Ethics of resuscitation at different stages of life: a survey of perinatal physicians.

Abstract

We surveyed US neonatologists and high-risk obstetricians about preferences for resuscitation in ethically difficult situations to determine whether (1) their responses adhered to traditional ethical principles of best interests and patient autonomy and (2) physician specialty seemed to influence the response. In an electronic survey, we presented 8 vignettes with varying prognoses for survival and long-term outcome. Respondents were provided outcome data for mortality and morbidity in each vignette. We asked whether resuscitation was in the patient's best interest and whether the physician would accede to requests for nonresuscitation. We analyzed surveys for 587 neonatologists and 108 high-risk obstetricians (15% overall response rate, 75% of web site visitors). There were no statistically significant differences in responses between the 2 physician subspecialty groups. As expected, in most cases there were inverse relationships between valuation of best interest and deferred resuscitation at the family's request. For example, for the oldest patient (an 80-year-old), 9.9% found resuscitation to be in the patient's best interest and 94.3% would allow nonresuscitation; for a 2-month-old, 93.9% found resuscitation to be in the patient's best interest and 24.5% would allow nonresuscitation. However, this pattern was not observed in the 2 newborn cases, in which resuscitation and nonresuscitation were both acceptable. In the triage scenario, the 7-year-old with cerebral palsy and acute trauma was consistently resuscitated first despite others having equivalent or better short- and long-term prognoses. On the basis of our results, physicians' decisions to resuscitate seem to be context-specific, rather than based on prognosis or consistent application of best-interest or autonomy principles. Despite their different professional perspectives, neonatologists and high-risk obstetricians seemed to converge on these judgments.