9899Clinical Outcomes in Birth Centers

JULIA C. PHILLIPPI AND KATHLEEN DANHAUSEN

LEARNING OBJECTIVES

Upon completion of this chapter, the reader will be able to:

1.  Compare maternal and neonatal outcomes in birth center and hospital settings

2.  Explain difficulties with data aggregation across birth center studies

3.  Identify common indications for antepartum, intrapartum, postpartum, and neonatal transfers from birth center care

4.  Identify risk factors for complications while receiving birth center care

The effect of the location of birth on perinatal outcomes is an essential component of informed consent for women. In this chapter, we synthesize data from a variety of trials to provide an overview of clinical outcomes in birth centers. These studies are international in scope and include data from birth centers located in the United States, Europe, and Australia. Most of the birth centers studied in the United States are freestanding or not located within or adjacent to a hospital. However, the international birth center studies also include birth centers that function as independent units within a hospital. We rely primarily on four well-designed, large, and recent studies of birth center care with samples that allow clear inferences to today’s clinical practice. In the National Birth Center Study II, Stapleton, Osborne, and Illuzzi (2013) analyzed data from 22,403 U.S. women planning birth center birth, including 15,574 who were admitted to a freestanding birth center in labor. In the Birthplace Study, 100Birthplace in England Collaborative Group (2011) reported on 11,666 British women planning birth center births, of which 11,282 were admitted to freestanding birth centers in labor. Waldenström and Nilsson (1997) reported on 928 Swedish women receiving in-hospital birth center care in one of the only randomized controlled trials in the birth center literature. In 2004, Gottvall, Grunewald, and Waldenström reported clinical outcomes of 3,256 pregnancies (among 2,534 women) cared for by the same birth center over a 10-year span.

In addition to this contemporary research, we examine all relevant studies of birth center care published since 1980, limiting the search to studies of birth locations with admission criteria similar to American Association of Birth Centers (AABC standards; see Chapter 8). We acknowledge that there are limitations to this approach, especially as we attempt to describe and combine studies in the distant past and those in countries with different health care systems. However, the gestalt of the literature on care in birth centers allows for evaluation of these centers for birth and practice.

SYNTHESIS OF MATERNAL OUTCOMES

Maternal Outcome Data Overview

The data used to describe maternal outcomes in birth centers were gathered from 23 publications using 14 total data sets. Manuscripts were initially identified through a search of Google Scholar, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and PubMed databases using “birth center”/”birthing center”, and “outcomes.” Articles published in English-language, peer-reviewed journals since 1980 were selected. Forty-three articles were identified for full review and manuscripts were excluded from further analysis if the research had a singular focus (e.g., perineal integrity) and did not provide comprehensive data on maternal outcomes, if they did not define the birth center practice model, or if the practice model did not meet AABC criteria for birth center practice. For example, manuscripts were excluded if they referred to practices prohibited in AABC birth centers, including the use of misoprostol or Pitocin for labor induction or augmentation, vacuum or forceps-assisted births, or the use of continuous fetal monitoring. In addition, at least one study was excluded because its outcome data included women who would have been considered too high risk for most birth centers (e.g., women with multiple fetuses and women with a fetus in a breech presentation).

101These 14 data sets included a total of 84,300 women planning to give birth in a birth center, using a variety of study designs. Nine studies involved a comparison between outcomes in birth centers and outcomes among a matched sample of women giving birth in hospitals (Benatar, Garrett, Howell, & Palmer, 2013; David, von Schwarzenfeld, Dimer, & Kentenich, 1999; Feldman & Hurst, 1987; Jackson et al., 2003; Overgaard, Fenger-Grøn, & Sandall, 2012a, 2012b; Overgaard, Møller, Fenger-Grøn, Knudsen, & Sandall, 2011; Scupholme & Kamons, 1987; Scupholme, McLeod, & Robertson, 1986). In nine studies, researchers used an observational design to describe outcomes among women utilizing birth center services without a direct comparison group (Fullerton et al., 1997; Nguyen et al., 2009; Roberts & Sward, 2001; Rooks, Weatherby, & Ernst, 1992a, 1992b, 1992c; Rooks et al., 1989; Rowe et al., 2013; Stapleton et al., 2013). Two large studies of low-risk women reported outcomes across home, birth center, and hospital births (Brocklehurst et al., 2011; Wax, Pinette, Cartin, & Blackstone, 2010). Finally, the outcomes from one randomized controlled trial of women assigned to birth center versus hospital care were reported in three publications (Waldenström & Nilsson, 1993, 1994, 1997).

These data were collected over 30 years, and from five countries. Of the 14 data sets, 10 were collected in the United States (Benatar et al., 2013; Feldman & Hurst, 1987; Fullerton et al., 1997; Jackson et al., 2003; Nguyen et al., 2009; Roberts & Sward, 2001; Rooks et al., 1989, 1992a, 1992b, 1992c; Scupholme & Kamons, 1987; Scupholme et al., 1986; Stapleton et al., 2013; Wax et al., 2010); and one each from Denmark (Overgaard et al., 2012a, 2012b), England (Brocklehurst et al., 2011; Rowe et al., 2013), Germany (David et al., 1999), and Sweden (Waldenström & Nilsson, 1993, 1994, 1997). Four of the 14 data sets included women who were attended by a diversity of intrapartum providers, including physicians, certified nurse-midwives (CNMs), and other legally practicing midwives such as certified professional midwives (CPMs) and licensed midwives (LMs); the providers in these data sets were all located in the United States (Jackson et al., 2003; Nguyen et al., 2009; Rooks et al., 1989, 1992a, 1992b, 1992c; Wax et al., 2010). The women represented in the remaining data sets were served exclusively by midwives during their birth center care, with U.S. studies specifying that these providers were CNMs.

Serious Maternal Complications

Studies utilizing comparison groups indicate that women in birth centers have no increased statistical risk of serious maternal outcomes as 102compared with women in hospital care. Events that constitute “serious maternal outcomes” are not defined consistently across studies. However, they generally include maternal death and severe morbidity such as the need for admission to an intensive care unit (ICU), hysterectomy, or blood transfusion. Most studies reported no serious adverse maternal events in either birth center or hospital groups. Waldenström and Nilsson report that one woman from each group (hospital and birth center) was admitted to the ICU and both had full recoveries (Waldenström & Nilsson, 1997). David et al. (1999) noted one death in the hospital group but did not report the cause. Finally, in the Birthplace Study, Brocklehurst et al. (2011) noted a statistically significant increase in the number of women requiring blood transfusion and transfer to higher acuity care among those planning hospital births as compared with those who intended birth in freestanding centers.

Mode of Birth

Women planning to give birth in a birth center at the time they begin labor are more likely to have a spontaneous vaginal birth when compared with women planning birth in a hospital. Five studies used matched cohorts of low-risk women planning to labor in a birth center or a hospital (Brocklehurst et al., 2011; David et al., 1999; Jackson et al., 2003; Overgaard et al., 2011; Scupholme et al., 1986), whereas the other studies compared their birth center sample to national averages (Rooks et al., 1989, 1992a, 1992b, 1992c; Stapleton et al., 2013). In all of the studies cited, women who started labor planning to give birth in a birth center were included in the birth center sample, regardless of whether they subsequently transferred to the hospital. All studies found higher rates of spontaneous vaginal birth in the birth center groups.

Six studies found that women receiving birth center care were significantly less likely to experience assisted vaginal birth with forceps or vacuum (Benatar et al., 2013; Brocklehurst et al., 2011; David et al., 1999; Feldman & Hurst, 1987; Jackson et al., 2003; Overgaard et al., 2011), with an additional study reporting lower rates of assisted birth without statistical significance (Scupholme et al., 1986). (Use of forceps and vacuum devices is not permitted in Commission for the Accreditation of Birth Centers [CABC]-accredited birth centers; however, they may be used following hospital transfer.) Among the five studies with matched samples of low-risk women, the cesarean rate within the birth center cohort was lower but did not reach significance (David et al., 1999; Feldman & Hurst, 1987; Jackson et al., 2003; Scupholme et al., 1986; Waldenström & Nilsson, 1997); 103three additional analyses involving unmatched low-risk comparison groups found a significantly lower cesarean birth rate among women who received birth center care as compared with women who solely received hospital-based care (Benatar et al., 2013; Brocklehurst et al., 2011; Overgaard et al., 2011).

Oxytocin Use and Length of Labor

The three research teams that measured length of labor reported significantly longer labors among mothers beginning care in birth centers as compared with the hospital (Feldman & Hurst, 1987; Scupholme et al., 1986; Waldenström & Nilsson, 1997). Wax and colleagues (2010) analyzed more than 745,000 U.S. birth certificates and found that the 4,661 women who gave birth in a freestanding birth center were significantly more likely to have prolonged or precipitous labor. (The study authors did not provide a definition of prolonged labor.) Although birth center staff may have a greater appreciation for the range of normal duration of physiologic labor, this finding most likely also reflects the greater in-hospital use of oxytocin. The use of intrapartum oxytocin is prohibited in CABC-accredited birth centers, and therefore was not used in any of the birth centers in this analysis; prolonged labor and “failure to progress” is a primary indication for hospital transfer. However, birth center clients who transferred to the hospital had significantly lower rates of hospital oxytocin use during labor in all six studies reporting this outcome (Brocklehurst et al., 2011; Feldman & Hurst, 1987; Jackson et al., 2003; Overgaard et al., 2011; Scupholme et al., 1986; Waldenström & Nilsson, 1997).

Pain Management

Pain relief options vary across birth centers, countries, and time periods. Water immersion is a primary pain relief method used in birth centers; studies have also described women receiving narcotic analgesia, pudendal or paracervical blocks, nitrous oxide, and sterile water papules in birth center settings (Feldman & Hurst, 1987; Jackson et al., 2003; MacVicar et al., 1993; Roberts & Sward, 2001; Rooks et al., 1989, 1992b; Scupholme & Kamons, 1987; Scupholme et al., 1986; Waldenström & Nilsson, 1994, 1997). Moreover, additional methods are often available to women following hospital transfer, including epidural anesthesia. Women in hospitals are significantly more likely to receive an epidural (Brocklehurst et al., 2011; Feldman & Hurst, 1987; Jackson et al., 2003; Overgaard et al., 2011; Waldenström & Nilsson, 1994, 1997).

104When a comparison group was included in the study design, researchers found that birth center clients were less likely to use pharmacological methods of pain relief as compared with those giving birth in a hospital (Feldman & Hurst, 1987; Jackson et al., 2003; Scupholme et al., 1986; Waldenström & Nilsson, 1994, 1997). One exception is Waldenström (1994), who found that women in birth centers were more likely to use sterile water papules than a matched group in the hospital. However, the last reported pain relief outcome data were collected in the late 1990s. This may represent declining research interest around pharmacological pain relief, fewer pharmacological options offered in birth centers, or both. In the 1980s and 1990s, when data were collected, rates of narcotic analgesia use by birth center clientele ranged from 43% to 13.1%. Rooks et al. (1989) stratified by parity and noted that 24% of nulliparas used pharmacological relief compared with 6.2% of multiparas.

Perineal Integrity

Women in birth center care experience fewer episiotomies and spontaneous perineal lacerations than women in hospital care (Brocklehurst et al., 2011; David et al., 1999; Feldman & Hurst, 1987). Although episiotomy rates have declined over the last 20 years in all settings, birth centers have consistently reported lower rates of episiotomy when compared with hospitals, and three studies found statistically significant differences (Brocklehurst et al., 2011; David et al., 1999; Feldman & Hurst, 1987). Data collected in New York City (NYC) showed a 47.2% birth center episiotomy rate compared with 78.1% in a matched, low-risk hospital cohort (Feldman & Hurst, 1987). A German study found a 15.7% episiotomy rate in the birth center compared with a 54.8% rate in area hospitals (David et al., 1999). The data from the Birthplace Study showed an episiotomy rate of 8.6% in freestanding birth centers compared with 19.3% in a hospital obstetric unit not staffed by midwives (Brocklehurst et al., 2011).

Three studies reported that women cared for in birth centers were significantly more likely to maintain intact perinea during birth. In these studies (located in Denmark, Germany, and the United States), women birthing in hospitals had significantly higher rates of first- and second-degree lacerations (David et al., 1999; Feldman & Hurst, 1987; Overgaard et al., 2011). In a likely reflection of the trend away from episiotomies, the rate of intact perinea experienced in birth centers increased over time, with 25% reported in the early 1980s versus 61.3% in the mid-2000s (Feldman & Hurst, 1987; Overgaard et al., 2011). Two studies reported differences in third- and fourth-degree lacerations; although rates in birth 105centers were lower, the overall incidence was too low to assess statistical significance (Brocklehurst et al., 2011; Overgaard et al., 2011).

Transfer of Care

Out-of-hospital care is primarily appropriate for low-risk women; a change in risk status requires transfer to higher level care. Maternal transfer from birth center care can occur at any time during the antepartum, intrapartum, or postpartum period, with antepartum transfers occurring for both medical and nonmedical reasons. (Examples of nonmedical reasons include the woman moving or choosing a different provider.)

Data on birth center transfers are complicated by the fact that studies define and measure transfers differently. For example, some studies divided antepartum transfer rates into medical and nonmedical, whereas others did not differentiate, and one included a separate category for women experiencing first-trimester loss (Stapleton et al., 2013). Comparing intrapartum transfer rates across studies is even more difficult. Some authors determine this rate as the number of women transferring while in labor divided by all women entering prenatal care at a birth center (Feldman & Hurst, 1987; Jackson et al., 2003; Waldenström & Nilsson, 1993, 1994, 1997), whereas others calculate rates based on a denominator of women admitted to the birth center in labor (David et al., 1999; Fullerton et al., 1997; Overgaard et al., 2011, 2012a; Roberts & Sward, 2001; Rooks et al., 1989, 1992b, 1992c; Scupholme & Kamons, 1987; Stapleton et al., 2013). Still others include in the denominator those women admitted in labor as well as women in labor who are assessed by center providers and transferred to hospital care prior to admission (Brocklehurst et al., 2011; Nguyen et al., 2009; Rowe et al., 2013; Scupholme et al., 1986).

Eighteen studies reported transfer rates. From entry to prenatal care through the postpartum period, up to 54.7% of women intending to give birth in a birth center experienced a transfer of care (Jackson et al., 2003). Across studies, the majority of transfers were for nonemergency conditions.

Antepartum Transfers

Antepartum transfer rates ranged up to 27.2% in a sample of approximately 1,800 U.S. women seeking birth center care in the mid-1990s (Jackson et al., 2003). Antepartum transfer indications include medical complications precluding birth center care (Table 4.1). In addition, preterm births and pregnancies that extend past 42 gestational weeks are not considered appropriate for out-of-hospital care in more recent studies. Some 106studies include nonmedical transfers such as maternal geographic relocation in the antepartum transfer category and have higher rates than studies that separate out nonmedical reasons for ending birth center care. Stapleton and colleagues (2013) reported a 13% antepartum transfer rate for medical reasons. Multiparous women are more likely than nulliparous women to be transferred antepartum (Waldenström & Nilsson, 1994).

Intrapartum Transfers

The most recent studies report intrapartum transfer rates ranging from 11.6% in a sample of 839 Danish women to 16.5% in the Birthplace Study (Brocklehurst et al., 2011; Overgaard et al., 2011). Similarly, the National Birth Center Study II (Stapleton et al., 2013) reported that 12.4% of the approximately 15,000 women admitted in labor were transferred intrapartum. The National Birth Center Study II also reported that 4.5% of women initially evaluated in labor were transferred prior to admission to the birth center (Stapleton et al., 2013). Studies providing information on intrapartum transfers noted that the most common indications included 107lack of progress in labor, rupture of membranes without labor, and prolonged labor (Table 4.2).

Three studies reported rates for emergent versus nonemergent maternal intrapartum transfers and noted that the majority of transfers were nonemergencies (Rooks et al., 1992c; Rowe et al., 2013; Stapleton et al., 2013). Two of these studies were published in the last 5 years and included large samples of more than 10,000 women (Rowe et al., 2013; Stapleton et al., 2013). Of the 12.4% of women experiencing intrapartum transfer in the National Birth Center Study II, less than 2% were transferred emergently (Stapleton et al., 2013). Rates of emergent transfer differ by parity. In the secondary analysis by Rowe et al. (2013) of the birthplace data, 27% of all nulliparas planning a birth center birth were transferred to the hospital during labor, with 9.5% of all nulliparas transferred for a “potentially urgent” indication. In comparison, 5% of multiparous women were transferred to the hospital during labor, whereas only 1.5% of multiparous women were urgently transferred. Of note, nulliparous and multiparous women had similar rates of “potentially urgent” hospital transfer immediately following a birth. Across studies, the leading reason for emergent transfer was nonreassuring fetal heart rate (Fullerton et al., 1997; Nguyen et al., 2009; Stapleton et al., 2013).

In studies reporting intrapartum transfer rates, approximately one third of nulliparous women were transferred during labor, with rates ranging from 27.3% to 29.6% (Brocklehurst et al., 2011; Rowe et al., 2013). 108Nulliparous women were transferred at approximately five times the rate of multiparous women; they accounted for 81.6% of the laboring women transferred in the National Birth Center Study II and 77% of women transferred from freestanding birth centers in the Birthplace Study (Brocklehurst et al., 2011; Overgaard et al., 2011; Rowe et al., 2013; Stapleton et al., 2013; Waldenström & Nilsson, 1997).

Postpartum Transfers

Eleven studies reported postpartum transfer statistics, which ranged from 0.5% to 4.5% (Brocklehurst et al., 2011; David et al., 1999; Feldman & Hurst, 1987; Overgaard et al., 2011; Roberts & Sward, 2001; Rooks et al., 1992c; Rowe et al., 2013; Scupholme et al., 1986; Stapleton et al., 2013; Waldenström & Nilsson, 1994, 1997). In the postpartum period, women were most commonly transferred for hemorrhage or retained placenta (Brocklehurst et al., 2011; Rooks et al., 1989; Stapleton et al., 2013; see Table 4.3). As with other transfer rates, various denominators were used including all women planning birth center birth, all those admitted in labor, or all those who gave birth in the center. The National Birth Center Study II (Stapleton et al., 2013) provides further insight into these transfers, reporting that 2.4% of women who gave birth in a birth center required postpartum transfer, and 0.5% of these women were transferred emergently.

Vaginal Birth After Cesarean

Few of the studies described provide information or analysis related to the clinical outcomes of women attempting a vaginal birth after cesarean (VBAC) in a birth center setting, primarily because many birth centers include a uterine scar as an exclusion criterion from birth center care. However, several studies have assessed attempted VBAC outcomes in birth centers and have found high rates of vaginal birth along with slightly increased risk to the mother and neonate, especially among women with more than one cesarean or neonates past the 42nd week of gestation. These studies were excluded from the larger analysis due to their narrow focus. Although women giving birth in birth centers have rates of uterine rupture that are similar to or less than those experienced in a hospital setting, the time between recognition of distress and expedited birth may increase the perinatal mortality rate. As with many measures of adverse perinatal outcomes, the incidence of uterine rupture and subsequent maternal or neonatal death is small across few studies, and thus conclusive statements cannot be made about the safety of VBAC in the birth center setting.

109Two relatively recent German studies used a national data set with information required from all midwives practicing out of hospital (Beckmann, Barger, Dorin, Metzing, & Hellmers, 2014; David, Gross, Wiemer, Pachaly, & Vetter, 2009). These authors compared two groups of women experiencing a second pregnancy. The experimental group had a history of cesarean, whereas the control group had a vaginal birth with their first pregnancy. Both studies found a VBAC rate of approximately 75% compared with a vaginal birth rate of more than 95% among women with no history of cesarean. Both authors reported a high transfer rate (~40%) for women attempting VBAC out of hospital, compared with a 5% transfer rate for women in the control group without a previous cesarean. Women with prior cesareans were more likely than women in the control group to be transferred for arrest of labor or suspected cephalopelvic disproportion (Beckmann et al., 2014; David et al., 2009).

David et al. noted that women with prior cesareans were significantly more likely to give birth to neonates with Apgar scores less than 7 at 5 minutes; however, this may be clinically insignificant as there were no differences in rates of neonatal intensive care unit (NICU) admission or infant mortality (David et al., 2009). There were no uterine ruptures or postpartum hysterectomies/laparotomies due to birth injury. Although both David et al. (2009) and Beckman et al. (2014) observed a higher rate of postpartum transfer among women with a prior cesarean, these women did not have higher rates of retained placenta or hemorrhage, which are the most common reasons for transfer in non-VBAC–related studies.

Two studies provide insight into VBAC in U.S. birth centers. Harrington, Miller, McClain, and Paul (1997) provide data from a hospital-based birth center in Los Angeles, California. They report a vaginal birth rate of 98% in 298 women with a prior cesarean compared with a vaginal birth rate of 99% among a control group of women without prior cesarean 110matched for age, parity, and gestational age. It is important to note that in this study 84% of the women with prior cesareans also had a history of prior vaginal birth. Women attempting VBAC with no prior vaginal birth were significantly more likely to transfer into hospital care compared to women attempting VBAC having had a prior vaginal birth (25% versus 6%). Among women who were transferred, women with a prior cesarean were more likely to give birth by cesarean (19% versus 6%); however, the difference was not statistically significant due to the low overall number of cesareans. No uterine ruptures were reported (Harrington et al., 1997).

The most comprehensive study of outcomes among women attempting VBAC in U.S. birth centers was published by Lieberman, Ernst, Rooks, Stapleton, and Flamm (2004). Forty-one birth centers offering VBAC services participated in a 10-year data collection effort and provided data for 1,453 women with a prior cesarean who presented to a birth center in labor. Of these women, 7% had more than one prior cesarean and 46% had a prior vaginal birth. The VBAC rate was 87%, with a significantly higher rate of vaginal birth observed in the subgroup of women with a prior vaginal birth (94% versus 81%).

Approximately one quarter of women were transferred into hospital care. Of these women, 7% were transferred immediately upon arrival to the birth center and 11% were otherwise transferred emergently. Women with a prior vaginal birth were significantly less likely to be transferred as compared with women with no prior vaginal birth (11% versus 35%), and the most common reason for intrapartum transfer was lack of progress in labor. Approximately 4% of women were transferred postpartum, with similar proportions transferred for maternal and neonatal indications. The most common maternal indication was laceration repair, and infants were most commonly transferred for respiratory support.

Six women experienced uterine rupture, with five transferring intrapartum for fetal indications and giving birth by cesarean, and one transferring postpartum for hemorrhage. Women with more than one prior cesarean were significantly more likely to experience uterine rupture (3% versus 0.2%). Prior vaginal birth was not protective against uterine rupture, nor was rupture associated with length of labor. Data were collected on the time elapsed from recognition of a complication to arrival at the hospital, and on the time until surgical incision. Four of the five women with intrapartum uterine ruptures arrived at the hospital within 15 minutes. However, only two women had a recognition-to-incision time of 30 minutes or less.

There were seven perinatal deaths, two related to uterine rupture. Women in their 42nd week of pregnancy were more likely to experience 111infant death as compared with women with a fetal estimated gestational age of less than 42 weeks (4.3% versus 0.4%); however, gestational age and birth weight were not associated with risk of rupture. Overall, 1.7% of women experienced a serious outcome (perinatal death (n = 7), Apgar less than 7 at 5 minutes (n = 15), uterine rupture (n = 6), or hysterectomy (n = 1); women in their 42nd week of pregnancy were more likely to experience a serious outcome (6.5% versus 1.6%).

Women who were transferred immediately upon arrival were included in the birth center statistics because presenting to the birth center ultimately delayed initiation of hospital care. However, when these women were excluded from analysis, serious outcomes occurred in 0.9% of births. Furthermore, the authors note that if women who were greater than 42 weeks in pregnancy and women with more than one prior cesarean were excluded from analysis, the rates of both uterine rupture and perinatal death were two per 1,000 births for women in birth center care. Unfortunately, the 10% of women who had either a history of more than one cesarean or a fetus past 42 gestational weeks experienced 50% of the uterine ruptures and 57% of the perinatal deaths observed among the women planning birth center care.

For comparison, 20% of U.S. women with one prior cesarean attempted a trial of labor in 2013, and 70% of these women gave birth vaginally, for an overall VBAC rate of 14% of all women with one prior cesarean (Curtin, Gregory, Korst, & Uddin, 2015). Women attempting VBAC in birth centers have higher rates of vaginal birth; however, they also have high rates of transfer to hospital care, compared with both multiparous and nulliparous women laboring in birth centers. Although absolute rates of uterine rupture are similar among women attempting VBAC in birth centers and in hospitals, there is a delay in accessing surgical resources when uterine rupture occurs in a birth center setting. However, this same delay may also occur in hospitals where surgical or anesthesia staff are not continually in-house.

SYNTHESIS OF NEONATAL OUTCOMES

Neonatal Outcome Data Overview

Although many of the studies described previously in this chapter included information on perinatal and infant outcomes, some focused exclusively on maternal outcomes and were omitted in this section, whereas additional studies on neonatal outcomes were also obtained. 112The resulting 24 manuscripts included 19 data sets on neonatal outcomes in birth centers with guidelines consistent with AABC standards. Publication dates ranged from 1986 to 2014. Twelve manuscripts (using 11 data sets) reported on births in the United States (Feldman & Hurst, 1987; Fullerton et al., 1997; Jackson et al., 2003); four studies were set in the United Kingdom, with three analyzing one data set (the Birthplace Study; Brocklehurst et al., 2011; Hollowell et al., 2011; Hundley et al., 1995; Rowe et al., 2013); three studied Australian data (Laws, Tracy, & Sullivan, 2010; Stern et al., 1992; Tracy et al., 2007); three studies analyzed the same Swedish data set (Gottvall et al., 2004; Gottvall, Winbladh, Cnattingius, & Waldenström, 2005; Waldenström & Nilsson, 1993, 1997); and individual studies from Denmark (Overgaard et al., 2011) and Germany (David et al., 1999). Many international studies involved birth centers located within a hospital but adhering to the birth center philosophy of limited technology and intervention, whereas most of the U.S. birth centers studied were freestanding.

Authors used a breadth of study designs, so direct comparisons of data across studies remain difficult. Of the studies included in this review, 12 provided neonatal comparison data from matched or unmatched low-risk women giving birth in hospitals (Brocklehurst et al., 2011; David et al., 1999; Feldman & Hurst, 1987; Gottvall et al., 2004, 2005; Hollowell et al., 2011; Hundley et al., 1995; Jackson et al., 2003; Overgaard et al., 2011, 2012b; Scupholme & Kamons, 1987; Scupholme et al., 1986); six were observational and provided a description of birth center outcomes with no selected comparison group (Fullerton et al., 1997; Roberts & Sward, 2001; Rooks et al., 1992b, 1992c; Stapleton et al., 2013; Stern et al., 1992); one used a randomized controlled design (Waldenström & Nilsson, 1997); and five used birth certificate or other national perinatal tracking data systems (Grünebaum et al., 2013, 2014; Laws et al., 2010; Tracy et al., 2007; Wax et al., 2010). Most of the studies using birth certificate data provided outcomes by location of birth, regardless of where labor was planned or started (e.g., hospital outcome data would include women who transferred from a birth center). All other studies used an intent-to-treat design, and outcomes of newborns of women in the birth center group who transferred to hospital care are included in the birth center statistics. Although most studies limited the birth center sample to women who were admitted in labor, several studies included all women who received antepartum care at the birth center. We did exclude Grünebaum and colleagues’ (2013) analysis of Apgar score data as there was a mathematical error in the calculation of the rate of low 5-minute Apgar scores for births with 113freestanding birth center midwives, and the rate was erroneously elevated by a factor of 100.

Apgar Scores

Thirteen studies reported infant Apgar scores (David et al., 1999; Feldman & Hurst, 1987; Hollowell et al., 2011; Jackson et al., 2003; Laws et al., 2010; Overgaard et al., 2011; Roberts & Sward, 2001; Rooks et al., 1989, 1992c; Scupholme et al., 1986; Stern et al., 1992; Waldenström & Nilsson, 1997; Wax et al., 2010). Although Apgar scores have limited validity as an indicator of long-term outcomes of infants, it is a widely used measure of infant transition to extrauterine life (American Academy of Pediatrics Committee on the Apgar Score, 2015). As these scores are a readily available measure of infant status, researchers have included them within many studies. In particular, it is common to study the percentage of infants who had an Apgar score less than 7 at 5 minutes of life. The incidence of this outcome ranged from 0.4% to 3% among infants born in birth centers, with three quarters of studies reporting the occurrence of low Apgar scores in less than 1% of infants. No study reported a significant difference between infants born in a hospital versus those born to mothers admitted to a birth center in labor. Laws et al. (2010) performed subgroup analysis (population) by parity and found that 1% of primiparas gave birth to infants with low 5-minute Apgar scores, as compared with 0.5% of multiparous women.

Neonatal Transfer

Six studies reported newborn transfer rates ranging from 2% to 5% (Fullerton et al., 1997; Roberts & Sward, 2001; Rooks et al., 1992c; Scupholme & Kamons, 1987; Scupholme et al., 1986; Stapleton et al., 2013). The most common reason for newborn transfer across studies was respiratory distress, which accounted for 60% to 100% of newborn transfers. Additional newborn transfer indications are reported in Table 4.4.

Two studies commented on the urgency of newborn transfer. Roberts and Sward (2001) reported that among the 4% of infants who required transfer after birth (all due to respiratory issues), 75% were transferred by ambulance, which was considered a proxy measure for urgency. The National Birth Center Study II (Stapleton et al., 2013) reported that of the 13,030 infants who were born in a birth center, 2.6% required hospital transfer, and 0.7% (n = 94) of those transfers were emergent. Rowe et al. (2013) 114analyzed the circumstances surrounding transfer in the U.K. Birthplace Study, and noted that all low 5-minute Apgar scores (n = 6) were among infants of women transferred for “potentially urgent” indications.

Stay updated, free articles. Join our Telegram channel

Join