Degree with credit for life experience

ABBREVIATIONS. CHW, Children's Hospital of Wisconsin; NICU, neonatal intensive care unit; PICU, pediatric intensive care unit; CPR, cardiopulmonary resuscitation; DNR, do not resuscitate.

There is a higher death rate among children who are younger than 1 year than in any other pediatric age group.[1] The infant mortality rate in the United States remains unchanged since the mid-1980s at approximately 7.2 deaths per 1000 live births.[2] Although some of these deaths are unexpected deaths as a result of sudden infant death syndrome or accidents, most of these infants die from a lethal anomaly, prematurity, or some other incurable condition. Despite advances in both prenatal and neonatal care, infants are still born prematurely, with lethal anomalies, or with conditions that lead to death. Understanding how these infants are being treated as they die can provide information that can lead to the development of ways in which families and their infants might be helped at the end of the infant's short life.

End-of-life care for infants has undergone relatively little analytic evaluation. In 1973, Duff and Campbell[3] questioned the practice of aggressively treating all critically ill neonates until death and suggested that withdrawing or withholding treatment may be a reasonable option. Approximately 20 years later, Wall and Partridge[4] found that physicians were willing not to treat marginally viable or severely handicapped infants. It is unclear what end-of-life care is offered to these dying infants. There have been a number of theoretical publications about how best to treat infants who cannot be cured,[5-7] but there are few analytic studies that clarify the type of end-of-life care that is being provided to them.[8-10] Statements about various types of end-of-life care for infants as being "good" or "bad" often are speculation. The Institute of Medicine states that a "decent or good death is one that is: free from avoidable distress and suffering for patients, families, and caregivers; in general accord with patients' and families' wishes; and reasonably consistent with clinical, cultural, and ethical standards."[11] Emanuel and Emanuel[12] put forward a framework of a "good death" that delineates dimensions that require interventions. These dimensions consist of the clinical status (including physical, psychological, and cognitive symptoms), social relationships and support, economic and caregiving needs, hopes and expectations, and spiritual and existential beliefs. Quantitative and qualitative analyses of current end-of-life care for infants, based on this framework of a good death, are necessary to determine what interventions should be offered to infants and their families. Establishing what constitutes good end-of-life care for infants requires that studies that describe and analyze end-of-life care be conducted.

Children's Hospital of Wisconsin (CHW) has a pediatric palliative care consultation service, making possible the exploration of how these consultations affected end-of-life care at our hospital. The goal of this study was to describe the end-of-life care received by families and their infants who were younger than 1 year at the time of their death at CHW. We hypothesized that a palliative care consultation alters the end-of-life care provided to these infants and families.

METHODS

Setting and Participants

A retrospective chart review was performed on infants who died at CHW during the period between January 1, 1994, and December 31, 1997. CHW is academically affiliated with the Medical College of Wisconsin and has a level III neonatal intensive care unit (NICU) and a 24-bed pediatric intensive care unit (PICU). CHW is the primary regional children's hospital that provides comprehensive, tertiary care services for Southeast Wisconsin, Northern Michigan, and parts of Northern Illinois. Included in this study were all infants who were younger than 1 year at the time of their death in the hospital. We chose to review only deaths that occurred in the hospital because that is the population for whom the chart review tool was validated. We do know that terminally ill infants with similar diagnosis were discharged from the hospital with the aid of our palliative care service; however, we do not have similar charting and we do not know what percentage of the total home deaths they represent. Therefore, infants whose families received palliative care consultations and went home to die were excluded from this study.

The palliative care service at CHW is an inpatient consultation service that has been described elsewhere.[13] It consists of 2 clinical nurse specialists and a physician medical director. The nurses respond to requests for consultations by working directly with the existing medical team and the patient's family. They participate by making recommendations about the environment, advanced directive planning, and medical interventions and by providing emotional support as the families cope with grief and bereavement. They report to the supervising palliative care physician regarding consultation requests and recommendations, but this physician does not become involved directly in management issues unless specifically requested to by the primary physician. The decision to implement any recommendations is made by the patient's attending physician. Once consulted, the palliative care staff follows the patient with the medical team for the duration of the admission. If the patient does not die during the hospitalization, then the palliative care service participates in discharge planning and home care through consultation with a home care agency. Whether the patient dies at home or in the hospital, the palliative care staff often help with funeral arrangements and bereavement counseling. During all 4 years of our study, the same 2 nurse specialists were the consultants for the pediatric palliative care service. This study was performed with the approval of CHW's Institutional Review Board.

Data Analysis

The data collection instrument used to perform the chart review was a modified version of the End of Life Chart Review, devised by the Center to Improve Care for the Dying.[14] The survey originally was created for adults, and our modifications consisted predominantly of replacing patient assessments and decisions with parental assessments and decisions. The categories covered in the chart review include the patients' demographics, place of death (intensive care unit or general floor), medical interventions and medicines (including analgesia) given in the last 48 hours of life, decisions to withdraw or withhold medical interventions, and psychosocial support. The chart review focuses on the description of interventions and care in the last 48 hours of life, whether it is ongoing care or withdrawing or withholding care. Specific psychosocial support included referrals to chaplains or social services and documentation of the "emotional needs of the family" when either a physician or a nurse documented anything about the families' or loved ones' emotional state. The degree of attention paid to emotions and the number of notes that mentioned this issue did not effect whether credit was given. In addition, we recorded whether a palliative care consultation was requested, including the timing of the consultation and the requesting physician.

To analyze the data gathered from the chart review, we formed 3 groups of infants. Group 1 consisted of infants whose families had a palliative care consultation, and group 2 consisted of all of the infants whose families did not have a palliative care consultation. A subgroup of group 2 (group 2a) consisted of infants who had a matching primary diagnosis to an infant from group 1 but whose families did not have a palliative care consultation.

Statistical Analysis

Data were abstracted and entered into a Microsoft Access database and analyzed with Statistical Analysis System Version 6.12 for Windows 95 (SAS Institute, Cary, NC). The student's t test for difference of means was used for interval-level variables, whereas proportions were analyzed by [chi square] test for difference of proportions. We compared group 1 first with group 2 and then with group 2a. Group 2 was not compared with group 2a because the latter is a subset of the former. Because this was an exploratory study and we analyzed many variables measured across the same study participants, we used the Bonferroni correction. The Bonferroni correction is a conservative statistical tool that adjusts the P value for statistical significance. For example, in Table 3, where the results of 12 separate analyses are reported, to achieve [Alpha] [is less than] 0.05, an adjustment of 0.05/12 is made. Therefore, with this test, a more conservative P value of P [is less than] .004 is required to achieve confidence in a statistical difference.