2

Context and Background

Childbirth and birth setting trends in the United States have changed significantly over the past century. This chapter summarizes the Panel 1 workshop presentations that focused on demographic and health trends in childbirth in the United States; birth setting trends (i.e., who is giving birth where); and the essential role of U.S. birth certificate data in analyzing these trends. See Box 2-1 for a summary of key points made by individual speakers. This panel was moderated by Sherin Devaskar, M.D., University of California, Los Angeles. A summary of the panelists’ discussion with the audience is included at the end of Chapter 3.

HISTORICAL AND RECENT TRENDS IN CHILDBIRTH IN THE UNITED STATES1

Brady Hamilton described several key demographic and health trends in childbirth in the United States, both recent and historical. All of the data he described were based on information obtained from the birth certificates filed in the United States for each year.2

______________________________________

1This section summarizes information presented by Brady Hamilton, Ph.D., National Center for Health Statistics (NCHS), Reproductive Statistics Branch, Washington, DC.

2Data obtained from U.S. birth certificates are compiled in the National Vital Statistics System, a data-sharing system maintained by the Centers for Disease Control and Prevention’s (CDC’s) NCHS.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 7
2 Context and Background C hildbirth and birth setting trends in the United States have changed significantly over the past century. This chapter summarizes the Panel 1 workshop presentations that focused on demographic and health trends in childbirth in the United States; birth setting trends (i.e., who is giving birth where); and the essential role of U.S. birth certificate data in analyzing these trends. See Box 2-1 for a summary of key points made by individual speakers. This panel was moderated by Sherin Devaskar, M.D., University of California, Los Angeles. A summary of the panelists’ discus- sion with the audience is included at the end of Chapter 3. HISTORICAL AND RECENT TRENDS IN CHILDBIRTH IN THE UNITED STATES1 Brady Hamilton described several key demographic and health trends in childbirth in the United States, both recent and historical. All of the data he described were based on information obtained from the birth certificates filed in the United States for each year.2 1  his section summarizes information presented by Brady Hamilton, Ph.D., National Center T for Health Statistics (NCHS), Reproductive Statistics Branch, Washington, DC. 2  ata obtained from U.S. birth certificates are compiled in the National Vital Statistics D System, a data-sharing system maintained by the Centers for Disease Control and Prevention’s (CDC’s) NCHS. 7

OCR for page 7
8 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS BOX 2-1 Context and Background: Key Points Made by Individual Speakers •  rady Hamilton noted that demographic trends among pregnant women in the B United States have changed over the past few decades. For example, while the majority of births are to non-Hispanic white women, the number and per- centage of births to groups other than non-Hispanic white women has been increasing. •  amilton presented some of the changes in health trends in the United States H over the past few decades, such as the increasing rate of Cesarean deliveries between 1996 and 2009. However, the rate of Cesarean deliveries appears to have abated somewhat in the past few years. •  arian MacDorman emphasized that birth setting trends have been chang- M ing as well. Most notably, the percent of out-of-hospital births has recently increased, by 36 percent since 2004, but still with only 1.2 percent of all births in the United States occurring outside of hospital settings. The increase in out-of-hospital birth rate is occurring much more quickly for non-Hispanic white women. •  acDorman pointed out that risk factors associated with different birth settings M have also been changing over time, with women in a home or birth center set- ting much less likely than women in hospital settings to deliver preterm and low-birth-weight infants. This trend suggests to MacDorman that selection of low-risk women as candidates for home and birth center births has improved over time. •  ll of the trends described by Hamilton and MacDorman and summarized in A this chapter are based on U.S. birth certificate data. In their opinion, U.S. birth certificate data are vital to gaining a better understanding of demographic and health trends among pregnant women and of birth setting trends. •  igel Paneth remarked that he was more impressed by how little things have N changed over the past few decades than by how much they have changed. The most notable changes since 1982, in his opinion, are decreased birth rates but steady fertility rates, the older age of most mothers, a shift in the birth popula- tion (decrease in the percent of births among non-Hispanic white women), and increased interventions (especially Cesarean sections). Demographic Trends in Childbirth The number of births in the United States has been generally rising over the last 9 decades, from 2.95 million in 1920 to 3.95 million in 2011 (see Figure 2-1). This overall increase has been punctuated by several periods of decline, including in the 1920s through the early 1930s, the 1960s through the early 1970s, the early 1990s, and over the past few years. The general rising trend is a product of the increasing size of the U.S. female population of reproductive age, changes in the composition of the reproductive age population, and changing fertility patterns.

OCR for page 7
CONTEXT AND BACKGROUND 9 Year FIGURE 2-1  The number of births per year in the United States, 1920-2011, final 1920-2010 and preliminary 2011. NOTES: Beginning with 1959, trend lines are based on registered live births; trend lines for 1920-1958 are based on live births adjusted for underregistration. SOURCE: Hamilton et al., 2012. Figure 2-1 With respect to distribution of births by population group, in 2011 Bitmapped the majority of births were to non-Hispanic white women (2,150,926), followed by Hispanic women (912,290) and non-Hispanic black women (583,079). These three groups are also the largest race and Hispanic-origin groups in the United States by population size. Births to non-Hispanic American Indian or Alaska Native and to non-Hispanic Asian or Pacific Islander accounted for about 45,000 and 250,000 births, respectively, in 2011. The number and percentage of births to groups other than non- Hispanic whites has been increasing over the past few decades, as has the number and percentage of births to parents of different races. In 2010, slightly more than 2 percent of U.S. births were to women who reported more than one race (i.e., multiracial mothers). As Hamilton noted, the number of births is a product of the size of the population, specifically the number of females of reproductive age, as well as fertility patterns. One way to assess fertility patterns is with fertility measures, such as total fertility rate, which estimates the number of births that a group of 1,000 women would have over their lifetimes based on the birth rates by age of mother in a given year. Fertility rate can also be ex- pressed as the expected number of births per woman. Generally, the trend

OCR for page 7
10 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS in total fertility rate (not shown) has followed the trend in the number of births. However, the rate has been fairly level over the past three decades, ranging between 1.8 and 2.1 births per woman from 1980 to 2011. In 2011, the average number of expected births per woman differed markedly by race and ethnicity, with Hispanic women having the highest rate (2.2), and American Indian or Alaska Native women having the lowest rate (1.4). Although overall fertility patterns have remained fairly stable over the past three decades, there have been some marked shifts in birth rates by age of mother. The most noticeable shift is a decline in rates for women under age 30 and a rise in rates for women over age 30. In 2011, 40 percent of U.S. births were to women age 30 and over, up from 20 percent in 1980. The age of first-time mothers has been increasing as well, from 21.4 in 1970 to 25.4 in 2010. Health Trends in Childbirth Data from U.S. birth certificates can be used to assess not just demo- graphic trends, but also health trends in childbirth. Hamilton described several of these trends, starting with the rate of Cesarean delivery, which increased from 1996 to 2009 (see Figure 2-2). However, the trend appears 35 33 Non-Hispanic Black All races and origins 31 29 Non-Hispanic White Percent 27 25 Hispanic 23 21 19 0 17 1996 2001 2006 2011 Year FIGURE 2-2  Cesarean delivery rates in the United States, 1996-2011, by selected race and Hispanic origin (final 1996-2010, preliminary 2011). NOTE: Singleton births only. SOURCE: Hamilton et al., 2012.

OCR for page 7
CONTEXT AND BACKGROUND 11 to have abated, with the rate declining slightly from 2009 to 2010 and remaining unchanged from 2010 to 2011. Non-Hispanic black women are more likely than other women to have a Cesarean delivery. While Cesarean delivery rates increased for all age groups, Cesarean rates by age of mother have decreased slightly over the past couple of years but still remain well above what they were in 1996. In addition to Cesarean delivery rates, another key health trend that can be assessed using U.S. birth certificate data is preterm births. Since 2006, preterm birth rates have declined significantly for infants in each of the three largest ethnic groups (i.e., Hispanic, non-Hispanic black, and non- Hispanic white). Despite the declines for all groups, disparities persist. In 2011, the preterm birth rate for non-Hispanic black infants was 60 percent higher than for non-Hispanic white infants. Another significant health shift is in birth by gestational age. From 1990 to 2006, the overall distribution of gestational age shifted to earlier gestations, with the proportions of birth at 36 and 39 completed weeks in- creasing. From 2006 to 2011, gestational ages shifted to longer gestations, with more births occurring at 39 or more weeks and fewer births occurring at less than 39 weeks. In more specific terms, the percentage of births at 37, 38, and 39 completed weeks of gestation increased from the 1990s through the mid- 2000s, while the percentage of births at 40 completed weeks of gestation decreased. However, starting in 2006 and continuing to 2011, the percent- age of births at 37 and 38 weeks decreased while the percentage of births at 39 weeks increased rapidly and the percentage of births at 40 weeks increased slightly. With respect to trends in low birth weight, the percentage of infants born weighing less than 2,500 grams increased by more than 20 percent from the mid-1980s through 2006, but has declined slightly since then (down by 2 percent from 2006 to 2011). As with the preterm birth rate, low birth weight varies considerably by race and ethnicity. The rate for non-Hispanic black infants is the highest (11.46 in 2010) and more than two times higher than the lowest rate, which is for non-Hispanic white infants (5.22 in 2010). Trends in weight gain during pregnancy have shifted as well, with the percentage of women gaining more than 40 pounds (i.e., more than the recommended amount of weight gain during pregnancy) increasing by more than 50 percent between 1990 and 2010 and the percentage of women gaining less than 16 pounds (i.e., less than the recommended amount of weight gain during pregnancy) nearly doubling over the same time period. Gestational diabetes and gestational hypertension rates both vary sig- nificantly by race and ethnicity of mother. In 2010, the highest rate for gestational diabetes was among non-Hispanic Asian women (7.9 percent),

OCR for page 7
12 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS compared to 4.1 percent among non-Hispanic white women, 3.6 per- cent among non-Hispanic black women, and 4.4 percent among Hispanic women. Patterns are quite different for gestational hypertension. In 2010, non-Hispanic black women were more likely to have gestational hyperten- sion (5.3 percent) than any other group (non-Hispanic white women, 4.6 percent; Hispanic women, 3.1 percent; and non-Hispanic Asian women, 2.3 percent). Finally, both the number and the rate of twin births have been rising until very recently. The number of twin deliveries doubled from 68,339 to 137,217 between 1980 and 2009. In 2010, however, for the first time in several de- cades, both the number and the rate of twins declined slightly. The rate of twin births increased 76 percent from 1980 to 2009, rising by nearly 3 per- cent a year in the 1990s but by less than 1 percent per year in the mid-2000s. The rise of triple and higher-order births was even more dramatic, with the rate increasing by more than 400 percent during the same time period and peaking in 1998. Since peaking in 1998, both the rate and the number of triple and higher-order births has declined, with the lowest number since 1995 recorded in 2010 (5,503). Summary In summary, there have been substantial increases in the number and percentage of births to groups other than non-Hispanic white women, particularly among Hispanic women, and to women age 30 and over. The percent of women gaining more than 40 pounds during pregnancy has also increased. While Cesarean delivery rates have increased over the past several decades, the rates have decreased slightly in recent years. Similarly, while low-birth-weight rates have increased over the past several decades, the rates have decreased slightly in recent years. But marked disparities in both rates among racial and ethnic groups persist. There has been a long and sustained decrease in preterm birth rates, although, again, with marked disparity. Finally, twin birth rates appear to have stabilized, while triple and higher-order birth rates are clearly declining. In conclusion, Hamilton noted the several new items that will be added to the U.S. birth data files in 2013 (for the 2009, 2010, and 2011 data years): body mass index, tobacco use in 3 months prior to pregnancy (i.e., whether the mother quit prior to pregnancy), whether the mother received food through the Special Supplemental Nutrition Program for Women, Infants, and Children during pregnancy, whether pregnancy resulted from infertility therapy, infections present during pregnancy (e.g., Chlamydia), source of payment for the delivery (e.g., Medicaid), interval since last live birth, maternal morbidities (e.g., ruptured uterus), and infant breastfeeding.

OCR for page 7
CONTEXT AND BACKGROUND 13 WHO ARE THE WOMEN GIVING BIRTH IN VARIOUS SETTINGS?3 Birthing patterns in the United States have changed significantly over the past century (MacDorman et al., 2012). In her presentation, Marian MacDorman described how. As with the demographic and health trends observed by Brady Hamilton, all of the patterns observed by MacDorman were based on data from all birth certificates filed in the United States each year and compiled by the National Vital Statistics System. Importantly, U.S. birth certificate data show only the number of births actually delivered in each location (e.g., home, birthing center, hospital), not where women intended to deliver (e.g., women who planned to deliver at home but were transported to a hospital). Place of Birth In 1900, nearly all U.S. births occurred at home. By 1940, only 44 percent of U.S. births occurred outside a hospital. By 1969, only 1 percent of U.S. births occurred outside a hospital. The percent of out-of-hospital births has remained around 1 percent for several decades. In 1990 there were about 47,000 out-of-hospital births in the United States, a number that gradually declined to a low of about 35,500 in 2004. Recently, the percent of out-of-hospital births has increased—by 36 percent since 2004—with just over 47,000 U.S. babies born outside of a hospital in 2010, representing 1.2 percent of the U.S. births (see Figure 2-3). Despite this substantial increase in out-of-hospital births, they still represent “a drop in the bucket” compared to the nearly 4 million in-hospital births in the United States each year. Not until the 1989 revisions of the birth certificate was it possible to distinguish, for the first time, between types of out-of-hospital births, that is, whether the births occurred in homes or in birthing centers. As with total out-of-hospital births, both home and birthing center births declined gradu- ally from 1990 to 2004 and then increased rapidly from 2004 to 2010. Home births increased by 41 percent from 2004 to 2010, with 10 percent of the increase occurring in the last year; birthing center births increased by 44 percent over the same time period, with 14 percent of the increase oc- curring in the last year. In 2010, there were 31,500 home births and 13,166 birthing center births in the United States. Among out-of-hospital births, 67 percent are home births, 28 percent occur in birthing centers, and 5 percent are identified as “other” (which has an unclear meaning). U.S. birth certificate data indicate that 29 percent of out-of-hospital 3  his T section summarizes information presented by Marian MacDorman, Ph.D., NCHS, Reproductive Statistics Branch, Washington, DC.

OCR for page 7
14 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS 1.6 1.4 47,028 46,956 (1.13%) (1.18%) 1.2 35,587 1.0 (0.87%) Percent 0.8 0.6 0.4 0.2 0.0 1990 1995 2000 2005 2010 Year FIGURE 2-3  Number and percent of out-of-hospital births in the United States, 1990-2010. SOURCE: CDC, 2013. Figure 2-3 births are delivered by certified nurse midwives (CNMs) or certified mid- wives (CMs), 41 percent by other midwives (including certified professional midwives, licensed midwives, and direct entry midwives), 6 percent by physicians, and 24 percent by “other” (e.g., emergency responders, family members). Overall trends in out-of-hospital births disguise large variation by race and ethnicity. Even from 1990 to 2004, when overall out-of-hospital births were declining, out-of-hospital births for non-Hispanic white women increased by 5 percent (see Figure 2-4). Out-of-hospital births for all other race and ethnic groups declined during that period. More recently, from 2004 to 2010, out-of-hospital births increased by 46 percent for non-Hispanic white women, from 1.2 percent to 1.75 per- cent of births. In 2010, for non-Hispanic white women, 1 out of every 57 births in the United States was an out-of-hospital birth. From 2004 to 2010, out-of-hospital births increased more slowly for other racial and ethnic groups such that, by 2010, the percent of out-of-hospital births was four times higher for non-Hispanic white women than for other racial and ethnic groups. About 90 percent of the total increase in out-of-hospital births from 2004 to 2010 was due to an increase among non-Hispanic white women.

OCR for page 7
CONTEXT AND BACKGROUND 15 2.0 1.5 Hispanic Non-Hispanic white Percent 1.0 Non-Hispanic black 0.5 API American Indian 0.0 1990 1995 2000 2005 2010 Year FIGURE 2-4  Percent of out-of-hospital births in the United States by maternal race/ ethnicity, 1990-2010. NOTES: Non-Hispanic white, non-Hispanic black, and Hispanic data exclude New Hampshire in 1990-1992 and Oklahoma in 1990, as these states did not report Hispanic origin on their birth certificates for those years. API denotes Asian or Pacific Islander. SOURCE: CDC, 2013. Characteristics and Risk Factors Associated with Birth Setting MacDorman explained, “We know that only low-risk women should deliver outside of a hospital, but the precise definition of low risk remains controversial.” In an effort to describe the risk status of home versus birth center versus hospital births, MacDorman examined trends in various characteristics and risk factors associated with the different birth settings. These included maternal age, parity, smoking history, marriage status, and select medical risk factors. Based on 2010 data, with respect to maternal age, about 9 percent of hospital births are to teen mothers, compared to 2 to 3 percent of home and birthing center births. At the other end of the age spectrum, about 14 percent of hospital births are to women age 35 and older, compared to 15 percent of birth center and 21 percent of home births. The majority of births in all settings are to women between 20 and 34 years of age (76.5

OCR for page 7
16 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS percent of home births, 82.7 percent of birthing center births, 76.2 percent of hospital births). Women having a home or birth center birth are less likely to be hav- ing their first birth and are more likely to have had three or more previous children. More specifically, 22.5 percent of women having a home birth are having their first child, 45.9 percent their second or third child, and 31.6 percent their fourth or greater child; 36.1 percent of women having a birth center birth are having their first child, 44.3 percent their second or third child, and 19.5 percent their fourth or greater child; and 40.6 percent of women having a hospital birth are having their first child, 48.0 percent their second or third child, and 11.4 percent their fourth or more child. With respect to smoking, in 2010, 2 to 3 percent of women giving birth at home or in a birth center identify as smokers, compared to 9.3 percent of women giving birth in a hospital. With respect to marriage status, also in 2010, 14 to 15 percent of women giving birth at home or in a birth center are unmarried, compared to 41.1 percent of women with a hospital birth. MacDorman reported 2010 trends for five select medical risk factors: preterm birth, low birth weight, multiple births, diabetes, and hyperten- sion. Women with a home or birth center birth are much less likely to deliver preterm (5.4 percent for home births and 2.2 percent for birth center births, compared to 12.1 percent for hospital births) and to deliver low-birth-weight infants (3.9 percent for home births and 1.1 percent for birth center births, compared to 8.2 percent for hospital births). They are also much less likely to have multiple births (1.0 percent for home births and 0.3 percent for birth center births, compared to 3.5 percent for hospital births). The prevalences of diabetes and hypertension are also much lower among women delivering at home (1.1 percent for diabetes, 0.3 percent for hypertension) and in birth centers (1.1 for diabetes, 0.1 percent for hypertension) compared to hospitals (5.1 percent for diabetes, 1.4 percent for hypertension). The lower rates of medical risk factors for out-of-hospital births sug- gest that appropriate risk selection of low-risk women as candidates for out-of-hospital births is occurring. However, the possibility that these dif- ferences reflect differences in risk factor reporting between out-of-hospital and hospital settings cannot be ruled out. Another way to view risk factors is to examine changes over time. A comparison of 2004 and 2010 data reveals a decline in percent of home births to teens (3.9 percent in 2004, 2.2 percent in 2010) and to women age 35 and older (22.0 percent in 2004, 21.3 percent in 2010). The percent of home births to unmarried women also declined during this same time period (from 20.4 percent in 2004 to 14.9 percent in 2010), as did the percent of home births with live birth order of four or more (33.1 percent in 2004, compared to 31.6 percent in 2010). These same risk factors show

OCR for page 7
CONTEXT AND BACKGROUND 17 similar declines for birth center births, with 4.5 percent of birth center births to teens in 2004 dropping to 2.5 percent in 2010, 15.8 percent of birth center births to women age 35 and older in 2004 dropping to 14.8 percent in 2010, 16.0 percent of birth center births to unmarried women in 2004 dropping to 14.2 in 2010, and 22.8 percent of birth center births with live birth order of four or more in 2004 dropping to 19.5 percent in 2010. With respect to the selected medical risk factors described previously, a comparison between 2004 and 2010 indicates declines for the percent of home births born preterm (7.1 percent in 2004, compared to 5.4 percent in 2010) and for the percent of home births with low birth weight (5.3 percent in 2004, compared to 3.9 percent in 2010), but not much change for the percent of home births that are multiple births (1.1 percent in 2004, 1.0 percent in 2010). For birth center births, a similar decline occurred with percent of births born preterm (2.7 percent in 2004, compared to 2.2 percent), but not for low birth weight (1.0 percent in 2004, 1.1 percent in 2010). As with home births, the percent of birth center births with mul- tiple births also remained more or less the same (0.2 percent in 2004, 0.3 percent in 2010). Changes in reporting of some items on the U.S. birth certificate, such as smoking, make it difficult to examine trends over time. Meanwhile, the observed declines in percent of births born preterm or with low birth weight suggest, again, that selection of low-risk women as candidates for home and birth center births has improved over time. Planning Status of Home Birth Planning status of home birth is considered an important indicator of risk for home births. Studies suggest that most home births are planned home births and that unplanned home births usually result from an emer- gency situation or a woman not being able to get to the hospital in time. According to MacDorman, unplanned home births may be at a higher risk for poor birth outcomes, with the births taking place in environments unprepared for delivery. In 2010, planning status of home birth was reported in 31 states and in the District of Columbia, representing 60 percent of U.S. births. Although the data are not completely representative of the U.S. population, they can suggest national trends. In 2010, 88 percent of home births were planned and 12 percent were unplanned. However, the percent of home births that were planned versus unplanned varied by care provider. Among home births delivered by physicians, very few were planned (36 percent), which MacDorman stated is “in keeping with the preference of most physicians to attend births in hospitals.” In contrast, 98 to 99 percent of home births delivered by midwives (CNM/CM and other midwives) were planned.

OCR for page 7
18 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS Surprisingly, in MacDorman’s view, 70 percent of home births delivered by what was identified as “other” on the birth certificate were reported to be planned home births. MacDorman proposed the 70 percent figure suggests that either some women are planning to deliver at home without a trained care provider or, more likely, that fathers are signing birth certificates in states where some types of midwives may not be licensed. Planning status of home births varies considerably by race and ethnic- ity of the mother. Among non-Hispanic white women, 93 percent of home births in 2010 were planned. This is in contrast to non-Hispanic black women, among whom only 33 percent of home births were planned. For Hispanic and Native American women, 67 to 68 percent of home births in 2010 were planned; for Asian or Pacific Islander women, 75 percent of home births in 2010 were planned. Geographic Differences Home birth trends vary geographically, with the percent of home births being higher in the Pacific Northwest and lowest in the South (see Figure 2-5). In 2010, more than 2 percent of total births in Oregon, Montana, and Vermont occurred at home, and between 1.5 and 2 percent of total births in Alaska, Idaho, Maine, Pennsylvania, Utah, Washington, and Wisconsin oc- curred at home. For 16 states, at least 1 percent of births occurred at home. Another way to view the geographic variation in home birth trends is to examine how the 41 percent increase in home births that occurred nation- wide between 2004 and 2010 played out at a state level (see Figure 2-6). Overall, 35 states experienced statistically significant increases in the per- cent of births that occurred at home, including 19 states where the percent of home births increased by 41 percent or more. The 2004-2010 increase was widespread and involved states from every region of the country. Ver- mont was the only state that showed a significant decline in the percent of home births between 2004 and 2010; despite the decline, Vermont remains one of the highest states for home births, with 2 percent of Vermont babies born at home in 2010. Birth center births show similar geographic variation, with the highest percent of birth center births in Alaska (4.4 percent in 2010) and Idaho (2.1 percent) and with four additional states having 1 percent or more of their births occurring in birth centers (Montana, Oregon, Pennsylvania, Washington). In contrast, the percent of birth center births was less than 0.1 percent in 23 states. Six states had increases of 300 or more births occurring in birth centers between 2004 and 2010 (Florida, Oregon, Penn- sylvania, South Carolina, Texas, and Washington), together accounting for more than three-fourths of the increase in birth center births in the United States during that time period. Some of the geographic variation in birth

OCR for page 7
CONTEXT AND BACKGROUND 19 WA MT ME ND OR MN VT NH ID WI NY MA SD WY MI CT RI IA PA NE NJ NV OH IL IN DE UT CO WV MD DC VA CA KS MO KY NC TN AZ OK NM AR SC MS AL GA TX LA FL 2.00% or more AK 1.50-1.99% 1.00-1.49% HI 0.50-0.99% <0.50% U.S. average=0.79% FIGURE 2-5  Percentage of home births in the United States by state, 2010. SOURCE: CDC, 2013. WA MT Figure 2-5 ND ME OR MN VT NH ID WI NY MA SD WY MI CT RI IA PA NE NJ NV OH OH IL IN DE UT CO WV MD DC VA CA KS MO KY NC TN AZ OK NM AR SC MS AL GA TX LA FL AK Significant increase 41% or more HI Significant increase <41% No significant change Significant decrease Data not available U.S. average=41% increase FIGURE 2-6  Change in percentage of home births in the United States by state, 2004-2010. SOURCE: CDC, 2013. Figure 2-6

OCR for page 7
20 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS BOX 2-2 Summary of Key Trends: Who Is Giving Birth Where? •  fter a gradual decline from 1990 to 2004, the percent of total out-of-hospital, A home, and birthing center births increased rapidly from 2004 to 2010. The increase was widespread and involved states from every region of the country. • In 2010, 1.2 percent of U.S. births were out-of-hospital births. •  inety percent of the increase in out-of-hospital births from 2004 to 2010 oc- N curred among non-Hispanic white women. •  n 2010, 1 in 57 births to non-Hispanic white women were out-of-hospital I births. •  n 2010, 88 percent of home births in 31 states and in the District of Columbia I were planned. Among non-Hispanic white women, 93 percent were planned. In contrast, only 33 percent of home births were planned for non-Hispanic black women. •  n 2010, out-of-hospital births were more prevalent (>2.5 percent of births) I in the Pacific Northwest, Alaska, and Pennsylvania and least prevalent in the South. •  n 2010, home and birthing center births had a lower risk profile than hospital I births for a variety of risk factors, including teen births, nonmarital births, pre- term or low-birth-weight babies, multiple births, maternal smoking, hyperten- sion, and diabetes. •  onversely, out-of-hospital births had higher percentages of older mothers and C mothers having a fourth or higher-order birth, compared to hospital births. •  he risk profile for out-of-hospital births improved from 2004 to 2010, suggest- T ing that appropriate risk selection of low-risk women is occurring and improving. However, the possibility that these differences reflect differences in risk factor reporting between out-of-hospital and hospital settings cannot be ruled out. center births is due to geographic variation in access to birth centers. As of January 2013, 13 states did not have freestanding birth centers listed with the American Association of Birth Centers. Box 2-2 provides a summary of key trends in childbirth. BIRTH SETTINGS: ANYTHING NEW SINCE ’82?4 Nigel Paneth offered some perspective on the analyses of vital data presented by Brady Hamilton and Marian MacDorman. As the only rep- resentative of the Institute of Medicine (IOM)/National Research Council (NRC) committee responsible for authoring the 1982 report Research 4  his T section summarizes information presented by Nigel Paneth, M.D., M.P.H., Michigan State University, East Lansing, Michigan.

OCR for page 7
CONTEXT AND BACKGROUND 21 ­ ssues in the Assessment of Birth Settings (IOM and NRC, 1982), Paneth I also offered some perspective on what has been learned—and what remains to be learned—since then. He opened by commenting on the significance of vital statistic data gathered from U.S. birth statistics, which all state vital registrars are re- quired to submit to the NCHS. He described the data as an “extraordinary resource,” one that must be maintained and sustained. Without such data, analyses such as those presented by Hamilton and MacDorman would not be possible. Yet, in Paneth’s opinion, the public, including the medi- cal public, is “woefully uneducated” about the value of vital data, with many people not even knowing what a birth certificate is. The amount of resources currently being dedicated to the maintenance and sustenance of U.S. birth certificate (and death certificate) data is inadequate, at both the state and national levels. He urged medical professionals to be more out- spoken in their support of the need for keeping good birth certificate data, as well as other vital data. Key Changes Since 1982 Paneth identified several key changes since 1982: • Decreased birth rates, but steady fertility rates.5 While fertility rates have shifted to higher age brackets, overall fertility rates have declined only 6.4 percent since 1982 (from 67.3 to 63.2 percent). • Mothers are older. Mothers age 20 to 24 years old were once the first-place age bracket, but dropped to second place in 1997 and third place in 2007. Today, 40 percent of mothers are over the age of 30, with a mean maternal age of 28. • There has been a shift in the birth population, with the percent of births among non-Hispanic white women dropping from 80 percent in 1982 to 54 percent in 2011. Paneth described this is a “substantial demographic shift.” • Births are characterized by increased interventions, with Cesarean delivery rates now more than 30 percent, compared to 5 percent in 1972, and with interventions occurring earlier during gestation (such that the 39th week has replaced the 40th week as the modal gestational week, the week with the largest number of births). There has also been a significant increase in multiple births, which are associated with infertility treatments. 5  irth rate is defined as the number of children born in a year as a proportion of the total B population. Fertility rate is defined as the number of children an average woman is likely to have during her reproductive years.

OCR for page 7
22 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS • New information exists that was not available in 1982. In 1982, one of the committee recommendations was related to the need to know who is born at a birthing center versus at home and whether the birth was planned as such. In 1989, the U.S. birth certificate was revised to distinguish between home and birthing center births. For 31 states and the District of Columbia, data have also been collected on planning status. Key Nonchanges Since 1982 Paneth expressed that he was more impressed by how little things have changed than by how much they have changed. He identified several key “nonchanges” since 1982: • The total number of births in the United States has remained fairly constant, around 3.5 million to 4 million per year, even though the population size of the country has increased 35 percent from 232 million (in 1982) to 313 million. This trend does not reflect a decreased fertility rate, as fertility rate has not decreased much, but rather a decrease in the fraction of the population that are women of reproductive age (as boomers age out of fertility). • The percentage of out-of-hospital births has remained relatively steady, around 1 percent, since 1969. However, since 2005, the percentage of out-of-hospital births among non-Hispanic white women has been increasing—to nearly 2 percent. Paneth described the choice of out-of-hospital births in a subset of the U.S. popula- tion as a “notable recent trend.” • The diversity of birth settings—and variation in data being col- lected on the different U.S. birth certificates circulating—continues to make it difficult to make generalizable statements. For example, “home” is anything someone defines as “home.” Home settings range from places with easy ambulance access to five-story walk- ups where it is difficult to transport via ambulance. It is very dif- ficult to know what “home” on a birth certificate means. A Closer Look at Out-of-Hospital Births Reviewing some of the data that MacDorman reported, Paneth said that he was surprised that about two-thirds of out-of-hospital births are home births and about one-third birth center births (and about 5 percent “other”). He expected the proportion of birth center births to be greater, given the increased number of birth centers nationwide (about 200 to 250, compared to about 150 in 1982). He found it “striking” that nearly 90

OCR for page 7
CONTEXT AND BACKGROUND 23 percent of home births are planned, but that only about 30 percent of non- Hispanic black women’s home births are planned. Paneth suggested the increased premature rates among home births probably reflect the proportion of home births that are unintended or un- planned. He reflected that, generally and very encouragingly, both home births and birth center births involve mothers at generally lower risk of adverse outcomes of pregnancy (i.e., women who are predominantly non- Hispanic white, older, of higher parity, married, and nonsmoking). Paneth concluded with an anecdotal story about a local birth center 4 miles from his home, in Okemos, Michigan. On their website, the birth center announced: “It with great sadness that we announce the closing of the birth center on September 30, 2012. We have been blessed to have attended over 700 births since 2003. It has been a great pleasure to work with our families and help them give birth in a safe, comfortable and sup- portive environment. Our belief in and support of women seeking natural childbirth in a setting of their choosing is unwavering. We are so sorry that our community will no longer have a birth center to serve those that want that option. Thank you so much for inviting us into your lives.” According to a local newspaper, the Lansing State Journal, the birth center closed amid a legal battle with a couple whose newborn son died following a breech vaginal delivery. Unsure of the actual legal status, Paneth observed the end result: “the closing of a birth center and an option for mothers.” Paneth urged that vital data (from both birth and death certificates) be used to monitor planned out-of-hospital births and compare planned out-of-hospital births with hospital births in terms of risk factors for prob- lem births (e.g., such as those presented by MacDorman), both neonatal and maternal mortality, and both neonatal morbidity (e.g., ventilation, transfer, and Apgar scores) and maternal morbidity (e.g., lacerations and transfusions). He also urged surveillance for sentinel events. State programs already exist that identify maternal deaths and, in some locations, infant deaths. These programs should be expanded to identify and investigate individual events that should not be found in planned out-of-hospital deliveries (e.g., breech vaginal deliveries). Finally, he urged an assessment of the cost-effectiveness, satisfaction, and benefits (e.g., rate of breastfeeding) of home and birth center deliveries among low-risk women compared to hospital deliveries among women at similar risk. The Challenge of Analyzing Mortality Data Following Paneth’s presentation, MacDorman commented on the chal- lenge of analyzing mortality data and comparing out-of-hospital versus

OCR for page 7
24 RESEARCH ISSUES IN THE ASSESSMENT OF BIRTH SETTINGS hospital birth neonatal mortality given that the planning status of hospital births is unknown. For example, there are no data on the number or per- centage of women who begin laboring at home but are transferred to hos- pitals because of complications. Compounding the challenge is variation in risk. Vital data do not provide enough information about risk. MacDorman expressed reluctance to analyze mortality data given the apples-to-oranges comparison involved when planning status and risk are unknown. In response, Paneth stressed that vital data are only a starting point, but said, “I would not go so far as to say that because we cannot really fully answer, you should not answer at all.” He suggested that analyzing actual deliveries provides at least a sense of what those rates are in the different settings and whether there are any unusual mortalities. While it may be difficult to select comparison groups for studies (e.g., women in different settings but with similar risk profiles), vital data on U.S. birth certificates nonetheless provide enough descriptive information such that unusual events that warrant further investigation, what Paneth calls “senti- nel events,” should stand out. He said, “Even without a denominator, their existence is of interest.”