Chapter 18. Third-Trimester Vaginal Bleeding

Vaginal bleeding in the third trimester can be very worrisome to a patient and clinician. When evaluating a patient with this problem, it is important to consider all the possible diagnoses in order to arrive at the appropriate conclusion and treatment. The most common causes of third-trimester vaginal bleeding are:

• Cervical bleeding associated with cervical change
• Abruptio placentae
• Placenta previa
• Vasa previa

Cervical bleeding associated with cervical change will be discussed in the section on the evaluation of preterm labor and labor at term.

Abruptio placentae (placental abruption) is defined as the premature separation of the normally implanted placenta from the uterine wall after 20 weeks of gestation but prior to the delivery of the infant. It is diagnosed retrospectively, evident only when the inspection of the placenta reveals a clot over the placental bed with disruption of the underlying placental tissue. The placental tissue may not show overt evidence of disruption if the abruption-to-delivery interval is short. One-third of all antepartum bleeding in the third trimester is due to placental abruption, and it will occur in 1 in 75–225 deliveries. About 1 in 830 abruptions end in fetal demise.

Essentials of Diagnosis

• Bleeding from the vagina
• Uterine activity
• Fetal heart rate abnormalities
• Changes in maternal hemodynamic status

Pathogenesis

Placental abruption may be the end of a chronic vascular pathologic process or may be due to a single inciting event. Bleeding due to vascular disruption accumulates and tracks along the decidua, separating the placenta from the remaining decidual layer. This may result in a partial abruption, referring to a self-limited hematoma that does not dissect the placental attachment further, or it may proceed to a complete abruption, leaving no decidual interface intact.

Abruption can be classified into 3 broad categories that allow for a description of the clinical and laboratory findings (Table 18–1).

• Grade 1: A small amount of vaginal bleeding and abnormal uterine activity or irritability are usually noted. The fetal heart rate tracing is within normal limits. Maternal hemodynamic status is normal, and all coagulation studies and laboratory values are within normal limits.
• Grade 2: A mild to moderate amount of vaginal bleeding is noted. Uterine activity may be tetanic or with frequent palpable and painful contractions. The fetal heart rate tracing may show decreased variability or late decelerations. Maternal hemodynamic status shows signs of compensation, including orthostatic hypotension and tachycardia, while maintaining overall blood pressure. Maternal fibrinogen may be decreased.
• Grade 3: External uterine bleeding may range from mild (likely concealed) to severe. The uterus is typically painful and tetanic. Fetal death has occurred. Maternal hemodynamic status is unstable, showing signs of severe volume depletion with hypotension and tachycardia. Thrombocytopenia and coagulation panel value abnormalities are present. Fibrinogen concentration level is typically <150 mg/dL.

The primary cause of placental abruption is unknown in most cases; however, it has been linked to several risk factors. Mechanical force or trauma is sometimes implicated, usually as a result of domestic violence or a motor vehicle accident. The clinical picture following a traumatic event can be severe in nature, and symptoms typically present within 24 hours. Blunt abdominal trauma can compress the placental interface, allowing for a shearing effect when decompression occurs. Rapid deceleration is the typical inciting event during a motor vehicle accident. Prolonged continuous fetal monitoring is critical in assuring the safety of the pregnancy. The mechanical force of rapid decompression of the uterus is an uncommon etiology. This can occur after the delivery of the first infant during the vaginal delivery of a twin gestation or after rupture of membranes in a patient with polyhydramnios.

Several factors can put a patient at risk for this disease. Maternal hypertension (>140/90 mm Hg) has been strongly associated with the occurrence of placental abruption. In fact, all hypertensive disorders of pregnancy have been implicated as risk factors. Forty to fifty percent of patients with grade 3 abruptions are found to have hypertensive disease of pregnancy.

Smoking is associated with a significant increase in the incidence of placental abruption, with a 2.5 relative risk of abruption severe enough to cause fetal death. Hypertension with concomitant cigarette smoking during pregnancy further increases the risk of abruption.

The relationship between age and the risk of placental abruption is not clear when other confounding factors, such as hypertension and multiparity, are corrected. Most authors feel that there is no direct correlation; however, a Norwegian study done over 15 years showed a strong relationship between placental abruption and maternal age among all levels of parity.

Increasing parity is a risk factor for placental abruption. The incidence in primigravidas is 1%, and this incidence increases to 2.5% in grand multiparas, correcting for confounding factors. This may be due to impaired decidualization after the implantation of multiple past placentas on the uterine wall.

Acquired or inherited thrombophilias may be associated with an increased risk of abruption. Factor V Leiden and prothrombin gene mutations have been implicated in multiple studies. Less common inherited thrombophilias, including protein C, protein S, and antithrombin deficiencies, may also be associated.

Approximately 2–5% of pregnancies that are complicated by preterm premature rupture of membranes will also be complicated by placental abruption. There is an even more substantial risk in patients who subsequently develop chorioamnionitis after rupturing membranes. It is difficult to ascertain whether the abruption is the cause or an effect of the ruptured membranes.

There is a well-known association between cocaine abuse and placental abruption, and up to 10% of women who use cocaine during pregnancy will incur this complications. It is likely due to acute vasoconstriction and disruption of vascular integrity directly as a result of cocaine use.

A history of placental abruption significantly predisposes a patient to another placental abruption. Five to 17% of all pregnancies complicated by abruption will recur in subsequent pregnancies. After 2 abruptions, the risk of recurrence increases to 25%. The reason for this association is unknown.

Clinical Findings

Symptoms & Signs

Most placental abruptions will present with the clinical triad, although many will not fill all 3 of the following categories:

• Fetal distress or fetal death: Fetal distress is typically the first clinical sign among patients who have continuous fetal monitoring during the abruption. A nonreassuring fetal heart rate tracing or poor biophysical profile score may indicate fetal compromise from a decrease in the placental exchange surface area or from severe maternal hypotension due to a large blood loss.
• Tetanic uterine activity (ie, contractions).
• Uterine bleeding, external or concealed.

Prompt evaluation is critical in caring for the patient with third-trimester uterine bleeding and possible placental abruption. All other common and potentially life-threatening causes of bleeding should be ruled out, including placenta previa, vasa previa, vaginal trauma, and vaginal or cervical malignancy. Once all other causes have been ruled out, placental abruption becomes the most likely diagnosis. Nearly 80% of patients with placental abruption will present with the complaint of vaginal bleeding. The 20% who do not exhibit this hallmark will commonly be diagnosed with labor or preterm labor. Patients occasionally complain of symptoms consistent with uterine tenderness and will be found to have increased uterine tone upon physical exam. Patients with these symptoms upon presentation will have a more severe abruption.

Laboratory Findings

Blood type and Rh status, hemoglobin, hematocrit, platelet count, coagulation studies, and fibrinogen level should be sent. A Kleihauer-Betke test should be sent for all women who are Rh negative.

Imaging Studies

Ultrasound has become important in the diagnosis and characterization of placental abruption. More than 50% of patients with confirmed abruption will have evidence of hemorrhage on ultrasound. Echogenicity, size, and location of the hemorrhage can be described, allowing the clinician to better understand the timing and severity of the abruption. If the ultrasound is performed during the early phases of the abruption, the area of hemorrhage will appear isoechoic or hyperechoic compared with the echogenicity of the placenta. The hematoma becomes hypoechoic within 1 week and sonolucent within 2 weeks of the initial hemorrhage.

The size and location of the hematoma are important in evaluating the severity of the abruption. A larger hematoma is associated with a worse prognosis than a smaller hematoma. A retroplacental hemorrhage has a worse prognosis than a subchorionic hemorrhage, which is defined as a collection of blood between the chorion and the decidua. A retroplacental hemorrhage that is >60 mL in volume has at least a 50% morality rate associated with it.

Prevention

Although no intervention has been shown to prevent placental abruption, known risk factors for placental abruption include poorly controlled maternal hypertension, smoking, and cocaine use. Counseling patients against smoking and cocaine use and helping them find appropriate cessation programs may reduce the risk of abruption, as may antihypertensive medication for women with poorly controlled hypertension.

Treatment

The hemodynamic status of the mother should be immediately evaluated and stabilization performed if necessary. Two large-bore intravenous lines should be placed, and the fetal heart rate should be monitored continuously. Crystalloid infusion should be started to rapidly correct the volume deficit, and packed red blood cells should be given if severe anemia is evident or if there is continued uterine bleeding. Urine output should be maintained above 30 mL/h. If no transfusion is required immediately, 4 units of packed red blood cells should be crossed and held nearby. Fresh frozen plasma should be administered for a fibrinogen level <100 mg/dL, and platelets should be given if the platelet count is <20,000 or <50,000 for a patient with severe continued hemorrhage or a requirement for emergent caesarean delivery. The clinician should not be falsely reassured by normal blood pressure and pulse, as the patient may have had hypertension previously and has had too rapid a volume loss to produce tachycardia.

Mode and timing of delivery depend greatly on the severity of the placental abruption and the gestational age of the pregnancy. If a grade 1 abruption has occurred and the gestational age is >37 weeks, the most appropriate course of action is induction or augmentation of labor with very close monitoring of the maternal and fetal status at all times. In the case of a preterm fetus with a grade 1 abruption, conservative management is indicated. Expectant management with short-term hospitalization has been shown to prolong the length of the pregnancy and has not been shown to increase morbidity or mortality of the fetus or the mother. If conservative management is pursued, administration of corticosteroids to promote fetal lung maturity is indicated if the gestational age is <34 weeks.

If the decision has been made to deliver the infant, close monitoring throughout labor and delivery is imperative. The hemodynamic status of the mother needs to be known at all times, and the fetal heart rate should be evaluated continuously. Serial coagulation studies and complete blood count should be obtained. If possible, an intrauterine pressure catheter should be use to evaluate uterine tone. If it appears that the patient's labor curve is progressing normally, a vaginal delivery is the preferable method of delivery.

An emergent caesarean section may be necessary at any point during labor. An increase in the resting tone of the myometrium may indicate a worsening of the abruption, and the possible compromise of blood flow to the fetus is an indication for emergent delivery. Other indications include a nonreassuring fetal heart rate tracing, severe hemorrhage, and disseminated intravascular coagulation. If an indication for emergent delivery has been identified, the maternal hemodynamic status should be quickly stabilized and correction of coagulopathy should be performed before proceeding. If the fetal status is reassuring and the caesarean is for maternal reasons, the correction of hypovolemia and coagulopathy can be done in a more controlled manner to avoid fluid overload prior to the surgery.

Prognosis

Outcome depends significantly on the gestational age of the fetus and the severity of the abruption. Prematurity, intrauterine growth restriction, caesarean delivery, and perinatal mortality are all increased in pregnancies that are complicated by placental abruption. In 1999, a review found that a pregnancy complicated by abruption, of any level of severity, had a 9-fold increase in the incidence of perinatal death, a 2-fold increase in the incidence of intrauterine growth restriction, and a 4-fold increase in the incidence of preterm birth.

Placental abruption is also associated with a substantial recurrence risk in future pregnancies of approximately 5–17%. After a history of 2 pregnancies affected by placental abruption, the risk of recurrence is approximately 25%. Unfortunately, there is no intervention that has been demonstrated to reduce the risk of recurrence. Preconception counseling should be targeted toward eliminating known risk factors such as smoking or cocaine use.

When the placenta implants such that the placental tissue is located adjacent to or overlying the internal cervical os, it is called placenta previa. Placenta previa is the leading cause of third-trimester bleeding, complicating 4 in 1000 pregnancies over 20 weeks. The incidence is higher in early pregnancy prior to the development of the lower uterine segment, and most of these previas resolve as the pregnancy progresses.

There are 3 types of placenta previa (Fig. 18–1):

• Marginal placental previa: Characterized by the placenta being proximate to the margin of the internal os. It does not cover the os.
• Partial placenta previa: The placenta partially occludes the os, but does not completely cover it.
• Complete placenta previa: The internal os is fully covered by the placenta. This type is associated with the greatest risk of morbidity and mortality because it can cause the greatest amount of blood loss.

Essentials of Diagnosis

• Painless vaginal bleeding
• Ultrasonographic findings consistent with placenta previa

Pathogenesis

When a placenta implants in the lower part of the uterus, the pregnancy is at risk for placenta previa. There are several risk factors, including multiparity, increasing maternal age, history of prior caesarean section or uterine surgery, and multiple gestation.

Multiparas are at higher risk for placenta previa compared to nulliparas. The incidence of previa in nulliparas is 0.2%, whereas grand multiparas have an incidence of 5%. The theory behind this phenomenon is that once a placenta has implanted into a certain part of the uterine wall, it has permanently altered its constitution, making implantation at a different site more likely in subsequent pregnancies. Increasing maternal age has been a risk factor, and the cause of this association is unclear. The increased risk may be due to higher parity in older mothers, but it may also be an independent risk factor.

The risk of placenta previa increases directly with the number of uterine surgeries a patient has had in the past. This is mostly seen with increasing numbers of caesarean sections. The risk of placenta previa in second pregnancies after a first pregnancy delivered by caesarean section is 1–4%. The risk increases to nearly 10% in patients with 4 or more previous caesarean deliveries. Furthermore, it has been suggested that previas identified in the second trimester in patients with a previous caesarean delivery have a lower likelihood of resolving as the pregnancy progresses. Risk increases with previous curettage for spontaneous or induced abortion, thought to be due to a scarred active segment of the uterus.

Other risk factors for placenta previa include multiple gestation and smoking. This is due to the greater surface area of the placenta in these situations.

Placenta previa can be associated with several other conditions, including malpresentation, preterm premature rupture of membranes, and intrauterine growth restriction. There may also be an increased risk of congenital anomalies; however, there is no association with any specific anomaly.

Patients with placenta previa are at higher risk for developing placenta accreta, increta, or percreta.

• Placenta accreta: There is no decidua basalis, and the fibrinoid layer is incompletely developed.
• Placenta increta: The placenta invades the myometrium.
• Placenta percreta: The placenta penetrates the myometrium and may invade nearby viscera.

Previous uterine surgery is the risk factor most associated with placenta accreta. Patients with no prior uterine surgery and placenta previa will have accreta 4% of the time. Patients with 1 prior uterine surgery and placenta previa will have accreta 10–35% of the time. Multiple prior caesarean deliveries and placenta previa incurs a 60–65% risk of accreta. Two-thirds of patients with placenta accreta will require a caesarean-hysterectomy.

Clinical Findings

Symptoms & Signs

Placenta previa typically presents with painless vaginal bleeding, usually in the third trimester. Lack of pain with the presence of bleeding is what distinguishes placenta previa from placental abruption. The bleeding occurs in conjunction with the development of the lower uterine segment. As the myometrium becomes thinner, the placenta–decidua interface is disrupted, causing bleeding. The thinness of the lower uterine segment prevents it from contracting to minimize the bleeding from the uterine surface of the implantation site; however, sometimes the bleeding itself can irritate the myometrium and precipitate contractions.

The clinician should have a high index of suspicion for placenta previa in all patients who present with bleeding after 24 weeks. One-third of patients with placenta previa will present with bleeding before 30 weeks, one-third will present between 30 and 36 weeks, and one-third will present after 36 weeks. Ten percent of all women with previa will reach full term without an episode of bleeding. On average, a patient's first episode of bleeding will occur at 34 weeks, with delivery at 36 weeks. Risk of perinatal mortality and morbidity decreases linearly as gestational age increases.

Laboratory Findings

Baseline admission labs including blood type and Rh status, hemoglobin, hematocrit, and platelet count should be sent. Coagulation studies and fibrinogen concentration are not as important in patients with previa as in patients with abruption; however, if there is any doubt of the diagnosis, these should also be sent. A Kleihauer-Betke test should be sent for all women who are Rh negative.

Imaging Studies

Prior to the advent of routine second-trimester ultrasound, patients with placenta previa were diagnosed at the onset of bleeding. Currently, most cases are diagnosed by ultrasound in the second trimester, although most of these will resolve. Five to fifteen percent of all patients will have placenta previa at 17 weeks. Ninety percent of these will resolve by 37 weeks. This occurs because as the lower uterine segment develops, more distance is created between the placenta and the cervix. Complete previa and marginal or partial previa diagnosed in the second trimester will persist in 26% and 2.5% of patients, respectively. All patients who have placenta previa diagnosed before 24 weeks should have a sonogram between 28 and 32 weeks to reassess the position of the placenta.

Treatment

The initial management of a patient with a bleeding placenta previa is very similar to the initial management of a patient with placental abruption. Hemodynamic status of the mother should be immediately evaluated and stabilization performed if necessary. Large-bore intravenous lines should be placed, and the fetal heart rate should be monitored continuously.

Hemodynamic stabilization should be performed immediately. Crystalloid infusion should be started in order to rapidly correct a volume deficit, and packed red blood cells should be given if severe anemia is evident or if there is continued uterine bleeding. The goal hematocrit is at least 30% if the patient is bleeding. If no transfusion is required immediately, 4 units of packed red blood cells should be crossed and held nearby. The urine output should be maintained above 30 mL/h.

Subsequent management depends on gestational age, stability of the mother and fetus, the amount of bleeding, and presentation of the fetus. Delivery is always indicated if there is a nonreassuring fetal heart rate pattern despite resuscitation efforts, including maternal supplemental oxygen, left-side positioning, or intravascular volume replacement; if there is life-threatening maternal hemorrhage; or if the gestational age is >34 weeks and there is known fetal lung maturity. If the fetus is ≥37 weeks of gestational age and there is persistent bleeding or persistent uterine activity, delivery is also indicated. Digital cervical exams should be avoided.

Between 24 and 36 weeks, if maternal and fetal stability and well-being are assured, conservative expectant management may be indicated. About 75% of patients with symptomatic placenta previa are candidates for conservative management, and 50% of these patients can prolong their pregnancy by at least 4 weeks. Thirty percent of patients treated this way will progress to term without bleeding again. Seventy percent will have at least 1 more episode of bleeding, and 10% of these patients will have a third episode.

Conservative management, after stabilization has occurred and little or no uterine bleeding is noted, consists of several steps. Hydration and blood transfusion are given if necessary. Continuous fetal heart rate monitoring is required in cases where there is continued uterine bleeding, contractile activity, or intrauterine growth restriction. Tocolytic agents, if there is no suspicion for placental abruption, may be given if membranes have not been ruptured and there is contractile uterine activity. The patient should be restricted to bed rest with bathroom privileges. She should be given stool softeners, iron supplementation, and vitamin C. Steroids to promote fetal lung maturity should be administered if the gestational age is <34 weeks.

After steroids have been administered, if there is little or no uterine activity or bleeding, the patient may be a candidate for home therapy. To be considered for home therapy, the patient must be very reliable, have 24-hour contact via telephone, and have the ability to return quickly to the hospital at any time. She should remain on bed rest with bathroom privileges and should continue to take stool softeners and vitamin therapy. Strict instructions should be given regarding returning to the hospital if she experiences contractions or another episode of bleeding.

Fetal growth, amniotic fluid index, and placental location should be assessed by ultrasound every 3 weeks. Most experts agree that for the patient with complete placenta previa that is otherwise uncomplicated, delivery is recommended at 36–37 weeks.

If uterine bleeding is not excessive, patients with marginal previa may be delivered vaginally. Upon descent, the fetal head should tamponade bleeding. Abdominal delivery is indicated in most cases of placenta previa and in all cases of complete previa. If a caesarean section is performed, care should be taken not to disrupt the placenta upon fetal delivery. If possible, a uterine incision away from the placental bed should be used. For example, if there is an anterolateral placenta, a vertical incision in the lower uterine segment opposite the site of placental implantation should be used. A high transverse incision may be necessary for a low anterior placenta. In all cases, the operating room should be prepared for the possibility of the necessity to perform a hysterectomy.

As noted earlier, 66% of patients with placenta accreta will require a caesarean-hysterectomy. There are several surgical options if uterine preservation is important. The placenta can be removed and the uterine defect can be oversewn to abate the active focus of bleeding. The area of accreta can be resected and the uterus repaired. The last option includes leaving the placenta in situ. This is only acceptable in patients who are not actively bleeding. The cord should be ligated and cut close to its base. The patient should be treated with antibiotics and possibly methotrexate postpartum. In the rare case where bladder invasion is evident, the placenta should not be removed. These patients will likely require hysterectomy and partial cystectomy. (See Chapter 21.)

Prognosis

Neonatal outcomes have much improved in the last 20 years, secondary to conservative management, the liberal use of caesarean section, improved neonatal care, and earlier diagnosis. The perinatal mortality rate has fallen from 60% to 10% over the past several decades due to the ability to resuscitate and support infants who are increasingly more premature. Most mortality is due to prematurity. An earlier of episode of bleeding brings with it a higher risk of prematurity and thus a higher risk of mortality. The maternal mortality rate has decreased from 25% to <1% in patients with access to health care. Maternal mortality remains high in developing countries.

Vasa previa is a condition in which fetal vessels traverse the membranes in the lower uterine segment, crossing over the cervical os on front of the fetal head. This can occur with a velamentous insertion of the umbilical cord, where Wharton's jelly is not present to protect the fetal vessels, or in the case where a succenturiate lobe of the placenta has a vascular communication traversing the os. Rupture of these vessels can occur with or without rupturing membranes, and the result is fetal exsanguination. Vasa previa complicates approximately 1 in 1000 to 1 in 5000 pregnancies.

Essentials of Diagnosis

• Ultrasonographic evidence of vasa previa
• Painless vaginal bleeding upon rupture of membranes
• Fetal heart rate abnormalities
• Positive Apt, Ogita, or Loendersloot tests

Clinical Findings

Symptoms & Signs

The typical presentation of vasa previa is vaginal bleeding that occurs upon rupture of membranes with concomitant specific changes in the fetal heart rate tracing. With the onset of bleeding, the fetal heart responds to loss of intravascular volume with reflex tachycardia. Reflex tachycardia is usually followed by bradycardia accompanied by occasional accelerations. With severe fetal anemia, a sinusoidal fetal heart rate pattern may be seen. The clinician must have a high index of suspicion in order to make the correct diagnosis in enough time to prevent fetal death.

Laboratory Findings

Very rarely, the fetal heart rate does not show signs of acute volume loss, and the blood can be analyzed to determine its origin. The Apt, Ogita, and Loendersloot tests take between 5 and 10 minutes to perform and are available to identify fetal hemoglobin. They have varying sensitivities depending on the amount of dilution of the blood with amniotic fluid. However, these tests are primarily of historical interest. If a patient is bleeding from a diagnosed or strongly suspected vasa previa, caesarean delivery should be expedited.

Imaging Studies

With improvement in ultrasound technology and an increased awareness among providers about this disorder, many cases are now being diagnosed in the antepartum period via ultrasound. The ultrasound finding of either velamentous cord insertion or succenturiate lobe of the placenta should prompt providers to evaluate for vasa previa. The use of color Doppler during transvaginal ultrasound can clearly identify fetal vessels coursing through the fetal membranes and can establish the diagnosis. If the vessel is a fetal artery, pulsed Doppler can help confirm and measurement of the pulse rate can help to ascertain the diagnosis. Vasa previa can be distinguished from funic presentation (free loops of umbilical cord resting adjacent to the cervix) by mapping the fetal vessels from the placenta or cord insertion. Additionally, changes in maternal position may allow for free cord as in funic presentation to move, whereas vessels in a vasa previa will remain in place.

Treatment

If a patient presents with the previously mentioned classic findings, the appropriate course of action is to deliver immediately via caesarean section to prevent impending fetal demise. If the patient has been identified antenatally, it is recommended that the patient be monitored very closely for any evidence of vaginal bleeding and to administer corticosteroids to promote fetal lung maturity. Several experts recommend inpatient observation starting at 32 weeks. Inpatient observation allows for emergent caesarean in event of nonreassuring fetal testing, preterm labor, or preterm premature ruptured membranes. Empiric delivery at approximately 35 weeks of gestation without confirmation of lung maturity via amniocentesis is also recommended by several experts. Oyelese et al report that the mean gestational age at delivery in patients who are diagnosed antenatally is approximately 35 weeks, and almost 30% of these patients still require emergent delivery after membranes are ruptured. Because the rates of morbidity and mortality are so high after ruptured membranes in these patients, they recommend delivery at 35 weeks in order to avoid a catastrophic event and feel that it is justified even in the light of possible complications due to prematurity.

Prognosis

Fetal mortality has been reported as >50% in pregnancies complicated by vasa previa. Antenatal diagnosis increases the chance of fetal survival significantly. In one study, the rates of survival for fetuses and neonates with and without antenatal diagnosis were 97% and 44%, respectively.