Economic burden of maternal mortality in the USA, 2018–2020

CDC Wide-ranging Online Data for Epidemiologic Research data extraction shows that 2018–2020, White mothers had the highest percentage of births at about 57% as compared with Hispanic (25%) and Black births (16%). The Hispanic population had the highest birth and fertility rates as compared with White and Black populations (Table 1).

National Center for Health Statistics data report that there were an estimated 658 maternal mortality cases in 2018 (total YPLL 32,824; total VSL $7.9 billion), 754 in 2019 (total YPLL 37,998; total VSL $9.1 billion) and 861 in 2020 (total YPLL 43,131; total VSL $10.4 billion). This represents 2273 cases of maternal mortality 2018–2020 with a total YPLL of 113,953 and total VSL of $27.4 billion. Table 2 shows the total YPLL and VSL by year and overall.

During this 3-year period from 2018 to 2020, mothers under 25 accounted for 321 deaths (total YPLL 19,472; total VSL $4.7 billion), 25–39 year-old mothers for 1609 deaths (total YPLL 80,569; total VSL $19.4 billion) and 40-and-over mothers 342 deaths (total YPLL 13,912; total VSL $3.3 billion). Table 3 shows the total YPLL and VSL by age group classification bin and year.

Additionally, from 2018 to 2020, Non-Hispanic White mothers accounted for 986 deaths (total YPLL 49,234; total VSL $11.8 billion), Non-Hispanic Black mothers for 740 deaths (total YPLL 37,164; total VSL $8.9 billion) and Hispanic mothers 375 deaths (total YPLL base case 19,055; total VSL $4.6 billion). Table 4 shows the total YPLL and VSL by race/ethnicity and year.

This study assesses the relationship between age and race/ethnicity and the financial burden of maternal mortality in the USA from 2018 to 2020, as estimated by economic cost-benefit measures:  YPLL and VSL. Estimations of YPLL represent the sum of years of life lost annually by persons who suffered early deaths, and VSL represents the amount of additional spending that would be commonly accepted as justifiable to prevent excess maternal mortality. We found that there was a dramatic increase in both YPLL and VSL due to maternal mortality over the time studied, increasing more than 30% from an estimated YPLL of 32,824 and VSL of $7.9 billion to 43,131 and $10.4 billion, respectively. Overall, during 2018–2020, the economic burden of maternal mortality was estimated at a total YPLL of 113,953 and total VSL of $27.4 billion. We also observed significant racial and ethnic disparities in the economic metrics of maternal mortality.

Our results should be placed into context of another previous economic analysis of maternal mortality and morbidity. The analysis by O'Neil et al. developed cost estimates for all 2019 maternal morbidity conditions, including medical and nonmedical costs associated with outcomes for birthing people and their offspring from pregnancy through post-delivery year five. Their analysis showed that the costs of child outcomes was $24.0 billion, and this was much greater than the costs of maternal outcomes at $8.3 billion (of which $30.8 million is attributed to maternal mortality). Their work, however, was limited by data availability and likely underestimated the costs of maternal outcomes.

Our analysis showed increases in maternal mortality for nearly every category of age over the years studied, aside from 2019 where there was a transient decrease in maternal mortality for births to women aged 40 years and over compared with 2018. During 2018–2020, more than 15% of maternal deaths were among patients aged 40 years or older, though the total number of births by women of the same category was only 3.5% [18]. Comparing the overall increase in maternal mortality from 2018–2020, however, there were significant increases in births by women aged 25–39 years and 40 years and over; therefore, women who give birth at age 40 or over not only continue to have the highest risk of mortality, but experience mortality at higher rates [1–3]. Analysis of 2016–2017 vital statistics mortality data with cause-of-death literal text indicated that the leading causes of death for women aged 35 years and over were obstetric hemorrhage (leading cause; four-times higher than for women <35 years), postpartum cardiomyopathy (three-times higher), obstetric embolism, eclampsia/preeclampsia and other complications of obstetric surgery and procedures [19].

There was an increase in maternal mortality for each category of race/ethnicity in each of the years studied, with an overall increase of 21% for White mothers, 32% for non-Hispanic Black mothers and 35% for Hispanic mothers. However, from 2019 to 2020 non-Hispanic Black and Hispanic women experienced significant increases in maternal mortality, whereas for non-Hispanic White women the increase was not significant. While our findings show total maternal mortality and the economic costs associated to be highest among White deliveries, it must be viewed in the context of the proportion of White births and total births by race/ethnicity as a denominator (Table 1). In this time period, percentages of birth by race/ethnicity were about 57% White, 25% Hispanic and 16% non-Hispanic Black. For comparison, White mothers constitute about 47% of the economic cost measures, while non-Hispanic Black mothers are about 35% of the costs and Hispanic mothers about 18%. Therefore, YPLL and overall rate of maternal mortality remains disparate among women of different racial backgrounds when compared with their respective birth population percentages, with the greatest increases among non-Hispanic Black women. These racial disparities are multifactorial [20] and have previously been attributed to prevalence of preexisting and chronic disease [21], access to prenatal care [22] and pervasive implicit bias that negatively impacts medical care [23]. Non-Hispanic Black women are more likely to experience differences in obstetrical and anesthesia care received, severe maternal morbidity (SMM), failure to rescue (FTR) from SMM to maternal mortality and hospital readmission after birth [20,24–27].

Research has indicated that hospital-level factors and variations in hospital quality can influence obstetrical outcomes, including maternal mortality, FTR and SMM [26,28–30]. National obstetrical care benchmarking systems allow for better promotion of best practices and identification of system-based issues driving disparities [30]. Adoption of enhanced recovery after surgery protocols promotes equality in care provided to all patients and thereby reduces racial and ethnic disparities [20]. Implementation of health information technology also promotes healthcare equity to address disparities via better documenting of patient clinical history and risk factors, minimizing provider bias, improving coordination of clinical care and enhancing adherence to clinical guidelines [31,32]. Using 2017 data from a Turkish tertiary hospital setting, Oğlak et al. [33] indicated that adopting maternal near-miss concepts into a health system to evaluate the quality of obstetrical healthcare facilities can serve as a method to prevent maternal mortality. Maternal near miss or FTR have likewise been shown to be associated with hospital-level factors and to be a contributor to healthcare disparities. Friedman et al. [34] using National Inpatient Sample data from 1998 to 2010 (n = 50,433,539 delivery hospitalizations) showed that both low and high annualized hospital delivery volume were associated with both SMM and FTR from SMM to mortality. Additionally, FTR was higher among non-Hispanic Black and Hispanic women than non-Hispanic White women. Further analysis by Guglielminotti et al. [25] focused on racial and ethnic disparities in FTR, using 1999–2017 National Inpatient Sample data (n = 73,934,559 delivery hospitalizations) which indicated a higher FTR rate ratio for racial and ethnic group mothers compared with White mothers (ratio of FTR in the racial and minority group to FTR rate in White women). For instance, the adjusted FTR rate ratio was 1.79 (95% CI 1.77–1.81) for Black women, 1.39 (95% CI 1.37–1.41) for women of other race and ethnicity, 1.43 (95% CI 1.42–1.45) for women with missing race and ethnicity data and 1.08 (95% CI 1.06–1.09) for Hispanic women.

Our study has several limitations. The use of our economic metrics, including the measure of an entity's willingness to pay for risk reduction (VSL) may not be appropriate for evaluating maternal mortality, considering the inherent age- and race/ethnicity-based disparities [11,15]. VSL was determined from a set actuarial value that did not account for patient-level social determinant of health heterogeneity [12–14]. Additionally, there exist multiple methods to calculate VSL, which may differ based on national wealth and average personal incomes [14]. Official national data on maternal mortality was not available prior to 2018 [1–3], and our study cannot provide a longitudinal assessment of maternal mortality costs and disparities. Our study period also included the height of the COVID-19 pandemic in the USA, which strained healthcare resources and presented particular health risks to pregnant patients. The lack of stratified data limited the scope of our analysis and precluded us from assessing potential economic burden variance, analyzing contributing factors and in ultimately establishing specific population-level interventions to reduce maternal mortality risk. Our analysis focuses on maternal mortality rates, and it should be noted that some types of maternal deaths, such as homicide, suicide and drug-related overdoses, are not automatically included in maternal mortality rate analysis. For example, accidents, suicides and homicides are not converted to ICD-10 O-codes (coding system used for identifying maternal mortality), and as a result, some deaths due to external causes may not be coded as maternal, thereby leading to underreporting. Accurate coding to capture these causes of death requires death certificate certifiers of pregnancy status to be included. Research has shown that homicide, suicide and drug-related overdose represent the three leading causes of pregnancy-associated death [35]. Therefore, our analysis may undercount and underestimate the economic impact of maternal mortality.

This study has several strengths, including external validity for the US population. Our analysis of 2018–2020 maternal mortality data from the CDC Wide-ranging Online Data for Epidemiologic Research database provides national mortality estimates and classifications consistent with those used in previous National Center for Health Statistics, Division of Vital Statistics reports. Furthermore, we used US Social Security Administration actuarial life tables and economic estimates generalizable to the US population. We acknowledge that our economic estimates may not be generalizable to other countries, individual states or cities/communities, considering inherent societal value differences or disparities in race/ethnicity- and age-based maternal mortality.

Maternal mortality and disparities in populations affected by maternal mortality are caused by a multitude of patient-, hospital- and community-level factors [1–3]. Systemic and structural racism, differences in access to care and quality of care received, social determinants of health and co-occurrence of medical comorbidities contribute to these disparities in race/ethnicity- and age-based maternal outcomes [20]. Our study estimated the economic burden associated with these disparities in maternal mortality and showed a substantial increase in years of potential life lost and statistical value of lives lost due to maternal mortality from 2018 to 2020. These economic cost measure increases follow broader trends in increases in national maternal mortality rates over the past two decades. The data presented here suggest that a significant increase in public spending to reduce US maternal mortality risk may be justified according to commonly accepted economic cost–benefit analysis metrics. Addressing maternal mortality will require unparalleled national, state-wide and local initiatives to improve perinatal medical care. Prospective economic estimations from this analysis may additionally inform public health officials on the value of policies aimed at reducing maternal mortality and closing the disparity gap.

R  White is the recipient of a FAER Mentored Research Training Grant [ID: MRTG-08-15-2021-White (Robert)]. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.