development, with lower scores reported by parents than by the siblings themselves.^19,20 Recommendations for siblings include providing information about their brother’s or sister’s illness early on, group discussions, and informing school personnel.²¹

Another aspect of families that we can consider concerns other social and cultural characteristics. Various studies indicate that racial and ethnic minorities may have difficulties with trust and communication with the medical profession, poverty, and different belief systems and family structures.^22–25 Religion and spirituality are very important to families, but their impact is not uniform nor prescriptive.^26,27

The work of care

The work of care consists of the effort, labor, and various tasks in which parents and other family members engage in to care for the child with a complex chronic condition and for the family system. Daily tasks include administering medications, feeding and bathing, providing emotional support, and monitoring any changes in the child’s health.^28–32 Parents also are responsible for managing their child’s medical care by arranging and bringing their child to appointments and filling out copious amounts of paperwork,³³ while creating an environment where their child can have as normal a life as possible.³⁴ These efforts can occupy a great deal of time.³⁵ One study of parents whose children had gastrostomy tubes, compared to parents whose children did not, found parents in the first group devoted more than 7 hours a day to technical and non-technical care of the child compared to 3 hours in the second group (Fig. 2-6).³⁶ How this work is organized, distributed, and completed likely varies across family structures, and some evidence also exists that stronger family function is related to better execution of disease management activities.^37–39

Fig. 2-6 Number of hours per day devoted by parents to caring for children with gastrostomy tubes.

(Data abstracted from Heyman MB, Harmatz P, Acree M, Wilson L, Moskowitz JT, Ferrando S, et al. Economic and psychologic costs for maternal caregivers of gastrostomy dependent children. J Pediatr. 2004;145(4):511-516.)

Coping

Parents and siblings have been observed to suffer from an array of psychological and physical forms of distress: parents report depression, grief, guilt, and anxiety as well as problems such as insomnia, headaches, and musculoskeletal pain. One study of parents of children with severe cardiac disease, compared to control parents of healthy children, found substantial decrements in the parents quality of life in general and specifically regarding emotions, physical health, and vitality (Fig. 2-7).⁴⁰ Siblings struggle with fears, traumatic reactions, isolation, difficulties in school and social settings, and the behavior problems that result.⁴¹

Fig. 2-7 Quality of life dimensions of parents of children with cardiac disease.

(Data abstracted from Arafa MA, Zaher SR, El-Dowaty AA, Moneeb DE. Quality of life among parents of children with heart disease. Health Qual Life Outcomes. 2008; 6:91.)

Few methodologically rigorous studies have examined the impact of a child’s death on family members or the family structure. Much of what is known about parental reactions to a child’s death is based on samples of those who present for psychological treatment; presumably, a much larger percentage of bereaved families experience essentially normal patterns of grief and other mental health consequences from the death of a child, without ever presenting for clinical care. Estimates of psychopathology in bereaved parents are therefore likely to be erroneously high.

Financial concerns

Related to both the work of care and coping is a family’s financial concerns. In particular, three areas of finances—insurance, paid work, and out-of-pocket expenses—are affected by caring for a child with a complex chronic condition. Due to the high costs of care, families may have difficulty finding or keeping insurance coverage, and may find that certain services are not covered or caps on spending prevent other services from being used.⁴² The result is that 38% of the U.S. population of parents of children with special health care needs are either uninsured or inadequately insured (Fig. 2-8),⁴³ a phenomenon that is especially pressing as the adolescents transition into adulthood.⁴⁴ Often in families caring for a child with a complex chronic condition, one family member, usually the mother, decides to stay at home, forsaking the income and benefits provided by paid work.^45,46 Out-of-pocket expenditures also are very high, and may further impoverish families.⁴⁷ Very poor families, however, may actually be at a slight advantage due to the infusion of additional resources and outside attention into the home.⁴⁸

Fig. 2-8 Levels of inadequate health insurance among children with special healthcare needs.

(Data abstracted from Newacheck PW, Houtrow AJ, Romm DL, Kuhlthau KA, BloomSR, Van Cleave JM, et al. The future of health insurance for children with special health care needs. Pediatrics. 2009;123(5):e940-947.)

Care in the home

Pediatric palliative and hospice home care is intended to increase the quality of life of patients and families, and some, if not most, families prefer being at home to care for the child.³⁰ From 1989 to 1993, only 12.3% of children with CCCs died in the home; in 1999 to 2003, 17.7% of children with CCCs died in the home. In the same period, hospital deaths declined and deaths at all other sites and care institutions remained comparably stable.⁴⁹ This pattern of a rising proportion of pediatric deaths attributed to CCCs occurring at home is traceable to 1979 (Fig. 2-9). Over the same time periods, stark differences in the proportion of death that occurred at home are evident and persistent among patients classified on their death certificates as white, Hispanic, or black (Fig. 2-10).⁴⁹

Fig. 2-9 Rising percentage of pediatric CCC-related deaths occurring at home, 1979-2002.

(Data from National Center for Health Statistics.)

Fig. 2-10 Differences in proportion of CCC-related deaths occurring at home by race and/or ethnicity.

(Data abstracted from Feudtner C, Feinstein JA, Satchell M, Zhao H, Kang TI. Shifting place of death among children with complex chronic conditions in the United States, 1989-2003. JAMA. 2007;297(24):2725-2732.)

Home care usually involves the partnering of skilled nurses with the parents, who are able to share knowledge and divide tasks related to the work of care.⁵⁰ Issues to be considered in home care include the work of care, impact on the family, the likelihood of adverse events, and cost.^30,51–54 In a study of parents of 48 children who died from cancer, while only 41% of parents provided palliative care in the home and 48% of children died at home, 88% of parents chose “at home” as the most appropriate location of death in hindsight.⁷

Hospice care

While in general, little is known about the hospice care services that dying children receive, these services are generally thought to be simultaneously useful but underused. Research has demonstrated rare differences in the use of hospice services, but whether these differences arise due to differential access, eligibility, or other reasons remains unknown. A study of Florida Medicaid pediatric patients who died revealed that only 11% received any hospice services, but those who did were more likely to be non-Hispanic whites than non-Hispanic black, have chronic conditions, and longer Medicaid enrollment than non-hospice users.⁵⁵ In a study of pediatric oncologists, physicians with access to hospice that accepted patients receiving chemotherapy had more patients die at home, but not all hospices admit patients actively receiving chemotherapy.⁵⁶ A broader survey of pediatric medical providers experienced in end-of-life care reported that 86% of responders thought hospice was beneficial, especially in the provision of non-medical services.⁵⁷

Hospitals

While there is a trend of increasing hospice and home-based pediatric palliative care, pediatric death still occurs mostly in the hospital setting. From 1989 to 2003, 80.1% of childhood deaths resulting from CCCs occurred in the hospital (down from 85.7% in 1989 to 2003), with considerable variation in the location of death by CCC category.⁴⁹ The same study also demonstrated that infants and racial and ethnic minorities, were more likely to die in the hospital than older children and non-minorities.

Hospitalizations of children with CCCs who subsequently died were characterized by long lengths of stay (Fig. 2-11), mechanical ventilation during their terminal admissions, and death in the intensive care unit. In a study that linked information from death certificates to hospital use records from 1990 to 1996 in Washington state, 458 infants with CCCs who died under a week of age spent 92% of their lives in the hospital, and 286 infants with CCCs who died during the second to twelve months of life spent 41% of all their lives in the hospital.⁵⁸ For children and young adults with CCCs, the median number of days spent in the hospital in the year preceding death was 18, and only a third of this group was hospitalized until the last week of life. Across all age groups, rates of hospital use increased closer to the child’s death. Half of hospitalized infants and 19% of children and young adults with CCCs were mechanically ventilated during their terminal admission. Studies from hospitals in a range of countries (the United States, Australia, and the Netherlands), suggest that although a number of deaths take place in the operating room and emergency departments, most pediatric deaths occur in the intensive care unit.^59–62 In a study of a United Kingdom children’s hospital, infants with congenital malformations and perinatal conditions were most likely to die in the ICU (OR 2.42, 95% CI 1.65 to 1.35) and older children with cancer were the most likely to die outside the ICU (OR 6.5, 95% CI 4.4 to 9.6).⁶³

Fig. 2-11 Length of hospital stay for pediatric patients discharged alive or who died, 2006.

(Data from KID 2006, Agency for Healthcare Quality and Research.)

Over the past decade, hospital-based palliative care services have begun in many pediatric hospitals across the country. An evaluation of the impact of one such interdisciplinary program in Boston documented, after the program’s establishment, an increased frequency of hospice discussions (76% vs. 54%, for an adjusted risk difference of 22%), earlier documentation of do-not-resuscitate orders (18 vs. 12 days), and fewer deaths in the intensive care unit or in other hospitals (decrease of 16%). Parents also reported that their child suffered less pain (adjusted risk difference of 19%), and that they felt more prepared during their child’s last month of life and at the death.⁶⁴ In another study on the impact of a pediatric palliative care program in Germany, 69% of families preferred to have their child at home for the death compared to 18% of families before the program was instituted.⁶⁵

Respite and other modalities of care

Beyond hospital and hospice care, other modes may also be useful for dying children. Broadening the definition of the population in need to encompass children with complex or special healthcare needs, respite care provides parents with time off from being in charge of the care. Respite care also has different meanings and purposes attributed to it by different families.⁶⁶ In a study among caregivers of children with cerebral palsy in Canada, 46% of caregivers reported using respite services in the past year, and 90% of caregivers indicated that respite was beneficial for both the child and family.⁶⁷ In another study of children with special healthcare needs, researchers demonstrated that a program in an emergency department setting to provide telephone advice and care coordination enhanced parental satisfaction with emergency care.⁶⁸

Programs

In general, pediatric palliative and hospice care programs are characterized as being interdisciplinary and focused holistically on both the child and the family, with important functions being pain management, coordination of care, and decision-making support. For example, the Seattle Pediatric Palliative Care Project enrolled 41 patients over a 2-year period (1999 to 2001) whose ages ranged from infancy to 22 years old, with 31 specific diagnoses in the population and 34% cancer diagnoses.⁶⁹ Another program in Florida, the Partners in Care: Together for Kids (PIC:TFK) program, which uses government subsidies to create partnerships between state-employed care coordinators and home and community-based hospice staff, enrolled 468 children in the 3-year period between 2005 and 2008.⁷⁰ The PIC:TFK program was intended to provide supportive services to families beginning at diagnosis, and 92% of the children were either newly diagnosed or mid-stage. Services received in years 1 and 2 were support counseling, 46%; respite, 22%; nursing care, 15%; activity therapies, 14%; personal care, 2.5%; bereavement care, 0.8%; and pain and symptom management, 0.1%. While pain and symptom management services usage was very low, the study authors suggested that these services may have been provided by physicians who were not associated with the PIC:TFK program. The PIC:TFK was funded by Children’s Hospice International in partnership with the U.S. government, other programs by the CHI partnership exist in Colorado, Kentucky, New York, Virginia, and Utah, and would benefit from being similarly profiled.

Policies

Both federal and state governments have been involved in crafting policies to address pediatric palliative care for families and providers. Federally, the best example remains the 1993 Family Medical Leave Act (FMLA), which protects parents who take time off to be with children receiving medical and psychiatric care from job loss. FMLA permits, among other things, both mothers and fathers to take unpaid leaves of absences from work while still receiving employer-provided health benefits. While the FMLA covers less than half of workers in the private sector, one study of employment statistics across the United States showed that the FMLA did increase leave coverage and usage without negatively impacting women’s employment or wages overall.⁷¹ In California, the “Nick Snow Children’s Hospice and Palliative Care Act of 2005” created a pilot waiver program to address provision of palliative care through home health and hospice agencies, funded by a partnership between Medi-Cal and California Children’s Services.

Payers and socioeconomic circumstances

The economics of health and healthcare have direct influences on the experiences of children and families receiving palliative and hospice care. According to data collected by the Agency for Healthcare Research and Quality for 2006 (Fig. 2-12), most children hospitalized in the United States who were discharged alive were insured either by government or private insurance, with less than 10% of patients having had no primary insurance (the slight exception being pediatric patients age 15 to 20 years, where 11.9% had no insurance). By contrast, among patients who died during their hospitalization, a larger percentage of patients in each age bracket had no primary insurance, and a smaller percentage had private insurance as their primary payer (with the exception again of the older adolescents, who upon reaching 18 years of age are no longer eligible for certain forms of government health insurance coverage).

Fig. 2-12 Primary payers of hospital care for children.

(Data from KID 2006, Agency for Healthcare Quality and Research.)

In the same U.S. hospitalization data, another economic aspect of childhood mortality is evident: the risk of dying during hospitalization is highest among patients from the poorest neighborhoods, and reduces across each step upward in median household income (Fig. 2-13). This observation, which accords with an extensive body of research connecting child wellbeing to socioeconomic circumstances,⁷² reveals the connections between all the layers of the systems that we have outlined. Those layers are from individual risk of having a life-threatening medical condition, through the impact on family finances of having a child with a CCC, to access to and usage of healthcare system services, to the programs and policies that shape the availability of payers and services.

Fig. 2-13 Risk of death during hospitalization and neighborhood wealth.

(Data from KID 2006, Agency for Healthcare Quality and Research.)

International pediatric palliative care

Within the international setting, many national and multinational nonprofit organizations provide a support structure for independent programs based around the world by assisting with fundraising and advocacy. Multinational organizations such as Children’s Hospice International and the International Children’s Palliative Care Network link palliative care programs and hospices in South Africa, Zimbabwe, Uganda, India, Singapore, Argentina, Thailand, Indonesia, Costa Rica, and Belarus.⁷³ In Europe, the European Association for Palliative Care provides a similar program for pediatric palliative care and hospice programs in Norway, Sweden, the Netherlands, Germany, France, Spain, Italy, and other European countries. Within the United Kingdom, the Association for Children’s Palliative Care (ACT) and the Association of Children’s Hospices (ACH) network with individual programs and the British government to advocate for the best possible care.

Issues in pediatric palliative care differ by geopolitical regions in at least two major ways. First, causes of childhood mortality differ drastically around the globe. For example, in a study of death in children under 15 years in Mozambique between 1997 and 2006, 73.6% of deaths were attributable to communicable diseases, with 21.8% caused by malaria, 9.8% by pneumonia, and 8.3% caused by HIV/AIDS.⁷⁴ In Africa, the large number of children living with HIV are intended recipients of many palliative care services.^73,75 In a study on causes of mortality in children younger than 5 years of age in Iraq, diarrhea was the leading cause of death in both infants (49.8%) and children 12 to 59 months (43.4%).⁷⁶ In the same study, researchers noted that cause of childhood death in Iraq often goes unreported when access to health services is limited, and most deaths occur at home. While causes of childhood mortality are certainly linked to broader geopolitical issues of overall morbidity and mortality, poverty, and health services availability and access, pediatric palliative care needs to adapt in each situation to meet the needs of local children.

Second, ethical perspectives about neonatal and childhood illness and death vary significantly around the world. The most obvious example concerns divergent views on euthanasia.^77,78 More broadly, variations in how societies and cultures conceptualize the nature of childhood, parental responsibility, the parent-child relationship, and death likely all affect how pediatric palliative care policies and decisions are framed, understood, and implemented.