|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Attitudes and Practices Among Pediatric Oncologists Regarding End-of-Life Care: Results of the 1998 American Society of Clinical Oncology SurveyFrom the Department of Pediatric Hematology/Oncology, Children’s Hospitals and Clinics—St Paul, St Paul, MN; Department of Clinical Bioethics, Warren G. Magnuson Clinical Center, National Institutes of Health, Bethesda, MD; Center for Research Methodology and Biometrics, AMC Cancer Research Center, Denver, CO; Department of Pediatrics, Stanford University School of Medicine, Palo Alto, CA; Pain and Palliative Care Service, Memorial Sloan-Kettering Cancer Center, New York, NY; and Center for Survey Research, University of Massachusetts/Division of Hematology and Oncology, Beth Israel/Deaconess Medical Center, and Department of Medical Adult Oncology, Dana-Farber Cancer Institute, Boston, MA. Address reprint requests to Joanne M. Hilden, MD, Department of Pediatric Hematology/Oncology, Children’s Hospitals and Clinics—St Paul, 345 N Smith Ave, St Paul, MN 55102; email hilde005{at}tc.umn.edu
PURPOSE: In 1998, the American Society of Clinical Oncology (ASCO) surveyed its membership to assess the attitudes, practices, and challenges associated with end-of-life care of patients with cancer. In this report, we summarize the responses of pediatric oncologists and the implications for care of children dying from cancer. METHODS: The survey consisted of 118 questions, covering eight categories. All ASCO members in the United States, Canada, and the United Kingdom were mailed a survey, which was completed by 228 pediatric oncologists. Predictors of particular attitudes and practices were identified using stepwise logistic regression analysis. Potential predictors were age, sex, religious affiliation, importance of religious beliefs, recent death of a relative, specialty, type of practice (rural or urban, academic or nonacademic), amount of time spent in patient care, number of new patients in the past 6 months, and number of patients who died in the past year. RESULTS: Pediatric oncologists reported a lack of formal courses in pediatric palliative care, a strikingly high reliance on trial and error in learning to care for dying children, and a need for strong role models in this area. The lack of an accessible palliative care team or pain service was often identified as a barrier to good care. Communication difficulties exist between parents and oncologists, especially regarding the shift to end-of-life care and adequate pain control. CONCLUSION: Pediatric oncologists are working to integrate symptom control, psychosocial support, and palliative care into the routine care of the seriously ill child, although barriers exist that make such comprehensive care a challenge.
MANY DESCRIPTIONS of the unique features of pediatric palliative care have been published, detailing the obstacles to good care of dying children.1-11 Because of the rarity of childhood death, pediatric providers lack experience in dealing with death.4,9,12,13 Some identified obstacles to good care are communication difficulties,2,6,14 families’ unrealistic expectations regarding medical outcome,2 inexperience in pediatrics among hospice providers,6,7 problems with provider continuity,1 the extreme emotional burden placed on parents who must make decisions about the transition from curative to palliative care or about withdrawal of support,1,2,7 difficulties with pain assessment and control,3,15 and the unaddressed impact on providers.5,7 The responses of pediatric oncologists to this survey are relevant to the field of pediatric end-of-life care in general, because cancer is the fourth leading cause of death among children age 1 to 19 years and the leading cause of nonacute death.16 The lessons learned can be applied by all caregivers who have the opportunity to provide compassionate care to dying children and their families.
Survey Development The survey was developed by a task force composed of oncologists, nurses, oncology social workers, palliative care providers, cancer survivors, medical ethicists, and a survey methodologist (Appendix A). A subcommittee worked to refine the questions, and a draft survey was subject to cognitive and behavioral pretesting with oncologists. On the basis of this pretest, the survey instrument was revised to its final form. That survey consisted of 118 questions in eight sections: (1) education and training in end-of-life care; (2) current practice in end-of-life care; (3) perceived barriers to the delivery of end-of-life care; (4) decision making in end-of-life care; (5) vignettes about the management of patients at the end of life, including specific pediatric cases; (6) individual experiences with terminal patients; (7) the role of the American Society of Clinical Oncology (ASCO) in improving end-of-life care; and (8) demographics and practice characteristics. To facilitate applicability of the survey to both adult and pediatric providers, the questions regarding end-of-life care practices, perceived barriers, decision making, and individual experiences with terminally ill patients were written in an age-neutral fashion. Pediatric vignettes (Appendix B) were written to permit assessment of pain management and patterns of recommendation for curative versus palliative therapy for patients whose disease had relapsed. To minimize confusion, we did not use the terms "euthanasia" and "physician-assisted suicide" (PAS) in the survey. Rather, previously reported descriptions of these terms were used17,18: "to inject the patient with medication to intentionally end the patient’s life" (euthanasia) and "to provide a patient with a prescription so the patient could end his or her life by an overdose" (PAS).
Physician Eligibility and Participation Physicians were not compensated for participation. All ASCO members in the United States, Canada, and the United Kingdom were mailed a survey with a postage-paid return envelope. A reminder letter with a second survey was mailed to nonresponders 4 weeks later. Because the rate of response to written surveys is generally low, randomly selected oncologists were contacted by telephone and by mail and reminded to complete the survey. Fifty-five percent of pediatric oncologists responded to the survey.
Statistical Analysis
Demographics Two hundred twenty-eight pediatric oncologists responded to the survey; mean age was 45 years (range, 30 to 72 years). Of the responding pediatric oncologists, 80.3% practiced in an academic setting and 70.8% reported spending more than half of their time in direct patient care. Most (93.2%) reported having had fewer than 25 patients die in the preceding 12 months (median, eight patients), whereas 61.4% of adult oncologists had had more than 25 patients die in that time period.
Education and Training Pediatric oncologists reported that role models were far more useful during fellowship training than in earlier phases of training. A majority of respondents (82.3%) reported that role models did a good job of teaching how to deliver the news of a terminal prognosis, and 68.6% reported that role models did a good job of teaching how to coordinate the care of dying patients. In contrast, formal training sessions were reported as less helpful with regard to communicating with or coordinating care for dying patients, with 26.8% and 21.4% reporting formal courses as helpful in these respective areas.
Self-Assessment of Competence in End-of-Life Care Pediatric oncologists rated their communication skills high, with rankings of 4 or 5 for "communicating with dying patients and their families" (96.3%), "coordinating care of dying patients" (86.3%), and "explicitly discussing transitioning from antineoplastic treatment to palliative care" (94.6%). In addition, 83.2% rated themselves high as mentors to trainees and colleagues in these areas. Yet 68.9% of pediatric oncologists reported anxiety in association with having to tell a patient and his or her parents that the child would likely die soon; 29.1% reported feeling "very anxious." Predictors for physicians who reported feelings of anxiety were being likely to feel time constraints in one’s practice (multiple logistic regression, P = .012), male sex (P = .023), and being less likely to report oneself as competent in symptom management (P = .005). Caring for children at the end of life was reported as emotionally and intellectually satisfying by 66.0% and 44.0% of pediatric oncologists, respectively. It was reported as "the worst part of the job" by 24.5% and "the best part of the job" by 23.3%. Approximately half (49.8%) of pediatric oncologists reported some feeling of failure at the prospect of a patient dying within 6 months. These feelings were reported as being "strong to very strong" by 16.3%. In multivariate analysis, oncologists who indicated strong feelings of failure were less likely to think that providing care for dying children was the most important part of their jobs (P < .001). These same physicians also reported feeling less effective in obtaining "all or almost all" the care that their dying patients needed (P = .013). Of the oncologists who indicated strong feelings of failure, 9.5% thought they could obtain "all the care needed," 12.5% thought they could get "almost all" of this care, and 25.6% thought they could get "most or some" of what their dying pediatric patients needed.
Barriers to High-Quality Care and Specific Experiences With Dying Patients
When asked what percentage of their patients died in pain, 32.2% of pediatric oncologists reported 0%, 34.4% reported 10% to 20%, and 10% reported that half of their patients died in pain. Pediatric oncologists reported using referrals to pain or palliative care specialists a median of 50% of the time. A pain service was reported as "readily available and easy to use" by 70.2% of pediatric oncologists; 29.8% reported it as "rarely, never, or only sometimes" available. Similarly, 35.8% of pediatric oncologists reported an available and easy-to-use palliative care team, but 48.3% reported such a team as "rarely or never" available. Pediatric oncologists maintained the primary responsibility for caring for their dying patients in the last few months of life, with 72.9% reporting that they do so for 75% to 100% of their patients. Physicians surveyed were asked what percentage of advance directives were initiated by the pediatric oncologist; 47.0% responded that they initiated no such directives.
Home hospice care was reported to be less available to pediatric oncologists (77.8%) than to adult oncologists (91.1%, P Pediatric oncologists did not find reimbursement rules a barrier to providing palliative chemotherapy and radiation therapy, parenteral nutrition, or patient-controlled analgesia pumps. Third-party refusal to cover unskilled home care was reported as a significant problem by 30.3% of pediatricians, and insufficient reimbursement for coordinating the care of dying patients was reported by 52.3%.
Decision Making in End-of-Life Care
Two vignettes (Appendix B) were presented to those surveyed, to assess decision making about the transition from curative to palliative care. For the child with stage IV neuroblastoma who did not respond to second-line chemotherapy, 85.5% of pediatric oncologists recommended symptom management and hospice care and 7.0% recommended further chemotherapy. Those who recommended chemotherapy were more often those who perceived families as having unrealistic expectations (P = .013) and those who reported that end-of-life care was "the worst part of the job" (P = .022). Almost all pediatric oncologists (94.2%) recommended continued therapy for the child with primitive neuroectodermal tumor that recurred after chemoradiotherapy, with 67% of those recommending a phase I trial. The vignette continued with a description of worsening symptoms and disease, with the patient’s Karnofsky performance status score (Appendix C) decreasing to 50. At this point, approximately three-quarters of pediatric oncologists recommended hospice care and 20.1% of pediatric oncologists recommended active therapy (phase I trial: 12.3%; high-dose chemotherapy with stem-cell support: 7.8%). Predictors of the recommendation of active therapy were respondent age more than 50 years (P = .011) and an academic center practice (P = .017). Greater personal satisfaction from delivering end-of-life care was reported more often (P = .026) in the group of physicians who recommended hospice care.
A third vignette was presented to assess decision making about pain management for a child terminally ill with acute lymphoblastic leukemia (Appendix B). Most respondents (72.0%) reported a willingness to increase the dose administered by morphine drip, to optimize comfort, regardless of the risk of respiratory depression. However, 19.1% recommended administration of opioid antagonists (12.1%), nonopioids (6.5%), or a single opioid bolus (0.5%); the only predictor of nonaggressive pain control was respondent age greater than 50 years (P
Euthanasia and PAS Respondents were then asked about their actual experience with euthanasia and PAS. In terms of PAS, 20.1% of pediatric oncologists reported between one and 10 requests for PAS in a career; 79.0% reported never experiencing such a request. When asked how often they had actually provided such a prescription, 95.7% reported never having done so. However, 4.2% reported having done so for one to 10 cases throughout their careers. When asked about requests to perform euthanasia, 26.1% of pediatric oncologists reported having received such a request at some time in their careers. When asked how many times they had actually performed euthanasia, a minority (8.6%) of pediatric oncologists reported having done so for between one and five patients. Among adult oncologists, the percentage was 3.7%. Of those pediatric oncologists who reported having performed PAS or euthanasia, 96% responded to the pediatric pain management vignette in a fashion compatible with aggressive pain management despite concern about respiratory depression.
The report of this survey of the attitudes and practices of pediatric oncologists regarding the care of dying children comes at a time when end-of-life medical care is being systematically evaluated. The survey provided information about areas that need improvement in end-of-life care delivery systems for children: education, symptom management, multidisciplinary-team building, communication, and earlier involvement of hospice providers. In addition, the data from this survey about euthanasia and PAS provide an opportunity to consider the importance of communicating accurately about pain control efforts. First, provider education needs attention. The reported experiences of the surveyed pediatric oncologists verify the lack of formal courses in pediatric palliative care, show a strikingly high reliance on trial and error, and indicate the need for strong role models in this area. Pediatric residency and subspecialty fellowship program directors must incorporate end-of-life care curricula unique to pediatric patients into their programs. Although pediatric oncologists reported that formal training in end-of-life care was not useful, few experienced such training. Formal training is likely to be useful if physicians understand its importance in clinical practice and if such training includes current palliative care research and philosophy. Better use should be made of palliative care and hospice rotations in the training of pediatric providers and in continuing education classes. The majority (91%) of pediatric oncologists believe that they are competent to manage the pain of dying children, and they report that most of their patients do not die in pain. Although data regarding the prevalence of pain in dying children are sparse, a recent survey of parents whose children died from cancer found that 82% of these children experienced some pain in the last month of life.19 Other studies found that pain was experienced by nearly 50% of the assessed children with cancer.20,21 These studies also noted a discrepancy between the physician’s perception of the child’s pain and the parents’ or child’s perception of that pain.19-21 Physicians may believe that they are controlling pain when the parents or patients themselves perceive otherwise. These differences in perception may impede physician education regarding pain control. Clearly there is a need for more data and better education in the area of symptom control for dying children. Second, the data on barriers to good care indicate a need to systematize a multidisciplinary approach to the care of the dying child, with regard to medical as well as psychosocial needs. One of the barriers identified by more than a third of respondents was the lack of a readily available and easy-to-use palliative care team or pain service. Parents indicate that they want their dying children cared for by a provider with whom they have had a continuous relationship,1 and studies by palliative care and hospice professionals long ago showed the value of a multidisciplinary approach to the care of dying children.2,7 When a palliative care philosophy is practiced throughout a pediatric cancer patient’s illness, the parents of the child might be more willing to use a hospice program. Palliative care or pain service teams should be added to the standard care system. In this way, oncologists could maintain their primary relationships with terminally ill patients and have ready access to the expertise of those on palliative care or pain service teams. The survey identified physician communication as a problem area. Pediatric oncologists perceive themselves as competent at communicating with dying children and their families and at discussing the transition from curative to palliative care. Yet researchers who have studied the psychosocial concerns of bereaved parents have reported that families find physician communications vague and confusing, which can lead to anger or a feeling of responsibility for the death of one’s child.2,22,23 Pediatric oncologists reported significant anxiety about discussing the death of their patients with families and with patients themselves. The survey indicates that 47% of pediatric oncologists do not initiate a discussion of advance directives; instead, they leave this to the family to initiate. Some report a sense of failure at the prospect of a patient’s death. Although it is likely that pediatric oncologists feel the intensely emotional nature of this process and thus experience anxiety, it is also likely that they believe they are doing their best (thus the report of competence). These discrepancies clearly have implications for pediatric oncologists’ need for training and receptiveness to education in these areas. In addition, because those reporting anxiety also reported feeling rushed, a system should be developed whereby practitioners have the time to communicate thoroughly. Although the present survey shows that pediatric oncologists have ready access to psychosocial staff, it suggests that a multidisciplinary approach to terminal care has not been incorporated effectively into mainstream, routine pediatric oncology practice. This multidisciplinary approach requires a significant time commitment by physicians and team members. However, such "counseling time" is relatively poorly reimbursed.2,24 Indeed, in this survey, pediatric oncologists reported that insufficient reimbursement for coordinating the care of dying patients is a significant obstacle. This issue needs to be addressed in more detail, for its implications are broad. Third, the very acknowledgement of impending death is an issue. Physicians cited the absence of an effective therapy as the greatest impetus toward a shift from curative to palliative intent. The function and condition of the child, the presence of severe pain, and the caregiving burden on the family were less of an influence than was the availability of an effective therapy (Table 2). Responses to the vignettes about decision making during the shift from curative to palliative intent confirmed that pediatric oncologists often recommend continued treatment for patients who have had multiple relapses. Thus, both physicians and parents tend to use all available therapy options, no matter what the condition of the child. In this survey, those who recommended continued chemotherapy for a child in poor condition who had experienced multiple relapses tended to be those who perceived families as having an "unrealistic expectation for cure." When such families desire continued oncologic treatment for their children, phase I therapies should be offered concurrently with palliative care philosophy and practice. The pediatric oncologists’ responses to questions about euthanasia and PAS may reflect the confusion in the literature about the difference between allowing a child to die and performing euthanasia.3,25,26 The frequency of reported practice of euthanasia among pediatric oncologists was more than twice that of medical oncologists. Almost all (96%) who reported having practiced pediatric euthanasia or PAS (admittedly small numbers) demonstrated a willingness to control a dying child’s pain with high-dose opioids, despite concern that this might cause respiratory depression. These data could suggest that pediatric oncologists believed that adequate pain control in those cases was the equivalent of euthanasia, even though it is well documented that significant respiratory depression does not occur in children older than 1 month who are receiving high doses of morphine.3,27 This is an unfortunate area of confusion. If physicians believe they are causing a patient’s death when they are simply relieving pain, it is possible that the child’s parents and others will believe this as well.28 The opportunity to incorporate multidisciplinary, compassionate end-of-life care into mainstream pediatric practice is timely. The (merging) Children’s Cancer Group and Pediatric Oncology Group formed a committee to accomplish this in a system already organized29 to conduct excellent pediatric oncology clinical trials. In "Ensuring Quality Cancer Care," the National Cancer Policy Board recommended additional studies and initiatives to ensure high quality of care at the end of life—in particular the management of cancer-related pain and timely referral to palliative care and hospice care for adults and children.30 The World Health Organization and the International Association for the Study of Pain31 published a monograph entitled "Cancer Pain Relief and Palliative Care in Children" that includes a series of recommendations to improve care for children at the end of life. These international groups have called on pediatricians to become aware of this patient population and the existing barriers to care and have encouraged them to advocate with their local and national organizations to prioritize pain relief and palliative care for children with cancer. At the same time, Children’s Hospice International, working closely with the Health Care Financing Administration and Congress, announced a million-dollar federal appropriation for Children’s Hospice Demonstration Model Programs. The Program for All-Inclusive Care for Children and Their Families will offer an opportunity to overcome existing roadblocks to the provision of appropriate care for children and their families and will allow the provision of a continuum of care from the time of diagnosis through bereavement. At the same time, various foundations, including the Robert Wood Johnson Foundation and the Project on Death in America of the Open Society Institute, have supported multiple local and national initiatives in pediatric pain and palliative care, including the Project on Death in America’s pediatric palliative care faculty scholars program. Families requesting better end-of-life care have increased public awareness, as is well demonstrated by the advocacy efforts of Children’s Hospice International and groups such as Candlelighters. Various grassroots organizations, including the American Pain Foundation, Partnership for Caring, and Americans for the Better Care of the Dying, have drawn attention to the special needs of children for appropriate palliative care. In short, nongovernmental, governmental, and medical groups have identified the needs of children with serious life-threatening illness and are advocating for improved care for this population. The challenge for caregivers is to integrate symptom control, psychosocial support, and palliative care into the routine care of the seriously ill child. To do so, there must be a better understanding of dying and the barriers to good care that exist for these patients and their families.
Members of the American Society of Clinical Oncology Subcommittee on Cancer Care at the End of Life are as follows: Jacob D. Bitran, MD, Lutheran General Hospital, Cancer Care Center, Park Ridge, IL; Linda L. Blank, American Board of Internal Medicine, Philadelphia, PA; Diane Blum, MSW, Cancer Care, Inc, New York, NY; Dan Brock, PhD, Department of Philosophy, Brown University, Providence, RI; Eduardo Bruera, MD, Department of Symptom Control and Palliative Care, University of Texas M.D. Anderson Cancer Center, Houston, TX; Ezekiel J. Emanuel, MD, PhD, Department of Clinical Bioethics, Warren G. Magnusen Clinical Center, National Institutes of Health, Bethesda, MD; Kathleen M. Foley, MD, Memorial Sloan-Kettering Cancer Center, New York, NY; David S. Gordon, MD, Gordon Strategic Ventures, the Woodlands, TX; Joanne M. Hilden, MD, Department of Pediatric Hematology/Oncology, Children’s Hospitals and Clinics, St. Paul, MN; Robert Kagan, MD, Department of Radiation Oncology, Kaiser Permanente, Los Angeles, CA; Michael H. Levy, MD, PhD, Supportive Oncology Program, Fox Chase Cancer Center, Philadelphia, PA; Allen S. Lichter, MD, University of Michigan Medical School, Ann Arbor, MI; Michael P. Link, MD, Department of Pediatrics, Stanford University School of Medicine, Palo Alto, CA; Robert J. Mayer, MD, Dana-Farber Cancer Institute, Boston, MA; W. Charles Penley, MD, Baptist Medical Plaza I, Nashville, TN; Lowell E. Schnipper, MD, Division of Oncology, Beth Israel Deaconess Medical Center, Boston, MA; Thomas J. Smith, MD, Massey Cancer Center, Medical College of Virginia, Richmond, VA; Karen J. Stanley, RN, MSN, AOCN, Kaiser Permanente Southern California, Claremont, CA; Ellen Stovall, National Coalition for Cancer Survivorship, Silver Spring, MD; Laurence Kevin Tokaz, MD, South Austin Cancer Center, Austin, TX; and William C. Wood, MD, Department of Surgery, Emory University School of Medicine, Atlanta, GA.
Decision Making Vignette 1 A 5-year-old boy is readmitted for the second time for nausea, vomiting, and dehydration after chemotherapy for a second recurrence of stage IV neuroblastoma (n-myc amplified). Scans demonstrate progressive disease. His parents report that for the last month, he has spent most of the day in bed. He complains of intermittent severe abdominal pain. After rehydration and therapy for nausea, he is ready for discharge to home. Vignette 2 A 16-year-old girl has primitive neuroectodermal tumor of the chest wall. Her tumor has recurred after primary chemotherapy and radiation. She is an honor student, and her stated goal, and that of her parents, is to beat this disease. She has chest pain but no other symptoms, and there are no distant metastases. (Therapy choices) After 4 months of this therapy, her symptoms worsen and scans show progressive disease. Her Karnofsky performance status score is 70. The patient and family indicate a desire to continue active antineoplastic therapy. Your recommendation to the family is (therapy choices). After 2 months of this therapy, her symptoms and disease worsen. Her Karnofsky score is 50. Your recommendation to the family is (therapy choices). (Therapy choices: chemotherapy, a phase I trial, or hospice care). Pain Control Vignette 3 A 10-year-old girl has refractory acute lymphoblastic leukemia, having relapsed for the third time after therapy that included sibling-donor bone marrow transplantation. She is hospitalized for pain control. Her Lansky Play-Performance Scale score is 10 (no play; does not get out of bed). Her resuscitation status, worked out with the girl’s parents 2 weeks ago and revalidated with this admission, is do not resuscitate/do not intubate, with full fluids, transfusions, and antibiotics. Her bone pain is not well controlled with a 50-mg/h morphine drip with patient-controlled boluses. You are concerned that as you increase the morphine dose to control additional pain, she might have sufficient respiratory depression to die. She states that the pain is excruciating and she wants to die. Euthanasia and PAS Adult vignette A 63-year-old man develops metastatic prostate cancer that invades the bones and causes excruciating pain. His disease is refractory to hormonal therapy. Morphine, radiation therapy, nerve blocks, and other palliative measures are failing to control the pain completely. The patient asks to have his life ended.
Karnofsky performance status is scored as follows:
Supported by the Pine Tree Apple Tennis Classic Oncology Research Fund. We gratefully thank Drs Sarah Friebert, Joanne Wolfe, William Carroll, and Bruce Himelstein for their thoughtful reviews of the manuscript. Thanks also go to Jan Watterson for assistance in the preparation of the manuscript.
1. Martinson IM: Improving care of dying children. West J Med 163: 258-262, 1995[Medline] 2. Whittam EH: Terminal care of the dying child: Psychosocial implications of care. Cancer 71: 3450-3462, 1993[Medline] 3. Liben S: Pediatric palliative medicine: Obstacles to overcome. J Palliat Care 12: 24-28, 1996[Medline] 4. Kopecky EA, Jacobson S, Joshi P, et al: Review of a home-based palliative care program for children with malignant and non-malignant diseases. J Palliat Care 13: 28-33, 1997[Medline] 5. Goldman A: Home care of the dying child. J Palliat Care 12: 16-19, 1996[Medline] 6. Faulkner KW: Talking about death with a dying child. Am J Nurs 97:64, 66, 68-69, 1997 7. Frager G: Pediatric palliative care: Building the model, bridging the gaps. J Palliat Care 12: 9-12, 1996[Medline] 8. Freyer DR: Children with cancer: Special considerations in the discontinuation of life-sustaining treatment. Med Pediatr Oncol 20: 136-142, 1992[Medline]
9.
Khaneja S, Milrod B: Educational needs among pediatricians regarding caring for terminally ill children. Arch Pediatr Adolesc Med 152: 909-914, 1998 10. Nelson LJ, Rushton CH, Cranford RE, et al: Forgoing medically provided nutrition and hydration in pediatric patients. J Law Med Ethics 23: 33-46, 1995[Medline] 11. American Academy of Pediatrics Committee on Bioethics: Guidelines on foregoing life-sustaining medical treatment. Pediatrics 93:532-536, 1994 12. Kelly T: When your patient dies. Pediatr Manage July:15-25, 1993 13. Brown DR, Meyer CP, Shenkin B, et al: Pediatrics and death: A compassion deficit? Pediatr Manage July:26-29, 1993 14. Goldman A, Christie D: Children with cancer talk about their own death with their families. Pediatr Hematol Oncol 10: 223-231, 1993[Medline] 15. Collins JJ, Stevens MM, Cousens P: Home care for the dying child: A parent’s perception. Aust Fam Physician 27: 610-614, 1998[Medline]
16.
Landis SH, Murray T, Bolden S, et al: Cancer statistics, 1999. CA Cancer J Clin 49: 8-31, 1999 17. Emanuel EJ: Euthanasia: Historical, ethical, and empiric perspectives. Arch Intern Med 154: 1890-1901, 1994[Abstract] 18. Emanuel EJ, Fairclough DL, Daniels ER, et al: Euthanasia and physician-assisted suicide: Attitudes and experiences of oncology patients, oncologists, and the public. Lancet 347: 1805-1810, 1996[Medline]
19.
Wolfe J: Symptoms and suffering at the end of life in children with cancer. N Engl J Med 342: 326-348, 2000 20. Miser AW, Dothage JA, Wesley RA, et al: The prevalence of pain in a pediatric and young adult cancer population. Pain 29: 73-83, 1987[Medline] 21. Collins JJ, Byrnes ME, Dunkel IJ, et al: The measurement of symptoms in children with cancer. J Pain Symptom Manage 19: 363-377, 2000[Medline] 22. Martinson IM, Cohen MH: Themes from a longitudinal study of family reactions to childhood cancer. J Psychosoc Oncol 6: 81-98, 1988 23. Foley GV, Whittam EH: Care of the child dying of cancer: Part I. CA Cancer J Clin 40: 327-354, 1990[Medline] 24. Levy MH: Doctor-patient communication: The lifeline to comprehensive cancer care, ASCO Educational Book Spring: 195-202, 1998 25. Visser HK, Aartsen HG, de Beaufort ID: Medical decisions concerning the end of life in children in the Netherlands. Am J Dis Child 146: 1429-1431, 1992[Medline] 26. Orlowski JP, Smith ML, Van Zwienen J: Pediatric euthanasia. Am J Dis Child 146: 1440-1446, 1992[Abstract] 27. Schechter NL: The undertreatment of pain in children: An overview. Pediatr Clin North Am 36: 781-794, 1989 28. Wolfe J, Grier HE, Klar N, et al: Physician-assisted suicide and euthanasia: Experiences and attitudes among parents of children who have died of cancer. Proc Am Soc Clin Oncol 18: 577a, 1999 (abstr 227) 29. Bleyer WA: The U.S. pediatric cancer clinical trials programmes: International implications and the way forward. Eur J Cancer 33: 1439-1447, 1997 30. Hewitt M, Simon J (eds): Ensuring Quality Cancer Care. National Cancer Policy Board. Washington, DC, National Academy Press, 1999 31. World Health Organization and International Association for the Study of Pain: Cancer Pain Relief and Palliative Care in Children. Geneva, Switzerland, World Health Organization, 1998 Submitted May 5, 2000; accepted August 2, 2000. This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
2 test for ordered categorical data, and Wilcoxon rank sum test for continuous outcomes. Type I errors were minimized by setting alpha to 0.001 for each test. In contrast, analyses of responses limited to pediatric oncologists used a Type I error rate of 0.05. Where indicated, predictors of particular attitudes and practices were identified using stepwise logistic regression analysis; the selection criteria for entry into the model were set at alpha = 0.01. In all analyses, aspects considered in the attempt to determine predictors were age, sex, religious affiliation, importance of religious beliefs, death of a relative within the last 5 years, specialty, type of practice (rural or urban, academic or nonacademic), amount of time in patient care, number of new patients in the past 6 months, the number of patients who died in the past year, and individual answers to questions regarding clinical practice. All analyses were performed using SAS version 6.12 software (SAS Institute, Cary, NC). 
.001); the same was true of hospital-based hospice services (38.6% v 51.1%, P 
