(C) Daily Yonder - Keep it Rural
This story was originally published by Daily Yonder - Keep it Rural and is unaltered.
. . . . . . . . . .



Obstetrics & Gynecology [1]

['Atwani', 'Rula Robbins', 'Lindsay Saade', 'George Kawakita', 'Author Information']

Date: 2025-06-26

Maternal health outcomes in the United States continue to fall behind those of other high-income countries, with maternal mortality and morbidity representing serious public health challenges.1 The U.S. maternal mortality rate rose by 89% since 2018, reaching 32.9 deaths per 100,000 live births in 2021.2 Structural inequities in health care access—particularly in rural and underserved areas—leave large segments of the population without adequate maternity care.3,4 These disparities disproportionately affect historically marginalized communities, exacerbating racism and socioeconomic inequities in maternal health outcomes.5–7

A key determinant of maternal health is the availability of qualified maternity care practitioners, including obstetricians, gynecologists, and certified nurse–midwives. Previous studies suggest that adequate access to these practitioners improves outcomes such as preterm birth, preeclampsia, and neonatal outcomes.8–10 Their uneven distribution has created gaps in care, with up to 40% of rural counties lacking maternity care practitioners.11,12

The concept of “maternity care deserts” has become a focus for policy makers and researchers, because it addresses inequities in access to maternal health care. These areas are defined as counties lacking a hospital or birth center offering obstetric care and the absence of maternity care practitioners.13 Maternity care deserts are associated with significantly higher rates of adverse maternal outcomes because limited access to skilled practitioners delays timely interventions during critical stages of pregnancy and childbirth.14,15 In this analysis, we aimed to evaluate the association between the level of maternity care access and maternal mortality.

METHODS

This was a cross-sectional study using county-level data from publicly available resources, specifically the Centers for Disease Control and Prevention (CDC) Wide-Ranging Online Data for Epidemiologic Research database.16 The analysis included data from January 2018 to December 2021. The CDC Wide-Ranging Online Data for Epidemiologic Research platform incorporates multiple data sets such as the Natality, Underlying Cause-of-Death, and Fetal Death databases. Information on live births was extracted from the Natality data set, which records all births occurring within the United States to U.S. residents.17 Mortality data were gathered from the Underlying Cause-of-Death data set, which is based on death certificates and provides both primary causes of death and detailed demographic information.18

To enhance accuracy and to mitigate misclassification, the study population was restricted to individuals aged 15–44 years in light of prior findings indicating a higher false-positive rate for maternal mortality among individuals aged 45 years or older.19 All data sets adhered to consistent inclusion and exclusion criteria. Lastly, we limited analyses to individuals who lived in 50 states and the District of Columbia. Because only publicly available, deidentified data were used, the study was determined to be exempt from human subjects review by the IRB. All reporting followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.

The primary exposure was a time-varying, county-level measure of maternity care access, adapted from the March of Dimes methodology.20 For each calendar year from 2018 to 2021, we calculated three components using year-specific data: the ratio of obstetric clinicians—obstetrician–gynecologists (ob-gyns), certified nurse–midwives, and certified midwives—to the annual number of live births, the presence of at least one birthing facility (hospital obstetric unit or accredited freestanding birth center), and the percentage of female individuals of reproductive age who were uninsured. Counties were classified each year into four mutually exclusive categories: maternity care desert, low access, moderate access, or full access (operational definitions are provided in Appendix 1, available online at https://links.lww.com/AOG/E187).20 Annual counts of clinicians, hospital obstetric units, and insurance coverage were obtained from the Area Health Resources File for 2018–2021,21 and data on accredited birth centers were abstracted from the Commission for the Accreditation of Birth Centers database.22

The primary outcome of interest was maternal mortality per 100,000 live births, defined as the death of a pregnant person during pregnancy or within 42 days of termination of pregnancy, regardless of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, excluding accidental or incidental causes.23 Deaths were identified with International Classification of Diseases, Tenth Revision codes A34, O00–O95, and O98–O99. The secondary outcome was pregnancy-related mortality, defined as death during pregnancy or within 1 year of the end of pregnancy from a pregnancy complication, a chain of events initiated by pregnancy, or the aggravation of an unrelated condition by the physiologic effects of pregnancy, which encompasses deaths occurring within 1 year postpartum (all O codes and A34).24 Outcomes were assigned to the county in which individuals resided to maintain consistency across data sets.

To visualize the geographic distribution of maternity care practitioners and maternal mortality rates, we created county-level maps after calculating the average rates over the study period. Demographic characteristics are reported stratified by the four access categories. Ordered trends across categories were assessed with the Jonckheere–Terpstra test for continuous variables and the Cochran–Armitage trend test for binary variables. Next, multivariable negative binomial mixed-effects models were used to examine the association between maternity care access and maternal health outcomes. A negative binomial distribution with a log-link function was chosen because of overdispersion. The log of the number of births was included as an offset to account for the differences in the number of births. The analysis included adjustments for random effects of states and counties, the proportion of maternal age 35 years or older, obesity (body mass index [BMI, calculated as weight in kilograms divided by height in meters squared] 30 or higher), chronic hypertension, pregestational diabetes, and Social Vulnerability Index—selected because of their well-established association with maternal mortality.25,26 These data were obtained from the Area Health Resources File and CDC.21,27 Birth year was included as a covariate to capture temporal trends, including the effect of the coronavirus disease 2019 (COVID-19) pandemic. Adjusted rates of outcomes were calculated from the marginal standardization form of predictive margins.28 Absolute risk differences were calculated by estimating contrasts of margins to improve interpretability by estimating adjusted probabilities. Absolute risk differences illustrate differences in maternal health outcomes across levels of maternity care practitioner density while mitigating misinterpretation risks for outcomes with low baseline incidence. Lastly, adjusted incident rate ratios and 95% CIs were calculated for maternal mortality and pregnancy-related mortality.

Robustness was assessed through sensitivity analyses. First, the associations were reanalyzed by examining the number of ob-gyn physicians only (MD or DO), excluding midwives. This approach was taken to isolate the effect of physician-based obstetric care and to account for variability in midwifery practice, scope, and accreditation across different regions, thereby providing a clearer understanding of potential differences in practitioner composition. Second, we analyzed outcomes based exclusively on maternity care practitioner density (0, 1–29, 30–59, and 60 or more maternity care practitioners/10,000 live births) instead of standardized definitions of maternity care access, which incorporate hospital availability and insurance coverage. This allowed us to isolate the effect of practitioner availability independently of other structural factors, offering a more targeted assessment of how practitioner distribution influences maternal outcomes. Third, a leave-one-out approach was applied, sequentially excluding each state to identify any disproportionate influence on the results. Finally, we examined the mortality of reproductive-aged female individuals (age 15–44 years) regardless of pregnancy, which would not be affected by the risk of misclassification surrounding the mortality of pregnant individuals. For this analysis, the log of the number of reproductive-aged female individuals was included as an offset.29 All analyses were conducted with Stata 18.5 and R 4.4.1. Statistical significance was defined as P<.05 or a 95% CI excluding the null.

RESULTS

The study included 14,772,210 live births across four categories of maternity care access: desert (n=720,858), low access (n=708,668), moderate access (n=431,188), and full access (n=12,911,496) (Appendix 2, available online at https://links.lww.com/AOG/E187). The annual distribution of counties within each category is provided in Appendix 3, available online at https://links.lww.com/AOG/E187. Figure 1 illustrates the distribution of maternity care practitioner densities and maternal mortality ratios across counties in the United States. Counties in the West and Northeast, including areas in California, Washington, and parts of the Northeast, exhibited the highest ob-gyn densities. These areas correlated with lower maternal mortality ratios, represented by light blue. In contrast, counties in the South and Midwest demonstrated lower maternity care practitioner densities. These regions also exhibited higher maternal mortality ratios, highlighting disparities in access to maternity care and maternal health outcomes across counties.

Fig. 1.: Distribution of obstetrician and gynecologist density and maternal mortality per 100,000 live births across counties. A. Maternal mortality rates per 100,000 live births across counties. Counties are color-coded on a gradient from gray (lower rate) to yellow and orange (higher rate), with orange highlighting areas with the highest rate. B. Density of obstetricians and gynecologists per 10,000 live births across counties. Counties are color-coded on a gradient from orange (lower density) to gray (higher density), with gray indicating areas with the highest density of physicians. MMR, maternal mortality rate.

Demographics according to the level of maternity access are presented in Table 1. All demographic characteristics were statistically significantly different across the level of maternity access except for Social Vulnerability Index. Individuals in maternity care desert counties were more likely to be younger and to have obesity, chronic hypertension, and pregestational diabetes. Adjusted rates of outcomes according to the level of maternity access are presented in Appendix 4, available online at https://links.lww.com/AOG/E187.

Desert counties compared with full-access counties had a significantly higher maternal mortality rate (32.25 vs 23.62; absolute risk difference 8.62 [95% CI, 4.63–12.61]; adjusted incidence rate ratio 1.36 [95% CI, 1.21–1.54]) and pregnancy-related mortality rate (43.82 vs 34.72; absolute risk difference 9.10 [95% CI, 4.28–13.93]; adjusted incidence rate ratio 1.26 [95% CI, 1.13, 1.41]). Outcomes according to level of maternity care access are presented in Fig 2. Low-access counties showed no significant difference in maternal mortality rate (24.04 vs 23.62; absolute risk difference 0.41 [95% CI, −3.68 to 4.51]; adjusted incidence rate ratio 1.02 [95% CI, 0.86–1.21]) or pregnancy-related mortality rate (35.63 vs 34.72; absolute risk difference 0.91 [95% CI, −4.06 to 5.88]; adjusted incidence rate ratio 1.03 [95% CI, 0.89–1.18]). Similarly, moderate-access counties had no significant differences in maternal mortality rate (25.59 vs 23.62; absolute risk difference 1.96 [95% CI, −4.50 to 8.42]; adjusted incidence rate ratio 1.08 [95% CI, 0.84–1.40]) or pregnancy-related mortality rate (37.48 vs 34.72; absolute risk difference 2.76 [95% CI, −5.72 to 11.24]; adjusted incidence rate ratio 1.08 [95% CI, 0.86–1.36]).

Fig. 2.: Outcomes according to level of maternity care access. Forest plot displays absolute risk differences (ARDs) for maternal and pregnancy-related mortality across four levels of maternity care access: desert, low access, moderate access, and full access. Adjusted rates, ARDs, and adjusted incidence rate ratios (IRRs) were adjusted for random effects of states and counties, proportion of maternal age 35 years or older, obesity, chronic hypertension, pregestational diabetes, Social Vulnerability Index, and year of birth. Unadjusted rates were obtained from mixed-effect models adjusted only for random effects of states and counties. Full access serves as the reference group. Horizontal bars represent 95% CIs around the point estimates for ARD. Adjusted rates are presented per 100,000 live births.

The results of the sensitivity analyses are presented in Appendices 4–7, available online at https://links.lww.com/AOG/E187. Overall, the findings remained consistent with the primary analyses, reinforcing the robustness of the results. Sensitivity analyses limited to obstetricians (MD or DO) only (Appendix 4, https://links.lww.com/AOG/E187) demonstrated similar patterns, with higher maternal and pregnancy-related mortality rates observed in desert counties compared with full-access counties, whereas low- and moderate-access counties showed no significant differences. Analysis considering only maternity care practitioner density (Appendix 5, available online at https://links.lww.com/AOG/E187) yielded comparable results, confirming higher risks in counties with no practitioners and no significant differences in counties with low-to-moderate practitioner densities. The leave-one-out analysis (Appendix 6, https://links.lww.com/AOG/E187) further supported the stability of findings, showing that the exclusion of any single state did not alter the estimates. Finally, the analysis of the mortality of female individuals of reproductive age (15–44 years) (Appendix 7, https://links.lww.com/AOG/E187) showed that desert counties, low-access counties, and moderate-access counties had an increased risk of mortality compared with full-access counties, suggesting that the misclassification of pregnant individuals in the mortality data would not affect the interpretation of the results.

DISCUSSION

In this cross-sectional analysis using national U.S. data sets, we found that maternity care desert counties had significantly higher maternal and pregnancy-related mortality compared with full-access counties. Low- and moderate-access counties had rates comparable with those of full-access counties for both maternal and pregnancy-related mortality.

Each year, approximately 750 individuals in the United States lose their lives to pregnancy-related causes, the majority of which are preventable.30,31 Maternal health outcomes in the United States have worsened significantly in recent years. Maternal mortality rates have risen from 17.4 per 100,000 live births in 2018 to 23.8 per 100,000 live births in 2020, with Black women experiencing nearly three times the mortality rate of White women.2 An estimated 2.2 million individuals of reproductive age lived in counties with no access to maternity care in 2022.13 Prior research in Louisiana has shown that individuals residing in maternity care deserts face significantly elevated risks of maternal mortality and pregnancy-related mortality.14 A national study by Barrera et al32 further demonstrated that county-level contextual indicators, including poverty, unemployment, food insecurity, and limited health care access, were significantly associated with increased pregnancy-related mortality, with particularly pronounced effects among Black and Hispanic populations. Unlike prior investigations that relied on state-level or socioeconomic proxies, our study directly quantifies the association between county-level maternity care access and maternal mortality nationwide. By integrating a nationally representative typology of maternity care deserts with mortality data, we furnish evidence that structural access to obstetric services is independently associated with maternal and pregnancy-related deaths.

National trends further reflect these disparities, with more than 400 maternity services closing between 2006 and 2020, particularly in rural areas, where 9% of counties lost hospital maternity services between 2004 and 2014 and 45% never had maternity services.33 The ongoing closures of maternity units not only reduce access to skilled practitioners but also exacerbate staffing shortages because obstetricians and nurse–midwives are forced to relocate, creating and perpetuating maternity care deserts.

The geographic maldistribution of maternity care practitioners necessitates targeted policy and health care delivery interventions. Strategies to increase maternity care practitioner density in desert counties, particularly in rural and low-resource settings, could include loan repayment programs, scholarships for trainees with rural backgrounds, and tax incentives for health care practitioners willing to practice in underserved counties.34 Expanding ob-gyn residency slots in underserved areas may also help mitigate these disparities over time. Moreover, building partnerships between urban teaching hospitals and rural health care facilities could further enhance access, enabling resource sharing and on-demand consultations for high-risk pregnancies.

In counties with persistent shortages, telemedicine offers an opportunity to extend maternity care services. Remote consultations, tele-ultrasonography, and wearable health technologies for monitoring maternal and fetal well-being could provide timely care while minimizing geographic barriers.35,36 Furthermore, mobile maternity care units equipped with diagnostic tools could be deployed to underserved counties to provide prenatal and postpartum care, addressing critical gaps in health care access.

The lack of a clear stepwise gradient in maternal mortality across low-, moderate-, and full-access counties likely reflects constraints in our access taxonomy. By design, the March of Dimes categories capture structural supply (facility and practitioner counts) but omit critical functional dimensions—quality of care, practitioner capacity, and intercounty resource sharing—that can offset lower supply. Consequently, counties with relatively few obstetric resources may still achieve favorable outcomes through robust referral networks, telehealth services, or proximity to higher-access neighbors. Further research is needed to investigate the mechanisms by which the level of maternity access influences maternal outcomes. Although this study highlights associations at the county level, longitudinal analyses are required to determine causal pathways. Key questions include whether improved outcomes result primarily from earlier prenatal care, enhanced emergency interventions, or better integration of multidisciplinary teams.

Socioeconomic disadvantage, limited health literacy, and restricted access to advanced prenatal diagnostics such as genetic testing or high-resolution ultrasonography may exert a stronger influence on maternal and perinatal outcomes than practitioner density alone.37 The endurance of maternity care deserts is also shaped by structural racism and policy-level forces. Obstetric unit closures often arise from reimbursement policies, minimum delivery-volume requirements, and the wider consolidation of rural health care systems.38,39 Incomplete Medicaid expansion, inequitable practitioner reimbursement, and elevated medicolegal liability further disincentivize hospitals from sustaining obstetric services. Clarifying how these socioeconomic, structural, and policy drivers intersect is essential for developing targeted, durable solutions.

Our study has several strengths. First, it leveraged large, nationally representative data sets, providing a comprehensive assessment of maternity care practitioner density and maternal outcomes at the county level. This localized focus enhances the relevance of our findings to policy and practice, particularly in underserved regions. Second, the use of established practitioner categories in this study ensures applicability to the broader maternity care landscape. Finally, the inclusion of multiple sensitivity analyses strengthens the reliability of our findings across different thresholds and modeling approaches.

Nonetheless, our study is not without limitations. Fatalities temporally linked to pregnancy such as homicide, suicide, or drug overdose are often coded as external causes on death certificates; as a result, they may be excluded from the CDC's pregnancy-related mortality classification. To account for possible misclassification, we also analyzed all-cause mortality among reproductive-aged female individuals; the findings mirrored those of our primary analysis. Furthermore, the Underlying Cause of Death data set is subject to known limitations, including misclassification of maternal deaths as a result of variability in death certificate reporting and International Classification of Diseases, Tenth Revision coding practices. Moreover, the surge in mortality during the COVID-19 pandemic may have accentuated existing disparities. To mitigate this temporal confounding, we included birth year as a covariate. Notably, as a cross-sectional analysis, it precludes causal inferences regarding the observed relationships between maternity care access and maternal outcomes. Although we accounted for socioeconomic and demographic factors, unmeasured confounders such as transportation barriers, health care infrastructure, and cultural factors may influence the results. Our definition of maternity care practitioners did not include family physicians, who may provide obstetric services in many rural areas, because the Area Health Resources File does not indicate whether they provide obstetric care. Lastly, although our maternity care access measure adopted the March of Dimes classification, it does not capture the intracounty spatial distribution of services. In geographically extensive or sparsely populated rural counties, residents may still face substantial travel distances to the nearest practitioner or delivery facility, even when overall density metrics suggest adequate coverage.

In conclusion, residing in a maternity care desert is significantly associated with higher rates of maternal and pregnancy-related mortality. Expanding access to comprehensive maternity services in these counties is a critical step toward improving maternal health outcomes and narrowing persistent inequities.

[END]
---
[1] Url: https://journals.lww.com/greenjournal/fulltext/2025/08000/association_of_maternity_care_deserts_with.2.aspx#:~:text=Residing%20in%20a%20maternity%20care,%2C%20and%20certified%20nurse%E2%80%93midwives.

Published and (C) by Daily Yonder - Keep it Rural
Content appears here under this condition or license: Creative Commons CC BY-ND 4.0 International.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailyyonder/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.