Hemodynamic assessment in a pregnant and peripartum patient

Shigeki Fujitani, MD; Marie R. Baldisseri, MD

Pregnancy complications con-tribute significantly to mater-nal and fetal morbidity andmortality. The overall preva-

lence of obstetric patients who may re-quire critical care during their pregnancyranges from 1 to 9 in 1,000 gestations (1,2). Overall maternal mortality was 11.5maternal deaths per 100,000 live birthsduring 1991–1997 (3). The goal for ma-ternal deaths per 100,000 live births inthe Unites States is �3.3 by the NationalCenter for Disease Prevention and HealthPromotion. Most maternal deaths havebeen attributed to complications fromhypertensive diseases, pulmonary illness,and cardiac diseases (4). The mortalityrate of critically ill obstetric patientsranges from 12% to 20% (5). A review of18 studies reporting on the main indica-tion for intensive care unit (ICU) admis-sion showed that hypertensive diseases(eclampsia, preeclampsia, HELLP syn-drome— hemolysis, elevated liver en-zymes, low platelet count, and hyperten-sive crisis) were the most frequent cause(30.8%), followed by hemorrhage (shock,

placental abruption, postpartum hemor-rhage, 20.3%) and pulmonary complica-tions (pulmonary edema, pneumonia,adult respiratory disease syndrome,asthma) (4).

Normal Cardiac Physiology inPregnancy

Hemodynamic studies of normal find-ings during pregnancy have demonstratedincreases in cardiac output, left ventricularstroke work, and oxygen consumption (Ta-bles 1 and 2). Cardiac output is principallyincreased by stroke volume, particularly inthe first two trimesters. In the third trimes-ter, cardiac output increases only mini-mally primarily because of the increase inheart rate as term approaches. The failureof pulmonary artery occlusion pressure orcentral venous pressure to increase in preg-nant patients, despite significant increasesin intravascular volume, reflects the de-creases in the systemic and pulmonary vas-cular resistances. The decreased vascularresistances allow both the systemic andpulmonary vasculature to accommodatehigher volumes at normal vascular pres-sures (6).

Structural Remodeling of theHeart

The heart is dramatically remodeledduring pregnancy. The remodeling of theheart causes enlargement of all fourchambers, particularly the left atrium,

which may predispose to supraventricu-lar and atrial arrhythmias. As the uterusenlarges and the diaphragm elevates, theheart is rotated upward and to the left,which results in left axis deviation on theelectrocardiogram and the appearance ofcardiomegaly on the chest radiographwithout underlying cardiac pathology.The valvular annular diameters increase,as does the thickness of the left ventric-ular wall. New murmurs often appearduring pregnancy. Mild pulmonic and tri-cuspid regurgitation occurs in �90% ofhealthy pregnant women (7, 8). One thirdof pregnant women have evidence of clin-ically insignificant mitral regurgitation(8). Left ventricular end-diastolic and leftatrial dimensions gradually return tonormal over a 2-wk period, whereas leftventricular wall thickness returns to nor-mal over a 24-wk period (8).

Pulmonary Adaptation toPregnancy

During pregnancy, there is an in-crease in tidal volume of approximately40%, a decrease in the functional resid-ual capacity by 25%, and an increase inoxygen consumption as a result of in-creased metabolic needs of the motherand the fetus. The combination of thedecreased functional residual capacityand the increased oxygen consumptiondiminishes the oxygen reserve of themother and subsequently increases the hy-poxic risk to both the mother and the fetus

From UCLA-VA Greater Los Angeles Program, In-fectious Disease Section 111F, Los Angeles, CA (SF);and the Department of Critical Care Medicine, Univer-sity of Pittsburgh Medical Center, Pittsburgh, PA(MRB).

Copyright © 2005 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/01.CCM.0000183156.73560.0C

Pathophysiology: Critical care in obstetrics has many similar-ities in pathophysiology to the care of nonpregnant women.However, changes in the physiology of pregnant woman neces-sary to maintain homeostasis for both mother and fetus, espe-cially during critical illness, result in complex pathophysiology.Understanding the normal physiologic changes during pregnancy,intrapartum, and postpartum is the key to managing critically illobstetric patients with underlying medical diseases and pregnan-cy-related complications.

Hemodynamic Monitoring: When the pathophysiology of criti-cally ill obstetric patients cannot be explained by noninvasive hemo-dynamic monitoring and the patient fails to respond to conservative

medical management, invasive hemodynamic monitoring may behelpful in guiding management. Most important, the proper interpre-tation of hemodynamic data is predicated on knowledge of normalvalues during pregnancy and immediately postpartum. Invasive he-modynamic monitoring with pulmonary artery catherization has beenused in the obstetric population, particularly in patients with severepreeclampsia associated with pulmonary edema and renal failure.(Crit Care Med 2005; 33[Suppl.]:S354–S361)

KEY WORDS: cardiac output; oxygen consumption; tidal volume;functional residual capacity; supine hypotensive syndrome; auto-transfusion; hemodynamic monitoring; cardiac comorbidity; mi-tral stenosis; pulmonary hypertension

S354 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

in response to maternal hypoventilation orapnea. Oxygen requirements increase byapproximately 30–40 mL/min in preg-nancy and are met by an increase in minuteventilation, primarily as a result of an in-creased tidal volume. The increase inminute ventilation results in a mild com-pensated respiratory alkalosis with a de-cline in the PaCO2 to �30 torr (4.0 kPa).The pH does not change (pH range 7.40–7.47) secondary to renal compensation thatresults in a decrease in the serum bicarbon-ate concentration (18–21 mEq/L) (5). Preg-nant women who present with a PaCO2 levelof 40 torr should prompt the clinician tolook for an imminent cause of respiratoryfailure.

Endocrinologic Adaptation

There are numerous endocrine andmetabolic alterations during pregnancythat principally affect the hypothalamus,pituitary, and adrenal glands. Both cortico-tropin and cortisol levels are elevated inpregnancy. The pituitary gland in preg-nancy enlarges, mainly due to proliferationof prolactin-producing cells in the anteriorpituitary. The enlargement of the pituitarymakes it more susceptible to alterations inblood supply and increases the risk of post-partum infarction when large amounts ofblood loss occur during delivery. In pa-tients with occult adrenal insufficiency, a

life-threatening adrenal crisis may be pre-cipitated by the stress of labor and delivery.Estrogen is a mediator that plays a role insome of the physiologic changes in preg-nancy, and it predominantly affects uterineblood flow (9. Progesterone and prostaglan-dins E2 and I2 have vasodilatory effects onthe uteroplacental vessels, which result inincreased cardiac output in a pregnantwoman (9).

Influence of Body Position

In the supine position, the enlargeduterus compresses the inferior vena cava,reducing venous return to the heart,which decreases stroke volume and car-diac output. However, before 24 wks theeffect of the supine position on cardiacoutput is not observed. The effect of thesupine position on cardiac output is mostmarked in late pregnancy. The decreasein cardiac output in the supine positioncompared with the lateral recumbent po-sition can be as high as 25–30%. The“supine hypotensive syndrome” of preg-nancy (severe hypoperfusion associatedwith hypotension and bradycardia) canfurther decrease cardiac output by 30–40%. The left lateral position quickly re-lieves compression of the inferior venacava. An alternate position, although lessoptimal than the left lateral position butpreferable to the supine position, is left

lateral tilt to 15° or manual displacementof the gravid uterus. The latter maneuverof left uterine displacement can be per-formed by manually moving the uterusaway from the midline to the left side whenthe patient is supine. A study to assess thematernal position during resuscitationshowed that at an angle of 27° on an in-clined board, 80% of maximum force wasobtained with cardiac compressions (10).

Clark et al. (11) observed the centralhemodynamic responses to positionchanges in ten normotensive pregnant pa-tients in the third trimester and at 11–3wks postpartum. It is of interest that theorthostatic response after pregnancy wasmuch more labile than that during thethird trimester (11). This implies a greaterhemodynamic stability in response to or-thostasis during pregnancy.

Normal Hemodynamic ProfilesDuring Pregnancy andPostpartum

Although cardiac output remains rel-atively constant in the latter half of preg-nancy, there is significant increase dur-ing active labor and immediately afterdelivery. Table 1 illustrates that duringlabor and delivery, cardiac output in-creases up to 50% and blood volume in-creases additional 300–500 mL with each

Table 1. Normal hemodynamic changes during pregnancy

Physiologic Variable Term Pregnancy Labor and Delivery Postpartum

Cardiac output Increases 30–50% Increases 50% Increases 60–80%within 15–20 min

Blood volume Increases 30–50% Additional 300–500 mL with each contraction Decreases to baselineHeart rate Increases by 15–20 beats/min Increases depend on stress and pain relief Decreases to baselineBlood pressure Decreases by 5–10 mm Hg in

midpregnancyIncrease depends on stress and pain relief Decreases to baseline

Systemic vascular resistance Decreases Increases Decreases to baselineOxygen consumption Increases by 20% Increases with stress of labor and delivery Decreases to baselineRed blood cell mass Increases by 15–20%

Table 2. Central hemodynamic changes in normal pregnancy

Measurement Nonpregnant PregnantChange FromNonpregnant

Cardiac output, L/min 4.3 � 0.9 6.2 � 10 43%Heart rate, beats/min 71 � 10 83 � 10 17%Systemic vascular resistance, dyne � cm � sec�5 1530 � 520 1210 � 266 �21%Mean arterial pressure, mm Hg 86.4 � 7.5 90.3 � 5.8 NSPulmonary artery occlusion pressure, mm Hg 6.3 � 2.1 7.5 � 1.8 NSCentral venous pressure, mm Hg 3.7 � 2.6 3.6 � 2.5 NSColloid oncotic pressure, mm Hg 20.8 � 1.0 18.0 � 1.5 �14%

NS, not significant. Reproduced from Ref. 6; Copyright © 1989, with permission from Elsevier.

S355Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

contraction in the second stage of labor.The hemodynamic changes seen duringlabor and delivery are influenced by an-esthetic and analgesic techniques, espe-cially if caudal anesthesia is used. Withinthe first 15–20 mins after delivery of thefetus and placenta, there is a substantialincrease in cardiac output because bloodis no longer diverted to the uteroplacen-tal vascular bed. Approximately 500 mL isredirected to the maternal circulation inthe so-called “auto-transfusion” effect ofpregnancy. This effect can cause cardiacoutput to increase by 60–80% after aor-tocaval compression is removed andblood volume is increased. Most of thephysiologic changes of pregnancy resolveand revert to normal within several daysafter delivery. Cardiac output remains atthe high values seen during pregnancyfor 2 days after delivery and then returnsto normal gradually within 2 wks to 3months after delivery as sodium and wa-ter balances normalize (12). Table 3 listspotential indications for hemodynamicmonitoring in obstetric patients.

Noninvasive HemodynamicMonitoring

Belfort et al. (13) assessed the corre-lation between noninvasive techniqueswith Doppler echocardiography and inva-sive hemodynamic measurements madewith the pulmonary artery catheter (PAC)in 11 critically ill obstetric patients. Theauthors showed a high correlation be-tween invasive and noninvasive tech-niques in the measurement of stroke vol-ume and cardiac output. Ventricularfilling pressures and pulmonary arterypressures also showed a similar signifi-cant correlation with invasive tech-niques. Belfort et al. (14) also reported ona series of 14 patients with an indicationfor invasive hemodynamic monitoring inwhom Doppler ultrasound was used as a

guide to clinical management. This pilotstudy concluded that the noninvasivemonitoring with Doppler ultrasound hadfacilitated management and only two pa-tients went on to have invasive monitor-ing to allow continuous monitoring (14).A study by Penny et al. (15) showed thatthe esophageal Doppler monitor consis-tently underestimated cardiac output by40% in patients with severe preeclampsia�35 yrs old. The reason for underestima-tion was uncertain, but it may have beendue to incorrect assumption of a fixedaortic diameter during systole and a fixedpercentage of blood perfusion. Bioimped-ance techniques are another noninvasivetool of central hemodynamic assessment.However, a study assessing cardiac indexin normal term pregnancy with thoracicelectrical bioimpedance and oxygen ex-traction techniques showed that the resultsof thoracic electrical bioimpedance wereinfluenced by maternal position (16, 17).

Urine Sodium

One study with seven oliguric patientswith preeclampsia evaluated volume sta-tus with fractional excretion of sodiumand invasive hemodynamic monitoring(18). The findings suggested that if urineoutput or urinary diagnostic indexes weresolely used to guide peripartum fluidmanagement, the majority of these testswould indicate hypovolemia (71.4%) inpatients with preeclampsia. The concor-dance of the reading of pulmonary arteryocclusion pressure and fractional excretionof sodium was only 57.1%. Thus, fractionalexcretion of sodium may be misleading ifused to guide fluid management in oliguricpatients with preeclampsia.

Invasive HemodynamicMonitoring

The utility of the PAC expanded fromcardiac functional diagnosis to the direc-

tion of therapy when Swan et al. (19)introduced the balloon-tipped catheter in1970. One year later, Ganz et al. (20)introduced a new technique for measure-ment of cardiac output by thermodilu-tion. Since then, the PAC has been widelyused in critically ill patients both for di-agnosis and to guide therapy. However,the clinical value of the PAC in criticallyill patients has been debated for decades.A multiple-center prospective cohortstudy demonstrated that patients withPAC had significantly increased 30-day,2-month, and 6-month mortality rate andincreased ICU and hospital length of stayby case matching multivariable regres-sion analysis (21). A multiple-center, ran-domized, blinded, controlled trial as-sessed the efficacy of PAC for high-riskperioperative patients with American So-ciety of Anesthesiologists class III or IV. Ahigher rate of pulmonary embolism wasdocumented in patients with PAC; how-ever, the survival rates at 6 and 12months did not differ among patients inthe standard care and PAC groups (22).Supportive data to evaluate the efficacy ofthe PAC in hemodynamically unstable pa-tients, such as acute respiratory distresssyndrome and septic shock, are still lack-ing.

The Pulmonary Artery Catheterin Obstetrics

In 1980s, the PAC was first used forinvasive hemodynamic monitoring incritically ill pregnant patients. Studieshave demonstrated the benefit of hemo-dynamic monitoring with placement of aPAC in patients with severe preeclampsiacomplicated with pulmonary edemaand/or renal failure (23). A summary of14 studies using the PAC in obstetricpatients demonstrated that 64.3% ofthese studies used PAC for hemodynamicassessment in patients with the diagnosisof preeclampsia or eclampsia (24). Hemo-dynamic studies of obstetric patients sug-gest placement of the pulmonary arterycatheter for severe valvular disease, car-diomyopathy, severe pulmonary and re-nal disease, and septic shock not respon-sive to standard therapies (25). Potentialindications for hemodynamic monitoringin obstetric patients are listed in Table 4.

Complications of the PulmonaryArtery Catheter

The benefit and risks of the PAC inobstetrics have been extensively reviewed

Table 3. Potential indications for hemodynamic monitoring in obstetric patients

Cardiac indicationsSevere valvular heart disease (aortic stenosis or mitral stenosis associated with pulmonary

hypertension)Cardiomyopathy with ejection fraction �15–20%Sudden cardiovascular collapse (suspected amniotic fluid embolism or pulmonary embolism)

Pulmonary indicationsAdult respiratory distress syndrome with positive end-expiratory pressure �15 mm HgSevere pulmonary disease with secondary pulmonary hypertensionPulmonary edema associated with severe preeclampsia

Renal indicationsPersistent oliguria despite fluid resuscitation (e.g., severe preeclampsia)

MiscellaneousSeptic shock refractory to fluid resuscitation and vasopressor therapy

S356 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

(1, 26–28). Complications associated withinvasive hemodynamic monitoring includepneumothorax, ventricular arrhythmias,air embolism, pulmonary infarction, pul-monary artery rupture, sepsis, local vascu-lar thrombosis, intracardiac knotting, andvalve damage. Complications have de-creased in part because of better physicianand nursing vigilance. The prevalence ofpneumothorax has decreased from 6% to1% in the early literature to �0.1%, pul-monary infarction from 7.2% in 1974 to 0to 1.3% in recent studies, and pulmonaryartery rupture from 0.1 to 0.2% to almost�0.1%. Local vascular thrombosis has de-creased with the use of heparin-bondedcatheters. Septicemia has decreased from2% to 0.5% (28).

Clinical Interpretation ofHemodynamic Data Obtained byMeans of Invasive Monitoring inPregnancy

Hemodynamic measurements ob-tained with the pulmonary artery cathe-ter provide particular information aboutthe circulation and the factors that gov-ern the pumping action of the heart: pre-load, afterload, contractility, and theheart rate. Pulmonary artery occlusionpressure and the stroke volume or leftventricular stroke work (stroke volumetimes the pressure gradient over the leftventricle) generate a Frank-Starlingcurve that allows rapid and accurate bed-side assessment of the patient’s hemody-namic condition and responses to treat-

ment (29). Central venous pressurereflects the preload of the right ventriclethat can fail as a result of left ventricularfailure and increasing pressures in theleft ventricle. In particular, in the pres-ence of cardiopulmonary disease and inconditions in which pulmonary capillarydamage with pathologic capillary wallpermeability may be expected (e.g., inseptic or severe hemorrhagic shock,eclampsia, or any other cause of adultrespiratory distress syndrome), measuredabsolute values of central venous pres-sure and pulmonary artery occlusionpressure may be extremely disparate (29).Pregnant women without any evidence ofleft heart failure and with normal mitralvalves have shown a marked discrepancybetween central venous pressure and pul-monary artery occlusion pressure values.Wallenburg (30) concluded that with theexception of very low or high readings,central venous pressure values cannot beexpected to give reliable information onhemodynamic conditions on the left sideof the heart and, for that reason, shouldnot be used in the intensive care manage-ment of critically ill pregnant patients.

Mixed venous oxygen saturation(SV̄O2) is a marker of the balance betweenoxygen delivery and oxygen consump-tion. The continuous measurement of ox-ygen saturation in mixed venous blood isperformed with specialized pulmonaryartery catheters equipped with fiberopticbundles that can transmit light to andfrom the catheter tip. SV̄O2 measured bythe PAC is within 1.5% of the SV̄O2 mea-

sured by co-oximeters (31). Continuousreading of SV̄O2 provides a generalscreening method for monitoring a groupof four variables: cardiac output, hemo-globin, arterial oxygen saturation, andoxygen consumption. Changes in SV̄O2

are extremely sensitive but are a nonspe-cific marker of cardiovascular stress. Ex-cessive oxygen consumption (shivering,spontaneous movements), decreased he-moglobin, and decreased arterial oxygensaturation all independently result in adecrease in SV̄O2. To the best of ourknowledge, there is no study to report theuse of SV̄O2 as a hemodynamic monitor inthe obstetric population.

Preeclampsia

Preeclampsia is the most common ill-ness diagnosed and treated in critical careobstetric medicine. Invasive hemody-namic monitoring with the PAC is mostcommonly used in patients with severepreeclampsia, so we will focus on theutility of the PAC in patients with severepreeclampsia. Preeclampsia is a pregnan-cy-related multiple-system disease thatusually occurs after the 32nd week ofgestation. The diagnosis of preeclampsiais confirmed by the development of hy-pertension with proteinuria, with orwithout generalized edema occurring af-ter the 20th week of gestation although itmay occur postpartum. Preeclampsia isgenerally categorized as mild or severeprimarily on the basis of the degree ofelevation in blood pressure, the degree ofproteinuria, or both. The incidence ofpreeclampsia was about 5% in one pro-spective multiple-center study in theUnited States (32). A retrospective studyinvestigating 32 obstetric patients requir-ing admission to ICU showed that pre-eclampsia was the most common reason(22%) for admission to the ICU (2).

It has been postulated that failure ofcardiac adaptations, such as increasedplasma volume and reduced systemic vas-cular resistance due to the failure of thesecond wave of trophoblastic invasioninto the spiral arteries of the uterus, isthe main pathophysiology of preeclamp-sia. Severe preeclampsia may rapidly de-teriorate the cardiopulmonary status ofthe mother and fetus. Regardless of theduration of gestation, the recommendedtreatment is prompt delivery of the fetus.For gestations �34 wks, prolongation ofpregnancy with close monitoring may beindicated to improve neonatal survival

Table 4. PAC measurements in renal failure, pulmonary edema, and eclampsia

Physiologic VariableRenal Failure

(n � 53)Pulmonary Edema

(n � 30)Eclampsia(n � 17)

Initial PAC readingsCVP, mm Hg, mean � SE 4.3 � 0.6 9.6 � 1.2a 6.5 � 1.5PAOP, mm Hg, mean � SE 10.3 � 0.7 21.0 � 2.0b 13.3 � 2.4SVR, dyne � sec � cm�5, mean � SE 1218 � 76 1348 � 103 1572 � 279CO, L/min, mean � SE 7.6 � 0.3 7.6 � 0.4 6.6 � 0.7LVSVI, g/(min � m2), mean � SE 56.3 � 3.5 50.9 � 3.8 49.8 � 7.9LVEF, % 49 66 55Intervention before PAC placement, %Volume expansion 100 60 100Diuretics 7.5 87 6Dopamine 85 10.3 47Intervention after PAC placement, %Volume expansion 90 40 88Diuretics 72 100 71

PAC, pulmonary artery catheter; CVP, central venous pressure; PAOP, pulmonary artery occlusionpressure; SVR, systemic vascular resistance; CO, cardiac output; LVSVI, left ventricular stroke volumeindex; LVEF, left ventricular ejection fraction.

ap � .001, pulmonary edema group vs. renal failure; bp � .0001, pulmonary edema group vs. renalfailure group, p � .03, pulmonary edema group vs. eclampsia group. Reproduced from Ref. 23; © 2005,with permission from Elsevier.

S357Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

and reduce short-term and long-termneonatal morbidity.

Preeclampsia with acute pulmonaryedema is a common reason for ICU ad-mission. Preeclampsia was the admittingdiagnosis in 18–28% of pregnant patientspresenting with pulmonary edema (33,34). Pulmonary edema has been noted inapproximately 2.9% of patients with se-vere preeclampsia (35). Preeclamptic pa-tients frequently have multiple causes ofpulmonary edema including increasedcapillary permeability due to endothelialcell injury, hypoalbuminemia, and after-load-induced left ventricular dysfunction.Benedetti et al. (36) noted that the devel-opment of pulmonary edema in patientswith preeclampsia seemed to be multifac-torial according to hemodynamic moni-toring. In this study, ten patients withsevere preeclampsia were evaluated he-modynamically with a pulmonary arterycatheter. Two of ten patients had cardio-genic pulmonary edema, three patientshad pulmonary capillary leak, and fivehad a reduced colloid oncotic pressure/pulmonary artery occlusion pressure gra-dient, which caused or contributed to thedevelopment of pulmonary edema.

The most frequently reported indica-tions for pulmonary artery catheterplacement in severe preeclampsia are forrenal failure and pulmonary edema Thecomplication rate in these patients is�5%. The incidence of complicationsmay be somewhat lower in pregnant pa-tients due to their young age and theabsence of underlying comorbid condi-tions (4). Pulmonary edema has been as-sociated with severe preeclampsia in ap-proximately 3% of cases (37). Patientswith preeclampsia associated with pul-monary edema were found to have signif-icant maternal (11%) and perinatal (9–23%) mortality rates (38, 39). Gilbert etal. (23) showed that patients with severepreeclampsia complicated with pulmo-nary edema had elevated right and leftfilling pressures. Their hemodynamicmeasurements influenced by therapy arelisted in Table 4. Diuretic use increasedfrom 87% to 100% after PAC was placedin these patients.

Gilbert et al. (23) found that renalfailure was the most common indicationfor pulmonary artery catheterization.Clark et al. (40) reported on nine patientswho had persistent oliguria complicatedwith severe preeclampsia. In this study,the authors described three hemody-namic subsets: a) women with a low pul-monary artery occlusion pressure and

high cardiac output and systemic vascu-lar resistance; b) women with a normal orincreased pulmonary artery occlusionpressure and cardiac output and normalsystemic vascular resistance; and c)women with an increased pulmonary ar-tery occlusion pressure and systemic vas-cular resistance and decreased cardiacoutput. These findings may underscorethe need for invasive hemodynamic mon-itoring in patients with renal failure as-sociated with severe preeclampsia if con-servative management fails. However,most patients with severe preeclampsia inthe absence of pulmonary edema and re-nal failure can be managed without inva-sive monitoring. No supporting data existdemonstrating an outcome benefit usingthis intervention, and, therefore, no evi-dence-based recommendations can bemade concerning the use of the PAC incritically ill pregnant patients.

Primary Pulmonary Hypertension

The incidence of primary pulmonaryhypertension is one to two per million(41). Women are affected four to fivetimes more often than men. Secondarypulmonary hypertension can develop as acomplication of cardiac and pulmonarydisease or as a complication of drugs suchas cocaine or appetite suppressants. Thematernal mortality rate associated withpregnancy and pulmonary hypertensionranges from 30% to 50% (42, 43).

Pulmonary artery catheterization pro-vides an early warning of rising pulmo-nary artery pressures as well as deterio-ration in right ventricular function. Itsuse has not been associated with im-proved survival (44). Because of the poorprognosis of both primary and secondarypulmonary hypertension in obstetrics,some clinicians have advocated the use ofpulmonary artery catheterization duringdeliveries and in the immediate postpar-tum period (45). In a case study of a26-wk pregnant patient with primary pul-monary hypertension, an epoprostenolinfusion was titrated using the pulmo-nary artery pressure and cardiac outputmeasurements obtained with a PAC, andthe patient had a favorable outcome (45).The author concluded that early recogni-tion and treatment with a pulmonary va-sodilator in conjunction with hemody-namic monitoring and anticoagulationtherapy for primary pulmonary hyperten-sion may reduce the likelihood of com-plications and mortality associated withpulmonary hypertension and pregnancy.

Pregnant Patients WithStructural Cardiac Disease

The incidence of significant cardiacdisease in pregnancy is �2% but is in-creasing (46, 47). Advances in medicaltherapy and in cardiac surgery have al-lowed female cardiac patients to surviveto childbearing age and to have a success-ful term pregnancy (48). Approximately90% of pregnant women with cardiac dis-ease are New York Heart Association classI or class II. These patients will effectivelytolerate the hemodynamic changes im-posed during normal pregnancy. The re-maining 10% of pregnant patient withNew York Heart Association class III or IVheart disease account for 85% of cardiacdeaths (49). In general, pregnant womentolerate valvular incompetence (such asmitral or aortic regurgitation) betterthan stenosis. This can be explained bythe reduced systemic vascular resistanceincreasing forward flow and thus decreas-ing the effects of regurgitation. In con-trast, the stenotic lesions (mitral stenosisand aortic stenosis) combined with tachy-cardia, especially during labor, result inincreases in blood volume and in cardiacoutput. This can result in elevated leftatrial pressures that increase the likeli-hood of atrial fibrillation and heart failure(50). Because of the high mortality rateassociated with cyanotic heart disease,Eisenmenger’s syndrome, or severe pul-monary hypertension, pregnancy in thesepatients should be discouraged (Table 5).

Mitral Stenosis

Mitral stenosis is the most commonrheumatic valvular lesion encountered inpregnancy. Complications of mitral ste-nosis include pulmonary edema, rightventricular failure, and atrial arrhyth-mias with risk of embolization. Severalpathophysiologic effects can exacerbatecardiac function in the setting of mitralstenosis. The expanded maternal bloodvolume can increase the risk of pulmo-nary congestion and edema, and thephysiologic tachycardia of pregnancy de-creases left ventricular filling time, whichleads to elevated left atrial pressures thatcause pulmonary edema and decreasedforward flow causing hypotension, fa-tigue, and syncope. Because of the rela-tively fixed cardiac output associated withmitral stenosis, this patient is at risk forthe development of acute pulmonaryedema. During labor, adequate pain con-trol with epidural anesthesia prevents

S358 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

tachycardia. �-blockade is also useful inthis setting to blunt the tachycardic re-sponse to pain and anxiety. Caution mustbe taken with epidural anesthesia in thesepatients since there may be a rapid de-crease in the systemic vascular resistancein a patient who has decreased ability toincrease her cardiac output to maintainblood pressure (51). In addition, thesepatients poorly tolerate the volume shiftsin the early postpartum period, precipi-tating acute pulmonary edema, andtherefore invasive hemodynamic moni-toring may be of value in the peripartummanagement (52).

Aortic Stenosis

Severe aortic stenosis is far less com-mon in the pregnant women. The great-est risk is imposed by the limited cardiacoutput, which may affect the coronaryartery perfusion. These patients do nottolerate the significant fluid shifts thatare seen during delivery and in the im-mediate postpartum period. The overallmortality rate is as high as 17%, andpulmonary artery catheterization may beuseful to maintain an adequately highocclusion pressure without precipitatingpulmonary edema (53).

Aortic and Mitral Regurgitation

In contrast, aortic and mitral insuf-ficiencies are generally well toleratedduring pregnancy. The hemodynamicchanges of pregnancy are beneficial to apatient who has aortic or mitral regur-gitation because of the increased bloodvolume and decreased systemic vascu-lar resistance that promotes forwardflow across the regurgitant valve (51).

Ischemic Heart Disease

Although coronary artery disease isuncommon in women of childbearingage, myocardial infarction may occur be-cause of increased myocardial oxygen de-mands of pregnancy and the hypercoag-ulable state. With the increase inmaternal age seen during the past fewdecades, there has been a slight increasein the number of women with peripartummyocardial ischemia and infarction. Post-partum women have a six-fold higher rateof myocardial infarction than age-matched nonpregnant women (54). Man-agement should be similar to the non-pregnant patient, with optimally at least a

2-wk period of convalescence before de-livery (55).

Congenital Heart Disease

Congenital cardiac abnormalities suchas uncomplicated atrial septal defect,ventricular septal defect, and patent duc-tus arteriosus are usually well toleratedin pregnancy. However, the developmentof pulmonary hypertension associatedwith these conditions markedly increasesthe risk, with maternal mortality re-ported at 30–50% and fetal loss of up to75% (53, 56, 57). The decreased systemicvascular resistance occurring in preg-nancy may cause right-to-left shunting,decreased pulmonary perfusion, andmarked hypoxemia. This is aggravated bydecreased filling pressures due to hemor-rhage or anesthesia and can result inmarked hypotension, hypoxemia, andsudden death (57). For these patients,maintaining high filling pressure may bevaluable during labor (56). On the con-trary, left-to-right shunts are well-tolerated in pregnancy. Left-to-right

shunts eventually can lead to pulmonaryhypertension and reversal of the shuntwith cyanosis.

Hypertrophic obstructive cardiomy-opathy may first manifest during preg-nancy with a decrease in the systemicvascular resistance and tachycardia thatreduce left ventricular filling, thus ag-gravating the outflow obstruction.Management with careful monitoringand the avoidance of hypotension, Val-salva maneuvers, and possibly �-block-ade generally produce a good outcome(58). Uncontrolled tetralogy of Fallotmay cause similar hemodynamic effectwith right-to-left shunting. Marfan’ssyndrome may occasionally be respon-sible for life-threatening aortic dissec-tion in pregnancy, especially associatedwith aortic root diameters �4.5 cm(59). Pregnancy increases the risk ofarterial rupture, and aortic dissection,intracerebral hemorrhage, and splenicartery rupture may occur (60). Hyper-tension and aortic root diameter shouldbe tightly monitored (Table 6) (61).

Table 5. Summary of maternal mortality rates associated with cardiac comorbidity

ConditionMaternal Mortality

Rate, %

All mitral stenosis 0–1.6NYHA class III or IV 5.0–7.0

Aortic stenosis 0–2.0Pulmonary hypertension 30–50Mechanical heart valve 1.0–4.0Coarctation of the aorta 0–2.0Marfan’s syndrome

All Marfan’s patients 0–1.1Patients with risk factors 50

Eisenmenger’s syndrome 36Cyanotic congenital heart disease 1.0Peripartum cardiomyopathy

In current pregnancy 18–50In previous pregnancy with persistent LV dysfunction 19

Myocardial infarction within 2 wks of delivery 50

NYHA, New York Heart Association; LV, left ventricular. Reproduced from Ref. 51; Copyright ©2004, with permission from Elsevier.

Table 6. Valvular heart lesions that are associated with high maternal risk or fetal risk duringpregnancy

1. Severe aortic stenosis with or without symptoms2. Aortic regurgitation with NYHA class III–IV symptoms3. Mitral stenosis with NYHA class II–IV symptoms4. Mitral regurgitation with NYHA class III–IV symptoms5. Aortic or mitral valve diseases that result in severe pulmonary hypertension (pulmonary

pressure �75% of systemic pressure)6. Aortic or mitral valve disease with severe left ventricular dysfunction (ejection fraction �40%)7. Mechanical prosthetic valves that require anticoagulation8. Aortic regurgitation in Marfan’s syndrome

NYHA, New York Heart Association. Reproduced from Ref. 61; Copyright © 1998, with permissionfrom The American College of Cardiology Foundation and American Heart Association, Inc.

S359Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

REFERENCES

1. Baskett TF, Sternadel J: Maternal intensivecare and near-miss mortality in obstetrics.Br J Obstet Gynaecol 1998; 105:981–984

2. Kilpatrick SJ, Matthay MA: Obstetric patientsrequiring critical care. A five-year review.Chest 1992; 101:1407–1412

3. Maternal mortality–United States,1982–1996: MMWR Morb Mortal Wkly Rep1998; 47:705–707

4. Anath CV: Epidemiology of critical illnessand outcomes in pregnancy. In: Critical CareObstetrics. Fourth Edition. Belfort MA, DildyGA, Saade GR, et al (Eds). Boston, Blackwell,2004, p 11

5. Lapinsky SE, Kruczynski K, Slutsky AS: Crit-ical care in the pregnant patient. Am J RespirCrit Care Med 1995; 152:427–455

6. Clark SL, Cotton DB, Lee W, et al: Centralhemodynamic assessment of normal termpregnancy. Am J Obstet Gynecol 1989; 161:1439–1442

7. Campos O, Andrade JL, Bocanegra J, et al:Physiologic multivalvular regurgitation dur-ing pregnancy: A longitudinal Doppler echo-cardiographic study. Int J Cardiol 1993; 40:265–272

8. Robson SC, Hunter S, Moore M, et al: Hae-modynamic changes during the puerperium:A Doppler and M-mode echocardiographicstudy. Br J Obstet Gynaecol 1987; 94:1028–1039

9. Naylor DF Jr, Olson MM: Critical care obstet-rics and gynecology. Crit Care Clin 2003;19:127–149

10. Rees GA, Willis BA: Resuscitation in latepregnancy. Anaesthesia 1988; 43:347–349

11. Clark SL, Cotton DB, Pivarnik JM, et al:Position change and central hemodynamicprofile during normal third-trimester preg-nancy and post partum. Am J Obstet Gynecol1991; 164:883–887

12. Capeless EL, Clapp JF: When do cardiovascu-lar parameters return to their preconceptionvalues? Am J Obstet Gynecol 1991; 165:883–886

13. Belfort MA, Rokey R, Saade GR, et al: Rapidechocardiographic assessment of left andright heart hemodynamics in critically illobstetric patients. Am J Obstet Gynecol1994; 171:884–892

14. Belfort MA, Mares A, Saade G, et al: Two-dimensional echocardiography and Dopplerultrasound in managing obstetric patients.Obstet Gynecol 1997; 90:326–330

15. Penny JA, Anthony J, Shennan AH, et al: Acomparison of hemodynamic data derived bypulmonary artery flotation catheter and theesophageal Doppler monitor in preeclamp-sia. Am J Obstet Gynecol 2000; 183:658–661

16. Clark SL, Southwick J, Pivarnik JM, et al: Acomparison of cardiac index in normal termpregnancy using thoracic electrical bio-impedance and oxygen extraction (Fick)techniques. Obstet Gynecol 1994; 83:669–672

17. Weiss S, Calloway E, Cairo J, et al: Compar-

ison of cardiac output measurements bythermodilution and thoracic electrical bio-impedance in critically ill versus no-criticallyill. Am J Emerg Med 1995; 13:626–631

18. Lee W, Gonik B, Cotton DB: Urinary diagnos-tic indices in preeclampsia-associated oligu-ria: Correlation with invasive hemodynamicmonitoring. Am J Obstet Gynecol 1987; 156:100–103

19. Swan HJ, Ganz W, Forrester J, et al: Cathe-terization of the heart in man with use of aflow-directed balloon-tipped catheter. N EnglJ Med 1970; 283:447–451

20. Ganz W, Donoso R, Marcus HS, et al: A newtechnique for measurement of cardiac out-put by thermodilution in man. Am J Cardiol1971; 27:392–396

21. Connors AF Jr, Speroff T, Dawson NV, et al:The effectiveness of right heart catheteriza-tion in the initial care of critically ill pa-tients. SUPPORT Investigators. JAMA 1996;276:889–897

22. Sandham JD, Hull RD, Brant RF, et al: Arandomized, controlled trial of the use ofpulmonary-artery catheters in high-risk sur-gical patients. N Engl J Med 2003; 348:5–14

23. Gilbert WM, Towner DR, Field NT, et al: Thesafety and utility of pulmonary artery cathe-terization in severe preeclampsia andeclampsia. Am J Obstet Gynecol 2000; 182:1397–1403

24. Nolan TE, Wakefield ML, Devoe LD: Invasivehemodynamic monitoring in obstetrics. Acritical review of its indications, benefits,complications, and alternatives. Chest 1992;101:1429–1433

25. Mabie WC, Sibai BM: Treatment in an obstet-ric intensive care unit. Am J Obstet Gynecol1990; 162):1–4

26. Rafferty TD, Berkowitz RL: Complications ofpulmonary artery catheterization of obstetricpatients. Int J Gynaecol Obstet 1980; 18:133–135

27. Crane-Elders AB, Nijhuis JG, van DongenPW, et al: Severe maternal morbidity andfetal mortality caused by a diagnostic Swan-Ganz procedure: A case report. Eur J ObstetGynecol Reprod Biol 1990; 37:95–98

28. Matthay MA, Chatterjee K: Bedside catheter-ization of the pulmonary artery: Risks com-pared with benefits. Ann Intern Med 1988;109:826–834

29. Wallenburg HC: Hemodynamics in Hyper-tensive Pregnancy. Amsterdam, Elsevier,1988

30. Wallenburg HC: Invasive hemodynamicmonitoring in pregnancy. Eur J Obstet Gy-necol Reprod Biol 1991; 42(Suppl):S45–S51

31. Armaganidis A, Dhainaut JF, Billard JL, et al:Accuracy assessment for three fiberoptic pul-monary artery catheters for SvO2 monitor-ing. Intensive Care Med 1994; 20:484–488

32. Sibai BM, Gordon T, Thom E, et al: Riskfactors for preeclampsia in healthy nullipa-rous women: A prospective multicenterstudy. The National Institute of Child Healthand Human Development Network of Mater-

nal-Fetal Medicine Units. Am J Obstet Gy-necol 1995; 172:642–648

33. DiFederico EM, Burlingame JM, KilpatrickSJ, et al: Pulmonary edema in obstetric pa-tients is rapidly resolved except in the pres-ence of infection or of nitroglycerin tocolysisafter open fetal surgery. Am J Obstet Gynecol1998; 179:925–933

34. Sciscione AC, Ivester T, Largoza M, et al:Acute pulmonary edema in pregnancy. Ob-stet Gynecol 2003; 101:511–515

35. Sibai BM, Mabie BC, Harvey CJ, et al: Pul-monary edema in severe preeclampsia-eclampsia: Analysis of thirty-seven consecu-tive cases. Am J Obstet Gynecol 1987; 156:1174–1179

36. Benedetti TJ, Kates R, Williams V: Hemody-namic observations in severe preeclampsiacomplicated by pulmonary edema. Am J Ob-stet Gynecol 1985; 152:330–334

37. Cotton DB, Lee W, Huhta JC, et al: Hemody-namic profile of severe pregnancy-inducedhypertension. Am J Obstet Gynecol 1988;158:523–529

38. Bhagwanjee S, Paruk F, Moodley J, et al:Intensive care unit morbidity and mortalityfrom eclampsia: An evaluation of the AcutePhysiology and Chronic Health Evaluation IIscore and the Glasgow Coma Scale score.Crit Care Med 2000; 28:120–124

39. Douglas KA, Redman CW: Eclampsia in theUnited Kingdom. Br Med J 1994; 309:1395–1400

40. Clark SL, Greenspoon JS, Aldahl D, et al:Severe preeclampsia with persistent oliguria:Management of hemodynamic subsets. Am JObstet Gynecol 1986; 154:490–494

41. Rubin LJ: Primary pulmonary hypertension.Chest 1993; 104:236–250

42. McCaffrey RM, Dunn LJ: Primary pulmonaryhypertension in pregnancy. Obstet GynecolSurv 1964; 19:567–591

43. Weiss BM, Zemp L, Seifert B, et al: Outcomeof pulmonary vascular disease in pregnancy:A systematic overview from 1978 through1996. J Am Coll Cardiol 1998; 31:1650–1657

44. Monnery L, Nanson J, Charlton G: Primarypulmonary hypertension in pregnancy: Arole for novel vasodilators. Br J Anaesth2001; 87:295–298

45. Stewart R, Tuazon D, Olson G, et al: Preg-nancy and primary pulmonary hypertension:Successful outcome with epoprostenol ther-apy. Chest 2001; 119:973–975

46. Clark SL: Cardiac disease in pregnancy. Ob-stet Gynecol Clin North Am 1991; 18:237–256

47. Metcalfe J, McNulty JH, Ueland K: Physiologyand Management. Second Edition. Boston,Little Brown, 1984

48. Danzell JD: Pregnancy and pre-existing heartdisease. J La State Med Soc 1998; 150:97–102

49. Ostheimer GW, Alper MH: Intrapartum an-esthetic management of the pregnant patientwith heart disease. Clin Obstet Gynecol 1975;18:81–97

50. Lupton M, Oteng-Ntim E, Ayida G, et al:

S360 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

Cardiac disease in pregnancy. Curr Opin Ob-stet Gynecol 2002; 14:137–143

51. Klein LL, Galan HL: Cardiac disease in preg-nancy. Obstet Gynecol Clin North Am 2004;31:429–452

52. Clark SL, Phelan JP, Greenspoon J, et al:Labor and delivery in the presence of mitralstenosis: Central hemodynamic observations.Am J Obstet Gynecol 1985; 152:984–988

53. Clark SL: Cardiac disease in pregnancy. CritCare Clin 1991; 7:777–797

54. Salonen Ros H, Lichtenstein P, Bellocco R, etal: Increased risks of circulatory diseases in

late pregnancy and puerperium. Epidemiol-ogy 2001; 12:456–460

55. Hankins GD, Wendel GD Jr, Leveno KJ, et al:Myocardial infarction during pregnancy: Areview. Obstet Gynecol 1985; 65:139–146

56. Clark SL: Structural Cardiac Disease in Preg-nancy. Second Edition. Boston, BlackwellScientific, 1991

57. Pitkin RM, Perloff JK, Koos BJ, et al: Preg-nancy and congenital heart disease. Ann In-tern Med 1990; 112:445–454

58. Oakley GD, McGarry K, Limb DG, et al: Man-agement of pregnancy in patients with hy-

pertrophic cardiomyopathy. BMJ 1979;1:1749–1750

59. Lipscomb KJ, Smith JC, Clarke B, et al: Out-come of pregnancy in women with Marfan’ssyndrome. Br J Obstet Gynaecol 1997; 104:201–206

60. Barrett JM, Van Hooydonk JE, Boehm FH:Pregnancy-related rupture of arterial aneu-rysms. Obstet Gynecol Surv 1982; 37:557–566

61. Bonow RO. ACC/AHA Guidelines for the Man-agement of Patients with Valvular Heart Dis-ease. J Am Coll Cardiol 1998; 32:1486–1588

S361Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)