CHAPTER 3
PATHWAYS TO A CHILD’S DEATH

. . . I ran downstairs . . . out into the pouring rain. . . . “For God’s sake, where is the ambulance?” . . . [At the hospital,] Dr. Stillman came back looking devastated, and utterly drained. . . . He said that Alexander had died of SIDS. . . . The whole thing was sick. I would know if Alexander was dead. Wasn’t I his mother?

Esmeralda Williamson-Noble, parent, no date

My twenty-year-old brother died 8 months ago from cancer. When he became ill, our whole family changed, and my parents didn’t have time to think of anything else. . . . It still seems as if Michael’s death is all my mother thinks about.

High school junior (Paulson, 2001)

Since birth [our son’s] medical needs have increased, and his health has deteriorated. . . . Our goal has been to try to provide him with the best quality of life he could have. . . . We live every day not knowing when will be our son’s last. . . . [For emergencies] we carry around a sheaf of papers—about 12 pages—that detail all of our wishes as we know them to this point, because we don’t know what every situation will bring.

Tina Heyl-Martineau, parent, 2001

Children who die follow many different pathways to death. Their families accompany them and then follow their own pathways of grief beyond the child’s death, even to the end of their own lives. Understanding the similarities and differences in pathways to a child’s death and in the experi-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 72
CHAPTER 3 PATHWAYS TO A CHILD’S DEATH . . . I ran downstairs . . . out into the pouring rain. . . . “For God’s sake, where is the ambulance?” . . . [At the hospital,] Dr. Stillman came back looking devastated, and utterly drained. . . . He said that Alexander had died of SIDS. . . . The whole thing was sick. I would know if Alexander was dead. Wasn’t I his mother? Esmeralda Williamson-Noble, parent, no date My twenty-year-old brother died 8 months ago from cancer. When he became ill, our whole family changed, and my parents didn’t have time to think of anything else. . . . It still seems as if Michael’s death is all my mother thinks about. High school junior (Paulson, 2001) Since birth [our son’s] medical needs have increased, and his health has deteriorated. . . . Our goal has been to try to provide him with the best quality of life he could have. . . . We live every day not knowing when will be our son’s last. . . . [For emergencies] we carry around a sheaf of papers—about 12 pages—that detail all of our wishes as we know them to this point, because we don’t know what every situation will bring. Tina Heyl-Martineau, parent, 2001 Children who die follow many different pathways to death. Their families accompany them and then follow their own pathways of grief beyond the child’s death, even to the end of their own lives. Understanding the similarities and differences in pathways to a child’s death and in the experi-

OCR for page 72
ences of their families is a helpful foundation for considering the range of palliative, end-of-life, and bereavement services needed to assist children and families. In addition, it is important to keep in mind that for life-threatening conditions that are not invariably fatal, children who eventually die and children who survive often cannot be predictably distinguished at the time of diagnosis, during initial treatment, or sometimes even after initial treatments have failed. This unpredictability increases the challenges and the importance of understanding how to integrate aspects of palliative care from the time of diagnosis. The first section of this chapter discusses prototypical trajectories of dying that depict in graphic form the different ways that death may come to children. The second section presents illustrative stories or vignettes that attempt to represent—in ways that epidemiologic data and charts cannot— the human dimensions of death in childhood. These stories, although they inevitably and greatly simplify real life, suggest how the varied circumstances that surround the deaths of children may affect the child, the family, and the health professionals who care for them. The third section discusses prototypical patterns of care that illustrate traditional and newer perspectives on the relationship between curative or life-prolonging care and palliative and end-of-life care. The newer perspective, stressed in this report, encourages the integration of certain aspects of palliative care from the time a child is diagnosed with a fatal or potentially fatal medical condition. The section also discusses how the emphasis of care may vary depending on the medical circumstances and may, given similar medical “facts,” be affected by differences in family values and circumstances as well as differences in the resources available to them. Documentation of the focus and adequacy of pediatric palliative and end-of-life care is very limited. The final section of the chapter reviews this small literature. TRAJECTORIES OF DYING To help illuminate differences in the paths that people follow to death, Glaser and Straus introduced the concept of a trajectory of dying. They proposed that “the dying trajectory of each patient has at least two outstanding [and variable] properties . . . duration and shape” (Glaser and Straus, 1965, p. 6). The shape of the trajectory depends on time and on the person’s level of functioning or health status. Figure 3.1 presents four simplified trajectories of death in childhood that depict time along the horizontal axis and health status along the vertical axis. These trajectories underscore the reality that no single model of care and support will apply to all dying children and their families. The trajectories do not, however, necessarily map in a straightforward fashion to specific models of palliative care.

OCR for page 72
FIGURE 3.1 Prototypical trajectories of child death. SOURCE: Adapted from IOM, 1997. The most abrupt trajectory, Figure 3.1a, represents the unexpected, sudden death—for example, that of a child killed instantly in a car crash or discovered dead in his crib at home. Such events account for more than a third of childhood deaths. Although children who have died can themselves experience no medical or supportive care, first-response emergency medical personnel commonly follow protocols that require efforts at resuscitation. When a child is involved, only rarely is death actually declared “in the field” rather than in the hospital. If the declaration is based on criteria for brain death, protocols require a period of assessments and observation that may take from 12 to 48 hours or longer (see Chapter 1).

OCR for page 72
Figure 3.1b depicts the experience of children who respond positively to initial, sometimes difficult and possibly life-threatening treatment for a potentially fatal condition, for example, certain brain cancers, but who eventually die of their disease or its further treatment. At diagnosis and during the early stages of initial treatment, these children may not be uniformly distinguished either from children who will not respond or from children who will eventually be cured. (Overall, 70 percent of children with cancer are cured.) Even when an initial positive response to treatment is followed by a return of the disease, determining prognosis is not automatic. From a palliative care perspective, at least three tasks follow diagnosis— each adapted to the situation of the individual child and family. One task is to provide the family—and the child, as appropriate—with accurate and timely information on diagnosis, prognosis, and treatment options and to help them make informed decisions and prepare for a future that includes the hope of cure. A second task is to provide comfort to the child (in particular, by preventing or relieving symptoms of the disease or its treatment) and support to the family. If initial or subsequent treatment fails the child, families will need information about the benefits and burdens of remaining treatment options and about prognosis. When further curative or life-prolonging interventions fail, the additional task for the care team is to help the child and family through the dying phase of the child’s illness and then support the family at death and after. Figure 3.1c depicts the brief trajectory of an infant born with problems incompatible with extended life. As described in Chapter 2, extreme prematurity, severe congenital anomalies, and other life-threatening conditions affecting infants account for a significant fraction of child deaths. With modern prenatal care, some families know in advance that their child will not survive, and some know that they are at significant risk of premature delivery. Extremely premature infants and infants with severe congenital anomalies have produced high-profile disagreements about appropriate care, but care to comfort the infant and the family is always appropriate. As illustrated later in this chapter and discussed further in Chapters 4 and 5, actual support for families may be very limited. Figure 3.1d abstracts the experience of children who suffer from a fatal progressive illness—for example, muscular dystrophy—that is characterized by periods of slowly deteriorating health status that are interrupted by potentially fatal medical crises, which the child repeatedly survives until one crisis ends in death. In some situations, a child’s past favorable response to treatment may leave parents and even clinicians unprepared for and surprised by a subsequent failure of treatment. In other situations, clinicians and parents recognize that the child has reached the final stage of his or her illness and focus on physical and emotional comfort, accepting

OCR for page 72
that resuscitation and other life-sustaining interventions will be more burdensome than beneficial. Various other trajectories for progressive conditions and for some chronic conditions could be mapped. For example, children with severe asthma and appropriate treatment may have mostly normal functioning without disease progression, but a sudden life-ending crisis may occur and be almost as unexpected as a death from injury. The illustrative vignettes presented next describe experiences of a child’s death that approximate the four trajectories presented above. In addition to the two dimensions of health status and time, they depict other dimensions of a child’s death including the individuality of the child and family and the competence and compassion of health care professionals. ILLUSTRATIVE STORIES OF CHILDREN AND THEIR FAMILIES During the course of its work, the committee heard many compelling stories from parents whose child had died. In the details of their experiences and in their expectations, the families differed, but they made clear that parents long remember how health care professionals helped or hurt them. Most parents reported a mix of good and bad experiences. Some told of exemplary care. Others reported insensitive and even cruel behavior from physicians, nurses, or others involved in their child’s care. Some found that insurers or health plans were supportive; some had no insurance; others had continuing and exhausting difficulties trying to get care approved—and then suffered with their child when coverage of hospice or other supportive services was denied. Often faced with uncertainty about their child’s prognosis and treatment options, some parents anguished over decisions about how far to pursue experimental treatments or whether and when mechanical ventilation or other life-support interventions should be withheld or withdrawn. Beyond his or her medical condition, a child’s experience of dying will depend on many factors, including the family’s circumstances, values, and responses to the illness and the kind of medical and other services that are available to meet the child’s physical, psychological, spiritual, and other needs. The brief stories presented below suggest how the complex interplay of these factors may affect the level of distress experienced by children and their families. None of the following stories depicts a specific child and family exactly, although each is based on real experiences. Some are fairly typical of childhood deaths, whereas others represent more unusual or particularly challenging situations. The stories are clearly simplified, but they highlight differences in the needs of children and families and differences in the

OCR for page 72
extent to which the care provided met these needs. Most depict deficits in some aspects of the care provided the child or family. “Ana Rivera” When death comes to children, it often is sudden and unexpected (Figure 3.1a). In this case, 3-month-old Ana’s father finds her lifeless body in her crib one morning. Her parents’ grief is complicated and intensified when they realize that a police officer’s questions are probing the possibility of homicide. This story also illustrates the special difficulties facing immigrant parents whose language abilities and resources are limited. Jorge and Maria Rivera had emigrated from El Salvador two years prior to Ana’s birth, leaving their three children behind with Maria’s mother and promising to send money to support the family. Maria’s first husband and two older brothers had been killed during the war, but her cousin had escaped to the United States and urged Maria and her second husband Jorge to follow him. Jorge got a job at a fast-food restaurant and Maria cleaned houses, and they scrimped on everything so that they could send a little money back to their family in El Salvador. They felt isolated but were overjoyed when they learned that Maria was pregnant. A neighbor brought Maria to a clinic that provided prenatal care to poor women. The Riveras were proud to give this new baby—Ana—such a good start. After an uneventful pregnancy, delivery, and first few months, Jorge found 3-month-old Ana cold, blue, and motionless in her crib one morning. The Riveras rushed Ana to the public hospital where she was immediately taken into a treatment room. In a short while, the Riveras were led into a small sitting room in the emergency department where a physician informed them, through a janitor who spoke both English and Spanish, that Ana had died. The janitor did not understand all of the words the physician said but assured the Riveras that everything had been done to help Ana. After the physician left, a police officer came into the room and began to ask Jorge questions about how he had found Ana, what he was doing prior to finding her, what her health had been like prior to that morning, and whether Jorge had ever hurt either Ana or Maria. Jorge remembered that he had been angry the night before and yelled at Ana when she wouldn’t stop crying, but he was afraid to tell the police officer this information. No charges were brought against Jorge and Maria, but they lived in fear that they would be arrested and sent back to El Salvador. Both had intense feelings of loss for Ana and for their three other children who were still in El Salvador. A hospital social worker told the Riveras about bereavement support groups, but none was nearby and none had groups for Spanish speakers. The social worker really helped in one way. She helped make sure that the Riveras qualified for free care under a community program for the

OCR for page 72
medically indigent, so at least they did not face medical bills that they could not possibly pay. “Jimmy Marshall” In this case, which involves another leading cause of child mortality, a car–bicycle collision set off a frantic effort to save an injured child’s life. To the parents, the emergency department and intensive care unit are frightening and strange. Despite the help of several social workers, there is much they do not understand and that is never explained. They are unwittingly misled by one clinician who is satisfied with the technical success of a procedure but not focused on what is still a very grim, likely fatal situation. One result is a father’s lifelong regret that he was not with his wife and son when the boy died. While riding his bicycle with a friend, Jimmy Marshall, an 11-year-old boy, was struck by a car. Emergency personnel found him semiconscious with obvious head injuries and rushed him to the nearest trauma center, which treated but did not specialize in pediatric trauma. In the emergency department (ED), Jimmy was taken to a large, open resuscitation bay while the trauma team, including nurses, surgeons, neurosurgeons, emergency physicians, and pediatricians, cared for him. His condition was clearly very serious. Friends rushed Jimmy’s mother to the hospital. In a small, windowless room, the social worker provided some brief information and comforting words. Then, the ED “attending” came in to explain that Jimmy had suffered a serious brain injury and they were doing everything they could to help him. Later, a “pediatric resident” said that Jimmy was being taken to x-ray for special scans of his brain and that the team was very worried. Fighting tears and hysteria, Mrs. Marshall asked, “But he is going to be all right isn’t he?” The resident appeared to almost shrug as she said, “I can’t tell you that.” The rest was a blur. Just after Jimmy was taken to the intensive care unit, his father arrived. The attending physician gave a brief, grim update. Both Marshalls started to weep. The family would have welcomed the support of a hospital chaplain, but no one thought to see if someone was available. As they waited, the conversation around them sounded ominous, but the terminology was mysterious, and both parents were too bewildered or intimidated to ask many questions. Told that neurosurgeons needed to put a hole in their son’s skull in order to monitor dangerous swelling, Mr. Marshall protested that this might injure Jimmy’s brain. A doctor replied, “You have more important things to worry about right now.” After the procedure, the neurosurgeon approached, smiling and commenting that things went “great.” For the first time, Jimmy’s parents felt

OCR for page 72
relieved and somewhat confident. The father, afraid that he would lose his job, then left to return to his construction work. However, when the doctor who appeared to be in charge returned, he told Mrs. Marshall that they needed to talk and that she should try to reach her husband. In the interim, the doctor told Jimmy’s mother that there had been increasing brain injury and that the task was to prevent the brain from swelling and “crushing itself.” Her son might die. This seemed impossible given that only a few minutes earlier someone else had seemed so cheerfully pleased. Suddenly, a nurse rushed in to say that the doctor was needed. Jimmy’s heart had stopped and CPR (cardiopulmonary resuscitation) had been initiated. The doctors knew that the likelihood of resuscitating Jimmy was negligible but were obligated to try. The doctors subsequently had to tell the Marshalls that they had done everything, but Jimmy had died. Later, Mr. Marshall agonized that he hadn’t been told enough to realize that he should stay with his wife and boy, and he sometimes felt angry at how the neurosurgeon misled them. He and his wife got a piece of paper with information about bereavement support, but it was laid aside and then lost in all the turmoil. There was no further contact from those who were with Jimmy when he died. At least, they had qualified for “free” hospital care. “Melissa Devane” As children mature, their intellectual and emotional understanding of serious illness and the prospect of death evolves. This story describes the complex relationships between a severely ill adolescent and her parents and physician and the different concerns she has as she moves from diagnosis and treatment, to recurrences and further treatment, to death (Figure 3.1b). The adolescent is in conflict with her parents and physician about undergoing burdensome experimental treatment but eventually persuades them to respect her wishes. When Melissa was 13 years old, she was an excellent softball player who hoped some day to play on the Olympic softball team. Late one summer, her knee began to hurt severely and kept hurting. When Melissa’s primary care physician thought an x-ray suggested a tumor, she referred her to Dr. Garcia, a pediatric oncologist. After a biopsy, Dr. Garcia diagnosed osteogenic sarcoma, an invasive bone cancer. As Dr. Garcia recommended, Melissa had several weeks of chemotherapy followed by surgical removal of the tumor and then several months more of chemotherapy plus physical therapy. Although the hospitalizations for chemotherapy and episodes of fever were no fun, what bothered Melissa most was losing the ability to play her sport ever again. Also, her hair fell out just as school was starting. The social worker on her oncology team helped Melissa to get a wig and, as

OCR for page 72
part of the team’s attention to school reentry issues, explained Melissa’s situation to classmates and teachers. The family celebrated the end of Melissa’s chemotherapy with a big party. Melissa was sad that only half of her former teammates made it to the party. She commented to the social worker that she had learned through the experience “who her real friends were.” The yearbook club and her younger sister’s softball games became new focal points. At a follow-up appointment 13 months after the end of chemotherapy, Dr. Garcia found a spot on the CT (computed tomography) scan of Melissa’s right lung. He recommended surgical removal of the suspected tumor to confirm the diagnosis. The surgery was uneventful, and Melissa was pleased there would be no chemotherapy. Six months later, however, scans showed tumors in both lungs. This time, intensive chemotherapy followed surgery. Family and friends and an occasional tutor supported Melissa through the months of hospitalization. Now almost 17 years old and hoping to plan for college, Melissa did her best to look to the future and regain her strength. By the winter of her senior year, she was back in school full time, working on the yearbook, and seeing a new boyfriend. Then, routine scans showed another recurrence in the lung, but this tumor was too big to remove surgically. Dr. Garcia proposed experimental chemotherapy that would require long hospitalization. Melissa reacted by saying, “I’d rather die than have more chemotherapy” and “you’re unreal—I’m going to die anyway.” This surprised everyone. Melissa had been “such a perfect patient.” Melissa’s dad felt the issue was settled. (“It’s our decision. She’s a child.”) Her mother felt that “since it is her body and her life,” Melissa should be part of the decision-making and that Dr. Garcia ought to be clearer about what could be expected. The subsequent discussions were emotional but less anger filled as they went along and Dr. Garcia acknowledged that it was unlikely that the experimental treatment would help meaningfully prolong her life and would be arduous. Melissa explained that she wanted to live what time she had left by doing what she wanted to do: attend the prom, finish the yearbook, and coach her sister in softball a bit more. Melissa was able to do those things. She and her parents and Dr. Garcia ultimately agreed on palliative chemotherapy that was given by mouth. The Devanes’ health plan quickly approved referral to a local hospice. Because Melissa was likely to die within six months and because the family had opted for palliative care only, there was no issue of appealing the plan’s limits on hospice coverage. The hospice care team gave the family the help it needed to keep Melissa at home with minimal pain. Dr. Garcia continued to keep in regular touch with Melissa and her parents. Melissa died peacefully at home in July in her mother’s

OCR for page 72
arms, having lived longer than Dr. Garcia predicted and having seen her sister have a championship season with Melissa’s support and coaching. “Sean Riley” This story illustrates the increasingly common situation in which a fatal congenital anomaly is diagnosed prenatally and the families await a birth that likely will be followed by death within hours or days (Figure 3.1c). Although many mothers and fathers opt for abortion when faced with such a tragic diagnosis, others—for religious, philosophical, or emotional reasons—choose to continue the pregnancy. A few innovative programs have recently been developed to provide extensive clinical information and preparation and emotional, spiritual, and practical support (e.g., help with coordinating services) to these families following diagnosis and through and after the infant’s death (see, e.g., Sumner, 2001). Most parents, however, have limited access to such information, preparation, comfort, and assistance from health care professionals. They may be able to rely on families and friends, and some discover parent-to-parent support groups or other resources. Catherine and Kevin Riley were delighted to learn that they were expecting a new baby to join their 4-year-old daughter, Caitlin. In Catherine’s sixteenth week of pregnancy, she had an ultrasound examination. From the reactions of the technicians, she knew immediately that something was wrong, but no one was willing to tell her anything except that she should contact her obstetrician. Told initially that the ultrasound results would be reviewed at Catherine’s next regularly scheduled visit in four weeks, Kevin called the obstetrician’s office to explain the extreme stress they were under and their need to know the results. The obstetrician explained that the fetus had anencephaly (failure of the brain to develop), that he had little training in managing such pregnancies, and that they should consider termination of the pregnancy “because the baby would not live.” The Rileys were unwilling to consider termination and requested a referral to a new obstetrician who specialized in high-risk pregnancies. The second obstetrician said that there was nothing to be done for mother or fetus and that if they did not want to terminate the pregnancy, he would see Catherine again when she was ready to deliver (in about five months). Without access to regular obstetrical visits for further information and preparation, the Rileys drifted in fear, uncertainty, and grief during the following weeks and months. They sought information from the Internet where they found stories from other families about anencephalic infants who had survived “for years” but learned little to answer their questions

OCR for page 72
about why this problem had happened and how they should prepare for what was to come. Catherine went into labor at 38 weeks’ gestation. The Rileys found delivery room nurses who were unaware of her fetus’s diagnosis and an unfamiliar obstetrician who was covering for the obstetrician she had last seen. As Catherine’s contractions increased in intensity, the nursing staff on the delivery floor became concerned about the fetus’s heart rate. The obstetrician suggested delivery by cesarean section. Kevin questioned this recommendation, but the obstetrician and nursing staff both insisted it was needed and that if they refused both parents would have to sign a form stating their refusal. Confused by these demands given the diagnosis and prognosis, Kevin and Catherine requested a second opinion. Before the opinion could be obtained, the fetus’s heart rate dropped to dangerously low levels, and the baby was delivered by cesarean section with Catherine under general anesthesia. The obstetrician and anesthesiologist confirmed the diagnosis of anencephaly. Kevin was not allowed in the delivery room because Catherine was under general anesthesia. Neither saw their son, Sean, during his brief moments of life. After delivery and her baby’s death, Catherine was placed on a floor in the hospital to recover where there were no babies or other postpartum women. She noticed that nursing staff avoided her, and she felt very isolated. Her obstetrician saw her once, inquired about her pain control, and said she could go home “when she was feeling better.” She was discharged three days later, after receiving an envelope with a lock of Sean’s hair, a handprint, and a Xeroxed paper about “grief.” Many of the Riley’s friends avoided any contact when news about Sean’s birth and death became known. Her employer refused postpartum leave or paid bereavement leave because “she had no baby.” Kevin never spoke about Sean again, although he noticed that he became extremely sad unexpectedly for years afterward. Daughter Caitlin had stayed with relatives during the delivery. No one told her much about Sean’s birth and death. She began to think that her parents were angry with her. Expectations of both Kevin and Catherine for Caitlin increased significantly, which added more pressure for an already vulnerable child. Five years after Sean’s death, the Rileys remain a sad and troubled, though loving, family. “Johnny Gabrielle” Although caring for a child with a serious chronic condition is always demanding, the burden is significantly increased when the child is developmentally delayed and cannot communicate his or her needs directly. Most children and families can be helped to lead fairly normal lives for extended

OCR for page 72
increased the use of sedatives and pain medications before the withdrawal of mechanical ventilation (which was usually one of the first therapies stopped for these children). Relief of suffering (e.g., air hunger, pain) was the justification for this increase, and the only patients not receiving such care were comatose and thought not to be suffering. The researchers reported no use of paralyzing agents, the use of which they considered indefensible. Parental and clinical assessments of children’s pain and other distress may differ. For example, as shown in Table 3.1, Wolfe and colleagues (2000a) found “considerable discordance” for certain symptoms including fatigue, poor appetite, constipation, and diarrhea. Although some observational tools have been developed to assess certain symptoms in children, clinicians or researchers often depend on parental assessments of children’s pain not only in retrospective studies following a child’s death but also for very young children, children with cognitive TABLE 3.1 Discordance Between the Reports of Parents and Physicians Regarding Children’s Symptoms in the Last Month of Life   Reported by Parent but Not by Physician (N = 92) Reported by Physician but Not by Parent (N = 92)   Symptom No. (%) No. (%) Kappa Statistic (95% CI)a P Valueb Fatigue 44 (48) 1 (1) –0.02(–0.07 to 0.02) <0.001 Pain 15 (16) 11 (12) 0.10(–0.12 to 0.32) 0.56 Dyspnea 19 (21) 10 (11) 0.10(–0.11 to 0.31) 0.14 Poor appetite 33 (36) 1 (1) 0.29(0.15 to 0.43) <0.001 Constipation 31 (34) 7 (8) 0.16(–0.02 to 0.33) <0.001 Nausea/vomiting 25 (27) 18 (20) 0.06(–0.14 to 0.26) 0.36 Diarrhea 20 (22) 8 (9) 0.31(0.12 to 0.51) 0.04 aData were missing for 10 children for whom there was no documentation of clinic or hospital visits in the last month of life and for 1 child whose records were not available for review. CI denotes confidence interval. bMcNemar’s test was used. SOURCE: Wolfe et al., 2000a. Used with permission.

OCR for page 72
impairments, and children with very advanced disease. One review of studies comparing parental and child assessments concluded that the evidence is mixed but parents tend to underestimate their child’s pain (Chambers and Craig, 1999). A Swedish study of children being treated for cancer interviewed children and parents (Ljungman et al., 1999). It reported that “treatment-related pain was constant and dominant,” that treatment- and procedure-related pain was more severe than pain due to malignancy, and that pain evaluations were unsystematic. An older study by McGrath and colleagues (1990) also reported that treatment- and procedure-related pain was common, with more than a third of the children reporting moderate to severe pain from chemotherapy and 61 and 78 percent reporting such pain from lumbar puncture and bone marrow aspiration, respectively. The study did not report on treatment strategies. A 1995 article reporting on severe pain in children dying of cancer concluded that following WHO pain guidelines was adequate for most children but that a subgroup required extraordinary measures (Collins et al., 1995). Individualized management and careful assessment of pain were essential. Studies by Ferrell and colleagues also suggest the inadequate management of cancer pain in children (Ferrell et al., 1994; Rhiner et al., 1994). The studies assessed the impact on families of living with and trying to manage a child’s pain and identified parents’ feelings of helplessness and stress. Additional studies have examined pain assessment and management for broader groups of children. A 1994 article reported on pain management practices in three Canadian neonatal intensive care units (NICUs). It concluded that procedural and disease-related pain is frequently untreated or undertreated (Fernandez and Rees, 1994). A study by Kazak and colleagues (1996) compared distress during invasive procedures for childhood leukemia as rated by mothers and nurses. Children receiving a pharmacological-only intervention were rated as having more distress than children receiving a combined pharmacological and psychological intervention. Younger children were rated as having more distress than older children. A 1999 review of the literature on psychological interventions for disease-related pain in children identified relatively few studies, most of which had methodological problems (e.g., small sample size, no control group) (Walco et al., 1999). Taken together, the studies suggested benefits of some interventions but they were not definitive. Unexplained variation in pain management and other practices is also a concern. One study of narcotics use for very low birth weight babies in NICUs reported a 29-fold variation in the use of opioids (Kahn et al, 1998). This kind of variation demands further investigation. Differences may re-

OCR for page 72
late, in part, to differences in the degree to which physicians feel legally vulnerable in administering pain medications or have misconceptions about addiction from long-term use of narcotics. Variation in methods used to assess the contribution of pain in prematurely born infants may also play a role. Pain assessment tools that have been validated in premature babies at different gestational ages and at different chronological ages should prove useful in evaluating the adequacy of neonatal intensive care, reduce unexplained variations in treatment, and reduce infants’ distress (see, e.g., Krechel and Bildner, 1995). Palliative Care Consultations and Discussions The extent of explicit palliative care consultations is little documented for children and families. A recently published study examined all 196 deaths in an NICU between 1994 and 1997 (Pierucci et al., 2001; see also Leuthner and Pierucci, 2001). The authors reported that palliative care consultations were associated with significantly more supportive services (e.g., referrals to chaplains or social workers) for infants and families, fewer medical procedures (e.g., blood draws, radiographs, endotracheal tubes but not intravenous fluids, nutritional support, narcotics), and documentation of family emotional needs and support. Consultations prior to the child’s death increased during the period studied from 5 to 38 percent. The authors could not assess whether lack of consultation reflected lack of physician or family readiness for palliative care or other factors. They did not attempt to measure the effectiveness of the supportive services provided in relieving child or family distress. In the study by Wolfe and colleagues (2000a) cited above, earlier discussion of hospice care was associated with parental reports that their child was peaceful and calm during the last month of life. For about two-thirds of the children who died of progressive disease (rather than treatment complications), charts noted a discussion of hospice care, which occurred, on average, about 58 days before death. Further, suffering was greater in children whose parents reported that the child’s physician was not involved in his or her end-of-life care. This pattern held after adjusting the analysis for cause of death, child’s age at death, place of death, interval between the child’s death and the parental interview, and physician clustering (Wolfe et al., 2000a).4 4   Authors “adjusted for the possibility that the parents of patients who had the same physician may have had similar responses” and called this “physician clustering” (Wolfe et al., 2000a, p. 327).

OCR for page 72
Limitations of Life-Support and Disease-Modifying Interventions A recent examination of the limitation of aggressive life support concluded that much attention had been paid to the question of whether and when to limit these interventions and argued that more research should focus on how to assess and improve the quality of care for patients once the decision has been made (Rubenfeld and Curtis, 2000). Studies of pediatric deaths in intensive care have found that deaths are often preceded by limitations of life support interventions—more than half of the deaths in one recent study (Keenan et al., 2000), more than one-third to one-half in an earlier study (Vernon et al., 1993;5 Levetown et al., 1994), and more than 70 percent in a neonatal intensive care unit (Wall and Partridge, 1997). The accompanying assessment and management of pain, air hunger, or other distress has been less documented as has the timing of decisions to limit life support. Even though death is often preceded by limitations of life-support interventions, studies also suggest that during their last month of life, the majority of children who die from chronic conditions receive some form of medical treatment that is regarded as curative or life prolonging. The earlier-cited study by Wolfe and colleagues (2000a) reported that 56 percent of children with cancer received some sort of cancer-directed therapy in the last month of their lives. More than one-third (36 percent) had undergone a bone marrow transplant, and for 60 percent of these children, the transplant was their last cancer-directed treatment. Two-thirds of children in the study had DNR orders. These orders were entered, on average, 33 days before death for children who died of their disease and 1.7 days before death for children who died of treatment-related complications. An earlier study of DNR orders in one children’s hospital also reported that most of these orders were written for children who were receiving “aggressive” medical therapy in intensive care units (Lantos et al., 1993, p. 52). Orders were more common for older children than for infants. Wolfe and colleagues suggest that when life-prolonging treatment is pursued to the last, more attention to concurrent palliative therapies might reduce children’s suffering. Possibly reflecting the difficulty of changing a long-term pattern of treatment, one small study suggests that children with chronic illness hospitalized in PICUs were less likely to have life-sustaining treatment withheld or withdrawn than were acutely ill patients (Keenan et al., 2000). Levetown’s 1994 study, however, reported no difference between the two 5   This article used the terminology “supportive care” to describe life-support interventions, whereas the term is used here and more generally to refer to care that provides comfort to patients or families.

OCR for page 72
groups on this aspect of care. Additional research on treatment patterns for acute versus chronic conditions could inform understanding of the factors influencing treatment choices. For children with serious, chronic conditions, repeated hospitalizations may be an almost routine part of life. Families may become used to seeing the ill child “bounce back” after numerous life-threatening acute events. Then, as Ann Goldman suggests, “After a lifetime of intensive treatment, it can be difficult to abandon past patterns of care” (Goldman, 1999, p. 30). With advances in organ transplantation, patients with conditions such as cystic fibrosis (often now young adults) and their families may regard what was previously seen as the end stage of their disease as an opportunity to be listed for a transplant, thus maintaining a hope for survival rather than preparing for death. Long-term survival, while improving, varies for different kinds of conditions (see Kurland and Orenstein, 2001), and a shortage of donated organs limits the number of transplants. Of the 45 patients who died of cystic fibrosis in Canada in 1996, an analysis found that 93 percent had been considered for transplant, 7 (17 percent) had been listed, and 2 (4 percent) had received the transplant (Mitchell et al., 2000). The analysis did not report how many received a transplant and did not die.6 Palliative care was never discussed for one-quarter of the patients and was discussed a month before death for 40 percent. Another study, this one in the United States, evaluated records for 44 patients more than 5 years of age who died of cystic fibrosis at one institution between 1984 and 1993. The majority of these patients were over age 20. Four died after lung transplants, and four died on mechanical ventilators in the PICU awaiting transplant (Robinson et al., 1997). Three-quarters of the patients were receiving therapeutic (e.g., intravenous antibiotics, chest physical therapy) or preventive (e.g., oral vitamins) interventions in the last 12 hours of life. Compared to surgical interventions, most cancer chemotherapy, and some other drug regimens, these therapies are less invasive and produce limited side effects. All patients had do-not-resuscitate orders, and nearly all received some opiates for pain (two of six patients who did not refused for religious reasons). Of the 44 patients, one died at 6   Three-year survival rates following lung or heart–lung transplantation for cystic fibrosis are generally reported (for the last decade) to be from 40 to 60 percent (see, e.g., Balfour-Lynn et al., 1997; Hosenpud et al., 1998; Mendeloff et al., 1998; Aurora et al., 1999; Vizza et al., 2000). Transplant survival rates are higher for certain other conditions and organs, most notably kidney transplantation for end-stage renal disease. Five-year survival rates above 50 percent have been reported for children with end-stage heart disease (see, e.g., Jenkins et al., 2000; Morrow et al., 2000; Williams et al; 2000). Survival rates improve with experience with the procedure and vary with medical condition (e.g., congenital anomaly, cardiomyopathy).

OCR for page 72
home under hospice care; the remainder died in the hospital with “many of the important psychological and emotional aspects of hospice” (Robinson et al., 1997, p. 208). The authors of this study noted that while the final outcome of the disease is certain, uncertainty about short-term prognosis and the waxing and waning nature of respiratory failure in cystic fibrosis make it difficult to abandon antibiotics, physical therapy, mechanical ventilation, and similar measures. Again, some of these interventions impose relatively little burden on patients. Despite prognostic complexities, improvements in the care of patients with unpredictable life-shortening conditions are possible. The strategy of “hoping for the best while preparing for the worst” allows clinicians and families to pursue life-prolonging interventions such as transplants with conviction and, at the same time, to be attentive to symptom management, personal goals (e.g., a family vacation), and other steps to help families make the most of the time they have. One study of the withdrawal of life support in adults reported that “forms of life support that were perceived as more artificial, scarce, or expensive were withdrawn earlier” and that “once the decision has been made to forgo life-sustaining treatment, the process remains complex and appears to target many different goals simultaneously” (Asch et al., 1999, p. 153). If anything, the complexity of the process and the existence of multiple goals of care are likely to be even more evident when decisions involve children. Inconsistency in clinical decisions about life-support and disease-modifying interventions is a concern. One survey study of pediatricians’ decisionmaking found considerable variation in management strategies about withholding or withdrawing life-support interventions for chronically ill children as described in eight clinical scenarios (Randolph et al., 1999). For six of the eight scenarios, less than half the respondents chose the same intensity of care (e.g., mechanical ventilation, dialysis, antibiotics). For three of the scenarios, at least 10 percent chose the most intensive level of management while at least 10 percent chose comfort measures only. The most important influences on choices were the probability of acute (short-term) survival and parental wishes. Those who rated parental wishes highly were more likely to choose aggressive care when parents wanted it, even when the long-term prognosis was poor. Also, those who rated themselves as more likely than their colleagues to withdraw life support were indeed more likely to make that choice. Although the authors note that decisions in actual practice needed investigation, they observed that “variability in decision-making may lead to unnecessary suffering, lack of fairness when making decisions about neurologically handicapped infants, and inappropriate use of scarce resources in futile cases” (Randolph et al.,

OCR for page 72
1999, p. e46). They recommended clearer guidelines for decisions about limitation of life support for children. A recent study of European neonatologists also found considerable variability in how quality-of-life was weighed in decisions in the NICU (Rebagliato et al. 2000). The authors concluded that the country in which a physician practices was the most powerful predictor of choices. The researchers also found that physician attitude, seniority of position, length of experience, and working in larger NICUs were significantly correlated with a greater likelihood of a physician’s choosing to limit life-sustaining medical interventions based on predicted quality of life. In the United States, studies examining the attitudes of neonatologists report that neonatologists frequently are willing to limit treatment in severely malformed, premature, or ill babies. One study examined the actual practices of neonatologists at a university medical center as documented in medical records. It found that 73 percent of all neonatal deaths at a tertiary NICU followed the withdrawal or withholding of life-sustaining treatment (Wall and Partridge, 1997). In contrast, a similar study in 1973 reported that only 14 percent of neonatal deaths were attributed to withdrawing or withholding treatment (Duff and Campbell, 1973). In the 1997 study, death attributable to withdrawing or withholding life-sustaining treatment was more common in severely neurologically damaged babies (more than 85 percent) than in babies with major congenital anomalies (67 percent). Neonatologists most commonly noted futility of treatment as the reason for limiting or withdrawing treatment (74 percent). They noted quality-of-life issues in 51 percent of cases (more often mentioning severe disability than unnecessary suffering as the issue). Quality of life, exclusive of any reference to the futility of treatment, was mentioned as the reason for withdrawing or withholding treatment in 23 percent of the deaths. The authors of the study observe that medical records might underdocument the importance of quality-of-life considerations based on concerns about federal “Baby Doe” regulations on treatment of infants with life-threatening conditions (see Chapter 8). Problems in Communication Hilden and colleagues observe that physicians report anxiety about discussing an impending death with children and their parents (Hilden et al., 2001a). Anxiety and dread may interfere with a thorough assessment of a child’s emotional and spiritual well being and careful evaluation of child and family circumstances (e.g., cultural values, religious beliefs). Inadequate assessment may, in turn, compromise timely efforts to identify and relieve physical and emotional distress.

OCR for page 72
Clinicians’ anxiety may also contribute to poor communication about the child’s situation and prognosis, which may deprive children and families of information they need to plan for the future and weigh the pain or other suffering expected from an intervention against its probability of meaningfully extending a child’s life. Certainly, most pediatric specialists have less experience than similar adult specialists in communicating a grim prognosis, although both groups have little or no training in this task (see Chapters 5 and 9). A survey of 122 experienced emergency room physicians published in 1997 reported that few of the respondents reported any training in telling parents about a child’s death or any familiarity with guidelines for such notifications (Ahrens and Hart, 1997; Maruyama, 1997). Two-thirds of the respondents described communicating with parents after a child’s death as the most difficult part of emergency medicine. A similar percentage said communicating with parents about a child’s death was more difficult than communicating with a family about an adult’s death. Even more striking in this small survey, more than two-thirds of the respondents said that they had prolonged resuscitation efforts solely to delay telling parents that their child had died. Khaneja and colleagues (1998) surveyed residents, fellows, and attending pediatricians at one institution to describe their general attitudes toward pediatric death and dying. Of those responding, 61 percent of all attending physicians and 73 percent of attending physicians in specialties with a higher predicted patient mortality rate reported that they sometimes consider a patient’s death to be a personal failure. Conversely, residents and fellows rarely felt this sense of personal failure when a patient died. Despite feelings of failure, when a patient died, older physicians not surprisingly felt more prepared to deal with issues of death and dying than younger physicians: 13 percent of residents, 56 percent of fellows, and 71 percent of attending physicians felt adequately trained to deal with end-of-life issues, but the study did not independently assess skills. However, almost all of the physicians agreed that there was a need for further support of health care workers in dealing with death and dying (100 percent of residents, 83 percent of fellows, and 90 percent of attending physicians wanted more support). Most also agreed that the existing support systems in the pediatric department of their hospital were inadequate. Many physicians expressed a desire to attend funerals of patients, for instance, but very few had actually done so. Most cited conflicting clinical duties as a reason they were unable to attend funerals. Other researchers as well as personal stories also point to physician anxiety and confusion about communicating bad news and discussing death and dying (see, e.g., Solomon, 1993; Nuland, 1994; ABIM, 1996a,b; IOM, 1997; Christakis and Iwashyna, 1998; Bruera et al., 2000; for pediatricians

OCR for page 72
specifically, see Ahrens and Hart, 1997; Vazirani et al., 2000). Although most of these studies involve adult patients and families, it seems reasonable that some of the problems they identify will apply to communication involving children with life-threatening conditions. For example, a study by Lamont and Christakis (2001) reported that clinicians may consciously withhold information or provide cancer patients and families with misleadingly optimistic assessments and that younger adult patients were more likely to receive such misleading assessments than older patients. A recent project to improve communication about organ donation in adult neurologic and pediatric intensive care units reported on training sessions that involved conversations between health care teams and standardized patient families. As reported in a conference poster session (Williams et al., 2001b), the researchers found that physicians were reluctant to use the “d” words (death, dying), that euphemisms confused family members, and that teamwork among team members (physicians, nurses, transplant coordinators, and clergy) contributed to effective communication. (Following training, organ donation rates increased on the study units.) One small study of residents’ competence in delivering bad news (a diagnosis of lung cancer) to adults concluded that they were not very good at it (Eggly et al., 1997). Using a 16-item instrument, the authors rated residents’ skills as displayed during taped interviews with simulated patients. Residents were rated weakest in eliciting the patient’s perspective. The authors concluded that the residency program was not doing an adequate job of teaching residents to use a patient-centered approach to discussions with patients. Another small study also used simulated parents as part of the process of assessing the parent counseling and information-giving skills of pediatric residents and emergency department fellows, providing feedback, and assessing improvement in subsequent discussions (Greenberg et al., 1999). The investigators reported improvements in two areas: reducing parents’ feelings of dependence and improving feelings of trust. A third small study of pediatric critical care fellows using a somewhat different methodology also showed improvements following feedback by the simulated parents (Vaidya et al., 1999). Other studies have concluded (through self-reports and clinician assessments) that communications training for medical students and residents can improve their self-confidence, comfort, or skills in breaking bad news (Cushing and Jones, 1995: Garg et al., 1997; Vetto et al., 1999). If physicians cannot face fully a child’s poor prognosis and then appropriately communicate their assessment to families, timely reexamination of the goals of care and corresponding adjustments in care plans may be delayed. One retrospective study of parent and physician perceptions of the end-of-life care of 103 child cancer patients found a significant disparity between the time the physician first documented that the child was entering

OCR for page 72
the end-of-life phase of his or her illness and the time the parents reported recognizing that there was no realistic chance for cure (mean difference of 101 days) (Wolfe et al., 2000b). The differences in physician and parent understanding of prognosis were not statistically significantly associated with differences in the care outcomes studied. However, when the physician’s and parents’ understanding of prognosis correlated more closely early on, the goals of care were more likely to be directed at lessening the child’s suffering. Hospice was introduced earlier. Parents also were more satisfied with home care during the end-of-life period. When a psychologist or social worker was involved in the child’s care, agreement between physician and parent assessments was more likely. IMPLICATIONS Children who die vary in the paths they follow, and some children with grave medical problems survive. Thus, pediatric palliative care—and if death occurs, end-of-life and bereavement care—must be flexible to meet the highly variable, changing, and sometimes difficult-to-predict needs and circumstances of children and families. Evidence about the adequacy of current care for children with fatal or potentially fatal medical problems is limited but generally consistent with that found for adults in pointing to shortfalls in many dimensions of care. Specifically, pain and other symptoms are inadequately and variably assessed and treated, despite the availability of treatments that are effective, especially for pain. Palliative care consultations appear to increase the use of supportive services, but late or absent consideration of the benefits of palliative or hospice care is a problem. This puts children at risk of needless suffering from invasive and uncomfortable procedures that promise little or no benefit. Routine discussion and re-evaluation of the goals of care, which assists parents in making medical and other decisions, does not appear to be a part of “normal” health care for children with complex medical conditions. More children with fatal conditions are dying after the limitation of life-support interventions, but the adequacy of care to manage any distress associated with stopping or not starting such interventions is little studied. Little research is likewise available on the prevalence, intensity, and management of symptoms other than pain and of psychological distress in children with fatal or potentially fatal conditions. Different studies point to problems in the communication of information to help families understand what may lie ahead and to make decisions about different courses of care. Barriers to these “just in case” conversations include physicians’ discomfort in discussing bad news, fragmentation of care, and financial disincentives.

OCR for page 72
The literature reviewed in this chapter as well as illustrative vignettes suggest several directions for improving the effectiveness and consistency of pediatric palliative and end-of-life care. These directions include better understanding of the dimensions of pediatric palliative and end-of-life care and better understanding and application of existing knowledge and resources. Other directions include changes in the organization of care that may increase the effectiveness and reliability of such care, better training of health professionals in the principles and techniques of palliative and end-of-life care (including accurate but sensitive communication), and more research on almost every dimension of such care and the medical and other criteria for identifying children and families who might benefit from it. The next several chapters expand on these topics and also review the financing, legal, and ethical context of care for children with fatal or potentially fatal medical conditions.