Why Do Neonatologists in Scandinavian Countries and the Netherlands Make Life-and-death Decisions So Different?

The Swedish 2016 guidelines clearly represent an interventionist approach.^9,10 These guidelines recommend an antenatal transfer to a level-3 NICU at 22 + 0/7 weeks’ gestation. Antenatal steroids are considered at 22 + 0/7 weeks’ gestation and are recommended at 23 + 0/7 weeks’ gestation. Neonatologists are present at deliveries at 22 + 0/7 weeks’ gestation, and resuscitation is considered at 22 + 0/7 weeks’ gestation and recommended for all infants at 23 + 0/7 weeks’ gestation. Guidelines, reports, and consensus meetings from Denmark, Norway, and other Scandinavian countries also recommend active treatment from 23 + 0/7 weeks’ gestation onward, although slight differences between centers and countries are noted.^11,–13 

In contrast, the Dutch 2005 guideline for the management of preterm deliveries was clearly more conservative in recommending against the resuscitation of newborns <25 weeks’ gestation.¹⁴ On the basis of 2 national outcome studies, the ethical justification of this restricted approach was that “the amount of ‘good’ (less than 10% intact survival) does not justify the amount of ‘harm’ done (10–20% disabilities, 70–80% mortality after a period of intensive treatment).”¹⁴ The authors argued that at 25 weeks’ gestation, the balance seemed more favorable, with 20% to 30% intact survival, 12% to 14% disabilities, and 52% to 56% mortality. In 2010, the revised guideline was published.¹⁵ It recommended resuscitation for all newborns who were born at ≥24 weeks’ gestation. The new cutoff point was justified by the absence of differences in outcomes between resuscitation at 24 weeks’ gestation or 25 weeks’ gestation in large international studies. Recently, the 2-year follow-up results of infants who were born at 24 weeks’ gestation in the first year after the implementation of the new guideline were published.¹⁶ The results revealed that of all infants who were born at 24 weeks’ gestation and survived to 2 years of age (46%), more than half had developed without disabilities. Of all children who were born at 24 weeks’ gestation, 25% survived to 2 years of age without disabilities. These results were interpreted by the authors as roughly comparable with outcomes of 24 weeks’ gestation in other international units. The comments in the public media and among health care providers were mixed: Some argued that the mortality and the morbidity were such that additional (downward) adjustments to lower the limit for active treatment could not be justified.¹⁷ Others argued that the results were in fact a self-fulfilling prophecy and suggested to stop using the label of gestational age to make life and death decisions.¹⁸ 

The authors of 2 recently published international vignette studies confirmed the conservative and restricted approach in regard to providing care of extremely premature infants in the Netherlands today.^19,20 In the first study, Wilkinson et al²⁰ compared resuscitation thresholds for extremely premature infants between neonatologists in the United Kingdom, Sweden, and the Netherlands. They reported that neonatologists from the Netherlands had the least proactive approach to the resuscitation of these infants.²⁰ Interestingly, they also showed that despite the different approaches to decisions in the 3 countries, there was relatively little difference between countries in neonatologists’ prognostic thresholds for resuscitation. Laventhal et al¹⁹ also reported that neonatologists from the Netherlands were least likely to resuscitate a premature infant of 24 weeks’ gestation in comparison with colleagues from Norway and many other non-Scandinavian countries. These results reveal a striking similarity with those from the European Concerted Action on Information to Parents and Ethical Decision-Making in Neonatal Intensive Care (EURONIC) study, in which researchers explored decision-making in a sample of NICUs in 11 European countries, including Sweden. When confronted with the birth of an extremely premature infant (24 weeks’ gestation, 560 g, and Apgar score of 1), most physicians in every country but the Netherlands would resuscitate and start intensive care.²¹ The data for this study were collected roughly 20 years ago. These findings and the time line suggest a consistent conservative approach.

Unfortunately, not much is published about how and when decisions to withhold or withdraw life-sustaining treatments (LSTs) in the Scandinavian NICUs are made. Keeping track of all the care and decisions that precede death and dying in the NICU is essential to understand how the population of infants being observed came into existence and what we can validly infer from comparisons of outcomes. Authors of reports from Sweden have described significant variations in mortality rates between Swedish health care regions, particularly between the most immature infants of 22 to 24 weeks’ gestation. According to the authors, this could be partly based on differences in approach to end-of-life decision-making.²² Additional analysis of that approach would be welcomed and helpful to promote a true comparison of outcomes. In several publications from Norway, researchers have mentioned their physician’s protreatment attitude and high acceptance rate for withdrawal of treatment in the NICU,^23,24 but as far as I know, details about reasons for withdrawal have not yet been published.

In contrast, the authors of several studies from the Netherlands have reported high rates of NICU deaths by withholding or withdrawing LSTs.^3,25,–27 Most recent studies from industrialized countries reveal that ∼35% to 80% of deaths in the NICU are by withholding or withdrawing LSTs.^3,28,–31 Rates in the Netherlands vary between 93% and 98%. A relatively large proportion of these decisions (∼50%) were based on quality-of-life arguments and concern for newborns who were physiologically stable. Guidelines and reports were developed to guide physicians in making these quality-of-life decisions.³² The argument that was most frequently used to withdraw artificial ventilation in the NICU was the fear of future suffering.²⁷ Dutch parents were always involved in the decision-making process, and conflicts between physicians and parents about the decision to withhold or withdraw treatment were relatively rare.³³ 

Reports about the practice of hastening the death of newborns who are severely ill in the Scandinavian delivery rooms or in NICUs are lacking as far as I could determine. However, in the Netherlands, this practice is well described and accepted in the context of end-of-life care.^25,34 For the Dutch, hastening of death entails life-shortening as a side-effect of medication use to treat pain and symptoms at the end of life. This is accepted only in situations in which palliative medication use is proportionate.³⁵ In addition, hastening of death entails the more controversial practice of neonatal euthanasia. This practice is legal in the Netherlands, where it is defined as the deliberate ending of life in a newborn with hopeless and unbearable suffering.³⁵ Important principles and values that serve as the moral justification of neonatal euthanasia are beneficence, parental determination, responsibility, and compassion or care.^35,36 An important step toward legalization was the publication and nationwide adoption of the Groningen Protocol (GP) in 2005.³⁷ One of the main goals in the GP was to make the existing practice of neonatal euthanasia transparent and subject to public review. In the protocol, the criteria that need to be fulfilled to make physicians’ deliberate life ending in newborns careful and legal is described. Examples of these criteria are the following: the required presence of hopeless and unbearable suffering, certainty about the diagnosis and prognosis, and full consent of both parents. After the adoption of the GP, euthanasia decreased from 15 cases to 2 cases over 5 years. Among health care providers and policy makers, the question on why this happened arose. The answer came from national surveys that revealed that the political decision to include an antenatal screening policy as a part of antenatal care in 2007 (for free) had resulted in increased terminations of pregnancies and fewer euthanasias.³⁸ Surprisingly, the effects of the GP were different from those that were predicted by either supporters or critics. Nevertheless, the number of euthanasia cases has continued to drop to 4 cases in 12 years.

The examination of decision-making at the beginning and at the end of life for premature newborns in the Scandinavian countries and in the Netherlands raises many questions. Why do neonatologists in these 2 neighboring regions that are presumably on an equal footing with respect to resources make life and death decisions so differently? Why are the Dutch so liberal in life-ending matters and conservative when an infant is born at the limits of viability? These questions are difficult to answer. However, maybe the ethical context in which the decision-making takes place can help us to understand the differences a little bit better. Among the central values of end-of-life practice in the Dutch NICUs is the recognition of quality of life and quality of dying as leading goals of care. Authors of reports and policy statements published in the last decades have all confirmed the basic belief among the Dutch that suffering of newborns and children should be avoided and/or alleviated at all times, even if the only way to achieve that would be by hastening death. The Dutch tend to trust their professionals, and entrusting this sensitive task to doctors has not seemed dangerous to them. Polls and surveys among the Dutch public and the medical profession in the last 50 years have consistently revealed support for euthanasia to end suffering. This, together with a history of tolerance and a culture of making taboo subjects discussable and preferring transparency over secrecy, may have created the cultural soil in which examining the hastening of death and regulating neonatal euthanasia could take place along the same lines as in the adult world.³⁹ In addition, the (Calvinistic) duty to be accountable is another central value deeply rooted in Dutch society that is relevant to neonatology. In reports by the pediatric profession published between 1980 and 2013, all recommend that when life support is initiated, continuation is conditional on subsequent developments in diagnosis and prognosis.³⁵ If it becomes clear that the outcome of NICU treatments might involve a damaged quality of life (suffering), not only withdrawal but also continuation of LST demands justification. We were taught that neonatologists carry the full responsibility for the suffering of patients and families if the outcome of NICU treatment is poor. And so, this is what many of us teach our residents today. The strong focus on responsibility might explain, at least partly, why Dutch neonatologists have remained rather conservative in their approach in regard to the active treatment of infants who are at the limits of viability.

GP

Groningen Protocol

LST

life-sustaining treatment