Overview of maternal mortality

INTRODUCTION — Globally, maternal mortality has been falling, with substantial variation among countries and within countries [1,2]. In contrast to the global trend, maternal mortality in the United States has been increasing in recent decades [3].

This topic will discuss issues related to maternal mortality, including approaches to risk reduction in resource-rich countries (sometimes categorized as "high income" or "developed" countries).

Resource-limited countries (sometimes categorized as "low-income" or "low- and middle-income" countries or "developing" countries) have additional issues that impact risk reduction, which are reviewed separately. (See "Approaches to reduction of maternal mortality in resource-limited settings".)

Issues related to severe maternal morbidity are also reviewed separately. (See "Severe maternal morbidity".)

TERMINOLOGY — The World Health Organization's (WHO) International Classification of Diseases, 10^th revision (ICD-10) and Centers for Disease Control and Prevention's (CDC) maternity mortality review committee (MMRC) definitions for maternal death and its subclassifications are as follows [4,5]:

●Maternal death – The death of a patient while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes [6].

●Late maternal death – The death of a patient from direct or indirect obstetric causes more than 42 days, but less than one year, after the end of pregnancy.

●Pregnancy-associated death – A death during or within one year of pregnancy, regardless of the cause. This is comprised of the following:

•Pregnancy-related death – A death during or within one year 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.

•Pregnancy-associated, but not related death – A death during or within one year of pregnancy from a cause that is not related to pregnancy.

●Direct obstetric death – A direct obstetric death results from obstetric complications of pregnancy, labor, delivery, or the postpartum period, and from interventions, omissions, incorrect treatment, or a chain of events related to the obstetric complication. Direct obstetric deaths are more common than indirect obstetric deaths (86 and 12 percent, respectively) [7,8].

●Indirect obstetric death – An indirect obstetric death results from preexisting disease (eg, diabetes, cardiac disease, malaria, tuberculosis, HIV) or a new disease that develops during pregnancy and is unrelated to pregnancy-related conditions but is aggravated by the physiologic effects of pregnancy (eg, influenza). The term indirect obstetric death is used interchangeably with the terms nonmaternal, nonobstetric, and indirect maternal death.

●Maternal mortality ratio (MMR) – The MMR refers to the number of maternal deaths during a given time period per 100,000 live births. This is the most commonly used measure of maternal mortality and serves as an indicator of the risk for death once a patient has become pregnant. The denominator is live births rather than all pregnancies because of the difficulty in ascertaining the number of miscarriages and abortions in the population.

●Maternal mortality rate – The maternal mortality rate is defined as the number of maternal deaths in a given period per 100,000 females of reproductive age (15 to 49 years of age) during the same time period. Since the frequency of pregnancy in females of childbearing age is a factor in calculating this rate, it is affected by differences in the frequency of pregnancy or birth in the population even though the risk of maternal death per pregnancy/birth remains unchanged.

●Pregnancy-related mortality ratio (PRMR) – The number of pregnancy-related deaths for every 100,000 live births where a pregnancy-related death is defined as the death of a patient while pregnant or within 1 year of the end of a pregnancy (regardless of the outcome, duration or site of the pregnancy) from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes. This is the metric reported by the Centers for Disease Control and Prevention (CDC) in the United States; it is different from the MMR reported by international organizations.

●Lifetime risk of maternal death – The lifetime risk of maternal death takes into account the cumulative probability of dying as a result of pregnancy across a female's reproductive years. It is calculated by multiplying the maternal mortality rate by the length of the reproductive period (approximately 35 years; (1 – [1-maternal mortality rate]) [9].

●Preventability – A death is considered preventable if there was at least some chance of the death being prevented by one or more reasonable changes to patient, family, provider, facility, system, and/or community factors. This definition is used by MMRCs to determine if a death they review is preventable.

The United States employs two national approaches to maternal mortality surveillance and reporting: The CDC's National Vital Statistics System (NVSS) and the Pregnancy Mortality Surveillance System (PMSS). The NVSS provides MMRs for national and international use based on information provided in death certificates and follows the WHO guidelines using death certificate information; late maternal deaths (occurring 43 days to one year postdelivery) are not included in these guidelines. The PMSS performs epidemiologic surveillance of pregnancy-related deaths; in the United States, all states voluntarily submit deidentified death certificate data for females ages 12 to 55 who died within one year of pregnancy, linked birth or fetal death certificates, and additional sources of data, when available [10]. Medical epidemiologists review the data to determine if a pregnancy-related death has occurred. For each pregnancy-related death, the PMSS includes pregnancy outcome, associated conditions, demographic, and obstetric variables.

METHODOLOGY — The absence of high-quality, population-based data is one of the challenges of measuring maternal mortality [11-13]. Maternal deaths are relatively rare events, even in countries where maternal mortality rates are high. Thus, large sample sizes are needed to estimate both maternal mortality risks and contributors to maternal deaths.

Globally, several measures and surveillance methods are used to ascertain the magnitude of maternal mortality in a given region. These include medical certification in vital registration, household surveys (including sisterhood method), census records, and reproductive age mortality studies (RAMOS) [6]. Although many of these methods are imprecise, they provide some baseline information on the causes and magnitude of maternal deaths.

Resource-limited countries often apply their limited resources toward implementation of processes to decrease maternal mortality rather than toward improvement of surveillance programs. Resource-rich countries have more options. In the United States, the Centers for Disease Control and Prevention's Pregnancy Mortality Surveillance System combines data from state health departments, maternal mortality review committees, media, and individual providers, in addition to death certificates, to provide the best available assessment of maternal mortality in the United States.

●Vital registration – Few resource-limited countries have the ability to employ vital registration systems to assess levels and trends in maternal mortality. In resource-rich countries, information about maternal mortality is determined primarily from vital registrations (eg, death certificates) of deaths by causes. Nevertheless, maternal mortalities are still underreported and frequently misclassified. In one study from the United States, 38 percent of maternal deaths were unreported on death certificates, and at least 50 percent of maternal deaths were unreported among patients who were undelivered at the time of death, experienced a fetal death or therapeutic abortion, died more than a week after delivery, or died as a result of a cardiovascular disorder [11]. Similarly, in a retrospective study from Hong Kong between 2000 and 2019, vital statistics failed to identify the majority (90.5 percent) of maternal deaths that occurred in eight public maternity hospitals [14]. Confidential enquiries, such as the longstanding surveillance system in the United Kingdom, are often used to determine the extent of misclassification of maternal death [15-19].

●Direct household surveys – In areas where vital registration data are not available or reliable, household surveys provide an alternative method of maternal mortality assessment. However, surveys for direct estimation of maternal death are expensive and require large sample sizes to provide statistically reliable estimates. The sisterhood method is the most common method for household survey.

The sisterhood method asks four simple questions about survival of the respondents' adult sisters:

•How many sisters have you ever had, born to the same mother, who ever reached the age 15 (or who were ever married), including those who are now dead?

•How many of these sisters reaching age 15 are alive now?

•How many of these sisters are dead?

•How many of these dead sisters died during pregnancy or during childbirth, or during the six weeks after the end of the pregnancy?

In high-fertility populations, this approach is useful because sample size requirements (and thus costs) can be reduced. At high levels of maternal mortality (over 500 maternal deaths per 100,000 live births), a sample size of ≤4000 households is acceptable [20]. The method is not appropriate for use in settings where the total fertility rate is less than four children per family; in areas of significant migration, civil strife, or war; or where other social disruptions exist. Because deaths occurring over a large time interval are being documented, the overall estimate of maternal mortality is determined for a period of 10 to 12 years before the survey [20]. Although this is a limitation, the method is still useful since maternal mortality generally changes slowly and it provides some data for settings where there are no alternative means of generating estimates.

Demographic health surveys may also employ the indirect sisterhood approach, which utilizes more in-depth questions. The indirect sisterhood method relies on fewer assumptions than the original sisterhood method but requires a larger sample size; information gathering and analyses are also more complex. This method also does not provide a current estimate of maternal mortality, but the greater specificity of information allows for calculation of ratios for seven years prior to the survey.

●RAMOS – The RAMOS method involves investigation of the cause of death for all females of reproductive age. It has been used to calculate the degree of misclassification of maternal deaths in countries with and without well-developed vital registration systems. Multiple methods and sources are applied to obtain a comprehensive sample. In countries without vital registration systems, interviews with household members and medical providers enable further classification of deaths as possibly maternal. This method is one of the most complete determinations of maternal deaths; however, it can be complicated and time-consuming to perform.

●Verbal autopsy – Where medical classification of cause of death is limited or unavailable, some studies establish cause of death using verbal autopsy methodology. However, the reliability and validity of this method for determining the cause of death have not been established. The method may fail to address certain maternal deaths, such as first trimester deaths (ectopic, abortion-associated) or medical causes resulting from complications of terminations (eg, sepsis) and indirect causes (eg, HIV infection).

●Census – Census level data may include questions on household deaths during an established reference period followed by more detailed questions that identify maternal deaths based on timing of death in relation to pregnancy, similar to verbal autopsy methodology. The advantage of this method is that it can generate national data relative to household characteristics.

●Death certificates – In the United States, a 2003 revision of the death certificate added a pregnancy checkbox to connect death certificates and ICD-10 coding for maternal death. The five standard questions asked in the death certificate (if female) include: (1) not pregnant within the past year; (2) pregnant at the time of death; (3) not pregnant, but pregnant within 42 days of death; (4) not pregnant, but pregnant 43 days to one year before death; (5) unknown if pregnant within last year. To be classified as a maternal mortality, the death must be related to the pregnancy and not be a result of an accidental or incidental cause. The addition of these questions resulted in increased ascertainment of maternal deaths, particularly late maternal deaths [21,22]. For example, when reported with and without use of the checkbox, maternity mortality rates in 2014 and 2015 more than doubled (20.9 versus 8.7 deaths per 100,000 live births) [23]. Thus, caution should be exercised when comparing maternal mortality rates calculated before and after implementation of the pregnancy checkbox, as maternal mortality was greatly underreported prior to its implementation.

Although the revisions were recommended in 2003, they were not universally adopted by all states until 2017, and one state does not use these standard questions [24]. The nationwide adoption of the checkbox enabled the National Center for Health Statistics to report a nationwide maternal mortality ratio in 2018 [23]. To further mitigate misclassification of maternal deaths using death certificate data alone, further criteria (eg, including decedents ages 10 to 44 years [rather than 10 to 54 years], assignment of maternal codes to underlying cause of death if the checkbox is the only indication of pregnancy in the death certificate) were applied [25,26].

Even with these additions, state surveillance programs suggest that maternal death is underreported and often misclassified. For example:

•A study in Maryland found that checkboxes on death certificates were effective in identifying deaths resulting from maternal causes but were far less effective in identifying deaths resulting from nonpregnancy-related causes, such as homicide, accidental death, and substance use, which represented three of the four leading causes of pregnancy-associated death in Maryland [27].

•An observational study of United States death certificate maternal mortality data reported the estimated maternal mortality rate rose by nearly 27 percent between 2000 and 2014 in 48 states and the District of Columbia (maternal mortality rate per 100,000 live births: 18.8 in 2000 to 23.8 in 2014) [28]. The authors concluded that their analysis of measurement change suggested that maternal mortality rates were higher than reported during the early 2000s.

●Other methods – In low-income countries, other benchmark or process indicators, such as the number of deliveries with skilled attendants and minimum (below 15 percent) cesarean birth rates, have been proposed as proxy measures for health system progress in maternal mortality reduction efforts. (See "Approaches to reduction of maternal mortality in resource-limited settings".)

INCIDENCE AND TRENDS — Maternal mortality ratios (MMR) worldwide for 2000 and 2020 have been compiled by the World Health Organization, United Nations International Children's Emergency Fund (UNICEF), United Nations Population Fund (UNFPA), World Bank Group, and the United Nations Population Division and are available online.

Globally, the MMR was 223 maternal deaths per 100,000 live births in 2020 [29]. MMR fell 34 percent between 2000 and 2020 but with significant variation by the country's income level: 12 deaths per 100,000 live births in high-income countries, 44 deaths per 100,000 live births in upper middle-income countries, 255 deaths per 100,000 live births in lower middle-income countries, and 430 deaths per 100,000 live births in low-income countries. Country-specific MMR estimates ranged from 2 (Norway) to 1223 (South Sudan) maternal deaths per 100,000 live births. In 2016, there were 24 countries with an MMR of more than 400 maternal deaths per 100,000 live births [8].

The need for continued efforts to eliminate preventable maternal deaths is highlighted by the gap between the Sustainable Development Goal of fewer than 70 maternal deaths per 100,000 live births globally by 2030 and the recent MMR of low-performing regions, which is nearly 20 times the goal MMR. However, the gap can be closed, even in resource-limited regions. As an example, from 1996 to 2015, the county-level MMRs across China, where 25 percent of all maternal deaths are due to hemorrhage, declined by an annualized rate of 8.5 percent per year (from 108.7 to 21.8 per 100,000 live births), with significant declines reported by both developed and less-developed counties [30]. Simulation training for management of hemorrhage has played a significant role in the reduction in maternal mortality due to hemorrhage in China [31].

In contrast to the global trend of decreasing MMR, the Centers for Disease Control and Prevention and National Center for Health Statistics reports that maternal mortality in the United States has been increasing in recent decades: from 10 deaths per 100,000 live births in 1990 to 32.9 deaths per 100,000 live births in 2021 [3,32,33]. The increase of maternal mortality in the United States from 2018 and 2021 was most likely related to COVID-19. The greatest increase in maternal mortality was in non-Hispanic Black females which rose from 37.3 to 69.9 deaths per 100,000 live births. This is 2.6 times the rate for non-Hispanic White females (maternal mortality rate 26.6) and 2.5 times the rate for Hispanic females (maternal mortality rate 28). This is also the largest increase in Black female maternal mortality since 2015.

Equally concerning, it has been estimated that up to 80 percent of maternal deaths are preventable (see 'Risk reduction' below). Improved identification of pregnancy-related deaths and changes in coding and classification may account for most of this rise [34-36]. However, increases in maternal age, body mass index, and comorbidities have increased and may also account for part of the rise. Although the cause(s) for increasing MMR are not clear, factors that do not appear to contribute significantly to the increase include changes in the proportion of rural females, immigration rate, cesarean birth rate, and rates of medical conditions [37].

TIMING — Worldwide, approximately one-quarter of maternal deaths occur antepartum, another one-quarter occur intrapartum and immediately postpartum, approximately one-third are subacute and delayed postpartum deaths, and the remainder are late deaths [34]. In a United States study of data from 2017 to 2019, approximately 22 percent of pregnancy-related deaths occurred during pregnancy, 13 percent on the day of delivery, 12 percent on days 1 to 6 postpartum, 23 percent occurred within days 7 to 42 of delivery, and 30 percent occurred 43 days to one year after delivery [38]. Thus, more than half of maternal deaths occurred during the postpartum period. In another study performed by the Centers for Disease Control and Prevention (CDC) including 1480 maternal inpatient deaths from 2017 to 2019, rates of death were highest during delivery (6.1 per 100,000 hospital deliveries) compared with postpartum (4.5) and antepartum (2.9) [39]. However, while antenatal and postpartum hospitalizations accounted for <10 percent of all perinatal hospitalizations, they contributed to over half (56 percent) of the inpatient deaths during these periods.

Patient safety initiatives and standardizing obstetric care for high-risk patients (eg, patient safety bundles for maternal severe hypertension, thromboembolism, obstetric hemorrhage) have likely contributed to the reduction in antepartum and intrapartum maternal deaths [40]. Additional initiatives directed toward postpartum hospitalizations are needed [39].

CAUSES

Overview — Circumstances leading to maternal death are both complex and multifactorial, often involving at least four contributing factors [41] and patient/family, provider, and/or facility delay. The "Three Delays" model has been widely applied in the global context to understand and investigate complex social, cultural, and medical events contributing to maternal deaths [41-43]. These delays involve:

●Delay in the decision to seek care (eg, lack of knowledge of warning signs, lack of knowledge of symptoms requiring health care assessment, unrecognized life-threatening illness, women needing to seek permission from family members before obtaining care).

●Delay in arrival to an appropriate medical care facility (eg, poor or no transportation, long distance from care facility).

●Delay in receiving adequate care once a patient arrives to the medical facility (eg, lack of assessment resulting in misdiagnosis, delayed or ineffective treatment, unrecognized or undertreated life-threatening condition, inadequate facilities for severity of disease, lack of patient care coordination and poor communication between providers). For example, in the United States, women residing in rural compared with nonrural areas experience higher maternal mortality due to lack of access to obstetric care or subspecialists (eg, maternal-fetal medicine specialists) [44]. In 2020, 35 percent of all counties in the United States were classified as a maternity care desert (ie, a county without a hospital or birth center offering obstetric care and without any obstetric providers), which included over 2.2 million females ages 15 to 44 years and over 146,000 births [45]. An additional 11 percent of counties were classified as having low access to maternity care services (ie, one hospital or less offering obstetric services, fewer than 60 obstetric providers per 10,000 births, or ≥10 percent of females ages 18 to 64 years without health insurance) which impacted 2.8 million females ages 15 to 55 years and 187,900 births.

This framework enables assessment of patient, provider, and social/cultural practices that contribute to maternal death, which are key components of maternal mortality surveillance audits [42,43].

Specific disorders

Worldwide — The proportion of maternal mortality attributable to various causes varies worldwide. In a systematic analysis performed by the World Health Organization in 2014, the leading causes of maternal death were [7]:

●Obstetric hemorrhage (27 percent)

●Hypertensive disorders (14 percent)

●Pregnancy-related sepsis (11 percent)

●Abortion (8 percent)

●Embolism (3 percent)

●Other direct causes (10 percent; complications of delivery, obstructed labor, and all other direct causes)

●Indirect causes (28 percent; preexisting medical disorders, HIV-related maternal deaths, and all other indirect causes)

In subgroup analysis of the "other direct causes" of death, complications of delivery and obstructed labor each accounted for 2.8 percent of all maternal deaths. Obstructed labor as a cause of death was much more common in resource-limited than resource-rich regions (2.9 versus 0.6 percent of all maternal deaths). Among "indirect causes of death," preexisting medical conditions accounted for 14.8 percent and HIV for 5.5 percent of all maternal deaths. Although an analysis of the preexisting medical disorders was not performed, an earlier systematic review noted that cardiovascular disease was the most common medical condition among the indirect causes of maternal death [46].

Obesity, which is increasing in prevalence globally, contributes to an increased risk for venous thromboembolism-related maternal death and deaths associated with cesarean birth, preeclampsia, and cardiovascular disease [18].

Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus the causes coronavirus disease 2019 (COVID-19) has become a new cause of maternal mortality worldwide. In a cohort study including 43 institutions in 18 countries evaluating the risks associated with COVID-19 on maternal and neonatal outcomes, pregnant patients with (706 patients) versus without (1424 patients) a diagnosis of COVID-19 were at higher risk for preeclampsia/eclampsia (relative risk [RR] 1.8, 95% CI 1.3-2.4), infection requiring antibiotic therapy (RR 3.4, 95% CI 1.6-7), intensive care unit admission (RR 5.1, 95% CI 3.1-8.1), and maternal mortality (RR 22.3, 95% CI 2.9-1.7) [47]. In addition, asymptomatic women with a diagnosis of COVID 19 remained at higher risk for maternal morbidity (RR 1.2, 95% CI 1-1.5) and preeclampsia (RR 1.6, 95% CI 1-2.6).

Further discussion regarding COVID-19 and pregnancy, and vaccines to prevent SARS-CoV-2 infection are discussed in detail separately. (See "COVID-19: Overview of pregnancy issues".)

Maternal mortality as it relates to specific cancers (eg, breast, ovary, cervix) is also discussed separately. (See "Gestational breast cancer: Treatment", section on 'Prognosis' and "Adnexal mass: Evaluation and management in pregnancy", section on 'Prognosis' and "Cervical cancer in pregnancy", section on 'Outcome'.)

United States — The Centers for Disease Control and Prevention (CDC) issued a report using data from over 1000 maternal deaths from 36 states from 2017 to 2019 and reported the causes of pregnancy-related death including pregnancy and up to one year postpartum; more than 80 percent of such deaths were determined to be preventable [38].

The most common causes of pregnancy-related death were [38]:

●Mental health conditions (eg, deaths of suicide, overdose/poisoning related to substance use disorder; 22.7)

●Hemorrhage (13.7 percent)

●Cardiovascular conditions (12.8 percent)

●Infection (9.2 percent)

●Embolism (8.7percent)

●Cardiomyopathy (8.5 percent)

●Hypertensive disorders of pregnancy (6.5 percent)

●Amniotic fluid embolism (3.8 percent)

●Injury (eg, homicide, overdose/poisoning deaths not related to substance use disorder; 3.6 percent)

●Cerebrovascular accidents (2.5 percent)

These data are the first to be released under the Enhancing Reviews and Surveillance to Eliminate (ERASE) Maternal Mortality program and reflect an updated, more robust set of data, with information from more Maternal Mortality Review Committees (MMRCs) than ever before.

The causes of maternal death have shifted in frequency since recording began in 1987 [32,35,48-50]. When compared with maternal mortality data from prior to 2017, data from 2017 indicate there was an increase in deaths from mental health conditions (eg, suicide, opioid use disorder) and hemorrhage, but a reduction in deaths from cardiovascular disease, hypertensive disorders of pregnancy, pulmonary embolism, and cardiomyopathy [10,32,48,49,51]. Data from the National Center for Health Statistics from 2019 and 2020 also indicate an increase in deaths from drug overdoses, motor vehicle accidents, and homicide, but not suicide [52].

Furthermore, racial differences in the causes of maternal death exist. In the CDC report discussed above, Black patients had a higher proportion of deaths from cardiac and coronary conditions, while White patients had more deaths from mental health conditions [38]. Differences in causes of maternal mortality are also seen in patients who identify as American Indian or Alaska Native. In a separate CDC report, American Indian and Alaska Native patients had a higher proportion of deaths from mental health conditions and hemorrhage; however, the data are limited because of the small number of patients included [53].

As discussed above, trauma (eg, suicide, homicide, motor vehicle accidents, homicide) is a major contributor to maternal mortality but is excluded as a cause from the calculation of most national maternal mortality ratios [6,54-58]. In addition to the CDC study reported above [38], representative studies include:

●In an analysis of the 2018 to 2019 National Center for Health Statistics database, homicide-related maternal deaths were a leading cause of pregnancy-associated deaths with 3.6 homicides per 100,000 live births during pregnancy or within one year postpartum and 2.2 homicides per 100,000 live births up to 42 days postpartum [59]. The prevalence of pregnancy-associated homicide was highest among non-Hispanic Black females and females <25 years of age. Overall, pregnancy was associated with a nearly doubled risk for homicide among non-Hispanic White and non-Hispanic Black females ages 10 to 24.

●In a cross-sectional analysis including pregnancy-related deaths from 2008 to 2017 and reviewed by 14 maternal mortality review committees (MMRCs), there were 421 pregnancy-related deaths of which 46 (11 percent) were due to maternal mental health conditions including suicide (63 percent), unintentional poisonings/overdoses (24 percent), and other means or fatal injuries (13 percent) [60]. These deaths were most likely to occur in non-Hispanic White females (86 percent) and between 43 and 365 days postpartum (63 percent). All pregnancy-related maternal mental health deaths were deemed preventable.

●In an analysis from the National Violent Death Reporting System examining deaths during pregnancy and up to one year postpartum among females ages 15 to 54 from 16 states reporting complete data from 2003 to 2007, there were 94 pregnancy-associated suicides and 139 pregnancy-associated homicides, yielding maternal death rates of 2.0 and 2.9 deaths per 100,000 live births [61].

●An MMRC review from Colorado from 2004 to 2012 reported self-harm (eg, accidental overdose, suicide) was the leading cause of pregnancy-associated death [62]. Of the 211 maternal deaths, 63 (30 percent) were classified as self-harm of which the majority (90 percent) occurred postpartum and were associated with substance use and/or psychiatric disorders (depression being the most common). The pregnancy-associated mortality ratio was 34 per 100,000 live births. Pregnancy-associated maternal death ratios for overdose and suicide were 5 and 4.6 per 100,000 live births, respectively.

●In a retrospective study including over 11,700 pregnancy-associated deaths from 2010 to 2019 in the United States, drug-related deaths, suicide, and homicide accounted for 22 percent of the pregnancy-associated deaths [63]. All three causes of death increased in frequency during the study period.

Specific disorders that may lead to maternal death are discussed in detail separately:

●(See "Pregnancy in women with congenital heart disease: General principles".)

●(See "Acquired heart disease and pregnancy".)

●(See "Peripartum cardiomyopathy: Etiology, clinical manifestations, and diagnosis" and "Peripartum cardiomyopathy: Treatment and prognosis".)

●(See "Sudden cardiac arrest and death in pregnancy".)

●(See "Critical illness during pregnancy and the peripartum period".)

●(See "Overview of postpartum hemorrhage".)

●(See "Deep vein thrombosis in pregnancy: Epidemiology, pathogenesis, and diagnosis".)

●(See "Venous thromboembolism in pregnancy: Prevention".)

●(See "Treatment of hypertension in pregnant and postpartum patients".)

●(See "Preeclampsia: Antepartum management and timing of delivery".)

●(See "Eclampsia".)

●(See "Amniotic fluid embolism".)

●(See "Acute respiratory failure during pregnancy and the peripartum period".)

●(See "Initial evaluation and management of major trauma in pregnancy".)

●(See "Anesthesia for cesarean delivery".)

●(See "COVID-19: Overview of pregnancy issues".)

MAJOR DEMOGRAPHIC RISK FACTORS

Race and ethnicity — There is significant disparity in maternal mortality in the United States for Black females compared with White females and females of other races, and this disparity appears to be worsening [32,37,48,51,64,65]. The Centers for Disease Control and Prevention's (CDC) Pregnancy Mortality Surveillance System reported that, from 2011 through 2015, the pregnancy-related mortality ratio (PRMR) per 100,000 live births were approximately 3.3 and 2.5 times higher for Black and American Indian/Alaska Native females, respectively, compared with White females (42.8 for Black females, 32.5 for American Indian/Alaska Native females, 14.2 for Asian/Pacific Islander females, 13 for White females, and 11.4 for Hispanic females) [32].

Subsequent reports confirm this finding [66,67]. In an expanded analysis from 2007 to 2016, the PRMR for Black and American Indian/Alaska Native women ≥30 years of age were four to five times greater than those for White women [66]. Similarly, in an analysis from the National Center for Health Statistics from 2019 to 2020, all-cause mortality rate ratios were higher for Black females (1.9, 95% CI 1.8-2.1) and American Indian/Alaska Native females (3.49, 95% CI 2.85-4.22) compared with White females [52]. There has also been a rise in Black maternal mortality from sickle cell disease. This is discussed in detail separately. (See "Sickle cell disease: Obstetric considerations", section on 'Maternal risks'.)

Mortality differences persist at all education levels, which suggests that factors beyond education are involved. In the expanded analysis discussed above, among women with a college education or higher, the PRMR for Black women was 5.2 time that of White women [66]. This racial disparity is one of the largest perinatal health disparities in the United States and does not appear to be related to changes in data collection [32,68,69] (see 'Methodology' above). Similarly, in an analysis of 2016 to 2017 vital statistics mortality data, the all-cause maternal mortality ratio was over three times higher for non-Hispanic Black females than for non-Hispanic White females (mortality rate ratio [MRR] 3.55, 95% CI 2.94-4.28); the MRR for the four most common causes of death were, in descending order [67]:

●Eclampsia and preeclampsia (MRR 5.06, 95% CI 3.16-8.21)

●Postpartum cardiomyopathy (4.86, 95% CI 2.93-8.12)

●Obstetric embolism (2.58, 95% CI 1.55-4.23)

●Obstetric hemorrhage (2.27, 95% CI 1.22-4.11)

Although increasing educational and socioeconomic levels and prenatal care are protective against maternal death, at all educational and socioeconomic levels, PRMR for Black women are three to five times higher than ratios for White women, and the reduction in maternal death associated with prenatal care is greater for White women than for Black women who receive prenatal care [66,68,70]. Although PRMR increase with maternal age for all women, Black women aged ≥25 years have a fourfold greater risk of dying from pregnancy than White women, and the excess risk of death is highest among Black women aged 40 years or older [48,66].

Social and structural determinants rather than biological factors contribute to these racial inequities and the disparity gap between Black and White individuals [71]. A disparity bundle has been added to the Alliance for Innovation on Maternal Health (AIM) safety bundle to address this disparity gap [72,73].

Maternal mortality and severe morbidity are also higher among non-White women in the United Kingdom and Canada [18,74-76]. Similar findings have been observed globally in countries where ethnicity and immigrant status are included in national surveillance systems and studies of adverse maternal health outcomes [13,77-79]. (See "Racial and ethnic inequities in obstetric and gynecologic care and role of implicit biases".)

Younger and older maternal age — Several studies have linked younger and older maternal age with an increased risk of maternal mortality [32,80-85]; patients must be counseled about these risks during preconception counseling visits.

●Younger maternal age – A large international multicenter study found that adolescents had a higher maternal mortality ratio (MMR; 260 maternal deaths per 100,000 live births) compared with females ages 20 to 24 years, who had the lowest MMR (190 maternal deaths per 100,000 live births) [81]. However, age-specific MMRs varied among countries and regions; in Southeast Asia, for example, adolescents had the lowest MMR of any age group.

●Older maternal age – In the large international multicenter study mentioned above, females >35 years had a higher MMR compared with females in any younger age group (>700 maternal deaths per 100,000 live births) [81]. In a second study, females ages ≥45 compared with females <35 years had a higher risk of death (odds ratio [OR] 9.90, 95% CI 5.6-15.98); cardiac events, including heart failure and myocardial infarction, were the primary contributing factors (OR 8.42, 95% CI 6.48-10.93 and OR 21.38, 95% CI 11.46-39.88, respectively) [84].

In a United Kingdom database study of maternal deaths for women age ≥35 years, each additional one-year increase in age was associated with a 12 percent increased odds of maternal death [86]. Adjusted risk factors associated with increased likelihood of death included, in descending order:

•Inadequate use of antenatal care (adjusted OR [aOR] 23.62, 95% CI 8.79-63.45)

•Medical comorbidities (aOR 5.92, 95% CI 3.56-9.86)

•Previous complication of pregnancy (aOR 2.06, 95% CI 1.23-3.45)

•Maternal smoking during pregnancy (aOR 2.06, 95% CI 1.13-3.75)

Furthermore, causes of maternal mortality in older females may differ from that of younger females. In the United States, the leading cases of maternal mortality for females ≥35 compared with <35 years between the years 2016 to 2017 included, in descending order [85]:

•Obstetric hemorrhage (increased risk: approximately fourfold)

•Postpartum cardiomyopathy (threefold)

•Obstetric embolism (twofold)

•Eclampsia/preeclampsia (twofold)

•Other complications of obstetric surgery (twofold)

RISK REDUCTION — Forty to 80 percent of maternal deaths are considered preventable [32,38,41,87-90]. One report suggested the following interventions to reduce maternal mortality, in order of significance [91]:

●Family planning with birth spacing and contraception (30 percent reduction)

●Safe abortion (13 percent reduction)

●Hemorrhage prevention and treatment (8 to 9 percent reduction)

●Cesarean section when indicated (7 percent reduction)

●Prevention of eclampsia and treatment of preeclampsia (7 percent reduction)

General approach — In the United States, deficient medical care, medical comorbidities, structural barriers and social determinants of health appear to be strong contributors to maternal mortality, especially for Black women who continue to have a higher case-fatality rate than White women (see 'Race and ethnicity' above). In the United States, better use of resources, rather than increase of resources, is a major focus for reducing maternal mortality. Efforts have focused on team and individual training; simulations and drills; development of protocols, guidelines, and checklists; use of information technology; and education [92-95]. The goal of many of these activities is to achieve early diagnosis and appropriate medical care of pregnancy complications [87]. (See "Reducing adverse obstetric outcomes through safety sciences" and "Patient safety in the operating room".)

For example, hospitals can implement multidisciplinary (eg, obstetric providers, anesthesia, nursing) huddles to assess and review each obstetric patient's risk factors, identify those at high risk for complications, and develop a shared mental model of how the needs of those patients can be met [96]. Huddles among the surgeon, anesthetist, nurses, and scrub technicians for all patients undergoing scheduled or nonurgent cesarean deliveries develop shared understanding of the patient and the procedure and clarify what additional resources might be needed for the surgery, especially in the event of an unexpected complication. As part of this process, safety concerns that are identified should be communicated to the patient. Shared decision making by the patient and the obstetric team may reduce any potential institutional biases that affect disparities in maternal morbidity and mortality.

Discharge planning and postpartum follow-up are essential for women with obstetric near-miss morbidity. For example, guidelines for follow-up of women with preeclampsia, especially those with severe features, should include blood pressure surveillance for up to 72 hours and at seven days, depending on severity [97]. In addition, the American College of Obstetricians and Gynecologists (ACOG) recommends that all postpartum patients have an initial encounter within three weeks after delivery followed by a comprehensive visit between four and 12 weeks postpartum and ongoing follow-up as needed [98]. (See "Overview of the postpartum period: Normal physiology and routine maternal care", section on 'Follow-up visits'.)

A prediction model for severe maternal morbidity and mortality has been developed [99], but further validation of the model is needed before it can be routinely used.

CMQCC — The California Maternal Quality Care Collaborative (CMQCC) was formed as a public-private partnership to lead maternal quality improvement activities. The CMQCC's review of 207 pregnancy-related deaths in California concluded there was a good-to-strong chance that 41 percent could have been prevented and that 60 to 70 percent of deaths related to hemorrhage and preeclampsia could be prevented [100]. Based on these and other data, they initiated several activities, including linking public health surveillance to actions; mobilizing a broad range of public and private partners; developing a rapid-cycle Maternal Data Center to support and sustain quality improvement initiatives; and implementing a series of data-driven, large-scale quality improvement projects [101]. These interventions have been associated with a 50 percent reduction in maternal deaths (three-year average of 7 maternal deaths per 100,000 live births) at a time when maternal deaths in other states were increasing.

Key clinical and systems issues are rapid and systematic responses to (1) hemorrhage, (2) severe hypertension, and (3) infection/fever/sepsis. The CMQCC has created toolkits to help providers reduce morbidity/mortality from obstetric hemorrhage, preeclampsia, venous thromboembolism, and cardiovascular disease, as well as to reduce cesarean birth rates and eliminate early elective deliveries. Other potential areas to reduce risk for maternal death include improvements in preconception counseling and pregnancy care (eg, availability of tertiary care, multidisciplinary care, specialized equipment) for women with medical comorbidities, including obesity [87,102].

AIM — The Alliance for Innovation on Maternal Health (AIM), a collaboration led by ACOG and involving 30 other organizations representing the spectrum of women's health care, has helped to implement consistent maternity care practices by creating bundles of best practices for improving safety in maternity care. The bundles help clinicians, the obstetric team, and facilities consistently manage the care of high-risk pregnant women, and include:

●Obstetric hemorrhage

●Severe hypertension in pregnancy

●Safe reduction of primary cesarean birth

●Cardiac conditions in obstetric care

●Obstetric care for women with substance use disorder

●Perinatal mental health conditions

●Postpartum discharge transition

●Sepsis in obstetrical care

Maternal mortality review committees — Maternal mortality review committees (MMRCs) use multiple sources of data from clinical and nonclinical sources (eg, vital records, social services information, police reports) to perform in-depth reviews to determine the following: (1) Was the death pregnancy-related? This determination is based on the response to the question, "Would she have died if she had not been pregnant?" After review of multiple data sources, if the answer to that question is no, the death is considered pregnancy related. The committees then address additional questions: (2) What was the underlying cause of death?, (3) Was the death preventable?, (4) What were the factors that contributed to the death?, (5) What are the recommendations and actions that address those contributing factors?, and (6) What is the anticipated impact of those actions if implemented? [41]. They also determine factors that contributed to death (eg, patient/family, provider, facility, systems of care, or community), and analyze the impact of social determinants of health on maternal outcomes [103]. MMRCs must be multidisciplinary and include health care providers beyond obstetrician-gynecologists, such as emergency medicine clinicians, pathologists, cardiologists, psychiatrists/mental health providers, advocacy groups, violence/injury prevention specialists, and state public health representatives engaged in state perinatal collaboratives.

Recommendations for maternal mortality reduction in resource-rich countries are based, in part, on confidential inquiries of maternal deaths and evaluations of maternal deaths and severe maternal morbidities by MMRCs [32,89,104,105]. Their findings can enable development of referral systems for appropriate transfer or consultation, in addition to providing education and feedback on management of scenarios or cases that result in high near-miss maternal morbidity and/or mortality. To facilitate discussions, the proceedings of committees reviewing severe maternal morbidities and maternal mortalities should have protection by state statute to shield them from liability or discovery. At least one review has suggested that the process of confidential inquiry and resultant change in clinical management have contributed to the substantial reduction in maternal mortality in the United Kingdom over the past 60 years (from 200 deaths in England and Wales in 1952 to 1954 to 3 deaths in the entire United Kingdom from 2013 to 2015) [106].

In 2018 the United States Preventing Maternal Deaths Act established federal legislation for states to create MMRCs to review all maternal deaths. The legislation authorizes the Centers for Disease Control and Prevention (CDC) to assist states to create or expand MMRCs, collect consistent data to help understand what causes maternal mortality, and recommend locally relevant strategies for state Departments of Public Health to prevent pregnancy deaths and reduce disparities. The legislation asked for a follow-up report to Congress on maternal mortality data to track successes and setbacks. Finally, Congress asked the Department of Health and Human Services to research disparities in maternal health outcomes.

After review of data from nine MMRCs in the United States, the authors of a 2018 report recommended the following interventions to reduce maternal deaths and estimated their potential impact, which varies by cause of death:

●Large impact likely:

•Adopt levels of maternal care/ensure appropriate level of care determination

•Improve policies regarding prevention initiatives, including screening procedures and substance use prevention or treatment programs

•Enforce policies and procedures

•Improve policies related to patient management, communication and coordination between providers, and language translation

•Improve access to care

●Small to medium impact likely:

•Improve patient management for mental health conditions

•Improve training

•Improve standards regarding assessment, diagnosis, and treatment decisions

•Improve procedures related to communication and coordination between providers

•Improve patient/provider communication

To build upon this momentum, the CDC has expanded its investment in efforts to eliminate preventable maternal mortality, with awards to support MMRCs in 39 states and one US territory [107].

Recommendations of national organizations — Recommendations of selected national organizations are described below.

●ACOG has developed many resources for its members regarding patient safety [108] and maternal mortality [109]. ACOG also supports use of Maternal Health Compacts in which a tertiary care hospital provides services to its referring lower-resource hospitals for high-risk patients [110]. The connection can be activated by lower-resource hospitals, especially in rural areas, to get immediate consultation in the event of an unexpected obstetric emergency whose care demands exceed their resources. The tertiary hospital may also run simulations of obstetric emergencies and assist with quality-improvement activities.

●Joint Commission – The United States Joint Commission suggestions to help hospitals reduce the risk of maternal death include [104]:

•Good communication between all members of the health care delivery team, consultants, the patient, and the patient's family.

•Education of clinicians about the potential additional risks in pregnant women with underlying medical conditions. These risks should be discussed during preconception care and counseling, and appropriate contraception should be offered. High-risk patients should be referred to obstetricians with expertise in and resources for caring for these patients.

•Development of written protocols and drills for promptly responding to changes in maternal vital signs with best practices. Common scenarios are management of severe hypertension or hypotension, treatment of pulmonary edema in preeclampsia, and early response to postpartum hemorrhage.

•Instituting measures (pneumatic compression devices, low molecular weight heparin) to reduce the frequency of pulmonary embolism in high-risk patients.

Hospitals should review and report severe maternal morbidities [111]. In addition, they should collect data on race/ethnicity and institute processes for individual providers and teams to recognize and address implicit bias and the impact of social determinants of health on maternal outcomes [72].

●UK Confidential Enquiries Commission – The Eighth Report of the Confidential Enquiries into Maternal Deaths in the United Kingdom had similar recommendations as those described above, and also included [88]:

•Access to professional interpreter services for ethnic minorities

•Emphasis on prevention and treatment of infection

•Consistent reporting of serious incidents and maternal deaths

•Performance of autopsies by pathologists with expertise in this area

Resource-limited countries — Specific recommendations for reducing maternal mortality in resource-limited countries are discussed in detail separately. (See "Approaches to reduction of maternal mortality in resource-limited settings".)

SUMMARY AND RECOMMENDATIONS

●Incidence – Maternal mortality ratios (MMR) worldwide are available online. In contrast to the global trend of decreasing MMR, maternal mortality in the United States has been increasing in recent decades. (See 'Incidence and trends' above.)

●Timing – Worldwide, approximately one-quarter of maternal deaths occur antepartum, another one-quarter occur intrapartum and immediately postpartum, and the remainder occur up to one year postpartum. (See 'Timing' above.)

●Causes – According to the World Health Organization, the leading causes of maternal death worldwide are (see 'Worldwide' above):

•Obstetric hemorrhage (27 percent)

•Hypertensive disorders (14 percent)

•Pregnancy-related sepsis (11 percent)

•Abortion (8 percent)

•Embolism (3 percent)

•Other direct causes (10 percent; complications of delivery, obstructed labor, and all other direct causes)

•Indirect causes (28 percent; preexisting medical disorders, HIV-related maternal deaths, and all other indirect causes)

●Risk factors

•Race and ethnicity – In the United States, Black, American Indian, and Alaska Native women have higher maternal mortality compared with White women. (See 'Race and ethnicity' above.)

•Older age – Pregnant patients of all races ages ≥45 years compared with <35 years are at greater risk for cardiac maternal mortality. (See 'Younger and older maternal age' above.)

●Risk reduction – Forty to 80 percent of maternal deaths are considered preventable. (See 'Risk reduction' above.)

•In the United States, deficient medical care, medical comorbidities, and social determinants of health appear to be strong contributors to maternal mortality. As such, better use of resources rather than lack of resources is a major focus for reducing maternal mortality. Efforts have focused on team and individual training; simulations and drills; development of protocols, guidelines, and checklists; use of information technology; and education. The goal of many of these activities is to achieve early diagnosis and appropriate medical care of pregnancy complications. Strategies that address implicit bias and social determinants of health focus on reducing racial and ethnic disparities in maternal mortality. (See 'Risk reduction' above.)