Journal of Perinatology (2015) 35, 218–222 © 2015 Nature America, Inc. All rights reserved /15 www.nature.com/jp
ORIGINAL ARTICLE
Impact of a palliative care program on end-of-life care in a neonatal intensive care unit N Younge, PB Smith, RN Goldberg, DH Brandon, C Simmons, CM Cotten and M Bidegain OBJECTIVE: Evaluate changes in end-of-life care following initiation of a palliative care program in a neonatal intensive care unit. STUDY DESIGN: Retrospective study comparing infant deaths before and after implementation of a Palliative Care Program comprised of medication guidelines, an individualized order set, a nursing care plan and staff education. RESULT: Eighty-two infants died before (Era 1) and 68 infants died after implementation of the program (Era 2). Morphine use was similar (88% vs 81%; P = 0.17), whereas benzodiazepines use increased in Era 2 (26% vs 43%; P = 0.03). Withdrawal of life support (73% vs 63%; P = 0.17) and do-not-resuscitate orders (46% vs 53%; P = 0.42) were similar. Do-not-resuscitate orders and family meetings were more frequent among Era 2 infants with activated palliative care orders (n = 21) compared with infants without activated orders (n = 47). CONCLUSION: End-of-life family meetings and benzodiazepine use increased following implementation of our program, likely reflecting adherence to guidelines and improved communication. Journal of Perinatology (2015) 35, 218–222; doi:10.1038/jp.2014.193; published online 23 October 2014
INTRODUCTION Approximately one third of pediatric deaths occur in the neonatal period, most commonly in the neonatal intensive care unit (NICU).1 Current data show that deaths in the NICU differ from deaths in the pediatric intensive care unit, with fewer NICU deaths occurring in children receiving mechanical ventilation.2 In fact, the majority of NICU deaths follow withdrawal or withholding of life-sustaining interventions.2–10 A recent policy statement by the American Academy of Pediatrics (AAP) defines pediatric palliative care as treatments that aim to (1) relieve suffering across multiple realms; (2) improve children’s quality and enjoyment of life while helping families adapt and function during the illness and through bereavement; (3) facilitate informed decision making by patients, families and health care professionals; and (4) assist with ongoing coordination of care among clinicians and across various sites of care.11 A decade ago the AAP and the Institute of Medicine identified an urgent need to improve and expand palliative care and end-of-life programs for all children.1,12 In response, pediatric palliative care programs have become increasingly common in United States children’s hospitals.13 Many benefits of pediatric palliative care have been reported, including a reduction in hospital days, reduction in medical costs and improvement in families’ quality of life.14 However, little is known about the impact and implementation of palliative care practices in the NICU and palliative care remains underutilized in neonates.15,16 Research that quantifies aspects of how palliative care programs optimize end-oflife care for infants and improve family outcomes is needed. The purpose of this study was to evaluate the impact of Neonatal Palliative Care Program on end-of-life care in the NICU at Duke University Medical Center. METHODS We identified all infants who died in the NICU at Duke University Medical Center, an academic tertiary care center in two different Eras. Era 1 (2003 to
2005) included infants who died before the initiation of the Neonatal Palliative Care Program in February 2007. Era 2 (2008 to 2009) included infants who died following the program’s implementation. The year immediately preceding and the first year following program implementation (2006 and 2007) were excluded to minimize contamination. The program includes a palliative care protocol, an electronic order set that can be individualized, a nursing plan of care, staff education and medication guidelines for palliative and end-of-life care including use of sedatives, analgesics and muscle relaxants (Table 1). Many aspects of our protocol were adopted from previously published neonatal guidelines.17,18 Our program was designed by an interdisciplinary team of neonatologists, advanced practice nurses, bedside nurses, social workers, pharmacologists and chaplains and included a neonatologist who is board-certified in Hospice and Palliative Medicine. The Duke NICU is a 67-bed unit that serves as a major referral center for premature and term infants with complications of prematurity, respiratory failure, persistent pulmonary hypertension, perinatal asphyxia and other complex medical and surgical problems. There are ~ 850 admissions per year of which 30% are outborn. We analyzed differences in demographics, cause of and age at death, number of family meetings, morphine dosage (total micrograms per kilogram received in the last 24 h of life), use of benzodiazepines (any use within the last 4 days of life), use of neuromuscular blockers (NMB; any use of paralytic agents within the last 24 h of life), do-not-resuscitate (DNR) orders (active DNR order in chart at the time of death) and withdrawal of life support (WLS) for infants who died in Era 1 vs those who died in Era 2. WLS was defined as withdrawal of respiratory support or other lifesustaining treatments, such as vasoactive medications. Clinical conditions at death were analyzed and compared between the Eras: active cardiopulmonary resuscitation; active DNR order in the chart; WLS; or brain death. Infants were excluded from the study if they had missing charts; if they died within 24 h of birth, as these deaths may occur quickly, mostly in the delivery room with limited time to establish a palliative care plan; or if they had a diagnosis of brain death, due to absence of neurologic function. Brain death was diagnosed by Duke’s Institutional Brain Death Criteria for newborn infants. We also performed a subgroup analysis of Era 2 infants with activated palliative care order sets. As part of the palliative care program, providers have an option to activate an electronic ‘palliative care order set’ at any
Division of Neonatal-Perinatal Medicine, Department of Pediatrics, Duke University Medical Center, Durham, NC, USA. Correspondence: Dr M Bidegain, Division of NeonatalPerinatal Medicine, Department of Pediatrics, Duke University Medical Center, Hock Plaza, Suite 504, Box 2739, 2424 Erwin Road, Durham, NC 27705, USA. E-mail: [email protected] Received 13 August 2014; accepted 11 September 2014; published online 23 October 2014
Impact of a neonatal palliative care program N Younge et al
219 Table 1.
Overview of Neonatal Palliative Care Program
Program component
Description
Neonatal palliative care protocol
-Describes method of identifying infants (prenatal and postnatal) and lists conditions in which palliative care may be beneficial. -Clinical guidelines for antenatal care, delivery planning and provision of palliative care in the neonatal period. -Provides recommendations for structure and content of family meetings, including a list of recommended topics for end-of-life care discussions. -Provides guidance on appropriate language for discussion of palliative and end-of-life care. -Includes recommended dosage, frequency and route of administration for medications commonly used in palliative and end-of-life care, including analgesics and sedatives. -The clinician providing palliative care for an infant can activate the electronic order set. -Includes recommendations on medication use, clinical and comfort care.
Medication guide for palliative and end-of-life care Palliative care order set
time during the patient’s admission. To activate the palliative care order set, the providers need prior agreement among themselves and with the family. Agreement to activate this order set is usually reached during weekly team morbidity and mortality conferences and/or during family meetings. We compared this group of infants with activated order sets to infants in Era 2 who did not have palliative care order sets activated. Finally, to evaluate the effect of the program on very low birth weight infants, we performed an analysis of deaths in infants weighing o1500 g at birth. Data analyses were performed with Wilcoxon rank-sum test for continuous data and Fisher’s exact tests for categorical data. The study was approved by the Institutional Review Board at the Duke University Hospital.
RESULTS Infant deaths: Era 1 vs Era 2 Overall mortality in the NICU did not change significantly over the study period. Deaths/number of admissions were 105/1982 (5.3%) in Era 1 compared with 85/1644 (5.1%) in Era 2 (P = 0.86). Of 105 infants who died during Era 1, 82 were included in the analysis. Eighteen were excluded due to death within 24 postnatal hours and five were excluded due to missing charts. Of 85 infants who died during Era 2, 68 were included for analysis; 16 infants were excluded due to death within the first 24 postnatal hours and 1 was excluded due to brain death. For the 150 infants included in the analysis, the most frequent cause of death was complications of prematurity (51%), followed by congenital anomalies (39%), hypoxic–ischemic encephalopathy (9%) and respiratory failure (1%). Infants from Era 1 and Era 2 did not differ in demographics, illness characteristics or cause of death (Table 2). Median Apgar scores were lower in Era 2. Postnatal age at death was similar between the two groups. There was no significant group difference in the incidence of WLS (Table 3). Withdrawal of respiratory support was the method of WLS in 101/102 (99%) of infants. Median time to death after WLS was 1 h (range 0, 62) and did not differ significantly between Eras. Most patients who died in both Eras had at least one family meeting to discuss end-of-life care. The incidence of DNR orders also did not change significantly following implementation of the program. Active cardiopulmonary resuscitation at the time of death, including chest compressions, occurred in 9/82 (11%) of infants in Era 1 and 9/66 (14%) of infants in Era 2 (P = 0.62). The amount of morphine administered in the last 24 h of life did not differ between groups (Table 3). In addition, the proportion of infants receiving morphine in the last 4 days of life remained similar. A greater proportion of infants in Era 2 received benzodiazepines in the 4 days preceding death. There was no significant difference in the number of infants receiving neuromuscular blockade on the day of death. Infant deaths in Era 2: use of the palliative care order set The palliative care order set was activated in 21 (31%) of the 68 infants who died in Era 2. When compared with infants in Era 2 © 2015 Nature America, Inc.
without activated order sets, this subgroup of infants did not differ significantly in birth weight, weight at death, gestational age, gender, inborn status, race, Apgar scores or cause of death (Table 2). There was a significantly higher incidence of DNR orders in the group with activated order sets (Table 3). The number of meetings to discuss end-of-life care was significantly higher among the infants with activated order sets and the time between the first end-of-life meeting and WLS was longer. Median time to death after WLS was longer for infants with activated order sets. There was no significant difference in morphine use between infants with and without activated order sets, but a significantly higher proportion of infants with activated order sets received benzodiazepines in the last 4 days of life (Table 3). Fewer infants received neuromuscular blockade in the last day of life in the activated order set group, but the difference did not reach statistical significance. Infant deaths in very low birth weight (birth weight o 1500 g) infants: Era 1 vs Era 2 A subgroup analysis of infants weighing o1500 g at birth showed no significant difference in the incidence of WLS between Era 1 and Era 2 (75% vs 59%; P = 0.10), DNR orders (45% vs 43%; P = 0.83) and median time between first end-of-life meeting and actual WLS (0 (0, 23) vs 0 (0, 29) days; P = 0.66). There was no significant difference in the use of morphine (88% vs 81%; P = 0.26), benzodiazepines (20% vs 36%; P = 0.08) or NMB (20% vs 19%; P = 0.95) at the end-of-life. There was a higher incidence of DNR orders in infants o1500 g who had activated palliative care order sets when compared with infants without activated order sets (80% vs 31%; P = 0.007), but no difference in incidence of WLS (56% vs 59%; P = 0.84), median morphine dose on the day of death [929 (0 to 2628) vs 121 (0 to 3699) μg/kg; P = 0.11], use of benzodiazepines in the last 4 days of life (50% vs 31%; P = 0.28), or use of NMB in the last 24 h of life (10% vs 22% P = 0.40). DISCUSSION Although implementation of Neonatal Palliative Care Programs has been described previously,17–19 our study is one of the first to evaluate the impact of such programs on end-of-life care in the NICU. In our center, implementation of a Neonatal Palliative Care Program did not significantly change the incidence of WLS or overall mortality in the NICU. Consistent with previous findings,20 there was no increase in the incidence of WLS among infants receiving palliative care. DNR orders were more frequent among infants for whom the electronic palliative care order set was activated, likely because the guidelines of our palliative care program recommend considering discussions of limitations of resuscitation in infants receiving end-of-life care. Previous studies have reported WLS to precede 42 to 95% of deaths in the NICU.3–8,10,15,21 The proportion of deaths in the NICU accompanied by withdrawal or withholding of support appears to have Journal of Perinatology (2015), 218 – 222
Impact of a neonatal palliative care program N Younge et al
220 Table 2.
Infant characteristics
Birth weight (g), median (range) Weight at death (g), median (range) GA (weeks), median (range) Male, n (%) Inborn, n (%)
Era 1 pre-PC program N = 82
Era 2 post-PC program N = 68
P-value
Era 2 Activated PC orders N = 21
Era 2 No PC orders N = 47
P-value
928 (390, 3960) 1435 (429, 6020) 28 (22, 40) 47 (59) 48 (59)
1195 (330, 3331) 1498 (466, 4870) 29 (23, 40) 40 (60) 43 (63)
0.96 0.61 0.63 40.99 0.62
1530 (400, 3331) 1670 (516, 4130) 34 (23, 39) 14 (67) 13 (62)
1100 (330, 3095) 1345 (466, 4870) 28 (23, 40) 26 (56) 30 (64%)
0.17 0.24 0.09 0.31 40.99
19 20 3 0
(45) (48) (7) (0)
0.15
Ethnicity, n (%) Black, not Hispanic White, not Hispanic Hispanic Other
37 34 7 3
(46) (42) (9) (3)
27 (44) 26 (43) 7 (11) 1 (2)
0.98
8 (42) 6 (32) 4 (21) 1 (5)
Apgar scores, median (range) 1 min 5 min
3 (0, 8) 7 (0, 9)
2 (0, 8) 6 (0, 9)
0.02 o0.01
4 (0, 8) 6 (0, 9)
2 (0, 8) 6 (1, 9)
0.31 0.23
45 (55) 30 (37) 6 (7) 1 (1) 13 (1, 287)
32 (47) 28 (41) 7 (10) 1 (2) 16 (1, 225)
0.81 0.55
6 (29) 13 (62) 2 (10) 0 (0) 12 (1, 137)
26 (55) 15 (32) 5 (11) 1 (2) 16 (1, 225)
0.09 0.80
Cause of death, n (%) Complications of prematurity Congenital anomalies HIE Respiratory failure Age at death (days), median (range)
Abbreviations: GA, gestational age; HIE, hypoxic–ischemic encephalopathy; PC, palliative care.
Table 3.
End-of-life care for infants before and after implementation of Neonatal Palliative Care Program
WLS incidence, n (%) DNR incidence, n (%) At least one family meeting, n (%) Number of end-of-life family meetings, median (range) Time between first end-of-life meeting and WLS (days), median (range) Time to death after WLS (h), median (range) Patients receiving opioids last 4 days, n (%) Morphine dose last 24 h (μg/kg), median (range) Benzodiazepines used last 4 days, n (%) Neuromuscular blocker use last 24 h, n (%)
Era 1 pre-PC program N = 82
Era 2 post-PC program N = 68
P-value
Era 2 Activated PC orders N = 21
Era 2 no PC orders N = 47
P-value
60 (73) 38 (46) 72 (88) 1 (0, 5) 0 (0, 23)
42 (63) 36 (53) 61 (91) 2 (0, 5) 1 (0, 29)
0.17 0.42 0.36 0.13 0.47
13 (65) 18 (85) 20 (95) 2 (1,4) 1 (0, 28)
29 (62) 18 (38) 41 (87) 1 (0,5) 0 (0, 29)
0.80 o 0.01 0.17 o 0.01 0.03
1 (0, 62) 72 (88) 480 (0, 2879) 21 (26) 14 (17)
1 (0, 41) 55 (81) 179 (0, 9492) 29 (43) 12 (18)
0.17 0.17 0.17 0.03 0.95
7 (0, 41) 19 (90) 302 (0, 2987) 13 (62) 1 (5)
1 (0, 20) 36 (77) 92 (0, 9492) 16 (34) 11 (23)
0.02 0.32 0.15 0.03 0.06
Abbreviations: DNR, do-not-resuscitate; PC, palliative care; WLS, withdrawal of life support.
increased over time.7–10,21 The most commonly cited reason for WLS is imminent or inevitable death, followed by concerns for future quality of life or suffering.4,5,22 End-of-life care requires complex decision making by parents and providers, and the importance of providing palliative care in this setting has been increasingly recognized.1,12 A policy statement on the care of high-risk newborn infants from the American Academy of Pediatrics emphasizes the importance of continuing comfort care measures even when treatment goals no longer involve intensive care interventions.23 However, reports indicate that the use of palliative care services in the NICU is uncommon. Single-center studies report palliative care interventions in only 13 to 24% of infant deaths.15,16,20 The reasons for this limitation are complex and multifactorial,24,25 but the results of our study provide assurance that offering palliative care services does not lead to an increased incidence of withdrawal of life support or death. Journal of Perinatology (2015), 218 – 222
One limitation of our study was exclusion of infants who died within the first 24 h of life. When we designed the study, we were concerned that there would be inadequate time to establish palliative care and utilize the palliative care order set in infants who died shortly after birth. However, since implementation of our program, we have seen that the use the palliative care protocol in infants with an antenatal diagnosis of a major congenital anomaly has been very helpful in early initiation of palliative care. Our study also evaluated the effect of a Neonatal Palliative Care Program on use of sedatives and analgesics at the end-of-life. Previous studies have reported sedative or analgesic use in 66 to 87% of infants at the time of or subsequent to WLS.8,26,27 In our study, the majority of infants received opioids before death in both Eras, at 88% and 81%, respectively. We noted an increase in the use of benzodiazepines in the post-implementation Era (26% © 2015 Nature America, Inc.
Impact of a neonatal palliative care program N Younge et al
vs 43%, P = 0.03). These results are consistent with a recent study from Janvier et al.27 reporting the use of comfort medications in the last 48 h of life in two US NICUs (Chicago and Milwaukee) with 66% and 72% of infants receiving opioids, and 52% and 53% receiving benzodiazepines, respectively. Our increase in the use of benzodiazepines was likely related to the medication guidelines recommended by our program, with recommendations on benzodiazepine use at the end-of-life. Although no studies have evaluated the effects of treatment with benzodiazepines in infants at the end-of-life, they are frequently used in adult palliative care to treat agitation and anxiety.28 Our study did not include evaluation of clinical response to comfort medications at the endof-life. NMB should not be given at the time of WLS as they are not comfort medications, they interfere with respiratory function and they could mask signs of pain and distress. In our study, a minority of infants received NMB in the last 24 h of life in both Eras (17% vs 18%, P = 0.95). However, only 3 of the 150 patients were on NMB at the time of withdrawal (2 patients from Era 1 and 1 patient from Era 2). Our palliative care program recommends discontinuing NMB for a period of hours before WLS to allow the paralytic effect to subside. However, in some cases the patient’s condition deteriorates rapidly with multisystemic compromise, making it impossible to completely wean the patient off NMB before WLS. Consistent with clinical practice in other NICUs in the United States, in no patients were NMB initiated at the time of withdrawal.27 Family meetings to discuss end-of-life care occurred more frequently in the subgroup of infants with activated palliative care order sets, likely reflecting improved communication among the health care team and family preceding the infant’s death. The fact that the time between the first end-of-life discussions and actual withdrawal was longer in the patients with activated order sets may also indicate that end-of-life discussions were initiated earlier in the course of disease, providing further evidence of improved communication. Although it is not clear why time to death after WLS was longer for the infants with activated order sets, we suspect that this trend was attributable to the higher incidence of congenital anomalies in this group of infants (Table 2). We hypothesize that WLS more commonly occurred owing to concern for future quality of life in the infants with congenital anomalies as opposed to severe illness or imminent death. Therefore, these infants may have been physiologically more stable at the time of WLS resulting in longer time to death. Alternatively, it is possible that the use of palliative care orders, including early use and appropriate dosing of sedatives and analgesics, prolonged survival in these infants. Previous studies in infants and adults have shown that use of these medications after WLS does not hasten time to death and, in some cases, has been associated with increased survival time.27,29 CONCLUSION Despite the frequency of death in the NICU, providing optimal care for infants at the end-of-life presents many challenges and complexities. Little is known about the impact of Neonatal Palliative Care Programs in the neonatal intensive care setting. Our study shows that establishment of a Neonatal Palliative Care Program did not impact the incidence of WLS or overall mortality, but did have an effect on the way medications were used at the end-of-life. We also found evidence of earlier communication with families and increased frequency of end-of-life family meetings. Although further studies are needed to assess the effect of Neonatal Palliative Care Programs on other metrics such as provider and parent satisfaction, our findings suggest that end-oflife care for newborn infants can be improved by adopting a teamdriven, patient- and family-centered, consistent approach. © 2015 Nature America, Inc.
221
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Kimberley A Fisher, PhD, and Carolyn Jones, MSN, NNP for their expert technical contributions (financial support provided by the Jean and George W Brumley, Jr Neonatal Perinatal Research Institute). Dr Smith receives salary support for research from the NIH and the US Department of Health and Human Services (government contract HHSN267200700051C, HHSN275201000003I, and UL1TR001117). Dr. Cotten received funding from NICHD: 5U10HD040492-10, from NHLBI: 1R01HL10570201A1 and from NIDDK: NIH-NIDDK203-2345. Dr. Brandon received funding from NIH-NINRR01NR010548.
REFERENCES 1 Field MJ, Behrman RE (eds). When Children Die: Improving Palliative and End-of-Life Care for Children and Their Families. National Academies Press: Washington, DC, USA, 2003. 2 Fontana MS, Farrell C, Gauvin F, Lacroix J, Janvier A. Modes of death in pediatrics: differences in the ethical approach in neonatal and pediatric patients. J Pediatr 2013; 162(6): 1107–1111. 3 Roy R, Aladangady N, Costeloe K, Larcher V. Decision making and modes of death in a tertiary neonatal unit. Arch Dis Child Fetal Neonatal Ed 2004; 89(6): F527–F530. 4 Verhagen AA, Dorscheidt JH, Engels B, Hubben JH, Sauer PJ. End-of-life decisions in Dutch neonatal intensive care units. Arch Pediatr Adolesc Med 2009; 163(10): 895–901. 5 Wall SN, Partridge JC. Death in the intensive care nursery: physician practice of withdrawing and withholding life support. Pediatrics 1997; 99(1):64–70. 6 Cook LA, Watchko JF. Decision making for the critically ill neonate near the end of life. J Perinatol 1996; 16(2 Pt 1): 133–136. 7 Barton L, Hodgman JE. The contribution of withholding or withdrawing care to newborn mortality. Pediatrics 2005; 116(6): 1487–1491. 8 Wilkinson DJ, Fitzsimons JJ, Dargaville PA, Campbell NT, Loughnan PM, McDougall PN et al. Death in the neonatal intensive care unit: changing patterns of end of life care over two decades. Arch Dis Child Fetal Neonatal Ed 2006; 91(4): F268–F271. 9 Hagen CM, Hansen TW. Deaths in a neonatal intensive care unit: a 10-year perspective. Pediatr Crit Care Med 2004; 5(5): 463–468. 10 Weiner J, Sharma J, Lantos J, Kilbride H. How infants die in the neonatal intensive care unit: trends from 1999 through 2008. Arch Pediatr Adolesc Med 2011; 165(7): 630–634. 11 Section on Hospice and Palliative Medicine and Committee on Hospital Care. Pediatric palliative care and hospice care commitments, guidelines, and recommendations. Pediatrics 2013; 132(5): 966–972. 12 American Academy of Pediatrics. Committee on Bioethics and Committee on Hospital Care Palliative care for children. Pediatrics 2000; 106(2 Pt 1): 351–357. 13 Feudtner C, Womer J, Augustin R, Remke S, Wolfe J, Friebert S et al. Pediatric palliative care programs in children's hospitals: a cross-sectional national survey. Pediatrics 2013; 132(6): 1063–1070. 14 Gans D, Kominski GF, Roby DH, Diamant AL, Chen X, Lin W et al. Better outcomes, lower costs: palliative care program reduces stress, costs of care for children with life-threatening conditions. Policy Brief UCLA Cent Health Policy Res 2012; (PB2012-3): 1–8. 15 Brandon D, Docherty SL, Thorpe J. Infant and child deaths in acute care settings: implications for palliative care. J Palliat Med 2007; 10(4): 910–918. 16 Leuthner SR, Pierucci R. Experience with neonatal palliative care consultation at the Medical College of Wisconsin-Children's Hospital of Wisconsin. J Palliat Med 2001; 4(1): 39–47. 17 Catlin A, Carter B. Creation of a neonatal end-of-life palliative care protocol. J Perinatol 2002; 22(3): 184–195. 18 Carter BS, Bhatia J. Comfort/palliative care guidelines for neonatal practice: development and implementation in an academic medical center. J Perinatol 2001; 21(5): 279–283. 19 Gale G, Brooks A. Implementing a palliative care program in a newborn intensive care unit. Adv Neonatal Care 2006; 6(1): 37–53. 20 Pierucci RL, Kirby RS, Leuthner SR. End-of-life care for neonates and infants: the experience and effects of a palliative care consultation service. Pediatrics 2001; 108(3): 653–660. 21 Singh J, Lantos J, Meadow W. End-of-life after birth: death and dying in a neonatal intensive care unit. Pediatrics 2004; 114(6): 1620–1626. 22 Hentschel R, Lindner K, Krueger M, Reiter-Theil S. Restriction of ongoing intensive care in neonates: a prospective study. Pediatrics 2006; 118(2): 563–569. 23 Bell EF. Noninitiation or withdrawal of intensive care for high-risk newborns. Pediatrics 2007; 119(2): 401–403.
Journal of Perinatology (2015), 218 – 222
Impact of a neonatal palliative care program N Younge et al
222 24 Davies B, Sehring SA, Partridge JC, Cooper BA, Hughes A, Philp JC et al. Barriers to palliative care for children: perceptions of pediatric health care providers. Pediatrics 2008; 121(2): 282–288. 25 Kain V, Gardner G, Yates P. Neonatal palliative care attitude scale: development of an instrument to measure the barriers to and facilitators of palliative care in neonatal nursing. Pediatrics 2009; 123(2): e207–e213. 26 Partridge JC, Wall SN. Analgesia for dying infants whose life support is withdrawn or withheld. Pediatrics 1997; 99(1): 76–79.
Journal of Perinatology (2015), 218 – 222
27 Janvier A, Meadow W, Leuthner SR, Andrews B, Lagatta J, Bos A et al. Whom are we comforting? An analysis of comfort medications delivered to dying neonates. J Pediatr 2011; 159(2): 206–210. 28 Henderson M, MacGregor E, Sykes N, Hotopf M. The use of benzodiazepines in palliative care. Palliat Med 2006; 20(4): 407–412. 29 Chan JD, Treece PD, Engelberg RA, Crowley L, Rubenfeld GD, Steinberg KP et al. Narcotic and benzodiazepine use after withdrawal of life support: association with time to death? Chest 2004; 126(1): 286–293.
© 2015 Nature America, Inc.
ORIGINAL ARTICLE
Impact of a palliative care program on end-of-life care in a neonatal intensive care unit N Younge, PB Smith, RN Goldberg, DH Brandon, C Simmons, CM Cotten and M Bidegain OBJECTIVE: Evaluate changes in end-of-life care following initiation of a palliative care program in a neonatal intensive care unit. STUDY DESIGN: Retrospective study comparing infant deaths before and after implementation of a Palliative Care Program comprised of medication guidelines, an individualized order set, a nursing care plan and staff education. RESULT: Eighty-two infants died before (Era 1) and 68 infants died after implementation of the program (Era 2). Morphine use was similar (88% vs 81%; P = 0.17), whereas benzodiazepines use increased in Era 2 (26% vs 43%; P = 0.03). Withdrawal of life support (73% vs 63%; P = 0.17) and do-not-resuscitate orders (46% vs 53%; P = 0.42) were similar. Do-not-resuscitate orders and family meetings were more frequent among Era 2 infants with activated palliative care orders (n = 21) compared with infants without activated orders (n = 47). CONCLUSION: End-of-life family meetings and benzodiazepine use increased following implementation of our program, likely reflecting adherence to guidelines and improved communication. Journal of Perinatology (2015) 35, 218–222; doi:10.1038/jp.2014.193; published online 23 October 2014
INTRODUCTION Approximately one third of pediatric deaths occur in the neonatal period, most commonly in the neonatal intensive care unit (NICU).1 Current data show that deaths in the NICU differ from deaths in the pediatric intensive care unit, with fewer NICU deaths occurring in children receiving mechanical ventilation.2 In fact, the majority of NICU deaths follow withdrawal or withholding of life-sustaining interventions.2–10 A recent policy statement by the American Academy of Pediatrics (AAP) defines pediatric palliative care as treatments that aim to (1) relieve suffering across multiple realms; (2) improve children’s quality and enjoyment of life while helping families adapt and function during the illness and through bereavement; (3) facilitate informed decision making by patients, families and health care professionals; and (4) assist with ongoing coordination of care among clinicians and across various sites of care.11 A decade ago the AAP and the Institute of Medicine identified an urgent need to improve and expand palliative care and end-of-life programs for all children.1,12 In response, pediatric palliative care programs have become increasingly common in United States children’s hospitals.13 Many benefits of pediatric palliative care have been reported, including a reduction in hospital days, reduction in medical costs and improvement in families’ quality of life.14 However, little is known about the impact and implementation of palliative care practices in the NICU and palliative care remains underutilized in neonates.15,16 Research that quantifies aspects of how palliative care programs optimize end-oflife care for infants and improve family outcomes is needed. The purpose of this study was to evaluate the impact of Neonatal Palliative Care Program on end-of-life care in the NICU at Duke University Medical Center. METHODS We identified all infants who died in the NICU at Duke University Medical Center, an academic tertiary care center in two different Eras. Era 1 (2003 to
2005) included infants who died before the initiation of the Neonatal Palliative Care Program in February 2007. Era 2 (2008 to 2009) included infants who died following the program’s implementation. The year immediately preceding and the first year following program implementation (2006 and 2007) were excluded to minimize contamination. The program includes a palliative care protocol, an electronic order set that can be individualized, a nursing plan of care, staff education and medication guidelines for palliative and end-of-life care including use of sedatives, analgesics and muscle relaxants (Table 1). Many aspects of our protocol were adopted from previously published neonatal guidelines.17,18 Our program was designed by an interdisciplinary team of neonatologists, advanced practice nurses, bedside nurses, social workers, pharmacologists and chaplains and included a neonatologist who is board-certified in Hospice and Palliative Medicine. The Duke NICU is a 67-bed unit that serves as a major referral center for premature and term infants with complications of prematurity, respiratory failure, persistent pulmonary hypertension, perinatal asphyxia and other complex medical and surgical problems. There are ~ 850 admissions per year of which 30% are outborn. We analyzed differences in demographics, cause of and age at death, number of family meetings, morphine dosage (total micrograms per kilogram received in the last 24 h of life), use of benzodiazepines (any use within the last 4 days of life), use of neuromuscular blockers (NMB; any use of paralytic agents within the last 24 h of life), do-not-resuscitate (DNR) orders (active DNR order in chart at the time of death) and withdrawal of life support (WLS) for infants who died in Era 1 vs those who died in Era 2. WLS was defined as withdrawal of respiratory support or other lifesustaining treatments, such as vasoactive medications. Clinical conditions at death were analyzed and compared between the Eras: active cardiopulmonary resuscitation; active DNR order in the chart; WLS; or brain death. Infants were excluded from the study if they had missing charts; if they died within 24 h of birth, as these deaths may occur quickly, mostly in the delivery room with limited time to establish a palliative care plan; or if they had a diagnosis of brain death, due to absence of neurologic function. Brain death was diagnosed by Duke’s Institutional Brain Death Criteria for newborn infants. We also performed a subgroup analysis of Era 2 infants with activated palliative care order sets. As part of the palliative care program, providers have an option to activate an electronic ‘palliative care order set’ at any
Division of Neonatal-Perinatal Medicine, Department of Pediatrics, Duke University Medical Center, Durham, NC, USA. Correspondence: Dr M Bidegain, Division of NeonatalPerinatal Medicine, Department of Pediatrics, Duke University Medical Center, Hock Plaza, Suite 504, Box 2739, 2424 Erwin Road, Durham, NC 27705, USA. E-mail: [email protected] Received 13 August 2014; accepted 11 September 2014; published online 23 October 2014
Impact of a neonatal palliative care program N Younge et al
219 Table 1.
Overview of Neonatal Palliative Care Program
Program component
Description
Neonatal palliative care protocol
-Describes method of identifying infants (prenatal and postnatal) and lists conditions in which palliative care may be beneficial. -Clinical guidelines for antenatal care, delivery planning and provision of palliative care in the neonatal period. -Provides recommendations for structure and content of family meetings, including a list of recommended topics for end-of-life care discussions. -Provides guidance on appropriate language for discussion of palliative and end-of-life care. -Includes recommended dosage, frequency and route of administration for medications commonly used in palliative and end-of-life care, including analgesics and sedatives. -The clinician providing palliative care for an infant can activate the electronic order set. -Includes recommendations on medication use, clinical and comfort care.
Medication guide for palliative and end-of-life care Palliative care order set
time during the patient’s admission. To activate the palliative care order set, the providers need prior agreement among themselves and with the family. Agreement to activate this order set is usually reached during weekly team morbidity and mortality conferences and/or during family meetings. We compared this group of infants with activated order sets to infants in Era 2 who did not have palliative care order sets activated. Finally, to evaluate the effect of the program on very low birth weight infants, we performed an analysis of deaths in infants weighing o1500 g at birth. Data analyses were performed with Wilcoxon rank-sum test for continuous data and Fisher’s exact tests for categorical data. The study was approved by the Institutional Review Board at the Duke University Hospital.
RESULTS Infant deaths: Era 1 vs Era 2 Overall mortality in the NICU did not change significantly over the study period. Deaths/number of admissions were 105/1982 (5.3%) in Era 1 compared with 85/1644 (5.1%) in Era 2 (P = 0.86). Of 105 infants who died during Era 1, 82 were included in the analysis. Eighteen were excluded due to death within 24 postnatal hours and five were excluded due to missing charts. Of 85 infants who died during Era 2, 68 were included for analysis; 16 infants were excluded due to death within the first 24 postnatal hours and 1 was excluded due to brain death. For the 150 infants included in the analysis, the most frequent cause of death was complications of prematurity (51%), followed by congenital anomalies (39%), hypoxic–ischemic encephalopathy (9%) and respiratory failure (1%). Infants from Era 1 and Era 2 did not differ in demographics, illness characteristics or cause of death (Table 2). Median Apgar scores were lower in Era 2. Postnatal age at death was similar between the two groups. There was no significant group difference in the incidence of WLS (Table 3). Withdrawal of respiratory support was the method of WLS in 101/102 (99%) of infants. Median time to death after WLS was 1 h (range 0, 62) and did not differ significantly between Eras. Most patients who died in both Eras had at least one family meeting to discuss end-of-life care. The incidence of DNR orders also did not change significantly following implementation of the program. Active cardiopulmonary resuscitation at the time of death, including chest compressions, occurred in 9/82 (11%) of infants in Era 1 and 9/66 (14%) of infants in Era 2 (P = 0.62). The amount of morphine administered in the last 24 h of life did not differ between groups (Table 3). In addition, the proportion of infants receiving morphine in the last 4 days of life remained similar. A greater proportion of infants in Era 2 received benzodiazepines in the 4 days preceding death. There was no significant difference in the number of infants receiving neuromuscular blockade on the day of death. Infant deaths in Era 2: use of the palliative care order set The palliative care order set was activated in 21 (31%) of the 68 infants who died in Era 2. When compared with infants in Era 2 © 2015 Nature America, Inc.
without activated order sets, this subgroup of infants did not differ significantly in birth weight, weight at death, gestational age, gender, inborn status, race, Apgar scores or cause of death (Table 2). There was a significantly higher incidence of DNR orders in the group with activated order sets (Table 3). The number of meetings to discuss end-of-life care was significantly higher among the infants with activated order sets and the time between the first end-of-life meeting and WLS was longer. Median time to death after WLS was longer for infants with activated order sets. There was no significant difference in morphine use between infants with and without activated order sets, but a significantly higher proportion of infants with activated order sets received benzodiazepines in the last 4 days of life (Table 3). Fewer infants received neuromuscular blockade in the last day of life in the activated order set group, but the difference did not reach statistical significance. Infant deaths in very low birth weight (birth weight o 1500 g) infants: Era 1 vs Era 2 A subgroup analysis of infants weighing o1500 g at birth showed no significant difference in the incidence of WLS between Era 1 and Era 2 (75% vs 59%; P = 0.10), DNR orders (45% vs 43%; P = 0.83) and median time between first end-of-life meeting and actual WLS (0 (0, 23) vs 0 (0, 29) days; P = 0.66). There was no significant difference in the use of morphine (88% vs 81%; P = 0.26), benzodiazepines (20% vs 36%; P = 0.08) or NMB (20% vs 19%; P = 0.95) at the end-of-life. There was a higher incidence of DNR orders in infants o1500 g who had activated palliative care order sets when compared with infants without activated order sets (80% vs 31%; P = 0.007), but no difference in incidence of WLS (56% vs 59%; P = 0.84), median morphine dose on the day of death [929 (0 to 2628) vs 121 (0 to 3699) μg/kg; P = 0.11], use of benzodiazepines in the last 4 days of life (50% vs 31%; P = 0.28), or use of NMB in the last 24 h of life (10% vs 22% P = 0.40). DISCUSSION Although implementation of Neonatal Palliative Care Programs has been described previously,17–19 our study is one of the first to evaluate the impact of such programs on end-of-life care in the NICU. In our center, implementation of a Neonatal Palliative Care Program did not significantly change the incidence of WLS or overall mortality in the NICU. Consistent with previous findings,20 there was no increase in the incidence of WLS among infants receiving palliative care. DNR orders were more frequent among infants for whom the electronic palliative care order set was activated, likely because the guidelines of our palliative care program recommend considering discussions of limitations of resuscitation in infants receiving end-of-life care. Previous studies have reported WLS to precede 42 to 95% of deaths in the NICU.3–8,10,15,21 The proportion of deaths in the NICU accompanied by withdrawal or withholding of support appears to have Journal of Perinatology (2015), 218 – 222
Impact of a neonatal palliative care program N Younge et al
220 Table 2.
Infant characteristics
Birth weight (g), median (range) Weight at death (g), median (range) GA (weeks), median (range) Male, n (%) Inborn, n (%)
Era 1 pre-PC program N = 82
Era 2 post-PC program N = 68
P-value
Era 2 Activated PC orders N = 21
Era 2 No PC orders N = 47
P-value
928 (390, 3960) 1435 (429, 6020) 28 (22, 40) 47 (59) 48 (59)
1195 (330, 3331) 1498 (466, 4870) 29 (23, 40) 40 (60) 43 (63)
0.96 0.61 0.63 40.99 0.62
1530 (400, 3331) 1670 (516, 4130) 34 (23, 39) 14 (67) 13 (62)
1100 (330, 3095) 1345 (466, 4870) 28 (23, 40) 26 (56) 30 (64%)
0.17 0.24 0.09 0.31 40.99
19 20 3 0
(45) (48) (7) (0)
0.15
Ethnicity, n (%) Black, not Hispanic White, not Hispanic Hispanic Other
37 34 7 3
(46) (42) (9) (3)
27 (44) 26 (43) 7 (11) 1 (2)
0.98
8 (42) 6 (32) 4 (21) 1 (5)
Apgar scores, median (range) 1 min 5 min
3 (0, 8) 7 (0, 9)
2 (0, 8) 6 (0, 9)
0.02 o0.01
4 (0, 8) 6 (0, 9)
2 (0, 8) 6 (1, 9)
0.31 0.23
45 (55) 30 (37) 6 (7) 1 (1) 13 (1, 287)
32 (47) 28 (41) 7 (10) 1 (2) 16 (1, 225)
0.81 0.55
6 (29) 13 (62) 2 (10) 0 (0) 12 (1, 137)
26 (55) 15 (32) 5 (11) 1 (2) 16 (1, 225)
0.09 0.80
Cause of death, n (%) Complications of prematurity Congenital anomalies HIE Respiratory failure Age at death (days), median (range)
Abbreviations: GA, gestational age; HIE, hypoxic–ischemic encephalopathy; PC, palliative care.
Table 3.
End-of-life care for infants before and after implementation of Neonatal Palliative Care Program
WLS incidence, n (%) DNR incidence, n (%) At least one family meeting, n (%) Number of end-of-life family meetings, median (range) Time between first end-of-life meeting and WLS (days), median (range) Time to death after WLS (h), median (range) Patients receiving opioids last 4 days, n (%) Morphine dose last 24 h (μg/kg), median (range) Benzodiazepines used last 4 days, n (%) Neuromuscular blocker use last 24 h, n (%)
Era 1 pre-PC program N = 82
Era 2 post-PC program N = 68
P-value
Era 2 Activated PC orders N = 21
Era 2 no PC orders N = 47
P-value
60 (73) 38 (46) 72 (88) 1 (0, 5) 0 (0, 23)
42 (63) 36 (53) 61 (91) 2 (0, 5) 1 (0, 29)
0.17 0.42 0.36 0.13 0.47
13 (65) 18 (85) 20 (95) 2 (1,4) 1 (0, 28)
29 (62) 18 (38) 41 (87) 1 (0,5) 0 (0, 29)
0.80 o 0.01 0.17 o 0.01 0.03
1 (0, 62) 72 (88) 480 (0, 2879) 21 (26) 14 (17)
1 (0, 41) 55 (81) 179 (0, 9492) 29 (43) 12 (18)
0.17 0.17 0.17 0.03 0.95
7 (0, 41) 19 (90) 302 (0, 2987) 13 (62) 1 (5)
1 (0, 20) 36 (77) 92 (0, 9492) 16 (34) 11 (23)
0.02 0.32 0.15 0.03 0.06
Abbreviations: DNR, do-not-resuscitate; PC, palliative care; WLS, withdrawal of life support.
increased over time.7–10,21 The most commonly cited reason for WLS is imminent or inevitable death, followed by concerns for future quality of life or suffering.4,5,22 End-of-life care requires complex decision making by parents and providers, and the importance of providing palliative care in this setting has been increasingly recognized.1,12 A policy statement on the care of high-risk newborn infants from the American Academy of Pediatrics emphasizes the importance of continuing comfort care measures even when treatment goals no longer involve intensive care interventions.23 However, reports indicate that the use of palliative care services in the NICU is uncommon. Single-center studies report palliative care interventions in only 13 to 24% of infant deaths.15,16,20 The reasons for this limitation are complex and multifactorial,24,25 but the results of our study provide assurance that offering palliative care services does not lead to an increased incidence of withdrawal of life support or death. Journal of Perinatology (2015), 218 – 222
One limitation of our study was exclusion of infants who died within the first 24 h of life. When we designed the study, we were concerned that there would be inadequate time to establish palliative care and utilize the palliative care order set in infants who died shortly after birth. However, since implementation of our program, we have seen that the use the palliative care protocol in infants with an antenatal diagnosis of a major congenital anomaly has been very helpful in early initiation of palliative care. Our study also evaluated the effect of a Neonatal Palliative Care Program on use of sedatives and analgesics at the end-of-life. Previous studies have reported sedative or analgesic use in 66 to 87% of infants at the time of or subsequent to WLS.8,26,27 In our study, the majority of infants received opioids before death in both Eras, at 88% and 81%, respectively. We noted an increase in the use of benzodiazepines in the post-implementation Era (26% © 2015 Nature America, Inc.
Impact of a neonatal palliative care program N Younge et al
vs 43%, P = 0.03). These results are consistent with a recent study from Janvier et al.27 reporting the use of comfort medications in the last 48 h of life in two US NICUs (Chicago and Milwaukee) with 66% and 72% of infants receiving opioids, and 52% and 53% receiving benzodiazepines, respectively. Our increase in the use of benzodiazepines was likely related to the medication guidelines recommended by our program, with recommendations on benzodiazepine use at the end-of-life. Although no studies have evaluated the effects of treatment with benzodiazepines in infants at the end-of-life, they are frequently used in adult palliative care to treat agitation and anxiety.28 Our study did not include evaluation of clinical response to comfort medications at the endof-life. NMB should not be given at the time of WLS as they are not comfort medications, they interfere with respiratory function and they could mask signs of pain and distress. In our study, a minority of infants received NMB in the last 24 h of life in both Eras (17% vs 18%, P = 0.95). However, only 3 of the 150 patients were on NMB at the time of withdrawal (2 patients from Era 1 and 1 patient from Era 2). Our palliative care program recommends discontinuing NMB for a period of hours before WLS to allow the paralytic effect to subside. However, in some cases the patient’s condition deteriorates rapidly with multisystemic compromise, making it impossible to completely wean the patient off NMB before WLS. Consistent with clinical practice in other NICUs in the United States, in no patients were NMB initiated at the time of withdrawal.27 Family meetings to discuss end-of-life care occurred more frequently in the subgroup of infants with activated palliative care order sets, likely reflecting improved communication among the health care team and family preceding the infant’s death. The fact that the time between the first end-of-life discussions and actual withdrawal was longer in the patients with activated order sets may also indicate that end-of-life discussions were initiated earlier in the course of disease, providing further evidence of improved communication. Although it is not clear why time to death after WLS was longer for the infants with activated order sets, we suspect that this trend was attributable to the higher incidence of congenital anomalies in this group of infants (Table 2). We hypothesize that WLS more commonly occurred owing to concern for future quality of life in the infants with congenital anomalies as opposed to severe illness or imminent death. Therefore, these infants may have been physiologically more stable at the time of WLS resulting in longer time to death. Alternatively, it is possible that the use of palliative care orders, including early use and appropriate dosing of sedatives and analgesics, prolonged survival in these infants. Previous studies in infants and adults have shown that use of these medications after WLS does not hasten time to death and, in some cases, has been associated with increased survival time.27,29 CONCLUSION Despite the frequency of death in the NICU, providing optimal care for infants at the end-of-life presents many challenges and complexities. Little is known about the impact of Neonatal Palliative Care Programs in the neonatal intensive care setting. Our study shows that establishment of a Neonatal Palliative Care Program did not impact the incidence of WLS or overall mortality, but did have an effect on the way medications were used at the end-of-life. We also found evidence of earlier communication with families and increased frequency of end-of-life family meetings. Although further studies are needed to assess the effect of Neonatal Palliative Care Programs on other metrics such as provider and parent satisfaction, our findings suggest that end-oflife care for newborn infants can be improved by adopting a teamdriven, patient- and family-centered, consistent approach. © 2015 Nature America, Inc.
221
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Kimberley A Fisher, PhD, and Carolyn Jones, MSN, NNP for their expert technical contributions (financial support provided by the Jean and George W Brumley, Jr Neonatal Perinatal Research Institute). Dr Smith receives salary support for research from the NIH and the US Department of Health and Human Services (government contract HHSN267200700051C, HHSN275201000003I, and UL1TR001117). Dr. Cotten received funding from NICHD: 5U10HD040492-10, from NHLBI: 1R01HL10570201A1 and from NIDDK: NIH-NIDDK203-2345. Dr. Brandon received funding from NIH-NINRR01NR010548.
REFERENCES 1 Field MJ, Behrman RE (eds). When Children Die: Improving Palliative and End-of-Life Care for Children and Their Families. National Academies Press: Washington, DC, USA, 2003. 2 Fontana MS, Farrell C, Gauvin F, Lacroix J, Janvier A. Modes of death in pediatrics: differences in the ethical approach in neonatal and pediatric patients. J Pediatr 2013; 162(6): 1107–1111. 3 Roy R, Aladangady N, Costeloe K, Larcher V. Decision making and modes of death in a tertiary neonatal unit. Arch Dis Child Fetal Neonatal Ed 2004; 89(6): F527–F530. 4 Verhagen AA, Dorscheidt JH, Engels B, Hubben JH, Sauer PJ. End-of-life decisions in Dutch neonatal intensive care units. Arch Pediatr Adolesc Med 2009; 163(10): 895–901. 5 Wall SN, Partridge JC. Death in the intensive care nursery: physician practice of withdrawing and withholding life support. Pediatrics 1997; 99(1):64–70. 6 Cook LA, Watchko JF. Decision making for the critically ill neonate near the end of life. J Perinatol 1996; 16(2 Pt 1): 133–136. 7 Barton L, Hodgman JE. The contribution of withholding or withdrawing care to newborn mortality. Pediatrics 2005; 116(6): 1487–1491. 8 Wilkinson DJ, Fitzsimons JJ, Dargaville PA, Campbell NT, Loughnan PM, McDougall PN et al. Death in the neonatal intensive care unit: changing patterns of end of life care over two decades. Arch Dis Child Fetal Neonatal Ed 2006; 91(4): F268–F271. 9 Hagen CM, Hansen TW. Deaths in a neonatal intensive care unit: a 10-year perspective. Pediatr Crit Care Med 2004; 5(5): 463–468. 10 Weiner J, Sharma J, Lantos J, Kilbride H. How infants die in the neonatal intensive care unit: trends from 1999 through 2008. Arch Pediatr Adolesc Med 2011; 165(7): 630–634. 11 Section on Hospice and Palliative Medicine and Committee on Hospital Care. Pediatric palliative care and hospice care commitments, guidelines, and recommendations. Pediatrics 2013; 132(5): 966–972. 12 American Academy of Pediatrics. Committee on Bioethics and Committee on Hospital Care Palliative care for children. Pediatrics 2000; 106(2 Pt 1): 351–357. 13 Feudtner C, Womer J, Augustin R, Remke S, Wolfe J, Friebert S et al. Pediatric palliative care programs in children's hospitals: a cross-sectional national survey. Pediatrics 2013; 132(6): 1063–1070. 14 Gans D, Kominski GF, Roby DH, Diamant AL, Chen X, Lin W et al. Better outcomes, lower costs: palliative care program reduces stress, costs of care for children with life-threatening conditions. Policy Brief UCLA Cent Health Policy Res 2012; (PB2012-3): 1–8. 15 Brandon D, Docherty SL, Thorpe J. Infant and child deaths in acute care settings: implications for palliative care. J Palliat Med 2007; 10(4): 910–918. 16 Leuthner SR, Pierucci R. Experience with neonatal palliative care consultation at the Medical College of Wisconsin-Children's Hospital of Wisconsin. J Palliat Med 2001; 4(1): 39–47. 17 Catlin A, Carter B. Creation of a neonatal end-of-life palliative care protocol. J Perinatol 2002; 22(3): 184–195. 18 Carter BS, Bhatia J. Comfort/palliative care guidelines for neonatal practice: development and implementation in an academic medical center. J Perinatol 2001; 21(5): 279–283. 19 Gale G, Brooks A. Implementing a palliative care program in a newborn intensive care unit. Adv Neonatal Care 2006; 6(1): 37–53. 20 Pierucci RL, Kirby RS, Leuthner SR. End-of-life care for neonates and infants: the experience and effects of a palliative care consultation service. Pediatrics 2001; 108(3): 653–660. 21 Singh J, Lantos J, Meadow W. End-of-life after birth: death and dying in a neonatal intensive care unit. Pediatrics 2004; 114(6): 1620–1626. 22 Hentschel R, Lindner K, Krueger M, Reiter-Theil S. Restriction of ongoing intensive care in neonates: a prospective study. Pediatrics 2006; 118(2): 563–569. 23 Bell EF. Noninitiation or withdrawal of intensive care for high-risk newborns. Pediatrics 2007; 119(2): 401–403.
Journal of Perinatology (2015), 218 – 222
Impact of a neonatal palliative care program N Younge et al
222 24 Davies B, Sehring SA, Partridge JC, Cooper BA, Hughes A, Philp JC et al. Barriers to palliative care for children: perceptions of pediatric health care providers. Pediatrics 2008; 121(2): 282–288. 25 Kain V, Gardner G, Yates P. Neonatal palliative care attitude scale: development of an instrument to measure the barriers to and facilitators of palliative care in neonatal nursing. Pediatrics 2009; 123(2): e207–e213. 26 Partridge JC, Wall SN. Analgesia for dying infants whose life support is withdrawn or withheld. Pediatrics 1997; 99(1): 76–79.
Journal of Perinatology (2015), 218 – 222
27 Janvier A, Meadow W, Leuthner SR, Andrews B, Lagatta J, Bos A et al. Whom are we comforting? An analysis of comfort medications delivered to dying neonates. J Pediatr 2011; 159(2): 206–210. 28 Henderson M, MacGregor E, Sykes N, Hotopf M. The use of benzodiazepines in palliative care. Palliat Med 2006; 20(4): 407–412. 29 Chan JD, Treece PD, Engelberg RA, Crowley L, Rubenfeld GD, Steinberg KP et al. Narcotic and benzodiazepine use after withdrawal of life support: association with time to death? Chest 2004; 126(1): 286–293.
© 2015 Nature America, Inc.
