Special Feature

The Critically Ill Pregnant Patient: Chapter I

By Saadia R. Akhtar, MD, MSc, Pulmonary and Critical Care Medicine, Yale University School of Medicine, is Associate Editor for Critical Care Alert

Care of the critically ill pregnant patient poses unique challenges. The normal physiology of a pregnant patient differs considerably from that of a non-pregnant patient and these differences may affect many aspects of routine care: resuscitation, mechanical ventilation, choice of drugs and use of diagnostic studies are some examples. Pregnancy may increase the risk of some life-threatening disorders (such as venous thromboembolic disease). In addition, there are some potentially serious diseases (such as eclampsia) that occur only during pregnancy. Although obstetric patients make up < 1% of ICU admissions, and ICU admissions are necessary for only about 0.5% of all pregnant women, the mortality for these patients can be as high as 20%.1-4 Finally, the condition and care of the mother directly impact the survival and health of the fetus. For these reasons, it is essential that all personnel working in an ICU familiarize themselves with the critical care issues of pregnancy. A multidisciplinary approach to care of these patients, involving an obstetrician and a neonatologist, is imperative.

This essay will begin the discussion on management of the critically ill pregnant patient, which will be continued in a later issue of this newsletter. It will review the cardiovascular and pulmonary physiology of normal pregnancy and ways in which this may impact ICU management. It will address maternal resuscitation in the event of cardiorespiratory arrest and comment on some routine ICU issues such as choice of vasopressors, sedatives and antibiotics, and the safety of diagnostic studies during pregnancy.

Cardiovascular Physiology of Normal Pregnancy5-6

A variety of physiological changes are necessary in order to ensure adequate perfusion and oxygenation of the growing fetus as well as to continue to meet the needs of the mother. Blood flow to the uterus must increase 10-fold (up to 500-700 mL/min) during pregnancy. This is accomplished by an increase in blood volume as well as a rise in cardiac output. Blood volume expands by about 40-50% (or 2 L): the increase in plasma volume is proportionally higher than the change in red cell mass, resulting in the relative anemia of pregnancy. Total body water also increases and plasma oncotic pressure decreases: the latter is one of the factors that predisposes pregnant women to edema formation. Simultaneously, the cardiac output rises by up to 30-50% (even further increases occur during labor). Initially, this rise is due primarily to an increase in stroke volume; as pregnancy proceeds, the heart rate also goes up by about 10-15 beats per minute on average. Echocardiography demonstrates a definite increase in left ventricular mass.

Peripheral vasodilation begins early in pregnancy and results in a 20-30% reduction in systemic vascular resistance. This may be the stimulus for the rise in cardiac output. It is also one of the factors contributing to venous stasis and increased risk for venous thrombosis during pregnancy. Other typical hemodynamic changes include a 20-30% reduction in pulmonary vascular resistance, a 10-20% drop in diastolic blood pressure and a smaller decrease in systolic blood pressure.

These cardiovascular changes peak at about the end of the second trimester and though the return to normal is fairly rapid (hemodynamic changes begin to resolve within days to 2 weeks post-partum and return to pre-pregnancy levels by 6 months), it is not immediate. Following delivery then, there is a large increase in maternal blood volume and preload: for women with any significant cardiac disease or pulmonary hypertension, this may lead to acute decompensation.

Finally, unique to pregnancy is the phenomenon of aortocaval compression by the enlarging uterus when the patient is in the supine position. This begins to occur by about 20 weeks of gestation and can result in a 25% reduction in cardiac output. Additionally, up to 30% of women (usually in the second half of pregnancy) will exhibit the more severe symptoms of the supine hypotension syndrome with marked hypotension, bradycardia, and syncope while supine.7 Thus it is important to place bed-bound pregnant patients in a left lateral tilt position. Note that caval compression also leads to increased venous stasis and is one of the factors predisposing pregnant women to thromboembolic disease.

In addition to usual ICU cardiopulmonary monitoring of the mother, continuous electronic fetal heart rate monitoring is essential in all circumstances. Non-reassuring readings (such as bradycardia, abnormal fetal heart rate variability, late decelerations or absence of appropriate spontaneous accelerations) should prompt further evaluation.

Pulmonary Physiology of Normal Pregnancy5-6

Pulmonary mechanics and gas exchange are altered during normal pregnancy. The enlarging uterus pushes upward on the diaphragm and leads to a reduction in functional residual capacity (FRC) of up to 25%. Total lung capacity changes only slightly, by 0 to -4%. Other lung volumes are unaltered. Airway mucosal edema and hyperemia are observed but do not impact airflow rates. These airway mucosal changes do however lead to symptomatic rhinitis in 30% of pregnant women. They may also be an important reason for the relatively high failed intubation rates reported in obstetric patients. It is essential to keep this in mind and downsize endotracheal and naso- or oro-gastric tubes.

In order to accommodate the 15-20% rise in oxygen consumption and carbon dioxide production during pregnancy, maternal minute ventilation must increase considerably. A change of up to 40-50% above baseline is seen and leads to the familiar complaint of dyspnea during the latter half of pregnancy. This increase in minute ventilation is accomplished primarily by an increase in tidal volume which may be caused by progesterone’s central respiratory stimulant effect. (Tachypnea is uncommon and, when present, is likely a sign of disease.) Interestingly, the resulting minute ventilation is beyond what is necessary to maintain neutrality. That is, a mild respiratory alkalosis and associated mild compensatory metabolic acidosis are found throughout pregnancy. Normal arterial blood gas values during pregnancy are: pH 7.40-7.47, pCO2 28-32 mm Hg, pO2 > 100 mm Hg, HCO3 18-21 mEq/L. Due to the reduced FRC, mild hypoxemia and elevated alveolar-to-arterial PO2 gradient [P(A-a)O2] may develop in the supine position.8 In addition, the low FRC in conjunction with baseline high maternal oxygen consumption leads to poor maternal oxygen reserve and makes pregnant patients prone to hypoxia with even short periods of apnea. Pre-oxygenation before sedation and intubation is thus particularly important for these patients.

When pregnant patients require mechanical ventilatory support, it is essential to adjust ventilator settings to target the usual arterial blood gas values described above.9 Further hyperventilation leads to uterine vasoconstriction, reduced placental and fetal perfusion and fetal distress. Similarly, although an arterial PCO2 up to 55-60 mm Hg appears to be tolerated well, more severe hypoventilation and hypercapnia clearly result in fetal acidemia and distress. The role of bicarbonate infusions for acidemia in pregnant patients is undefined. Animal studies suggest that bicarbonate does not cross the placenta but carbon dioxide does, thus implying that bicarbonate infusion may worsen fetal acidosis; it is unclear whether this finding holds for humans.10 In regard to airway pressures, while maternal plateau pressures may be falsely’ elevated due to increased abdominal pressure and decreased chest compliance during normal pregnancy, there is no evidence to support targeting or accepting higher plateau pressures than for non-pregnant patients. Thus, although there are no specific studies of low-tidal-volume ventilation or management of acute lung injury in pregnant patients, experts suggest applying this strategy for usual indications but limiting permissive hypercapnia to PCO2 no higher than 55-60 mm Hg.9

Maternal Resuscitation9,11

Fetal oxygenation is related to maternal oxygen delivery, placental function, and fetal extraction. Maternal oxygen delivery is determined by blood flow to the uterus and by the oxygen content of that blood: maternal anemia, hypoxemia, hypotension or uterine vasoconstriction will all impair fetal oxygenation. If necessary, blood flow to critical maternal organs will be maintained at the expense of the fetus. Thus, for a critically ill pregnant patient, rapid support and correction of maternal blood pressures and gas exchange are essential for fetal viability.6

Usual Advanced Cardiac Life Support (ACLS) guidelines apply for pregnant patients who experience cardiorespiratory arrest. As fetal oxygenation depends on maternal oxygenation and the latter has poor reserve, establishing an airway quickly is essential. As noted above, airway edema and hyperemia should prompt use of smaller-than-usual artificial airways. Manual lateral displacement of the uterus or positioning of the patient in a left lateral tilt position are essential to minimize aortocaval compression. Fetal monitoring is reasonable during resuscitation, although the findings generally correlate closely with maternal condition. It is essential, if fetal monitors are in place, to remove or disconnect them prior to defibrillation or cardioversion.

If initial efforts appear to be failing after 4 to 5 minutes of resuscitation and the patient is at or beyond about 24 weeks of gestation, emergent Cesarean section is recommended. ACLS interventions should be continued during delivery. Delivery may ease resuscitation of the mother by increasing preload and cardiac output and facilitating maternal ventilation. Good fetal neurological outcome is most likely if delivery occurs within 4 to 5 minutes of arrest but has been reported after as much as 45 minutes after maternal arrest (with continued aggressive cardiopulmonary resuscitation).12

Choice of Medications During Pregnancy6,13-16

The distribution and clearance of many drugs are greatly altered by the normal physiological changes of pregnancy. (These include the increases in total body water and plasma volume as discussed above. They also include a decrease in plasma protein concentration, slowing of gastric emptying and changes in renal perfusion and hepatic perfusion and metabolism.) These changes in pharmacokinetics may place the mother at increased risk of drug toxicity or under-dosing.’ Furthermore, almost all medications cross the placenta to some degree and may adversely affect the fetus. Thus, it is essential to be vigilant in the choice, dosing, and monitoring of medications during pregnancy.

The Federal Food and Drug Administration (FDA) classifies medications for use during pregnancy into 5 categories: A (controlled human studies demonstrate safety of the drug during pregnancy); B (presumed safe based on animal studies), C (see below), D (clear evidence of some risk to the fetus); or X (absolutely contraindicated in pregnancy). The majority of drugs (66% of those in the Physicians’ Desk Reference) are classified as category C (risk cannot be ruled out: human studies are lacking, and animal studies are either positive for fetal risk or lacking).17 Most vasopressors and sedatives used in the ICU fall into risk category C.

Untreated maternal hypotension is obviously detrimental to the mother and the fetus. However, vasopressors have the potential to constrict uterine vessels and reduce blood flow to the placenta and fetus despite improving maternal blood pressures. There is little literature on choice of vasopressors for pregnant patients with persistent shock. Most available studies address transient use of vasopressors for hypotension associated with spinal anesthesia or in the operative setting. These along with animal and in vitro studies suggest that ephedrine is least likely to cause uterine vasoconstriction and this remains the recommended first line agent for pregnant patients (again, particularly during obstetric anesthesia). Phenylephrine and metaraminol are also favored for short-term use. Norephinephrine is suggested for persistent shock in pregnant patients in the ICU setting but there are no specific data for or against this.

With respect to sedatives and paralytics, the only agents clearly shown to be teratogenic are benzodiazepines. When administered in the first trimester, benzodiazepines increase the risk of cleft deformities. There is weak evidence that they may also be associated with developmental delays. The primary risk of other sedatives (narcotics, propofol) or the paralytics is that of fetal and neonatal depression. When these agents are used, it is important to remember that the newborn infant may need cardiorespiratory support for a period of time following delivery.

Finally, antimicrobial choice should be guided first and foremost by what is appropriate treatment for the suspected source of infection or sepsis. When possible, drugs from the penicillin, cephalosporin or macrolide classes should be chosen because extensive experience has demonstrated their safety during pregnancy. One exception in the macrolide class is clarithromycin, which has adversely affected fetal development in animal studies. Quinolones are also FDA pregnancy class C due to arthropathy noted in animal studies of ciprofloxacin. Aminoglycosides may cause fetal ototoxicity but are otherwise safe. Sulfonamides alone are acceptable but trimethoprim does increase the risk of congenital malformations with first trimester exposure. The antifungal of choice during pregnancy is amphotericin B. Lastly, although the anti-herpetic agents are safe to use, those antivirals targeting influenza are not tested or approved in pregnancy.

Diagnostic Studies and Fetal Radiation Exposure6, 18-19

Diagnostic studies crucial for care of the critically ill pregnant patient should never be delayed or deferred: maternal survival and good health are essential for good fetal outcomes. With that in mind, it is reassuring that estimated fetal radiation exposure for most diagnostic studies is low and acceptable. With the appropriate use of an abdominal lead shield, the fetal radiation exposure (in rads) for common studies is: PA and lateral chest X-ray £ 0.001-0.008, ventilation-perfusion scan < 0.031, chest CT scan with pulmonary embolism protocol < 0.0131 and pulmonary angiogram 0.221-0.405. Not surprisingly, abdominal and pelvic CT scans are associated with the highest fetal radiation exposures, of up to 5 rads. Teratogenicity does not appear to occur until the total fetal radiation exposure over the course of gestation exceeds 10 rads. Cumulative exposure to < 5 rads increases the risk of childhood leukemia from 1/2800 to 2/2800 but is not teratogenic. (To put this into perspective, these cancer rates are less than the rates of spontaneous congenital malformations.) Thus, in general, it is quite safe to proceed with selected and clinically indicated radiological studies in pregnant patients.


Routine ICU care must be tailored to support the unique physiology and meet the special needs of the pregnant patient. An understanding of the normal cardiopulmonary physiology of pregnancy provides the foundation on which to develop further expertise and comfort in the management of this patient population. This essay has reviewed these normal changes and discussed mechanical ventilation, resuscitation, usual ICU medications and diagnostic studies during pregnancy. Other aspects of the critical care of pregnancy will be addressed in future reports.


  1. Kilpatrick SJ, Matthay MA. Chest. 1992;101:1407-1412.
  2. Lapinsky SE, et al. Can J Anaesth. 1997;44:325-329.
  3. Mahutte NG, et al. Obstet Gynecol. 1999;94:263-266.
  4. Hazelgrove JF, et al. Crit Care Med. 2001;29:770-775.
  5. Chesnutt AN. Crit Care Clin. 2004;20:609-615.
  6. Lapinsky SE, et al. Am J Respir Crit Care Med. 1995; 152:427-455.
  7. Kinsella SM, Lohmann G. Obstet Gynecol. 1994;83:774-788.
  8. Garcia-Rio F, et al. Eur Respir J. 1997;10:2371-2375.
  9. Campbell LA, Klocke RA. Am J Respir Crit Care Med. 2001;163:1051-1054.
  10. Ron-El R, et al. Acta Obstet Gynecol Scand. 1980;59:225-230.
  11. The American Heart Association in collaboration with the International Liaison Committee on Resuscitation. Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation. 2000;102 (8 Suppl):I229-I252.
  12. Yildirim C, et al. Yonsei Med J. 2004;45:561-563.
  13. Webster WS, Freeman JA. Expert Opin Pharmacother. 2003;4:949-961.
  14. Niebyl JR. Am J Perinatol. 2003;20:405-414.
  15. Lee A, et al. Anesth Analg. 2002;94:920-926.
  16. Chestnut DH, et al. Anesthesiology. 1988;68:363-366.
  17. Doering PL, et al. Am J Obstet Gynecol. 2002;187:333-339.
  18. Toppenberg KS, et al. Am Fam Physician. 1999;59:1813-1818.
  19. Winer-Muram HT, et al. Radiology. 2002;224:487-492.