U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Division of Behavioral and Social Sciences and Education; Board on Children, Youth, and Families; Committee on Assessing Health Outcomes by Birth Settings; Backes EP, Scrimshaw SC, editors. Birth Settings in America: Outcomes, Quality, Access, and Choice. Washington (DC): National Academies Press (US); 2020 Feb 6.

Cover of Birth Settings in America

Birth Settings in America: Outcomes, Quality, Access, and Choice.

Show details

6Maternal and Newborn Outcomes by Birth Setting

The previous chapter describes the challenges involved in studying the effects of birth settings on maternal and neonatal outcomes in the United States. In addition to the deficiencies in data sources and methodological limitations discussed in that chapter, the literature on birth settings compares a wide array of beneficial and nonbeneficial outcomes across and within settings. Often these studies do not report the same outcomes or use the same definitions or terminology, making it difficult to develop assessments or to draw useful conclusions across the existing body of literature (Khan, 2019). In addition, the overall small number of U.S. women giving birth in home and birth center settings (under 2%) compared with hospital settings (about 98%) complicates many studies of outcomes by setting (MacDorman and Declercq, 2019). The reason is that infrequent events such as maternal and infant death, while of great interest to the committee, tend to have unstable estimates with wide confidence intervals as a result of relatively small sample sizes for home and birth center subgroups.

Aspects of care during the childbearing year are also variable and dependent on such factors as the health of the mother and infant, models of prenatal and intrapartum care, the type of birth attendant, practice standards, and facility policies across and within birth settings and regions. Furthermore, as discussed in Chapter 2, the definition of what constitutes a birth center varies across the literature, both nationally and internationally, and most U.S. data sources cannot reliably track movement across birth settings or accurately attribute outcomes to the intended provider or place of birth (i.e., an intention-to-treat model). Where possible, the committee differentiates between outcomes for home births and outcomes for birth center births; however, some studies combine these births as “planned, out-of-hospital births” for analysis, which makes it impossible to compare the two (see, e.g., Snowden et al., 2015; Bovbjerg et al., 2017).

With these caveats in mind, this chapter provides a framework for understanding outcomes by birth setting and reviews the available research on maternal and newborn health outcomes for low-risk women for all three U.S. birth settings—home, birth center, and hospital—as well as data on outcomes by provider type where available. It then reviews studies of outcomes by birth setting internationally. Finally, the chapter concludes with a discussion of how interprofessional collaboration influences outcomes across and within birth settings. Where possible, we highlight the broad spectrum of outcomes that are of interest for this report and point to gaps in data and understanding that need to be filled by further research.

Importantly, we note that the literature reviewed focuses on outcomes by birth setting for low-risk women. As discussed in Chapter 3, maternal or fetal condition may have a significant influence on the choice of birth setting, as women who have medical or obstetric risk factors or comorbidities or are pregnant with fetuses at risk for complications are likely to give birth in a hospital. Conversely, healthy low-risk women living in regions with access to home and birth center birth will be overrepresented in these settings. In the same way, hospital level will influence the sample studied; for example, very high-risk women and fetuses are overrepresented in tertiary care facilities. These differences make direct comparisons across settings difficult. In the absence of adequate data to control for risk level and demographic differences, accurate comparisons are impossible.

UNDERSTANDING MATERNAL AND INFANT OUTCOMES

Miller and colleagues (2016) describe a continuum of global maternity care wherein two patterns result in excess of morbidity and mortality. The authors refer to these extremes as “too little, too late” (TLTL) and “too much, too soon” (TMTS). TLTL is used to describe care in which inadequate staffing, training, infrastructure, supplies, and medications (Austin et al., 2014, p. 2176) result in care that is withheld, below an evidence-based standard, or simply unavailable until it is too late. Severe morbidity and mortality result from this pattern of care. The converse system, TMTS, is characterized by routine overuse of interventions and the medicalization of healthy, uncomplicated pregnancies and births. Miller and colleagues (2016) argue that TMTS care often includes the unnecessary use of nonevidence-based interventions (e.g., continuous electronic fetal monitoring), as well as the overuse of interventions that can be lifesaving but are potentially harmful when applied routinely, without medical indication (e.g., cesarean section). In these systems, overintervention drives morbidity. As facility births have increased globally, so has the recognition that TMTS systems can produce harm, increase costs, and concentrate disrespect and abuse in childbirth (Freedman and Kruk, 2014; World Health Organization, 2014; International Confederation of Midwives, White Ribbon Alliance, International Pediatric Association, and World Health Organization, 2015; Miller and Lalonde, 2015). While TMTS systems are typically associated with high-resource nations and TLTL with low- and middle-resource ones, because of inequality, these extremes often coexist within a single nation as a result of inequality.

When preventable maternal (or fetal) death and severe morbidity occur in U.S. hospitals, these outcomes may result either from TMTS (as in the case of morbidity associated with higher-than-ideal cesarean rates) or from TLTL (as with unrecognized hemorrhage). When preventable death and suffering occur at home or in birth centers, these outcomes are likely the result of TLTL. In its review of the literature, the committee identified a number of maternal and infant outcomes of interest related to these TMTS and TLTL concepts. These outcomes include maternal and infant mortality and morbidity indicators, which have been the traditional focus of birth settings research, as well as psychosocial outcomes, including several measures of dignity in the childbirth process, such as bodily autonomy, maternal agency, respectful care, and empowerment. The committee chose to review as many of these outcomes as possible because, taken together, they provide a broader understanding of what “safety” and a “healthy mother, healthy baby, and healthy family” mean to childbearing families and to members of the care team. In short, this broader perspective recognizes that the experience of care cannot be separated from clinical outcomes; it is not a complement or secondary consideration, but an important aspect of ensuring high-quality childbirth care (World Health Organization, 2018). Moreover, the experience of care and the outcome of care are closely associated. Although the committee recognizes the importance of patient experience in childbirth care, the literature on women’s experiences in the maternity care system is limited. That literature is discussed in detail below.

Thus, in the discussion that follows, we examine the relationship between what we are calling intervention-related morbidity and birth setting, including outcomes and interventions reported in the literature, such as infection, induction, augmentation, postpartum hemorrhage, and genital tract tearing. We recognize that some of these interventions are necessary, unavoidable regardless of setting, or the result of maternal request. Nonetheless, because the desire to avoid unnecessary interventions is a primary reason families choose home and birth center births and because lower rates of these morbidities are also desirable from cost savings and quality-of-care perspectives, we review them in detail below. These morbidity outcomes are widely reported in the field, and thus are the focus of our review. However, we wish to emphasize that a broader conception of maternal morbidity is warranted. A more comprehensive view of morbidity would acknowledge the experiences of women and encompass such disorders as postpartum depression and anxiety, disrespect, unconsented care, coercion, and other forms of mistreatment that are starting to be documented in the United States (Vedam et al., 2019). These experiences impact not only the health of the person following childbirth but also that of the infant and family.

Additionally, the literature on health outcomes by birth setting would benefit from disaggregation of outcomes by race/ethnicity, socioeconomic status, and sexual orientation and gender identity, where possible. While the current literature on outcomes largely does not address differences by race and ethnicity or other subpopulations, as noted in Chapters 3 and 4, traditionally marginalized groups often accrue a disproportionate share of clinical and social risk factors for adverse outcomes during pregnancy and birth. In the sections below, we report variations in outcomes by subpopulations where available, but underscore the general paucity of evidence in this area and the need for future research to grapple with potential variation in outcomes by subpopulations, particularly for historically marginalized groups. Given the difficulty of studying outcomes by birth setting for the reasons outlined above, as well as the tendency toward confirmatory bias noted by Roome and colleagues (2016),1 the committee, whose membership is diverse professionally, grappled with multiple tensions. Ultimately, we believe this was a strength because it led us to better understand multiple perspectives and viewpoints in a way that reflects the wide range of views and preferences held by U.S. women regarding place of birth. In general, we concluded that each setting—home, birth center, and hospital—offers a set of risks and benefits that accrue to either the pregnant woman or the newborn. And while no setting can fully remove risk from birth, evidence suggests that many risks are modifiable at the level of systems, processes, providers, and policies.

FETAL AND NEONATAL OUTCOMES BY U.S. BIRTH SETTING

Neonatal outcomes include neonatal mortality and neonatal morbidity. Death before delivery of the fetus is termed an intrapartum death. Death after birth is termed early neonatal mortality (up to 7 days of life2). Death up to 28 days is termed neonatal mortality; death 29 days to 1 year of age, postneonatal mortality; and death up to 1 year of age, infant mortality. The majority of infant deaths occur during the first 7 days of life. This terminology is important as different studies refer to various timepoints and apply the term perinatal mortality inconsistently. Measures of neonatal morbidity include seizures, neonatal intensive care unit (NICU) admissions, hypoxic-ischemic encephalopathy, and low Apgar scores. The rates of neonatal mortality and morbidity across settings vary depending on the population studied and the parameters set by the researchers (e.g., whether researchers excluded congenital anomalies, or restricted the population to low-risk mothers).

As discussed in Chapter 3, unforeseen emergencies related to either the birth process or an unrecognized condition of the newborn may require immediate skilled intervention, including cesarean delivery or neonatal resuscitation. In these cases, the ability to access higher-level care without delay is critical for the safety of the fetus or newborn. Risk for the newborn in a home or birth center birth setting and in hospitals without these capabilities may be mitigated through various strategies, including selection of low-risk mothers; referral to an obstetric or maternal–fetal medicine provider for pregnancy complications; minimization of transfer times in case of a need to transfer; barrier-free transfer to a hospital for birth complications; collaborative professional models of care; formal training of skilled practitioners; and professional regulation, oversight, and accountability (see Chapter 7).

Several studies of health outcomes for newborns in home and birth center settings have been conducted in the United States. Hospital births attended by midwives are generally used as the baseline against which these birth settings are compared, and some studies report comparisons by provider type across and within setting as well. U.S. studies use registry and birth certificate data. The findings from these studies are shown in Tables 6-1 and 6-2.

TABLE 6-1. Rate and Percentage of Neonatal Mortality by U.S. Birth Setting.

TABLE 6-1

Rate and Percentage of Neonatal Mortality by U.S. Birth Setting.

TABLE 6-2. Rate and Percentage of Neonatal Morbidity by U.S Birth Setting.

TABLE 6-2

Rate and Percentage of Neonatal Morbidity by U.S Birth Setting.

Systematic Reviews

Two important systematic reviews have examined neonatal outcomes for home and birth center settings as compared with hospital settings: Phillippi and colleagues (2018) and Wax and colleages (2010). We discuss each study in detail below.

Phillippi and colleagues (2018) conducted a systematic review of 17 studies on neonatal outcomes; all of the studies evaluated neonatal mortality, and some also evaluated neonatal morbidity. Collectively, the studies included outcomes for the neonates of a total of 84,500 women admitted to a birth center in labor, including outcomes after transfer. The review found that in no study with a hospital comparison group was there a higher rate of neonatal mortality in the birth center group, and that nulliparous women and women older than 35 had a higher risk of poor neonatal outcomes in both birth centers and hospitals. In any studies that included births with a gestation more than 42 weeks, a higher risk of neonatal mortality was found for those pregnancies (Phillippi et al., 2018). All of these studies demonstrated selection of a favorable medical, obstetric, and social risk profile among women choosing home and birth center settings.

Wax and colleagues (2010) published a systematic review of birth outcomes for planned home and hospital births, which included all English-language, peer-reviewed publications from high-resource countries available at the time that reported fetal and neonatal outcomes by birth setting. Neonatal outcomes for planned home births included lower rate of low birthweight (less than 10% for gestational age or less than 2,500 grams) (1.3% vs. 2.2%; odds ratio [OR] 0.60; confidence interval [CI] 0.50–0.71), compared with the hospital sample. The neonatal mortality rate was nearly twice as high in planned home births as compared with planned hospital births (0.20% vs. 0.09%; OR 1.98; CI 1.19–3.28), and nearly three times as high when only nonanomalous infants were included in the analysis (0.15% vs. 0.04%; OR 2.87; CI 1.32–6.25). It is important to note that when sensitivity analyses were performed, which removed older studies and excluded those that had used matching, there was no significant difference in prematurity and neonatal death between the two birth settings (Wax et al., 2010).

Birth Registry Studies

A number of studies use birth registries, allowing for analysis on an intention-to-treat basis. As discussed in Chapter 4, however, they are limited in that reporting to registries is not mandatory. This means that findings come from samples rather than complete populations, and thus may not be generalizable. Moreover, because these studies are descriptive in nature, none has an explicit comparison group. Four descriptive studies using registry data—by Cox and colleagues (2015; 1.24/1,000 for women without a history of cesarean [ResQu: high, GRADE: poor]),3 Cheyney and colleagues (2014a; 0.85/1,000 [ResQu: high, GRADE: poor]),4 Johnson and Daviss (2005; 1–2/1,000 [ResQu: moderate, GRADE: poor]), and Stapleton and colleagues (2013; 0.47/1,000 [ResQu: high, GRADE: poor])—have documented low rates of perinatal mortality at home and in birth centers for healthy, low-risk women in the United States using birth registry data. Thornton and colleagues (2017 [ResQu: high, GRADE: poor]), similarly using registry data, compared midwife-led birth center and hospital groups and found no difference in the neonatal outcome composite5 (0.44% for both groups).6

Vital Statistics Studies

A number of other frequently cited studies report fetal and neonatal outcomes by birth setting using vital statistics. In 2012, the state of Oregon added new variables to the birth certificate (intended place of birth and the type of intended provider at the onset of labor), allowing Snowden and colleagues (2015) to use an intention-to-treat approach to examine outcomes by birth setting. The authors limited analyses to data from Oregon collected over the 2-year period following the change to the birth certificate, yielding a sample of 75,923 hospital births and 3,203 home and birth center births. Looking at singleton, term, nonanomalous births, this study found poorer outcomes at home and in birth centers—specifically, a perinatal mortality rate of 3.9 per 1,000 in home birth and birth center births compared with 1.8 per 1,000 in hospital births, and a neonatal mortality rate of 1.6 per 1,000 in home birth and birth center births compared with 0.6 per 1,000 in hospital births (Snowden et al., 2015).

Malloy (2010) conducted a retrospective cohort study measuring infant outcomes by setting and birth attendant using linked birth and death files from 2000 to 2004. The samples included hospital births attended by a certified nurse midwife (CNM), hospital births attended by another type of midwife, birth center birth attended by a CNM, home births attended by a CNM, and home births attended by another type of midwife. The neonatal mortality rate (measured as 0–27 days after birth) was 0.5/1,000 live births for hospital-CNM, 0.4/1,000 for hospital-other midwife, 0.6/1,000 for birth center-CNM, 1.0 for home-CNM, and 1.8/1,000 for home-other midwife. Births at home with a CNM or other midwife had a higher risk of neonatal death compared with hospital births with a CNM (2.02/1,000, CI 1.18–3.45; 3.63/1,000, CI 2.89–4.67, respectively). Home births attended by any midwife had a greater risk of a low 5-minute Apgar score <4 (home-CNM 7.83/1,000, CI: 6.09–10.1; home-other midwife 3.39/1,000, CI 2.70–4.24) compared with CNM-attended hospital deliveries.

Other studies also used vital statistics data (see, e.g., Cheng, 20137 [ResQu: moderate]; Grünebaum et al., 2013,82014,9 2015b10 [ResQu: low]; Bachilova et al., 201811 [ResQu: low, GRADE: poor]; and Wasden et al., 201712 [ResQu: low]). Collectively, these authors found worse neonatal outcomes for completed home births, including higher rates of neonatal mortality (Grünebaum and colleagues only), low 5-minute Apgar scores, and neonatal seizures. However, these studies were unable to track outcomes using an intention-to-treat model, and the impact of misclassification on the reliability of findings from studies based on vital statistics has not been conclusively studied. Similarly, as discussed in Chapter 4, it is difficult to ensure that all of the home births in a study sample were planned and attended because of variability in birth certificates from state to state whereby some accidental home births cannot be distinguished from planned ones (California), and planned, unassisted home births (also called “freebirths”) cannot be readily distinguished from those that are attended by a trained midwife (all states).

Comparative Risk of Neonatal Mortality and Morbidity

U.S. studies show elevated rates of neonatal mortality in home births compared with hospital births; see Tables 6-3 and 6-4. The relative risk to the infant may be two-fold, with absolute risks of about 1.2/1,000 versus 0.6/1,000 for home and hospital, respectively. The literature is not conclusive as to the magnitude of these rates because the available data make it difficult to distinguish between planned and unplanned or accidental home births, attended and intentionally unassisted births (also called “freebirths”), and provider type, if present.

TABLE 6-3. Comparative Risk of Neonatal Mortality by Birth Setting.

TABLE 6-3

Comparative Risk of Neonatal Mortality by Birth Setting.

TABLE 6-4. Comparative Risk of Neonatal Morbidity by Birth Setting.

TABLE 6-4

Comparative Risk of Neonatal Morbidity by Birth Setting.

Finding 6-1: Statistically significant increases in the relative risk of neonatal death in the home compared with the hospital setting have been reported in most U.S. studies of low-risk births using vital statistics data. However, the precise magnitude of the difference is difficult to assess given flaws in the underlying data. Regarding serious neonatal morbidity, studies report a wide range of risk for low-risk home versus hospital birth and by provider type. Given the importance of understanding these severe morbidities, the differing results among studies are of concern and require further study.

Research is critically needed to further evaluate neonatal outcomes among home and freestanding birth centers. Vital statistics studies of freestanding birth center outcomes show an increased risk of poor neonatal outcomes. Studies conducted in the United States using an intention-to-treat approach have demonstrated that births in birth centers and hospitals have similar to slightly elevated rates of neonatal and perinatal mortality.

Finding 6-2: Vital statistics studies of low-risk births in freestanding birth centers show an increased risk of poor neonatal outcomes, while studies conducted in the United States using models indicating intended place of birth have demonstrated that low-risk births in birth centers and hospitals have similar to elevated rates of neonatal mortality. Findings of studies of the comparative risk of neonatal morbidity between low-risk birth center and hospital births are mixed, with variation across studies by outcome and provider type.

Moreover, giving the interrelationship of midwife credentialing with birth settings, its mediating effect on perinatal outcomes cannot be ascertained with confidence from the current literature.

MATERNAL OUTCOMES BY U.S. BIRTH SETTING

Overuse and associated intervention-related maternal morbidity have been well documented in U.S. hospitals. Over the past several decades, cesarean birth rates have increased to 31.9 percent (Hamilton et al., 2019), which is higher than the generally recognized level at which lifesaving maternal and neonatal benefits outweigh the risks. Potential complications of cesarean births include a greater incidence of both maternal and infant outcomes including hemorrhage, infection, and admission to the intensive care unit; longer hospital stays; reduced breastfeeding success; and infant respiratory problems. A Healthy People 2020 goal is to reduce the nulliparous, vertex, term, singleton (NVTS) cesarean rate to 23.9 percent (Office of Disease Prevention and Health Promotion, 2019).

As discussed in Chapter 2, hospitals across the United States vary widely in terms of rural or urban setting, level of care, resources available, and staffing models. Researchers have examined some of these differences and how they impact maternal outcomes, including interventions during birth and morbidities. Kozhimannil (2014) looked at differences between urban and rural hospitals, and found that rates of non-indicated cesarean birth and non-indicated labor induction were not dramatically different between the two in 2010 (16.9% and 17.8% for cesarean birth and 16.5% and 12% for induction, respectively). However, the rates of cesarean birth rose in both types of hospital between 2002 and 2010, and the rate of non-indicated labor induction rose disproportionately faster in rural compared with urban hospitals. Snyder and colleagues (2011) compared the rates of labor intervention in university and community hospitals across Ohio. They found that women giving birth in community hospitals were more likely to be induced at 37 weeks (1.7 adjusted odds ratio [aOR]), at 38 weeks (1.8 aOR), and at 39–42 weeks (2.0 aOR) compared with women in university hospitals. However, rates of cesarean birth did not differ between the two types of hospitals. Fingar and colleagues (2018) looked at rates of severe maternal morbidity across multiple kinds of hospitals. They found the outcome to be more prevalent in safety net hospitals, minority-serving hospitals, teaching hospitals, public (compared with privately owned) hospitals, and hospitals in the Northeast and South.

Much of the research examining hospital-specific rates of intervention and morbidity consists of studies that compare hospital births with home or birth center births in order to draw conclusions about differences among settings. Because of their goal of making comparisons, these studies usually adjust the population studied to match the lower-risk profiles of women who give birth at home or in a birth center; for example, a study might look only at births that are singleton, vertex, and full-term. Thus, the rates gleaned from the hospital data cannot be used as representative of the rates for all hospital births, only of the rates for low-risk women in hospital settings. Recognizing this limitation of studies that compare home and birth center births with hospital births, these studies have consistently found higher rates of maternal intervention and morbidity in planned hospital births. We review this literature—systematic reviews, studies using birth registry data, and studies using vital statistics data—in detail below; the results are summarized in Table 6-5.

TABLE 6-5. Rate, Percentage, and Comparative Risk of Maternal Intervention-Related Morbidity by Birth Setting.

TABLE 6-5

Rate, Percentage, and Comparative Risk of Maternal Intervention-Related Morbidity by Birth Setting.

Systematic Reviews

Wax and colleagues (2010) published a systematic review of the literature on health outcomes following planned home and hospital births that provided data on morbidity in the home setting as well. The authors included all English-language, peer-reviewed publications from high-resource countries available at the time that reported maternal, fetal, and neonatal outcomes by birth setting. In alignment with the committee’s findings, these authors note that it is impossible to evaluate maternal mortality by birth setting, as these data are not reported in the literature, or small sample sizes do not allow for meaningful analysis. However, they were able to compare planned home and hospital births for a broad range of general morbidity and intervention-related morbidity indicators. They found that planned home births were associated with fewer maternal interventions, including epidural analgesia (9.0% vs. 22.9%; OR 0.24; CI 0.22–0.25); electronic fetal heart rate monitoring (13.8% vs. 62.6%; OR 0.10; CI 0.09–0.10); episiotomy (7.0% vs. 10.4%; OR 0.26; CI 0.24–0.28); operative delivery (3.5% vs. 10.2%; OR 0.26; CI 0.24–0.28); and cesarean birth in healthy, low-risk mothers (5.0% vs. 9.3%; OR 0.42; CI 0.39–0.45). Women who planned home births were also less likely to experience 3rd- and 4th-degree tears 1.2% vs. 2.5%; OR 0.35; CI 0.33–0.45); infection (0.7% vs. 2.6%; OR 0.27; CI 0.19–0.39); postpartum hemorrhage/bleeding (4.9% vs. 5.0%; OR 0.66; CI 0.61–0.71); vaginal lacerations (7.9% vs. 22.4%; OR 0.85; CI 0.78–0.93); and retained placenta (1.2% vs. 1.6%; OR 0.65; CI 0.51–0.83). This association between reduced morbidity and home birth settings may be attributable both to home birth models of care and to the fact that healthy women who are highly motivated to avoid interventions are proportionately overrepresented in home birth samples.

Vital Statistics Studies

More recent research on maternal outcomes by birth setting has largely upheld the findings of the Wax and colleagues (2010) systematic review. Using vital statistics data, Cheng and colleagues (2013 [ResQu: moderate]) found lower rates of interventions in the home compared with births in the hospital setting, including operative vaginal birth (0.1% vs. 6.2%; aOR 0.12; CI 0.08–0.17), labor induction (1.4% vs. 25.7%; aOR 0.19; CI 0.18–0.22), augmentation of labor (2.1% vs. 22.2%; aOR 0.29; CI 0.27–0.31), and use of antibiotics in labor (2.6% vs. 15.2%; aOR 0.40; CI 0.37–0.42). Three studies of birth center outcomes also used vital statistics data (MacDorman and Declercq, 201613; Li et al., 201714; Stephenson-Famy et al., 201815).

Birth Registry Studies

As noted above and in Chapter 5, studies using birth registries allow for analysis on an intention-to-treat basis; however, because reporting to registries is not mandatory these results may not be generalizable. In addition, these studies are descriptive and do not have an explicit comparison group. Cheyney and colleagues (2014a [ResQu: high, GRADE: poor]), using data from a national registry, describe outcomes of planned home births in the United States between 2004 and 2009. Among 16,924 women who went into labor intending to give birth at home, 89.1 percent completed their birth at home. The majority of intrapartum transfers from home to hospital were for slow, nonprogressive labors, and only 4.5 percent of the total sample required oxytocin augmentation and/or epidural analgesia. The rates of spontaneous vaginal birth, assisted vaginal birth, and cesarean birth were 93.6 percent, 1.2 percent, and 5.2 percent, respectively (Cheyney et al., 2014a).

Four studies of birth center outcomes use data collected prospectively through the Perinatal Data Registry (PDR), a national, validated, online data collection tool developed by the American Association of Birth Centers (AABC) (Stapleton et al., 201316; Jolles et al, 201717; Thornton et al., 201718; Alliman et al., 201919). An additional study evaluates outcomes of the Strong Start Initiative—a project developed by the Centers for Medicare & Medicaid Innovation (CMMI) and conducted between 2013 and 2017 (Hill et al., 2018; see Chapter 4). Birth center studies using registry data consistently indicate that women who participate in birth center care experience low cesarean rates (6–12%) and high breastfeeding initiation rates (92–95%) (Jolles et al., 2017; Stapleton et al., 2013; Thornton et al., 2017).

Summary

Maternal outcomes by birth setting are remarkably consistent: low-risk home and birth center births are associated with lower rates of perineal laceration; reduced rates of medical intervention, including cesarean delivery; and higher rates of breastfeeding initiation and exclusive breastfeeding at 6–8 weeks postpartum. Most of the published data are from observational cohort studies, but several of those studies are based on large samples (Bailey, 2017; Hill et al., 2018; Stapleton et al., 2013) or include most or all birth center births in a region or country for a period of time (Birthplace in England Collaborative Group, 2011; Hollowell et al., 2017; Bailey, 2017; Grigg et al., 2017; Sprague et al., 2018).

Lower rates of intervention and higher rates of breastfeeding are at least partially attributable to selection bias, wherein those who choose home or birth center birth are often highly motivated to achieve a physiologic birth and to breastfeed. The precise effect of selection bias on birth center outcomes is not known. However, the balance of evidence also suggests that there is something about the wellness-oriented, individualized, relationship-centered approach of midwifery care across home, birth center, and hospital settings that contributes to lower rates of medical interventions that can be dangerous when overused.

To find reliable comparison groups for lower-risk birth center and home birth participants, some studies used exclusion criteria to compile low-risk groups so that women with no risk factors in each model could be compared. Other studies used regression analysis to control for differing risk levels to achieve more comparable groups for analysis. Overall, birth center outcomes are consistent for low- or lower-risk women for increased odds of spontaneous vaginal birth, decreased risk for cesarean and assisted vaginal birth, increased initiation and continuation of breastfeeding, and similar intrapartum and neonatal outcomes relative to hospital birth outcomes.

Finding 6-3: In the United States, low-risk women choosing home or birth center birth compared with women choosing hospital birth have lower rates of intervention, including cesarean birth, operative vaginal delivery, induction of labor, augmentation of labor, and episotiomy, and lower rates of intervention-related maternal morbidity, such as infection, postpartum hemorrhage, and genital tract tearing. These findings are consistent across studies. The fact that women choosing home and birth center births tend to select these settings because of their desire for fewer interventions contributes to these lower rates.

Intervention-related maternal morbidity also varies greatly across hospital settings. There are promising strategies and approaches to lowering the rates of non–medically indicated, morbidity-related interventions in hospital settings (see Chapter 7 for further discussion of these models).

PATIENT EXPERIENCE AND SATISFACTION BY U.S. BIRTH SETTING

Factors in Maternal Satisfaction and Relationship to Outcomes

Maternal satisfaction across birth settings has typically been highest when women are supported in choosing the birth setting and provider type that align most closely with their value systems, individual pregnancy characteristics, and personal preferences. Multiple studies from Europe and Canada have measured maternal satisfaction (Janssen et al., 2006; Christiaens and Bracke, 2009; Lindgren and Erlandsson, 2010), but no study in the United States has systematically compared maternal satisfaction across birth settings. A systematic review with publications from multiple countries by Hodnett (2002) found the most critical predictors of satisfaction to be individual expectations, the amount of support received from caregivers, the quality of the caregiver–patient relationship, and maternal involvement in decision making.

Several recent studies note high rates of maternal satisfaction when care is received from midwives regardless of location (Sandall et al., 2010; Macpherson et al., 2016), when doula care is provided (Hardin and Buckner, 2004; Kozhimannil et al., 2016; Thomas et al., 2017), by mode of delivery (Bossano et al., 2017; Alderdice et al., 2019), and when care is midwife-led at home and in birth centers (Fleming et al., 2016). For additional discussion of outcomes associated with doula care, see Box 6-1.

Box Icon

BOX 6-1

Influence of Doulas on Outcomes Across Settings.

In addition, it is known that one-to-one nursing care during labor and birth influences women’s satisfaction with their birth experience (Hodnett et al., 2002). Type of nursing care is a major factor in how women perceive the birth experience. Numerous studies informed by the voices of new mothers have found that women value support, encouragement, physical presence, explanations, and respect for their need for control (Corbett and Callister, 2000; Tumblin and Simkin, 2001; Hodnett, 2002; Matthews and Callister, 2004; Brown et al., 2009; Lyndon et al., 2017). Labor and delivery nurses have likewise been explicit about how the quality and quantity of their care is affected by inadequate nurse staffing (Simpson et al., 2012, 2016; Simpson and Lyndon, 2017a). In the context of inadequate staffing, nurses report that they are unable to accomplish all aspects of nursing care required because they are balancing the most pressing demands of the clinical needs of their additional patients. Labor support and physical presence at the bedside are the first aspects of care suspended when an obstetric unit is short-staffed (Simpson et al., 2012, 2016), even though multiple studies have shown that these aspects of care are essential to positive birth experiences. Box 6-2 elaborates on nurses’ influence on labor and birth outcomes.

Box Icon

BOX 6-2

Nurses’ Influence on Labor and Birth Outcomes.

Some U.S. women report finding some aspects of their childbirth experience to have been negative or traumatic, including feeling inadequately supported during the perinatal period and having poor-quality relationships/interactions with their care provider. Lack of support—a poor outcome in itself—has been associated with other undesirable psychosocial sequelae, including increased rates of postpartum mood disorders (Bell and Andersson, 2016; Tani and Castagna, 2017), birth trauma (Simpson and Catling, 2016; Hollander et al., 2017; Reed et al., 2017), and cesarean regret (Porter et al., 2007; Burcher et al., 2016).

A recent article by Vedam and colleagues (2019) reports findings from a convenience sample survey, administered by a multidisciplinary team that included service users, that was designed to capture the lived experiences of maternity care among diverse populations and across U.S. births settings. Patient-designed survey items included questions about verbal and physical abuse, failure to meet professional standards of care, autonomy, discrimination, poor rapport with providers, and substandard conditions within the health system. The researchers found that 17.3 percent of women, or one in six, had experienced at least one form of mistreatment during labor and birth (n = 2,138). Forms of mistreatment included loss of autonomy; being shouted at, scolded, or threatened; and having requests for help ignored or refused. Women who transferred to a hospital from a planned home or birth center birth or whose opinion on the best course of action differed from their provider’s reported even higher rates of mistreatment. Women’s experiences also differed significantly by birth setting, with 5.1 percent of women who gave birth at home reporting mistreatment versus 28.1 percent of women who gave birth in a hospital. A reduced likelihood of mistreatment was associated with giving birth vaginally, giving birth in a community setting (home or birth center birth), and giving birth with a midwife as the primary attendant regardless of location of care. Being White, multiparous, and older than age 30 were associated with lower levels of mistreatment. Mistreatment rates among women of color were consistently higher than those among White women, and this relationship held even when the authors accounted for interactions between race and other characteristics, such as socioeconomic status. Any mistreatment was reported by 27.2 percent of low-income women of color versus 18.7 percent of low-income White women. Regardless of maternal race and ethnicity, having a Black partner was also associated with a higher rate of mistreatment. Experiences of care and perceived vulnerability to obstetric violence or obstetric racism appear to play important roles in shaping maternal decision making around where and with whom to give birth, as well as around what constitutes safety. (See also the discussion of institutional bias and discrimination in Chapter 4.)

Summary

Psychosocial outcomes, including several measures of dignity in the childbirth process, such as bodily autonomy, maternal agency, respectful care, and empowerment, are important. Some studies show that patient satisfaction is higher and reports of disrespectful care are lower among home and birth center births than among hospital births. Recent research has prompted greater understanding that various forms of disrespect and abuse can occur during the childbirth process in the United States and that rates and types of mistreatment vary by maternal race/ethnicity.

Finding 6-4: Some women experience a gap between the care they expect and want and the care they receive. Women want safety, freedom of choice in birth setting and provider, choice among care practices, and respectful treatment. Individual expectations, the amount of support received from caregivers, the quality of the caregiver–patient relationship, and involvement in decision making appear to be the greatest influences on women’s satisfaction with the experience of childbirth.

INTERNATIONAL STUDIES OF OUTCOMES BY BIRTH SETTING

The committee examined studies of outcomes by birth setting internationally that could provide comparisons with the United States (see, e.g., Hutton et al., 2009, 2016; Janssen et al., 2009, 2015; Birthplace in England Collaborative Group, 2011; Schroeder et al., 2012; de Jonge et al., 2013, 2015, 2017; Homer et al., 2014; Vedam et al., 2014b; Zielinski et al., 2015; Bolten et al., 2016; Scarf et al., 2016, 2018). The committee chose Australia, Canada, the Netherlands, and the United Kingdom because they are high-resource countries and have relatively robust data on birth settings and outcomes from their vital statistics systems, as well as a range of well-conducted studies. Table 6-6 provides a comparison by country of types of providers, birth settings, and selected outcomes.

TABLE 6-6. Comparison of Types of Providers, Birth Settings, and Selected Outcomes, by Country.

TABLE 6-6

Comparison of Types of Providers, Birth Settings, and Selected Outcomes, by Country.

Of course, it is important to note the deep differences among countries that shape the types of health care systems in each nation. The committee commissioned a study to identify these differences across the four identified countries (Kennedy et al., 2019). The paper authors note several important commonalities. First, the four countries share a commitment to integration of care across birth providers and systems. In these countries, out-of-hospital birth providers are part of an integrated, regulated maternity care system. For example, in Australia, the Netherlands, and the United Kingdom, almost all vaginal births include at least one midwife in attendance, usually as the only professional present if the birth is without complications. In all four countries, midwives are trained through a post-secondary education program and prepared to handle first-line complications. This integration translates to a second shared feature of maternity care systems in these countries: seamless transfer across settings. Strong systems are in place in all four countries to provide for collaboration, consultation, transfer, and transport when access to an obstetrician is needed.

A third difference Kennedy and colleagues identified is the presence of universal access to primary and maternity care, including access to different (risk-appropriate) provider options during pregnancy and birth. This universal access to care (including preconception care) means women are neither without coverage prior to becoming pregnant nor dropped from health care coverage after they have had their baby, as is the experience of many Medicaid recipients in the United States (Ranji et al., 2019). In addition to these features, all four countries have adopted a practice of respectful care, including respect for maternal autonomy. This culture of respect informs the evidence-based guidelines in place in each country. These guidelines are intended to support clinical decision making for women and their providers and include information on appropriate risk selection and assessment, as well as out-of-hospital birth options. For example, the UK NICE Guidelines for Intrapartum Care (National Institute for Health and Care Excellence, 2017) clearly define the risk factors and situations in which consultation and (or) transfer of a laboring woman is required. Importantly, the guidelines support women’s choice in birth setting, reflecting the practice of trusting women to make appropriate decisions for themselves, their babies, and their families (Kennedy et al., 2019).

Australia, Canada, the Netherlands, and the United Kingdom also provide additional social and welfare supports as compared with the United States, as well as increased availability of maternal and paternal maternity benefits. As a result, the level of disparities and inequity among childbearing people and risk propensities are different from those found among childbearing people in the United States. Accordingly, international comparisons are inherently limited, but they do provide insights into how changes in the structure of health care systems might affect birth outcomes.

International Studies of Home Birth Outcomes

When examining international studies of home birth outcomes, it is important to recognize that the context of the maternity care systems in the four countries the committee chose for comparison is very different from that of the current U.S. system, being characterized by universal health coverage and access, standardized high-level midwifery training, regulated risk-based selection of birth setting, and systems for transfer to a higher level of care when needed. Studies from the four comparison countries are often based on local or national registry data, allowing for an intention-to-treat approach to analysis. Although this minimizes selection bias, missing data are often treated as uninformative, an assumption that is likely incorrect (Wiegerinck et al., 2018). Overall, the international studies reviewed by the committee indicate that benefits result from fewer maternal interventions. They also generally find no difference in neonatal death between planned home and hospital birth cohorts. Notable exceptions exist, including infants born to primiparous women, for whom higher rates of perinatal mortality are seen in the United Kingdom (Birthplace in England Collaborative Group, 2011), and several studies from the Netherlands (Evers et al., 2010; Daysal, 2015; Wiegerinck et al., 2018) show both higher perinatal mortality and an effect of distance on outcome. Table 6-7 shows international studies of neonatal outcomes by birth setting; Table 6-8 shows international studies of maternal outcomes by birth setting.

TABLE 6-7. Rate, Percentage, and Risk of Neonatal Morbidity and Mortality in Australia, Canada, the Netherlands, and the United Kingdom.

TABLE 6-7

Rate, Percentage, and Risk of Neonatal Morbidity and Mortality in Australia, Canada, the Netherlands, and the United Kingdom.

TABLE 6-8. Rate, Percentage, and Risk of Maternal Morbidity and Mortality in Australia, Canada, the Netherlands, and the United Kingdom.

TABLE 6-8

Rate, Percentage, and Risk of Maternal Morbidity and Mortality in Australia, Canada, the Netherlands, and the United Kingdom.

While four large studies from the Netherlands have found no significant differences in intrapartum or neonatal mortality rates for planned home versus planned hospital births (de Jonge et al., 2009 [ResQu: high]20; van der Kooy et al., 2011 [ResQu: high, GRADE: fair]21; de Jonge et al., 2015 [ResQu: high, GRADE: fair]22; and de Jonge et al. (2013 [ResQu: high, GRADE: fair]),23 the evidence on outcomes by setting in the Netherlands is mixed. Wiegerinck and colleagues (2018 [ResQu: high, GRADE: poor]) compared intrapartum and neonatal mortality in low-risk term women starting labor in midwife-led versus obstetrician-led care (n = 57,396). Perinatal mortality occurred in 30 of 46,764 (0.064%) women in midwife-led care and in 2 of 10,632 (0.019%) women in obstetrician-led care (OR 3.4, 95% CI 0.8–14.3).

In Australia, Kennare and colleagues (2010 [ResQu: moderate, GRADE, poor]) found similar perinatal mortality rates between home births and hospital births (7.9 vs. 8.2 per 1,000 births) using a retrospective population-based design. However, they found a higher intrapartum fetal death rate in the home birth group (1.8 vs. 0.8 per 1,000 births), with significantly lower cesarean (9.2% vs. 27.1%) and episiotomy (3.6% vs. 21.7%) rates for home versus hospital births. Catling-Paul and colleagues (2013 [ResQu: low, GRADE: poor]) examined 12 publicly funded home birth programs in Australia (n = 1,807; 9 of the programs provided information, for a total of 97% of all home births nationally) and found, after excluding babies with fetal anomalies, a neonatal mortality rate of 1.7 per 1,000 births and a 5.4 percent cesarean rate. The largest study to date (n = 258,161, with 0.3% planning a home birth) within the country was a retrospective analysis of public birth data conducted by Homer and colleagues (2014 [ResQu: high, GRADE: fair]). They found no significant differences in composite perinatal and neonatal mortality/morbidity index score (7.1/1,000 for planned home births vs. 5.8/1,000 for planned hospital births), and a significant difference in cesarean delivery rates (3.3% vs. 10.6% for home and hospital births, respectively), but the authors also note they were unable to test the reliability of the differences because they did not have the statistical power.

In Canada, Janssen and colleagues (2009 [ResQu: high, GRADE: poor]) conducted a prospective cohort study and found no significant differences in perinatal mortality among three groups—midwife-attended home births (0.35/1,000), midwife-attended hospital births (0.57/1,000), and physician-attended hospital births (0.64/1,000). Maternal outcomes were significantly better in the home birth group than in the midwife-attended and physician-attended hospital groups, with the following specific outcomes, respectively: cesarean delivery rate: 7.2 percent vs. 10.5 percent vs. 11 percent, intact perineum rate: 54.4 percent vs. 46.1 percent vs. 43 percent, and postpartum hemorrhage rate: 3.8 percent vs. 6.0 percent vs. 6.7 percent. A second Canadian study (Hutton et al., 2009 [ResQu: high, GRADE: poor]) with a retrospective cohort design (N = 6,692 planned home births matched to 6,692 planned hospital births for comparable low-risk women) found no differences in combined perinatal/neonatal mortality rates (1/1,000 in both samples) or in composite perinatal and neonatal mortality/morbidity scores (2.4% for home vs. 2.8% for hospital). Cesarean delivery rates were significantly lower in the home group (5.2% vs. 8.1%), as was postpartum hemorrhage with more than 1,000 mL of blood loss.

The Birthplace in England Collaborative Group (2011 [ResQu: high, GRADE: good]) conducted a prospective cohort study comparing outcomes for births occurring in home, birth center, and midwifery and obstetric hospital units for 64,538 low-risk women at term. A composite outcome was created that combined stillbirth, early neonatal death, meconium aspiration, birth-related injuries, and encephalopathy. Overall, no significant differences in the composite outcome for the entire sample were found. When the sample was stratified by nulliparity, rates for the composite outcome were higher for home than for hospital for nulliparous women (9.3/1,000 vs. 5.3/1,000). A follow-up study focused on costs concluded that home birth was a cost-effective option for all low-risk women, including nulliparas (Schroeder et al., 2012).

Studies from Japan, New Zealand, Norway, and Sweden also found outcomes similar to those of the studies discussed above (see, e.g., Kataoka et al., 2013; Davis et al., 2011; Blix et al., 2012; Lindgren et al., 2008).

International Studies of Birth Center Outcomes

In other countries, including the United Kingdom, Australia, Canada, and the Netherlands, integration of birth center care into maternity care systems is much better than in the United States (Bailey, 2017; Birthplace in England Collaborative Group, 2011; Davis et al., 2012; Grigg et al., 2017; Hollowell et al., 2017; Sprague et al., 2018). In the United Kingdom, women have the option of choosing from multiple birth settings, including home, freestanding midwifery unit (FMU), a model similar to the U.S. birth center alongside midwifery unit (AMU) (a midwifery unit located in or collocated with a hospital), and obstetric hospital unit.

The Birthplace in England Collaborative Group (2011) was a prospective cohort study with a total sample size of 64,538, including 11,282 women planning births in an FMU. Researchers found no increased odds of poor outcomes for neonates born in FMUs versus obstetric units. Using the obstetric unit as the reference point, percentages and aORs for outcomes in the FMUs were spontaneous vaginal birth, 90.7 percent (aOR 3.38); cesarean birth, 3.5 percent (aOR 0.32); blood transfusion, 0.5 percent (aOR 0.48); and pitocin augmentation, 7.1 percent (aOR 0.26). Despite these outcomes, as of 2016, only about 2 percent of births in the United Kingdom occurred in FMUs (Walsh et al., 2018).24

A secondary analysis of the Birthplace in England Collaborative data (Hollowell et al., 2017 [ResQu: high, GRADE: fair]) compared outcomes in FMUs (n = 11,265) and AMUs (n = 16,673) and found no significant differences in adverse perinatal outcomes. Odds of “straightforward vaginal birth” were higher for planned FMU compared with planned AMU births (Hollowell et al., 2017). Compared with women with planned AMU births, women with planned FMU births had significantly lower odds of instrumental delivery and significantly lower rates of epistomy, epidurals, and augmentation of labor with syntocinon (Hollowell et al., 2017). In addition, one micro-costing study found a 40 percent reduction in costs for FMU care compared with hospital care for low-risk women, even when the cost of transfers was included (Schroeder et al., 2017).

A birth center study from Canada evaluated outcomes for all birth center admissions (n = 495) for the first year of operation in two Toronto-area facilities and compared them with outcomes among a low-risk group with midwife-led hospital births. No maternal deaths occurred; one fetal death was discovered at triage at the birth center, with immediate transfer to the hospital. Outcomes favored the birth center group, which evidenced lower rates of cesarean (7.7% vs. 12.1%) and synthetic oxytocin augmentation (12.5% vs. 24.5%) (Sprague et al., 2018 [ResQu: moderate, GRADE: poor]).

Appropriate maternity care in the Netherlands is determined based on risk level within an integrated system in which women have a choice of birth setting (Hermus et al., 2017; Schuit et al., 2016). All low-risk women have the option of being admitted when in labor for either home birth or primary labor care in a hospital with an independent midwife or a general practitioner. Birth center birth is also a primary care option for low-risk women in the Netherlands, though access to birth centers is limited in some regions. During the time period of the Schuit study, from 2000 to 2007, 52 percent of all women were admitted to primary hospital maternity care, and 32 percent of this group were transferred to secondary care during labor (Schuit et al., 2016; n = 746,642). During this time period, only 1 percent of women in the Netherlands planned to give birth in a birth center. For those women, odds of transfer from the birth center to a hospital were similar to the odds of transfer from primary to secondary care in the hospital, and greater than the odds of transfer from a planned home birth to a hospital (Schuitt et al., 2016).

In 2013, a national study of birth center care was conducted in the Netherlands, in which 21 of all 23 birth centers in the country participated (Hermus et al., 2017). Some of these birth centers were freestanding (3), but a majority were alongside a hospital (14). Outcomes were assessed using an optimality index, which focuses on processes and outcomes rather than adverse outcomes (Hermus et al., 2017). Of a total sample of 3,455 women, those in birth center care experienced optimality scores similar to those of women in midwife-led care in hospitals, but lower scores than those having home birth. Care in freestanding birth centers led to higher rates of spontaneous vaginal birth and fewer transfers relative to care in alongside hospital birth centers (Hermus et al., 2017).

A secondary analysis of matched cohort data in Denmark included 839 women each in birth center and hospital groups (Christensen and Overgaard, 2017 [ResQu: high, GRADE: poor]). Caesarean birth was lower in the birth center group, with odds ratios of 0.4 for nulliparous and 0.8 for multiparous women. For uncomplicated birth center births, the odds ratio was 2.2 for nulliparous and 2.9 for multiparous women (Christensen and Overgaard, 2017).

In Australia, the midwife-led units called “birth centers” are more similar to AMUs in the United Kingdom than to freestanding birth centers in the United States and New Zealand, usually being located within hospitals instead of being freestanding (Homer et al., 2014; Laws et al., 2014). Two cohort studies in Australia compared outcomes for women having planned births in midwife-led versus obstetric units. A study by Homer and colleagues (2014 [ResQu: high, GRADE: fair]) included 14,483 planning midwife unit care and 242,936 planning obstetric unit care, while a study by Laws and colleagues (2014 [ResQu: high, GRADE: poor]) included 15,742 participants in planned midwife unit care and 66,190 in obstetric unit care. In the midwife-led units, women were statistically more likely than those in the obstetric units to experience a normal vaginal birth, with cesarean rates of 4.8 percent versus 10.6 percent (aOR 0.36, 95% CI 0.34–0.39) in the Homer and colleagues study and 3.9 percent versus 12.6 percent (aOR 0.23, 95% CI 0.20–0.25) in the Laws and colleagues study. There were no statistically significant differences in stillbirth and early neonatal deaths among the three groups (Homer et al., 2014).

Studies from Japan, New Zealand, Norway, and Sweden also found outcomes similar to those of the studies discussed above (see, e.g., Kataoka et al., 2013; Suto et al., 2015; Christensen and Overgaard, 2017; Øian et al., 2018; Grigg et al., 2017; Davis et al., 2012; Bailey, 2017).

Summary

Two international studies suggest a small increase in adverse outcomes for the neonate in home versus hospital births for low-risk individuals (Caughey and Cheyney, 2019). The vast majority of international evidence, however, when limited to high-quality studies, particularly those from countries with well-integrated maternity care systems and clear collaboration guidelines, generally show no increase in neonatal morbidity or mortality for low-risk, planned home or birth center births versus low-risk hospital births, though notable exceptions exist. Taken together, U.S. and international studies suggest that for infants, home births for low-risk individuals may be as safe as hospital or birth center births when certain system-level features are in place, including collaboration and integration across birth settings, eligibility criteria for community birth, well-trained providers, appropriate risk selection, ability to manage first-line complications, interdisciplinary collaboration, choice among multiple birth settings that are covered through universal coverage policies, and low barriers to transfer. System-level features and the availability of institutional supports for physiologic childbearing and respectful care appear to play important roles in determining perinatal risk (Vedam et al., 2018). Conversely, in maternity care systems that lack these features, higher rates of infant morbidity and mortality are found. In the United States, integration of midwifery into maternity care delivery systems is fragmented, coverage is not available for all women or for all birth providers, and care is poorly coordinated, contributing to adverse perinatal outcomes.

Finding 6-5: International studies suggest that home and birth center births may be as safe as hospital births for low-risk women and infants when (1) they are part of an integrated, regulated system; (2) multiple provider options across the continuum of care are covered; (3) providers are well qualified and have the knowledge and training to manage first-line complications; (4) transfer is seamless across settings; and (5) appropriate risk assessment and risk selection occur across settings and throughout pregnancy. Such systems are currently not widespread in the United States.

As demonstrated by the international literature, risks in home and birth center settings may be mitigated by various strategies, including selection of low-risk mothers, referral to an obstetric or maternal fetal medicine provider for pregnancy complications, low thresholds for transfer, timely (20–30 minutes) transport when complications arise, barrier-free and mutually respectful transfers of care when needed, collaborative professional models of care, formal training of skilled practitioners (including in neonatal resuscitation), professional regulation, and oversight and accountability. Professional regulation may be particularly important. In the United States today, certified professional midwives, who attend a majority of U.S. home births, are regulated in only 33 states. In states where they are not regulated, there are no minimum standards for practice; regulated access to lifesaving, first-line medications, including antihemorrhagics; or access to continuing education, quality improvement opportunities, or professional development. In all international settings, where home and freestanding birth center births are better integrated into the maternity care system, these strategies have been widely implemented, and as a result, outcomes for both women and newborns tend to be better in these settings than in the United States, where health systems do not employ these strategies.

Finding 6-6: Lack of integration and coordination and unreliable collaboration across birth settings and maternity care providers are associated with poor birth outcomes for women and infants in the United States.

INTERPROFESSIONAL COLLABORATION ACROSS THE MATERNITY CARE TEAM AND BETWEEN BIRTH SETTINGS

Interprofessional teamwork is essential to the provision of high-quality maternity care (Guise and Segel, 2008). Previous research has shown that when professionals collaborate on decision making and coordination of care is seamless, fewer preventable intrapartum neonatal and maternal deaths occur during critical obstetric events (Cornthwaite et al., 2013). Poor communication, disagreement, and lack of clarity around provider roles are identified as primary determinant of these adverse outcomes (Guise and Segel, 2008; The Joint Commission, 2004; Cornthwaite et al., 2013). When differences around defining risk and responsibility exist among providers, interprofessional cooperation and access to options for care are reduced (Barclay et al., 2016; Coxon et al., 2016; Healy et al., 2016).

Conversely, collaboration among health professionals can improve safety and quality, particularly as different members of the health care team contribute differing perspectives and areas of expertise. Two studies used data from the National Institutes of Health’s (NIH’s) Consortium on Safe Labor to compare outcomes from obstetric units having physician-only care with those from units in which midwives and physicians practiced together (Carlson et al., 2019; Neal et al., 2019). They both found that women receiving care in the latter units were less likely than women at physician-only centers to experience induction, oxytocin augmentation, and cesarean birth. Collaboration among health professionals can also greatly improve safety when care must be transferred across birth settings as interprofessional teamwork has been shown to be central to providing safe, effective, and efficient obstetric care (Guise and Segel, 2008).

In the United States, access to maternity care that is coordinated among homes, birth centers, and hospitals is unreliable, uncommon (Shah, 2015), highly variable, and generally shaped by physician perspectives (Leone et al., 2016; Rainey et al., 2017). U.S. obstetricians lack clear protocols for determining when and how to transfer patients to hospitals offering a higher level of care and risk-appropriate providers. Moreover, U.S. hospitals and birth centers often lack formal referral relationships and may face financial disincentives to transfer patients (Cheyney et al., 2014c; Shah, 2015; Vedam et al., 2014a).

Cheyney and colleagues (2014c) examined the views of hospital- and home-based clinicians in the context of 50 home-to-hospital transfers through open-ended, semistructured interviews (n = 40), and engaged in a process of reciprocal ethnography whereby results were returned to participants for comment and critique. Six key themes (three from receiving providers and three from referring midwives) that emerged from the interviews highlighted differences in referring and receiving providers’ perspectives and experiences of transfer and interprofessional collaboration.

Hospital-based providers in this study described (1) the belief that home birth is substantially more dangerous than published studies suggest; (2) the experiences of fear and frustration generated when physicians or CNMs are forced to assume the risk of caring for another provider’s patient; and (3) challenges related to unfamiliar charting and strained interprofessional communication during the heightened emotions of a transfer (Cheyney et al., 2014c, p. 446). Further, the perception that out-of-hosptial midwives, regardless of credential (i.e., certified professional midwives [CPMs], CNMs, or licensed midwives [LMs]), had mismanaged the needs of the woman and infant made hospital practitioners question the quality of data and publications that conclude that home birth is safe under certain conditions. The physicians in this study expressed fear and vulnerability over having to take over the care of a woman in labor who was transferred, even with the knowledge that few transfers are emergent.

Out-of-hospital midwives’ transfer narratives focused on three key themes that differed from those of hospital-based colleagues: (1) midwives’ tendency to defend the midwifery model of care; (2) physicians’ tendency to judge midwives by “the exception, rather than the rule”; and (3) physicians’ failure to take responsibility for their roles in poor state and national maternal–child health outcomes (Cheyney et al., 2014c, p. 449), instead blaming midwives. Clients worried about “punitive” cesarean sections and humiliating “blaming and shaming” for attempting a home birth, and this often led to a refusal of transfer until minor complications (i.e., a slow, nonprogressive labor) developed into something more severe (i.e., fetal distress). Fear on the part of clients and the lack of collaboration between community midwives and hospitals/providers sometimes led to a delay in transfer that could be detrimental (Cheyney et al., 2014c).

Following their analysis of common themes, Cheyney and colleagues (2014c) outlined a larger set of sociopolitical mechanisms that restrict collaboration between community midwives and receiving physicians. The first is the ethical conflict of interest providers face because professional associations ignore the reality that in some instances, care must be shared. The second is restrictive legislation that prevents CPMs in many states from gaining the legal status that is a precursor to the training and regulation that are likely to improve the quality of care and facilitate integration. The third is the cycle of liability concerns and fear of adverse outcomes that lead to delays in care and fractured communication, which in turn contribute to the actualization of the feared increased liability and bad outcomes. These mechanisms impede efficient and mutually respectful interactions and can result in costly delays (Cheyney et al., 2014c).

The authors also argue, however, that these mechanisms could lead to possible solutions. Midwives requested that, if a transfer is required, the receiving hospital staff show respect for them and the woman and include them in dialogues regarding the best course of treatment, while the hospital-based providers hoped the midwives would provide them with timely and clear charting. Hospital-based providers requested that midwives prepare their clients for the possibility of a transfer prior to labor, and the midwives encouraged the hospital staff to not assume that someone who has attempted a home birth will necessarily decline hospital procedures. These solutions can assist in creating an integrated maternity system premised on mutual accommodation and smooth articulations across birth settings and provider types (Cheyney et al., 2014c).

Fox and colleagues (2014) published a metasynthesis aimed at developing a more nuanced understanding of women’s experiences of home-to-hospital transfer by synthesizing and interpreting the then existing body of qualitative research. Three categories emerged from their synthesis: (1) communication, connection, and continuity; (2) making the transition; and (3) making sense of events. Their review of four studies (n = 45) identified three factors that make transfers as seamless as possible from the perspective of the laboring woman: (1) quality and clarity of communication, (2) feeling connected to the backup hospital, and (3) continuity of midwifery care. Initial arrival at the hospital is a time of vulnerability and fear for clients who have spent their pregnancy planning for a community birth. Retaining the care of a known midwife is the core coping technique those women use to make the move—physically and ideologically—to a higher level of care. Receiving providers who are sensitive to women’s needs to be reassured and accepted greatly reduce the tension, fear, and stress that mark the transfer experience for the patient. In addition, the reasons for transfer must be clearly communicated, both at the time of transfer and then again in more detail following the birth whenever possible. Fox and colleagues argue that women need to talk through their experiences of transfer, and that they need to have their feelings of disappointment acknowledged in order to move forward to the next phase of parenting. Focusing on the fact of a healthy baby is not sufficient and can be counterproductive if it omits the process of grieving the lost experience. Continuity of care provider was found to be essential to this process because it enables understanding and coping alongside a known caregiver. Several additional studies on the experiences of transfer have also been published as part of the Australian Birthplace Study (see Fox et al., 2018a, 2018b), as well as in other high-resource countries that are currently experiencing a rise in planned home births (Rowe et al., 2012; Ball et al., 2016; Blix et al., 2016; Patterson et al., 2017).

Following the 2011 Home Birth Consensus Summit in the United States, scholars from family medicine, midwifery, nursing, health administration, obstetrics, public health, pediatrics, and ethics, as well as consumers and childbirth educators, formed a collaborative work group with the goal of translating the existing body of literature on transfers from home or birth center to hospital into an applied set of best-practice guidelines for all professionals involved when a transfer is necessary (Vedam et al., 2014a).

Members of the work group reviewed national and international exemplars of best-practice protocols and standards for effective communication and documentation during transfer, examined the literature on strategies for promoting interprofessional coordination and collaboration, and developed a rating system to assess the relevance and clarity of each resource. Findings highlighted the need for “increased commitment to shared decision making, mutually respectful communication between maternity care providers and health system staff, quality improvement processes and policies to ensure ongoing evaluation of outcomes of transfers, and expanded interprofessional education opportunities” (Vedam et al., 2014a, p. 631). The work group collated key components into Best Practice Guidelines for Transfer from Home to Hospital (Home Birth Summit, n.d.).

After describing the methods used in the development of these transfer guidelines, Vedam and colleagues (2014a, p. 632) make the following statement: “Regardless of one’s opinion of planned home birth, all clinicians and researchers can agree on the importance of improving interprofessional collaboration. Progress will require stakeholders with historically opposing views to find common ground within the contested space of home birth, especially when all share responsibility for care.” As Caughey and Cheyney (2019, p. 1042) have recently argued, “everyone involved shares the responsibility for reducing the chasms between community and hospital care and between obstetricians and midwives where these exist; a shorter distance to traverse literally, metaphorically, and ideologically could mean improved outcomes for all.” In fact, research suggests that integration of midwives into regional health systems is a key determinant of optimal maternal–newborn outcomes. A recent study of midwifery integration in the United States (Vedam et al., 2018) showed that states with midwifery-inclusive laws and regulations were correlated with better maternal and neonatal health outcomes and had higher rates of physiologic birth, breastfeeding, and vaginal birth after cesarean (VBAC). Conversely, poor coordination of care across providers and birth settings has been associated with adverse maternal–newborn outcomes. Yet prior to this study, the characteristics of an integrated system had not been described or linked to health disparities.

The study by Vedam and colleagues (2018) consisted of a multidisciplinary team of scholars who examined published regulatory data that described the regulations around the practice of midwifery and interprofessional collaboration across each state in the United States. The team used a modified Delphi process and selected 50 key items to create a weighted, composite Midwifery Integration Scoring system (MISS). These items measured the differences across jurisdictions in scope of practice, autonomy, governance, and prescriptive authority for midwives, as well as restrictions that can affect safety, quality, and access to providers across birth settings. States were ranked by MISS scores, and using reliable indicators in the Centers for Disease Control and Prevention’s (CDC’s) Vital Statistics Database, correlation coefficients were calculated between MISS scores and maternal–newborn outcomes by state. Hierarchical linear regression analyses were used to control for confounding effects of race and other factors.

MISS scores ranged from lowest at 17 (North Carolina) to highest at 61 (Washington) out of 100 points, indicating a wide range of integration across the United States, as well as generally low levels, as 61/100 was the highest score achieved. MISS scores correlated with the density of midwives and access to care across birth settings. States with higher MISS scores had significantly higher rates of spontaneous vaginal delivery, VBAC, and breastfeeding and significantly lower rates of cesarean, preterm birth, low-birthweight infants, and neonatal death. Significant differences in newborn outcomes, including preterm birth, low birthweight, and neonatal death, were accounted for by MISS scores and persisted after controlling for proportion of African American births in each state. Higher MISS scores were associated with significantly higher rates of physiologic birth, fewer obstetric interventions, and fewer adverse neonatal outcomes.

Overall, findings from this study suggest that states with higher levels of midwifery integration have improved outcomes for pregnant people and newborns, and states with lower levels of midwifery integration have poorer outcomes. Previous studies have also demonstrated similar relationships among midwifery care, systems integration, and improved maternity care outcomes (Cornthwaite et al., 2013; Manojlovich, 2014; Comeau et al., 2018; Reszel et al., 2018). The midwifery integration project is an ecological study, and as such, its main weaknesses is that it does not also assess patient-level data. Important individual-level confounders, moderators, and mediating factors could have affected the findings. In addition, as Vedam and colleagues (2018) are careful to note, the midwifery integration study has identified important correlations, but not necessarily causal relationships. Further research is needed to help clarify relationships among the policy environment, midwifery care across settings, individual-level risk factors, and birth outcomes in the United States.

CONCLUSION

There has yet to be a national, prospective, cohort study in the United States that utilizes an intention-to-treat model to compare outcomes by planned birth location and provider type, pays adequate attention to statistical power for rare outcomes, and controls for maternal risk factors and other confounders. Certainly, a better understanding of the true effect size of choice of birth setting with respect to fetal and neonatal outcomes would be achievable with a nationally validated, granular data registry (Caughey and Cheyney, 2019). With the rising rate of home and birth center births in the United States over the past several decades, it has become increasingly imperative to have a system that allows the tracking of outcomes by intended place of birth, provider type at the onset of labor, transfers over the course of care, and pregnancy characteristics.

Given both the acute and downstream risks of unnecessary interventions and the risks associated with potentially delayed access to lifesaving obstetric and neonatal interventions, birth setting decisions trade off some risks for others (Tilden et al., 2017), meaning there is no clear, risk-free option for giving birth (Caughey and Cheyney, 2019). In the United States, women and clinicians who desire access to medical intervetions generally prefer birth in a hospital (Tilden et al., 2017), while women and providers who do not want unneccesary interventions and who are focused on maternal autonomy and physiologic birth may consider home or birth center births. These distinctions are neither universal nor dichotomous: many hospital-based providers are committed to preventing unnecessary cesarean births and supporting physiologic birth, while some home birth and birth center midwives may overintervene (Caughey and Cheyney, 2019).

Taken together, the literature reviewed in this chapter makes clear that there are risks and benefits for pregnant women and newborns in each of the three birth settings in the United States. However, the literature (particularly from the international experience) also suggests that these risks are modifiable by systems through processes, policies, providers, and regulation.

CONCLUSION 6-1: In the United States, home, birth center, and hospital birth settings each offer risks and benefits to the childbearing woman and the newborn. While no setting is risk free, these risks may be modifiable within each setting and across settings.

The committee’s review of the relevant literature on health outcomes by birth setting revealed a dearth of evidence related to the possible connection between maternal mortality and severe maternal morbidity and birth settings, likely because these events are so rare. Indeed, only one case of either of these outcomes following a planned home or birth center birth (see Cheyney et al., 2014a) has been reported in the literature on safety by birth setting.

CONCLUSION 6-2: A lack of data and the relatively small number of home and birth center births prevent exploration of the relationship between birth settings and maternal mortality and severe maternal morbidity.

In the next chapter, we turn to the question: How can each birth setting work to improve outcomes and make birth safer, where safety encompasses both clinical and psychosocial outcomes?

Footnotes

1

Roome and colleagues (2016) reviewed the position statements on home birth for midwifery and obstetric colleges in the United States, the United Kingdom, Canada, Australia, and New Zealand in an effort to examine how the same body of research tends to lead to different positions on the acceptability of birth in the home setting. They found that midwifery organizations tend to support home birth as a viable option for healthy women, whereas physicians’ organizations have statements that oppose this option. In 2015, the United Kingdom was the only country reviewed where physician- and midwife-led organizations had issued a joint statement in support of home birth. Roome and colleagues found widely differing stances that they argue reflect traditional midwifery perspectives on childbirth as a physiologic process versus obstetric perspectives, which focus on the potential for pathology. Ultimately, these authors assert that the differences in position statements are largely the by-product of confirmatory bias (i.e., the tendency to process information by looking for, or interpreting, information that is consistent with one’s existing beliefs).

2

Causes of death in this group refer to failure to resuscitate or consequences of severe hypoxic ischemic encephalopathy (see Chapter 3).

3

The study by Cox and colleagues (2015), who used the same data registry, describes neonatal outcomes for women (n = 1,052) who planned a vaginal birth after cesarean (VBAC) at home with midwives who were contributing data to the Midwives Alliance of North America Statistics Project (MANA Stats) 2.0 data registry between 2004 and 2009. Five neonatal deaths (4.75/1,000) occurred in the prior cesarean group compared with 1.24/1,000 in multiparas without a history of cesarean (p = 0.015).

4

Cheyney and colleagues (2014a) found elevated rates of fetal intrapartum and neonatal mortality in a home birth sample (n = 16,924) when clients with higher-risk factors such as breech, twins, labor after cesarean section with no prior vaginal birth, gestational diabetes, and preeclampsia were included in the sample. Low Apgar scores (<7) occurred in 1.5 percent of newborns, and postpartum neonatal transfers were infrequent, occurring in only 0.9 percent of births. In terms of postnatal outcomes, 86 percent of newborns were breastfeeding exclusively at 6 weeks of age. Excluding lethal anomalies, the intrapartum, early neonatal, and late neonatal mortality rates were 1.30, 0.41, and 0.35 per 1,000, respectively, when higher-risk births were included, for a combined perinatal death rate of 2.06 per 1,000. When the sample was limited to low-risk women (term, singleton, vertex, no previous cesarean), the intrapartum mortality rate dropped to 0.85/1,000.

5

The neonatal outcome composite consisted of severe prenatal outcomes: intrapartum and newborn mortality, hypoxic neurologic injury, Apgar score <4 at 5 minutes, seizures, persistent pulmonary hypertension, positive pressure ventilation >10 minutes, and meconium aspiration syndrome.

6

Thornton and colleagues (2017), using Perinatal Data Registry data, used exclusion criteria to form low-risk groups admitted to birth centers (n = 8,776) and those that chose hospital admission (n = 2,527).

7

The study by Cheng and colleagues (2013) used 2008 vital statistics data from 27 states, which included the 2003 revision of the birth certificate that delineates the location of a birth as hospital, freestanding birth center, or home, and further as accidental, intended, or unknown if intended. Although the 2003 revision of the birth certificate includes information about where the birth took place and the planned status of that birth, it does not take into account intention-to-treat. Thus, the data do not account, for example, for planned home births that were transferred to hospitals. The authors compared outcomes of neonates whose mothers had planned home births and those who delivered in hospitals, and found that, compared with hospital births, more planned home births had 5-minute Apgar scores below 4 and a lower rate of NICU admission; they do not report neonatal mortality.

8

In their 2013 study, Grünebaum and colleagues used birth certificate data from the CDC’s National Center for Health Statistics to examine deliveries by physicians and midwives in and out of the hospital between 2007 and 2010 for a national sample of nearly 14 million singleton term births. Term was defined as 37 weeks or more gestation and a birthweight of 2,500 g or more. The majority of term singleton births (91%; n = 12,663,051) were physician-attended hospital births; midwife-attended hospital births constituted 8 percent of births (n = 1,118,678), 0.3 percent were midwife-led freestanding birth center births (n = 42,216), and 0.5 percent (n = 67,429) were midwife home deliveries. Grünebaum and colleagues (2013) found that, compared with hospital births attended by physicians or midwives, home births and births in freestanding birth centers attended by midwives had a significantly higher risk of a 5-minute Apgar score of 0 and neonatal seizures or serious neurologic dysfunction. This risk was greater for nulliparous women. In addition, women who gave birth at home with a midwife attending were significantly more likely to have macrosomic infants (birthweight greater than 4,000 g); see Tables 5-1 and 5-2. The distinction between causality and correlation should be noted here. Birth settings would not cause higher birthweight, but home births could be correlated with higher birthweight for multiple reasons, such as waiting until the natural onset of labor rather than undergoing induction (see, e.g., Zhang et al., 2010).

9

The 2014 study by Grünebaum and colleagues used the CDC-linked birth and infant death dataset from 2006–2009 for early and total neonatal mortality for nearly 14 million singleton, vertex, and term births without congenital anomalies. This dataset included births attended by midwives and physicians in the hospital and midwives at home and in birth centers. The authors used midwife-attended home births as a proxy for planned home births. Compared with deliveries by hospital midwives, home births were more likely to be postterm, and mothers were more likely to have macrosomic infants. Midwife-attended home births also had a significantly higher total neonatal mortality risk relative to deliveries attended by midwives in the hospital, and the risk of neonatal mortality increased for postterm births and nulliparous women. Similar results were observed for early neonatal mortality. The excess total neonatal mortality for midwife-attended home births compared with midwife-attended hospital births was estimated at 0.93 per 1,000 births, and the excess early neonatal mortality at 0.79 per 1,000 births. In birth center births, excess total neonatal mortality was reported as .26 per 1,000 births, and excess early neonatal mortality as .32 per 1,000 births (Grünebaum, 2014).

10

The 2015 study by Grünebaum and colleagues used a national sample of about 12 million deliveries from 2010–2012 CDC-linked vital records to analyze the frequency of four perinatal risk factors—breech presentation, prior cesarean delivery, twins, and gestational age 41 weeks or longer—that were associated with planned midwife-attended home births in the United States and compare them with deliveries performed in the hospital by CNMs. (Home births attended by others were excluded; only planned home births attended by midwives were included.) Compared with CNM-attended hospital births, all four risk factors were significantly higher among midwife-attended planned home births, and three were significantly higher for planned home births attended by midwives not certified by the American Midwifery Certification Board.

11

The 2018 study by Bachilova and colleagues used CDC-linked vital records for 2011–2013 to conduct a 3-year retrospective cohort study of 71,704 planned home births in the United States. The authors found an overall early neonatal mortality rate of 1.5 per 1,000 planned home births, with significantly elevated risk in some subgroups. The risk of early neonatal death was significantly higher among nulliparous women (adjusted odds ratio [aOR] 2.71; 95% CI 1.71–4.31), women with previous cesarean births (aOR 2.62; 95% CI 1.25–5.52), nonvertex presentations (aOR 4.27; 95% CI 1.33–13.75), plural births (aOR 9.79; 95% CI 4.25–22.57), preterm births (34– <37 weeks gestation) (aOR 4.68; 95% CI 2.30–9.51), and births at ≥41 weeks gestation (aOR 1.76; 95% CI 1.09–2.84). The authors conclude that early neonatal deaths occur more commonly when certain risk factors are present and that more careful patient selection may reduce adverse neonatal outcomes among planned home births.

12

Wasden and colleagues (2017) used vital statistics data from New York City as their control group to identify the risk of HIE (hypoxic ischemic encephalopathy) compared with infants who received head cooling for HIE at a New York City institution. Demographics, obstetric information, location of birth, and intended location of birth were obtained from the vital records for both the cases and controls. A total of 69 infants underwent head cooling for HIE and were matched with 276 normal controls. After adjusting for pregnancy characteristics and mode of delivery, the odds of having an infant requiring treatment for HIE was 44 (95% CI 1.7–256.4) for out-of-hospital births compared with infants without HIE, regardless of intended place of birth. For those who did plan a home birth, the odds of having an infant with HIE were 21 (95% CI 1.7–256.4) compared with infants not requiring treatment for HIE. The authors conclude that out-of-hospital births were associated with increased odds of having an infant requiring treatment for HIE.

13

MacDorman and Declercq (2016 [ResQu: moderate, GRADE: poor]) examined trends in out-of-hospital births (n = 59,674) that occurred between the years of 2004 and 2014. Data for this study came from 47 states and Washington, DC, using birth certificate data that had been revised after 2003. Results showed that out-of-hospital births increased by 72 percent over the 10-year period. Compared with women who had hospital births, out-of-hospital births had lower prepregnancy obestity and higher rates of breastfeeding initiation and vaginal birth after cesarean (VBAC). Results were significant at the p <0.05 level.

14

Li and colleagues (2017 [GRADE: poor]) conducted a population-based cohort study with matched birth certificate data (1996–2013) and Medicaid claims and hospital discharge abstracts in South Carolina to evaluate the validity of reports of neonatal seizures in infants born at home or in birth centers and then transferred to the hospital (n = 1,233). Their results showed birth certificates were not reliable as a sole source for analyzing the prevalence of neonatal seizures

15

Stephenson-Famy and colleagues (2018 [ResQu: moderate, GRADE: poor) performed a retrospective cohort study using birth certificate data (2004 to 2011) of women who planned to give birth in a birth center (n = 7,118 planned birth center births). A total of 93 percent of women gave birth at the birth center, and 7 percent gave birth in a hospital setting. Nulliparity was the most significant risk factor for hospital transfer (aOR 7.2; CI 5.3–9.8), followed by maternal age >40 (aOR 3.7; CI 2.1–6.7) and inadequate prenatal care (aOR 3.7; CI 2.7–5.0).

16

In a descriptive study, Stapleton and colleagues (2013 [ResQu: high, GRADE: poor]) evaluated outcomes of care for more than 15,500 women eligible for birth center admission in labor using the AABC’s data registry (called the Uniform Data Set, or UDS, at the time, now the PDR). The authors found a spontaneous vaginal birth rate of 93 percent and a cesarean birth rate of 6 percent; the remaining births were assisted vaginal births (Stapleton et al., 2013). The intrapartum transfer rate after admission to a birth center was 12.4 percent, and of those, 0.9 percent were considered emergency transfers. Intrapartum fetal deaths were 0.47 per 1,000, and neonatal deaths, excluding anomalies, were 0.40 per 1,000.

17

Jolles and colleagues (2017) analyzed data from Medicaid enrollees whose birth outcomes were recorded in the PDR. This study compared cesarean section rates between similar cohorts of healthy women who chose elective hospitalization versus a birth center birth. The authors found a significantly increased risk of cesarean section among planned hospital births. Cesarean rates for low-risk women admitted to a birth center were 2.7 percent, compared with 9 percent for low-risk women admitted to a hospital.

18

Using PDR data (Thornton et al., 2017 [ResQu: high, GRADE: poor]), exclusion criteria were used to form low-risk groups admitted to birth centers (n = 8,776) and those that chose hospital admission (n = 2,527). Comparing midwife-led birth center and hospital groups with midwifery care, the authors found a nonsignificant difference in cesarean birth rates (4.14% for birth centers vs. 4.99% for hospitals), a significant difference in breastfeeding initiation rates (94.5% for birth centers vs. 72.8% for hospitals), and no difference in the neonatal outcome composite (0.44% for both groups).

19

Alliman and colleagues (2019) found that Medicaid beneficiaries (n = 6,424) enrolled in AABC Strong Start birth center sites experienced preterm birth rates of 4.4 percent and low-birthweight rates of 3.7 percent, compared with CDC birth certificate rates (n = 3,945,875) of 9.9 percent and 8.2 percent, respectively. The total cesarean rate was 12.3 percent, with a primary cesarean rate of 8.7 percent for births at Strong Start birth centers. Breastfeeding initiation was 92.9 percent compared with a national rate of 83.1 percent. In the birth center arm of Strong Start, eligible women participated in birth center prenatal care, and experienced these improved outcomes even if they elected hospital delivery. (Refer to Box 4-1).

20

The study compares 529,688 low-risk women with uncomplicated pregnancies who intended to have midwife-led care at the onset of labor (n = 321,307 planned home and n = 163,261 hospital births) (de Jonge et al., 2009 [ResQu: high]).

21

The study reports a retrospective analysis of intention-to-treat and perfect guidelines approaches (n = 679,952 low-risk women) (van der Kooy et al., 2011). The perfect guideline approach “includes the subset of women within the natural prospective approach population who in retrospect were compliant with the guidelines, which define low risk at the onset of labor and therefore are allowed to choose between a home or hospital birth under supervision of a midwife” (van der Kooy et al., 2011, p. 1039).

22

The study compares low-risk women planning midwife-led care for home versus hospital births (n = 466,112 planned home births and n = 276,958 planned hospital births) (de Jonge et al., 2015).

23

The study retrospectively analyzes national perinatal registry and maternal morbidity data, finding no ignificant differences in severe maternal morbidity (admission to intensive care unit [ICU], or hemolysis, levatenzymes, low platelet count [HELLP] syndrome) between home births (n = 92,333) and hospital births (n = 54,419) for low-risk, term, singleton pregnancies.

24

Some confusion results from the fact that FMUs and AMUs are often reported together as “midwifery units,” and in places where midwifery units exist, the percentage of women choosing them as their birth setting varies from 4 percent to 31 percent across England.

Copyright 2020 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK555483
Contents

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (3.8M)

In this Page

Other titles in this collection

Recent Activity

ClearTurn OffTurn On

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...