Purpose: To characterize end-of-life decisions for surgical neonates and compare similarities and differences in practice between pediatric surgeons in The Netherlands and the United States.
Methods: The authors evaluated the deaths of all neonates admitted to the surgical intensive care unit (SICU) of two major children's hospitals: Sophia Children's Hospital (SCH) in The Netherlands and Columbus Children's Hospital (CCH) in the United States. Between January 1990 and July 1993, neonatal SICU admissions totaled 362 (SCH) and 125 (CCH). Neonates who died were classified as follows: group 1 = poor prognosis, expected death, and group 2 = good prognosis, expected survival.
Results: The mortality rates were comparable for each SICU: 12% (SCH) and 14% (CCH). The average survival period was shorter in group 1 (1.5 days) than in group 2 (26.5 days). Criteria for assignment to group 1 differed, with "expected poor quality of life" used at SCH, and "futility" at CCH. Criteria for group 2 were similar and included significant postoperative complications. Although the percentages with a do-not-resuscitate (DNR) status were comparable (SCH, 51%; CCH, 55%), the application of the DNR order differed in each SICU. The majority of neonates at SCH had either withholding or withdrawal of life support, whereas no further escalation in treatment was offered for infants with a DNR order at CCH. The average survival period after the DNR order was 4 days at SCH and 7 days at CCH.
Conclusion: DNR orders were used for more than half the surgical neonates with critical illness. Criteria for DNR status and implementation of patient care after the DNR order differed between the SICUs.