If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Elective deliveries increased sharply until the mid-2000s in 2 U.S. states.
•
The rates of elective deliveries were higher for more educated and White women.
•
Trends in elective deliveries were similar across socioeconomic and racial groups.
Introduction
Cesarean section deliveries in the U.S. increased from 5% of births in 1970 to 32% in 2020. Little is known about trends in cesarean sections and inductions in low-risk pregnancies (i.e., those for which interventions would not be medically necessary). This study addresses the following questions: (1) what is the prevalence of elective deliveries at the population level?, (2) how has that changed over time?, and (3) to what extent do the rates of elective deliveries vary across the population?
Methods
We first documented long-term trends in cesarean sections in the U.S., California, and New Jersey. We then used linked birth and hospital discharge records and an algorithm based on Joint Commission guidelines to identify low-risk pregnancies and document trends in cesarean sections and inductions in low-risk pregnancies in California and New Jersey over a recent 2-decade period, overall and by maternal characteristics and gestational age.
Results
In low-risk pregnancies in California and New Jersey, rates of cesarean sections and inductions increased sharply from the early 1990s through the mid-2000s, peaked at 33% in California and 41% in New Jersey in 2007, and then declined somewhat, and the proportions of inductions that were followed by cesarean sections increased from fewer than 1 in 5 to about 1 in 4. More education, non-Hispanic White race/ethnicity, U.S.-born status, and non-Medicaid were associated with higher rates of interventions. Trends were similar across all socioeconomic groups, but differences have been narrowing in California. Among early-term (gestational age of 37–38 weeks) births in low-risk pregnancies, the rates of elective deliveries increased substantially in both states until the mid/late-2000s, peaked at about 35% in California and over 40% in New Jersey, and then decreased in both states to about 20%.
Conclusions
Given established health risks of nonmedically necessary cesarean sections, that a nontrivial share of induced deliveries in low-risk pregnancies result in cesarean sections, and that interventions in low-risk pregnancies have not substantially declined since their peak in the mid-2000s, the trends documented in this paper suggest that sustained, even increased, public health attention is needed to address the still-too-high rates of cesarean sections and inductions in the U.S.
The rates of cesarean section (C-section) delivery in the U.S. increased from 5% of births in 1970 to a peak of 33% of births in 2009 and then declined only minimally to 32% in 2020.
When medically necessary, C-sections can save lives; for example, increased rates of preterm C-sections in the U.S. between 1990 and 2004 were associated with perinatal survival, mostly because of decreases in stillbirths.
and pose surgical risks to mothers. Induced vaginal deliveries do not appear to be associated with such risks, but inductions of labor are often followed by C-section deliveries.
Since the mid-2000s, professional guidelines and public health campaigns have encouraged pregnant women and healthcare providers to wait until 39 weeks of gestation before intervening and appear to have attenuated the sharp increases in C-sections and inductions in the U.S.
However, the rates remain close to historic highs, and whereas increases in C-sections before 2000 were accompanied by reductions in infant and fetal mortality, increases afterward coincided with no such improvements,
Little is known about trends in C-sections and inductions in low-risk pregnancies (i.e., those for which interventions would not be medically necessary). In this study, we document trends in C-sections and inductions in low-risk pregnancies in 2 populous and diverse U.S. states, California (CA) and New Jersey (NJ). We address the following questions: (1) what is the prevalence of elective deliveries at the population level?, (2) how has that changed over time?, and (3) to what extent do rates of elective deliveries vary across the population?
METHODS
The study was approved by the IRB of Columbia University; the University of California, Los Angeles; Rutgers University; the CA Health and Human Services Agency; and the NJ Department of Health.
Study Sample
We used linked birth and hospital discharge records from the states of CA and NJ (linked birth and hospital discharge data are not available at the national level). For CA, we used CA's birth master files of all state-registered births in 1992–2012 (>11.3 million births), 96% of which were linked to maternal hospital discharge records by the CA Office of Statewide Health Planning and Development. For NJ, we linked records from NJ's Electronic Birth Certificate & Perinatal Database of all births in the state in 1997–2015 (>2.1 million births) to maternal hospital discharge records; the linkage rate was 93%. In both states, the linkages were performed by probabilistic matching using identifying variables, including the mother's name, dates of birth of the mother and child, and the child's birth weight. The birth records included the method of delivery, gestational age (GA), and demographic factors. The discharge records included maternal hospital discharge diagnoses and procedures codes from the ICD-9. The observation period for CA was 1992–2012 on the basis of the linked data made available to us by the CA Office of Statewide Health Planning and Development. The observation period for NJ was from 1997, the first full year of data available from the NJ Electronic Birth Certificate & Perinatal Database, through 2015. We had access to more recent years of data in NJ but found that the transition to ICD-10 codes after 2015 resulted in irregularities in trend lines that no doubt reflected changes in coding patterns rather than changes in obstetric practice. The 2003 revision of the U.S. Standard Certificate of Live Birth resulted in inconsistent measures of race/ethnicity across states between 2003 and 2016. CA transitioned to the 2003 revised form early in the observation period, and the data were harmonized across years. NJ began transitioning to the new form in July 2014; observations based on the revised form were dropped from all analyses (10% of births in 2014 and 50% of births in 2015).
Measures
Following recent studies that investigated the impacts of elective deliveries on neonatal and child outcomes,
we created low-risk samples for each state using Joint Commission (JC) guidelines for the perinatal core outcomes of decreasing the rate of elective deliveries and decreasing the cesarean delivery rate in nulliparous women with a term, singleton baby in a vertex position. From all births, we excluded those who had any of the ICD-9 codes on the JC lists for Conditions Possibly Justifying Elective Delivery Prior to 39 Weeks Gestation or Contraindications to Vaginal Delivery in the maternal delivery discharge record (Appendix Table 1, available online). The codes in the lists, which were compiled by a panel of perinatal experts and implemented in 2014 as part of an effort to monitor perinatal outcomes for the purposes of hospital accreditation, cover pregnancy and birth complications (including preeclampsia, diabetes, breech/malposition, fetal abnormalities, fetal distress/abnormal heart rate, and prolonged labor), multiple and preterm/post-term births, and most small-for-GA births.
We further limited the sample to first births and infants with GA between 37 and 40 completed weeks. The first restriction is consistent with the JC's focus on nulliparous women; having a CS for one delivery increases the likelihood that subsequent children will be delivered by CS, making first-time mothers a vital focus.
Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop.
GA was calculated as the number of completed weeks between the date of the first day of the mother's last menstrual period (LMP) and the child's birthday. We considered C-sections and inductions in low-risk pregnancies to be nonmedically necessary or elective interventions.
Statistical Analysis
Statistical analysis was performed in 2021–2022. First, we compiled and documented long-term trends in the rates of C-sections and inductions in the U.S., CA, and NJ from 1960‒1970s through 2020 from various published sources and our own data to place the trends for CA and NJ in context. Then, we documented trends over time in the following. (1) The rates of low-risk births (GA of 37–40 weeks and not having any of the conditions listed by the JC as justifying interventions) among all singleton first births in CA and NJ during the observation period of our study. (2) The rates of C-sections and inductions and maternal sociodemographic characteristics of births in the U.S., CA, and NJ for the observation period of our study for CA and NJ and for the year 2002 for the U.S; the year 2002 was the last year before the roll out of the 2003 revision of the U.S. Standard Certificate of Live Birth, which resulted in inconsistent measures of many characteristics across states between 2003 and 2016 and precluded us from providing national characteristics of births for the entire observation period. We present the characteristics both overall (for the U.S., CA, and NJ) and for the subsamples of low-risk births in CA and NJ. Sociodemographic characteristics were maternal education (less than high school, high school graduate or some college, college education or more), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian), nativity (foreign born, U.S. born), and Medicaid coverage for the birth (MediCal in CA). (3) Trends in rates of C-sections, vaginal deliveries after inductions, and either intervention in low-risk pregnancies in CA (n=1,654,888) and NJ (n=270,192). (4) Trends in any elective intervention in low-risk pregnancies in each state by maternal sociodemographic characteristics listed earlier as well as by GA (37–38, 39–40 weeks).
We also conducted (5) sensitivity analyses that did not consider certain ICD-9 codes with large increases over time in both states compared with the other JC risk factors as criteria that would eliminate cases from the low-risk samples and (6) supplementary analyses that used a clinical measure of GA to define the sample, which was available for NJ but not for CA, to compare trends in NJ using the LMP-based and clinical measures of GA to define the samples.
RESULTS
Over the past 5+ decades, C-section rates in CA were very similar to those in the U.S. (Appendix Figure 1, available online). The rates in NJ were very similar to those in the U.S. until the 1990s but afterward exceeded the national rates.
The proportion of singleton first births characterized as low risk using our JC-based algorithm was stable during the observation period (1992–2012) in CA and from 1997 to 2012 in NJ (Appendix Figure 2, available online). The rate decreased somewhat in NJ between 2012 and 2015, the last 3 years of our observation period for that state. Overall, there is no indication that the large increases in C-section rates in CA and NJ in the 1990s and 2000s reflected increases in obstetric risk factors that would justify C-sections according to the JC criteria (which are listed in Appendix Table 1, available online) because these have not changed in the aggregate among all singleton first births (including high risk) over time (as shown in Appendix Figure 2, available online), and the decrease in low-risk births in NJ in the later years occurred at a time when C-sections were declining, not increasing.
The overall rates of C-sections and inductions were higher in NJ (34% and 24%, respectively) than in CA (27% and 15%, respectively) and the U.S. (26% and 21%, respectively, in 2002) (Table 1).
Rates of interventions were also higher in NJ than in CA among low-risk births. Maternal characteristics differed across states; for example, mothers in NJ were more educated and less likely to be covered by Medicaid than those in CA and those in the U.S., and the percentages of mothers that were Hispanic and foreign born were higher in CA. However, within each state, maternal characteristics of the overall and low-risk samples were similar.
Table 1Obstetric Interventions and Maternal Characteristics of Births in the U.S. (2002), CA (1992–2012), and NJ (1997–2015)
Data for the U.S. are from the studies by Martin et al.2 and Curtin and colleagues15 Figures are for 2002 only because the slow roll out across states of the 2003 revision of the U.S. Standard Certificate of Live Birth resulted in inconsistent measures of education and race/ethnicity across states between 2003 and 2016, precluding us from providing national characteristics of births over the entire observation period.
Data for CA were compiled from our own data. CA transitioned to the 2003 revision of the U.S. Standard Certificate of Live Birth early in the observation period, and the data were harmonized across years.
Data for NJ were compiled from our own data. NJ began transitioning to the 2003 revision of the U.S. Standard Certificate of Live Birth in July 2014; observations based on the revised form were dropped (10% of births in 2014 and 50% of births in 2015). CA, California; C-section, cesarean section; NJ, New Jersey.
All births (2002)
All births 1992–2012
Low-risk births 1992–2012
All births 1997–2015
Low-risk births 1997–2015
Obstetric interventions
C-sections
26
27
13
34
19
Inductions
21
15
12
24
20
Education
≤12 years
52
58
53
41
38
13–15 years
21
21
21
21
21
≥16 years
26
22
25
38
41
Race/ethnicity
Non-Hispanic White
57
32
32
50
50
Asian
5
11
13
9
11
Hispanic
22
49
47
24
24
Non-Hispanic Black
14
6
6
16
14
Nativity
U.S. born
77
55
58
67
66
Foreign born
23
45
42
33
34
Insurance status
Medicaid
36
45
44
25
24
Non-Medicaid
64
55
56
75
76
n
4,021,726
11,300,257
1,654,888
1,869,635
270,192
Note: Figures are column percentages.
a Data for the U.S. are from the studies by Martin et al.
Figures are for 2002 only because the slow roll out across states of the 2003 revision of the U.S. Standard Certificate of Live Birth resulted in inconsistent measures of education and race/ethnicity across states between 2003 and 2016, precluding us from providing national characteristics of births over the entire observation period.
b Data for CA were compiled from our own data. CA transitioned to the 2003 revision of the U.S. Standard Certificate of Live Birth early in the observation period, and the data were harmonized across years.
c Data for NJ were compiled from our own data. NJ began transitioning to the 2003 revision of the U.S. Standard Certificate of Live Birth in July 2014; observations based on the revised form were dropped (10% of births in 2014 and 50% of births in 2015).CA, California; C-section, cesarean section; NJ, New Jersey.
The rates of elective deliveries (i.e., interventions in low-risk pregnancies) showed trends similar to those for C-sections overall (from Appendix Figure 1, available online), almost doubling in both states from the beginning of the observation period until 2007, from 17% to 33% in CA and from 22% to 41% in NJ (Figure 1). Increases in inductions and C-sections contributed equally to the increase in elective deliveries in CA. In NJ, the rate of C-sections increased considerably more than that of inductions.
Figure 1Percentage of low-risk births with inductions and C-sections, California and New Jersey, 1990s through 2010s.
The trends shown in Figure 1 reflect the final delivery method, not the initially attempted delivery method. Throughout the observation period, a nontrivial proportion of inductions in low-risk pregnancies ended in C-section deliveries (which could have been because of failure to progress, delivery complications after induction, or decisions to move on to C-section after brief attempts at induction), and the proportion increased over time, from <1 in 5 to about 1 in 4 (Figure 2). In fact, most of the elective deliveries in both states involved induction of labor.
Figure 2Percentages of low-risk births with C-sections after inductions, CA and NJ.
In both states, maternal socioeconomic advantage (more education, non-Hispanic White race/ethnicity, U.S.-born status, and non-Medicaid status) was associated with higher rates of elective deliveries, but the trends in elective deliveries over time were similar across all groups (Figure 3). In CA, most socioeconomic differences have decreased since the mid-2000s, when professional guidelines and public health campaigns started encouraging pregnant women and healthcare providers to wait until 39 weeks to intervene. For example, at the beginning of the observation period, the rate of elective deliveries was almost 50% higher among college-educated women than among women with less than a high school education, but by the end of the period, the rates of elective deliveries were only about 10% higher among the college-educated group. Racial/ethnic differences likewise narrowed during this period, and the nativity and poverty differences were virtually eliminated by the end of the period. Overall, the convergence across SES in CA after the mid-2000s reflected larger declines in the rates of elective deliveries among more economically advantaged groups. Socioeconomic differences in the rates of elective deliveries have narrowed much less in NJ.
Figure 3Percentages of low-risk births that were elective (C-sections or inductions) by maternal characteristics, California and New Jersey.
Trends in interventions at early term (GA of 37–38 weeks) were similar to those at full term (GA of 39–40 weeks) until the mid-2000s, when guidelines to delay until 39 weeks when possible became widespread; after that inflection point, interventions at early term declined substantially but remained at approximately 20% of all elective deliveries in both states.
Sensitivity analyses that reclassified fetal distress (ICD-9 code 65631) and abnormal heart rate (ICD-9 code 6597)—the only 2 JC conditions in Appendix Table 1 (available online) that increased substantially over time in both states—as low-risk conditions revealed trends in elective interventions virtually identical to those presented in Figures 1 and 3 (not shown). Trends for NJ using clinical GA to define the sample were very similar to those for NJ in Figures 1 and 3, which used LMP-based GA to define the samples in both states. Clinical GA was not available in the CA data.
DISCUSSION
This study found that the rates of C-sections and inductions in low-risk pregnancies increased sharply in NJ and CA from the early 1990s to the mid-2000s; peaked at 33% in CA and 41% in NJ in 2007, when efforts by the American College of Obstetrics and Gynecology, March of Dimes, and other organizations to wait until 39 weeks to intervene were implemented; and then decreased somewhat, particularly at early term. We also found that since the mid-2000s, socioeconomic differences in the rates of elective deliveries have been narrowing in CA but not NJ and that overall, the large increases in elective deliveries observed in the earlier period were not confined to groups with higher SES, suggesting that maternal demand-side factors (e.g., C-sections on request) did not play a major role. Rather, it is more likely that for most of the period, institutional (e.g., medical practice) factors were driving up the rates. It is noteworthy that among early-term births, the rates of elective deliveries increased substantially in both states until the mid/late-2000s, peaked at about 35% in CA and over 40% in NJ, and then decreased in both states to about 20%, which according to recommendations at the time was still too high.
The narrowing of socioeconomic differences in the rates of elective deliveries in CA since the mid-2000s is noteworthy because that state accounts for 12% of the U.S. population
Hospital Quality improvement interventions, statewide policy initiatives, and rates of cesarean delivery for nulliparous, term, singleton, vertex births in California.
CA could even be heading toward a point at which socioeconomically advantaged women are less rather than more likely to have elective deliveries. This potential crossover in risk reduction by SES owing to increases in medical knowledge would be consistent with previous observations
and concerning from a public health equity perspective.
As noted earlier, past research has found that associations between C-sections and neonatal/child morbidities are larger than those between induced deliveries and adverse neonatal/child morbidities. However, we found that an increasing fraction of inductions in low-risk pregnancies (most recently around 25% in both NJ and CA) have led to C-section deliveries, and other studies found that the risks to children in those cases are the same as those of planned elective C-sections.
In other words, whereas elective-induced vaginal deliveries do not appear to be detrimental to children's health, elective inductions are often followed by C-sections (for any possible number of reasons), which do confer health risks to children.
Strengths of our study include the use of linked administrative data on all births in 2 populous and diverse states; the focus on a time period with large changes in delivery methods in the U.S.; and the documentation of trends at the population level, which provides important context for clinicians and practitioners.
Limitations
Limitations include our inability to document trends in elective deliveries in the U.S. as a whole owing to a lack of available data, and the use of administrative data instead of clinical reviews of cases. Another limitation is the possibility that changes in practice and recording of data confounded the trends. However, in supplementary analyses, we found that the prevalence of each individual high-risk condition was stable over time, with the exception of fetal distress (ICD-9 code 65631) and abnormal heart rate (ICD-9 code 6597), and when we reclassified those conditions as low risk, the trends in elective deliveries did not change.
CONCLUSIONS
Given the evidence that elective C-sections increase risks for children and their mothers, that a nontrivial share of induced deliveries results in C-sections, and that the rates of elective deliveries have not substantially declined since their peak in the early 2000s, the trends documented in this paper suggest that sustained or even increased public health attention is needed to address the still-too-high rates of elective C-sections and inductions in the U.S.
ACKNOWLEDGMENTS
The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the NIH, HHS/Health Resources and Services Administration, or the Robert Wood Johnson Foundation.
This work was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (Award Numbers R01HD090119 and P2CHD058486). The research was also supported by the National Center for Advancing Translational Sciences, a component of NIH under Award Number UL1TR003017; the HHS/Health Resources and Service Administration under Award Number U3DMD32755; and the Robert Wood Johnson Foundation through its support of the Child Health Institute of New Jersey (Grant 74260).
Declarations of interest: none.
CRediT AUTHOR STATEMENT
Julien Teitler: Conceptualization, Funding acquisition, Methodology, Supervision, Writing - original draft. Valentina Chegwin: Data curation, Writing - review & editing. Linda Li: Data curation, Writing - review & editing. Kayuet Liu: Conceptualization, Formal analysis, Funding acquisition, Methodology, Supervision, Writing - original draft. Peter Bearman: Conceptualization, Funding acquisition, Writing - review & editing. Marilyn A. Gorney-Daley: Writing - review & editing. Nancy E. Reichman: Conceptualization, Funding acquisition, Methodology, Supervision, Writing - original draft.
Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop.
Hospital Quality improvement interventions, statewide policy initiatives, and rates of cesarean delivery for nulliparous, term, singleton, vertex births in California.
00050-5&id=ga1.jpg){kind=link}
00050-5&id=gr1.jpg){kind=link}
00050-5&id=gr2.jpg){kind=link}
00050-5&id=gr3.jpg){kind=link}