Ethical Issues

What/Who Defines the Limits of Viability, and Must All These Infants Born at the Limits of Viability be Resuscitated?

Continued improvement in neonatal care has shifted the limits of viability to younger, more immature gestational ages. In 2014, an executive summary of a workshop conducted by the National Institutes for Child health and Human Development (NICHD), the Society for Maternal-Fetal Medicine, the American Academy of Pediatrics (AAP), and the American College of Obstetricians and Gynecologists (ACOG) that focused on periviable birth (defined as birth period from 20 0/7–25 6/7 weeks’ gestation) was published. This is a critical period of gestation during which brief pregnancy prolongation can significantly alter the potential for survival and long-term complications. This consensus group encouraged more intensive treatment at younger gestational ages than previously recommended, adding significantly to the ongoing debate about specific gestational age–based recommendations.

Currently in 2018, 23 weeks’ gestation remains the accepted lower limit of viability, although this threshold continues to be the focus of intense deliberation. There are many institutions where medical professionals are initiating resuscitation and providing neonatal intensive care for fetuses born at 22 weeks’ gestation, whereas others strictly adhere to this gestational age range for no or limited options for intervention. The gray zone time period of periviability most often quoted is between 22 to 25 weeks’ gestation; it is a gestational age spectrum during which optimal treatment options are unclear, and treatments offered vary based on provider and/or parental preference. The outer thresholds of less than 22 weeks and greater than 25 completed weeks’ gestation represent points at which treatment variation is limited, and so, too, are parental authority and provider preference. It has been generally accepted that fetuses less than 22 weeks’ gestation receive comfort care only and no intensive treatment, and that infants born after 25 weeks’ gestation generally receive intensive treatment, providing there are no anomalies or conditions inconsistent with survival. Advances in obstetric management of preterm and periviable pregnancies and significant enhancements in neonatal treatments over the past 30 years have resulted in lowering of the threshold for potential survival to 22 to 23 weeks’ gestation, and raised concerns about what is appropriate practice at the upper end of the periviability threshold at 25 weeks’ gestation.

These parameters have been established based primarily on morbidity and mortality statistics collected by various neonatal research networks throughout North America, Europe, and Australasia. Consensus guidelines have been promulgated by a variety of perinatal organizations, including the National Institutes of Health, ACOG, AAP, British Association of Perinatal Medicine, World Association of Perinatal Medicine, and the Nuffield Council on Bioethics in the United Kingdom. The NICHD neonatal research network reports survival rates of 6% at 22, 26% at 23, 55% at 24, and 72% at 25 weeks’ gestation, respectively. Morbidity rates that encompass mild, moderate, and severe neurodevelopmental impairment are reported as: 100% at 22, 92% at 23, and 91% at 24 weeks’ gestation, respectively. Survival data and neurodevelopmental outcomes reported by the Vermont Oxford Neonatal Network, a consortium of NICUs across the United States and including neonatal units worldwide, are similar. A systematic review of guidelines for the management of extremely premature deliveries published in 2015 by Guillén et al concluded that there continues to be wide variation in treatment recommendations (comfort care, parental wishes, individualized care, or active care), particularly between 23 and 24 weeks’ gestation, although there is general agreement across highly developed countries in the world and among international professional organizations that comfort care is appropriate at 22 weeks’ gestation, with active care more commonly recommended at 25 weeks of gestation. It is of interest that at 23 weeks’ gestation, these guidelines emphasize comfort care, deferring fully to parental wishes, individualized care (allows for physician assessment at the time of birth, informing the treatment options), or offer no recommendation. In contrast, at 24 weeks’ gestation, these guidelines focus recommendations on following parental wishes, individualizing care or active care; several guidelines offer no specific recommendations. AAP and ACOG guidelines are in agreement, recommending that care be individualized across the period of 22 through 25 weeks’ gestation. The American Heart Association neonatal resuscitation guidelines supports comfort care at 22 weeks, deferring to parental wishes at 23 and 24 weeks’ gestation, and providing active care at 25 weeks of gestation.

Some clinicians argue that the data on which these recommendations are based are biased. The potential for bias may be because of the reporting of older birth cohorts born during the 1980s and early 1990s. These studies offer true longitudinal information in terms of outcomes into adulthood, but are not reflective of the outcomes for the beneficiaries of advances in obstetric and neonatal practices that have occurred since 1990. In addition, some health, cognitive, and neurologic morbidities have remained relatively static, whereas others have improved. Although survival rates have increased for infants born between 22 and 25 weeks’ gestation and there have been some shifts in the severity of disability experienced by these extremely preterm survivors to more moderate and mild neurodevelopmental impairment, on the whole, overall impairment rates have remained stable since the mid-1990s, when use of antenatal corticosteroid therapy and postnatal exogenous surfactant administration became standard practice. Reduction in severe morbidity has been greatest among infants born at or greater than 25 weeks’ gestation. Other clinicians make the case that if one creates an arbitrary cutoff line at 23 weeks’ gestation for active care and doesn’t attempt to resuscitate fetuses at 22 weeks’ gestation, then a self-fulfilling prophecy is sustained, and there will be no further improvements in survival or developmental outcomes. There are data to support this. Data from Japan where active care is provided to fetuses at 22 weeks’ gestation demonstrate lower mortality rates of 64% before hospital discharge for this group of infants. An important caveat is that many of the infants in the Japanese studies were lost to follow-up, so better outcome data is needed to more accurately assess the impact of active intervention at this gestation. In conjunction with increased survival, the study investigators reported an increase in the number of infants with cerebral palsy, indicating that improved survival does not necessarily equate to better outcomes.

Tyson and colleagues reported that gestational age alone does not independently predict outcomes for these extremely immature fetuses; other characteristics such as weight, sex, administration of antenatal steroids, and whether the infant is a multiple or a singleton are also important predictors and should be considered when counseling a family. Although we have outcome data in terms of numbers of children who have died or who experience a particular outcome, there are significant variations in outcomes influenced by myriad sociodemographic and management practices that confound our ability effectively predict an individual child’s expected outcome. This makes antenatal counseling and discussions about treatment decisions for the periviable fetus very challenging. A 2015 report in the New England Journal of Medicine shows significant between-hospital variation in the treatment of neonates at the edge of viability. Practice variations and differences in available resources likely account for some of the discrepancies in outcomes, adding further complexity to antenatal counseling. One example cited a range of active treatments of infants born at 22 weeks’ gestation from 7.7% to 100%, depending on the birth hospital. There is much less interhospital variation in active treatment of infants born at or greater than 25 weeks’ gestation, but the authors still report differences in outcomes, so these outcome differences are not solely attributable to differences in treatment decisions between hospitals.

A framework centered on probability-based thresholds has been proposed by Tyson and colleagues as a strategy to help families and clinicians approach decision making when outcomes are variable or unclear. The model proposes that when the probability of survival without profound impairment is 10% or less and the probability of death or profound impairment (developmental score <50 or physical impairments that limit voluntary control of movement and involve all areas of motor function) is 90% or greater, intensive care should not be offered; instead, we should support families through the process of comfort care for their infant. For fetuses for whom the expected outcome would be a 10% to 20% chance of survival without profound impairment and an 80% to 90% chance of death or profound impairment, comfort care is an acceptable choice, although parental wishes for treatment can be honored with an understanding that treatment should be considered experimental. As the probability of survival without profound impairment increases and the odds of the combined outcome of death or survival with profound disability decreases, intensive care becomes optional or automatic (mandatory). The continuum for which intensive care would be optional encompasses the probability of survival without profound impairment ranging from 20% to 65%, with a concurrent probability of death or profound impairment of 35% to 80%. Fetuses for whom intensive care would be deemed mandatory are those whose unimpaired survival ranges from 65% to 100%, and risk of death or profound impairment is less than 35%. Using this framework provides clinicians and families with a means for incorporating changing outcomes associated with advances in treatment.

The fetus in case 1 is estimated to be 22 4/7 weeks’ gestation, 500 g, female, and a singleton. The pregnancy is a product of in vitro fertilization; consequently, the estimated gestational age is likely accurate. When the infant’s characteristics of female gender, 22 weeks’ gestation, singleton pregnancy are put into the NICHD calculator, the results predict a 91% chance of death or moderate to severe neurodevelopmental impairment if antenatal steroids have been given before delivery. Without antenatal steroid administration, the death or moderate to severe disability rate rises to 95%. It is clear that when counseling families of periviable fetuses, the information shared and recommendations offered are in part guided by the available outcome data that is utilized.