What the Anesthesiologist Should Know before the Operative Procedure

Obstetric hemorrhage accounts for 25% to 30% of all maternal deaths worldwide and causes death in underdeveloped countries at a rate 100 times that in developed countries. Most deaths are due to postpartum hemorrhage. In developed countries, postpartum hemorrhage is the leading cause of major maternal morbidity and is the leading reason for obstetric related intensive care unit admission. Rates of postpartum hemorrhage are rising due to an increasing rate of abnormal placental implantation associated with an increase in cesarean delivery (CD) rates and due to an increase in cases of uterine atony. The reasons for the increase in post-delivery uterine atony following both vaginal and CD are unclear. Recent reviews of closed anesthesia malpractice claims associated with patient hemorrhage show that 30% of claims involve obstetric patients.

Antepartum hemorrhage (APH) affects 5% to 20% of pregnancies, defined as genital tract bleeding after 20 or 24 weeks. Antepartum hemorrhage often presents greater threat to the life of the fetus than to the mother. Obstetric and anesthetic management depends on assessing the status of both the mother and fetus, assessing the cause of hemorrhage, and planning for delivery, if necessary. Approximately one-third of the cases of APH are due to placenta previa, one-fourth to placental abruption, and rarely to vasa previa. The remaining 40% are due to other obstetric causes such as placental marginal sinus bleeding (60%), heavy show (20%), genital trauma (5%), and vaginal varicosities, tumors, infection, and cervicitis, rarely. These later causes rarely affect pregnancy outcome. Uterine rupture causes catastrophic bleeding and fetal compromise and occurs very rarely. In cases where placenta previa is associated with prior CD, especially if placenta accreta is suspected, anesthesiologists should be involved in pre-delivery planning and prepare for significant intra and postoperative hemorrhage.

Postpartum Hemorrhage (PPH) can be life threatening. Estimating the true incidence is difficult as the currently accepted definitions of greater than 500 mL for vaginal delivery and greater than 1000 mL for CD are near average values for blood loss, and caregivers frequently underestimate the amount of obstetric bleeding. An abnormal rate and duration of bleeding should be used to diagnose PPH and its significance determined by observing maternal hemodynamic changes as well as the amount of blood loss. Uterine atony is the most common etiology of postpartum hemorrhage following vaginal or cesarean delivery. Other causes of PPH include: genital trauma, retained products of conception, abnormal placentation, and very rarely, uterine inversion. Anesthesiologists should be prepared to anesthetize women to remove retained placental products, repair genital track trauma, uterine exploration, and possible emergency hysterectomy. Monitoring and preparation for treatment massive hemorrhage and dilutional coagulopathy are necessary as it occurs commonly in cases of large blood loss.

Continue Reading

1. What is the urgency of the surgery?

What is the risk of delay in order to obtain additional preoperative information?

Urgency for the surgical treatment of APH is dictated by the threat to the life of mother and fetus and the cause of hemorrhage, with an attempt to decide the timing of delivery in non-emergent cases. The urgency of surgical treatment for PPH depends on the obstetric cause, amount of prior and ongoing blood loss, and an evaluation of the efficacy of pharmacologic intervention.

Emergent: If bleeding is rapid and the fetus is in distress, CD is immediately indicated for patients who have APH, regardless of the etiology. In cases of placental abruption, bleeding may be concealed, but immediate delivery is indicated for fetal reasons (usually fetal bradycardia). In patients with PPH, emergent surgical or radiological intervention is indicated if bleeding is rapid and the patient has not responded to prior therapy.

Urgent: If bleeding is not severe, delivery may be postponed in order to mobilize resources. However, surgery for the treatment of PPH is usually emergent and cannot wait for several hours.

Elective: Surgery for treatment of APH can be elective for cases of placenta previa or vasa previa with minimal pre-delivery bleeding and no evidence of fetal stress. Surgery for treatment of PPH is not elective.

2. Preoperative evaluation


A diagnosis of the obstetrical cause for APH is necessary. Transvaginal ultrasound examination of the uterus is safe and almost 100% diagnostic for the diagnosis of placenta previa. If the patient has had a previous CD and a placenta previa is identified, the risk of placenta accreta increases to over 40% with two prior abdominal deliveries. Other risk factors for placenta accreta include prior uterine surgery, such as myomectomy, submucosal uterine fibroids, and age greater than 35 years.

Hemorrhage due to placenta accreta accounts for 50% of all cesarean hysterectomies, the majority of which are not planned for. In the patient with previous CD, a screening ultrasound examination should be performed to determine the presence of an accreta; however, a negative ultrasound examination should not be relied upon to totally exclude accreta, and appropriate plans to treat large hemorrhage should be made. The use of color Doppler techniques may increase the positive predictive value over routine ultrasonography. Magnetic resonance imaging improves the sensitivity for diagnosis when the placenta lies posteriorly, but is most probably no more sensitive in making the diagnosis than an ultrasound examination when the placenta is located elsewhere.

Vasa previa can also be diagnosed with reasonable accuracy with ultrasound examination.

Uterine rupture is diagnosed by a sudden deterioration in the fetal status (usually fetal bradycardia) and is not necessarily accompanied by an abrupt increase in maternal hemorrhage or pain. Uterine rupture without a prior uterine scar is unusual and most often related to maternal trauma.

Risk factors for placental abruption include maternal trauma, hypertensive disorders of pregnancy (chronic hypertension with superimposed pregnancy induced hypertension), smoking, premature rupture of membranes, cocaine abuse, and multiparity. Placental abruption occurs in 1% of pregnancies, but has an associated fetal mortality of up to 20%. Ultrasound is not sensitive or specific for the diagnosis of placental abruption.


PPH is most often due to uterine atony, diagnosed by physical examination. The diagnosis of retained placental products can be made by ultrasound examination but often requires an examination under anesthesia and curettage. Placenta accreta may be diagnosed prior to delivery, but the diagnosis is usually made following delivery when the placenta fails to separate from the uterine wall. Ultrasound examination can confirm the diagnosis. The diagnosis of genital trauma is made by examination of the cervix and vagina in a patient with a firmly contracted uterus. Postpartum uterine inversion occurs rarely and is diagnosed by presence of a vaginal mass on physical examination.

Pre-existing coagulation disorders (congenital disorders such as von Willebrand’s disease or hemophilia, acquired disorders due to preeclampsia, or anticoagulation) are rare causes for PPH. Although classic cases of amniotic fluid embolism that involve severe hemodynamic compromise and pulmonary failure are accompanied in most cases by disseminated intravascular coagulation (DIC), recent reviews of this topic suggest that a milder syndrome without significant cardiovascular compromise and pulmonary failure can cause DIC.

The patient who refuses blood products should be identified and other providers should be made aware of the patient in the antepartum period or upon admission. The hemoglobin level should be optimized in the outpatient setting and a written plan for management of hemorrhage created, to include a list of acceptable blood products and consent for cell salvage, if allowed.

Medically unstable conditions warranting further evaluation prior to emergent delivery in cases of APH are extremely rare, since such hemorrhage is life threatening, primarily to the fetus. Urgent cases can allow the evaluation of prior maternal medical conditions. Aggressive treatment of PPH should never be delayed as many cases are life threatening.

Delaying surgery may be indicated in urgent cases of APH if maternal bleeding is minimal, the mother is stable hemodynamically, and the fetus shows no signs of distress. Delay to allow the administration of steroids to enhance fetal lung maturity is appropriate. Surgical therapy for PPH most often cannot be delayed.

3. What are the implications of co-existing disease on perioperative care?

a. Cardiovascular system

Perioperative evaluation

Assessment of the patient’s intravascular volume status is essential. Pregnant patients can tolerate a blood loss of up to 15% of their blood volume (approximately 1 L) without hemodynamic change, and often 1500 mL (20% to 25% of blood volume) before the signs of mild shock appear (tachycardia [heart rate up to 120 bpm], mild hypotension [systolic blood pressures of greater than 80 mm Hg], and patient restlessness). Blood losses of 30% to 35% of blood volume (2 to 2.5 L) create altered consciousness, severe tachycardia (heart rate greater than 130 bpm), and severe hypotension (systolic blood pressure less than 60 mm Hg). Circulatory collapse, after the limit for hemodynamic compensation has been reached, can be rapid and profound.

Perioperative risk reduction strategies

The potential for massive blood loss and the need for transfusion should be anticipated in all obstetric patients. The American College of Obstetricians and Gynecologists (ACOG) recommends that all facilities that care for laboring women be able to manage severe obstetric hemorrhage. A multidisciplinary team of providers, including obstetricians, anesthesiologists, blood bank, nursing, and interventional radiology personnel should be formed, particularly in institutions which care for high-risk pregnancies. Protocols by which the team can be activated in cases of severe maternal hemorrhage should be developed. Multiple large-bore intravenous cannulas should be placed. A dilutional or shock-induced coagulopathy should be anticipated and tests for coagulopathy instituted, although laboratory tests for same are often unsatisfactory for guiding immediate therapy. The degree of blood loss is often not well estimated by providers, which can lead to delayed volume replacement or delays in intervention, the most frequent causes of significant maternal morbidity. Poor communication between obstetric team members has been emphasized as a significant contributor to maternal morbidity in 60% of closed anesthesia obstetric claims. Recently, work on team-based approaches has emphasized the need for regular simulation of obstetric hemorrhage scenarios in which standardized treatment algorithms are practiced. The protocol for the treatment of hemorrhage should outline the interventions by providers and a time frame in which they occur. This protocol should be immediately available for use on all labor and delivery units (Figure 1). In addition, pre-transfusion blood testing (i.e. sending a specimen to the blood bank, type and screen, type and crossmatch) should be performed based on assessment of the patient’s risk for hemorrhage. Application of a standardized approach in large health care systems has been shown to reduce maternal morbidity and transfusion, emergent hysterectomy, and resource use.

Figure 1

Suggested management of obstetric hemorrhage.


In the patient with placenta previa, tocolysis may quiet uterine contractions and decrease bleeding. Tocolysis may also allow a delay in delivery to allow better fetal lung maturity. In the patient with suspected placenta accreta, a pre-delivery, multidisciplinary plan should be prepared. Both the ACOG and the Royal College of Obstetrics and Gynaecology strongly recommend that such patients be cared for only in institutions where adequate surgical and anesthesia personnel and blood banking facilities are available to conduct emergency cesarean hysterectomy and rapid transfusion. The patient should be counseled about the likelihood for transfusion and the loss of fertility if a hysterectomy is needed. Cell saver technology use is appropriate and delivery should occur when all elements of the surgical and anesthesia team are assembled. The surgical team should have experience in performing an immediate cesarean hysterectomy. Although balloon occlusion of pelvic vessels may reduce bleeding, this treatment may not be effective. After delivery, the balloon catheters should be retained for a period of time as they can be reinflated if bleeding occurs.

In the patient with placental abruption, the amount of vaginal bleeding may not reflect the amount of maternal blood lost. Retro-placental bleeding is often concealed and small amounts of vaginal bleeding may lead a practitioner to grossly misestimate blood loss and the true maternal volume status.


Although there are many risk factors for PPH, they lack clinical utility in predicting hemorrhage in a given patient with the exception of the pre-delivery diagnosis of placenta previa/accreta. Precipitous labor is associated with uterine atony and genital/vaginal lacerations. Instrumental delivery can cause lacerations as well. General anesthesia with a volatile agent can interfere with uterine tone. Both uterine atony and genital trauma are associated with uterine fibroids, fetal macrosomia, and multiple gestation. Chorioamnionitis, multiparity, an augmented labor with oxytocin, and prior use of tocolytics such as magnesium or nifedipine also predispose to uterine atony. Placenta accreta is associated with uterine fibroids, multiparty, prior cesarean delivery, and prior uterine scar from hysterotomy or curettage. Both uterine atony and coagulopathies following delivery are associated with amniotic fluid embolism or following delivery of a fetal demise.

Active obstetric management of the third stage of labor will reduce the risk of PPH. Early detection of abnormal bleeding will reduce maternal morbidity, although its detection may be difficult, especially intra-abdominal bleeding after cesarean section. A predetermined definition of PPH should trigger a predetermined treatment algorithm and should prompt activation of a multidisciplinary team to avoid treatment delay (Figure 1). In the first stage, the obstetric team should look for the three most common causes of PPH listed above and the anesthesiologist should begin resuscitation and provide analgesia for obstetric procedures. The second stage should occur after the first-interventions are deemed ineffective (30 minutes) or if blood loss exceeds 2 liters. Second line uterotonic agents, such as prostaglandins (15-methyl-prostaglandin F2α IM or misoprostol rectally) or methylergonovine should be administered if indicated. Uterine tamponade (Bakri balloon®) should be considered because subsequent surgical interventions can be delayed or avoided. More advanced monitoring (arterial line and possibly central venous monitoring) may be required. In stage 3, radiological intervention with arterial embolization or laparotomy may be considered. A hemodynamically unstable patient should not be routinely sent to the radiologic suite, and then only if personnel and resources to continue resuscitation are available. Laparotomy with uterine compression suturing (B-Lynch sutures) or ligation of blood supply to the uterus should be considered. Hysterectomy should be considered early as it may be the only therapy that can control bleeding and prevent significant morbidity. Uterine atony and placenta accreta refractory to therapy usually dictate hysterectomy.

Baseline coronary artery disease or cardiac dysfunction:.

Pregnant women rarely have significant prior cardiac disease.

Patients with chronic hypertension or hypertension in association with pregnancy should have their antihypertensive medications maintained. Magnesium therapy in the patient with preeclampsia should be maintained despite the side effect of interfering with the action of uterotonics.

b. Pulmonary

Pregnant women rarely have prior chronic obstructive pulmonary disease.

Reactive airway disease: In the patient with asthma with hemorrhage, the usual history to determine severity (frequency of attacks, triggers, recent medication use, emergency department visits, and history of prior intensive care unit admissions) will usually suffice to determine therapy. Continuation of usual inhaler medications, administration of increased doses of steroids in cases of exacerbation, and use of regional techniques will reduce intraoperative risk for exacerbations. 15-Methyl-PGF administration is contraindicated in patients with asthma as life-threatening bronchospasm has been reported following its administration. Work in asthmatic volunteers shows that the increase in airway resistance is idiosyncratic and often profound, and is not related to the severity of underlying bronchospastic disease.

c. Renal-GI:

If time allows, the administration of pharmacologic prophylaxis to include a non-particulate antacid and an H2 receptor blocker or protein pump inhibitor (to reduce stomach content acidity) and metoclopramide (to promote stomach emptying) should be administered to all pregnant women undergoing surgical procedures to reduce the risk of aspiration pneumonitis.

d. Neurologic:

Not applicable

e. Endocrine:

Pregnant women with gestational or chronic diabetes should have their pharmacologic therapy continued and blood sugars checked frequently as blood sugars can vary greatly.

which may be of concern in a patient undergoing this procedure and are relevant for the anesthetic plan (e.g. musculoskeletal in orthopedic procedures, hematologic in a cancer patient)

If the patient is a Jehovah’s Witness or has other objections to transfusion therapy, discussions of the patient’s wishes to include the types of blood products and which intraoperative cell salvage techniques or normovolemic hemodilution techniques are acceptable should occur. In cases of placenta previa, iron and erythropoietin therapy should be offered during pregnancy to maximize maternal hemoglobin.

4. What are the patient's medications and how should they be managed in the perioperative period?

Except for women who have chronic hypertension, pregnant women rarely receive medications for the treatment of chronic disease.

In general, the patient’s usual chronic medications should be continued. In the patient who has APH due to placenta previa, magnesium sulfate may be used as a tocolytic to allow the administration of betamethasone to mature the fetal lung and to allow the obstetrician to better time delivery. In patients with preeclampsia, magnesium therapy should be continued to prevent seizures. Magnesium, however, interferes with the action of uterotonics and may aggravate hemorrhage.

g. Are there medications commonly seen in patients undergoing this procedure and for which should there be greater concern?

In women who have APH, magnesium sulfate may be administered for tocolysis to allow further administration of glucocorticoids for fetal lung maturity, since many women who have APH deliver preterm. It may cause less hypotension than beta agonists in the face of maternal hemorrhage and may improve neonatal neurologic outcome. However, work in pregnant ewes suggests that magnesium therapy may worsen maternal hypotension during hemorrhage in the presence of neuraxial blockade. Magnesium will prolong the effects of nondepolarizing neuromuscular blockers and repeated succinylcholine administration. Magnesium interferes with the action of agents used to treat uterine atony.

Oxytocin (Pitocin®) is the first-line treatment for both the prophylaxis and treatment of uterine atony; 20 to 60 U/L is administered at a rate up to 1 L/hr. Bolus dosing is controversial, with 10 U administration associated with a significant incidence of hypotension and electrocardiographic evidence of myocardial ischemia. Recent work suggests that the ED90 of oxytocin when administered by bolus dose following elective CD is 0.35 U. However, the ED90 is approximately 3-4 times that for women who have labored prior to CD. Hypotension associated with its use may be accentuated in the patient with hemorrhage. Bolus doses of or the use of infusions that administer less than 1 U/min will significantly reduce the incidence. The administration of oxytocin for labor augmentation for prolonged periods will increase the risk of postpartum hemorrhage. One recent case controlled study showed that the odds ratio of hemorrhage increased by 1.58 for each 4-hour period that oxytocin was infused at the maximum rate used in most obstetric units.

Ergot alkaloids such as ergonovine and methylergonovine (Methergine®) can be administered in doses of 0.2 mg IM and are typically the second-line drugs used for atony. They have been reported to cause vasoconstriction and severe hypertension in rare patients, and thus their use is cautioned in patients with pre-existing hypertension, preeclampsia, or pulmonary hypertension and immediately following use of a vasopressor. Bolus IV administration is not recommended, although infusion of a dilute concentration with careful attention to maternal blood pressure may be indicated in cases of severe hemorrhage. Maternal nausea and vomiting are common.

Prostaglandins, primarily 15-methyl-PGF2α (Hemabate®) can be administered IM or intramyometrially. In one study, 85% of women with an inadequate response to an oxytocin infusion responded to Hemabate® administration. Doses of 0.25 mg can be given every 15 minutes up to a total dose of 2 mg. Maternal side effects can be hypertension and diarrhea. The drug should be given cautiously in patients with reactive airway disease as life-threatening bronchospasm has been reported. Acute increases in pulmonary vascular resistance leading to acute pulmonary hypertension have also been reported.

Misoprostol (Cytotec®) may be effective when other agents have failed to control hemorrhage. Small case series report that rectal administration of 1000 ug improves uterine response when prior oxytocin and methylergonovine have failed. One randomized controlled trial showed it to be less effective than the combination of oxytocin and ergometrine for the control of PPH. Like other prostaglandins, maternal fever, chills, nausea and vomiting may accompany its use. It may be used when 15 -methyl prostaglandin F2 administration is contraindicated.

Prostaglandin E2 (dinoprostone) is a third prostaglandin agent that can be used in face of PPH, but it is only available as a 20 mg vaginal suppository in the United States, which makes it less useful during vaginal bleeding. It causes systemic vasodilation with increases in cardiac output and bronchodilation.

Although methotrexate has been used to speed placental involution in the patient with placenta accreta when the obstetrician elects to retain the placenta at CD, its efficacy has been recently questioned. Weekly ultrasound and plasma beta human chorionic gonadotropin monitoring will be required to determine the efficacy of therapy. The dose is unclear.

Tranexamic acid, an antifibrinolytic, is inexpensive and may reduce the amount of blood loss by inhibiting fibrinolysis in the face of major blood loss. In case reports and small case series, its prophylactic use in patients at high risk for PPH following CD has been associated with reduced blood loss 24 hours postpartum in patients who developed major hemorrhage, without side effects. However, only one small randomized controlled trial supports its use in the treatment of PPH and one non-randomized trial using historical controls failed to show a significant difference in reducing blood loss in patients with PPH. The recommended dose is 1 g with re-dosing at 4 hours. An ongoing clinical trial may offer further guidelines for its use.

Recombinant factor VIIα is extremely expensive and timing of administration, and optimal dose in parturients, has not been determined. Small case series and case reports support its effectiveness in many hemorrhaging obstetric patients, but no randomized controlled trials have evaluated its use. Therefore, its use in the treatment of hemorrhage should be considered only after failure of other conventional therapies. It may be most useful in transiently reducing blood loss in an unstable patient to allow definitive surgical or radiological therapy. Despite its use, conventional transfusion therapy should not be stopped. Doses up to 50 mcg/kg are suggested with little increased efficacy above 90 mcg/kg. Thrombosis is a feared complication, reported in small case series.

h. What should be recommended with regard to continuation of medications taken chronically?

Cardiac, pulmonary, renal, neurologic, and psychiatric medications should be continued. Antiplatelet agents should be discontinued in the bleeding patient.

See other sections as the management is not affected by the presence of maternal hemorrhage.

– If the patient has a sensitivity to latex (e.g. rash from gloves, underwear, etc.) versus anaphylactic reaction, prepare the operating room with latex-free products.

See other sections as the management is not affected by the presence of maternal hemorrhage.

(common antibiotic allergies and alternative antibiotics)

See other sections as the management is not affected by the presence of material hemorrhage.

(common antibiotic allergies and alternative antibiotics)

Regarding malignant hyperthermia and local anesthetics/muscle relaxants, see other chapters as the management is not affected by the presence of maternal hemorrhage.

Hemoglobin levels should be obtained upon presentation and then followed serially during hemorrhage. In cases of a several blood unit transfusion, serum calcium levels should be monitored. The routine laboratory coagulation tests of partial thromboplastin time, prothrombin time, and platelet count are often abnormal as a dilutional coagulopathy develops frequently in cases of large blood loss and fluid replacement. Serial fibrinogen levels should be followed during acute hemorrhage. Since up to 20% of women with placental abruption will have evidence of a consumptive coagulopathy, coagulation studies and a fibrinogen level are indicated.


In cases of APH, transvaginal ultrasound examination of the uterus is safe and 87.5% sensitive and 99% specific for the diagnosis of placenta previa. If the patient has had a previous CD and a placenta previa is identified, the risk of placenta accreta increases from 3% in the patient with no previous CD, to 11% with one prior section, and over 40% with two or more. In the patient with previous CD, presence of a placenta accreta should be screened by more discriminative ultrasound examination (sensitivity of 77% and specificity of 96% for placenta accreta) or magnetic resonance imaging (sensitivity of 88% and specificity of 100%). Vasa previa can also be diagnosed with reasonable accuracy with ultrasound examination.

Blood products

The American Society of Anesthesiologists Practice Guidelines for Obstetric Anesthesia and the ACOG do not recommend a routine type and screen or cross match prior to vaginal or operative delivery. The decision to obtain either should be based on the likelihood of maternal hemorrhagic complications. This may be especially important in women who have known antibodies which could make transfusion with homologous blood difficult.

Autologous prepartum donation can also be considered, as it has minimal effect on maternal hemodynamics; however, in cases of severe maternal hemorrhage, it has not been shown to reduce the quantity of homologous blood that is transfused.

Intraoperative cell salvage has been used safely in many patients and should be considered in cases where severe blood loss is anticipated and in patients who might be difficult to cross-match or object to homologous transfusion for religious and other reasons. The risk for amniotic fluid embolism is minimal and recent work suggests that the concentration of fetal squamous cells in washed red cells is similar to that found in parturients prior to delivery. Women who receive salvaged blood may become alloimunized to fetal hemoglobin, so Rh negative women should receive anti-D immune globulin.

Arterial blood gas monitoring for oxygenation and acidosis and serum lactic acid can help assess the adequacy of resuscitation during massive transfusion.

During periods of massive hemorrhage, routine laboratory testing of coagulation measures (partial thromboplastin time, prothrombin time, international normalized ratio, platelet count, fibrinogen level), measures of adequate oxygen carrying capacity (hemoglobin), and adequacy of fluid resuscitation (guided by rising or falling lactic acid level) may not be timely enough to help guide blood product therapy during a rapidly changing clinical picture. Indeed, they were never designed for this purpose. Point of care testing with rapid hemoglobin testing (Hemocue®) and rapid tests of coagulation based on thromboelastography (Rotem®) have been introduced into clinical practice and may offer rapid assessment of coagulation status during massive transfusion. Calcium levels may become low due to citrate binding of calcium during massive transfusion; serum calcium levels should be obtained.

Studies of women with PPH have shown that mean fibrinogen levels fall rapidly. Since fresh frozen plasma is low in fibrinogen, cryoprecipitate transfusion will often be necessary to restore fibrinogen levels to > 2 g/L.

Platelet transfusion to achieve a platelet count > 50 × 109 / L may be required.

Many institutions have extended their use of massive transfusion protocols to the treatment of obstetric hemorrhage. Such protocols deliver combinations of packed red blood cells, fresh frozen plasma, and platelets in fixed ratios without use of laboratory values to guide their administration. Outcome studies in surgical trauma patients and smaller studies in hemorrhaging obstetric patients have shown reduced overall morbidity.

Intraoperative Management: What are the options for anesthetic management and how to determine the best technique?


The decision to provide neuraxial anesthesia or general anesthesia for delivery of the patient with APH depends on assessment of the maternal intravascular volume status, prior administration of obstetric medications, assessment of the fetal environment, and a careful airway examination. If there is minimal or no active bleeding and no evidence of coagulopathy, neuraxial anesthesia is preferred in the patient with placenta previa or placental abruption. Intravascular volume depletion or a nonreassuring fetal heart rate pattern may accompany APH and necessitate a general anesthetic.


Invasive procedures to control PPH can be done under local anesthesia with sedation or inhalational analgesia in cases of genital trauma and retained placenta. If neuraxial blockade has been performed for labor analgesia, the block may be extended. Procedures to control uterine atony can be performed under a newly placed regional block if the mother is hemodynamically stable. General anesthesia will most often be required in cases where hysterectomy is performed to control hemorrhage due to placenta accreta or uterine atony refractory to pharmacologic therapy. General anesthesia may also be used to provide uterine relaxation needed to successfully replace an inverted uterus.

a. Regional anesthesia

Neuraxial anesthesia

Benefits: In cases of CD of the patient with placenta previa and placental abruption, it allows the mother to be awake during delivery. The risks of aspiration and difficult endotracheal intubation, which occurs at rates much higher than in general surgery patients, are most often avoided. During cases of severe bleeding, blood loss and transfusion requirements have been shown to be less with neuraxial compared to general anesthesia. In the patient undergoing a vaginal delivery, neuraxial analgesia can be rapidly it can be rapidly extended to anesthesia in the hemodynamically stable patient.

Drawbacks: Epidural anesthesia may adversely affect the compensatory hemodynamic responses to untreated hemorrhage if hemorrhage becomes severe. It is not an option in the patient who has had sufficient blood loss to create hemodynamic instability. Induction of general anesthesia and airway control may be required after the procedure is under way. Since difficult/impossible endotracheal intubation is encountered more frequently in the pregnant patient than in nonpregnant women, difficult airway management may be more likely to further complicate an emergency situation. Pulmonary aspiration may occur if the patient becomes hypotensive in the face of significant hemorrhage with loss of airway reflexes.

Issues: The decision to convert from regional anesthesia to general anesthesia is critical and should be done early when significant hemorrhage is identified. This will avoid airway management during a time of potential task stress due to the need for maternal volume resuscitation, decrease the hemodynamic risk associated with anesthetic induction in a hypovolemic patient, and avoid increased difficulty in airway management due to airway edema associated with fluid resuscitation.

Peripheral nerve block

Local infiltration is an anesthetic option for genital trauma repair and for arterial embolization. For all other procedures, peripheral nerve block is not an option.

b. General anesthesia

Benefits: General anesthesia is indicated in the patient with APH and fetal compromise. It is required in the face of rapid and/or significant hemorrhage with maternal hemodynamic compromise.

Drawbacks: The mother and support partner do not participate in the birth of the child and blood loss is greater than in patients who undergo neuraxial block. In cases of peripartum hysterectomy, its use is associated with greater blood loss than if a neuraxial block is used.

Airway concerns: Since difficult or impossible endotracheal intubation occurs more frequently than in nonpregnant women (about 1:250 parturients), plans for managing the difficult airway should be made.

c. Monitored anesthesia care

Cases of mild/moderate PPH (retained placental products, genital/vaginal lacerations) may be managed with local infiltration and moderate sedation. Arterial embolization procedures can be done under local anesthesia and sedation as well. All other procedures require neuraxial or general anesthesia.

6. What is the author's preferred method of anesthesia technique and why?

Based on current recommendation from the ACOG, preoperative cefazolin 2 to 3 g IV based on body weight within 60 minutes of skin incision is associated with the greatest reduction in postoperative wound infection and post-delivery endometritis. The addition of a broad-spectrum antibiotic such as azithromycin 250 mg administered prior to and during the surgical procedure may further decrease the risk. In cases where prior administration of gentamicin and ampicillin has been instituted for the treatment of chorioamnionitis, the administration of cefazolin prior to skin incision is also recommended. Administration of gentamycin plus clindamycin is indicated as a replacement for cephalosporin administration in persons with documented allergy to cephalosporins or penicillin allergy. Redosing of antibiotics will be required if significant blood loss has occurred or during procedures lasting longer than 4 hours.


Placenta previa: Patients with placenta previa are always delivered via cesarean. Choice of anesthetic technique depends on maternal volume status, urgency of delivery due to ongoing bleeding, and fetal status, with consideration for the possibility of placenta accreta in the patient who has had a prior CD. Case series support that neuraxial anesthesia can be safely undertaken even in the face of potential large blood loss, although the support person should be informed that he/she may need to leave from the operating room if general anesthesia is required.

General anesthesia will be necessary in cases of rapid ongoing blood loss and obtaining an adequate history, and performing resuscitation in such circumstances can be quite daunting. At least two large-bore intravenous catheters should be placed; arterial line monitoring should be considered routine. Blood products should be available in the operating room at the time of delivery, as the blood loss can be very large. The need for fresh frozen plasma, cryoprecipitate, and platelets to treat a dilutional coagulopathy should be anticipated. Induction of anesthesia with 0.5 to 1.0 mg/kg of ketamine or etomidate 0.3 mg/kg should be undertaken in the hypovolemic parturient.

If a placenta accreta has been diagnosed preoperatively, then retention of the placenta in place with closure and expectant management may offer the best alternative for treatment as blood loss can become massive if placental removal is attempted. Even if the placenta separates, the insertion site over the lower uterine segment will often not contract and becomes a site for significant hemorrhage. Peripartum hysterectomy (discussed below) will most often be needed or, if the patient strongly wishes to retain fertility, transfer to interventional radiology suite for embolization of arterial vessels may be undertaken (discussed below under postpartum hemorrhage).

Vasa previa: This is a true obstetric emergency if hemorrhage is detected as fetal blood loss is dramatic. CD is always performed, often under general anesthesia, to save the baby’s life.

Placental abruption: Placental abruption is suspected when fetal compromise is present. An increase in maternal abdominal pain is often not reported, even in absence of neuraxial analgesia. Vaginal bleeding does not correlate well with the amount of abruption. The degree of maternal and fetal distress, maternal coagulation status, and severity of the abruption will dictate whether neuraxial anesthesia can be safely performed if a block has not been previously placed. A coagulopathy can be present in 10% of patients, so general anesthesia may be required in the face of urgent delivery.

Laboratory studies should include hemoglobin, type and screen, partial thromboplastin time, prothrombin time, platelet count, and fibrinogen and fibrin degradation products. These are routinely drawn since a consumptive coagulopathy can occur within the placental clot. Vaginal delivery is possible if blood loss is moderate or has ceased, and if the fetal status is reassuring. Neuraxial anesthesia can be safely performed under such circumstances.

Uterine rupture: Uterine dehiscence in the lower uterine scar in the patient who has had a previous CD should be differentiated from true uterine rupture that involves the muscle wall of the uterus. The former often presents without significant bleeding or pain and with fetal heart rate abnormality as the only sign. The later presents with fetal compromise, evidence of significant maternal hemorrhage, and emergency CD and post-delivery laparotomy uterine repair is required. Most cases of uterine rupture occur in association with a trial of labor following a previous CD; however, neuraxial analgesia does not interfere with the diagnosis of uterine dehiscence/rupture as an increase in maternal pain is often not present. Hysterectomy may be needed if uterine repair or arterial ligation of the blood supply to the uterus is ineffective (see section on postpartum hemorrhage below).


For significant hemorrhage, a multidisciplinary response is necessary. Anesthesiologists should direct fluid resuscitation and provide analgesia. Communication with the obstetric team of maternal hemodynamics and laboratory data is of extreme importance to decide on the next steps for treatment (see Figure 1).

Uterine atony: Oxytocin is routinely administered to prevent atony, but if not effective, other uterotonics are administered quickly. Oxygen is administered, adequate venous access is ensured, additional neuraxial or IV analgesia is given, laboratory studies are drawn, and the blood bank is notified. If hemorrhage is not quickly controlled, further treatment is performed in the obstetric operating room as placement of uterine tamponade devices and more invasive therapy may be required.

Genital trauma/retained placental products: For significant lacerations or for manual uterine exploration for removal of retained products, an anesthesiologist is present to administer analgesia and prepare for resuscitation. Local infiltration with IV opioid supplementation or ketamine in doses of 0.25 to 0.5 mg/kg in the patient without an epidural block is effective in many cases, and if exploration is done in the operating room, inhaled 50% nitrous oxide in oxygen analgesia is helpful. If the patient is hemodynamically stable, re-dosing an existing neuraxial block is appropriate.

Placenta accreta: Previously unrecognized placenta accreta is often identified at CD when the placenta fails to separate from the uterine wall. Unfortunately, this often results in massive hemorrhage. Additional obstetric and anesthesia help should be summoned, additional IV access and arterial line monitoring placed, and induction of general anesthesia initiated early on during resuscitation efforts. In the patient in whom placenta previa and possible accreta were identified preoperatively and in whom internal iliac arterial balloons were placed, preparations for massive hemorrhage should not be postponed as blood loss is often large despite their use. Recent work describing obstetric management by leaving the placenta in situ shows a reduction in acute operative blood loss and acute maternal morbidity; however, longer term morbidity may be increased, and further work is needed to balance reductions in acute morbidity vs. an increase in longer term consequences. Methotrexate, administered postoperatively, was once thought to improve placental involution and reduce risk, but recent work suggests that this drug is ineffective. Monitoring for complications of infection and secondary hemorrhage should occur. Later hysterectomy is often required for most patients (see below).

Invasive procedures:

Uterine compression sutures, uterine/internal iliac arterial ligation: Uterine compression sutures (B-Lynch) are most effective for uterine atony, but have a 30% failure rate. Arterial ligation of the uterine, ovarian, and internal iliac arteries is often difficult and time consuming as collateral circulation and the enlarged uterus interfere with a stepwise uterine devascularization. It may be 85% effective in skilled hands. Both of these procedures can be performed by extending a pre-existing epidural block, but preparations for immediate general anesthesia should occur.

Arterial embolization: This can be performed before laparotomy, during a CD procedure, or after failure of other invasive surgical procedures. Although only local anesthesia is needed, anesthesia personnel should expect to accompany the patient if the need for resuscitation is anticipated. Although small case series report success rates of 85% to 95% for arterial embolization, no controlled trials exist and a recent systematic review showed that some reports show little reduction in the rate of hysterectomy.

Peripartum hysterectomy: Preparation should be similar to placenta previa and accreta above. If performed for uterine atony, general anesthesia should be first choice, even if a neuraxial block is in place as conversion from neuraxial anesthesia will be most often be necessary after accreta identification at CD. Transfusion requirements in excess of 10 units of packed red blood cells should be anticipated.

Uterine inversion: Treatment requires replacement of the uterus by the obstetrician, which may require uterine relaxation. Case reports suggest that intravenous nitroglycerin in 100- to 500-mcg boluses can be effective while using a vasopressor to counteract hypotension (phenylephrine 100 mcg). General anesthesia with a volatile agent may also be required both for pain control and uterine relaxation.

Intraoperative care

Transfusion therapy to achieve a hematocrit of 30%, platelet count of greater than 50,000 /mm3, and fibrinogen greater than 1 g/L are optimal goals, although this fibrinogen level may seriously underestimate the level required to stop significant hemorrhage. A level of 2 g/L is advocated by some and a low fibrinogen level is the most sensitive indicator for the risk of ongoing bleeding. The early use of cryoprecipitate for fibrinogen replacement after the first 4 units of fresh frozen plasma (FFP) is controversial as FFP has varying amounts of fibrinogen. A dose of 4 units of FFP (with the just noted qualification), two cryoprecipitate pools, or 3 g of fibrinogen concentrate will raise the plasma fibrinogen by about 1 g/L. The decision to transfuse packed red blood cells, FFP, and platelets in ratios that approach those in whole blood should not wait for laboratory confirmation of coagulopathy if blood loss has reached one-half to one blood volume and is occurring rapidly. Although the support for such an approach comes from studies in non-obstetric patients, reports show improved survival, decreased instances of pneumonia, and abdominal compartment syndrome.

Normovolemic hemodilution is possible in cases when the potential for blood loss has been determined preoperatively. A few case series describe good maternal and fetal outcomes, but its utility and cost effectiveness have not been demonstrated. Current literature does not support that it reduces the amount of autologous transfusion during obstetric hemorrhage.

Intraoperative cell salvage has been used safely; however, its use is controversial for fear that transfused fetal products in the washed cells from processed blood may cause an amniotic fluid embolism-type syndrome. The lack of well-designed prospective trials of its use and questions of its cost effectiveness have lead the ACOG to recommend its use only when sufficient banked blood is not available or when the patient refuses autologous transfusion.

Tranexamic acid administration is indicated if other measures fail. Some administer it when transfusion of blood products in fixed ratios is started. Its prophylactic use may reduce blood loss during CD in patients identified to be a high risk for hemorrhage.

Factor VIIα administration should be considered only after failure of other conventional transfusion therapy.

Intraoperative complications

Disseminated intravascular coagulation (DIC) is often impossible to differentiate from dilutional coagulopathy. In most cases differentiation is moot if treatment with transfusion of blood products in large amounts at fixed ratios of red cells to FFP plasma to platelets has been started. Scoring systems that help diagnose DIC in other clinical settings (platelet counts less than 50,000/mm3, strong increases in fibrin markers [fibrin degradation products, D-dimer], prolonged PT, and fibrinogen level less than 1 g/L) are not helpful in treating the hemorrhaging obstetric patient.

Myocardial ischemia occurs with significant frequency in the parturient with ongoing hemorrhage. If recognized by electrocardiographic monitoring in the face of rapid hemorrhage, it is an absolute indication to give type O blood without obtaining hemoglobin level. Cases of myocardial infarction have not been reported.

Acute respiratory distress syndrome occurs rarely following resuscitation. Pulmonary edema due to over resuscitation may occur. The patient who survives the initial hemodynamic insult from an amniotic fluid embolism will most often require post resuscitation mechanical ventilation.

a. Neurologic:

Arterial ligation procedures and arterial embolization procedures have been associated with rare instances of lower extremity neuropathy and ischemia.

Ensure the patient is awake and alert and that the adult respiratory distress syndrome associated with shock and coagulopathy is either not present or improving.

c. Postoperative management

To control postoperative pain, catheter techniques involving epidural infusion of dilute local anesthetics can be performed in patients in whom a neuraxial catheter was placed preoperatively, but maternal volume status should be adequate to avoid hypotension. A coagulopathy may be present, which may delay removal of previously placed catheters.

The type of post-delivery monitoring depends on the degree of blood loss, state of maternal hemodynamics, and need for postoperative ventilation. Most patients with greater than a blood volume blood loss will require intensive care unit care to monitor for possible recurrence of bleeding, even if the patient does not require postoperative ventilation.

Postoperative complications

Complications of blood transfusion to include transfusion-related lung injury and infectious complications of transfusion have not been reported as significant complications in obstetric patients. Infectious complications, re-bleeding, ongoing acidosis, and a persistent coagulopathy in the patient whose hemorrhage has been previously controlled suggest sepsis and need for surgical re-exploration. A coagulopathy should resolve relatively quickly once bleeding has been controlled. Thromboembolic events occur rarely after factor VIIα therapy in obstetric patients. The multiorgan failure syndrome occurs rarely and may be due to irreversible shock. Monitoring for site infection, bladder and ureteral injury, and intra-abdominal hemorrhage should occur if the placenta is left in situ or arterial embolization or hysterectomy has been performed.

What's the Evidence?

Flood, KM, Said, S, Geary. “Changing trends in peripartum hysterectomy over the last 4 decades”. Am J Obstet Gynecol. vol. 200. 2009. pp. 632.e1-632.e6.

Bateman, BT, Berman, MF, Riley, LE, Leffert, LR.. “The epidemiology of postpartum hemorrhage in a large, nationwide sample of deliveries”. Anesth Analg. vol. 110. 2010. pp. 1368-73.

Porreco, RP, Clark, SL, Belfort, MA. “The changing specter of uterine rupture”. Am J Obstet Gynecol. vol. 200. 2009. pp. 269.e1-269.e4.

Patterson, JA, Roberts, CL, Bowen, JR. “Blood transfusion during pregnancy, birth, and the postnatal period”. Obstet Gynecol. vol. 123. 2014. pp. 126-33.

Mhyre, JM, Shilkrut, A, Kuklina, EV. “Massive blood transfusion during hospitalization for delivery in New York State, 1998-2007”. Obstet Gynecol. vol. 122. 2013. pp. 1288-1294. (These reports note the change in etiologies and the increase in rate of peripartum hemorrhage that have occurred over the last several decades. Rates of obstetric blood transfusion have increased.)

Mercier, FJ, Van de Velde, M.. “Major obstetric hemorrhage”. Anesthesiol Clin. vol. 26. 2008. pp. 53-66.

Ahonen, J, Stefanovic, V, Lassila, R.. “Management of post-partum haemorrhage”. Acta Anesthesiol Scand. vol. 54. 2010. pp. 1164-78.

Shields, LE, Wiesner, S, Fulton, J. “Comprehensive maternal hemorrhage protocols reduce the use of blood products and improve patient safety”. Am J Obstet Gynecol. vol. 212. 2015. pp. 272-80.

Kacmar, RM, Mhyre, JM, Scavone, BM. “The use of postpartum hemorrhage protocols in United States academic obstetric anesthesia units”. Anesth Analg. vol. 119. 2014. pp. 906-10. (These articles stress the importance of a team-based approach and prior agreed-on algorithms as elements in improving care of the patient with peripartum hemorrhage. Hemorrhage protocols appear readily available in 67% of units, but the use of massive transfusion protocols is greater than 90%.)

Toledo, P, McCarthy, RJ, Hewlett, BJ. “Accuracy of blood loss estimation after vaginal delivery”. Anesth Analg. vol. 105. 2007. pp. 1736-1740.

Toledo, P, McCarthy, RJ, Burke, CA. “The effect of live and web-based education on the accuracy of blood-loss estimation in obstetric scenarios”. Am J Obstet Gynecol. vol. 202. 2010. pp. 400.e1-5(These reports verify that providers significantly overestimate blood loss when it is small and underestimate it when it is large.)

Mercier, FJ, Bonnet, MP.. “Use of clotting factors and other prohemostatic drugs for obstetric hemorrhage”. Curr Opin Anaesthesiol. vol. 23. 2010. pp. 310-6.

Ickx, BE.. “Fluid and blood transfusion management in obstetrics”. Eur J Anaesthesiol. vol. 27. 2010. pp. 1031-35.

Cotton, BA, Au, BK, Nunez, TC. “Predefined massive transfusion protocols are associated with a reduction in organ failure and postinjury complications”. J Trauma. vol. 66. 2009. pp. 41-8.

Phillips, LE, McLintock, C, Pollock, W. “Recombinant activated factor VII in obstetric hemorrhage: experiences from the Australian and New Zealand haemostasis registry”. Anesth Analg. vol. 109. 2009. pp. 1908-15. (These results support that the early use of clotting factor and platelet replacement is associated with better outcomes in patients who have sustained massive hemorrhage and summarize the current recommendations for use of factor VIIα.).

Butwick, AJ.. “Postpartum hemorrhage and low fibrinogen levels: the past, present, and future”. Int J Obstet Anesth. vol. 22. 2013. pp. 87-91.

Mallaiah, S, Barclay, P, Harrod, I. “Introduction of an algorithm for ROTEM-guided fibrinogen concentrate administration in major obstetric haemorrhage”. Anaesthesia. vol. 70. 2015. pp. 166-75.

Karlsson, O, Jeppsson, A, Hellgren, M. “Major obstetric hemorrhage: monitoring with thromboelastography, laboratory analyses or both?”. Int J Obstet Anesth. vol. 23. 2014. pp. 10-7. (These studies suggest that point of care coagulation testing and monitoring fibrinogen levels during maternal hemorrhage leads to less transfusion requirement than in patients in whom standard laboratory coagulation testing is used.)

Jump to Section