Septic shock is a life-threatening disorder secondary to bacteremia. The American College of Obstetricians and Gynecologists defines septic shock as sepsis with hypotension despite adequate fluid resuscitation, with the presence of perfusion abnormalities including (but not limited to) lactic acidosis and oliguria. The incidence of bacteremia in obstetric patients has been estimated to be between 0.7% and 10%. Although gram-negative bacteria are usually responsible for most of these infections, septic shock may also result from infection with other bacteria, fungi, protozoa, or viruses. The most common cause of obstetric septic shock is postoperative endometritis (85%). Other commonly associated conditions include antepartum pyelonephritis, septic abortion, and chorioamnionitis.
Sepsis may lead to a systemic inflammatory response that can be triggered not only by infections, but also by noninfectious disorders, such as trauma and pancreatitis. However, there is strong evidence to support the concept that endotoxin is responsible for the pathogenesis of gram-negative septic shock. Escherichia coli has been implicated in 25–50% of cases of septic hypotension, but a variety of other organisms may be causative, including Klebsiella, Enterobacter, Serratia, Proteus, Pseudomonas, Streptococcus, Peptostreptococcus, Staphylococcus, Fusobacterium, Clostridium, and Bacteroides. The gram-negative endotoxin theory does not explain gram-positive shock, although an understanding of the proposed mechanisms will serve to exemplify the multisystemic effects of this disorder.
Endotoxin is a complex lipopolysaccharide present in the cell walls of gram-negative bacteria. The active component of endotoxin, lipid A, is responsible for initiating activation of the coagulation, fibrinolysis, complement, prostaglandin, and kinin systems. Activation of the coagulation and fibrinolysis systems may lead to consumptive coagulopathy. Complement activation leads to the release by leukocytes of mediators that are responsible for damage to vascular endothelium, platelet aggregation, intensification of the coagulation cascade, and degranulation of mast cells with histamine release. Histamine will cause increased capillary permeability, decreased plasma volume, vasodilatation, and hypotension. Release of bradykinin and b-endorphins also contributes to systemic hypotension. Early stages of septic shock involve low SVR and high cardiac output with a relative decrease in intravascular volume. Late or cold shock subsequently involves an endogenous myocardial depressant factor that has not been isolated. This factor is associated with decreased cardiac output and continued low SVR in the absence of pressor agents. Recent studies suggest that tumor necrosis factor (TNF) may lead to depressed myocardial function during septic shock. Monocytes and macrophages incubated with endotoxin produce this 17-kDa polypeptide within 40 minutes. Direct injection of TNF into animals leads to many of the changes seen in endotoxic shock. Other possible factors include interleukin (IL)-1, IL-6, IL-8, interferon gamma, and granulocyte stimulating factor.
Septic shock can be divided into 3 stages: preshock, early shock (warm shock), and late (or cold) shock. In preshock, patients present with tachypnea and respiratory alkalosis (above the mild respiratory alkalosis seen in normal pregnancy). Their condition is best described as a moderate hyperdynamic state, with elevated cardiac output, decreasing SVR, and normal blood pressures. Response to therapy will be greatest at this stage. Early shock is a more hyperdynamic state. Blood pressure drops (systolic blood pressure < 60 mm Hg), and SVR decreases dramatically (< 400 dynes · s · cm−5). Altered mental status, temperature instability, and sinusoidal fluctuations in arterial blood pressure may be seen at this stage. As this condition progresses into late shock, activation of the sympathetic nervous system with release of catecholamines will lead to intense vasoconstriction, which serves to shunt blood from the peripheral tissues to the heart and brain (cold shock). The compensatory vasoconstriction results in increased cardiac work. Lactic acidosis, poor coronary perfusion, and the influence of myocardial depressant factor may also contribute to poor cardiac performance (Fig. 23–5). The fetus is more resistant to the effects of endotoxin than the mother; however, alterations in uteroplacental flow can lead to hypoxia, acidosis, placental abruption, intracranial hemorrhage, and fetal demise.
Presenting left ventricular function of 10 pregnant women with septic hypotension. LVSWI, left ventricular stroke work index; PCWP, pulmonary capillary wedge pressure.
(Reproduced, with permission, from Lee W, Clark SL, Cotton DB, et al. Septic shock during pregnancy. Am J Obstet Gynecol 1988;159:410.)
The clinical manifestations of septic shock depend on the target organs affected. The most common cause of death in patients with this condition is respiratory insufficiency secondary to ARDS.
Complete blood cell count, serum electrolytes, urinalysis, baseline arterial blood gases, chest radiograph, and a coagulation profile are laboratory studies important in the management of these patients. Hematologic findings may include significant anemia, thrombocytopenia, and leukocytosis. Serum electrolytes are often abnormal because of acidosis, fluid shifts, or decreased renal perfusion. Urinalysis permits evaluation of renal involvement. In addition to urine cultures, aerobic and anaerobic blood cultures may be helpful to confirm the diagnosis and guide antibiotic therapy.
Arterial blood gas measurements and a chest radiograph will facilitate clinical assessment of the ventilatory and oxygenation status. Early stages of septic shock will be associated with respiratory alkalosis, which later progresses to metabolic acidosis.
A baseline electrocardiogram (ECG) should be performed to rule out myocardial infarction or cardiac dysrhythmia. Abdominal radiographic studies may be useful to rule out other intrapelvic or intra-abdominal sources of obstetric sepsis (eg, bowel perforation, uterine perforation, tubo-ovarian abscess). Significant disseminated intravascular coagulation (DIC) will be identified by abnormal PT, PTT, or fibrinogen levels.
The differential diagnosis should include other hypovolemic and cardiogenic shock syndromes. Additional causes of acute cardiopulmonary compromise include amniotic fluid embolism, pulmonary thromboembolism, cardiac tamponade, aortic dissection, and diabetic ketoacidosis. The history, physical examination, and laboratory studies will usually be sufficient to distinguish between these diagnoses.
Successful management of obstetric septic shock depends on early identification and aggressive treatment focused on stabilization of the patient, removal of underlying causes of sepsis, broad-spectrum antibiotic coverage, and treatment of associated complications. Febrile patients with mild hypotension who respond rapidly to volume infusion alone do not require invasive monitoring. In other cases, the pulmonary artery catheter should be used to guide specific therapeutic maneuvers for optimizing myocardial performance and maintaining systemic cardiac output and blood pressure. A hemodynamic approach for stabilizing pregnant women with septic shock should include (1) volume repletion and hemostasis, (2) inotropic therapy with dopamine on the basis of left ventricular function curves, and (3) addition of peripheral vasoconstrictors (phenylephrine first, then norepinephrine) to maintain afterload (Fig. 23–6).
Hemodynamic algorithm for treatment of obstetric septic shock. SVRI, systemic vascular resistance index; PCWP, pulmonary capillary wedge pressure; SBP, systolic blood pressure.
(Reproduced, with permission, from Lee W, Clark SL, Cotton DB, et al. Septic shock during pregnancy. Am J Obstet Gynecol 1988;159:410.)
Septic shock during pregnancy should be treated with a broad-spectrum antibiotic regimen such as ampicillin, gentamicin, and clindamycin. Aminoglycoside maintenance doses should be titrated in relation to serum peak and trough levels, or a 24-hour dosing regimen may be used. Newer antibiotics such as imipenem, cilastatin, vancomycin, and the extended spectrum penicillins (eg, ticarcillin) are also proving to be effective therapies. There must be a careful search for infected or necrotic foci that can result in persistent bacteremia, and surgical intervention may be necessary. In one study, 40% of septic obstetric patients required surgical removal of infected products of conception, and all survived. If chorioamnionitis is present in the septic obstetric patient, prompt delivery is necessary. However, if the pregnancy is not the cause of infection, immediate delivery is usually not required. Supportive care should also include control of fever with antipyretics, hypothermic cooling blankets, or both. Correction of maternal acidosis, hypoxemia, and systemic hypotension will usually improve any abnormalities in the fetal heart tracing.
Aggressive treatment of obstetric septic shock must rapidly and effectively reverse organ hypoperfusion, improve oxygen delivery, and correct acidosis. Priority should be given to cardiopulmonary support with the additional understanding that other major organ systems can also be severely affected.
A sequential hemodynamic approach for stabilizing obstetric septic shock with volume repletion, inotropic therapy, and peripheral vasoconstrictors is recommended. Volume therapy initially begins with 1–2 L of lactated Ringer's solution infused over approximately 15 minutes. However, it is important that volume infusion not be withheld in a hypotensive patient pending placement of a pulmonary artery catheter. The total amount of crystalloid administered should be guided by the presence or absence of maternal hypoxemia secondary to pulmonary edema and left ventricular filling pressures, as estimated by PCWP.
In general, myocardial performance will be optimized according to the Starling mechanism at a PCWP of 14–16 mm Hg. Such preload optimization is mandatory before the initiation of inotropic therapy. Blood component therapy can also be an important adjunctive measure if the patient has experienced significant hemorrhage and has developed an associated coagulopathy.
If the shock state persists despite volume replacement and adequate hemostasis, efforts should be directed toward improving myocardial performance and vascular tone. Inotropic agents such as dopamine, dobutamine, or isoproterenol are excellent choices for improving myocardial contractility in an obstetric patient with a failing heart (Table 23–4). We recommend dopamine as the first-line drug of choice for treating septic hypotension when inotropic therapy is indicated. This substance is a chemical precursor of norepinephrine that has alpha-adrenergic, beta-adrenergic, and dopaminergic receptor stimulating actions. The dopamine infusion is initiated at 2–5 μg/kg/min and titrated against its effect on improving cardiac output and blood pressure in patients with obstetric septic shock. At low doses (0.5–5.0 μg/kg/min), this sympathomimetic amine acts primarily on the dopaminergic receptors, leading to vasodilation and improved perfusion of the renal and mesenteric vascular beds. Higher dopamine doses (5.0–15.0 μg/kg/min) are associated with predominant effects on the b receptors of the heart. The beta-adrenergic effects are responsible for improved myocardial contractility, stroke volume, and cardiac output. Much higher dopamine dosages (15–20 μg/kg/min) will elicit an alpha-adrenergic effect, similar to a norepinephrine infusion, and result in generalized vasoconstriction. Vasoconstrictive action associated with high doses of infused dopamine can actually be detrimental to organ perfusion and will rarely be useful under these clinical circumstances. Although myocardial performance after dopamine therapy is best evaluated by ventricular function curves, it is reasonable to maintain a systemic cardiac index above 3 L/min/m2.
Table 23–4. Sympathomimetic and Vasopressor Drugs Useful for Therapy of Obstetric Septic Shock. ||Download (.pdf)
Table 23–4. Sympathomimetic and Vasopressor Drugs Useful for Therapy of Obstetric Septic Shock.
|Agent||Maintenance Dose Range1||Therapeutic Goals|
|Dopamine||2–10 μg/kg/min||Cardiac index ≥ 3 L/min/m2 SBP ≥ 80 mm Hg|
|Dobutamine||2–10 μg/kg/min||Optimize left ventricular function curves|
|Phenylephrine||1−5 μg/min||SVRI ≥ 1500 dynes · 5 · cm−5· Cm−2|
If satisfactory ventricular function is not achieved with dopamine, a second inotropic agent such as dobutamine (2–20 μg/kg) should be added to the dopamine regimen. Dobutamine is a direct myocardial b1 stimulant that increases cardiac output with only minimal tachycardia. Isoproterenol should be considered a third-line agent, which can be titrated at 1–20 μg/min. This drug acts primarily on beta-adrenergic receptors to increase contractility and heart rate. However, potential side effects may include ventricular ectopy, excessive tachycardia, and undesired vasodilatation. Digoxin is commonly added to the previously described regimen to improve the force and velocity of myocardial contraction. This agent is given in a loading dose of 0.5 mg IV, followed by 0.25 mg every 4 hours for a total dose of 1.0 mg. IV digoxin should be given under continuous ECG monitoring with special attention to serum potassium levels. The usual maintenance dosage during pregnancy is 0.25–0.37 mg/dL depending on plasma drug levels.
A peripheral vasoconstrictor may be initiated if there is a reduced systemic vascular resistance index (SVRI; less than 1500 dynes · s · cm−5) accompanied by a systolic blood pressure of less than 80 mm Hg despite inotropic therapy. It should be emphasized that maintenance of afterload appears to be a major hemodynamic determinant associated with maternal survival. Because of its pure alpha-adrenergic activity (which increases SVR), phenylephrine (1–5 μg/kg/min) is the initial drug of choice. Norepinephrine is only indicated for septic shock patients with decreased afterload who do not respond to volume loading, inotropic therapy, and phenylephrine. This drug is a mixed adrenergic agonist with a primary effect on the alpha receptors, which leads to generalized vasoconstriction and increased SVR. Although the therapy of septic shock should focus primarily on stabilization of maternal factors, vasopressor agents should be administered cautiously during pregnancy because they have been reported to decrease uterine blood flow in animals with experimentally induced spinal hypotension.
Some investigators have advocated large doses of corticosteroids for the acute management of septic shock, but human clinical trials have failed to demonstrate any conclusive benefit.
Newer investigational agents include corticosteroids and antiendotoxin therapy. Multicenter trials of endotoxin antibodies have suggested a possible improvement in mortality rate and organ failure in some subgroups of nonpregnant septic patients.
- Sudden, unexplained peripartum respiratory distress, cardiovascular collapse, and coagulopathy
- Bleeding secondary to coagulopathy or uterine atony (common)
- Amniotic fluid debris in right side of the heart on autopsy
Amniotic fluid embolism is a rare but potentially devastating complication of pregnancy that often results in poor obstetric outcome. Most of the information about amniotic fluid embolism has been derived from clinical reports, because the rarity of the disorder does not allow for clinical trials, and no suitable animal model exists. The first major review of the literature regarding this condition was by Morgan in 1979. This evaluated 272 cases. Since that time, a national registry was initiated by Clark. The incidence of amniotic fluid embolism is difficult to estimate and may be anywhere from 1 in 8000 to 1 in 30,000.
The basic mechanism of disease is related to the effects of amniotic fluid on the respiratory, cardiovascular, and coagulation systems. One of the classic theories hypothesized that the following 3 primary acute events occur: (1) pulmonary vascular obstruction, leading to sudden decreases in left ventricular filling pressures and cardiac output; (2) pulmonary hypertension with acute cor pulmonale; and (3) ventilation-perfusion inequality of lung tissue, leading to arterial hypoxemia and its metabolic consequences.
Only a small volume of amniotic fluid (1–2 mL) is transferred to the maternal circulation during normal labor. Thus enhanced communication between the amniotic fluid sac and the maternal venous system is necessary for amniotic fluid embolism to occur. Sites of entry may include endocervical veins lacerated during normal labor, a disrupted placental implantation site, and traumatized uterine veins. Squamous cells and trophoblastic tissue are often found in the maternal pulmonary vasculature of patients who underwent pulmonary artery catheterization. However, one must see more specific material such as mucin, fetal debris, vernix, lanugo, and squamous cells coated with white blood cells and granular debris to confirm the diagnosis. If meconium is present, a more dramatic response is seen. Fetal demise has also been shown to worsen this condition. Once amniotic debris enters the venous system, it travels rapidly to the cardiopulmonary circulation, leading to shock and arterial hypoxemia. Myocardial dysfunction may result from ischemic injury or right ventricular dilatation. Some experimental evidence suggests that amniotic fluid may have a direct myocardial depressant effect. Endothelin, a vasoconstrictive peptide found in vascular endothelial cells, has been implicated. Other factors that may play a role include proteolytic enzymes, histamine, prostaglandins, complement, and biogenic amines (eg, serotonin). These mediators are seen in other shock states such as sepsis and anaphylaxis, leading Clark to suggest that amniotic fluid embolism be termed “anaphylactoid syndrome of pregnancy.” The effects of systemic hypotension and hypoxemia may lead to cardiopulmonary collapse, renal insufficiency, hepatic failure, seizures, and coma.
Amniotic fluid embolism is almost always associated with some form of DIC. The etiology of coagulopathy associated with amniotic fluid embolism is incompletely understood, but it is known that amniotic fluid has potent total thromboplastin and antifibrinolytic activity, both of which increase with advancing gestational age. Once clotting is triggered in the pulmonary vasculature, local thrombin generation can cause vasoconstriction and microvascular thrombosis.
Limited hemodynamic observations with pulmonary artery catheterization suggest that in humans with amniotic fluid embolism, left ventricular dysfunction is the only significant hemodynamic alteration that is consistently documented. The response to amniotic fluid embolus in humans may be biphasic, initially resulting in intense vasospasm, severe pulmonary hypertension, and hypoxia. The transient period of right heart failure with hypoxia is later followed by a secondary phase of left heart failure, as reflected by elevated pulmonary artery pressure with subsequent return of right heart function. This biphasic theory may account for the extremely high maternal mortality rate within the first hour (25–34%) and explains why pulmonary hypertension can be difficult to document in patients with this disorder.
In his classic review of 272 patients with amniotic fluid embolus, Morgan characterized the main presenting clinical features: 51% presented with respiratory distress and cyanosis, 27% with hypotension, and only 10% with seizures. The Clark national registry noted 30% of patients presented with seizures or seizure-like activity, 27% with dyspnea, 17% with fetal bradycardia, and 13% with hypotension. Between 37% and 54% of patients exhibited an associated bleeding diathesis. Risk factors identified in the Morgan study included multiparity, tumultuous labor, or tetanic uterine contractions. Other studies have noted risk factors including advanced maternal age, use of uterine stimulants, caesarean section, uterine rupture, high cervical lacerations, premature separation of the placenta, and intrauterine fetal demise. Clark, however, was unable to identify any notable risk factors. Other presenting signs that have been described include tachypnea, peripheral cyanosis, bronchospasm, and chest pain.
Arterial blood oxygen tension typically indicates severe maternal hypoxemia. This hypoxemia may result from ventilation-perfusion inequality with atelectasis and associated pulmonary edema. The diagnosis of significant coagulopathy is manifested by the presence of microangiopathic hemolysis, hypofibrinogenemia, prolonged clotting times, prolonged bleeding time, and elevated fibrin split products. The chest radiograph is nonspecific, although pulmonary edema is often noted. The ECG typically reveals unexplained tachycardia, nonspecific ST- and T-wave changes, and a right ventricular strain pattern. Lung scans occasionally identify perfusion defects resulting from amniotic fluid embolism even though chest radiographic findings are normal.
Many conditions may mimic the effects of amniotic fluid embolism on the respiratory, cardiovascular, and coagulation systems. Pulmonary thromboembolism can result in severe hypoxemia with pulmonary edema. In contrast to amniotic fluid embolism, chest pain is a relatively common finding. Congestive heart failure due to fluid overload or preexisting heart disease may mimic the cardiorespiratory compromise observed during amniotic fluid embolism. Hypotension may result from several disorders, including septic chorioamnionitis or postpartum hemorrhage. Pulmonary aspiration (Mendelson's syndrome) is associated with tachycardia, shock, respiratory distress, and production of a frothy pink sputum, but is usually also associated with bronchospasm and wheezing. Other conditions in the differential diagnosis include air embolism, myocardial infarction, anaphylaxis, placental abruption, eclampsia, uterine rupture, transfusion reaction, and local anesthesia toxicity.
Amniotic fluid embolism remains one of the most devastating and unpreventable conditions complicating pregnancy. Therapeutic measures are supportive and should be directed toward minimizing hypoxemia with supplemental oxygen, maintaining blood pressure, and managing associated coagulopathies. Patients with poor oxygenation often require intubation and positive end-expiratory pressure. Adequate oxygenation will minimize related cerebral and myocardial ischemia and acidosis-induced pulmonary artery vasospasm. Pulmonary artery catheterization should be considered in the absence of coagulopathy to guide inotropic therapy with dopamine. If invasive hemodynamic monitoring is not available, rapid digitalization should be considered. Finally, the development of consumptive coagulopathy may require replacement of depleted hemostatic components in cases with significant uncontrollable bleeding or abnormal clotting parameters.
Maternal mortality rates range from 60–80%; however, a recent study quoted a 26.4% mortality rate. Of those patients who do not survive, 25% die within the first hour, and 80% within the first 9 hours. Correspondingly high perinatal morbidity and mortality rates would be expected.
- Unexplained chest pain and dyspnea (most frequent presenting symptoms)
- History of pulmonary embolism, deep venous thrombosis, prolonged immobilization, or recent surgery
- Physical examination: usually nonspecific, depending on extent of cardiopulmonary involvement, but may include tachycardia, wheezing, pleural friction rub, and pulmonary rales
- Laboratory evaluation: decreased arterial blood oxygen tension to less than 90 mm Hg in the sitting position
- Diagnostic studies: pulmonary radionuclide ventilation-perfusion scanning, spiral computed tomography, and angiography
Pulmonary thromboembolism is a rare complication of pregnancy (0.09%) but is a significant cause of maternal morbidity and mortality. Mortality has been documented as 12.8% if untreated and 0.7% if therapy is instituted. The diagnosis of deep venous thrombosis (DVT) occurs in the antepartum period approximately half the time and is evenly distributed throughout each trimester. Pulmonary embolism has a higher incidence in the postpartum period. Predisposing factors commonly include advanced maternal age, obesity, traumatic delivery, abdominal delivery, thrombophlebitis, and endometritis. Patients with underlying thrombophilias or previous thrombotic events are at greater risk for this condition.
More than 100 years ago, Virchow postulated that the basic mechanism of thrombus formation is related to a combination of vessel injury, vascular stasis, and alterations in blood coagulability. Venous thrombi consist of fibrin deposits and red blood cells with varying amounts of platelet and white blood cell components. In most cases, lower extremity and pelvic thrombi are responsible for the pathologic sequelae.
Ordinarily, the vascular endothelium does not react with either platelets or the blood coagulation system unless it is disrupted by vessel injury. Such injury exposes subendothelial cells to blood elements responsible for activation of the extrinsic coagulation cascade. Disruption of the vascular endothelium may occur during traumatic vaginal delivery or caesarean section.
Pregnancy is associated with venous stasis, especially in the lower extremities, because the enlarging uterus reduces blood return to the inferior vena cava by direct mechanical effects. Hormonal factors may also contribute to vasodilatation and stasis during pregnancy. Stasis prevents the hepatic clearance of activated coagulation factors and minimizes mixing of these factors with their serum inhibitors. In this manner, venous stasis becomes another predisposing factor for the formation of thrombi. Stasis secondary to prolonged bedrest for medical or obstetric complications will predispose a pregnant woman to increased venous stasis and formation of vascular thrombi. The period of greatest risk for thrombosis and embolism appears to be the immediate postpartum, especially after caesarean delivery.
The maternal circulation becomes hypercoagulable from alterations in the coagulation and fibrinolytic systems. Serum concentrations of most coagulation proteins, such as fibrinogen and factors II, VII, VIII, IX, and X, increase during pregnancy. These changes are also associated with decreased fibrinolytic activity, which is responsible for the conversion of plasminogen to the active proteolytic enzyme plasmin.
Women with congenital or acquired thrombophilias are at increased risk for thrombosis; in fact, up to half of women who have these events in pregnancy may have an underlying disorder. The most commonly recognized thrombophilia in the white population is factor V Leiden mutation (5%). Other less common but significant disorders include prothrombin gene mutation G20210A (2–4%), antithrombin III deficiency (0.02–0.2%), protein C deficiency (0.2–0.5%), protein S deficiency (0.08%), and hyperhomocysteinemia (1%). The antiphospholipid antibody syndrome also significantly increases maternal risk of thromboembolism and other pregnancy complications.
Once a venous thrombus is formed, it may dislodge from its peripheral vascular origin and enter the central maternal circulation. Propagation of the original venous clot or recurrent pulmonary emboli are possible. DVTs limited to the calf rarely embolize, but approximately 20% extend to the proximal lower extremity.
The subsequent cardiopulmonary effects of pulmonary embolus will depend on the location and size of thrombi in the lung. A patient with a large embolus affecting the central pulmonary circulation may present with acute syncope, respiratory distress, and shock. Smaller emboli may not have significant clinical sequelae.
No single symptom or combination of symptoms is sensitive or specific for the diagnosis of pulmonary embolus. Classic triads (hemoptysis, chest pain, and dyspnea; or dyspnea, chest pain, and apprehension) are rarely seen (Table 23–5). Chest pain and dyspnea were the most common symptoms in patients with angiographically documented pulmonary emboli (more than 80%). Physical findings include tachycardia, tachypnea (rate >16/min), pulmonary rales, wheezing, and pleural friction rub.
Table 23–5. Symptoms and Signs in 327 Patients with Pulmonary Embolus Confirmed by Angiography. ||Download (.pdf)
Table 23–5. Symptoms and Signs in 327 Patients with Pulmonary Embolus Confirmed by Angiography.
|Symptom or Sign||Frequency (%)|
|Respiration more than 16/min||92|
|Pulse more than 100/min||44|
|Fever (>37.8 °C [99.7 °F])||43|
There are no specific routine laboratory findings associated with the diagnosis of pulmonary embolus, although arterial blood gas measurements will often reveal significant hypoxemia. In the upright position, almost all healthy young pregnant women will have an arterial blood oxygen tension greater than 90 mm Hg. An alveolar-atrial (A-a) gradient of greater than 20 is suspicious for pulmonary embolus. The ECG may reveal unexplained tachycardia associated with cor pulmonale (right axis deviation, S wave in lead I, Q wave plus T wave inversion in lead III). A chest roentgenogram may be normal or may show infiltrates, atelectasis, or effusions. Thirty percent of patients with a pulmonary embolus will have a normal chest x-ray.
It is generally accepted that a normal radionuclide perfusion study can effectively rule out pulmonary embolus. Perfusion studies are occasionally equivocal, and ventilation scanning may be required to clarify the diagnosis. Ventilation scanning will improve the specificity of the perfusion study, because this will rule out airway disorders that may be responsible for reduced pulmonary perfusion. The radiation exposure is minimal (<0.1 rad). Unfortunately, a V/Q scan can only confirm a diagnosis if it is normal or indicates high probability of embolus. Therefore, 40–60% of patients will require further testing.
Spiral computed tomography (CT) is a newer form of imaging that has a sensitivity and specificity of 94% in the nonpregnant patient. Spiral CT may also be helpful in detecting other abnormalities causing pulmonary symptoms (eg, pleural effusions, consolidation, emphysema, pulmonary masses). However, this study may miss emboli below the segmental level. Magnetic resonance imaging has limited value in diagnosing pulmonary embolism in pregnancy because it has not been well studied.
If the previously mentioned studies are equivocal, pulmonary angiography should be considered. Subsequent exposure of the fetus to the relatively low levels of ionizing radiation from angiography can be minimized with appropriate pelvic shielding and selective angiography on the basis of prior radionuclide scanning.
Noninvasive Doppler should be considered as an initial diagnostic test for suspected DVT involving the lower extremities. Compression ultrasound uses firm compression with the transducer probe to detect intraluminal filling defects. Imaging is most useful for the distal iliac, femoral, and popliteal veins. Doppler is also useful for the proximal iliac veins. Sensitivity is 95%, with a 96% specificity. Impedance plethysmography measures impedance flow with pneumatic cuff inflation. Sensitivity and specificity are 83% and 92%, respectively. Compression of the inferior vena cava by the gravid second- or third-trimester uterus may cause false-positive results.
If the above noninvasive tests are inconclusive, it may be helpful to confirm the extent of the original thrombotic event by venography with pelvic shielding. The soleal calf sinuses and the valves involving the popliteal and femoral veins are the sources of most deep venous thrombi. Venography is associated with induced phlebitis in approximately 3–5% of procedures performed. Radiofibrinogen methods to detect thrombus formation will result in placental transfer of radioactive iodine and are contraindicated in pregnant or nursing women.
Any condition potentially related to cardiopulmonary compromise during pregnancy should be included in the differential diagnosis. This includes amniotic fluid and air emboli, spontaneous pneumothorax, septic shock, and preexisting heart disease.
Once predisposing risk factors to pulmonary embolus are identified, it is important to minimize the possibility of further complications. In patients at higher risk for DVT, prophylactic measures should be directed toward preventing venous stasis that leads to clot formation. Mechanical maneuvers such as raising the lower extremities 15 degrees above the horizontal, keeping the legs straight rather than bent at the knees when sitting, or performing calf flexion exercises may be useful, as may external pneumatic compression. One method used to prevent perioperative thrombophlebitis includes minidose heparin prophylaxis, 5000 U subcutaneously 2 hours before surgery and every 12 hours until routine ambulation is achieved. Minidose heparin prophylaxis significantly decreases not only the incidence of DVT but also the incidence of fatal pulmonary emboli. Subcutaneous minidose heparin may be reinstituted approximately 6 hours after delivery. Postpartum or postoperative ambulation is important in minimizing thromboembolic complications during this high-risk period. Some women may require therapeutic anticoagulation during pregnancy to prevent a thromboembolic event. Included in this category are women with artificial heart valves, antithrombin III deficiency, antiphospholipid antibody syndrome, history of rheumatic heart disease and atrial fibrillation, homozygosity for factor V Leiden or prothrombin gene mutation, and recurrent thromboembolic disease. Therapeutic anticoagulation can be achieved by using subcutaneous heparin 2–3 times a day, adjusting for a PTT of 2.0–3.0 times normal. Low molecular weight heparin (LMWH) can also be used. LMWH does not cross the placenta, and it has been shown to be relatively safe in pregnancy. In addition, complications of heparin therapy (osteoporosis, thrombocytopenia) are less common with this medication, and dosing in pregnancy usually may require fewer adjustments. The PTT does not need to be followed; instead, peak antifactor Xa levels can be checked every 4–6 weeks. It is controversial whether other disorders, such as protein C or S deficiency, or a family history of thrombophilias, require anticoagulation therapy. These patients may benefit from minidose heparin prophylaxis. LMWH can also be used for prophylaxis. The typical dose is either weight-based or empiric.
Treatment of Documented Pulmonary Embolism
Once pulmonary embolism is documented, therapeutic intervention should be directed to correction of arterial hypoxemia and any associated hypotension. Other measures should prevent clot propagation or recurrent emboli. Supplemental oxygen should be given to achieve an arterial oxygen tension of at least 70 mm Hg. A loading dose of 5000–10,000 U of heparin should be given intravenously by continuous infusion, followed by a maintenance dose of approximately 1000 U/h. The PTT should be maintained at 1.5–2.5 times control values. Other investigators recommend the use of heparin levels for monitoring anticoagulation therapy. Heparin levels may be measured on the third or fourth day and should be approximately 0.2 μg/mL, not to exceed 0.4 μg/mL. Alternatively, LMWH can also be used for therapeutic anticoagulation. Leg elevation, bedrest, and local heat will be beneficial to patients who have associated DVT. Intravenous morphine may be helpful in alleviating anxiety and ameliorating chest pain.
Intrapartum care of pulmonary embolus is complicated, and individual treatment approaches may vary. Selected patients with recent pulmonary thromboembolism, ileofemoral DVT, or heart valve prosthesis should probably continue full anticoagulation with high-dose heparin during labor or surgical procedures. Under these circumstances, the risk for potential bleeding complications from anticoagulant needs to be balanced against the risk of thromboembolism. Although there is a higher incidence of wound hematomas associated with peripartum anticoagulation, there is no clear evidence that this regimen is associated with excessive postpartum hemorrhage after normal vaginal delivery.
Postpartum patients receiving heparin may be switched over to warfarin once oral intake is tolerated. Warfarin is considered safe during breastfeeding. Heparin should be continued for the first 5–7 days of warfarin therapy. By the time heparin is discontinued, the international normalized ratio should be 2.0–3.0 times the normal value. Alternatively, it may be desirable to continue moderate doses of subcutaneous heparin (10,000 U twice daily) or LMWH. Postpartum anticoagulation should be continued for at least 3 months if the patient developed pulmonary embolus in the third trimester.
Complications of Treatment
The major complication of anticoagulant therapy is maternal or fetal hemorrhage. Heparin does not cross the placenta due to its large molecular weight, but it has been associated with maternal thrombocytopenia and osteoporosis. These effects can be reduced with LMWH. Warfarin is known to cross the placental barrier, and its use in the first trimester has been associated with embryopathy (nasal hypoplasia and stippled epiphyses) in approximately 5–8% of fetuses. Fetal nervous system abnormalities (eg, hydrocephalus) have also been noted with the use of warfarin during pregnancy.
A small percentage of patients will experience recurrent pulmonary emboli despite full anticoagulation. These patients may be candidates for vena caval ligation by a transabdominal approach under general or regional anesthesia. If the pelvis is suspected as the source of embolus, the right ovarian vein should also be ligated. It has been estimated that approximately 95% of patients with pulmonary embolism massive enough to cause hypotension eventually die. In this context, pulmonary artery embolectomy may be life-saving.
Placement of a vena caval umbrella via the internal jugular vein is an option for unstable patients with recurrent emboli who would not be prime surgical candidates. Although abdominal radiography is required for this procedure, placement of the umbrella filter does not require general anesthesia. This strategy will prevent larger emboli from reaching the pulmonary circulation.
Pulmonary embolus, with a mortality rate of 12–15% if left untreated, will develop in approximately one-fourth of untreated patients with antenatal DVT. In a review of pregnancies complicated by DVT treated with anticoagulant therapy, the incidence of pulmonary embolus was 4.5% of patients, with a maternal mortality rate of less than 1%.
American College of Obstetricians and Gynecologists. Thromboembolism in Pregnancy. ACOG Practice Bulletin No. 19. Washington, DC: ACOG; August 2000.
Disseminated Intravascular Coagulation
- History of recent bleeding diathesis, especially concurrent with placental abruption, amniotic fluid embolism, fetal demise, sepsis, preeclampsia–eclampsia, or saline abortion.
- Clinical evidence of multiple bleeding points associated with purpura and petechiae on physical examination.
- Laboratory findings classically include thrombocytopenia, hypofibrinogenemia, and elevated PT, elevated D-dimer, and fibrin split products that do not easily correct with replacement.
DIC is a pathologic condition associated with inappropriate activation of coagulation and fibrinolytic systems. It should be considered a secondary phenomenon resulting from an underlying disease state. The most common obstetric conditions associated with DIC are intrauterine fetal death, amniotic fluid embolism, pre-eclampsia-eclampsia, HELLP (hemolysis, elevated liver enzymes, and low platelet count syndrome), placenta previa, and placental abruption. Saline-induced abortion is also a cause. It is a separate condition from dilutional coagulopathy, which often follows hemorrhage and volume replacement with crystalloid and red blood cells. Dilutional coagulopathy represents a diminished amount of factors and can be easily corrected b replacing these factors. DIC, however, represents a cascade of events leading to the activation of the coagulation cascade and continued consumption of the clotting factors.
The most widely accepted theory of blood coagulation entails a “cascade theory” (Fig. 23–7). Basically, the coagulation system is divided into intrinsic and extrinsic systems. The intrinsic system contains all the intravascular components required to activate thrombin by sequential activation of factors XII, XI, IX, X, V, and II (prothrombin). The extrinsic system is initially activated by tissue thromboplastin, leading to sequential activation of factors VII, X, V, and prothrombin. Both the intrinsic and extrinsic pathways converge to activate factor X, which subsequently reacts with activated factor V in the presence of calcium and phospholipid to convert prothrombin to thrombin.
Coagulation cascade mechanism.
Thrombin is a proteolytic enzyme responsible for splitting fibrinogen chains into fibrinopeptides, leading to the formation of fibrin monomer. This central enzyme is capable of activating factor XIII to stabilize the newly formed fibrin clot and will enhance the activity of factors V and VIII.
Activation of the coagulation system also stimulates the conversion of plasminogen to plasmin as a protective mechanism against intravascular thrombosis. Plasmin is an enzyme that inactivates factors V and VIII and is capable of lysing fibrin and fibrinogen to form degradation products. Thus the normal physiologic hemostatic mechanism represents a delicate and complex balance between the coagulation and fibrinolytic systems.
Pregnancy is considered to represent a hypercoagulable state. With the exception of factors XI and XIII, there is an overall increase in the activity of coagulation factors. Fibrinogen rises as early as 12 weeks' gestation and reaches a peak level of 400–650 mg/dL in late pregnancy. The fibrinolytic system is depressed during pregnancy and labor but returns to normal levels within 1 hour of placental delivery. The early puerperium is accompanied by a secondary rise in fibrinogen, factors VIII, IX, X, and anti-thrombin III; a return to nonpregnant levels occurs by 3–4 weeks postpartum.
The complex pathophysiology of DIC is characterized by (1) procoagulant system activation, (2) fibrinolytic system activation, (3) inhibitor consumption, (4) cytokine release, (5) cellular activation, and (6) resultant end-organ damage. DIC occurs as a secondary event in a wide variety of illnesses associated with excess production of circulating thrombin. The pathophysiologic factors responsible for inappropriate activation of the clotting mechanism include endothelial cell injury, liberation of thromboplastin from injured tissue, and release of phospholipid from red cell or platelet injury. All these mechanisms may contribute to development of a bleeding diathesis resulting from increased thrombin activity. Additionally, widespread DIC will cause increased platelet aggregation, consumption of coagulation factors, secondary activation of the fibrinolytic system, and deposition of fibrin into multiple organ sites, which can result in ischemic tissue damage. The associated thrombocytopenia and presence of fibrin split products will impair hemostasis.
Specific obstetric conditions associated with DIC include the following.
DIC may occur in placental abruption involving liberation of tissue thromboplastin or possible intrauterine consumption of fibrinogen and coagulation factors during the formation of retroplacental clot. This leads to activation of the extrinsic coagulation mechanism. Placental abruption is one of the most common obstetric causes of DIC.
Retained Dead Fetus Syndrome
Another cause of DIC is retained dead fetus syndrome involving liberation of tissue thromboplastin from nonviable tissue. This cause is less common in recent years due to advanced ultrasound technology and the earlier detection of this condition.
This involves not only the release of tissue thromboplastin, but also the intrinsic procoagulant properties of amniotic fluid itself. It is likely that the associated hypotension, hypoxemia, and tissue acidosis will encourage the activation of coagulation factors.
This condition is associated with chronic coagulation abnormalities that may lead to thrombocytopenia and elevation of fibrin degradation products. It is uncertain whether endothelial damage activates procoagulant proteins and platelets or the reverse, although the former is more likely. Eclampsia is associated with DIC 11% of the time; with HELLP syndrome this increases to 15%. Preeclampsia together with placental abruption also significantly increases this association.
Saline or Septic Abortion
Saline-induced abortion has been associated with subclinical DIC. Severe cases of DIC have occurred in 1 in 400 to 1 in 1000 cases. Disease may be related to the release of tissue thromboplastin from the placenta. Septic abortion may also cause release of tissue thromboplastin or release of bacterial endotoxin (phospholipids).
Other triggers of DIC include septicemia, viremias (eg, HIV, varicella, cytomegalovirus, hepatitis), drugs, and acidosis.
Acute clinical manifestations of DIC are variable and include generalized bleeding, localized hemorrhage, purpura, petechiae, and thromboembolic phenomena. Also, fever, hypotension, proteinuria, hypoxia, hemorrhagic bullae, acral cyanosis, and frank gangrene have been described. Widespread fibrin deposits may affect any organ system, including the lungs, kidneys, brain, and liver. Chronic DIC (eg, fetal demise) is associated with slower production of thrombin and may be associated with minimal or absent clinical signs and symptoms.
Although histologic diagnosis of fibrin deposits is the only definitive manner by which DIC may be confirmed, there are a host of indirect tests suitable for the clinical evaluation of coagulopathy.
Platelets are decreased (<100,000/μL) in more than 90% of cases. In the absence of other causes, spontaneous purpura usually does not occur when platelet counts are greater than 30,000/μL.
PT measures the time required for clotting by the extrinsic pathway and is dependent on the ultimate conversion of fibrinogen to fibrin. It is prolonged in only 50–75% of patients with DIC. The explanations for the normal PTs are, first, the presence of circulating activated clotting factors such as thrombin or factor Xa, that accelerate the formation of fibrin; and second, the presence of early degradation products, which are rapidly clottable by thrombin; these may cause the test to register a normal or fast PT.
Partial Thromboplastin Time
PTT is frequently normal in DIC (40–50% of the time) and is not as helpful for establishing the diagnosis. This test measures the function of the intrinsic and final common pathways of the coagulation cascade.
TT is elevated in 80% of patients with DIC. It is affected only by the amount of circulating fibrinogen or the presence of thrombin inhibitors such as fibrin degradation products and heparin. This test specifically measures the time necessary for conversion of fibrinogen to fibrin.
Fibrinogen is often decreased, with approximately 70% of patients with DIC having a serum level less than 150 mg/dL. The normal physiologic increase of serum fibrinogen levels during pregnancy may mask a pathologic decrease in this parameter.
Values greater than 40 μg/mL are suggestive of DIC. These are elevated in 85–100% of patients with DIC. These degradation products are diagnostic of the plasmin biodegradation of fibrinogen or fibrin, so indicate only the presence of plasmin.
Clotting Time and Clot Retraction
Observation of clotting time and ability of the clot to retract can be performed by using 2 mL of blood in a 5-mL glass test tube. These are relatively simple bedside tests that can provide qualitative evidence of hypofibrinogenemia. When the clot forms, it is usually soft but not reduced in volume (adding celite will hasten this reaction). Over the next half hour, the clot should retract, with the volume of serum exceeding that of the formed clot. If this phenomenon does not occur, low serum fibrinogen levels can be suspected.
A peripheral blood smear reveals schistocytes in approximately 40% of patients with DIC.
The time required for hemostasis after skin puncture will become progressively prolonged as the platelet count falls below 100,000/μL. Spontaneous continuous bleeding from puncture sites may develop if the platelet count falls below 30,000/μL.
Several of these laboratory findings are more reliable than the classic studies.
This is a neoantigen formed as a result of plasmin digestion of cross-linked fibrin when thrombin initiates the transition of fibrinogen to fibrin and activates factor XIII to cross-link the fibrin formed. The test is specific for fibrin (not fibrinogen) degradation products and is abnormal in 90% of cases.
This is abnormal in 89% of cases.
This is abnormal 75% of the time.
Most acute episodes of generalized bleeding in obstetric patients will be related to pregnancy, but other rare causes of congenital or acquired coagulopathies need to be considered. These include idiopathic thrombocytopenic purpura, hemophilia, and von Willebrand's disease. Placental abruption is often associated with uterine tenderness, fetal bradycardia, and uterine bleeding. DIC associated with fetal demise usually does not become apparent until at least 5 weeks after the absence of heart tones has been documented. Amniotic fluid embolus is typically associated with acute onset of respiratory distress and shock. Preeclampsia is characterized by hypertension and proteinuria, which may lead to eclamptic seizures.
In addition to the potential complications of uncontrolled hemorrhage previously discussed, widespread fibrin deposition may affect any major organ system. This may include the liver (hepatic failure), kidneys (tubular necrosis), and lungs (hypoxemia).
Although individual measures will be dictated by the specific obstetric condition, the primary, most important treatment of pregnancy-related DIC is correction of the underlying cause. In most cases, prompt termination of the pregnancy is required. Moderate or low-grade DIC may not be associated with clinical evidence of excessive bleeding and often will require close observation but no further therapy.
Supportive therapy should be directed to the correction of shock, acidosis, and tissue ischemia. Cardiopulmonary support, including inotropic therapy, blood replacement, and assisted ventilation, should be implemented with the patient in close proximity to a delivery suite. Fetal monitoring, careful recording of maternal fluid balance, and serial evaluation of coagulation parameters are extremely important. If sepsis is suspected, antibiotics should be employed. Central monitoring with a pulmonary artery catheter is relatively contraindicated due to potential bleeding complications. Vaginal delivery, without episiotomy if possible, is preferable to caesarean section. Failure of improvement in the coagulopathy within several hours after delivery suggests sepsis, liver disease, retained products of conception, or a congenital coagulation defect.
Blood component therapy should be initiated on the basis of transfusion guidelines reported by the National Institutes of Health. Criteria for red cell transfusions were discussed earlier (see Hypovolemic Shock). Fresh-frozen plasma has only limited and specific indications, which include massive hemorrhage, isolated factor deficiencies, reversal of warfarin, antithrombin II deficiency, immunodeficiencies, and thrombocytopenic purpura. Although most cases of severe obstetric hemorrhage will lead to laboratory evidence of coagulation abnormalities, transfusion of fresh-frozen plasma may not always benefit these patients; the amount transfused is usually insufficient for replacing coagulation factors lost by dilution or clot formation. Even with massive obstetric hemorrhage, most procoagulant levels are above 30% of normal values, which is sufficient for maintaining clinical hemostasis in most patients. Specific replacement of fibrinogen should be accomplished by cryoprecipitate. Each unit of cryoprecipitate carries approximately 250 mg of fibrinogen. Platelets should only be administered in the face of active bleeding with a platelet count < 50,000/μL or prophylactically with platelet count 20–30,000/μL or less or after massive transfusion (>2 times blood volume). Platelets should be transfused on the basis of 1 U/10 kg body weight to raise the cell count above 50,000/μL. However, it should be noted that clotting factors containing fibrinogen may be associated with enhanced hemorrhage and also with thrombosis when given to patients with DIC. For this reason, they should be administered with extreme caution. Obstetricians should remember that Rh immune globulin should be given to Rh-negative recipients of platelets from Rh-positive donors.
Subcutaneous low-dose heparin or LMWH may be effective in treating the intravascular clotting process of DIC. Heparin acts as an anticoagulant by activating antithrombin III but has little effect on activated coagulation factors. Anticoagulation is contraindicated in patients with fulminant DIC and central nervous system insults, fulminant liver failure, or obstetric accidents. The one instance, however, in which heparin has been demonstrated to benefit pregnancy-related DIC is in the case of the retained dead fetus with an intact vascular system, in which case heparin may be administered to interrupt the coagulation process and thrombocytopenia for several days until safe delivery may be implemented.
Most cases of obstetric DIC will improve with delivery of the fetus or evacuation of the uterus. The maternal and fetal prognosis will be more closely related to the associated obstetric condition than to the coagulopathy.
Bick RL. Syndromes of disseminated intravascular coagulation in obstetrics, pregnancy, and gynecology. Hematol Oncol Clin North Am
Ginsberg JS, Greer I, Hirsh J. Use of antithrombotic agents during pregnancy. Chest 2001;119:122S.
Acute Respiratory Distress Syndrome (ARDS)
- History of gastric aspiration, infection/sepsis, preeclampsia-eclampsia, seizures, hemorrhage, coagulopathy, or amniotic fluid embolism
- Progressive respiratory distress with decreased lung compliance
- Severe hypoxemia refractory to oxygen therapy
- Diffuse infiltrates on chest roentgenogram
- Normal PCWP, with absence of radiographic evidence of congestive heart failure
ARDS is a severe form of lung disease with acute onset, characterized by bilateral infiltrates on chest x-ray, no evidence of intravascular volume overload (PCWP no greater than 18 mm Hg), and severely impaired oxygenation, demonstrated by a ratio of arterial oxygen tension (PaO2) to the fraction of inspired oxygen (FIO2) of less than 200 mm Hg. ARDS appears to occur more commonly in obstetric patients than in the general population. Its incidence in the nonpregnant population is 1.5 per 100,000, but it has been estimated to occur in between 1 in 3000 and 1 in 10,000 pregnant patients. ARDS has many causes, including gastric aspiration, amniotic fluid embolism, sepsis, coagulopathy, massive blood transfusion, and shock. It can be easily confused with cardiogenic pulmonary edema secondary to alterations in preload, myocardial contractility, or afterload. A basic understanding of the differences between cardiogenic and noncardiogenic pulmonary edema is essential before rational therapeutic intervention may be implemented.
The basic underlying pathologic change responsible for ARDS is lung injury that results in damage to the pulmonary epithelium and endothelial tissue. This, in turn, leads to enhanced vascular permeability. Factors determining the net flux of lung fluid between the capillary lumen and interstitial space are quantitatively related by the Starling equation:
Net fluid flux = k[(Pcap – Pis) – (πcap – πis)]
(k = filtration coefficient, Pcap = pulmonary capillary hydrostatic pressure, Pis = interstitial space hydrostatic pressure, πcap = pulmonary capillary serum colloid osmotic pressure, πis = interstitial space fluid colloid osmotic pressure)
Normally, fluid flows from the capillary system to the interstitial space and is returned to the systemic circulation by the pulmonary lymphatic system. An increase in left atrial pressure is observed when the left ventricle is unable to pump all the returning blood into the left atrium. Accordingly, the pulmonary capillary hydrostatic pressure increases, facilitating net movement of lung fluid into the interstitial space. When capillary fluid efflux into the interstitial space exceeds lymphatic resorption, the clinical presentation of pulmonary edema will occur. Although colloid osmotic pressure in the interstitial space and serum also plays a role in pulmonary edema, the most common factor is increased capillary hydrostatic pressure secondary to increased preload (fluid overload), afterload (severe hypertension), and decreased myocardial contractility (postpartum cardiomyopathy).
Capillary membrane permeability plays a much larger role in the genesis of noncardiogenic pulmonary edema (ARDS). Such injury due to hypoxic ischemia, vasoactive substances, chemical irritation, or microthrombi facilitates further efflux of capillary fluid and plasma proteins into the interstitium. This increase in permeability acutely produces atelectasis and diminished compliance of the lung, and damage is usually nonuniform. As the functional capability of atelectatic bronchioles diminishes, shunting and hypoxemia develop.
Maternal physiologic changes can contribute to the severity of ARDS. It has been suggested that decreased extrathoracic compliance, decreased functional residual capacity, higher oxygen deficit, limited cardiac output increases, and anemia may adversely affect the clinical presentation and course of ARDS during pregnancy.
Classic signs of respiratory distress are tachypnea, intercostal retractions, and even cyanosis, depending on the degree of hypoxemia. Fetal tachycardia or late decelerations may reflect maternal hypoxemia and uteroplacental insufficiency. Pulmonary rales in noncardiogenic pulmonary edema will be indistinguishable from those of cardiogenic pulmonary edema, but physical findings consistent with the cardiogenic disorder (ventricular gallop, jugular venous distention, and peripheral edema) are not typical features of ARDS. Unfortunately, the physiologic changes of pregnancy may mask the significance of these physical findings during the more subtle stages of respiratory distress.
Arterial blood gas determinations will reveal a progressive moderate to severe hypoxemia despite oxygen therapy. Depending on the obstetric cause of ARDS, other laboratory findings will be variable or nonspecific. The initial chest roentgenogram will often be normal, even in the presence of clinically significant respiratory distress. Within the next 24–48 hours, patchy or diffuse infiltrates will progress to prominent alveolar infiltrates (Fig. 23–8). Unlike in cardiogenic pulmonary edema, the heart will most likely be of normal size in a patient with ARDS. PCWP measured by right heart catheterization is the procedure most helpful in differentiating ARDS and pulmonary edema. The PCWP is elevated (>20 mm Hg) in cardiogenic pulmonary edema but is often normal in ARDS.
Sequence of chest radiographs from a 21-year-old woman during her first pregnancy with antepartum pyelonephritis and acute respiratory distress syndrome (ARDS). A: Normal chest film. B: Bilateral patchy pulmonary densities have developed, consistent with the diagnosis of ARDS. Much of the apparent increase in heart size is related to shallow inspiration and supine technique. C: ARDS has improved dramatically, with only minimal residual pulmonary densities.
Measurement of endobronchial fluid COP has also been used to differentiate capillary permeability-induced pulmonary edema from hydrostatic or cardiogenic pulmonary edema. In pulmonary edema secondary to capillary permeability, the COP of endobronchial fluid obtained from endotracheal tube suctioning is usually greater than 75% of the simultaneously obtained plasma COP. In cardiogenic pulmonary edema, the COP of the endobronchial fluid is usually less than 60% that of plasma.
Histopathologically, idiopathic pulmonary fibrosis and ARDS are remarkably similar. Both show evidence of acute alveolar injury, which is characterized by interstitial inflammation, hemorrhage, and edema. This is followed by a hypercellular phase, loss of alveolar structure, and pulmonary fibrosis.
ARDS should be differentiated from infectious pneumonitis and cardiogenic causes of pulmonary edema. Cardiogenic pulmonary edema will usually respond more rapidly to diuretic therapy than will ARDS, in which abnormalities in capillary membrane permeability are not quickly resolved by such intervention.
Therapy should be directed toward the prevention of hypoxemia, correcting acid–base abnormalities, removal of inciting factors, and hemodynamic support appropriate for the specific cause (eg, amniotic fluid embolus, DIC). Cardiogenic pulmonary edema is usually treated with a combination of diuretics, inotropic therapy, and afterload reduction. If a hemodynamic profile is not immediately available by pulmonary artery catheter, the clinician may elect to begin oxygen and furosemide (20 mg IV) for the presumptive diagnosis of cardiogenic pulmonary edema. By contrast, it should be apparent that the basic therapy for ARDS is supportive. Endotracheal intubation with mechanical ventilation is almost always required. The pulmonary artery catheter will be helpful in guiding fluid management and optimizing cardiac performance. Additionally, mixed venous oxygen saturation from the distal port of the pulmonary artery catheter will provide an index of oxygen utilization.
In obstetric patients, reasonable therapeutic goals for cardiorespiratory support include a mechanical ventilator tidal volume of less than 10 mL/kg, PCWP 8–12 mm Hg, arterial blood oxygen tension greater than 60 mm Hg, and mixed venous oxygen tension greater than 30 mm Hg. If unable to maintain PaO2 of at least 60 mm Hg on 50% or less inspired oxygen, positive end-expiratory pressure (PEEP) in amounts of up to 15 cm H2O may be helpful. However, it is important to avoid barotrauma to the remaining functional alveolar units, so high tidal volumes and pressures should be avoided. If the mixed venous tension is low, transfusion of red blood cells or inotropic therapy may improve oxygen transport and delivery.
Since the presence of capillary membrane abnormalities in ARDS is associated with rapid equilibration of proteinaceous material between the capillaries and interstitial spaces, intravenous colloid replacement should be discouraged in lieu of crystalloid resuscitation. A policy of relative fluid restriction should be followed, but only if the following criteria are met: stable fetus, no evidence of metabolic acidosis, normal renal function, and no need for vasopressor therapy or PEEP. Sedation and pain relief should be used liberally and may help to decrease oxygen consumption. Nutritional support for patients on prolonged mechanical ventilation must be considered; enteral feeding is preferred, as it may reduce the translocation of gut bacteria into the body. Prospective controlled studies have not demonstrated the benefit of steroid therapy for ARDS. Once therapy for cardiopulmonary support has been implemented, a thorough search for predisposing factors to ARDS must be identified for specific intervention.
Potential future therapies for ARDS include high-frequency ventilation, extracorporeal membrane oxygenation, intravenous oxygen, inhaled nitric oxide, surfactant replacement, oxygen-free radical scavengers, arachidonic acid metabolite inhibitors, antiprotease agents, antiendotoxin antibodies, anti-tumor necrosis factor antibodies, and other immunologic therapies for sepsis.
The timing of delivery in these patients is unclear from the literature. Based on the high rates of fetal death, preterm labor, fetal heart rate abnormalities, and perinatal asphyxia, most authorities recommend delivery after a gestational age of 28 weeks. In one review, only 10 of 39 patients with antepartum ARDS were discharged undelivered, and all had pylelonephritis or Varicella. Caesarean section should be reserved for standard obstetrical indications.
Older series suggested a mortality rate as high as 50–60% for patients with ARDS. More recent reviews show rates of 39–44%. One study of 41 patients demonstrated a 24.4% mortality rate; this has been attributed to possible differences in patient population as well as improvements in critical care. Many affected patients developed pulmonary complications that included barotrauma and pneumothorax. Fortunately, survivors of ARDS usually do not demonstrate permanent long-term pulmonary dysfunction.
Catanzarite V, Willms D, Wong D, et al. Acute respiratory distress syndrome in pregnancy and the puerperium: Causes, courses, and outcome. Obstet Gynecol
- Cardiopulmonary arrest can be caused by a number of conditions during pregnancy.
- Without treatment, mortality is high for both the patient and the fetus.
- Treatment should follow standard basic and advanced cardiac life support algorithms, with some minor modifications due to pregnancy.
- Perimortem caesarean delivery may assist resuscitation by relieving aortocaval compression and increasing venous return to the heart. If necessary, it should be performed within 5 minutes of arrest for maximal effect.
Many of the critical conditions discussed in this chapter can lead to cardiopulmonary arrest. Cardiopulmonary arrest during pregnancy poses a unique challenge in that both the patient and the fetus are acutely at risk of severe morbidity and mortality. The prevalence ranges from 1 in 20,000 to 1 in 50,000 pregnancies, but this prevalence is likely increasing due to the increased prevalence of advanced maternal age and obesity in pregnancy, as well as women with other chronic medical problems becoming pregnant with the assistance of assisted reproductive technologies.
Cardiopulmonary arrest in pregnancy presents similar to nonpregnant women, beginning with symptoms such as chest pain, weakness, shortness of breath, and diaphoresis, and eventually leading to cardiac arrest. However, due to the vague nature of some of the early symptoms and overlap with typical symptoms of pregnancy, early warning signs may be missed more frequently in pregnant women.
The differential diagnosis of cardiopulmonary arrest includes conditions specific to pregnancy, and those present in the general population. The most common causes of cardiopulmonary arrest in pregnant women are:
Amniotic fluid embolism
Preexisting cardiac disease
Mortality is very high for pregnant women with cardiopulmonary arrest. Fetal demise is also very common. Due to this, treatment should be initiated immediately with the goal to maximize maternal cardiac output and ventilation.
Due to the high mortality associated with cardiopulmonary arrest during pregnancy, resuscitation should follow standard algorithms of basic and advanced cardiac life support, with some modifications for the pregnancy. It is not proper to defer potential life-saving treatment for the patient due to a fear of exposing the fetus to medications and therapies. The most effective interventions for saving the fetus are those that save the mother's life. Therefore, medications used in resuscitation protocols should be used without hesitation.
It may be difficult to perform cardiac compressions due to a large uterus and engorged breasts. Compressions should not be performed in the supine position, as the gravid uterus may cause aortocaval compression, diminished venous return, and subsequent decreased cardiac output. Patients should be positioned with a left lateral tilt before compressions are applied. This can be accomplished using a moving table or a wedge or with manual displacement of the uterus. Defibrillation and cardioversion have been successfully used during pregnancy without disturbance of the fetal cardiac conduction system. It is important, however, to remove fetal monitors to prevent arcing.
The decision to perform a perimortem caesarean section should be made rapidly, within 4–5 minutes of cardiac arrest, to optimize both maternal and neonatal survival. This extreme measure can maximize maternal survival by relieving aortocaval compression and increasing blood flow back to the heart. Although the minimum gestational age for perimortem caesarean delivery is controversial, aortocaval compression begins as early as 20 weeks; therefore, hysterotomy should be considered as part of resuscitative measures in pregnancies of at least 20–22 weeks or greater.
Unfortunately, prognosis is poor, but it depends on the etiology of the cardiopulmonary arrest, coexisting morbidities, and the ability to begin resuscitation expeditiously and effectively.
American College of Obstetricians and Gynecologists. Critical Care in Pregnancy. ACOG Practice Bulletin No. 100. Washington, DC: ACOG; February 2009.