Evaluating rates and factors associated with cesarean section and inpatient cost among low-risk deliveries in selected U.S. states

Abstract

Introduction

Cesarean sections are commonly performed in the United States, including among patients for whom vaginal delivery may be clinically feasible. This study aimed to evaluate rates and factors associated with cesarean section use and inpatient cost among low-risk deliveries in selected U.S. states.

Methods

This was a retrospective, cross-sectional analysis using Healthcare Cost and Utilization Project (HCUP) State Inpatient database for Maryland, Florida, and Wisconsin between January 1, 2017, and December 31, 2020. American Hospital Association (AHA) data and median household income quartiles based on the Agency for Healthcare Research and Quality’s (AHRQ) 2018 estimates were included in this study to assess hospital and patient neighborhood characteristics. AHA data was linked to HCUP data using the hospital identifier number. Median household income quartiles were linked to HCUP using ZIP codes. A multivariable generalized estimating equations regression model including a random intercept for hospitals was used to identify patient- and hospital-level characteristics associated with the use of cesarean section.

Results

245,383 women who underwent a delivery between 2017 and 2020 were included in the analysis. Of these women, 8.1% had cesarean section and 91.9% had vaginal delivery. Mean age was 26.9 (SD ± 4.41) years for cesarean section and 26.9 (SD± 4.37) years for vaginal delivery. An increasing rate of cesarean section was detected during the study period. Higher rates of cesarean section were found among Black and Hispanic women compared to White and Asian, and among women with lower income. Hospitals in Florida had the highest cesarean section rate of 9.4% among low-risk women while Maryland and Wisconsin had rates of 6.3% and 5.3%, respectively. Being Hispanic or Black, having private insurance, and giving birth in a for-profit hospital were associated with higher cesarean section utilization after controlling patient- and hospital-level factors.

Discussion

A range of clinical and policy interventions have been implemented over the past decade to reduce cesarean sections among low-risk deliveries; however, we still identified an increasing rate of cesarean section among low-risk women between 2017 and 2020 in select U.S. states. There is an emergent need to revisit policies and interventions that impact cesarean section in these states. Women with low socioeconomic status were more vulnerable to have cesarean sections. Identifying variation in cesarean delivery rates among low-risk populations may inform future efforts to improve maternal care quality.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-025-08148-0.

Introduction

Cesarean sections are frequently performed in the United States, including in cases where vaginal delivery is clinically feasible. However, decisions around delivery mode are often influenced by a combination of clinical guidelines, patient preferences, provider practices, and institutional norms [1]. The World Health Organization recommends that cesarean section be performed only when medically necessary. However, rates of primary cesarean sections have been increasing nationally over the past few decades, without evidence to support that all are medically necessary [2]. Prior studies have suggested that provider decision-making around cesarean delivery may be influenced by non-clinical factors such as institutional practices, time constraints, and perceived convenience, particularly when clinical indications are ambiguous [3].

Cesarean section delivery rates in the U.S. have remained high, accounting for approximately 32% of all births in 2023 [4]. Studies have identified substantial variation in cesarean section use across hospitals and regions, particularly among low-risk deliveries, suggesting potential overuse in certain settings [5–7]. The overuse of cesarean sections can pose several complications for both mother and infant. While cesarean delivery is generally safe, it is associated with higher risks of surgical complications, such as infection, hemorrhage, and longer hospital stays, compared to vaginal delivery. Cesarean delivery among low-risk patients is associated with increased maternal risks, including surgical site infections, hemorrhage, injury to surrounding organs (such as the bowel or bladder), adhesions, and longer recovery times compared to vaginal delivery [8, 9]. Mothers who have a primary cesarean delivery are more likely to have repeat cesareans in subsequent pregnancies; however, vaginal birth after cesarean is successful in approximately 60–80% of appropriately selected cases [10]. Risks to infants delivered through cesarean sections include higher rates of respiratory problems, which may lead to increased NICU admissions [2].

The cesarean section risks should be considered on an individual basis; however, a pattern of frequently performing cesarean sections in low-risk deliveries represents a pattern of harm. This overuse disproportionately impacts women of lower socioeconomic status (SES) [11, 12]. Women from lower SES groups are also more likely to experience complications due to cesarean sections, as research studies have shown that Medicaid-insured women had a higher risk of infection after cesarean sections than privately insured women [13]. Additionally, Black and Asian women are more likely to undergo primary cesarean sections compared to their White counterparts, showing a racial disparity in delivery methods among women of color [14, 15].

Factors associated with increasing cesarean section utilization remains unclear especially among women without indications for cesarean delivery. To address this gap, we conducted a study to evaluate the rates and factors associated with cesarean section and hospital utilization cost from 2017 to 2020 among low-risk women with no prior births using hospital discharge records of three U.S. states.

Methods

Study design and study population

This retrospective cross-sectional study utilized the Healthcare Cost and Utilization Project (HCUP) State Inpatient database for the states of Maryland, Florida, and Wisconsin. These states were chosen due to their large population estimates, diverse demographic characteristics, and availability of all necessary data components for our analysis.

The study population included women who underwent either a vaginal delivery or cesarean section between January 1, 2017, and December 31, 2020, as captured in HCUP and identified using International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS) codes (Appendix Table 1). Because parity information is not directly available in HCUP data, we applied a one-year lookback period to minimize the likelihood of including women with a prior delivery. However, we acknowledge that this approach may not completely exclude multiparous women. We used 2016 as a buffer year to ensure consistency in diagnosis and procedure coding following the transition to ICD-10 in the United States in late 2015. To create a low-risk population, we excluded participants with gestational age below 37 weeks, multiple pregnancies, abnormal fetal weight, and certain pregnancy-associated complications such as pre-eclampsia, placenta previa, HELLP syndrome, and eclampsia. Additional exclusions included high risk conditions (e.g. failed labor, multiple fetus), moderate risk conditions (e.g. long labor, nuchal cord, and fetus abnormality), and any diagnosis of Sever Maternal Morbidity (SMM) according to CDC Maternal Infant Health [16]. We also excluded women with HIV, diabetes, cardiovascular diseases, or other conditions that might indicate high risk deliveries, as determined through ICD-10 codes (Appendix Table 2) [14, 17, 18]. Our exclusion criteria were informed by guidance from the American College of Obstetricians and Gynecologists (ACOG) regarding common medical indications for cesarean delivery [19]. Subjects with missing or inapplicable information were also removed from the study. We report cesarean use among a population without documented clinical indications for surgical delivery based on administrative data to assess potential overuse of cesarean sections in low-risk populations.

To assess hospital and patient neighborhood characteristics, we utilized the American Hospital Association (AHA) data and median household income quartiles based on the Agency for Healthcare Research and Quality’s (AHRQ) 2018 estimates. AHA data was linked to HCUP data by AHA hospital unique identifier, while median household income quartiles data were linked based on the patient’s ZIP code of residence.

Institutional Review Board (IRB) approval was obtained by the institutional review board at the Johns Hopkins University School of Medicine (IRB # 00005170).

Independent variables

The study obtained patient demographic characteristics from HCUP records, which included age, race/ethnicity (i.e., White, Black, Asian, Hispanic, and Other/Unknown), patient insurance type (i.e., Medicaid, private, or other), and Median household income identified using the patient’s ZIP code. Diagnostic codes were not considered as an independent variable due to the very low number of low-risk women with comorbidities that were included in the study after applying the exclusion criteria (0.024% of the population). The reference categories used for demographic variables in the multivariable analysis were: White for race/ethnicity, Medicaid for insurance type, and the lowest income quartile for median household income.

Furthermore, the study obtained hospital characteristics such as hospital size (i.e., number of beds), hospital type (i.e., governmental, for-profit, and not-for-profit), and hospital teaching status (i.e., yes or no) from AHA data. The reference categories for hospital characteristics were: <200 beds for hospital size, governmental nonfederal for hospital type, and non-teaching hospital for teaching status.

Outcome

The primary outcome of this study was cesarean sections among low-risk deliveries. We explored the rates in cesarean section utilization over time, by race and SES, and compared utilization rate and cost between cesarean section and vaginal delivery. We also identified differences and similarities among the three states regarding the percentage of hospitals with cesarean sections among low-risk deliveries in each state.

Statistical analysis

We described woman and hospital characteristics by each procedure. Mean, median, standard deviation, and interquartile range (IQR) were reported as appropriate. Chi-squared tests were used for categorical variables and student t-test, or Mann-Whitney U test were used for continuous variables to assess differences in baseline characteristics. We plotted the cesarean section utilization rate by year and by racial group and different median household income. A bar graph was used to display the distribution of cesarean section utilization by hospitals in Maryland, Florida and Wisconsin. A generalized estimating equation (GEE) model was applied to account for patient clustering within hospitals and identifying patient- and hospital-level characteristics associated with the use of cesarean section. The GEE model incorporated hospital ID as the clustering variable. Patient covariates such as age, race, insurance type and median household income quartiles were included in the model, along with hospital covariates like hospital size/beds, hospital type, and teaching status. The selection of covariates was informed by data availability and standard demographic and hospital-level covariates commonly used in health services research [20, 21]. Different correlation structures were explored, and after considering the theoretical structure of the data, the independent correlation structure was chosen based on having the lowest QIC (quasi-likelihood under the independence model criterion) score [22], indicating the best model fit.

A sensitivity analysis was conducted by restricting patients to one state at a time. Statistical analyses were performed using R (version 4.1.2) [23–26]. Significance level was set at p < 0.05.

Results

We identified a total of 245,383 women with a delivery hospitalization between 2017 and 2020 in Maryland, Florida, and Wisconsin, with 8.1% (n = 19,817) undergoing cesarean section and 91.9% (n = 225,566) experiencing vaginal delivery. (Appendix Fig. 1) The median age at the time of delivery did not differ significantly between the two groups (27 years for both groups).

In comparison to the vaginal delivery group, the cesarean section group demonstrated a higher proportion of Black (22.2% vs. 18.8%) and Hispanic (26.8% vs. 21.6%) patients, and a lower proportion of White (45.2% vs. 53.0%) and Asian (2.7% vs. 3.4%) patients. The cesarean section group also included a higher proportion of women in the lowest quartile of median household income (29.8% vs. 25.9%). Insurance coverage was similar between the groups, with the majority having Medicaid (43.7% vs. 43.8%) or private insurance (51.8% vs. 51.3%).

For hospital characteristics, 41.0% (n = 8,132) of cesarean sections and 37.5% (n = 84,574) of vaginal deliveries occurred in hospitals with 200–399 beds. Additionally, 23.4% (n = 4,629) of cesarean sections and 16.1% (n = 36,279) of vaginal deliveries took place in for-profit hospitals. The average cost across all years was $31,200 (SD ± $19,700) for cesarean section and $16,000 (SD ± $10,100) for vaginal delivery (Table 1; Post-hoc pairwise comparisons with Bonferroni-adjusted p-values are presented in Appendix Table 3).

Table 1.

Characteristics of low-risk woman with no prior birth who underwent delivery between 2017 and 2020 in maryland, florida, and Wisconsin

Personal Characteristics	Cesarean sections	Vaginal Deliveries	P-value
N (%)	19,817 (8.1%)	225,556 (91.9%)	-
Age (yrs)			0.147
Mean (STD)	26.9 (4.41)	26.9 (4.37)
Median [Q1,Q3]	27.0 [23.0, 30.0]	27.0 [24.0, 30.0]
Race			< 0.001
White	8960 (45.2%)	119,575 (53.0%)
Black	4396 (22.2%)	42,431 (18.8%)
Asian	530 (2.7%)	7760 (3.4%)
Hispanic	5304 (26.8%)	48,768 (21.6%)
Other/unknown	627 (3.2%)	7032 (3.1%)
Insurance type			0.0422
Medicaid	8651 (43.7%)	98,805 (43.8%)
Private	10,263 (51.8%)	115,634 (51.3%)
Other	903 (4.6%)	11,127 (4.9%)
Median Household income			< 0.001
1 st Quartile ($0-45999)	4441 (22.4%)	56,797 (25.2%)
2nd Quartile ($46000–60999)	4602 (23.2%)	56,419 (25.0%)
3rd Quartile ($61000–81999)	5196 (26.2%)	56,638 (25.1%)
4th Quartile ($82000+)	5578 (28.1%)	55,712 (24.7%)
Cost			< 0.001
Mean (STD)	31,200 (19700)	16,000 (10100)
Median [Q1, Q3]	28,300 [18600, 39700]	13,500 [8890, 20600]

Hospital Characteristics	Cesarean sections	Vaginal Deliveries	P-value
Number of Hospital Beds			< 0.001
<200	4004 (20.2%)	50,117 (22.2%)
200–399	8132 (41.0%)	84,574 (37.5%)
400–599	2757 (13.9%)	34,337 (15.2%)
600–999	1825 (9.2%)	22,400 (9.9%)
>= 1000	3099 (15.6%)	34,138 (15.1%)
Hospital type			< 0.01
Governmental	2103 (10.6%)	20,779 (9.2%)
For profit	4629 (23.4%)	36,279 (16.1%)
Non-profit	13,085 (66.0%)	168,508 (74.7%)
Teaching status			0.0178
No	17,570 (88.7%)	198,703 (88.1%)
Yes	2247 (11.3%)	26,863 (11.9%)

Cesarean section rates were higher among Black and Hispanic women compared to White and Asian women. Cesarean sections were also more prevalent among women residing in lower income ZIP codes from 2017 to 2020 (Fig. 1).

Fig. 1.

Annual percentage of cesarean sections among low-risk deliveries utilization by race and socioeconomic status

Among the three states, Wisconsin had the highest proportion of hospitals in the 0–3% cesarean sections among low-risk deliveries category. Maryland had the highest proportion of hospitals within the 3–6% cesarean sections among low-risk deliveries category and Florida had the highest proportion of hospitals within the 6–9%, 9–12% and 12%+ categories (Fig. 2).

Fig. 2.

Percentage of hospitals with cesarean sections among low-risk deliveries across three states, grouped by 3% incremental bins

The regression model revealed that Black (aOR = 1.16, p < 0.001), Hispanic (aOR = 1.23, p < 0.001), and women of Other racial/ethnic backgrounds (aOR = 1.12, p < 0.001) had significantly higher odds of undergoing cesarean sections among low-risk deliveries compared to White women. Women with insurance types categorized as Other (aOR = 1.12, p < 0.001) or Private Insurance (aOR = 1.19, p < 0.001) exhibited significantly higher odds of receiving cesarean sections among low-risk deliveries relative to those with Medicaid. Women in the highest (4th) median household income quartile displayed significantly lower odds of undergoing cesarean sections among low-risk deliveries (aOR = 0.93, p < 0.0001) compared to those in the lowest quartile (Table 2).

Table 2.

Multivariable generalized estimating equations model assessing individual and hospital characteristics associated with Cesarean section among low-risk women utilization, 2017–2020

Characteristic	aOR	95% CI	p-value
Age	1.00	1.00, 1.00	0.02
Race
White	Ref
Asian	1.02	0.99, 1.06	0.2
Black	1.16	1.12, 1.19	< 0.001
Hispanic	1.23	1.19, 1.26	< 0.001
Others	1.12	1.07, 1.17	< 0.001
Insurance Type
Medicaid	Ref
Other	1.12	1.07, 1.16	< 0.001
Private Insurance	1.19	1.16, 1.21	< 0.001
Median Household income quartile
1 st Quartile ($0-45999)	Ref
2nd Quartile ($46000–60999)	1.00	0.97, 1.02	0.8
3rd Quartile ($61000–81999)	0.97	0.95, 1.00	0.10
4th Quartile ($82000+)	0.93	0.90, 0.97	< 0.001
Number of Hospital Beds
< 200	Ref
200–399	1.08	1.02, 1.14	0.009
400–599	0.87	0.81, 0.94	< 0.001
600–999	0.90	0.82, 0.98	0.012
>= 1000	0.91	0.85, 0.99	0.020
Hospital Type
Government, Nonfederal	Ref
Investor-owned (for-profit)	1.14	1.05, 1.23	0.002
Nongovernment, not-for-profit	0.87	0.81, 0.93	< 0.001
Teaching Hospital
No	Ref
Yes	1.01	0.94, 1.09	0.8

The analysis showed that women who delivered in hospitals with 200–399 beds had significantly higher odds of receiving cesarean sections among low-risk deliveries (aOR = 1.08, p = 0.01) compared to those who delivered in hospitals with fewer than 200 beds. In contrast, women who delivered in larger hospitals with 400–599 beds (aOR = 0.87, p < 0.001) and 600–999 beds (aOR = 0.90, p = 0.012) had significantly lower odds of undergoing cesarean sections. Additionally, women who delivered at investor-owned (for-profit) hospitals had significantly higher odds of receiving cesarean sections (aOR = 1.14, p = 0.002) and women who delivered at non-governmental, not-for-profit hospitals showed lower odds of unnecessary cesarean section (aOR = 0.87, p-value < 0.001) compared to women who delivered at nonfederal government hospitals.

To further investigate factors associated with cesarean sections in each state, we conducted separate regression models for Florida, Wisconsin, and Maryland. Median household income was not significantly associated with cesarean sections in Florida and Wisconsin. However, in Maryland, women in the highest median household income quartile had significantly lower odds of undergoing cesarean sections (OR = 0.89, p = 0.026) compared to those in the lowest quartile. (Appendix Tables 4, 5 and 6)

Discussion

Cesarean section, as the alternative method to vaginal delivery, has been increasing over time, which raises concerns about its overuse and potential harm. In our study of 245,383 women with no clinical indication for cesarean delivery, we observed an increasing rate in cesarean section rates from 2017 to 2019 in three U.S. states. Using HCUP inpatients data, we identified factors that are associated with cesarean section including being Black or Hispanic, having private insurance coverage, and living in a lower median household income ZIP code.

The prevalence of cesarean section among low-risk delivery utilization was 8.1% in our study, which is lower than the national rates of all cesarean sections estimated at 31.9% in 2019[27] and 31.8% in 2020[28]. Our lower utilization rate is mainly due to the restricted definition that we used to exclude necessary cesarean sections. We used this definition to evaluate factors associated with cesarean sections among low-risk women thus identifying the overuse of unnecessary cesarean section. Nonetheless, we found it challenging to identify well-defined criteria for medically necessary cesarean sections. Future studies should develop useful computational phenotypes (e.g., diagnosis and procedure codes) to identify medical conditions that require cesarean section in conventional clinical data sources such as electronic health records or hospital discharge data.

Our findings show a general increase in cesarean delivery rates in 2020 across all income and racial/ethnic groups. This pattern coincides with the onset of the COVID-19 pandemic, which substantially disrupted maternity care delivery. Pandemic-related factors, such as reduced prenatal care visits, altered hospital policies, provider shortages, and efforts to minimize labor duration and exposure, may have influenced decision-making around delivery mode, possibly leading to increased cesarean use, particularly among vulnerable populations. Prior studies have noted that low-income and racially minoritized groups experienced greater barriers to care and higher rates of COVID-related complications [29, 30], which may have contributed to both increased cesarean section rates and widened disparities. While our study cannot directly attribute the observed increase to pandemic-related factors, this temporal association underscores the importance of examining how emergency response policies and systemic disruptions can differentially affect obstetric outcomes across populations.

We found racial and socioeconomic disparities in the methods of delivery among women of color and/or from low socioeconomic backgrounds. For instance, compared to Wisconsin, which is majority White (~ 80%), Maryland and Florida, which have lower percentages of Whites (~ 48% and ~ 52%), had higher unnecessary cesarean section rates (5.4% compared to 6.3% and 9.4%, respectively) [31]. Notably, Black make over 30% of the overall population in Maryland, while Florida has around 26% Hispanics[31]. Our finding is consistent with previous retrospective studies [32, 33]. A randomized trial found that non-Hispanic Black and Hispanic people had higher relative risk for cesarean section compared to White population [34].

We found that women residing in lower median income ZIP codes had a higher prevalence of cesarean sections across the study population. For example, 9.2% of all deliveries were cesarean sections for women residing in the lowest quartile of income ZIP codes, while this rate was 6.7% for the highest quartile. Similarly, patients residing in the highest income ZIP codes, which were mainly in Maryland, had significantly lower cesarean section utilizations across the three states (e.g., women in the top 25% ZIP median income of Maryland had 5.9% unnecessary cesarean section). Despite these findings, when cesarean section utilization was compared across ZIP code median income of each state separately, statistical significance was not detected.

Our findings on ZIP code median income and rates of unnecessary cesarean section contrast with rates reported in low- and middle-income countries (LMICs). A study using survey data from 72 LMICs found that cesarean section rate varies substantially between and within countries but cesarean section rates increased with raising economic status. The same study showed a median cesarean section rate of 3.7% among the poorest fifth and 18.4% among the richest fifth LMICs [27]. Moreover, in LMICs, affluent women exhibit a frequency of cesarean section usage that is over five times greater compared to their economically disadvantaged counterparts [27, 35]. For example, a study conducted in Brazil found that socioeconomic status of monthly family income per capita was positively associated with cesarean Sect. [36] This discrepancy can be partially attributed to the fact that impoverished women are more likely to give birth at home and/or not being able to afford a cesarean section. Nevertheless, even when these women manage to access healthcare facilities for delivery, they are still approximately 2.5 times less likely to undergo a cesarean Sect. [37] The difference between our results and the LMIC studies might be due to differences in healthcare policies (e.g., insurance coverage, healthcare affordability), variations in healthcare resources and access, economic priorities, and cultural preferences.

We identified a variation in the cesarean section rates among hospitals which suggest potential practice differences in performing labor and delivery. We found a range of 0 to 29.1% variation of unnecessary cesarean section among hospitals which is three times higher than an earlier study using Medicaid data, which found rates ranging from 2.4 to 36.5% among low-risk pregnancies [38]. These findings suggest a need to address the underlying hospital-level factors that contribute to the overuse of unnecessary cesarean section, which may lead to healthcare inequity.

Reducing unnecessary cesarean section interventions had been implemented in different settings. One successful example was the program implemented in Californica between 2012 and 2015, which reduced cesarean section rate from 26 to 22.8% in subsequent years [7]. Policy interventions, such as reducing financial incentives for cesarean delivery or introducing institutional penalties, have been associated with declines in cesarean rates in some settings. However, our findings suggest that reductions in cesarean use may not occur uniformly across patient groups. Future research is needed to explore how provider- and hospital-level incentives may contribute to persistent disparities in cesarean delivery rates, particularly among racially and socioeconomically minoritized populations.

In states with restrictive abortion laws or no legal abortion policies, it is possible that a higher rate of cesarean sections could be observed due to various reasons. First, women may have less access to comprehensive reproductive healthcare services. This can lead to inadequate prenatal care, which may contribute to increased rates of cesarean sections. Second, there may be fewer medical professionals with training in reproductive health, and as a result, they may not be as skilled in determining when a cesarean section is medically necessary. Third, cultural attitudes and beliefs about childbirth, as well as the stigma surrounding abortion, could play a role in influencing unnecessary cesarean section rates. For instance, in areas where there is strong opposition to abortion, there may be a higher likelihood of performing cesarean sections even when not medically necessary, as this could be perceived as a more socially acceptable option. Lastly, healthcare providers may practice “defensive medicine” due to fear of malpractice lawsuits or other legal repercussions which could lead to a higher number of cesarean sections, as medical providers may choose to perform the procedure to minimize any perceived risk.

This study has several limitations. First, we only included 3 states in the analysis. We chose these states based on their population estimate, age, gender, and race distribution, as well as the variety of median household income. Second, we only used HCUP inpatient data thus capturing cesarean sections performed in a hospital setting. We did not have data on out-of-hospital facilities that may also perform delivery. Meanwhile, we used logistic regression via GEE to account for clustering, we acknowledge that odds ratios may overestimate effect sizes when the outcome is common; however, the overall cesarean rate in our low-risk cohort was 8.1%, which partially mitigates this concern. Third, we developed our exclusion criteria based on previous studies. Our results may change if a different set of criteria was used to define the study population. Our analysis did not distinguish between spontaneous and assisted vaginal deliveries (e.g., forceps- or vacuum-assisted). This may limit our ability to assess variation in delivery type with greater clinical specificity. Further research should focus on official guidelines and computational phenotypes to define various delivery methods and reasons to choose them. In addition, our study could not directly identify primiparous deliveries, as the HCUP inpatient dataset does not capture birth certificate variables such as “Previous Live Births Now Living.” This variable, available in state vital statistics records, would allow for more precise ascertainment of first deliveries. We applied a one-year lookback period to reduce, but not fully eliminate, the inclusion of multiparous women. Therefore, some misclassification of parity is possible. Future studies linking inpatient data with birth certificate files may better address this limitation. Lastly, we cannot identify the willingness or preference of women getting a cesarean section in the data. A variety of reasons may affect an individual to prefer a cesarean section such as fear of labor pain, fear of pelvic floor damage, fear of urinary incontinence, and fear of negative effects on sexuality or sexual relationships [39]. Thus, we could not study the role of patient preference in the observed cesarean sections.

Conclusion

A multitude of healthcare policies and clinical interventions have been rolled out in the past decade to reduce cesarean sections; however, we still identified an increasing rate of cesarean section among low-risk women between 2017 and 2020 in three U.S. states. Policies and interventions to reduce unnecessary cesarean section should be revisited in these states, and perhaps other states too. The study also showed that women of color and women residing in low-income communities are more vulnerable to having cesarean sections. Efforts to reduce cesarean delivery rates should be informed by further research on the underlying factors contributing to disparities in cesarean use among low-risk populations.

Authors’ contributions

Dun and Zhang contributed equally to this manuscript as joint first authors. Dun and Zhang had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Dun, Zhang, Wei, Kharrazi. Acquisition, analysis, and interpretation of data: Dun, Zhang, Wei, Kharrazi. Drafting of the manuscript: Dun, Zhang, Wei. Critical revision of the manuscript of important intellectual content: Dun, Zhang, Wei, Aziz, Kharrazi. Statistical analysis: Dun, Zhang, Wei. Administrative, technical, or material support: Dun, Zhang, Wei, Aziz, Kharrazi. Study supervision: Kharrazi.

Funding

No funding was reported for this study.

Data availability

Data is public available and can be downloaded at https://hcup-us.ahrq.gov/databases.jsp.

Declarations

Ethics approval and consent to participate

Institutional Review Board (IRB) approval was obtained by the institutional review board at the Johns Hopkins University School of Public Health.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.