OBJECTIVE: To determine whether parental characteristics affect estimates of best interests and intervention decisions for preterm infants.
DESIGN AND METHODS: The study consisted of an anonymous questionnaire given to nurses, physicians, and students. The study included scenarios of 3 sets of parents, including a 16-year-old teenager, a couple who were lawyers, and a couple with a history of in vitro fertilization, about to deliver at 22 weeks, 24 weeks, or 27 weeks. Respondents were asked whether active intervention is in the infant's best interests and whether they would comply with family decisions.
RESULTS: A total of 1105 questionnaires were sent out, with 829 respondents in Canada and the United States. At 22 weeks' gestation, 21% of the respondents thought that resuscitation was in the infant's best interest; among respondents who did not agree, 59% would intervene if the parents wished. At 27 weeks' gestation, 95% of respondents thought that resuscitation was in the infant's best interest, yet 34% would accept comfort care. Estimates of best interest, and willingness to comply, varied significantly by parental characteristics. At 22 weeks' gestation, 17% of respondents believed that resuscitation was in the best interest of the teenaged mother's infant compared with 26% of respondents who believed that resuscitation was in the best interest for the infants of the others; this difference persisted at 24 weeks. At 22 and at 24 weeks' gestation, compliance with active care despite believing that it not in the infant's best interest was significantly more frequent for the in vitro fertilization couple and the lawyers than for the teenaged mother. At 27 weeks' gestation, more than 93% of respondents complied for all parents.
CONCLUSIONS: Caregivers frequently are ready to intervene actively, or not, despite believing that it is against the infant's best interest. Willingness to do so varies according to parental characteristics.
- assisted reproductive technologies
- end-of-life decisions
- best interest
- comfort care
WHAT'S KNOWN ON THIS SUBJECT:
Although the best-interest standard is considered to be the paradigm for decision making for the extremely preterm infant, caregivers are prepared to give, or withhold, active care against their own evaluation of a baby's best interest.
WHAT THIS STUDY ADDS:
Evaluation of a baby's best interest, willingness to intervene actively or not, and compliance with parental wishes varied according to the age and socioeconomic status of the parents.
Survival rates for extremely preterm infants have dramatically improved over the last 5 decades, but these advances have led to new ethical dilemmas, many of which involve decisions regarding withholding or withdrawing life-sustaining interventions in situations where such interventions could save or prolong the life of an infant at risk for permanent and serious complications.
Professional guidelines suggest that doctors should share decision-making authority with parents for extremely preterm infants. However, interpretation of risks and benefits vary depending on country of origin, religious beliefs, cultural factors, age, and underlying values. Parents are expected to use their children's best interest as the guiding principle for decision making.1,2 However, if the health care providers believe “that the designated surrogate threatens the patient's best interests, the decision should be overridden;”1 such overriding of family decision making has been supported in the courts.1,3,4 Our previous research has demonstrated that assessments of whether resuscitation is in the best interests of a patient are not closely related to survival rates or disability5; resuscitation of newborn infants, and particularly the extremely preterm, was desired much less frequently than for older individuals, even those with much worse predicted outcomes.5 Accepting a family's refusal of resuscitation, even when physicians believed that it was not in the patient's best interest, was much more common for neonates than for older pediatric patients.5
Presumably, during the complicated neonatal decision-making process, and when caring for patients, physicians should not be biased by the parent's characteristics such as age, socioeconomic status, education, marital status, and ethnicity. However, demographic factors do influence the care given to women in preterm labor.6 We wondered if parental demographics (age, profession, fertility history, education, and marital status) would influence the assessments of best interests and desire for intervention for extremely preterm infants. To determine whether the proportion of respondents who believed it to be in the best interest of an extremely premature infant to receive active intervention changes according to the age and background of the parents.
DESIGN AND METHODS
Between May 2008 and March 2009, we administered an anonymous questionnaire to health care providers involved in the care of the preterm infants. Nurses, residents, and attending physicians at McGill University (Montreal, Canada), Thomas Jefferson University Hospital (Philadelphia, Pennsylvania), Christiana Care Health System (Newark, Delaware), and the Alfred I DuPont Hospital for Children (Wilmington, Delaware) participated in the study as well as medical students at Thomas Jefferson and McGill Universities. Questionnaires were administered during a group activity, such as a staff meeting or teaching session, or on an individual basis. The first question was a request for consent; if consent was refused, the blank questionnaire was collected along with the other questionnaires. The final page of the questionnaire consisted of demographic information (ie, occupation, gender, years of experience, religion, and whether they had children).
Each questionnaire had 9 possible scenarios: 3 different scenarios of parents about to have an extremely preterm infant at 3 possible gestational ages (22, 24, and 27 weeks).
Adolescent mother (teenager): A 16-year-old mother with an unplanned pregnancy and a history of 2 previous abortions. The identity of the infant's father is unknown.
Lawyers: First infant of a trial-lawyer couple, both aged 30 years, with no fertility problems.
Infertile couple (in vitro fertilization [IVF]): First infant of a couple, aged 40 years, with a fifth IVF attempt paid for by remortgaging their home. The mother works in a day care for disabled children.
Outcome statistics were explicitly described for each gestational age (see Table 1)
For each of the 9 combinations, the following questions were asked: (1) Do you think it is in the infant's best interest to be resuscitated? (2) Would you comply with the parent's decision for resuscitation? (3) Would you obtain a court order if you disagreed with the parent's decision to resuscitate? (4) Would you comply with the parent's decision for comfort care? (5) Would you obtain a court order if you disagreed with the parent's decision for comfort care? Answers were given using a modified Likert scale (always, generally, exceptionally, or never).
Statistical analysis was performed using a 2-tailed χ2 test with Yates correction. Because of the large number of participants, small differences in responses would likely be significant at a level of P < .05; therefore, we considered a P value less than 0.02 as significant. In addition, multivariate analysis of the respondents' demographic information was performed for each question, which demonstrated statistical significance. This study was approved by the institutional review boards at each institution.
A total of 1105 individuals were targeted, with 820 respondents and an overall response rate of 74%. In Canada, there were 289 of 385 responders (75%), with 129 students, 48 pediatric and obstetric residents, 14 perinatal attending physicians, 51 NICU nurses, and 47 obstetric nurses. In the United States, there were 531 of 720 responders (74%), with 216 students, 26 pediatric residents, 15 perinatal physicians (fellows and attendings), 126 NICU nurses, 88 obstetric nurses, 35 PICU nurses, and 25 surgical neonatal nurses. Response rates varied among groups (from 59% to 99%). US students had the highest response rate, at 99%. Students were younger and had fewer children than residents, who were younger and had fewer children than attending physicians and nurses (P < .05). A total of 59% of residents were seniors. A total of 64% of nurses had more than 5 years of practice. A total of 62% of students, 76% of residents, and 97% of nurses were female.
The largest proportion (95%) of respondents stated that it was always or generally in the best interest of the 27 weeks' gestation infant to be resuscitated, followed by the 24 weeks' gestation infant (57%), and the 22 weeks' gestation infant (21%). However, responses among the 3 parent scenarios differed. At 22 and at 24 weeks' gestation, the lowest gestational ages, a smaller proportion felt that it was in the best interest of the child of the teenaged mother compared with the older parents (Fig 1). At 27 weeks' gestation, over 90% of respondents believed that it was in the best interest of the infant to be resuscitated; this did not differ between parent scenarios.
A higher proportion of students (in both Canada and the United States) than nurses and physicians thought active care to be in the infant's best interests at 22 and 24 weeks' gestation. A total of 32% of students answered that resuscitation was in the best interest of a 22 weeks' gestation infant, compared with 4% of nurses and physicians (P < .001). A total of 68% of students thought that resuscitation was in the best interest of a 24 weeks' gestation infant, compared with 57% of physicians and 52% of nurses (P < .001 students compared with other groups; P < .01 physicians compared with nurses).
We analyzed the data to determine the proportion of respondents for each scenario who were prepared to comply with the parents even when this implied that they would act against their own determination of whether active care was in the infant's best interests. The results are shown in Table 2. At 22 weeks' gestation, respondents who believed that active care was in the best interest of the infant were much more likely to comply with parental wishes (>82% for all 3 scenarios) than were the respondents in the contrary situation (<65% for all 3 scenarios). At each gestational age, a smaller proportion of respondents were prepared to intervene actively against their estimate of best interest for the teenaged mother than for the IVF couple or the lawyers. At 24 weeks' gestation, but not at the other 2 ages, active intervention against a determination of best interest was accepted more frequently for the IVF couple than for the lawyers (P < .01). Students were more likely to comply at 22 and 24 weeks' gestation, compared with other groups (P < .01). Nurses and students were more likely to comply at 22 weeks' gestation than physicians (P < .01). Overall, Canadian respondents were less likely to comply at 22 weeks' gestation (53%) than Americans (69%; P < .01).
At 22 weeks' gestation, respondents would seek a court order to enforce comfort care for the infant of the teenaged mother more often than the lawyers or IVF couple (34%, 23%, and 18%, respectively; the teenaged mother compared with lawyer or IVF couple: P < .001; IVF couple compared with the lawyers: P < .02). Canadian respondents were more likely to seek a court order to impose comfort care for the infant of the teenaged mother than Americans (50% vs 27%; P < .01). At 24 weeks' gestation, the differences were less marked but remained significant; a court order was more often sought against the teenaged mother than the lawyers or IVF couple. (17%, 13%, and 12%, respectively; P < .02) There were no differences between the parent scenarios at 27 weeks' gestation. At 24 and 27 weeks' gestation, none of the surveyed neonatologists would seek a court order to withhold intervention.
The proportion of those accepting comfort care despite their answer that active care was in the infant's best interest differed significantly by group; students were much less likely to comply with comfort care (69%) compared with nurses (91%) and physicians (85%) at 22 weeks' gestation (P < .01). At 24 weeks' gestation, this also was significant (students 52%, nurses 73%, and physicians 69%).
Seventeen to 22% of respondents would seek a court order for intervention at 22 and 24 weeks' gestation, with no significant differences among scenarios. Differences between the parents were only apparent at 27 weeks' gestation. Respondents were most likely to seek a court order for the lawyers compared with the teenaged mother (59% and 49%, respectively; P < .001) and for the lawyers versus the IVF couple (59% and 52%, respectively; P < .01). At 27 weeks' gestation, 71% of physicians would seek a court order to intervene against parental wishes, compared with 45% of nurses and 50% of students (P < .01). No neonatologists would seek a court order at 22 weeks' gestation, 15% would seek a court order at 24 weeks' gestation, and 82% would seek a court order at 27 weeks' gestation.
During informed consent at very low gestational ages, physicians are expected to inform parents who can then participate in the decision-making process. It generally is accepted that treatments considered to be in a patients' best interest should be pursued.1,–,4 In this study, respondents frequently were willing to act in ways that were not in accordance with their own determination of the infant's best interest. Although a minority (21%) thought resuscitation was in the best interest of a 22 weeks' gestation infant, 62% would resuscitate if parents wished. In contrast, a majority (57%) thought resuscitation was in the best interest of the 24 weeks' gestation preterm infant, yet 70% would accept comfort care at parental demand. These abrogations of the best-interest principle were most striking at 27 weeks' gestation, where 95% thought that resuscitation was in the best interest of the infant, yet 34% would give comfort care and accept the death of the infant at parental request. Respondents' willingness to act was, therefore, not determined solely by their estimation of the infant's best interest. It seems that, in addition, they consider that the parents' wishes are important in determining how to proceed, but how much weight they give to those wishes varies according to family characteristics, with a young, single mother's wishes being given less weight than older married parents.
Perhaps this willingness to include more than their own evaluation of the infant's best interest in decision making is a reflection of a lack of certainty in respondents' evaluation of the best interest of the infant, which made them consider this to be a gray zone. In this gray zone, respondents might have been unsure how beneficial resuscitation might have been, and erred on the side of life until parental wishes were added into the equation. We did not ask why respondents were willing to act against their assessment of best interest, and they were only given 1 opportunity to respond regarding the infant's best interests. If we had asked that question twice, before and after a description of parents' desire, or unwillingness, to intervene, perhaps the infant's best interest would have been evaluated in 2 different ways.
Resuscitation was considered less frequently in the best interest of the infant of the adolescent mother than the older married parents. This translated into a larger proportion wishing to intervene for the infants of the older parents. At 22 and 24 weeks' gestation, respondents were more likely to seek a court order to give comfort care against the parents' wishes for the infant of the teenaged mother compared with the other parents. In contrast, at 27 weeks, when almost all the respondents thought that intervention was in the infant's best interest, they were more likely to seek a court order to intervene for the infants of the older parents. Clearly, it is not just parental desires that affect these decisions but also the characteristics of those parents.
Do the best interests of infants born between 22 and 25 weeks' gestation vary according to family characteristics? This really depends on the definition of best interests. Long-term developmental outcomes of extremely immature infants born to single, unmarried mothers who have not finished high school are worse than those who are born to married highly educated parents.7,–,9 That children of the rich and educated have better health outcomes than the poor uneducated is well known; however, many of our respondents were medical students who may well have been unaware of the specific outcome literature regarding extremely preterm infants. Even if they were aware of this literature, one could still ask whether it is unfairly discriminatory to allow socioeconomic status, age of parents, or their marital status to influence the care we give children. If we transform the social characteristics into a biological one (percentage of children with developmental impairments), it may seem acceptable. But using the knowledge that disadvantaged populations have worse health care outcomes to further discriminate against those who already are disadvantaged is morally problematic.
One also could question whether the best-interest paradigm is the most appropriate way to determine the appropriate course of action. Other ways of formulating these ethical dilemmas, such as minimizing harm, have been suggested.10,11 Perhaps the respondents considered that greater harm would be done to the infant if he or she were born into a family with lower socioeconomic status or that less harm would be done by letting such an infant die. It seems to us that these alternative formulations of the principles involved still leave the same unsettling questions about how the families' characteristics should be incorporated into the decision-making process.
In addition to socioeconomic status, other characteristics modify the outcome of a preterm infant, such as gender, place of birth, birth weight, and antenatal steroids. A 23 weeks' gestation female born after antenatal steroids to 2 university professors in a tertiary care center has better predicted outcomes (both survival and long-term disability) than a 25 weeks' gestation male born in a primary care setting without prenatal steroids to a single mother of lower socioeconomic status. Which of these factors should appropriately influence decision making? The position statements of many professional societies recommend a decision on the basis of gestational age alone,12,–,16 which we have argued is too simplistic.17 If we are to base our decisions, and the range of acceptable decisions that we will discuss with parents, on predicted outcomes, then, at most gestational ages, the likelihood of resuscitating a boy would be less than for a girl; is this acceptable? Is it acceptable to include in our calculations of risk the family's social background or financial resources, even if good scientific data demonstrate their importance in determining long-term outcomes? Is this different from including in our decision making a family's religious background or value system?
There is in, contrast, no reliable evidence that previous infertility affects long-term outcomes, but at lower gestational ages (22 and 24 weeks), respondents were more likely to accept a parental decision to resuscitate for the IVF couple than for the lawyers. Perinatologists and neonatologists previously have been shown to be more willing to intervene when a mother is older or when she conceived by IVF.18,19 These infants may be referred to as precious infants, who are less replaceable. These concepts, unfortunately, denigrate the status or the preciousness of other infants.
Students were more willing to resuscitate, and to seek legal action to resuscitate, at lower gestational ages. This may be because students are not familiar with the NICU and have not seen the long NICU stays that infants of less than 25 weeks' gestation may experience. On the other hand, medical students experience many rotations where survival similar to that demonstrated in Table 1 is considered acceptable.
Because of our good response rate, these results are likely to accurately reflect the opinions of our target population, but these cannot be directly extrapolated to other groups or cultures. Few respondents were neonatologists, which makes it hard to draw conclusions for this group. It would be interesting to explore whether the responses given on such scenarios correlate well or poorly with actual clinical practice decisions. The scenarios in the questionnaire were extremes at identical gestational ages placed side by side, a situation rare in reality. This probably made respondents realize their biases and might have led to a reduction of the difference one could have seen between the 3 parent scenarios in practice. On the other hand, respondents may truly try to limit the influence family circumstances may have on their life-or-death decisions for preterm infants. It is important that we further understand health care providers' opinions regarding intervention because they influence real intervention decisions.20,21 Our results show that family circumstances may well influence attitudes toward active intervention for a preterm newborn infant.
- Accepted January 3, 2011.
- Address correspondence to Annie Janvier, MD, PhD, Department of Neonatology, 5ème Bloc 5, Hôpital Sainte Justine, 3175 Chemin de la Côte Sainte Catherine, Montréal, H3T-1C5, Quebec, Canada. E-mail: or
FINANCIAL DISCLOSURE: The authors have indicated that they have no personal financial relationships relevant to this article to disclose.
- IVF =
- in vitro fertilization
- Beauchamp TL,
- Childress JF
- Sneiderman B,
- Irvine JC,
- Osborne PH
- 3.↵American Academy of Pediatrics Committee on Bioethics. Guidelines on forgoing life-sustaining medical treatment. Pediatrics. 1994;93(3):532–536
- 4.↵President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. Deciding to Forego Life-Sustaining Treatment. Washington, DC: US Government Printing Office; 1983
- Janvier A,
- Leblanc I,
- Barrington KJ
- Patrianakos-Hoobler AI,
- Msall ME,
- Marks JD,
- Huo D,
- Schreiber MD
- 11.↵American Academy of Pediatrics Committee on Bioethics. Religious objections to medical Care. Pediatrics. 1997;99:279–281
- 12.↵Danish Council of Ethics. Extreme prematurity, ethical aspects. Copenhagen, Denmark: Danish council of Ethics; 1995
- 13.↵Swiss Society of Neonatologists, Guidelines. Recommendation for the care of infants born at the limit of viability (GA 22–26 weeks). (accessed April 3, 2007at www.neonet.ch/assets/doc/Infants_born_at_the_limit_of_viability_-_english_final.pdf)
- Salle B,
- Sureau C
- 15.↵Nuffield council on bioethics. Critical care decisions in fetal and neonatal medicine: ethical issues, England 2006. (Accessed April 3 2007 at www.nuffieldbioethics.org/go/ourwork/prolonginglife/publication_406.html)
- 16.↵Management of the woman with threatened birth of an infant of extremely low gestational age. Canadian Medical Association Journal. 1994;151(5):547–551
- Morse SB,
- Haywood JL,
- Goldenberg RL,
- Bronstein J,
- Nelson KG,
- Carlo WA
- Copyright © 2011 by the American Academy of Pediatrics