CHAPTER 35: Stillbirth

In Williams’ time, absolute documentation of fetal death was frustrating for both patient and obstetrician. Now, sonography provides prompt confirmation, which allows expedient induction of labor and delivery. However, epidemiologically, defining and reporting fetal deaths was—and continues to be—a challenge. In response, efforts to standardize the definition of stillbirth and analyze varying reports for application into clinical practice and public health policy are now being emphasized. Moreover, stillbirth research and prevention within the United States and abroad has expanded. Global public health efforts were stimulated in part by a six-part series in The Lancet. This compilation was considered a call to action after the recognition that an estimated 2.65 million stillbirths occur each year and that 98 percent of stillborn fetuses are from low- and middle-income countries (The Lancet’s Stillbirth Series Steering Committee, 2011a–f). Unfortunately, progress in improving these rates has been slow, as outlined in The Lancet’s subsequent five-part progress report, which emphasized the need for dedicated leadership, measured effects of interventions, and investigation into knowledge gaps (The Lancet’s Ending Preventable Stillbirths Series Study Group, 2016a–e).

In the United States, an estimated 1 million fetal losses are reported each year, and most occur before 20 weeks’ gestation. Fetal mortality data from the National Vital Statistics system are usually presented for fetal deaths after the 20-week threshold (MacDorman, 2015). Using this definition, numbers of fetal deaths in the United States in 2013 slightly surpassed numbers of infant deaths (Fig. 35-1). As shown in Figure 35-2, fetal death rates are highest at the earliest and latest gestational ages, which suggests etiological differences.

The current definition of fetal death adopted by the Centers for Disease Control and Prevention National Center for Health Statistics is based on a definition recommended by the World Health Organization (MacDorman, 2015). It states that “Fetal death means death prior to complete expulsion or extraction from the mother of a product of human conception irrespective of the duration of pregnancy and which is not an induced termination of pregnancy. The death is indicated by the fact that after such expulsion or extraction, the fetus does not breathe or show any other evidence of life such as beating of the heart, pulsation of the umbilical cord, or definite movement of voluntary muscles. Heartbeats are to be distinguished from transient cardiac contractions; respirations are to be distinguished from fleeting respiratory efforts or gasps.”

Reporting requirements for fetal deaths in the United States are determined by each state, and thus, criteria differ significantly (Chap. 1, Definitions). Most states mandate reporting of deaths of fetuses that are 20 weeks’ gestation or older or have a minimum birthweight of 350 g (roughly equivalent to 20 weeks) or some combination of these two. However, several states require reporting of fetal deaths at all periods of gestation, and one sets the threshold at 16 weeks. Alternatively, two states require reporting of deaths for fetuses with birthweights of 500 g, which approximates that at 22 weeks. There is substantial evidence that not all fetal deaths for which reporting is required are actually recorded (MacDorman, 2015). This is most likely for those at earlier gestational ages.

Comparisons of rates among countries are limited by incomplete fetal death data. Namely, internationally, less than 5 percent of neonatal deaths have formalized documentation (The Lancet’s Ending Preventable Stillbirths Series Study Group, 2016d). Further, comparative analyses using birthweight versus gestational age among countries do not provide equivalent results. For example, in the United States, if stillbirth would be defined by a birthweight ≥500 g, the stillbirth rate would be reduced by 40 percent compared with a 22-week-age defined cohort (Blencowe, 2016). To address nomenclature differences, some have called for changes to the current definition (Joseph, 2015).

Overall, fetal mortality rates in the United States have remained relatively unchanged since 2006. However, the infant mortality rate has declined 11 percent, and both rates are now essentially equal (MacDorman, 2015). Three fetal mortality epochs are generally described: early (less than 20 completed weeks’ gestation); intermediate (20 to 27 weeks); and late (28 weeks or more). The fetal mortality rate at 20 to 27 weeks in 2013 declined 3 percent from the prior year. Between 2006 and 2012, rates were essentially unchanged in this age group. The late fetal mortality rate has been relatively unchanged since 2006 (Fig. 35-3).

The Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) created the Stillbirth Collaborative Research Network to ascertain stillbirth causes in racially and geographically diverse populations in the United States. From this, the Stillbirth Collaborative Research Network Writing Group (2011b) examined reasons for death at 20 weeks’ gestation or later between 2006 and 2008 in 59 tertiary care and community hospitals in five states. Standardized evaluations included autopsy, placental histology, and testing of maternal or fetal blood/tissues, including fetal karyotyping. Evaluations were performed in 500 women with 512 stillbirths. Of these losses, 83 percent were before labor. Causes of stillbirth were divided into eight categories shown in Table 35-1. These categories were then classified as probable, possible, or unknown. As an example, diabetes was considered a probable cause if the fetus had diabetic embryopathy with lethal anomalies or the mother had diabetic ketoacidosis. It was a possible cause if the mother had poor glycemic control and the fetus had abnormal growth. Overall, a probable or possible source was identified in 76 percent of cases.

This Network study is unprecedented in the United States for several reasons. It was a population-based cohort of stillbirths, in which all underwent systematic and thorough evaluation. Each assigned cause of fetal death is reasonably straightforward and comprehensible except for “placental abnormalities.” This category contains “uteroplacental insufficiency” and a few other less clearly defined placental entities. This aside, the leading reasons for fetal death were obstetrical and primarily included abruption, multifetal gestation complications, and spontaneous labor or ruptured membranes before viability. Importantly, this study illustrated that systematic evaluation may identify a likely cause in approximately three fourths of stillbirths. This rate is considerably higher than those in most prior analyses and serves to emphasize the importance of careful examination.

Many factors are associated with an increased risk of stillbirth. Among others, these include advanced maternal age; African-American race; smoking; illicit drug use; maternal medical diseases—such as overt diabetes or chronic hypertension; assisted reproductive technology; nulliparity; obesity; and prior adverse pregnancy outcomes—such as prior preterm birth or growth-restricted newborn (Reddy, 2010; Varner, 2014).

Two major studies have assessed whether stillbirth risk factors could be identified either before or shortly after pregnancy confirmation. In the first, Reddy and colleagues (2010) analyzed data from the NICHD Consortium on Safe Labor. Briefly, the pregnancy outcomes of 206,969 women delivered between 2002 and 2008 at 19 hospitals in the United States were analyzed. When the distribution of stillbirths according to gestational age was studied, the tragedy of stillbirth occurred primarily in term pregnancies. These investigators concluded that their results did not support routine antenatal surveillance for any demographic risk factors.

The second analysis of stillbirth risk factors was included in the Stillbirth Collaborative Research Network study described earlier. The validity of stillbirth prediction was assessed based on risks identified in early pregnancy. They found that pregnancy factors known at the start of pregnancy accounted for only a small proportion of stillbirth risk. Except for prior stillbirth or pregnancy loss from causes such as preterm birth or fetal-growth restriction, other risks had limited predictive value (Stillbirth Collaborative Research Network Writing Group, 2011a). The importance of prior stillbirth as a risk for recurrence has been emphasized by Sharma and associates (2006). Specifically, the stillbirth risk was fivefold higher in women with a prior stillbirth. From another report, prior preterm birth, fetal-growth restriction, preeclampsia, and placental abruption were strongly associated with subsequent stillbirth (Rasmussen, 2009). Table 35-2 lists estimates of stillbirth risk according to maternal factors.

Determining the cause of fetal death aids maternal coping, helps assuage any perceived guilt, permits more accurate counseling regarding recurrence risk, and may prompt therapy or intervention to prevent a similar outcome in subsequent pregnancies (American College of Obstetricians and Gynecologists, 2016a). Identification of inherited syndromes also provides useful information for other family members.

Clinical Examination

Important tests in stillbirth evaluation are neonatal autopsy, chromosomal analysis, and examination of the placenta, cord, and chorioamnionic membranes (Pinar, 2014). Page and coworkers (2017) found placental pathology and fetal autopsy to be the most useful. One algorithm from the American College of Obstetricians and Gynecologists (2016a) is shown in Figure 35-4. Findings are documented in the medical record, and relevant prenatal events are delineated. Photographs are taken whenever possible, and a full radiograph of the fetus—a fetogram—may be performed. Postnatal magnetic resonance (MR) imaging or sonography may be especially important in providing anatomical information if parents decline a full autopsy (McPherson, 2017; Shruthi, 2017).

Laboratory Evaluation

If autopsy and chromosomal studies are performed, up to 35 percent of stillborn fetuses are discovered to have major structural anomalies (Faye-Petersen, 1999). Approximately 20 percent have dysmorphic features or skeletal anomalies, and 8 percent have chromosomal abnormalities (Pauli, 1994; Saller, 1995). In the absence of anatomic dysmorphology, up to 5 percent of stillborn fetuses will have a chromosomal abnormality (Korteweg, 2008). Although the American College of Obstetricians and Gynecologists (2016a) previously recommended karyotyping all stillborn fetuses, technological advancements of high-resolution, whole-genome sequencing—such as with chromosomal microarray analysis (CMA)—are now replacing standard karyotyping for chromosomal analysis of stillborn fetuses (Chap. 13, Chromosomal Microarray Analysis). CMA does not require dividing cells and is reported to be more useful in the evaluation of fetal death. This is especially true because the culturing of macerated fetal tissue is frequently unsuccessful (Reddy, 2012). Both the American College of Obstetricians and Gynecologists (2016c) and the Society for Maternal-Fetal Medicine (2016) now endorse the use of CMA for stillborn fetuses.

Appropriate consent must be obtained to take fetal samples, including tissue or fluid obtained postmortem by needle aspiration. As recently outlined by the American College of Obstetricians and Gynecologists (2016c), any type of fetal or placental tissue or amnionic fluid can be submitted for genetic testing by CMA. Contamination with maternal tissue or blood is ideally avoided. If fetal blood cannot be obtained from the umbilical cord or by cardiac puncture, the American College of Obstetricians and Gynecologists (2016a) recommends at least one of the following samples: (1) a placental block measuring about 1 × 1 cm taken below the cord insertion site in the unfixed specimen; (2) umbilical cord segment approximately 1.5 cm long; or (3) internal fetal tissue specimen such as costochondral junction or patella. Tissue is washed with sterile saline before being placed in lactated Ringer solution or sterile cytogenetic medium. Notably, placement in formalin or alcohol kills viable cells. If conventional karyotyping is the only test available and the timing of death is recent, amnionic fluid can be obtained by amniocentesis, as those cells obtained in a sterile fashion provide a greater likelihood of cell growth and eventual result compared with tissue obtained after delivery. Maternal blood is obtained for Kleihauer-Betke staining; for antiphospholipid antibody and lupus anticoagulant testing if indicated; and for serum glucose measurement to exclude overt diabetes (Silver, 2013).

In cases with significant growth restriction, with a family or personal history of thrombosis, or with severe placental pathology, testing for factor V Leiden mutation, prothrombin mutation, antithrombin level, and protein C and S activity may provide some information that could affect future pregnancy management (American College of Obstetricians and Gynecologists, 2016a). Our interpretation of relevant placental pathology includes derangements that stem from maternal vessel obstruction, which are described in Chapter 6. Although some have recommended routine evaluation of heritable thrombophilias, no evidence supports the clinical or financial efficiency of screening in an unselected population. Silver and colleagues (2016), with data from the Stillbirth Collaborative Research Network, found that most maternal and fetal thrombophilias were not associated with stillbirth and recommended against routine testing.

Autopsy

Parents are offered and encouraged to allow a full autopsy. That said, valuable information can still be obtained from limited studies. Pinar and coworkers (2012) described the autopsy protocol used by the Stillbirth Collaborative Research Network. As an alternative, gross external examination combined with photography, radiography, MR imaging, bacterial cultures, and selective use of chromosomal and histopathological studies often aids determination.

A complete autopsy is more likely to yield valuable data. An analysis of 400 consecutive fetal deaths in Wales showed that autopsy altered the presumed cause of death in 13 percent and provided new information in another 26 percent (Cartlidge, 1995). Other investigators have found that autopsy results changed the recurrence risk estimates and parental counseling in 25 to 50 percent of cases (Faye-Petersen, 1999; Silver, 2007). For example, Miller and associates (2016) recently demonstrated that placental examination with autopsy altered future medical management in 45 percent of cases.

According to the survey by Goldenberg and coworkers (2013), most hospitals do not audit stillbirths. In other centers, however, maternal records and autopsy findings are reviewed on a monthly basis by a stillbirth committee composed of obstetricians, maternal-fetal medicine specialists, neonatologists, clinical geneticists, and perinatal pathologists. If possible, the cause of death is assigned based on available evidence. Most importantly, parents are then contacted and offered counseling regarding the reason for the death, the potential recurrence risk, and possible strategies to avoid recurrence.

Fetal death is psychologically traumatic for a woman and her family. Further stressors are an interval of more than 24 hours between the diagnosis of fetal death and labor induction, not seeing her infant for as long as she desires, having no tokens of remembrance, and poor communication (Radestad, 1996; Siassakos, 2017). The importance of seeing and holding a stillborn fetus for parental psychological well-being was recently summarized by Kingdon and colleagues (2015). As discussed in Chapter 61 (Postpartum Depression), a woman experiencing a stillbirth or early miscarriage is at increased risk for depression and should be closely monitored (Nelson, 2013).

Nuzum and coworkers (2014) reported that few obstetrical providers receive formal training in perinatal bereavement care. At Parkland Hospital, this care includes time with the infant, keepsake items, photographs, chaplaincy consultation, and bereavement support information. Care is coordinated through a dedicated nursing team affiliated with labor and delivery.

Table 35-3 lists an outlined approach for women with prior stillbirth. Importantly, these recommendations are based primarily on limited or inconsistent scientific evidence or on expert opinions. Unfortunately, few studies address management of affected women. Those with modifiable risk factors for stillbirth, such as hypertension or diabetes, warrant specific prevention strategies. Given that obesity has been identified as a risk factor for stillbirth and other obstetrical complications, preconceptional weight loss would seem prudent. Logically, women with a prior fetal death due to placental vascular events, that is, placental insufficiency, are also at increased risk for subsequent adverse perinatal outcomes (Monari, 2016). According to Reddy (2007), because almost half of fetal deaths are associated with growth restriction, fetal sonographic anatomical assessment beginning at midpregnancy is recommended. This is followed by serial growth studies beginning at 28 weeks. Supplementation with vitamin C or E in pregnancy has not been demonstrated to reduce the risk of fetal death (Rumbold, 2015a,b).

Weeks and associates (1995) evaluated fetal biophysical testing in 300 women whose only indication was prior stillbirth. There was one subsequent stillbirth, and only three fetuses had abnormal testing results before 32 weeks. Notably, no relationship was found between the gestational age of the previous stillborn fetus and the incidence or timing of abnormal test results or fetal jeopardy in the subsequent pregnancy. These investigators concluded that antepartum surveillance should begin at 32 weeks or later in the otherwise healthy woman with a history of stillbirth. This recommendation is supported by the American College of Obstetricians and Gynecologists (2016a) with the caveat that it increases the iatrogenic preterm delivery rate. Although fetal movement counting strategies are routinely employed as described in Chapter 17 (Clinical Application), few data guide its use in clinical practice for those with a prior stillbirth (Mangesi, 2015).

Delivery at 39 weeks’ gestation is recommended. Labor induction is suitable, and cesarean delivery is elected for those with a contraindication to induction. This timing minimizes fetal mortality rates, although the degree of risk reduction may be greater for older women (Page, 2013).

Following declines between 2000 and 2006, the United States fetal mortality rate has been relatively unchanged since 2006 (MacDorman, 2015). Interpretation of these fetal mortality rates in the context of changing national healthcare strategies has spawned considerable debate. One example is the effort to prevent non-medically indicated deliveries before 39 weeks and its subsequent effect on term stillbirth rates. The value of this practice for neonatal outcome is described in Chapter 26 (Techniques). To analyze whether implementation of this “39-week rule” has altered the term stillbirth rate, Nicholson and coworkers (2016) examined data from 45 states and the District of Columbia during a 7-year period. The proportion of births before 39 weeks progressively declined from 2007 and 2013, but the term stillbirth rate rose. This suggested that the 39-week rule may cause unintended harm. MacDorman and associates (2015) also evaluated trends in stillbirth rates by gestational ages in the United States between 2006 and 2012. They used a “traditional stillbirth rate,” which was calculated using a denominator composed of the live-birth number plus the stillbirth number at a given gestational age. They found increased rates at 24 to 27, 34 to 36, and 38 weeks’ gestation. Alternatively, no differences were found in “prospective stillbirth rates.” These rates were calculated using a denominator composed of the number of women who are pregnant at a given gestational age for weeks 21 through 42. The discrepancies in stillbirth rates appear to be primarily due to the decline in the number of births in the preterm and early-term gestational ages.

To summarize, implementation of the 39-week rule has reduced the number of elective births before 39 weeks’ gestation, although an unintended consequence may be an increase in term stillbirths—especially among women with medical complications. The importance of induction at less than 39 weeks in pregnant women with complications to prevent stillbirth is underscored by Little and colleagues (2015). These authors performed a retrospective multistate analysis of early-term deliveries (37^0/7 to 38^6/7 weeks) from 2005 to 2011. They noted a decline in the number of early-term deliveries during this time but not a significant change in the term stillbirth rates. There was, however, a 25-percent rise in the rate of term, singleton stillbirths among women with diabetes, and this was attributed to clinicians misapplying early-term delivery policies to high-risk women. Undoubtedly, continued surveillance of stillbirth rates is warranted for both high- and low-risk pregnancies at a state and national level.