Hemorrhage , hypertensive disorders of pregnancy , and sepsis account for more than one-half of global maternal deaths and for slightly more than one-third of deaths in high-income countries (HICs) ( Fig. 39.1 ). Hemorrhage is the leading direct cause of maternal death worldwide, followed by hypertensive disorders, causing an estimated 27.1% and 14.0% of all deaths, respectively. Infection and sepsis may be substantially underestimated in regions where laboratory diagnostic tests are unavailable. In one Malawian hospital with full laboratory capabilities, infection played a primary role in almost three-fourths of all maternal deaths.

Estimated Distribution for the Main Causes of Maternal Death Worldwide, and in Selected Regions. The relative size of each circle is illustrative but not directly proportional to the relative numbers of deaths in the three illustrated regions. *First-trimester deaths include those from induced abortion, miscarriage, and ectopic pregnancy.

Data from Say L, Chou D, Gemmill, et al. Global causes of maternal death: a WHO systematic analysis. Lancet Glob Health . 2014;2:e323–e333.

Rarely, anemia can cause lethal congestive heart failure in pregnancy, but more commonly, it increases the risk for maternal death from other complications, particularly hemorrhage and infection. Risk factors associated with anemia include (1) iron and other micronutrient deficiencies; (2) pregnancy interval < 1 year; (3) adolescent pregnancy; (4) hemoglobinopathy; (5) urinary tract infection; (6) human immunodeficiency virus (HIV) infection; (7) parasitic infections, including malaria; and (8) recurrent antepartum hemorrhage.

Obstructed labor causes less than 10% of maternal deaths worldwide, but it is an important cause of mortality in communities where early adolescent pregnancy is common, childhood malnutrition leads to small maternal pelvises, and operative delivery is unavailable. Death from obstructed labor is largely the result of hemorrhage caused by uterine rupture or sepsis caused by ascending genital tract infection, and deaths may be coded under those categories. Prolonged pressure on the pelvic outlet can lead to tissue necrosis and obstetric fistula, a complication that is thought to affect at least 2 million women worldwide.

HIV/acquired immunodeficiency syndrome (AIDS) increases vulnerability to both nonobstetric infection (e.g., tuberculosis, malaria) and obstetric complications (e.g., pregnancy-related sepsis, septic abortion) (see Chapter 36 ). Global mortality attributed to HIV/AIDS peaked in 2005 at 2.5%; by 2023 the WHO attributed approximately 0.5% of maternal deaths to HIV/AIDS. In countries most severely affected (Botswana, Malawi, South Africa), the MMR increased between 1990 and 2000, but has subsequently declined with increasing availability of antiretroviral therapy.

Deaths attributed to unsafe abortion account for 5% to 13% of maternal deaths worldwide. Unsafe abortion is highest in countries with both restrictive laws and limited healthcare infrastructure. Worldwide, 45% of abortions were unsafe between 2010 and 2014, compared with 44% in 1995. Risk is highest in sub-Saharan Africa; between 2010 and 2014, 77% of the region’s abortions were considered unsafe. Fortunately, increasing availability of medication abortion and postabortion care has improved abortion safety over the past 20 years; the case-fatality rate has decreased from 315 to 185 maternal deaths per 100,000 abortions. A WHO evidence-based protocol offers guidance for pregnant women less than 12 weeks’ gestational age to manage their own medical abortions using mifepristone and misoprostol in combination or misoprostol alone.

Early marriage (before 18 years of age) has been identified as a major health risk for girls, increasing their exposure to domestic violence, coercion, pregnancy, and sexually transmitted diseases such as HIV/AIDS. Girls younger than 15 years of age are four times more likely to die in childbirth than women in their 20s, and pregnancy is among the leading causes of death worldwide for girls 15 to 19 years of age. Early childbearing also increases the likelihood of high parity birth (≥5) later in life. With a threefold increase in the MMR, high parity remains an important demographic risk for maternal death because it remains so common in countries with high maternal mortality rates. Advanced maternal age is less common globally, but the MMR increases 3-fold by 35 years, 7-fold by 40 years, and 15-fold after 45 years of age.

Cesarean delivery is the most common major surgical procedure in Africa, but anesthesia providers working in LMICs often must contend with profound limitations in staffing, equipment, and other resources. Cesarean delivery rates should be approximately 15% to 20% to optimize maternal and neonatal outcomes, but are less than 10% in more than half of African countries. Low workforce density compounds resource deficiencies. An inflection point suggests that in settings with less than four specialist anesthesiologists per 100,000 population, maternal mortality increases dramatically. Yet only five African countries have more than one physician anesthesia provider per 100,000 population.

In addition, patients who labor at home may face a variety of social and environmental obstacles to reach a facility with the capacity to provide comprehensive emergency obstetric care, and many arrive at these facilities in septic or hemorrhagic shock. While complication rates following cesarean delivery are 2 to 3 times higher in LMICs, maternal mortality rates are 50 times higher than in HIC, indicating high rates of failure to rescue.

Perioperative maternal mortality is estimated to be between 1.2% and 2%. Based on a 2016 metaanalysis of studies that assessed anesthesia-related maternal death in LMICs, an estimated one in seven deaths associated with cesarean delivery is attributed to anesthesia. Although general anesthesia is associated with a threefold increased risk for death compared with neuraxial anesthesia (5.9 per 1000 and 1.2 per 1000, respectively), reports of high spinal anesthesia and hemodynamic collapse highlight potential hazards with neuraxial anesthesia. Failed airway management (including bronchospasm and aspiration of gastric contents) accounts for three-quarters of anesthesia-related deaths reported from LMICs. Peripartum deaths have also been attributed to limited availability or affordability of blood products, cardiac arrest at induction of anesthesia, drug overdose, adverse medication reaction, and drug error. The increased use of tranexamic acid for postpartum hemorrhage has led to an apparent increase in the number of cases of inadvertent intrathecal administration of tranexamic acid with planned spinal anesthesia, resulting in the release of a WHO drug safety alert. Overall, the number of maternal perioperative deaths likely pales in comparison with maternal deaths that result from the unmet need for lifesaving obstetric procedures, including cesarean delivery.

Strategies to reduce global maternal mortality include (1) improvement in family planning services and a reduction in the performance of an unsafe abortion ; (2) community-based education focused on safe birth practices and indications for transfer to a higher level of care ; and (3) development of the infrastructure needed to provide timely emergency obstetric care, including the performance of indicated cesarean delivery (and safe administration of anesthesia) by trained healthcare providers who can also provide resuscitation of women in whom shock develops secondary to hemorrhage or infection. Cluster randomized trials and cost-effectiveness analyses to evaluate these strategies are beginning to be published. When deployed as part of an integrated, context-specific, and culturally sensitive program, interventions to reduce maternal and fetal mortality can be highly cost effective.

According to a 2010 systematic review of data from HICs, hypertensive disorders of pregnancy, embolic disorders , and hemorrhage together account for 43% of maternal deaths in HICs (see Fig. 39.1 ). Indirect deaths account for another 25%.

Indirect deaths have exceeded direct deaths in the United Kingdom since 1997, with cardiac disease being the most common category of death. Similar patterns are seen in the United States, where the combination of cardiovascular conditions and cardiomyopathy comprised 20.6% of all pregnancy-related deaths in 2023, and cerebrovascular accidents and noncardiovascular medical conditions represented another 15.1%. Cardiomyopathy, including both peripartum and other types of cardiomyopathy, is the leading diagnosis underlying maternal cardiac death in the United States. In the United Kingdom, sudden arrhythmia, ischemic cardiac disease, and cardiomyopathy are the leading causes of maternal cardiac death.

Since 2015, psychiatric diseases have emerged in the United Kingdom, the United States, and France as a leading cause of death, equaling cardiac disease and cardiomyopathy in some reports. In the United States, 36 state maternal mortality review committees following the standardized CDC review process identified the six most frequent underlying causes of pregnancy-related death for 2017 to 2019: mental health conditions (22.7%), hemorrhage (13.7%), cardiac and coronary conditions (12.8%), infection (9.2%), thrombotic embolism (8.7%), and cardiomyopathy (8.5%). A preventability determination was made in 996 deaths; 839 (84%) were determined to be preventable. Data from France for the period between 2013 and 2015 revealed that the two leading causes of maternal deaths (suicides and cardiovascular disease) were also those with the highest proportion of preventable mortality, at 91.3% and 65.7% respectively.

The COVID-19 pandemic had a significant impact on the global MMR (see Chapter 36 ). Increases in MMR were due to either an increase in indirect obstetric deaths (due to the SARS-CoV-2 infection) or a decrease in global healthcare as a result of disruption to services caused by the pandemic, leading to a failure to manage the complications of pregnancy. In the United States, the COVID-19 pandemic seemed to heighten the existing racial disparities in MMR. A rapid systematic review using data from Mexico, Peru, Uganda, South Africa, and Kenya also showed increases in MMR in these locations. Early studies suggested pregnant patients were at increased risk of both severe morbidity and mortality following infection with SARS-CoV-2. However, concerns about vaccine safety led to vaccine hesitancy, which persisted despite a lack of evidence for adverse outcomes linked to the vaccine. A growing body of evidence subsequently confirmed that immunization protects both the mother and the neonate from adverse outcomes.

Detailed descriptions of hypertensive disorders of pregnancy, infection, hemorrhage, embolic disorders, cardiovascular diseases, and psychiatric disorders in pregnancy are provided in Chapters 35, 36, 37, 38, 41, and 49 , respectively. Injury-related deaths are considered pregnancy-associated but not pregnancy-related (see Chapter 53 ). Common clinical and sociodemographic factors that increase the risk for maternal death from all these complications are discussed in the following sections.

Advanced maternal age increases maternal risk, with a linear trend evident for each 5-year increase in maternal age beyond 34 years. The MMR for women over 40 was 7.8 times higher than for women under 25 years of age in the United States between 2018 and 2020, and three times higher in the United Kingdom than for women aged 20 to 24 years old. Significant differences were seen between ethnic groups in both countries with respect to advanced maternal age. In the United States, the MMR among women 40 years of age and older was 223.5 per 100,000 births for Black women (compared with 20.8 per 100,000 births for Black women less than 25 years of age) and 81.4 per 100,000 births for White women (compared with 11.7 per 100,000 births for White women less than 25 years of age). Based on data from the PMSS for 1991 to 1999, the association between age and mortality persisted after data were controlled for parity, prenatal care, race, and education. Among older Black women (≥40 years), the excess risks were greatest for hypertensive disorders of pregnancy, infection, cerebrovascular accident, and other medical conditions. Among older White women, the greatest excess risks for death were caused by hemorrhage, cardiomyopathy, embolic disorders, and other medical conditions.

In the United States, Black or African-American race significantly correlates with the risk for death. Non-Hispanic Black women experience a MMR that is 2.9-fold higher than non-Hispanic White women. In a case series of anesthesia-related maternal deaths in Michigan published in 2007, six of eight deaths occurred among non-Hispanic Black women in Detroit, suggesting a profound concentration of maternal risk. The disparity in maternal mortality between Black women and White women persists after data are controlled for maternal age, income, and receipt of prenatal care and appears to be related to a higher case-fatality rate. In the United Kingdom, the relative risk (RR) for Black women was 3.68 (95% confidence interval [CI], 2.32 to 5.65) compared with White women.

Other racial and ethnic groups also experience increased risk. In the United States between 2011 and 2016, the MMR was 240% higher for American Indian or Alaskan Native women compared with non-Hispanic White women. In England, women from Asian backgrounds were at higher risk than White women (RR, 1.75; 95% CI, 1.13 to 2.62), as were women from mixed ethnic backgrounds (RR, 1.32; 95% CI, 0.35 to 3.47).

Immigrants , asylum seekers, and nonnative language speakers appear to be particularly vulnerable to both maternal death and substandard care, based on data from the United Kingdom and the Netherlands. In the United States, Hispanic and Asian/Pacific Islander immigrants face increased risk compared with women of these same racial/ethnic groups born in the United States. Significant regional variation has been identified in France, with increased risk noted for women delivering in Paris (adjusted odds ratio [aOR], 1.6; 95% CI, 1.2 to 2.0) and the overseas districts (aOR, 3.5; 95% CI, 2.4 to 5.0) compared with continental France. Differences in behavior, biology, environmental conditions, social circumstances, and the quality of clinical care may contribute to disparities in outcomes in sociodemographically vulnerable populations.

Rural residence is an important risk factor for maternal mortality in the United States, with impacts across all age cohorts and racial and ethnic groups. Workforce shortages and economic stress have accelerated obstetric unit closures in rural communities, increasing the distances rural women need to travel for obstetric care. Higher rates of substance use and mental health disorders further contribute to population risk.

Maternal obesity increases the risk for maternal death from a variety of causes, including pulmonary embolism, infection, preeclampsia, and anesthesia-related complications. Remarkably, there are limited epidemiologic data to establish this connection. Obesity is a common feature in case series of maternal deaths. In the 2018 to 2020 CEMD report from the United Kingdom, 51% of women who died were overweight or obese, and admission to an ICU was more common in women with a body mass index (BMI) > 35 kg/m ² .

Multifetal pregnancies increase maternal risk for a variety of complications, including preeclampsia, venous thromboembolism, heart failure, myocardial infarction, peripartum hemorrhage, and maternal death. Compared with twin pregnancies, triplet and higher-order multiple pregnancies further increase maternal risk for preeclampsia, hemorrhage, and emergency peripartum hysterectomy. In the United States between 1979 and 2000, the relative risk for death associated with a multifetal pregnancy compared with a singleton pregnancy was 3.6 (95% CI, 3.1 to 4.1) with threefold to fourfold increases in the cause-specific relative risk for death for embolism, hemorrhage, hypertensive disorders of pregnancy, infection, cardiomyopathy, and other medical conditions.

Cesarean delivery has also been associated with an increased risk for maternal death; however, the association does not always reflect a causal relationship. Death can be a consequence of the indication for the operation rather than the mode of delivery itself. Confidential enquiry reports in the United Kingdom and United States do not routinely report maternal mortality data based on cesarean delivery versus vaginal delivery, possibly due to difficulties adjusting for the confounding influence of the indication for operative delivery. In an attempt to estimate the relative risk for death caused by cesarean delivery, a population-based case-control study from France focused on 65 maternal deaths after singleton births among low-risk women in whom complications developed only after delivery. Cases were identified by the French National Confidential Enquiry on Maternal Deaths from 1996 through 2000. These cases were compared with 10,244 singleton births to low-risk women identified through the French National Perinatal Survey conducted in 1998. After data were controlled for maternal age, nationality, parity, and preterm delivery, the aOR for increased risk for death with cesarean delivery was 3.64 (95% CI, 2.15 to 6.19). The increased risk was most dramatic for intrapartum cesarean deliveries (aOR, 4.58; 95% CI, 2.30 to 9.09) but was also present when cesarean delivery preceded labor (aOR, 2.42; 95% CI, 1.14 to 5.13). Among women who underwent cesarean delivery, there was an increased risk for cause-specific maternal mortality from venous thromboembolism, puerperal infection, and complications of anesthesia, but not for risk for death from postpartum hemorrhage or amniotic fluid embolism. A more recent study in Brazil used similar methodology to examine the relationship between cesarean delivery and postpartum maternal mortality. The authors showed that, after controlling for indication bias and confounders, cesarean delivery was associated with a higher risk of maternal death (OR, 2.87; 95% CI, 1.63 to 5.06). This risk was predominantly due to postpartum hemorrhage and complications of anesthesia.

A cohort study of deliveries in Canada between 1991 and 2005 compared 46,766 planned cesarean deliveries for breech presentation with 2,292,420 planned vaginal deliveries in which labor was either spontaneous or induced. Planned cesarean delivery increased the risk for postpartum cardiac arrest (aOR, 5.1; 95% CI, 4.1 to 6.3), major puerperal infection (aOR, 3.0; 95% CI, 2.7 to 3.4), anesthesia complications (aOR, 2.3; 95% CI, 2.0 to 2.6), and puerperal venous thromboembolism (aOR, 2.2; 95% CI, 1.5 to 3.4).

A medical record review of all in-hospital maternal deaths that occurred in a sample of US hospitals between 2000 and 2006 sought to identify evidence for a causal connection between mode of delivery and the mechanism of maternal death. Among 1,461,270 live births, there were 95 maternal deaths (6.5 per 100,000). Although 61% of the deaths were associated with cesarean delivery, one-third of these were perimortem cesarean deliveries in which the surgical procedure followed maternal cardiac arrest. Four deaths were thought to have been directly caused by cesarean delivery (attributed to hemorrhage or infection), with an additional seven deaths attributed to pulmonary thromboembolism after cesarean delivery. Two deaths were thought to have been causally related to vaginal delivery (one case of uterine inversion and one case of rupture of an unrecognized cerebral berry aneurysm during labor), and two deaths were attributed to pulmonary thromboembolism after vaginal delivery. Another 16 deaths were thought to have been potentially preventable had a cesarean delivery or an earlier cesarean delivery been performed (12 caused by preeclampsia, 3 caused by hemorrhage, and 1 caused by sepsis). The investigators concluded that a policy of universal thromboprophylaxis for all patients undergoing cesarean delivery would eliminate the increased risk for maternal death caused by cesarean delivery compared with vaginal delivery.

Regardless of mode of delivery, risk appears to be particularly concentrated in women with preexisting medical conditions; pulmonary hypertension, malignancy, systemic lupus erythematosus, sickle cell disease, and major cardiovascular and renal disease all confer substantially increased risk for maternal death or end-organ injury. Given the preponderance of indirect deaths documented in recent CEMD reports, assessors have repeatedly recommended both preconception counseling and intensive multidisciplinary antepartum and intrapartum care for women with serious medical or mental health conditions that may be aggravated by pregnancy. These conditions include congenital or acquired cardiac disease, obesity with a BMI ≥ 30 kg/m, epilepsy, diabetes, asthma, autoimmune disorders, renal or liver disease, HIV infection, and a personal or family history of severe mental illness. The “One Key Question Initiative” encourages clinicians to ask women the question “Would you like to become pregnant in the next year?” as a way to open conversations about contraception and the need for preconception counseling. It is further recommended that women at highest risk for adverse health events in future pregnancies be offered long-acting reversible contraception (LARC) as a strategy to improve maternal safety. The placement of LARC in the immediate postpartum period may help avoid systemic barriers to contraceptive use.

Some health system characteristics have been associated with a higher risk for maternal death; they include (1) low maternal-fetal medicine specialist density and (2) a single physician functioning as both the obstetrician and the anesthesia provider.

Death is considered the extreme outcome of the following continuum of adverse pregnancy events: normal pregnancy → morbidity → severe morbidity → near-miss → death. Approximately one-half of women experience some morbidity during pregnancy, most commonly anemia, urinary tract infection, mental health conditions, hypertensive disorders, and pelvic or perineal trauma. Research has focused on severe morbidity, near-miss events, and maternal deaths to elucidate the patient, provider, and health system factors that lead to these adverse outcomes.

Severe maternal morbidity includes unexpected complications of childbirth that result in significant short- or long-term consequences to a woman’s health ; its estimated incidence depends on the health system evaluated, the method of ascertainment, and the definition of severe morbidity used.

Near-miss morbidity occurs when a woman survives a life-threatening complication that occurred during pregnancy, childbirth, or within 42 days of termination of pregnancy. The concept of a near-miss event evolved from early studies of pregnant patients who required intensive care. The WHO Working Group on Maternal Mortality and Morbidity Classification determined that end-organ injury represents the most epidemiologically sound way to identify near-miss morbidity. In 2009 the WHO proposed a panel of clinical, laboratory-based, and management-based criteria to identify maternal end-organ injury. Cecatti et al. applied the WHO near-miss criteria to all ICU admissions in a tertiary care facility in Brazil between 2002 and 2007 and found a near-miss morbidity ratio of 14.7 events per 1000 live births, an MMR of 125 deaths per 100,000 live births, and a maternal near-miss–to–mortality ratio of 10.7:1. Subsequently, in 2011, the WHO released a maternal near-miss approach for “monitoring the implementation of critical interventions in maternal healthcare,” which aims to provide a systematic process for assessing and improving the quality of healthcare. The inclusion criteria for baseline assessment of quality of care are shown in Box 39.2 .

A review assessing the applicability of the WHO near-miss criteria to sub-Saharan Africa showed that the tool is not uniformly applied, and suggested challenges with the laboratory and management-based criteria. A broader systematic review focusing on LMICs confirmed multiple recommendations to adapt these laboratory and management-based criteria.

Despite these concerns, a systematic review published in 2022 evaluated global maternal near-miss frequencies using the WHO tool from 2011 to 2018. The global pooled near-miss estimate was 1.4% (95% CI, 0.4% to 2.5%). While there was not substantial between-study heterogeneity, the criteria were modified in some LMICs to exclude some laboratory tests and interventions due to resource restrictions.

Preventability has emerged as an important concept in maternal mortality and near-miss morbidity reviews because the opportunities for prevention identified from these reviews can be used to prioritize changes in clinical policy and health system improvements. The Perinatal and Maternal Mortality Review Committee in New Zealand proposed a comprehensive list of contributory factors, including those attributed to the healthcare organization, personnel, technology and equipment, the environment and geography, and patient-level barriers to accessing or engaging with care. If all these contributory factors are addressed, experts estimate that up to 80% of deaths could be prevented, but if changes are limited to modifications in clinical care, only 20% to 45% are likely preventable.

Preventability is a key strategy aimed at achieving the SDG by 2030. The WHO has released a document entitled “Ending Preventable Maternal Mortality (EPMM): A Renewed Focus for Improving Maternal and Newborn Health and Well-being.” This report outlines a systems approach that is aimed to be context-specific, people-centric, and which prioritizes equity. Five targets were set as goals for 2025, including (1) antenatal care contacts, (2) skilled health personal attending births, (3) early routine postnatal care, (4) emergency care coverage, and (5) the proportion of women informed and empowered to make their own decisions regarding sexual relations, contraceptive use, and reproductive healthcare.