Definitions
There are numerous terms used to refer to preterm delivery at a very early gestational age (eg, extremely low gestational age, extreme preterm, very preterm), at a very small size (extremely low birthweight, micro preemie), of a very immature fetus (immature born, marginally viable), or near the limit of potential survival (margin of viability, border of viability, threshold of viability, periviable). Each encompasses a gestational age spectrum that includes high rates of mortality and severe morbidities among survivors at the lower end and significantly higher rates of survival and survival without major disabilities at the upper end. We recognize that there is no ideal definition and that no phrase precisely reflects all components of the epidemiology and the dilemmas associated with decision-making during this time frame. We have opted for the phrase “periviable birth,” defined as delivery occurring from 20 0/7 through 25 6/7 weeks of gestation, to reflect the gestational age range in which survival rates range from 0% at 20 to >50% at 25 weeks of gestation.
Morbidity and mortality rates and ethical dilemmas
Approximately 0.4–0.5% of all births occur ≤27 weeks of gestation, and these account for >40% of infant deaths and most neonatal deaths. Infants born at 20 and 21 weeks of gestation do not survive, irrespective of resuscitation efforts. The survival data for births at 22, 23, 24, and 25 weeks of gestation (excluding infants with birthweight lower than 401 g, >1000 g, or higher than the 97th percentile for gestational age; infants with ambiguous genitalia or major anomalies; and survivors not requiring mechanical ventilation) reported from the National Institute of Child Health and Human Development Neonatal Research Network were 6%, 26%, 55%, and 72%, respectively, at initial discharge from the hospital. Investigators from the same network later reported an additional 2.2% of extremely low gestational age infants (<27 weeks of gestation and birthweight <1000 g) died after discharge and at <22 months of age. Many studies have noted that, from the mid 1980s through the late 2000s, there has been an increasing rate of survival after periviable birth. However, survival data for periviable births from the reports published since 2000 show remarkable variability in outcomes among studies ( Table 1 ). There are multiple reasons for this variation, some of which include nonmodifiable factors (eg, fetal sex and weight, singleton birth), modifiable factors (eg, intent to intervene, antenatal corticosteroid administration, life-sustaining interventions at birth), and study design and reporting features (eg, single-center, regional, or national data; definition of mortality; inclusion of all live births vs resuscitated newborn infants vs only those admitted to a NICU; Table 2 ). Other factors that might affect the reported survival are local practices and protocols for withholding or withdrawing intensive care after birth.
| Authors | Time frame | Study setting | Inclusions | Exclusions | Sample size | Survival, % | |||
|---|---|---|---|---|---|---|---|---|---|
| 22 wk | 23 wk | 24 wk | 25 wk | ||||||
| Donohue et al | 1993-2001 | Retrospective cohort, single center | Women admitted with a live fetus | None | 104 (2001-2003) | 31 | |||
| Petrova et al | 1998-2001 | Retrospective cohort, single center | Neonatal intensive care unit admissions | Congenital malformations | 114 | 40 | |||
| Tyson et al | 1998-2003 | Retrospective cohort, multicenter | Live births | Birthweight <401g, >1000 g, or >97th percentile; ambiguous genitalia; major anomalies; survival without mechanical ventilation | 4446 | 51 | |||
| Mehler et al | 2000-2007 | Prospective cohort, single center | Live births | None | 208 | 41 | 76 | 82 | 80 |
| Kyser et al | 2000-2009 | Retrospective cohort, single center | Neonatal intensive care unit admissions | Major anomalies; birthweight <401 g or >1000 g | 134 | 33 | 58 | 87 | |
| Stoll et al | 2003-2007 | Retrospective cohort, multicenter | Live births | None | 4160 | 6 | 26 | 55 | 72 |
| Lee et al | 2005-2008 | Prospective cohort, multicenter state-level | Live births | Birthweight <401 g or >1000 g, or >97th percentile; major anomalies; survival without mechanical ventilation. | 3048 | 5 | 28 | 60 | |
| Variable | Effect |
|---|---|
| Data source | International, national, regional, single institution data reflect variations in regional and local practices. |
| Cohort selection | Postnatal transportation and exclusion of inborn newborn infants not surviving to neonatal intensive care unit admission results in inclusion of those with higher potential for survival. Inclusion of nonresuscitated infants reduces overall rates of survival. Inclusion of stillbirths reduces survival potential due to lack of resuscitation. Inclusion of anomalous infants may decrease survival estimate due to inclusion of lethal anomalies. |
| Gestational age assignment | In vitro fertilization and ovulation induction provide accurate gestational age assignment. Last menstrual period estimates conception on day 14. Ultrasound initially performed at <24 weeks of gestation estimates gestational age within 5-10 days. |
| Nonmodifiable risk factors | Race/ethnicity, plurality (singleton vs multiple gestation), sex, birthweight for gestational age all affect newborn infant survival but are not modifiable. |
| Modifiable obstetric practices | Antenatal interventions to improve outcomes (eg, corticosteroids, tocolysis, antibiotics, magnesium for neuroprotection, fetal monitoring, willingness to perform cesarean delivery for fetal benefit). |
| Modifiable neonatal practices | Initial resuscitation and care (eg, approaches to ventilation and oxygenation, nutritional support, and treatment of newborn infant infections) |
| Approaches to comfort care | Influenced by institutional and physician philosophies, parental wishes, and religious convictions. |
| Regional/hospital legal and practice guidelines | Policies concerning neonatal resuscitation; “self-fulfilling prophecy” of nonintervention/initial comfort care. |
Data regarding long-term outcomes are likewise complex. Woods et al reported the outcomes at 30 months of age for infants born ≤23, at 24, and at 25 weeks of gestation in 1995 in the United Kingdom and Ireland. These authors found a progressive decline in the proportion of children with “severe” disability (Bayley Scales of Infant Development-II score ≤54) with increasing gestational age, ranging from 27% at 23 weeks to 19% at 24 weeks and 17% at 25 weeks of gestation. In the same cohort followed to 6 years of age, similar trends in the rates of “overall disability” (a combined measure of cognition, neuromotor function, hearing, and vision) were evident; 25%, 29%, and 18% at 23, 24, and 25 weeks of gestation, respectively. A recent systematic review, published since the workshop was held, evaluated long-term neurodevelopmental impairment at 4–8 years among survivors born between 22 and 25 weeks of gestation. The risk of moderate-to-severe neurodevelopmental impairment decreased 6% (95% confidence interval [CI], 1.7–10.3%) for each week gained in gestational age. However, morbidities were common and the CIs for point estimates were wide: 43% (21–69%), 40% (27–54%), 28% (18–41%), and 24% (17–32%) at 22, 23, 24, and 25 weeks of gestational age, respectively. Importantly, survival after birth at 22 and 23 weeks of gestation was uncommon (n = 12 and =75, respectively). Severe neurodevelopmental impairment did not significantly decrease with increasing gestational age at birth.
When counseling parents, it is appropriate to present the data regarding the rate of survival and long-term disabilities separately, since the parents’ perspectives and the importance they give these may be different. Physicians should recognize that the parents’ views on what is a “severe” disability may be different from those of the researchers or clinicians, who traditionally report the combination of death and severe disability together. Coping with a child’s behavior problem, considered a “minor” disability in the published literature, may be difficult to handle for some families, while other families may be able to adapt more readily to disabilities typically considered to be major (eg, cerebral palsy).
An appreciation of the complex issues discussed above will help health care providers to engage in counseling efforts without being biased by one’s personal values and experiences. Future studies are needed to develop accurate prediction models to permit better counseling of families based on their individual risks, rather than based on gestational age alone.
Obstetric care
Ideally, all periviable births should occur in tertiary care centers with expertise in maternal-fetal medicine and the availability of the highest level of neonatal intensive care services. Maternity hospitals without such resources should develop partnerships with a tertiary care center, and the latter should maintain requisite resources, including timely availability of needed experts to care for the mother, her fetus, and the newborn infant. Protocols should be developed to clarify the processes for consultation and transfer as well as management prior to and during transfer (eg, antenatal corticosteroid administration for fetal maturation, magnesium sulfate for neuroprophylaxis, antibiotics for infection and group B streptococcus [GBS] prophylaxis).
Obstetric interventions at periviable gestations have included measures to delay delivery and to improve newborn infant outcomes when delivery is anticipated, for example: emergent cerclage, tocolytic therapy to delay delivery for antenatal steroid benefit, antibiotics for GBS prophylaxis or to reduce infection and prolong latency after preterm premature rupture of the membranes (PROM), antenatal corticosteroids to enhance fetal maturation, and a willingness to intervene to prevent stillbirth or fetal trauma at delivery. The potential effect of these interventions on newborn infant outcomes and maternal well-being varies according to individual circumstances and with advancing gestational age.
Evidence-based data regarding obstetric interventions for those delivering at 20-25 weeks of gestation are limited, since these gestational ages were considered nonviable in the 1970s and early 1980s when many studies on these interventions for anticipated preterm births were conducted. Further, because a small number of women actually deliver at ≤25 weeks of gestation, most studies and metaanalyses involving these patients lack power to assess the effect of such interventions.
The workshop panel reviewed available literature and considered interventions and the levels of supporting evidence across the spectrum of periviable gestations. Some of these interventions pose little risk to the mother while others impose a significant burden.
Tocolytic therapy is proposed to reduce uterine activity for the purpose of delaying delivery to increase the time for antenatal corticosteroid effects. However, data regarding currently available therapeutic tocolytic agents fail to consistently demonstrate either pregnancy prolongation beyond 24-48 hours or newborn infant benefits, and no studies specifically address women with preterm labor or PROM at 20-25 weeks of gestation.
Physical examination–indicated cervical cerclage is performed when the fetal membranes are seen to bulge to or past the external cervical os in the absence of contractions. Observational studies describe that physical examination–indicated cerclage, performed at an average gestational age of 22 weeks, can be associated with a mean pregnancy prolongation of 7-9 weeks, compared with 2-3 weeks for those treated without cerclage placement, as well as increased live birth and neonatal survival.
Antenatal corticosteroid administration is one of the most effective antenatal interventions to improve infant outcomes. Lung tissue in explant culture from 12–24-week human fetuses has been shown to respond to corticosteroids with an increase in epithelial maturation and the appearance of lamellar bodies. Data from the Eunice Kennedy Shriver Neonatal Research Network observational cohort revealed death or neurodevelopmental impairment at 18-22 months to be lower for infants exposed to antenatal corticosteroids and born at 23 weeks of gestation (83.4% vs 90.5%; adjusted odds ratio [AOR], 0.58; 95% confidence interval [CI], 0.42–0.80), at 24 weeks (68.4% vs 80.3%; AOR, 0.62; 95% CI, 0.49–0.78), and at 25 weeks (52.7% vs 67.9%; AOR, 0.61; 95% CI, 0.50–0.74), but not at 22 weeks of gestation (90.2% vs 93.1%; AOR 0.80; 95% CI, 0.29–2.21). Death, intraventricular hemorrhage (IVH), or periventricular leukomalacia and death or necrotizing enterocolitis were also significantly less frequent among infants born at 23, 24, and 25 weeks of gestation after antenatal corticosteroid exposure. The benefits regarding reduced death persisted through 18-22 months (OR, 0.59; 95% CI, 0.53–65) for infants born at 22-25 weeks of gestation. Similarly, Mori et al reported that infants exposed to antenatal corticosteroids before birth at 24-25 weeks of gestation had less frequent respiratory distress syndrome (OR, 0.77; 95% CI, 0.60–0.98), less frequent severe IVH (OR, 0.49; 95% CI, 0.36–0.67), and lower mortality rates (OR, 0.65; 95% CI, 0.5–0.86) compared with unexposed infants at the same gestations. In fact, even among the infants born at 22-23 weeks of gestation, antenatal corticosteroid exposure decreased mortality rates (OR, 0.72; 95% CI, 0.53–0.97). In 2008, Tyson et al estimated that antenatal corticosteroid administration increased the “functional” (in terms of maturity) gestational age of those born at 22-25 weeks of gestation by 1.1, 1.2, and 1.3 weeks for survival, death or profound impairment, and death or any impairment, respectively.
Magnesium sulfate for neuroprotection has been studied among women at risk for imminent early preterm birth in 5 randomized controlled trials, including women recruited at 24-25 weeks of gestation. Overall, magnesium sulfate treatment reduced cerebral palsy (relative risk [RR], 0.68; 95% CI, 0.54–0.87) and substantial gross motor dysfunction (RR, 0.61; 95% CI, 0.44–0.85) among survivors without increasing mortality rates (RR, 1.04; 95% CI, 0.92–1.17). Similar effects were seen with administration at <30 weeks of gestation for any cerebral palsy (RR, 0.69; 95% CI, 0.52–0.92), moderate-to-severe cerebral palsy (RR, 0.54; 95% CI, 0.36–0.80), and death (RR, 1.00; 95% CI, 0.87–1.15). However, data specific to those treated at 20-25 weeks of gestation are not available.
Intrapartum antibiotic prophylaxis against GBS reduces newborn infant infection, and antibiotic treatment during conservative management of preterm PROM both prolongs pregnancy and reduces newborn infant infections. Studies of these interventions have included only limited numbers of women near the limit of viability and specific data for those at 20-25 weeks of gestation are lacking.
Cesarean delivery
If a decision is made to provide intensive interventions for an infant in the periviable period, a second decision is required regarding the mode of delivery. This second decision can be divided in 2: is routine cesarean delivery appropriate for all such pregnancies? And, if not, will emergency cesarean delivery be considered to prevent trauma, stillbirth, or fetal asphyxia for specific indications? The published literature regarding cesarean delivery for periviable birth is limited by a lack of adequate data reflecting the causes, interventions, and contribution of current practices on outcomes related to cesarean delivery for periviable births. Further, randomized controlled trials of adequate size regarding planned cesarean delivery compared with planned vaginal delivery for periviable births have not been performed. Currently available data do not consistently support routine cesarean delivery to improve perinatal mortality rates or neurologic outcomes for early preterm infants. There is no clear evidence that routinely performed cesarean delivery improves survival or long-term outcomes with growth restriction, and data suggesting improved outcomes with cesarean delivery for fetal malpresentation are limited. Alternatively, cesarean delivery in the periviable period incurs greater maternal morbidity, both immediately postoperatively and for future pregnancies, which must be considered in the risk-benefit balance when counseling women.
Within this framework, a team approach to counseling is recommended for those presenting at ≥20 weeks of gestation. The use of different obstetric interventions should be based on an individual analysis of the risks and benefits. When death is anticipated, the parents should be informed about the option of termination of pregnancy if this is consistent with regional statutes. A plan for reevaluation and follow-up counseling should also be in place. Importantly, providers and families should understand that initiation of intervention to enhance outcomes (eg, antibiotics for preterm PROM, antenatal corticosteroid administration) does not mandate that all other aggressive interventions (eg, cesarean delivery) be undertaken regardless of clinical circumstances in the periviable period.
Optimally, guidance regarding perinatal management of anticipated or imminent periviable birth would be offered based on a firm knowledge of the likelihood of infant survival and a known likelihood of long-term morbidities. Gestational age alone and currently available predictive algorithms do not provide information that is sufficiently accurate or generalizable. Regardless of local or regional differences, there are substantial current data supporting that infants born ≤21 weeks of gestation do not survive after birth, regardless of aggressive intervention, and that most infants born ≥24 weeks of gestation do survive if live-born and resuscitated. Alternatively, at 22-23 weeks of gestation, most live-born resuscitated infants will not survive, and it is likely that local and other individual factors will have the greatest effect on outcomes for these infants born at the cusp of viability. In certain circumstances (eg, unknown or irregular menstrual history with late prenatal care), gestational age cannot be determined accurately. In this circumstance, gestational age is estimated based on data available when periviable birth is imminent, and the accuracy of this estimate should be considered during counseling and decision-making.
Table 3 offers guidance based on current evidence and expert opinion. In many cases, data specific to the periviable period are lacking, so guidance is offered based on extrapolation from available data regarding interventions at more advanced gestational ages. Interventions to delay delivery prior to 22 weeks of gestation may not succeed in prolongation of pregnancy. In such circumstances, it is appropriate to withhold continued intrapartum interventions for fetal benefit, neonatal resuscitation, or both, despite initiation of aggressive therapy. With delivery at <22 weeks of gestation, interventions that significantly increase maternal morbidities (eg, cesarean delivery) should be avoided, where possible, and the live-born infants should be offered comfort care. Because most newborn infants at 24-25 weeks of gestation will survive if resuscitated, efforts to prolong pregnancy, intrapartum interventions for fetal benefit, and neonatal resuscitation should generally be offered, if appropriate. At 22-23 weeks of gestation, management decisions will need to be made based on whether the fetus is considered potentially viable based on individual clinical circumstances and whether the family desires aggressive measures to improve the potential for newborn infant survival after birth. In general, those born at 23 weeks of gestation should be considered potentially viable, as survival with resuscitation is 26-28% or more. Those considered nonviable at 22-23 weeks of gestation can be treated similarly to pregnancies at 20-21 weeks of gestation, while those considered potentially viable should be treated in a manner consistent with similar pregnancies at 24-25 weeks of gestation. If feasible, delivery of potentially viable infants should be undertaken in settings in which resources are available to care for extremely small and immature infants. This approach has the potential to increase the opportunity for survival and reduce morbidities among survivors.
| Variable | Weeks of gestation b | ||
|---|---|---|---|
| <22 0/7 | 22 0/7-22 6/7 | ≥23 0/7 | |
| Antenatal corticosteroids | Not recommended | Consider if delivery at ≥23 0/7 is anticipated | Recommended |
| Tocolytics to enhance latency for potential steroid benefit | Not recommended | Not recommended unless concurrent with antenatal steroids | Consider |
| Magnesium sulfate for neuroprotection | Not recommended | Not recommended | Recommended |
| Antibiotics for preterm premature rupture of membranes to enhance latency | Consider if delivery not imminent | Consider if delivery not imminent | Recommended if delivery not imminent |
| Intrapartum antibiotics for group B streptococcus prophylaxis c | Not recommended | Not recommended | Recommended |
| Continuous intrapartum electronic fetal monitoring | Not recommended | Not recommended | Recommended |
| Cesarean delivery for fetal indication d | Not recommended | Not recommended | Recommended |
| Aggressive newborn infant resuscitation | Not recommended, comfort care only | Not recommended unless considered potentially viable based on individual circumstances | Recommended unless considered nonviable based on individual circumstances |
a Survival of infants born in the periviable period is dependent on resuscitation and support. Between 22 weeks and 25 weeks of gestation, there may be mitigating factors (eg, intrauterine growth restriction, small fetal size, the presence of fetal malformations or aneuploidy, and pulmonary hypoplasia due to prolonged membrane rupture) that will affect the potential for survival and the determination of viability ( Table 2 ). The majority of survivors born at 25 6/7 weeks of gestation or less will incur major morbidities, regardless of gestational age at birth
b Infants born at <22 0/7 weeks of gestation are generally considered nonviable. Data from recent large studies suggest survival with delivery at 22 0/7 through 22 6/7 weeks of gestation to be 5-6%. With survival rates of approximately 26-28% and higher, infants born at 23 0/7 weeks through 25 6/7 weeks of gestation are generally considered potentially viable ( Tables 1 and 2 )
c Group B streptococcus carrier or carrier status unknown
d For example, persistently abnormal fetal heart rate patterns or biophysical testing (category II–III).
Obstetric care
Ideally, all periviable births should occur in tertiary care centers with expertise in maternal-fetal medicine and the availability of the highest level of neonatal intensive care services. Maternity hospitals without such resources should develop partnerships with a tertiary care center, and the latter should maintain requisite resources, including timely availability of needed experts to care for the mother, her fetus, and the newborn infant. Protocols should be developed to clarify the processes for consultation and transfer as well as management prior to and during transfer (eg, antenatal corticosteroid administration for fetal maturation, magnesium sulfate for neuroprophylaxis, antibiotics for infection and group B streptococcus [GBS] prophylaxis).
Obstetric interventions at periviable gestations have included measures to delay delivery and to improve newborn infant outcomes when delivery is anticipated, for example: emergent cerclage, tocolytic therapy to delay delivery for antenatal steroid benefit, antibiotics for GBS prophylaxis or to reduce infection and prolong latency after preterm premature rupture of the membranes (PROM), antenatal corticosteroids to enhance fetal maturation, and a willingness to intervene to prevent stillbirth or fetal trauma at delivery. The potential effect of these interventions on newborn infant outcomes and maternal well-being varies according to individual circumstances and with advancing gestational age.
Evidence-based data regarding obstetric interventions for those delivering at 20-25 weeks of gestation are limited, since these gestational ages were considered nonviable in the 1970s and early 1980s when many studies on these interventions for anticipated preterm births were conducted. Further, because a small number of women actually deliver at ≤25 weeks of gestation, most studies and metaanalyses involving these patients lack power to assess the effect of such interventions.
The workshop panel reviewed available literature and considered interventions and the levels of supporting evidence across the spectrum of periviable gestations. Some of these interventions pose little risk to the mother while others impose a significant burden.
Tocolytic therapy is proposed to reduce uterine activity for the purpose of delaying delivery to increase the time for antenatal corticosteroid effects. However, data regarding currently available therapeutic tocolytic agents fail to consistently demonstrate either pregnancy prolongation beyond 24-48 hours or newborn infant benefits, and no studies specifically address women with preterm labor or PROM at 20-25 weeks of gestation.
Physical examination–indicated cervical cerclage is performed when the fetal membranes are seen to bulge to or past the external cervical os in the absence of contractions. Observational studies describe that physical examination–indicated cerclage, performed at an average gestational age of 22 weeks, can be associated with a mean pregnancy prolongation of 7-9 weeks, compared with 2-3 weeks for those treated without cerclage placement, as well as increased live birth and neonatal survival.
Antenatal corticosteroid administration is one of the most effective antenatal interventions to improve infant outcomes. Lung tissue in explant culture from 12–24-week human fetuses has been shown to respond to corticosteroids with an increase in epithelial maturation and the appearance of lamellar bodies. Data from the Eunice Kennedy Shriver Neonatal Research Network observational cohort revealed death or neurodevelopmental impairment at 18-22 months to be lower for infants exposed to antenatal corticosteroids and born at 23 weeks of gestation (83.4% vs 90.5%; adjusted odds ratio [AOR], 0.58; 95% confidence interval [CI], 0.42–0.80), at 24 weeks (68.4% vs 80.3%; AOR, 0.62; 95% CI, 0.49–0.78), and at 25 weeks (52.7% vs 67.9%; AOR, 0.61; 95% CI, 0.50–0.74), but not at 22 weeks of gestation (90.2% vs 93.1%; AOR 0.80; 95% CI, 0.29–2.21). Death, intraventricular hemorrhage (IVH), or periventricular leukomalacia and death or necrotizing enterocolitis were also significantly less frequent among infants born at 23, 24, and 25 weeks of gestation after antenatal corticosteroid exposure. The benefits regarding reduced death persisted through 18-22 months (OR, 0.59; 95% CI, 0.53–65) for infants born at 22-25 weeks of gestation. Similarly, Mori et al reported that infants exposed to antenatal corticosteroids before birth at 24-25 weeks of gestation had less frequent respiratory distress syndrome (OR, 0.77; 95% CI, 0.60–0.98), less frequent severe IVH (OR, 0.49; 95% CI, 0.36–0.67), and lower mortality rates (OR, 0.65; 95% CI, 0.5–0.86) compared with unexposed infants at the same gestations. In fact, even among the infants born at 22-23 weeks of gestation, antenatal corticosteroid exposure decreased mortality rates (OR, 0.72; 95% CI, 0.53–0.97). In 2008, Tyson et al estimated that antenatal corticosteroid administration increased the “functional” (in terms of maturity) gestational age of those born at 22-25 weeks of gestation by 1.1, 1.2, and 1.3 weeks for survival, death or profound impairment, and death or any impairment, respectively.
Magnesium sulfate for neuroprotection has been studied among women at risk for imminent early preterm birth in 5 randomized controlled trials, including women recruited at 24-25 weeks of gestation. Overall, magnesium sulfate treatment reduced cerebral palsy (relative risk [RR], 0.68; 95% CI, 0.54–0.87) and substantial gross motor dysfunction (RR, 0.61; 95% CI, 0.44–0.85) among survivors without increasing mortality rates (RR, 1.04; 95% CI, 0.92–1.17). Similar effects were seen with administration at <30 weeks of gestation for any cerebral palsy (RR, 0.69; 95% CI, 0.52–0.92), moderate-to-severe cerebral palsy (RR, 0.54; 95% CI, 0.36–0.80), and death (RR, 1.00; 95% CI, 0.87–1.15). However, data specific to those treated at 20-25 weeks of gestation are not available.
Intrapartum antibiotic prophylaxis against GBS reduces newborn infant infection, and antibiotic treatment during conservative management of preterm PROM both prolongs pregnancy and reduces newborn infant infections. Studies of these interventions have included only limited numbers of women near the limit of viability and specific data for those at 20-25 weeks of gestation are lacking.
Cesarean delivery
If a decision is made to provide intensive interventions for an infant in the periviable period, a second decision is required regarding the mode of delivery. This second decision can be divided in 2: is routine cesarean delivery appropriate for all such pregnancies? And, if not, will emergency cesarean delivery be considered to prevent trauma, stillbirth, or fetal asphyxia for specific indications? The published literature regarding cesarean delivery for periviable birth is limited by a lack of adequate data reflecting the causes, interventions, and contribution of current practices on outcomes related to cesarean delivery for periviable births. Further, randomized controlled trials of adequate size regarding planned cesarean delivery compared with planned vaginal delivery for periviable births have not been performed. Currently available data do not consistently support routine cesarean delivery to improve perinatal mortality rates or neurologic outcomes for early preterm infants. There is no clear evidence that routinely performed cesarean delivery improves survival or long-term outcomes with growth restriction, and data suggesting improved outcomes with cesarean delivery for fetal malpresentation are limited. Alternatively, cesarean delivery in the periviable period incurs greater maternal morbidity, both immediately postoperatively and for future pregnancies, which must be considered in the risk-benefit balance when counseling women.
Within this framework, a team approach to counseling is recommended for those presenting at ≥20 weeks of gestation. The use of different obstetric interventions should be based on an individual analysis of the risks and benefits. When death is anticipated, the parents should be informed about the option of termination of pregnancy if this is consistent with regional statutes. A plan for reevaluation and follow-up counseling should also be in place. Importantly, providers and families should understand that initiation of intervention to enhance outcomes (eg, antibiotics for preterm PROM, antenatal corticosteroid administration) does not mandate that all other aggressive interventions (eg, cesarean delivery) be undertaken regardless of clinical circumstances in the periviable period.
Optimally, guidance regarding perinatal management of anticipated or imminent periviable birth would be offered based on a firm knowledge of the likelihood of infant survival and a known likelihood of long-term morbidities. Gestational age alone and currently available predictive algorithms do not provide information that is sufficiently accurate or generalizable. Regardless of local or regional differences, there are substantial current data supporting that infants born ≤21 weeks of gestation do not survive after birth, regardless of aggressive intervention, and that most infants born ≥24 weeks of gestation do survive if live-born and resuscitated. Alternatively, at 22-23 weeks of gestation, most live-born resuscitated infants will not survive, and it is likely that local and other individual factors will have the greatest effect on outcomes for these infants born at the cusp of viability. In certain circumstances (eg, unknown or irregular menstrual history with late prenatal care), gestational age cannot be determined accurately. In this circumstance, gestational age is estimated based on data available when periviable birth is imminent, and the accuracy of this estimate should be considered during counseling and decision-making.
Table 3 offers guidance based on current evidence and expert opinion. In many cases, data specific to the periviable period are lacking, so guidance is offered based on extrapolation from available data regarding interventions at more advanced gestational ages. Interventions to delay delivery prior to 22 weeks of gestation may not succeed in prolongation of pregnancy. In such circumstances, it is appropriate to withhold continued intrapartum interventions for fetal benefit, neonatal resuscitation, or both, despite initiation of aggressive therapy. With delivery at <22 weeks of gestation, interventions that significantly increase maternal morbidities (eg, cesarean delivery) should be avoided, where possible, and the live-born infants should be offered comfort care. Because most newborn infants at 24-25 weeks of gestation will survive if resuscitated, efforts to prolong pregnancy, intrapartum interventions for fetal benefit, and neonatal resuscitation should generally be offered, if appropriate. At 22-23 weeks of gestation, management decisions will need to be made based on whether the fetus is considered potentially viable based on individual clinical circumstances and whether the family desires aggressive measures to improve the potential for newborn infant survival after birth. In general, those born at 23 weeks of gestation should be considered potentially viable, as survival with resuscitation is 26-28% or more. Those considered nonviable at 22-23 weeks of gestation can be treated similarly to pregnancies at 20-21 weeks of gestation, while those considered potentially viable should be treated in a manner consistent with similar pregnancies at 24-25 weeks of gestation. If feasible, delivery of potentially viable infants should be undertaken in settings in which resources are available to care for extremely small and immature infants. This approach has the potential to increase the opportunity for survival and reduce morbidities among survivors.
