The exact measures of maternal death are defined differently by the different agencies collecting the statistics. The World Health Organization (WHO) defines the maternal mortality ratio (MMR) as the number of direct and indirect maternal deaths per 100,000 live births; it does not include late maternal deaths occurring more than 42 days but less than 1 year after pregnancy termination. In the United States, the pregnancy-related mortality ratio tallies the pregnancyrelated deaths per 100,000 live births; it includes maternal deaths while pregnant and within
1 year of the termination of pregnancy that occur from any cause related to or aggravated by the pregnancy regardless of the duration or site of that pregnancy but excludes accidental or incidental causes. The maternal mortality rate used by the United Kingdom Confidential Enquiry includes direct and indirect maternal deaths per 100,000 maternities, defined as pregnancies resulting in a live or stillbirth ≥20 gestational weeks; late maternal deaths are not included.

Developed versus developing countries have a markedly different incidence of maternal death; 99% of maternal deaths occur in developing countries. According to the WHO in 2010, there were approximately 210 maternal deaths per 100,000 live births globally with a range of uncertainty of 170 to 300 deaths per 100,000 live births. This is a decrease from the 1990 WHO estimates of a worldwide MMR of 400. The 2010 MMR’s were 16 (range of uncertainty 14 to 18) and 240 (range of uncertainty 190 to 330) for developed and developing regions, respectively. Hemorrhage, hypertensive disorders of pregnancy, and sepsis cause more than half of the maternal deaths globally and about one-third of the deaths in the developed world. Obstetric hemorrhaging is the single most common cause of maternal mortality worldwide and accounts for about 25% of all postpartum deaths. Anemia, obstructed labor, HIV/AIDS, and unsafe abortions also play a role. Early marriage (a girl <18 years old) is a major health risk for girls; for girls, ages 15 to 19 years, pregnancy is the leading cause of death worldwide. Women in their 20s are five times less likely to die in childbirth than girls younger than 15 years old. In developed countries, hypertensive disorders of pregnancy, embolic disorders, and hemorrhage account for slightly less than 50% of maternal deaths. Fourteen percent are attributed to indirect deaths due most commonly to cardiovascular disease, and the risk of maternal death is increased by cesarean section delivery, advanced maternal age, non-white race, maternal obesity, and multifetal gestation.

In the United States, based on data from the Pregnancy Mortality Surveillance System of the Centers for Disease Control and Prevention, the aggregate 1998 to 2005 pregnancy-related mortality ratio (PRMR) of 14.5 maternal deaths per 100,000 live births was higher than those reported for any period in the previous 20 years. Hemorrhage accounted for 12% of these pregnancy-related maternal deaths, whereas hypertensive disorders of pregnancy, sepsis, thrombotic pulmonary embolism, cardiomyopathy, other cardiovascular disorders, and noncardiovascular medical disorders each accounted for 10% to 13% of the deaths. Some data suggests that the MMR in the United States may be increasing; the reason for this increase is unclear. It may be that more actual deaths are occurring, that reporting has improved, or that there is an artificial increase due code changes with the new International Statistical Classification of Diseases, 10th Revision (ICD-10) that was begun in 1999. Clark et al. evaluated maternal deaths from 2007 to 2012 after introducing disease-specific protocols. Their MMR was 6.4 of 100,000 live births in greater than 1.2 million deliveries, and the disease-specific protocols aided in the reduction of maternal deaths from hypertensive disease and postoperative pulmonary embolism but not hemorrhage.

In the United Kingdom, the Confidential Enquiry into Maternal Deaths found that from 2006 to 2008, the overall maternal mortality rate was 11.29 per 100,000 maternities with a significant decline in direct deaths rates from 6.24 in 2003-2005 to 4.67. The reduction was due mainly to a decrease in deaths due to thromboembolism and to a lesser extent from hemorrhage. Mortality from indirect causes, most commonly cardiac disease, remained unchanged. Substandard care was found in 70% of the direct deaths and 55% of the indirect deaths.

In developed regions, the estimated anesthesia-related MMR ranges between 1 and 3 deaths per 1,000,000 live births. If a maternal death can be attributed to an anesthetic complication and if an ideal anesthetic would have likely avoided this death, the death is considered anesthetic-related. Based on data from the United States, historically, cesarean sections under general anesthesia were more hazardous than those under neuraxial anesthesia, but case fatality rates for general anesthesia have decreased considerably, whereas those for neuraxial anesthesia have risen slightly leading to no significant difference between the two in the 1997 to 2002 time period. Looking at 257,000 parturients receiving neuraxial or general anesthesia between 2004 and 2009, the Serious Complication Repository Project of the Society for Obstetric Anesthesiology and Perinatology found that approximately 1 of 3,000 patients had serious anesthesia-related complications with high neuraxial block,
respiratory arrest in the labor suite, and unrecognized spinal catheters being the most common of these complications.

Up to one-fourth of pregnant women may have antepartum vaginal bleeding, but only a small fraction experience antepartum hemorrhage. Most of this bleeding occurs in the first trimester, and bleeding in the first 20 weeks is presumed to be a threatened abortion or an ectopic pregnancy. Even if pregnancy loss does not occur, the risk for adverse outcomes later in pregnancy is increased. The rest of the discussion will be confined to pregnancies with a viable living fetus. Antepartum bleeding generally poses a greater threat to the fetus than the mother. The most common causes of antepartum hemorrhage are placenta previa and placental abruption. From the Canadian Perinatal Network, Sabourin et al. looked at 806 women with antepartum hemorrhage between 22 and 28 weeks of gestation. The most common causes were placental abruption (31.7%), placenta previa (21.2%), and cervical bleeding (6.6%), whereas no cause for hemorrhage was found in a third of the cases. Placenta previa is the abnormal implantation of the placenta over or adjacent to the cervical os and in front of the fetal presenting part. Placental abruption, or abruptio placentae, refers to the premature separation of a normally implanted placenta from the uterus’ decidua basalis before delivery of the fetus.

Rare causes of antepartum bleeding include velamentous insertion of the umbilical cord with the involved placental vessels overlying the cervix, which is known as vasa previa, uterine rupture, cervicitis, and local genital tract lesions.

In the remaining cases, it is not always possible to identify the source of antepartum bleeding. Usually, this type of bleeding begins with minimal symptoms, if any, and stops; no anatomic cause is identified at delivery, and it is likely attributable to marginal placenta separation. However, these pregnancies have a higher risk of poor neonatal outcome and preterm delivery.

With antepartum uterine bleeding after midpregnancy, placenta previa and placenta abruption should always be considered in the differential diagnosis. Digital examination of the
cervix should be avoided until the diagnosis of placenta previa has been excluded because digital examination of a placenta previa can cause immediate, severe hemorrhaging. Classically, it has been taught clinically that painless uterine bleeding is indicative of placenta previa, while painful uterine bleeding signifies placental abruption. The differential diagnosis, however, is not actually that straightforward; for instance, labor or placental separation accompanying a previa may cause pain, whereas pain from an abruption may mimic normal labor. Transabdominal sonography is a simple and safe method of locating the placenta in a suspected placenta previa. Transvaginal sonography has been shown to be superior to transabdominal sonography and is currently the imaging modality of choice. Recently, transperineal sonography has been shown to have a 98% positive-predictive value and 100% negative-predictive value of placenta previa. On the other hand, sonography infrequently aids in the diagnosis of an acute placental abruption because fresh clot and placenta appear sonographically similar. Negative findings on sonographic examination do not rule out a placental abruption. Magnetic resonance imaging (MRI) has been used by numerous investigators to visualize placental abnormalities, including placenta previa and placental abruption, but is unlikely to replace sonography any time soon due to availability and cost limitations.

Based on the accuracy and safety of transvaginal sonography in detecting placental position, the “double setup” is rarely, if ever, used. The double setup was a vaginal examination done in an operating room with the patient and medical personnel completely prepared to do an emergency cesarean section if hemorrhaging should occur.

The incidence of placenta previa in the United States is about 1 in 300 to 400 deliveries. The risk of having placenta previa increases with advancing maternal age, multiparity, multifetal gestations, previous cesarean delivery, prior uterine surgery, cigarette smoking, cocaine use, infertility treatments, and unexplained increases in maternal serum α-fetoprotein levels. In approximately 90% of the patients with sonographic evidence of placenta previa at 20 weeks, the placenta moved away from the cervix and out of the lower uterine segment by term.

Placenta previa is associated with an increased risk of antepartum bleeding, intrapartum and postpartum hemorrhage, abnormalities of placental adherence, peripartum hysterectomy, blood transfusions, preterm birth with perinatal morbidity and mortality, congenital malformations, and septicemia. In the presence of a prior cesarean section, the patient with placenta previa is at greater risk for having a placenta accreta; this risk increased as the number of cesarean sections increased. Silver et al. found that in women with placenta previa, the risk of placenta accreta was 3%, 11%, 40%, 61%, and 67% for the first, second, third, fourth, and fifth or greater repeat cesarean deliveries, respectively.

Classifications and terminology for placenta previa based on the degree to which the internal cervical os is covered have been confusing; they have been simplified to two terms. Currently, placenta previa is a term used to describe a placenta that partially or totally covers the internal cervical os, whereas the term low-lying placenta refers to a placenta implanted in the
lower uterine segment but remains at least 2 cm away from the os; this was previously called a marginal placenta previa. Normally, the placenta implants in the upper uterine segment. Clearly, the classification in some cases of placenta previa depends on the degree of cervical dilation at the time of cervical assessment. In vasa previa, the fetal vessels course through the membranes and cover the internal cervical os.

Obstetric management of placenta previa patients depends upon three things: fetal age and therefore maturity, presence or absence of labor, and severity of bleeding. The goal of management is to maximize fetal growth and minimize the possibility of antepartum hemorrhage. With many previas, the first episode of bleeding usually begins without warning or pain, stops spontaneously, and rarely causes maternal or fetal compromise, but bleeding often recurs. A mature fetus with or without labor should be delivered; if bleeding occur at or beyond 36 weeks, immediate delivery is indicated. If hemorrhaging is severe or there is concern about the fetal condition, delivery should take place regardless of gestational age. Otherwise, conservative management may safely prolong the pregnancy by an average of 4 weeks after the initial bleed. Between 24 and 34 weeks of gestation, betamethasone, a corticosteroid, often is given to the patient with placenta previa to accelerate fetal lung maturity.

With a preterm fetus and no persistent active uterine bleeding, management consists of close observation for bleeding and activity restriction in an inpatient setting or, in certain selected patients, an outpatient setting. For stable patients with no bleeding in the previous 48 hours who have ready access to hospital care, maternal and fetal outcome did not differ between outpatient and inpatient management. Maternal vital signs and hemoglobin (Hgb) are checked at regular intervals depending on when the bleeding, if it exists, subsides. The American Association of Blood Banks recommends that a blood type and screen be done every 3 days in actively bleeding pregnant women because of the risk for developing new alloantibodies. Fetal well-being also is assessed regularly. The use of tocolytic therapy in placenta previa patients with preterm labor and bleeding is controversial; it has been used to prolong gestation if both the mother and fetus are stable.

Elective delivery for women with placenta previa should take place at 36 to 37 weeks, and if a placenta accreta is suspected, it has been recommended by Spong et al. that delivery take place after 34 weeks. As gestational age increases, there is an increased risk of significant bleeding that would necessitate delivery: 4.7% at 35 weeks, 15% at 36 weeks, 30% at 37 weeks, and 59% at 38 weeks. If an indication for delivery is present, the use of amniocentesis to assess fetal lung maturity would not aid in guiding management.

Patients with placenta previa almost always are delivered by cesarean section, and some contend that all women whose placentas lie within 2 cm of the cervical os should be delivered by cesarean section. When the placenta is greater than 1 cm away from the cervical os, however, Vergani et al. and Bronsteen et al. suggest that vaginal delivery may be safe. Normally, with placenta removal, the myometrium of the uterus contracts and reestablishes vascular integrity by constricting the open vessels in the placental bed. In patients with placenta previa, bleeding after placenta removal can be brisk because the lower uterine segment contracts poorly and therefore does not constrict the avulsed vessels. There is a known association between placenta previa and placenta accreta that may further exacerbate uterine bleeding and require additional surgical and/or radiologic interventions to staunch the bleeding.

As needed, adequate intravenous access and laboratory studies, including a complete blood count, coagulation profile, and a type and screen/crossmatch should be obtained as circumstances dictate. Rh (anti-D) immune globulin is routinely administered to Rh-negative women at 28 weeks of gestation when the fetal blood type is unknown or known to be Rhpositive and again after delivery if the neonate is Rh-positive. With an acute bleeding episode
in an Rh-negative parturient, a qualitative rosette test or the quantitative Kleihauer-Betke test or flow cytometry should be done to determine if a fetal-maternal transfusion has occurred and if additional immune globulin is required.

Vasa previa refers to intramembranous fetal vessels, unprotected by placenta or umbilical cord, coursing through the membranes over the cervical os and under the fetal presenting part. The incidence is approximately 1 in 2,500 deliveries and associated risk factors include a second trimester low-lying placenta/placenta previa, a placenta with accessory lobes, velamentous umbilical cord insertions, multiple pregnancies, and pregnancies from in vitro fertilization. An undiagnosed vasa previa is associated with a high fetal mortality (about 60%). Because fetal blood volume is small, any bleeding puts the fetus in jeopardy. When membranes rupture, either spontaneously or artificially, the fetal vessels running through these membranes often rupture. Vaginal bleeding with fetal distress or death when the membranes rupture is seen in an undiagnosed vasa previa and necessitate an emergency cesarean section based on fetal distress. If vaginal blood is present, the presence of fetal (RBCs) can be determined using a Kleihauer-Betke test or flow cytometry, although there is rarely time for this. Fortunately, the majority of cases of vasa previa can be diagnosed prenatally by routinely evaluating placental cord insertion by ultrasound examination and transvaginal sonographic color Doppler if the insertion site cannot be delineated clearly. Good outcomes in patients with vasa previa require a prenatal diagnosis and delivery by cesarean section before the membranes rupture. Robinson and Grobman suggest that women with vasa previa be delivered by 34 to 35 weeks of gestation to achieve the best balance between the risks of perinatal morbidity and mortality; at any given gestational age, amniocentesis for fetal lung maturity verification did not improve outcomes.

The frequency of placental abruption is about 1 in 200 deliveries. The premature separation of the normally implanted placenta can be total or partial and can lead to external or concealed bleeding. Concealed hemorrhage is associated with greater maternal and fetal risks. The most common conditions associated with placental abruption are gestational hypertension, preeclampsia, and chronic hypertension. The incidence of placental abruption is increased with increased maternal age, African-American or Caucasian race, prematurely ruptured membranes, cigarette smoking, cocaine use, leiomyomas, history of previous abruption, external trauma, and possibly lupus anticoagulant and high parity. The bleeding almost always is maternal because the placental separation is within the maternal decidua.

The signs and symptoms of placental abruption vary considerably. Ultrasonography often does not confirm the diagnosis as the placenta and fresh clots have similar appearances. In severe placental abruptions, the diagnosis often is obvious with fetal distress and demise common. With a live fetus and vaginal bleeding, milder abruptions may be difficult to diagnose. If knowing would change the patient’s management, MRI is a highly sensitive test for detecting placental abruption and should be considered.

Treatment depends on fetal gestational age and fetal and maternal condition. If the diagnosis is uncertain and the fetus is uncompromised, inpatient expectant management with close observation can be practiced. However, further placental separation can occur at any time and lead to serious fetal compromise or death. The use of tocolytics to prolong pregnancy is controversial. Rapid delivery of the compromised, but alive, fetus is by cesarean delivery. Kayani et al. looked at 33 pregnancies with a clinically overt placental abruption and fetal bradycardia. Of the 22 of the neurologically intact neonates, 15 were delivered within a 20-minute decision to delivery time, whereas 8 of the 11 neonates who died or had cerebral palsy on 1 year follow-up had a >20-minute interval.

If placental separation has led to fetal death, a vaginal delivery usually is preferred; there are two exceptions: a massive hemorrhage that cannot be managed with vigorous blood replacement and obstetric complications that prevent vaginal delivery will necessitate an operative delivery.

Maternal complications with severe abruption can include postpartum hemorrhage, hemorrhagic shock, a consumptive coagulopathy, acute renal injury, and Sheehan syndrome (pituitary failure). Placental abruption can lead to fetal compromise or fetal demise.

Primary uterine rupture is defined as a rupture occurring in a previously intact unscarred uterus, whereas a secondary rupture is associated with a preexisting myometrial incision, injury, or anomaly. In developed countries, Getahun et al. report the uterine rupture incidence to be 1 in 4,800, whereas in developing countries, the incidence is higher. Uterine rupture is typically classified as either complete, with all of the uterine wall layers separated, or incomplete (uterine dehiscence), with the uterine muscle separated but the visceral peritoneum intact. Uterine dehiscence is more common; it is less likely to cause adverse maternal and fetal derangements or necessitate cesarean delivery. The morbidity and mortality rates are higher when the rupture is complete. Prior cesarean delivery is the greatest risk factor associated with uterine rupture. Fortunately, uterine ruptures are rare. Other risk factors associated with uterine rupture include previous uterine surgery involving the myometrium, a tumultuous labor, difficult forceps delivery, uterine trauma, breech extraction, external and internal version, labor stimulation with oxytocin or prostaglandins, inappropriate uterotonic medication use, uterine anomalies, placenta increta/percreta, connective tissue defects such as Marfan or Ehlers-Danlos syndrome, and uterine overdistention by fetal macrosomia, hydramnios, or multifetal pregnancy.

Uterine rupture can be difficult to diagnose because of the varied symptoms and physical findings that can occur. The most consistent sign of uterine rupture is a nonreassuring fetal heart rate pattern. In some patients, the symptoms of uterine rupture will be indistinguishable from those of placental abruption. Other signs that may be present include hypotension, vaginal bleeding, abdominal pain, change in the uterine contour, changes in the uterine
contraction pattern, and cessation of labor. There may be little pain, and because most of these ruptures occur during labor, any pain may be masked by the analgesics given to treat the discomfort of labor.

With a complete uterine rupture, fetal distress and maternal bleeding occur; an emergent cesarean delivery is imperative to deliver the fetus and stabilize the mother. After delivery of the fetus, the uterus can be repaired, the uterine arteries can be ligated, or a hysterectomy can be performed, depending on the size of the rupture and the patient’s condition. Fetal distress and/or death are common while maternal death in the developed world is not.

Yes, this patient has a higher chance of postpartum bleeding because of her known placenta previa and the increased risk of having an associated placenta accreta syndrome. There is a strong association between placenta previa, prior uterine surgery, and placenta accreta syndromes. Clark et al. found that placenta previa patients had a 5%, 24%, and 67% risk of having an associated placenta accreta with no, one, and four or more prior cesarean sections, respectively. Similarly, in patients with placenta previa, Silver et al. found a 3%, 11%, 40%, 61%, and 67% incidence of placenta accreta with zero, one, two, three, and four or more prior cesarean sections, respectively.

Placenta accreta syndromes refer to a placenta that adheres abnormally to the myometrium of the uterus due to a partial or total absence of the decidual basalis and an incomplete development of the fibrinoid layer. The normal placenta has three general layers: the decidua, which is maternal in origin and adherent to the uterus; the chorion, including the chorionic villi; and the amnion. Both the chorion and amnion are of fetal origin. Normally, in the first few minutes after delivery of the neonate, the placenta separates from its implantation site at a physiologic cleavage line through the spongy layer of the decidua basalis. In placenta accreta syndromes, evidence suggests that not only is the decidua basalis scanty or absent but also that the cytotrophoblasts may exert some control of over decidua invasion via angiogenesis and growth expression allowing chorionic villi to attach and invade uterine myometrium. Placenta accreta occurs when the placental chorionic villi are attached directly to the uterine myometrium without a decidual layer. With placenta increta, the villi invade into the myometrium, and with placenta percreta, the villi penetrate through the myometrium and serosa of the uterus and, sometimes, into adjacent organs. Based on histologic evidence in patients with placenta accreta syndromes, placenta accreta occurs most frequently, increta occurs much less frequently, and percreta occurs the least. The abnormal adherence may involve the entire placenta (total placenta accreta), several placental lobules (partial accreta), or part or all of a single lobule (focal placenta accreta). The term placenta accreta often is used as a general term to describe all of these conditions.

The incidence of placenta accreta syndromes is increasing due mainly to rising cesarean section rates. The incidence was approximately 1 in 2,510 deliveries in the 1980s versus 1 in 533 deliveries for the period from 1982 to 2002. Decidual formation is commonly defective over a previous cesarean section scar, in the lower uterine segment, and/or after uterine curettage. Increasing cesarean section rate and placenta previa are by far the two most important risk factors. Maternal age greater than 35 years, increasing parity, endometrial defects (Asherman syndrome), submucous leiomyomata, and otherwise unexplained elevations in maternal serum α-fetoprotein concentration and B-human chorionic gonadotropin (B-hCG) are other risk factors associated with placenta accreta.

In patients with conditions such as placenta previa and previous uterine surgery that are highly associated with placenta accreta, it is important to have a high index of suspicion and to look carefully for signs of accreta beginning at 20 to 24 weeks of gestation. Prenatal diagnosis of placenta accreta allows for effective planning that helps minimize maternal morbidity. Transabdominal and transvaginal ultrasonography are useful modalities for evaluating placental position and implantation; the loss of the hypoechoic boundary between the placenta and bladder, placental continuity with the bladder wall, and intraplacental sonolucent spaces, venous lakes, or placental lacunae adjacent to the involved uterine wall are strongly suggestive of placenta accreta. The sensitivity and specificity of grayscale ultrasonography in detecting placenta accreta are approximately 77% to 87% and 96% to 98%, respectively. Color flow Doppler and MRI do not significantly improve the diagnostic sensitivity of grayscale ultrasonography alone. However, they can be used as adjunctive diagnostic tools. When the diagnosis is uncertain or the suspected placenta accreta is posteriorly placed, MRI may be helpful. Although antepartum diagnosis of placenta accreta would be ideal, currently, there is no single diagnostic modality that determines the diagnosis of placenta accreta with absolute accuracy. Unfortunately, the diagnosis of placenta accreta is occasionally made postpartum during the third stage of labor when separation of the placenta does not occur. The first clinical manifestation of placenta accreta is usually profuse, potentially life-threatening bleeding that occurs at the time of attempted manual placental separation. During vaginal delivery, uterine inversion can occur. Placenta accreta has become one of the most common reasons for peripartum hysterectomy.

Optimal management of placenta accreta has yet to be defined. Care should be individualized for each patient, and a multidisciplinary approach to delivery planning should be taken. Prenatal diagnosis and management appear to improve outcome. The problems associated with placenta accreta depend upon the site of implantation, depth of myometrial penetration, and the number of placental lobules involved. Only a focal placenta accreta may not require aggressive surgical intervention. Otherwise, surgical management can be difficult and requires a multidisciplinary approach in order to minimize maternal complications. If placenta accreta is known to exist antenatally, it is generally accepted that it is ideally treated with an elective cesarean section and hysterectomy; the placenta should be left in place during surgery. The optimal timing for the delivery of the patient with placenta accreta remains controversial
and should be individualized. In a retrospective series, Warshak et al. found that patients with a predelivery diagnosis of accreta, who had a planned en bloc hysterectomy without the removal of the placenta at 34 to 35 weeks of gestation after betamethasone administration, had decreased maternal hemorrhagic morbidity and no significant increase in neonatal morbidity. Optimal maternal and neonatal outcome in a stable patient seems to occur with a planned delivery at 34 weeks of gestation without amniocentesis. For women with suspected placenta accreta, some suggest preoperative placement of balloon catheters into the internal iliac arteries. After delivery of the fetus, the catheters may be inflated intermittently during hysterectomy, which may potentially decrease blood loss and provide optimum exposure of the operative field. They also may be used for intraoperative or postoperative embolization of persistent bleeders. The placement of these catheters is not risk-free; thrombosis and thromboemboli have been reported. The American College of Obstetricians and Gynecologists (ACOG) found that data on the efficacy of this intervention are unclear. Some investigators suggest that the placement of preoperative ureteral stents reduced morbidity in cesarean hysterectomies; others use stents intraoperatively as needed.