Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jan 1.
Published in final edited form as: Matern Child Health J. 2023 Nov 8;28(1):165–176. doi: 10.1007/s10995-023-03828-9

State-Level Indicators of Structural Racism and Severe Adverse Maternal Outcomes During Childbirth

Jean Guglielminotti 1, Goleen Samari 2, Alexander M Friedman 3, Ruth Landau 1, Guohua Li 1,4
PMCID: PMC11299521  NIHMSID: NIHMS2008444  PMID: 37938439

Abstract

Objectives

Structural racism (SR) is viewed as a root cause of racial and ethnic disparities in maternal health outcomes. However, evidence linking SR to increased odds of severe adverse maternal outcomes (SAMO) is scant. This study assessed the association between state-level indicators of SR and SAMO during childbirth.

Methods

Data for non-Hispanic Black and non-Hispanic white women came from the US Natality file, 2017–2018. The exposures were state-level Black-to-white inequity ratios for lower education level, unemployment, and prison incarceration. The outcome was patient-level SAMO, including eclampsia, blood transfusion, hysterectomy, or intensive care unit admission. Adjusted odds ratios (aORs) of SAMO associated with each ratio were estimated using multilevel models adjusting for patient, hospital, and state characteristics.

Results

A total of 4,804,488 birth certificates were analyzed, with 22.5% for Black women. SAMO incidence was 106.4 per 10,000 (95% CI 104.5, 108.4) for Black women, and 72.7 per 10,000 (95% CI 71.8, 73.6) for white women. Odds of SAMO increased 35% per 1-unit increase in the unemployment ratio for Black women (aOR 1.35; 95% CI 1.04, 1.73), and 16% for white women (aOR 1.16; 95% CI 1.01, 1.33). Odds of SAMO increased 6% per 1-unit increase in the incarceration ratio for Black women (aOR 1.06; 95% CI 1.03, 1.10), and 4% for white women (aOR 1.04; 95% CI 1.02, 1.06). No significant association was observed between SAMO and the lower education level ratio.

Conclusions for Practice

State-level Black-to-white inequity ratios for unemployment and incarceration are associated with significantly increased odds of SAMO.

Keywords: Maternal morbidity, Structural racism, Childbirth, Epidemiology, Racial and ethnic disparities

Objectives

Addressing racial and ethnic disparities in severe maternal morbidity (SMM) and maternal mortality has been recognized as an urgent public health priority (Howell et al., 2018). SMM refers to unintended adverse outcomes of labor and delivery and is associated with substantially increased risk of maternal death (American College of Obstetricians Gynecologistset al., Medicine, Kilpatrick SK & Ecker JL, 2016). The reported incidence of SMM during childbirth and the postpartum has increased during the last two decades in all racial and ethnic groups, from 0.8% in 1999 to 1.6% in 2017 (Guglielminotti et al., 2021a). SMM affects about 50,000 women annually, with up to 20% of SMM cases occurring during the postpartum (Chen et al., 2021; Declercq et al., 2022). Compared with non-Hispanic white women, non-Hispanic Black women are up to three times more likely to experience SMM (Admon et al., 2018; Guglielminotti et al., 2021a). Furthermore, racial and ethnic disparities vary markedly across states, ranging from a minimum 1.2-fold increase in SMM among non-Hispanic Black women in Iowa to a maximum 3.9-fold increase in Utah (Admon et al., 2023; Healthcare Cost and Utilization Project. HCUP fast stats: severe maternal morbidity (SMM) among in-hospital deliveries 2022) Of all pregnancy-related deaths, about 70% are preventable and up to 50% occur during the postpartum (Petersen et al., 2019a). Like SMM, non-Hispanic Black women are up to three times more likely to die from pregnancy-related causes than non-Hispanic white women (MacDorman et al., 2021; Petersen et al., 2019b), and the racial and ethnic disparity in maternal mortality has worsened during the COVID-19 pandemic (Hoyert, 2020). Because of the continuum between SMM and maternal mortality, addressing racial and ethnic disparities in SMM will help reduce racial and ethnic disparities in maternal mortality.

Structural racism is viewed as a root cause of racial and ethnic disparities in healthcare access, utilization, and outcomes, independent of socioeconomic determinants of health (Bailey et al., 2021; Crear-Perry et al., 2021; Taylor, 2020). Structural racism refers to a system where public policies, institutional practices, cultural representations, and other norms work together to maintain and perpetuate racial group inequities in housing, education, employment, earnings, benefits, credit, media, health care, or criminal justice (Bailey et al., 2017; Devakumar et al., 2022; Yearby, 2020). In maternal and child health research, structural racism has been assessed using indicators of residential segregation (e.g., Indice of Concentration at the Extremes) and, more recently, using Black-to-white inequity ratios in domains encompassing the systematic exclusion of Black women from resources and opportunities (e.g., unemployment) or their unfair judiciary treatment (e.g., incarceration). A large body of research reports an association between these indicators measured at the ZIP code, county, or state levels and premature birth, low birth weight, and infant or child mortality (Chambers et al., 2018; Krieger et al., 2018; Vilda et al., 2021; Wallace et al., 2015, 2017, 2019). However, research assessing the association of these indicators with SMM or maternal mortality is more limited (Dyer et al., 2022; Janevic et al., 2020; Liu et al., 2019; Mari et al., 2023). To help close this research gap, we conducted this nationwide study using birth certificate data to assess the association between state-level Black-to-white inequity ratios and patient-level severe adverse maternal outcomes (SAMO) during childbirth in non-Hispanic Black and in non-Hispanic white women.

Methods

The study protocol was deemed exempt by the Institutional Review Board of the authors’ institution. The study is reported according to STROBE guidelines.

Data System

Data for this study were abstracted from US birth certificates 2017–2018 contained in the restricted access Natality File of the National Vital Statistics System, National Center for Health Statistics, Centers for Diseases Control and Prevention. The Natality File is based on the 2003 revised US Standard Certificate of Live Birth. As of January 2015, the 2003 revised US Standard Certificate of Live Birth was implemented in the 50 US states and the District of Columbia. The Natality file is a census of US births and contain comprehensive information on the pregnant woman, labor, and delivery. It also provides county and state identifiers for the woman’s residence and for the delivery hospital that allow to abstract county and state characteristics from other data systems.

Study Sample

The study sample included birth certificates in 2017 and 2018 for non-Hispanic white and non-Hispanic Black women in the 50 states and the District of Columbia. US territories were not included. Exclusion criteria were: (1) woman not residing in the US; (2) maternal state of residence not corresponding to state of delivery; (3) birth not occurring in a hospital; (4) maternal race and ethnicity not corresponding to non-Hispanic white or to non-Hispanic Black; (5) missing information on maternal outcomes; and (6) missing information on county or state identifiers for maternal residence or hospital of delivery.

Exposure

Based on previous research, the exposures of interest were state-level indicators of structural racism, defined as non-Hispanic Black to non-Hispanic white inequity ratios (Lukachko et al., 2014). For the purpose of the study, three Black-to-white inequity ratios were assessed (1) ratio for lower education level defined as less than high-school diploma; (2) ratio for unemployment rate; and (3) ratio for prison incarceration rate. A higher ratio value indicates more structural racism. Details of the calculation of the ratios, data sources, and data years are presented in Appendix 1. Ratios were available for the 50 states and DC except for the prison incarceration ratio in DC. We did not include Black-to-white inequity ratios for political participation (e.g., persons not registered to vote) because of a high proportion of missing values. We did not include indicators of residential segregation (e.g., Indices of Concentration at the Extremes) because our analysis was at the state-level and residential segregation is a relevant indicator at the county- or ZIP code-level but not at the state level.

For incarceration, we deliberately limited the analysis to prisons, which are under state or federal control, and did not include jails, which are under local or county control, to allow analysis of state-level indicators of structural racism. In addition, since prisoners can be imprisoned in a state different from the state where they reside and were sentenced, we conducted 2 sensitivity analyses. First, we limited the study sample to 35 states with available information on the proportion of prisoners coming from another state (Kaufman, 2020). Second, we excluded 4 states (Hawaii, Vermont, New Hampshire, and Wyoming) from these 35 states because these 4 states had a proportion of prisoners coming from another state greater than 5%.

At the time of this study, data for calculating the three Black-to-white inequity ratios were available only for years 2013–2017. We chose to limit our study period to 2017–2018 to ensure that the assessment of the exposure (inequity ratios) occurred before the assessment of the outcome (SAMO).

Outcome

The outcome was SAMO defined as the presence of at least one of four diagnoses or procedures: (1) blood transfusion, (2) eclampsia, (3) hysterectomy, and (4) intensive care unit (ICU) admission. Birth certificates do not contain codes of the International Classification of Diseases (ICD), precluding the assessment of severe maternal morbidity as defined by the US Centers for Disease Control and Prevention (CDC) (Centers for Disease Control & Prevention. How does CDC identify severe maternal morbidity, 2019).

The four conditions and procedures are recorded in specific check boxes on the birth certificate. The reported sensitivity of the individual components in a study conducted in Massachusetts in 2011–2013 and using administrative hospital discharge data as the gold standard, ranges from 12% for blood transfusion to 51% for hysterectomy; the positive predictive value ranges from 29% for ICU admission to 73% for blood transfusion (Luke et al., 2018). We are not aware of a validation study using national data.

Maternal, Hospital, and State Characteristics

Maternal characteristics and comorbidities directly recorded from birth certificate data included: age (≤ 19, 20–29, 30–39, or ≥ 40 years); education level (less than high school, high school with no diploma, high school graduate or GED, or college and higher); health insurance (Medicaid, private, self-pay, or other); pre-pregnancy body mass index (continuous, kg/m2); and preexisting or gestational diabetes or hypertension (binary, yes/no). The following maternal characteristics were estimated at the county of residence level and abstracted from the Area Health Resource File (AHRF): residence location (urban, suburban, or rural), proportion of persons in poverty (continuous, %), and proportion of persons unemployed (continuous, %).

Pregnancy, labor, and delivery characteristics directly recorded from birth certificate data included: previous cesarean section (binary); month of gestation prenatal care began (1st -3rd, 4th–6th, ≥ 7th, or no prenatal care); number of prenatal visits (continuous); delivery during a weekend (binary); mother transferred in (binary); nulliparous (binary); gestational age at delivery (≤ 33 completed weeks, 34–38 completed weeks, or ≥ 39 completed weeks); multiple gestation (binary); non-cephalic presentation (binary); induction of labor (binary); use of labor neuraxial analgesia (binary); attendant at birth (doctor of Medicine, doctor of Osteopathy, midwife, or other); delivery mode (vaginal spontaneous, vaginal assisted (vacuum or forceps), or cesarean); and birth weight (continuous, grams).

The following hospital characteristics were estimated at the hospital county level and abstracted from the AHRF: hospital location (urban, suburban, or rural), number of hospital births (continuous), and number of obstetricians and gynecologists (continuous, per 1000 hospital births).

The state-level proportions of persons below poverty level, persons with less than high school diploma, and persons unemployed were directly abstracted from the AHRF data. The state-level proportion of non-Hispanic Black residents was calculated as the ratio of the number of non-Hispanic Black residents to the total population estimate abstracted from the AHRF data. The state imprisonment rate was directly abstracted from the Sentencing Project data and expressed per 10,000 state residents. The state Medicaid income eligibility thresholds for a pregnant person and for a family of 3 were directly abstracted from the State Health Facts of the Kaiser Family Foundation data and expressed as a percent of the Federal Poverty Level (FPL).

Statistical Analysis

Statistical analysis was performed with R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria) and the package ‘lme4’ for mixed-effect modeling. No a priori study power was performed.

Descriptive Statistics

The three inequity ratios were compared between women with SAMO and women without SAMO using the absolute standardized difference (ASD), independently among Black and white women. An ASD greater than 10% is indicative of a significant imbalance between groups. The relationship between each of the three inequity ratios, categorized into deciles, and the incidence of SAMO was analyzed visually using plots independently for Black and white women.

Crude Odds Ratios of SAMO Associated with Structural Racism Indicators

Crude odds ratios (OR) and 95% confidence interval (CI) of SAMO associated with each of the three inequity ratios were estimated using univariate mixed-effect logistic regression models, independently for Black and white women. In these models, SAMO was the dependent variable, the inequity ratio was the independent variable, and the hospital county nested within the hospital state was the random effect (random intercept and constant slope). The significance of the OR was assessed using the 95% CI.

Adjusted Odds Ratios of SAMO Associated with Structural Racism Indicators

To identify variables required to adjust the odds ratio of SAMO, we developed a multivariable prediction model for SAMO using data from both Black and white women. Candidate variables included in the model were characteristics with an ASD > 10% in the comparison of women with and without SAMO or deemed clinically relevant (Appendix 2 and Appendix 3). To take into consideration current recommendations on adjustment, race and ethnicity were not included as a candidate variables (Kleinman & Howell, 2022). We used mixed-effect logistic regression modeling with SAMO as the dependent variable, the candidate variables as independent variables, and the hospital county nested within the hospital state as the random effect. Selection used a backward procedure with a significance threshold at 0.05. We conducted a complete case analysis with 308,643 certificates excluded (6.4%). Results of the multivariable prediction model are presented in Appendix 4.

Adjusted ORs of SAMO and of its 4 individual components associated with each of the three inequity ratios were estimated independently for Black and white women using multivariate mixed-effect logistic regression models, adjusting for the variables identified in the multivariable prediction model for SAMO developed in the previous step. The significance of the OR was assessed using the 95% CI. Black and white women were compared using the regression coefficient (β) of an interaction term between race and ethnicity and the inequity ratio examined.

Number of Potentially Preventable SAMO Cases

For the Black-to-white ratio for the prison incarceration rate, states were categorized into terciles, with states in the first tercile corresponding to those with the lowest values of the ratio and states in the third tercile to those with the highest values. Then, we estimated the potential reduction in SAMO cases in states in the second and third tercile if the SAMO rate in these states was reduced to the SAMO rate in states in the first tercile. The number of potentially preventable SAMO cases was calculated as the difference between the number of observed SAMO cases in states in the second and third tercile and the expected number of SAMO cases if these states had an adjusted SAMO rate similar to states in the first tercile. The adjusted SAMO rate was calculated using the multivariable prediction model for SAMO developed in the previous step.

Results

Of the 4,804,488 birth certificates included, 1,081,078 were for Black women (22.5%) (Appendix 5). The incidence of SAMO in white women was 72.7 per 10,000 (95% CI: 71.8, 73.6) (Table 1). Compared to white women, the incidence of SAMO was higher in Black women (106.4 per 10,000; 95% CI: 104.5, 108.4). A higher incidence in Black women was observed for each of the four diagnosis or procedure reported for SAMO (eclampsia, blood transfusion, hysterectomy, and ICU admission).

Table 1.

Incidence of severe adverse maternal outcomes during childbirth (United States, 2017–2018)

Non-Hispanic black women (n = 1,081,078)
 Non-Hispanic white women (n = 3,723,410)
Number of cases Incidence (Per 10,000; 95% CI) Number of cases Incidence (Per 10,000; 95% CI)
Severe adverse maternal outcomes 11,506 106.4 (104.5, 108.4) 27,065 72.7 (71.8, 73.6)
Blood transfusion 5520 51.1 (49.7, 52.4) 13,906 37.3 (36.7, 38.0)
Eclampsia 3969 36.7 (35.6, 37.9) 9404 25.3 (24.7, 25.8)
Intensive care unit admission 2569 23.8 (22.8, 24.7) 5254 14.1 (13.7, 14.5)
Hysterectomy 568 5.2 (4.8, 5.7) 1561 4.2 (4.0, 4.4)
Open in a new tab

CI confidence interval

A marked variation in the three Black-to-white inequity ratios was observed across states (Fig. 1A–C). The greatest variation was for the ratio for the prison incarceration rate that ranged from a minimum of 2.4 in the state of Hawaii to a maximum of 12.1 in the state of New Jersey (Fig. 1C).

Fig. 1.

Fig. 1

Open in a new tab

Distribution of the state-level Black-to-white inequity ratios in the 50 US states and the District of Columbia. A higher ratio value indicates more structural racism. The thin dashed red lines indicate the mean value. The information for the Black-to-white ratio for the prison incarceration rate is missing for the District of Columbia. Abbreviations: BWR Black-to-white ratio. A BWR for lower education level (less than high-school diploma). B BWR for unemployment rate. C BWR for prison incarceration rate

Compared to Black women without SAMO, Black women with SAMO had higher values (ASD > 10%) for two of the three Black-to-white inequity ratios (unemployment and incarceration) (Table 2), with no difference (ASD ≤ 10%) for lower education level. For white women, indicators values did not differ significantly between women with and without SAMO. For Black women, the relationship between the deciles of the inequity ratios and the incidence of SAMO revealed an almost linear increase in the incidence of SAMO with increased inequity ratios for unemployment and incarceration (Appendix 6). No obvious relationship was observed between the inequity ratio for lower education level and the incidence of SAMO. For white women, no obvious relationship was observed between the inequity ratios and the incidence of SAMO.

Table 2.

Comparison of the state-level indicators of structural racism between women with and without severe adverse maternal outcomes (United States, 2017–2018)

Black-to-white inequity ratio Non-Hispanic Black women (N = 1,081,078)
 Non-Hispanic white women (N = 3,723,410)
Women without SAMO (N = 1,069,572) Women with SAMO (N = 11,506) ASD (%) Women with-out SAMO (N = 3,696,345) Women with SAMO (N = 27,065) ASD (%)
Lower education level (less than high-school diploma) 2.07 (1 sd, 1.96) 2.00 (1 sd, 1.48) 4.2 2.02 (1 sd, 0.99) 1.99 (1 sd, 0.77) 4.0
Unemployment rate 2.10 (1 sd, 0.29) 2.13 (1 sd, 0.29) 10.0 2.09 (1 sd, 0.30) 2.11 (1 sd, 0.32) 6.9
Prison incarceration rate 5.54 (1 sd, 2.42) 5.91 (1 sd, 2.48) 15.2 6.23 (1 sd, 2.49) 6.37 (1 sd, 2.46) 5.6
Open in a new tab

Bold values for the absolute standardized difference indicate a value greater than 10%

Indicators are non-Hispanic Black to non-Hispanic white inequity ratios in the three domains described above. A higher ratio value indicates more structural racism. Results are expressed as mean (one standard deviation). An absolute standardized difference greater than 10% indicates a clinically relevant imbalance between groups

ASD absolute standardized difference, SAMO severe adverse maternal outcomes, sd standardized deviation

For Black women (Table 3 and Fig. 2), the adjusted OR of SAMO associated with a 1-unit increase in the inequity ratio for unemployment was 1.35 (95% CI 1.04, 1.73), and with a 1-unit increase in the inequity ratio for incarceration 1.06 (95% CI 1.03, 1.10). Increased odds associated with the inequity ratios for unemployment and incarceration was observed for blood transfusion, hysterectomy, and ICU admission but not for eclampsia (Table 4). No significant association was observed between SAMO and the inequity ratio for lower education level (adjusted OR 0.99; 95% CI 0.87, 1.14).

Table 3.

Crude and adjusted odds ratios of severe adverse maternal outcomes associated with state-level indicators of structural racism (United States, 2017–2018)

Black-to-white inequity ratio Crude OR (95% CI) (a) Adjusted OR (95% CI) (b) Adjusted regression coefficient (95% CI) for the interaction between the inequity ratio and maternal race and ethnicity
Lower education level(less than high school diploma)
 Non-Hispanic white 0.99 (0.95, 1.03) 1.01 (0.94, 1.09) –
 Non-Hispanic Black 0.98 (0.89, 1.08) 0.99 (0.87, 1.14) Not applicable
Unemployment rate
 Non-Hispanic white 1.19 (1.11, 1.27) 1.16 (1.01, 1.33) –
 Non-Hispanic Black 1.38 (1.25, 1.53) 1.35 (1.04, 1.73) 0.095 (0.005, 0.184)
Prison incarceration rate
 Non-Hispanic white 1.04 (1.02, 1.05) 1.04 (1.02, 1.06) –
 Non-Hispanic Black 1.05 (1.03, 1.08) 1.06 (1.03, 1.10) 0.030 (0.020, 0.040)
Open in a new tab

Odds ratios are per 1-unit increase in the indicator. Indicators are non-Hispanic Black to non-Hispanic white inequity ratios in the three domains described above

CI confidence interval, OR odds ratio

(a)

Estimated using univariate mixed-effect logistic regression with the hospital county nested within the hospital state as the random effect

(b)

Adjusted for maternal age, health insurance, body mass index, residence (rural, suburban, urban), preexisting diabetes, preexisting hypertension, gestational hypertension, previous cesarean section, month prenatal care began, number of prenatal visits, mother transferred in, gestational age at delivery, multiple gestation, non-cephalic presentation, attendant at birth, delivery mode, birth weight, hospital location (rural, suburban, urban), and state Medicaid income eligibility threshold

Fig. 2.

Fig. 2

Open in a new tab

Adjusted relationship between state-level Black-to-white inequity ratios and severe adverse maternal outcomes (in a logit form) in non-Hispanic Black women (red color; left-hand side panels) and in non-Hispanic white women (green color; right-hand side panels). Each grey point represents one patient. Beta (β) is the regression coefficient from a mixed-effect logistic regression model adjusting for patient, hospital, and state characteristics. The thick line is the corresponding regression line, and the thin dotted line is the identity line. BWR: Black-to-white ratio. A BWR for lower education level (less than high-school diploma). B BWR for unemployment rate. C BWR for prison incarceration rate

Table 4.

Adjusted odds ratios of the four individual adverse maternal outcomes associated with state-level indicators of structural racism (United States, 2017–2018)

Black-to-white inequity ratio Blood transfusion aOR (95% CI) (a) ICU admission aOR (95% CI) (a) Hysterectomy aOR (95% CI) (a) Eclampsia aOR (95% CI) (a)
Lower education level(less than high school diploma)
 Non-Hispanic white 1.06 (0.96, 1.16) 1.12 (1.03, 1.23) 1.19 (1.05, 1.36) 0.92 (0.79, 1.07)
 Non-Hispanic Black 1.07 (0.89, 1.30) 1.08 (0.92, 1.28) 1.25 (0.97, 1.61) 0.84 (0.64, 1.01)
Unemployment rate
 Non-Hispanic white 1.54 (1.30, 1.84) 1.36 (1.15, 1.61) 1.24 (0.97, 1.59) 0.75 (0.57, 0.98)
 Non-Hispanic Black 2.07 (1.44, 2.97) 1.51 (1.10, 2.06) 1.61 (0.97, 2.67) 0.68 (0.42, 1.10)
Prison incarceration rate
 Non-Hispanic white 1.08 (1.05, 1.10) 1.06 (1.03, 1.08) 1.06 (1.03, 1.10) 0.99 (0.95, 1.03)
 Non-Hispanic Black 1.10 (1.10, 1.10) 1.05 (1.01, 1.09) 1.11 (1.05, 1.17) 0.97 (0.91, 1.03)
Open in a new tab

Odds ratios are per 1-unit increase in the indicator. Indicators are non-Hispanic Black to non-Hispanic white inequity ratios in the three domains described above

aOR adjusted odds ratio, CI confidence interval, ICU intensive care unit

(a)

Odds ratios are estimated using mixed-effect logistic regression with the hospital county nested within the hospital state as the random effect, and adjusted for maternal age, health insurance, body mass index, residence (rural, suburban, urban), preexisting diabetes, preexisting hypertension, gestational hypertension, previous cesarean section, month prenatal care began, number of prenatal visits, mother transferred in, gestational age at delivery, multiple gestation, non-cephalic presentation, attendant at birth, delivery mode, birth weight, hospital location (rural, suburban, urban), and state Medicaid income eligibility threshold

For white women, the adjusted OR of SAMO associated with a 1-unit increase in the inequity ratio for unemployment was 1.16 (95% CI 1.01, 1.33), and with a 1-unit increase in the inequity ratio for incarceration 1.04 (95% CI 1.02, 1.06). Increased odds associated with the inequity ratios for unemployment and incarceration was observed for blood transfusion, hysterectomy, and ICU admission but not for eclampsia (Table 4). No significant association was observed between SAMO and the inequity ratio for lower education level (adjusted OR 1.01; 95% CI 0.94, 1.09).

For the inequity ratio for incarceration, results were robust after exclusion of states with a proportion of prisoners from another state greater than 5% (Appendix 7).

For Black women, reducing the SAMO rate in states in the second and third terciles of the inequity ratio for incarceration to the SAMO rate in states in the first tercile could potentially prevent 4597 SAMO cases or 34% of observed SAMO cases (Appendix 8). For white women, it could potentially prevent 3569 SAMO cases or 20% of observed SAMO cases.

Conclusions for Practice

Principal Findings

In this nationwide study using recent birth certificate data, we evaluated three state-level indicators of structural racism to assess the association between Black-to-white inequalities and SAMO. We found that higher Black-to-white inequalities in unemployment and incarceration measured at the state level were associated with significantly increased odds of SAMO in Black women and, to a lesser extent, in white women. For Black women, reducing the SAMO rate to the level observed in states with the lowest indicators of structural racism could prevent up to 1 in 3 SAMO cases, and for white women up to 1 in 5 SAMO cases.

Results in the Context of What is Known, and Implications

To date, empirical evidence linking indicators of structural racism to racial and ethnic disparities in adverse maternal health outcomes is limited to four studies. In the first study conducted in New York state, Liu et al. report that a high Black-to-white inequality in education achievement measured at the county of residence level is associated with a significantly increased risk of severe maternal morbidity, as defined by the CDC, in non-Hispanic Black women (Liu et al., 2019). In the second and third studies conducted in New York city and in Philadelphia, Janevic et al. and Mari et al. found that residential segregation measured using the Indices of Concentration at the Extremes was associated with significantly increased risk of severe maternal morbidity among non-Hispanic Black women (Janevic et al., 2020; Mari et al., 2023). In the fourth study conducted in Louisiana, Dyer et al. found that non-Hispanic Black women living in racially and economically polarized neighborhoods had a significantly increased risk of maternal mortality (Dyer et al., 2022). Our study confirms the association between state-level indicators of structural racism and racial and ethnic disparities in SAMO, and extends this association to the entire United States.

A notable finding from our study is the increased risk of SAMO associated with structural racism indicators for non-Hispanic white women. Indeed, structural racism is hypothesized to confer a benefit to the non-Hispanic white population. For example, Lukachko et al. report that non-Hispanic white people living in high structural racism states experienced null or lower odds of myocardial infarction compared to non-Hispanic white people living in low structural racism states (Lukachko et al., 2014). However, structural racism is now suggested to also negatively impact the non-Hispanic white population (Malat et al., 2018; McGhee, 2021). A potential explanation is that structural inequities impact all people that enter the healthcare system, because the system is not operating at an optimal level when racism undermines policies, practices, and procedures. In other words, even white women could be harmed by a broken system.

A possible link between the Black-to-white ratio for unemployment and increased risk of SAMO in Black women reported in our study may be the higher rate of uninsured among unemployed mothers and the subsequent lower access to healthcare. In the United Sates, health insurance coverage is heavily dependent on employment, and previous research has repeatedly reported higher uninsurance rates in non-Hispanic Black people, including during the perinatal period. Policies aiming at increasing insurance coverage for low-income people such as the 2014 Medicaid expansion under the Affordable Care Act have been associated with decreased racial and ethnic disparities in health insurance coverage and healthcare utilization (Bellerose et al., 2022; Sun et al., 2022), and in improved maternal outcomes (Guglielminotti et al., 2021b, 2023). Mass incarceration disproportionally affects non-Hispanic Black men and having a family member in prison harms the mental and physical health of the non-incarcerated partner and children, as well as the community health (Sundaresh et al., 2021; Wildeman, 2012; Yi et al., 2021). Mechanisms linking paternal incarceration to the health of the pregnant partner are still poorly understood and may include income and job loss, housing instability, parenting and relationship strain, and financial costs such bail (Dyer et al., 2019; Jahn et al., 2020; Wildeman & Wang, 2017). The absence of a significant association between the Black-to-white ratio for low education level and SAMO is perplexing and is inconsistent with the results of the Liu et al. study conducted in New York, in which the inequity ratio was measured at the residence county level (Liu et al., 2019). The conflicting findings might also be partially explained by the wide variation in the rates of adverse maternal outcomes across states (Admon et al., 2023; Healthcare Cost and Utilization Project, 2022)

Strengths and Limitations

The national birth certificate data allowed us to assess the association of structural racism with SAMO at the state level. The large sample size also enabled meaningful statistical comparisons for relatively rare outcomes across racial and ethnic groups. However, our findings should be interpreted in the context of several limitations. First, our study is observational in nature, and the association between indicators of structural racism and SAMO is not necessarily causal. Second, we were unable to analyze the association between indicators of structural racism and maternal mortality because the information on maternal death is not available in the birth certificate data. Third, the data do not provide the exact indication for blood transfusion, ICU admission, and hysterectomy but previous research suggest that obstetric hemorrhage is the leading indication for these three procedures (Friedman et al., 2016; Gyamfi-Bannerman et al., 2018; Wanderer et al., 2013). Fourth, hospital identifier data is not available and we only included a limited number of hospital characteristics at the hospital county level in our analysis (e.g., number of births), and were not able to adjust for volume of delivery or teaching status, hospital level characteristics that have previously been associated with maternal morbidity (Friedman et al., 2016). Fifth, we limited our analysis to non-Hispanic Black women and excluded other racial and ethnic minority groups. These choices were based on the fact that the information required to compute indicators for some racial and ethnic groups (e.g., prison incarceration rate for Asians) is either not available or missing for too many states. However, non-Hispanic Black women are the racial and ethnic group experiencing the most severe disparities in maternal morbidity and maternal mortality. Last, the sensitivity of using birth certificate data to detect SAMO is fairly low (Luke et al., 2018). However, the underreporting should be non-differential across racial and ethnic groups and thus be unlikely a major source of bias for the estimated ORs.

Conclusion

State-level structural racism indicators of Black-to-white inequalities in unemployment and incarceration are associated with significantly increased odds of SAMO. Eliminating structural racism could prevent one third of SAMO cases, underscoring the urgent need to address structural racism in maternal care.

Supplementary Material

Appendix

Significance.

What is already known on this subject?

Structural racism is viewed as a root cause of racial and ethnic disparities in maternal health outcomes, independent of socioeconomic determinants of health. However, supporting evidence is scant.

What this study adds?

In this population-based study in 2017–2018, state-level Black-to-white inequities in unemployment and incarceration are associated with significantly increased odds of severe adverse maternal outcomes (i.e., blood transfusion, hysterectomy, and ICU admission) during childbirth for non-Hispanic Black women, and to a lesser extent, for non-Hispanic white women. Addressing structural racism in maternal care may contribute to improving maternal health equity.

Funding

Jean GUGLIELMINOTTI is currently supported by grants from the National Institute on Minority Health and Health Disparities (R01 MD018410) and from the National Institute of Mental Health (R21 MH126096), National Institutes of Health, not related to this study.

Footnotes

Competing Interests The authors report no conflict of interest.

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10995-023-03828-9.

References

  1. Admon LK, Auty SG, Daw JR, Kozhimannil KB, Declercq ER, Wang N, & Gordon SH (2023). State variation in severe maternal morbidity among individuals with Medicaid insurance. Obstetrics and Gynecology, 141(5), 877–885. 10.1097/AOG.0000000000005144 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Admon LK, Winkelman TNA, Zivin K, Terplan M, Mhyre JM, & Dalton VK (2018). Racial and ethnic disparities in the incidence of severe maternal morbidity in the United States, 2012–2015. Obstetrics and Gynecology, 132(5), 1158–1166. 10.1097/AOG.0000000000002937 [DOI] [PubMed] [Google Scholar]
  3. American College of Obstetricians Gynecologists, the Society for Maternal-Fetal Medicine, Kilpatrick SK, Ecker JL. (2016). Severe maternal morbidity: Screening and review. American Journal of Obstetrics and Gynecology, 215(3), B17–22. 10.1016/j.ajog.2016.07.050 [DOI] [PubMed] [Google Scholar]
  4. Bailey ZD, Feldman JM, & Bassett MT (2021). How structural racism works - Racist policies as a root cause of U.S. racial health inequities. The New England Journal of Medicine, 384(8), 768–773. 10.1056/NEJMms2025396 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bailey ZD, Krieger N, Agenor M, Graves J, Linos N, & Bassett MT (2017). Structural racism and health inequities in the USA: Evidence and interventions. Lancet, 389(10077), 1453–1463. 10.1016/S0140-6736(17)30569-X [DOI] [PubMed] [Google Scholar]
  6. Bellerose M, Collin L, & Daw JR (2022). The ACA Medicaid expansion and perinatal insurance, health care use, and health outcomes: A systematic review. Health Affairs (millwood), 41(1), 60–68. 10.1377/hlthaff.2021.01150 [DOI] [PubMed] [Google Scholar]
  7. Centers for Disease Control and Prevention. (2019). How does CDC identify severe maternal morbidity? Retrieved August 16, 2023, from https://www.cdc.gov/reproductivehealth/maternalinfanthealth/smm/severe-morbidity-ICD.htm
  8. Chambers BD, Erausquin JT, Tanner AE, Nichols TR, & Brown-Jeffy S (2018). Testing the association between traditional and novel indicators of county-level structural racism and birth outcomes among Black and White women. Journal of Racial and Ethnic Health Disparities, 5(5), 966–977. 10.1007/s40615-017-0444-z [DOI] [PubMed] [Google Scholar]
  9. Chen J, Cox S, Kuklina EV, Ferre C, Barfield W, & Li R (2021). Assessment of incidence and factors associated with severe maternal morbidity after delivery discharge among women in the US. JAMA Network Open, 4(2), e2036148. 10.1001/jamanetworkopen.2020.36148 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Crear-Perry J, Correa-de-Araujo R, Lewis Johnson T, McLemore MR, Neilson E, & Wallace M (2021). Social and structural determinants of health inequities in maternal health. Journal of Women’s Health, 30(2), 230–235. 10.1089/jwh.2020.8882 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Declercq ER, Cabral HJ, Cui X, Liu CL, Amutah-Onukagha N, Larson E, Meadows A, & Diop H (2022). Using longitudinally linked data to measure severe maternal morbidity. Obstetrics and Gynecology, 139(2), 165–171. 10.1097/AOG.0000000000004641 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Devakumar D, Selvarajah S, Abubakar I, Kim SS, McKee M, Sabharwal NS, Saini A, Shannon G, White AIR, & Achiume ET (2022). Racism, xenophobia, discrimination, and the determination of health. Lancet, 400(10368), 2097–2108. 10.1016/S0140-6736(22)01972-9 [DOI] [PubMed] [Google Scholar]
  13. Dyer L, Chambers BD, Crear-Perry J, Theall KP, & Wallace M (2022). The index of concentration at the extremes (ICE) and pregnancy-associated mortality in Louisiana, 2016–2017. Maternal and Child Health Journal, 26(4), 814–822. 10.1007/s10995-021-03189-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dyer L, Hardeman R, Vilda D, Theall K, & Wallace M (2019). Mass incarceration and public health: The association between black jail incarceration and adverse birth outcomes among black women in Louisiana. BMC Pregnancy and Childbirth, 19(1), 525. 10.1186/s12884-019-2690-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Friedman AM, Ananth CV, Huang Y, D’Alton ME, & Wright JD (2016). Hospital delivery volume, severe obstetrical morbidity, and failure to rescue. American Journal of Obstetrics and Gynecology, 215(6), 795 e1–e14. 10.1016/j.ajog.2016.07.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Friedman AM, Wright JD, Ananth CV, Siddiq Z, D’Alton ME, & Bateman BT (2016). Population-based risk for peripartum hysterectomy during low- and moderate-risk delivery hospitalizations. American Journal of Obstetrics and Gynecology, 215(5), 640 e1–e8. 10.1016/j.ajog.2016.06.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Guglielminotti J, Daw JR, Friedman AM, Landau R, & Li G (2023). Medicaid expansion and risk of eclampsia. American Journal of Obstetrics & Gynecology MFM, 5(8), 101054. 10.1016/j.ajogmf.2023.101054 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Guglielminotti J, Landau R, & Li G (2021b). The 2014 New York State Medicaid expansion and severe maternal morbidity during delivery hospitalizations. Anesthesia and Analgesia, 133(2), 340–348. 10.1213/ANE.0000000000005371 [DOI] [PubMed] [Google Scholar]
  19. Guglielminotti J, Wong CA, Friedman AM, & Li G (2021a). Racial and ethnic disparities in death associated with severe maternal morbidity in the United States: Failure to rescue. Obstetrics and Gynecology, 137(5), 791–800. 10.1097/AOG.0000000000004362 [DOI] [PubMed] [Google Scholar]
  20. Gyamfi-Bannerman C, Srinivas SK, Wright JD, Goffman D, Siddiq Z, D’Alton ME, & Friedman AM (2018). Postpartum hemorrhage outcomes and race. American Journal of Obstetrics and Gynecology, 219(2), 185 e1–e10. 10.1016/j.ajog.2018.04.052 [DOI] [PubMed] [Google Scholar]
  21. Healthcare Cost and Utilization Project. (2022). HCUP fast stats: severe maternal morbidity (SMM) among in-hospital deliveries. Retrieved August 16, 2023, from https://datatools.ahrq.gov/hcup-fast-stats?count=3&tab=hcupfsse&type=subtab
  22. Howell EA, Brown H, Brumley J, Bryant AS, Caughey AB, Cornell AM, Grant JH, Gregory KD, Gullo SM, Kozhimannil KB, Mhyre JM, Toledo P, D’Oria R, Ngoh M, & Grobman WA (2018). Reduction of peripartum racial and ethnic disparities: A conceptual framework and maternal safety consensus bundle. Obstetrics and Gynecology, 131(5), 770–782. 10.1097/AOG.0000000000002475 [DOI] [PubMed] [Google Scholar]
  23. Hoyert DL (2020). Maternal mortality rates in the United States, 2020. Retrieved August 16, 2023, from https://stacks.cdc.gov/view/cdc/113967
  24. Jahn JL, Chen JT, Agenor M, & Krieger N (2020). County-level jail incarceration and preterm birth among non-Hispanic Black and white U.S. women, 1995–2015. Social Science & Medicine, 250, 112856. 10.1016/j.socscimed.2020.112856 [DOI] [PubMed] [Google Scholar]
  25. Janevic T, Zeitlin J, Egorova N, Hebert PL, Balbierz A, & Howell EA (2020). Neighborhood racial and economic polarization, hospital of delivery, and severe maternal morbidity. Health Affairs (millwood), 39(5), 768–776. 10.1377/hlthaff.2019.00735 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kaufman E (2020). The prisoner trade. Harvard Law Review, 133(6), 1815–1878. [Google Scholar]
  27. Kleinman LC, & Howell EA (2022). Equity and the hazard of veiled injustice: A methodological reflection on risk adjustment. Pediatrics. 10.1542/peds.2020-045948G [DOI] [PubMed] [Google Scholar]
  28. Krieger N, Kim R, Feldman J, & Waterman PD (2018). Using the Index of Concentration at the Extremes at multiple geographical levels to monitor health inequities in an era of growing spatial social polarization: Massachusetts, USA (2010–14). International Journal of Epidemiology, 47(3), 788–819. 10.1093/ije/dyy004 [DOI] [PubMed] [Google Scholar]
  29. Liu SY, Fiorentini C, Bailey Z, Huynh M, McVeigh K, & Kaplan D (2019). Structural racism and severe maternal morbidity in New York State. Clinical Medicine Insights: Women’s Health, 12, 1–8. 10.1177/1179562X198547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lukachko A, Hatzenbuehler ML, & Keyes KM (2014). Structural racism and myocardial infarction in the United States. Social Science and Medicine, 103, 42–50. 10.1016/j.socscimed.2013.07.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Luke B, Brown MB, Liu CL, Diop H, & Stern JE (2018). Validation of severe maternal morbidity on the US certificate of live birth. Epidemiology, 29(4), e31–e32. 10.1097/EDE.0000000000000828 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. MacDorman MF, Thoma M, Declcerq E, & Howell EA (2021). Racial and ethnic disparities in maternal mortality in the United States using enhanced vital records, 20162017. American Journal of Public Health, 111(9), 1673–1681. 10.2105/AJPH.2021.306375 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Malat J, Mayorga-Gallo S, & Williams DR (2018). The effects of whiteness on the health of whites in the USA. Social Science and Medicine, 199, 148–156. 10.1016/j.socscimed.2017.06.034 [DOI] [PubMed] [Google Scholar]
  34. Mari KE, Yang N, Boland MR, Meeker JR, Ledyard R, Howell EA, & Burris HH (2023). Assessing racial residential segregation as a risk factor for severe maternal morbidity. Annals of Epidemiology, 83, 23–29. 10.1016/j.annepidem.2023.04.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. McGhee HC (2021). The sum of us: What racism costs everyone and how we can prosper together. One World. [Google Scholar]
  36. Petersen EE, Davis NL, Goodman D, Cox S, Mayes N, Johnston E, Syverson C, Seed K, Shapiro-Mendoza CK, Callaghan WM, & Barfield W (2019a). Vital signs: Pregnancy-related deaths, United States, 2011–2015, and strategies for prevention, 13 states, 2013–2017. Morbidity and Mortality Weekly Report, 68(18), 423–429. 10.15585/mmwr.mm6818e1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Petersen EE, Davis NL, Goodman D, Cox S, Syverson C, Seed K, Shapiro-Mendoza C, Callaghan WM, & Barfield W (2019b). Racial/ethnic disparities in pregnancy-related deaths-United States, 2007–2016. Morbidity and Mortality Weekly Report, 68(35), 762–765. 10.15585/mmwr.mm6835a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sun EP, Guglielminotti J, Chihuri S, & Li G (2022). Association of Medicaid expansion under the affordable care act with perinatal care access and utilization among low-income women: A systematic review and meta-analysis. Obstetrics and Gynecology, 139(2), 269–276. 10.1097/AOG.0000000000004647 [DOI] [PubMed] [Google Scholar]
  39. Sundaresh R, Yi Y, Harvey TD, Roy B, Riley C, Lee H, Wildeman C, & Wang EA (2021). Exposure to family member incarceration and adult well-being in the United States. JAMA Network Open, 4(5), e2111821. 10.1001/jamanetworkopen.2021.11821 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Taylor JK (2020). Structural racism and maternal health among Black women. The Journal of Law, Medicine & Ethics, 48(3), 506–517. 10.1177/1073110520958875 [DOI] [PubMed] [Google Scholar]
  41. Vilda D, Hardeman R, Dyer L, Theall KP, & Wallace M (2021). Structural racism, racial inequities and urban-rural differences in infant mortality in the US. Journal of Epidemiology and Community Health, 75(8), 788–793. 10.1136/jech-2020-214260 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wallace ME, Crear-Perry J, Green C, Felker-Kantor E, & Theall K (2019). Privilege and deprivation in Detroit: Infant mortality and the index of concentration at the extremes. International Journal of Epidemiology, 48(1), 207–216. 10.1093/ije/dyy149 [DOI] [PubMed] [Google Scholar]
  43. Wallace M, Crear-Perry J, Richardson L, Tarver M, & Theall K (2017). Separate and unequal: Structural racism and infant mortality in the US. Health & Place, 45, 140–144. 10.1016/j.healthplace.2017.03.012 [DOI] [PubMed] [Google Scholar]
  44. Wallace ME, Mendola P, Liu D, & Grantz KL (2015). Joint effects of structural racism and income inequality on small-for-gestational-age birth. American Journal of Public Health, 105(8), 1681–1688. 10.2105/AJPH.2015.302613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Wanderer JP, Leffert LR, Mhyre JM, Kuklina EV, Callaghan WM, & Bateman BT (2013). Epidemiology of obstetric-related ICU admissions in Maryland: 1999–2008. Critical Care Medicine, 41(8), 1844–1852. 10.1097/CCM.0b013e31828a3e24 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wildeman C (2012). Imprisonment and (inequality in) population health. Social Science Research, 41(1), 74–91. 10.1016/j.ssresearch.2011.07.006 [DOI] [PubMed] [Google Scholar]
  47. Wildeman C, & Wang EA (2017). Mass incarceration, public health, and widening inequality in the USA. Lancet, 389(10077), 1464–1474. 10.1016/S0140-6736(17)30259-3 [DOI] [PubMed] [Google Scholar]
  48. Yearby R (2020). Structural racism and health disparities: Reconfiguring the social determinants of health framework to include the root cause. The Journal of Law, Medicine & Ethics, 48(3), 518–526. 10.1177/1073110520958876 [DOI] [PubMed] [Google Scholar]
  49. Yi Y, Kennedy J, Chazotte C, Huynh M, Jiang Y, & Wildeman C (2021). Paternal jail incarceration and birth outcomes: Evidence from New York City, 2010–2016. Maternal and Child Health Journal, 25(8), 1221–1241. 10.1007/s10995-021-03168-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix
NIHMS2008444-supplement-Appendix.doc (3.1MB, doc)

RESOURCES