Neonatal Survival at the Threshold of Viability: Does Location Matter?

Maureen A. O'Reilly, DNP, NNP-BC


August 24, 2018

'Will You Try to Save Our Baby?'

Graduate students Caroline and Dan were surprised to find they were expecting their first child. Nevertheless, they began preparing for their unexpected newborn by doing what they did best: research. Reading about pregnancy and delivery, they anticipated a healthy full-term baby. But at 22 3/7 weeks, Caroline awoke to uterine cramping and bleeding. Birth was imminent. Emotional and apprehensive, their reaction was to seek information online. They read studies based on 22-week survival and watched a news video: "Born Before 22 Weeks, 'Most Premature' Baby Is Now Thriving."[1] When a neonatologist arrived to talk with them, Dan asked, "Will you try to save our baby? They can do it in Japan."

Extremely Premature Newborn Survival in First-World Countries

The neonatologist will share with Caroline and Dan key statistics relevant to the birth of their 22-week infant, and the difficult decisions they must make. But what difference did it make if a 22-week infant had a better chance of survival if born in Japan?

Consider the range of 22 weeks' gestation survival rates among first-world countries (Table):

Table. Neonatal Survival at 22 Weeks[2]

Country Rate of Neonatal Survival
France 0.7%
United Kingdom 2%
United States 5.1%
Sweden 9.8%
Japan 33.1%

Wide ranges are also found at 23 weeks (from 1% to 52%) and 24 weeks (from 31% to 67%), with Japan again reporting the highest survival rates. These data suggest that survival rates for extremely premature infants born in Japan are the best in the world. Is that true, and if so, why?

In "An International Comparison of Death Classification at 22 to 25 Weeks' Gestational Age," Smith and colleagues[2] attempted to reconcile large, and—on the surface—surprising differences in extremely premature infant survival rates among first-world countries. Using national birth registry data, they used five denominators—all births, live births, births alive at the onset of labor, live births surviving to 1 hour, and live births surviving to 24 hours—in the United States, Canada, Japan, United Kingdom, Norway, Sweden, and Finland.

They found that variations in survival rates at 22-23 weeks' gestation are partially caused by differences in how each country defines and reports "live births." Although all countries had mandatory registration of live births, stillbirth registration differed by nation.[2] When stillbirths are included in the denominator, differences in international neonatal survival rates were substantially lower.[2]

Survival rates of infants born at 22 weeks (using live births as the denominator) varied from 3.7% to 56.7% among the seven nations. However, the range shrank to 1.8% to 22.3% when the denominator changed to include stillbirths and fetuses alive at the onset of labor. Survival rates also changed when early deaths were excluded or were limited to infants surviving at least 12 hours. The trend was similar for survival at 23 weeks' gestation.[2]

Influences on Survival Rates

Survival rates were influenced by factors beyond mathematical manipulations. Reported survival rates in all seven countries were affected by differences in the quality of care, gestation-based decision-making on whether to resuscitate, and how to register the birth. This death classification study also notes a relationship between survival rates and provision of "active care."[2]

Obstetric practice differences among the seven countries contributed to infant survival rates. These include whether to terminate a pregnancy, aggressive treatment of periviable infants, corticosteroid administration, antenatal transfer to a level III perinatal center, and use of cesarean section for fetal indications.[2]

Neonatologist Jeffrey Garland, MD, urges caution in interpreting neonatal survival data based solely on the live birth rate: "I'd continue to use the NICHD [National Institute of Child Health and Human Development] data when I talk with parents of a 22- to 23-week baby. You have to remember your location. Where you are, how evidence-based the guidelines are in the hospital where you deliver, whether all the resources are there for supporting a 22-weeker—it all matters. Japan may have better survival numbers, but the question is whether they count their births the way we do."

Getting the Right Denominator

Kyle Mounts, deputy executive director of programs for the Wisconsin Association for Perinatal Care (WAPC), agrees that the choice of denominator can make all the difference in meaning. "The denominator really matters. You can't compare data readily without agreement on the types of data and how they are defined—it skews how you perceive actual outcomes." WAPC, one of the oldest perinatal collectives in the United States, uses Wisconsin's PeriData.Net®, a comprehensive perinatal database used by birth hospitals in Wisconsin.[3]

Deborah Ehrenthal, MD, MPH, associate professor and director of the Division of Reproductive and Population Health at the University of Wisconsin, commented on Smith and colleagues' findings, saying, "Other significant factors may be involved in extremely preterm birth survival rates. We must acknowledge differences in both numerator and denominator data and ultimately, differences in perinatal and fetal factors." In 2011, she looked at variations among all 50 states in how outcomes were classified for infants born weighing 500 g. Fetal and early infant death outcomes for these infants reflected differences in classification, ultimately affecting vital statistics at the state level.[4]

Five European countries (Belgium, France, Italy, Portugal, United Kingdom) reported large international differences in survival rates for extremely preterm and very low birth weight births. In another study, Smith and colleagues[5] evaluated 1449 live births and fetal deaths between 22 and 25 weeks' gestation born in 2011-2012, looking at international similarities and differences. In all five countries, the number of births recorded as "live born" was consistently low at 22 weeks and consistently high at 25 weeks. At 22 weeks, survival to discharge was universally poor, as it was for any infant born weighing less than 500 g, regardless of treatment regimens. The use of antenatal steroids and respiratory support at 22-24 weeks varied among these nations, but both were uniformly high at 25 weeks. As a result, the authors recommended including birth weight along with gestational age in ethical guidelines for decision-making.[5]

Is Japan the Best Place to Deliver an Extremely Premature Infant?

Inoue and colleagues[6] analyzed demographic and clinical data of 1473 live-born infants weighing 500 g or less at gestation age 22 weeks or greater, treated in the 204 affiliated Neonatal Research Network hospitals in Japan between 2003 and 2012. Their review was mixed. Improvements in perinatal-neonatal medicine increased survival rates in infants of 500 g or less from 40% to 68%, but most survivors had major morbidities that might affect their long-term prognosis.[6]

Morisaki and colleagues[7] examined survival among 22- to 24-week infants born in Japan. Japan's national vital statistics database of stillbirths and live births from 2003-2011 included 14,726 singleton births at 22-24 weeks. They credited institutional factors (eg, hospital protocols) as having some impact on the decision to resuscitate extremely premature infants. But did parental factors, such as income, affect those decisions? Their results related socioeconomic factors to the rate of survival at delivery and the first hour post-birth. Death at 22-24 weeks was twice as likely to occur with low-income multiparous teenage mothers versus older, wealthier primiparous women. Socioeconomic factors substantially influenced whether infants of 22-24 weeks survived delivery and the first hour of life.[7]

Interventions and Survival

Kollée and colleagues[8] compared obstetric interventions for extremely premature birth, assessing their impact on mortality and short-term morbidity. The study used data from 10 regions in nine European countries. Three interventions were identified—corticosteroids, antenatal transfer, and cesarean section for fetal indications—and the study looked at outcomes associated with the number of interventions. The rate of interventions for 22- to 23-, 24- to 25-, and 26- to 27-week infants were compared with the rate for 28- to 29-week infants. Measured outcomes included stillbirth and in-hospital mortality, as well as neonatal complications. Large differences were seen between European regions in interventions offered at 22-23 and 24-25 weeks. Obstetric interventions had the most impact on outcomes among infants born at 24-25 weeks.[8]

The Effective Perinatal Intensive Care in Europe [EPICE] for very preterm births project is a multinational population-based cohort from 19 regions in 11 European countries. Zeitlin and colleagues[9] evaluated four high-evidence practices (admission to an appropriate level of care nursery; use of antenatal corticosteroids; prevention of hypothermia; surfactant within 2 hours of birth or early nasal continuous positive airway pressure), assessing the potential reduction of morbidity and mortality in very preterm infants born before 28 weeks. Only 58.3% of infants received all interventions for which they were eligible; infants of low gestational age were less likely to receive evidence-based care.[9]

Chen and colleagues[10] asked why infant mortality is higher in the United States than in Europe. Using economics theory, they questioned how mortality rate interpretation affects national policy. After adjustment for potential differences in reported births near the threshold of viability, higher infant mortality numbers persisted for the United States. Compared with Europe, the United States had similar neonatal mortality (< 1 month of age) but higher post-neonatal mortality (1-12 months of age).[10]

Dan and Caroline, Continued

Dan and Caroline wanted an answer: Would their baby survive in the United States if a 22-week infant could survive in Japan? Answering this question isn't simple. The death classification study comparing differences according to geography makes it clear that neonatal survival numbers may not tell the whole story.

The lower premature birth rate and higher survival rate in Japan are attributed to a well-developed healthcare system covering prenatal care and all costs for premature infants, readily available specialized equipment, and focused training. Moreover, specialists are present at all times in the neonatal intensive care unit. This doesn't sound drastically different from what occurs in most countries with similarly well-developed healthcare systems, so the survival differences, if real, may be a consequence of a different stance toward the resuscitation and care of a 22-week newborn.

After several well-publicized cases of survival at 22-week gestation, a 1990 law requires Japanese doctors to resuscitate babies from the age of 22 weeks. Hiroshi Nishida, a neonatologist involved in drafting the law, considers physicians' attitudes as a decisive factor in Japan's success with extreme prematurity, emphatically stating that preemies are human beings and everything must be done to help them survive.[11] He believes that the argument that a child might be disabled isn't valid. "We should not engage in social Darwinism," he says. "All life is equal."[11]

Survival rate data need reevaluation in light of the conclusions of the study by Smith and colleagues.[2] We lack common denominators to compare rates of survival at the edge of viability in the United States with those in Europe and Japan. Until definitions are established, discussing an extremely preterm baby's potential for survival with parents leaves our predictions dependent on where the baby is born and what resources are available. We don't really know whether our survival rates are comparable to Japan's or Sweden's if we are measuring apples while they count oranges.

If our goal is to lower the age of viability, we need national evidence-based guidelines on resuscitating extremely premature infants. And we've never held a national debate on social justice and the distribution of resources. In a market-based healthcare system, would the American public agree with Nishida that we must not practice social Darwinism?


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: