CHAPTER 43: Postterm Pregnancy

The above passage from Williams shows that pregnancies exceeding the expected normal length were problematic more than 100 years ago. These postterm pregnancies remain so today.

The adjectives postterm, prolonged, postdates, and postmature are often loosely used interchangeably to describe pregnancies that have exceeded a duration considered to be the upper limit of normal. We eschew use of the term postdates because the real issue in many postterm pregnancies is “post-what dates?” Postmature is reserved for the relatively uncommon specific clinical fetal syndrome in which the newborn has recognizable features indicating a pathologically prolonged pregnancy. Therefore, postterm or prolonged pregnancy is our preferred expression for an extended pregnancy.

The international definition of prolonged pregnancy, endorsed by the American College of Obstetricians and Gynecologists (2016b,d) is one that exceeds 42^0/7 weeks, namely, 294 days or more from the first day of the last menstrual period. Importantly, this is 42 “completed weeks,” as pregnancies between 41 weeks 1 day and 41 weeks 6 days, although in the 42nd week, do not complete 42 weeks until the seventh day has elapsed. The method that we use widely in this book is to divide the 42nd week into 7 days, that is, 42^0/7 through 42^6/7 weeks.

The current definition of postterm pregnancy assumes that the last menses was followed by ovulation 2 weeks later. That said, some pregnancies may not actually be postterm. Instead, the calculation may reflect an error in gestational age estimation because of faulty menstrual date recall or delayed ovulation. Thus, the two categories of pregnancies that reach 42 completed weeks are those truly 40 weeks past conception and those of less-advanced gestation but with inaccurately estimated gestational age. Even with exactly recalled menstrual dates, there still is imprecision, and the American College of Obstetricians and Gynecologists (2016d, 2017b) considers first-trimester sonography to be the most accurate method to establish or confirm gestational age. Several clinical studies support this practice (Bennett, 2004; Blondel, 2002; Joseph, 2007).

Of the 3.93 million neonates born in the United States during 2015, 0.4 percent were delivered at 42 weeks or later (Martin, 2017). In the past, the proportion was much higher. This trend suggests earlier intervention, however, the added accuracy from earlier sonographic dating of gestational age is another factor.

To identify potential predisposing factors for postterm pregnancy, Olesen and associates (2006) analyzed various characteristics in the Danish Birth Cohort. Only prepregnancy body mass index (BMI) ≥25 and nulliparity were significantly associated with prolonged pregnancy. Mission (2015) and Arrowsmith (2011) and their coworkers also reported similar associations. In nulliparas, those whose cervical length at midpregnancy is longer, that is, in the third or fourth quartile, are twice as likely to deliver after 42 weeks (van der Ven, 2016).

The tendency for some mothers to have repeated postterm births suggests that some prolonged pregnancies are biologically determined. Oberg and colleagues (2013) reported that when mother and daughter had a prolonged pregnancy, the risk for the daughter to have a subsequent postterm pregnancy was significantly increased. Laursen and associates (2004) found that maternal, but not paternal, genes influenced prolonged pregnancy. As discussed in Chapter 5 (Placental Estriol Synthesis), rare fetal-placental factors that predispose to postterm pregnancy include anencephaly, adrenal hypoplasia, and X-linked placental sulfatase deficiency (Ayyavoo, 2014; MacDonald, 1965).

Rates of stillbirth, neonatal death, and infant morbidity all rise after the expected due date has passed. This is best seen when perinatal mortality rates are analyzed from times before widespread intervention for postterm pregnancies. In two large Swedish studies shown in Figure 43-1, after reaching a nadir at 39 to 40 weeks, the perinatal mortality rate rose as pregnancy duration exceeded 41 weeks. This trend is also reported for the United States (Cheng, 2008; MacDorman, 2009). As shown in Table 43-1, the major cause of death in these studies includes gestational hypertension, prolonged labor with cephalopelvic disproportion, birth injuries, and hypoxic-ischemic encephalopathy. Similar outcomes were reported by Olesen and colleagues (2003) in 78,022 women with postterm pregnancies delivered before routine labor induction was adopted in Denmark. Moster and associates (2010) found higher rates of cerebral palsy in postterm births, and Yang and coworkers (2010) reported lower intelligence quotient (IQ) scores at age 6.5 years in children born ≥42 weeks’ gestation. Conversely, autism was not associated with postterm birth (Gardener, 2011).

Alexander and colleagues (2000a) reviewed 56,317 consecutive singleton pregnancies delivered at ≥40 weeks between 1988 and 1998 at Parkland Hospital. Labor was induced in 35 percent of pregnancies completing 42 weeks. The rate of cesarean delivery for dystocia and fetal distress was significantly greater at 42 weeks compared with earlier deliveries. More newborns of postterm pregnancies were admitted to intensive care units. Importantly, the incidence of neonatal seizures and deaths was doubled at 42 weeks. Smith (2001) has challenged analyses such as these because the population at risk for perinatal mortality in a given week consists of all ongoing pregnancies rather than just the births in a given week. He calculated perinatal mortality rates calculated using only births in a given week of gestation from 37 to 43 completed weeks compared with the cumulative probability—the perinatal index—of death when all ongoing pregnancies are included in the denominator. Using this computation, delivery at 38 weeks had the lowest risk index for perinatal death.

Postmaturity Syndrome

The postmature newborn is unique, and features include wrinkled, patchy, peeling skin; a long, thin body suggesting wasting; and advanced maturity in that the infant is open-eyed, unusually alert, and appears old and worried (Fig. 43-2). Skin wrinkling can be particularly prominent on the palms and soles. The nails are typically long. Most postmature neonates are not technically growth restricted because their birthweight seldom falls below the 10th percentile for gestational age (Chap. 44, Fetal-Growth Restriction). On the other hand, severe growth restriction—which logically must have preceded completion of 42 weeks—may be present.

The incidence of postmaturity syndrome in newborns at 41, 42, or 43 weeks, respectively, has not been conclusively determined. From data, the syndrome complicates 10 to 20 percent of pregnancies at 42 completed weeks (American College of Obstetricians and Gynecologists, 2016d). Associated oligohydramnios substantially raises the likelihood of postmaturity. Trimmer and associates (1990) reported that 88 percent of fetuses were postmature if there was oligohydramnios defined by a sonographic maximal vertical amnionic fluid pocket that measured ≤1 cm at 42 weeks.

Placental Dysfunction

Many believe that postterm pregnancy is an abnormal state. Redman and Staff (2015) posit that limited placental capacity, which is characterized by dysfunctional syncytiotrophoblast, explains the greater risks of the postmaturity syndrome.

Clifford (1954) proposed that the associated skin changes were due to loss of the protective effects of vernix caseosa. He also attributed the postmaturity syndrome to placental senescence, although he did not find placental degeneration histologically. Still, the concept that postmaturity stems from placental insufficiency has persisted despite an absence of morphological or significant quantitative findings (Larsen, 1995; Redman, 2015; Rushton, 1991). There are findings that the rate of placental apoptosis—programmed cell death—is significantly greater at 41 to 42 completed weeks compared with that at 36 to 39 weeks (Smith, 1999). Several proapoptotic genes such as kisspeptin are upregulated in postterm placental explants compared with the same genes in term placental explants (Torricelli, 2012). The clinical significance of such apoptosis is currently unclear.

Jazayeri and coworkers (1998) investigated cord blood erythropoietin levels in 124 appropriately grown newborns delivered from 37 to 43 weeks. The only known stimulator of erythropoietin is decreased partial oxygen pressure. Thus, they sought to assess whether fetal oxygenation was compromised due to placental aging in postterm pregnancies. All women had an uncomplicated labor and delivery. These investigators found that cord blood erythropoietin levels were significantly higher in pregnancies reaching 41 weeks or more. Although Apgar scores and acid-base studies were normal, these researchers concluded that fetal oxygenation was decreased in some postterm gestations.

Another scenario is that the postterm fetus may continue to gain weight and thus be unusually large at birth. This at least suggests that placental function is not severely compromised. Indeed, continued fetal growth is the norm—albeit at a slower rate beginning at 37 completed weeks (Fig. 43-3). Nahum and colleagues (1995) confirmed that fetal growth continues until at least 42 weeks. However, Link and associates (2007) showed that umbilical blood flow did not increase concomitantly.

Fetal Distress and Oligohydramnios

The principal reasons for increased risks to postterm fetuses were described by Leveno and associates (1984). Both antepartum fetal jeopardy and intrapartum fetal distress were found to be the consequence of cord compression associated with oligohydramnios. In their analysis of 727 postterm pregnancies, intrapartum fetal distress detected with electronic monitoring was not associated with late decelerations characteristic of uteroplacental insufficiency. Instead, one or more prolonged decelerations such as shown in Figure 43-4 preceded three fourths of emergency cesarean deliveries for nonreassuring fetal heart rate tracings. In all but two cases, there were also variable decelerations. Another common fetal heart rate pattern, although not ominous by itself, was the saltatory baseline. As described in Chapter 24 (Prolonged Deceleration), these findings are consistent with cord occlusion as the proximate cause of the nonreassuring tracings. Other correlates included oligohydramnios and viscous meconium. Schaffer and colleagues (2005) implicated a nuchal cord in abnormal intrapartum fetal heart rate patterns, meconium, and compromised newborn condition in prolonged pregnancies.

The volume of amnionic fluid normally continues to decline after 38 weeks and may become problematic. Moreover, meconium release into an already reduced amnionic fluid volume results in thick, viscous meconium that may cause meconium aspiration syndrome (Chap. 33, Neonatal Encephalopathy and Cerebral Palsy).

Trimmer and coworkers (1990) sonographically measured hourly fetal urine production using sequential bladder volume measurements in 38 postterm pregnancies. Diminished urine production was found to be associated with oligohydramnios. They hypothesized that decreased fetal urine flow was likely the result of preexisting oligohydramnios that limited fetal swallowing. Oz and associates (2002), using Doppler waveforms, concluded that fetal renal blood flow is reduced in those postterm pregnancies complicated by oligohydramnios. As a possible cause, again, the study by Link and associates (2007) showed that umbilical blood flow did not increase past term.

Fetal-Growth Restriction

In the late 1990s, the clinical significance of fetal-growth restriction in the otherwise uncomplicated pregnancy became more fully appreciated. Divon (1998) and Clausson (1999) and their coworkers analyzed births between 1991 and 1995 in the National Swedish Medical Birth Registry. Stillbirths were more common among growth-restricted newborns who were delivered after 42 weeks. Indeed, a third of postterm stillborn neonates were growth restricted. During this time in Sweden, labor induction and antenatal fetal testing usually commenced at 42 weeks. In a study from Parkland Hospital, Alexander and colleagues (2000d) analyzed outcomes for 355 neonates from pregnancies ≥42 weeks and whose birthweights were <3rd percentile. They compared these with outcomes of 14,520 similarly aged newborns above the 3rd percentile and found that morbidity and mortality rates were significantly increased in the growth-restricted neonates. Notably, a fourth of all stillbirths associated with prolonged pregnancy were in this comparatively small number of growth-restricted fetuses.

In the event of a medical or other obstetrical complication, it is generally not recommended that a pregnancy be allowed to continue past 42 weeks. Indeed, in many such instances, earlier delivery is indicated. Common examples include gestational hypertensive disorders, prior cesarean delivery, and diabetes. Other clinically important factors include amnionic fluid volume and potential fetal macrosomia.

Oligohydramnios

Most clinical studies support the view that diminished amnionic fluid determined by various sonographic methods identifies a postterm fetus with increased risks. Indeed, decreased amnionic fluid in any pregnancy signifies increased fetal risk (Chap. 11, Pregnancy Outcomes). Unfortunately, lack of an exact method to define “decreased amnionic fluid” has limited investigators, and many different criteria for sonographic diagnosis have been proposed. Fischer and colleagues (1993) attempted to determine which criteria were most predictive of normal versus abnormal outcomes in postterm pregnancies. As shown in Figure 43-5, the smaller the amnionic fluid pocket, the greater the likelihood that there was clinically significant oligohydramnios. Importantly, normal amnionic fluid volume did not preclude abnormal outcomes. Alfirevic and coworkers (1997) randomly assigned 500 women with postterm pregnancies to assessment using either the amnionic fluid index (AFI) or the deepest vertical pocket described in Chapter 11 (Measurement). They concluded that the AFI overestimated the number of abnormal outcomes in postterm pregnancies.

Regardless of the criteria used to diagnose oligohydramnios in postterm pregnancies, most investigators have found a higher incidence of some measure of “fetal distress” during labor. Thus, oligohydramnios by most definitions is a clinically meaningful finding. Conversely, reassurance of continued fetal well-being in the presence of “normal” amnionic fluid volume is tenuous. This may be related to how quickly pathological oligohydramnios develops. Although such cases are unusual, Clement and coworkers (1987) described six postterm pregnancies in which amnionic fluid volume diminished abruptly over 24 hours, and in one of these, the fetus died.

Macrosomia

The velocity of fetal weight gain peaks at approximately 37 weeks (see Fig. 43-3). Although growth velocity slows at that time, most fetuses continue to gain weight. For example, the percentage of fetuses born in 2009 whose birthweight exceeded 4000 g was 8.2 percent at 37 to 41 weeks and increased to 11.0 percent at 42 weeks or more (Martin, 2011). According to Duryea and associates (2014), the 95th percentile at 42 weeks is 4475 g. Even so, in some studies, brachial plexus injury was not related to postterm gestation (Walsh, 2011). Intuitively, it seems that both maternal and fetal morbidity associated with macrosomia would be mitigated with timely induction to preempt further growth. This, however, does not appear to be the case. The American College of Obstetricians and Gynecologists (2016c) has concluded that current evidence does not support such a practice in women at term with suspected fetal macrosomia. Moreover, the College concluded that in the absence of diabetes, vaginal delivery is not contraindicated for women with an estimated fetal weight up to 5000 g (Chap. 27, Shoulder Dystocia). Obvious problems with all such recommendations are substantive variations in fetal weight estimation.

Although some intervention is indicated for prolonged pregnancies, the method and timing of this are not unanimous. The decision focuses on whether labor induction is warranted or if expectant management with fetal surveillance is best. In a survey done more than 10 years ago, Cleary-Goldman and associates (2006) reported that 73 percent of members of the American College of Obstetricians and Gynecologists routinely induced women at 41 weeks. Most of the remainder performed twice weekly fetal testing until 42 weeks.

Induction Factors

Although all obstetricians know what an “unfavorable cervix” is, the term unfortunately defies precise objective definition. Thus, investigators have used differing criteria for studies of prolonged pregnancies. Harris and coworkers (1983) defined an unfavorable cervix by a Bishop score <7 and reported this in 92 percent of women at 42 weeks (Chap. 26, Preinduction Cervical Ripening). Hannah and colleagues (1992) found that 40 percent of 3407 women with a 41-week pregnancy had an “undilated cervix.” In a study of 800 women undergoing induction for postterm pregnancy at Parkland Hospital, Alexander and associates (2000b) reported that women in whom there was no cervical dilation had a twofold higher cesarean delivery rate for “dystocia.” Yang and coworkers (2004) found that cervical length ≤3 cm measured with transvaginal sonography was predictive of successful induction. In a similar study, Vankayalapati and associates (2008) found that cervical length ≤25 mm was predictive of spontaneous labor or successful induction.

Several investigators have evaluated prostaglandin E₂ (PGE₂) and E₁ (PGE₁) for induction in women with an unfavorable cervix and prolonged pregnancies. A study by the Maternal–Fetal Medicine Units Network (1994) found that PGE₂ gel was not more effective than placebo. Alexander and associates (2000c) treated 393 women with a postterm pregnancy with PGE₂, regardless of cervical “favorability,” and reported that almost half of the 84 women with cervical dilatation of 2 to 4 cm entered labor with PGE₂ use alone. In another study, mifepristone was reported to increase uterine activity without uterotonic agents in women beyond 41 weeks (Fasset, 2008). Prostaglandins and other agents used for cervical ripening are discussed in Chapter 26 (Pharmacological Techniques).

Sweeping or stripping of the membranes to induce labor and thereby prevent postterm pregnancy was studied in 15 randomized trials during the 1990s. Boulvain and coworkers (2005) performed a metaanalysis of these and found that membrane stripping at 38 to 40 weeks lowered the frequency of postterm pregnancy. Although maternal and neonatal infection rates were not increased, this practice did not modify the cesarean delivery rate. Since then, randomized trials by Wong (2002), Kashanian (2006), Hill (2008), and their coworkers found that sweeping membranes did not reduce the need to induce labor. Drawbacks of membrane stripping included pain, vaginal bleeding, and irregular contractions without labor.

The station of the fetal head within the pelvis is another predictor of successful postterm pregnancy induction. Shin and colleagues (2004) studied 484 nulliparas who underwent induction after 41 weeks. The cesarean delivery rate was directly related to station. The rate was 6 percent if the vertex before induction was at –1 station; 20 percent at –2 station; 43 percent at –3 station; and 77 percent at –4 station.

Induction versus Fetal Testing

Because of the marginal benefits from induction with an unfavorable cervix, as just discussed, some clinicians prefer instead to implement a strategy of fetal testing beginning at 41 completed weeks. For example, in a Canadian study, 3407 women were randomly assigned at 41 or more weeks to induction or to fetal testing (Hannah, 1992). In the surveillance group, evaluation included: (1) counting fetal movements during a 2-hour period each day, (2) nonstress testing three times weekly, and (3) amnionic fluid volume assessment two to three times weekly, with pockets <3 cm considered abnormal. Labor induction resulted in a small but significant reduction in the cesarean delivery rate compared with fetal testing—21 versus 24 percent, respectively. This difference was due to fewer procedures for fetal distress. There were only two stillbirths in the fetal testing group.

The Maternal–Fetal Medicine Network performed a randomized trial of induction versus fetal testing beginning at 41 weeks (Gardner, 1996). Fetal surveillance included nonstress testing and sonographic estimation of amnionic fluid volume performed twice weekly in 175 women. Perinatal outcomes were compared with those of 265 women also at 41 weeks randomly assigned to induction with or without cervical ripening. There were no perinatal deaths, and the cesarean delivery rate was not different between management groups. The results of this study could be used to support the validity of either management scheme.

In an analysis of 22 trials, Gulmezoglu and colleagues (2012) found that induction after 41 weeks rather than surveillance was associated with significantly fewer perinatal deaths and meconium aspiration syndrome cases and a lower cesarean delivery rate. In a review of two metaanalyses and a randomized study, similar conclusions were reached (Mozurkewich, 2009).

In most studies, labor induction at 42^0/7 weeks has a higher cesarean delivery rate compared with spontaneous labor. From Parkland Hospital, Alexander and coworkers (2001) evaluated pregnancy outcomes in 638 such women in whom labor was induced and compared them with outcomes of 687 women with postterm pregnancies who had spontaneous labor. Cesarean delivery rates were significantly increased—19 versus 14 percent—in the induced group because of failure to progress. When these investigators corrected for risk factors, however, they concluded that intrinsic maternal factors, rather than the induction itself, led to the higher rate. These factors included nulliparity, an unfavorable cervix, and epidural analgesia.

A large study from Denmark by Zizzo and associates (2017) is also instructive. In 2011, the Danish national guidelines were changed from labor induction at 42^0/7 weeks with no fetal surveillance to labor induction at 41^2/7 to 41^6/7 weeks with fetal surveillance beginning at 41^0/7 weeks. They compared two 3-year epochs—one before and one after 2011—and the results are shown in Table 43-2. The rate of pregnancies that progressed past 42^0/7 weeks decreased from 2.85 to 0.62 percent. Concurrently, as expected, the induction rate rose significantly, and this was accompanied by a drop in the perinatal mortality rate—22 to 13 per 1000 births. The cesarean delivery rate was not changed. A similar before-and-after observational study reported that induction at ≥42 weeks was associated with a significantly lower cesarean delivery rate—15 versus 19.4 percent (Bleicher, 2017).

From the foregoing, evidence to substantiate intervention—whether induction or fetal testing—commencing at 41 versus 42 weeks is limited. Most evidence used to justify intervention at 41 weeks is from the randomized Canadian and American investigations cited earlier. No randomized studies have specifically assessed intervention at 41 weeks versus an identical intervention used at 42 weeks. A large Swedish multicenter randomized trial of more than 10,000 women at 41^0/7 weeks has been designed to address the question (Elden, 2016).

Management Strategies

The American College of Obstetricians and Gynecologists (2016a) defines postterm pregnancies as having completed 42 weeks, namely, beyond 42^0/7 weeks. There is insufficient evidence to mandate a management strategy between 40 and 42 completed weeks. Thus, although not considered mandatory, initiation of fetal surveillance at 41 weeks is a reasonable option. After completing 42 weeks, recommendations are for labor induction as summarized in Figure 43-6.

When gestational age is uncertain, the American College of Obstetricians and Gynecologists (2017b) recommends delivery at 41 weeks’ gestation using the best clinical estimate of gestational age. The College also recommends against amniocentesis for fetal lung maturity.

At Parkland Hospital, based on results from the trials just discussed, we consider 41-week pregnancies without other complications to be normal. Thus, no interventions are practiced solely based on fetal age until 42 completed weeks. With complications such as hypertension, decreased fetal movement, or oligohydramnios, labor induction is carried out. It is our view that large, randomized trials should be performed before otherwise uncomplicated 41-week gestations are routinely considered pathologically prolonged. In women in whom a certain gestational age is known, labor is induced at the completion of 42 weeks. Almost 90 percent of such women are induced successfully or enter labor within 2 days of induction. For those who do not deliver with the first induction, a second induction is performed within 3 days. Almost all women are delivered using this management plan, but in the unusual few who are not delivered, management decisions involve a third—or even more—induction versus cesarean delivery. Women classified as having uncertain postterm pregnancies are managed with weekly nonstress fetal testing and assessment of amnionic fluid volume. Women with an AFI ≤5 cm or with reports of diminished fetal movement undergo labor induction.

Labor is a particularly dangerous time for the postterm fetus. Therefore, women whose pregnancies are known or suspected to be postterm ideally come to the hospital as soon as they suspect labor. While being evaluated for active labor, we recommend that fetal heart rate and uterine contractions be monitored electronically for variations consistent with fetal compromise.

During labor, the decision to perform amniotomy is problematic. Further reduction in fluid volume following amniotomy can enhance the possibility of cord compression. Conversely, after membrane rupture, a scalp electrode and an intrauterine pressure catheter can be placed. These usually provide more precise data concerning fetal heart rate and uterine contractions. Amniotomy also aids identification of thick meconium.

Thick meconium in the amnionic fluid is particularly worrisome. The viscosity probably signifies the lack of liquid and thus oligohydramnios. Aspiration of thick meconium may cause severe pulmonary dysfunction and neonatal death (Chap. 33, Neonatal Encephalopathy and Cerebral Palsy). Because of this, amnioinfusion during labor has been proposed as a way of diluting meconium to lower the incidence of aspiration syndrome (Wenstrom, 1989). As discussed in Chapter 24 (Management Options), the benefits of amnioinfusion remain controversial. In a large randomized trial by Fraser and colleagues (2005), amnioinfusion did not reduce the risk of meconium aspiration syndrome or perinatal death. According to the American College of Obstetricians and Gynecologists (2016a), amnioinfusion does not prevent meconium aspiration, however, it remains a reasonable treatment approach for repetitive variable decelerations.

The likelihood of a successful vaginal delivery is reduced appreciably for the nullipara who is in early labor with thick, meconium-stained amnionic fluid. Therefore, if the woman is remote from delivery, strong consideration should be given to prompt cesarean delivery, especially when cephalopelvic disproportion is suspected or either hypotonic or hypertonic dysfunctional labor is evident. Some practitioners choose to avoid oxytocin use in these cases.

Until recently, it was taught—including at Parkland Hospital —that aspiration of meconium could be minimized but not eliminated by suctioning the pharynx as soon as the head was delivered. According to the American Heart Association guidelines, this in no longer recommended (Wyckoff, 2015). The American College of Obstetricians and Gynecologists (2017a) does not recommend routine intrapartum suctioning. Alternatively, if the depressed newborn has meconium-stained fluid, then intubation is carried out.