Critical Care Obstetrics part 23 doc

Non-Invasive Monitoring
209
that increased atrial distension in pre - eclampsia triggered a
diuretic response.
These data have been contested. The most detailed study, to
date, by Borghi et al. [15] described detailed cardiac fi ndings
among 40 women with mild pre - eclampsia compared to a control
cohort of pregnant women and non - pregnant controls. This
study showed a progressive rise in left ventricular mass between
non - pregnant women compared to normal pregnancy with a
further increase in mass among women with pre - eclampsia.
Ejection fraction and fractional shortening decreased in normal
pregnancy while not reaching statistical signifi cance. However,
women with pre - eclampsia had a signifi cant reduction in both
these parameters in comparison to non - pregnant women. In
addition, left ventricular end - diastolic volume rose signifi cantly
in pre - eclampsia. Together with a fall in cardiac output in the
pre - eclamptic group, these fi ndings suggest a compensatory
increase in ventricular size to maintain cardiac output against an
elevated systemic vascular resistance.
The latter study also showed changes in the peak fi lling veloci-
ties of the left ventricle during diastole. The E/A ratio fell signifi -
cantly during pregnancy, partly refl ecting increased preload. In
pre - eclamspsia further augmentation of the A - wave peak velocity
resulted in further signifi cant reduction in the ratio. Collectively
these data support the notion of changes in both cardiac systolic
and diastolic function. The authors also measured ANP levels. In
keeping with previous studies elevated levels of ANP were found
in pregnancy with further increments occurring in pre - eclampsia.
These could not be accounted for by differences in atrial size
although a signifi cant correlation was found between left ven-

tricular mass and volume in women with pre - eclampsia [15] .
Doppler ultrasound and cardiomyopathy
Doppler ultrasound has an important role in the management
and evaluation of women with impaired ventricular function.
Echocardiography is used to delineate impaired left ventricular
systolic function in women with suspected peripartum cardiomy-
opathy. It plays a further role in the ongoing evaluation of women
once this diagnosis has been made. Specifi cally the prognosis has
been related to the normalization of left ventricular size and func-
tion within 6 months of delivery [16] . Currently accepted opinion
is that approximately 50% of affected women will recover normal
function. Those who have persistently impaired function face a
signifi cant risk of mortality [16] .
Subsequent pregnancies in women with a prior diagnosis of
cardiomyopathy demand careful echocardiographic assessment.
Although no clear agreement exists regarding risk, those with
persistently abnormal left ventricular function have been advised
against pregnancy. Confl icting reports have been made concern-
ing those who become pregnant. De Souza et al. report on the
evaluation of seven women who became pregnant after develop-
ing peripartum cardiomyopathy in a previous pregnancy. All
pregnancies were well tolerated without signifi cant change in
patients. The fi ndings of this study showed a high correlation
between invasive and non - invasive techniques in the measure-
ment of stroke volume and cardiac output. Ventricular fi lling
pressures and pulmonary artery pressures also showed a similar
signifi cant correlation with invasive techniques [11] .
The specifi c choice of echocardiographic technique for esti-
mating stroke volume and ejection fraction was explored in the
same group of patients. Comparisons between M - mode and

two - dimensional Doppler techniques revealed similar fi ndings,
although M - mode echocardiography was not possible in 2 out of
11 subjects secondary to body habitus and paradoxical motion of
the intraventricular septum. This study also allowed calculation
of the ejection fraction by dividing the stroke volume by the
end - diastolic volume. Using this equation, similar results were
obtained by all the methods employed for estimating left ven-
tricular function in pregnant women [12] .
Belfort et al. have reported a series of 14 patients with an indi-
cation for invasive hemodynamic monitoring in whom Doppler
ultrasound was used as a guide to clinical management. These 14
women had a spectrum of pathologies ranging from intractable
hypertension to complex cardiac lesions and included women
with oliguria and pulmonary edema. This pilot study concluded
that the non - invasive monitoring had facilitated management
and only two patients went on to have invasive monitoring in
order to allow continuous monitoring. Large volumes of fl uid
were administered to some of these patients (up to 8 L of crystal-
loid) without the development of fl uid overload or pulmonary
edema. To date this is the only study that has indicated the poten-
tial utility of routine rapid echocardiographic assessment of left
ventricular function in critically ill obstetric patients [13] .
Doppler ultrasound and pre - eclampsia
Doppler echocardiography has provided a ready means of study-
ing women at risk for developing hypertensive complications
during pregnancy. Longitudinal studies have demonstrated that
women with non - proteinuric or gestational hypertension main-
tain a hyperdynamic circulation with a high cardiac output
throughout pregnancy. By contrast women destined to develop
pre - eclampsia have signifi cantly elevated cardiac output without

any change in systemic resistance in the preclinical phase of the
disease. This is followed by a fall in cardiac output and increasing
resistance coincident with the onset of clinical disease [14] .
More recently, studies have focused on echocardiographically
described cardiac structure and function in pre - eclampsia, espe-
cially in relation to levels of atrial and brain natriuretic peptide
(ANP, BNP). Initial work had related elevated ANP levels to
increased left atrial dimensions following delivery in normal
pregnancies. These increased ANP levels did not lead to any
demonstrable diuresis in normal postpartum women. Women
with pre - eclampsia had bigger atria and higher ANP levels in the
early puerperium and these changes were associated with natri-
uresis and diuresis. The hypothesis related by these fi ndings was
Chapter 15
210
abosorbs the infrared light more strongly (Figure 15.1 ). This
allows the simultaneous acquisition of peripheral signals from
which the ratio of oxy - to deoxyhemoglobin can be calculated
and expressed as a percentage of oxyhemoglobin saturation.
Oximetry may be based on transcutaneous measurements or
can be derived from mixed venous blood via a probe located in
a pulmonary artery catheter. The peripheral pulse oximetry
devices rely on detection of pulsed alterations in light transmitted
between transmitter and a photodetector. This fi ltered signal is
necessary to eliminate the signal arising from venous blood that
would contain more deoxyhemoglobin.
Although oximetry is regarded as an effective method of moni-
toring oxygenation, some limitations are recognized. They
include the assumptions that methemoglobin and carboxyhemo-
globin are not present in signifi cant concentrations. Mixed

venous oxygen saturation monitoring is less frequently used than
peripheral oxygen saturation monitoring. It also shows greater
spontaneous variation than peripheral monitors but has a clinical
role to play in determining the balance between peripheral oxygen
delivery and peripheral oxygen consumption. This is a robust
measurement that will refl ect changes in cardiac output, hemo-
globin concentration, arterial and venous hemoglobin oxygen
saturation. This provides useful clinical information in many
clinical circumstances. A number of the determinants of the ulti-
mate mixed venous oxygen saturation value have the potential to
change at any given moment (hemoglobin, oxygen saturation,
and cardiac output). It is therefore important to understand that
it is only when all other parameters remain stable that changes in
the mixed venous oxygen saturation refl ect changes in cardiac
output.
Capnometery
Exhaled gas can be evaluated using an infrared probe and a pho-
todetector set to detect carbon dioxide. This is usually found in
symptomatology. Echocardiographic studies showed no change
in left ventricular end - diastolic diameters, with an increase occur-
ring in left ventricular fractional shortening [17] . Other studies
have reported similarly successful pregnancies [18] . However,
there are papers suggesting a risk of recurrent cardiomyopathy
and impaired contractile reserve, even in those with apparently
normal left ventricular function before pregnancy [19,20] .
Doppler ultrasound and other medical disorders
Echocardiocardiography is an essential investigation in women
with structural heart disease due to valvular damage or congenital
malformation [21 – 27] . Echocardiography will also contribute to
the diagnosis of Libman – Sacks endocarditis, which occurs,

though not frequently, among women suffering from systemic
lupus erythematosus, with or without antiphospholipid anti-
bodies [28,29] .
The management of Marfan ’ s syndrome also requires echocar-
diographic assessment because of the risk of catastrophic aortic
dissection. Transesophageal echocardiography is the preferred
method for evaluating the ascending aorta. The risk of dissection
correlates with an aortic root diameter greater than 4 cm [30] .
Aortic dissection may also occur under other circumstances and
may follow the use of crack cocaine [31,32] .
The role of transesophageal D oppler
Esophageal Doppler monitoring of hemodynamic data has been
carried out in adult intensive care units and found to be equiva-
lent to data derived from pulmonary artery catheter measure-
ments [33] . Pregnancy data are few, and to date only one study
has reported the use of transesophageal Doppler monitoring in
pregnancy compared to pulmonary artery catheters. This study
showed that the Doppler consistently underestimated cardiac
output by 40% in women under the age of 35 years [34] . This
error may be due to the assumptions implicit in the algorithm
used to calculate output. These assumptions include a fi xed aortic
diameter during systole and a fi xed percentage of blood perfusing
upper and lower parts of the body. Pregnancy physiological
changes probably invalidate these assumptions. The authors nev-
ertheless conclude that esophageal Doppler may contribute to the
estimation of trends in cardiac output over time in pregnancy.
Oximetry in the intensive care environment
Spectrophotometry is the detection of specifi c light frequencies
refl ected by a range of molecules. Specifi c molecules refl ect spe-
cifi c frequencies and their refl ective properties differ with changes

in molecular conformation. Oximetry is the detection of oxygen-
ated and deoxygenated blood. Deoxygenated hemoglobin absorbs
more light at 660 nm whereas at 940 nm oxygenated hemoglobin
600 700
(RED)
660
nm
(INFRARED)
910 nm
800 900 1000 WAVELENGTH (nm)
Absorbance
Hb
HbO
2
0.1
10
Figure 15.1 Oximetry in the intensive care environment. The oxygenated
hemoglobin refl ects more light at 660 nm whereas at 940 nm deoxyhemoglobin
refl ects infrared light more strongly .
Non-Invasive Monitoring
211
dinally (at 4 - week intervals) during normal gestation. The resis-
tance index (RI), pulsatility index (PI), and cerebral perfusion
pressure (CPP) were calculated using the velocity and blood pres-
sure data. The mean value, and the 5% and 95% percentiles, were
defi ned and it was noted that the middle cerebral artery (MCA)
velocities and the resistance and pulsatility indices decrease, while
the CPP increases, during normal pregnancy. Figure 15.2 shows
the CPP change during normal pregnancy. This study defi ned the
normative ranges for middle cerebral artery velocity, resistance

indices, and cerebral perfusion pressure during normal human
pregnancy using longitudinally collected data.
Women with pre - eclampsia and hypertensive women with
superimposed pre - eclampsia have been studied using TCD
ultrasound. Findings among these subjects include globally
elevated cerebral perfusion pressures and lower cerebral vascular
resistance compared to normotensive controls [38,39] . The
increased pressures were not directly related to blood pressure
alone [39] .
Doppler ultrasound, bias and confounders
There are a number of confounding infl uences that can affect the
interpretation of Doppler cerebral velocity data. These include
any factors that may: (i) increase the CO
2
or H
+
tension in the
cerebral circulation; (ii) decrease or increase the hemoglobin con-
centration; (iii) independently alter the diameter of the vessel
being studied at the point of insonation; and (iv) introduce error,
such as cigarette smoking and changes in posture. In pregnant
women (v) gestational age is another important factor that
requires consideration, since as the pregnancy progresses there
are signifi cant hemodynamic changes.
Increased CO
2
tension leads to cerebral vasodilation, as does
acidosis. Patients undergoing cerebral Doppler studies should,
ideally, be studied in a steady state or should have their end - tidal
CO

2
tension measured in order to control for fl uctuations. Even
the expiratory limb of a ventilator circuit. Expired gas shows a
pattern of increasing carbon dioxide concentration related to the
sequential expiration of air in the upper airway followed by air
from the alveoli. The end - expiratory (or end - tidal) carbon
dioxide concentration should approximate the partial pressure of
carbon dioxide in arterial blood. The development of a gradient
between these measurements refl ects an increase in anatomical
or physiological dead space. In the latter event, low cardiac output
and pulmonary embolism may both affect the measurement.
Changes in end - tidal partial pressure of carbon dioxide have been
correlated to changes in cardiac output and may be used as a
means of monitoring the effi cacy of resuscitation.
Transcranial D oppler ultrasound
Compared to the physiologic alterations in other vascular beds
during gestation the normal cerebral blood fl ow changes of preg-
nancy are poorly documented. This is due, partly, to technical
diffi culties associated with in vivo studies of blood fl ow in the
human brain. Angiography, the gold standard in the evaluation
of the cerebral vasculature, is an invasive test and presents obvious
ethical concerns for its use in normal pregnant women. Very little
data exist on the physiologic adaptations of the brain to preg-
nancy in the current literature and most texts dealing with the
changes of pregnancy do not address this issue at all. There are
also ethical problems with using angiography and other method-
ologies involving radiation, as well as magnetic resonance imaging
during pregnancy. The advent of Doppler ultrasound, and in
particular transcranial Doppler (TCD) ultrasound, has changed
this. It is now possible to acquire Doppler - derived velocity infor-

mation from most of the basal brain arteries (including almost
all of the circle of Willis branches) using a non - invasive tech-
nique. Using these data, it is possible to diagnose arterial malfor-
mations, functional abnormalities, and physiological changes in
brain blood velocity. One can detect direction and velocity of
blood fl ow, and from this infer the presence of distal or proximal
arterial constriction or dilatation. In addition, TCD can be used
to determine real - time changes over very short time intervals and
to continuously monitor cerebral blood velocity during surgical
procedures, or experimental drug protocols. TCD has been
extensively used in the clinical scenario by neurologists and neu-
rosurgeons to detect and follow cerebral vasospasm in patients
with subarachnoid hemorrhage [35] . TCD has also been used for
neurological monitoring during cardiopulmonary bypass in
pediatric cardiac surgeries [36] . Investigators are beginning to use
TCD to defi ne pregnancy - induced/associated changes in the
cerebral circulation.
Belfort et al. [37] have recently defi ned the hemodynamic
changes, specifi cally velocity, resistance indices, and cerebral per-
fusion pressure, in the middle cerebral artery distribution of the
brain during normal pregnancy. TCD ultrasound was used to
determine the systolic, diastolic, and mean blood velocities in the
middle cerebral arteries in non - laboring women studied longitu-
Figure 15.2 Cerebral perfusion pressure (CPP) changes during a normal
pregnancy as detected by middle cerebral artery (MCA) velocity.
Chapter 15
212
tone also tends to close the arterioles when pressure falls during
the pulse cycle. Under conditions of low vascular resistance, the
arterioles remain open throughout the pulse cycle and the active

smooth muscle tone never causes them to close completely.
However, even a slight increase in arteriolar tone will narrow the
diameter of open arterioles and, in some cases, cause them to
close completely when the pressure within them falls at the end
of the pulse cycle. The pressure at which an arteriole closes is
called its “ critical closing pressure ” [35,42] . Critical closing pres-
sure explains why arterioles close as pressure falls during the pulse
cycle and why fewer arterioles are open at the end of the pulse
cycle than earlier, when pressure is at its systolic maximum. Thus,
pressure at the end of a pulse cycle is less effective in perfusing
the capillary bed than that early in the cycle. In the brain, CPP is
reduced as arteriolar resistance rises abruptly due to more and
more arterioles reaching their critical closing pressure. Another
feature of arteriolar tone is its effect in delaying the fl ow of blood
from arteries to capillaries. When arteriolar tone is high it reduces
the rate of blood fl ow from arteries to capillaries. This maintains
the arterial blood pressure at a higher level for a longer portion
of the pulse cycle than if the arteriolar tone was low and there
was a rapid run - off of blood. Blood pressure distends the arterial
segments and blood is effectively stored in the arteries while the
pressure decays during the pulse cycle. The amount stored in each
segment depends on the compliance of the artery and the pres-
sure gradient between the lumen and the region outside the
artery. The result of storing blood in the arteries and reducing the
rate of fl ow through arterioles is to slow the deceleration of blood
fl ow during the pulse cycle. The more compliant the arterial
segment, the slower the deceleration during the pulse cycle. This
feature of arteriolar tone interacting with arterial pressure and
arterial compliance affects the shape of the velocity profi le during
the pulse cycle. When arteriolar tone is low, blood velocity rapidly

rises to a maximum and falls quickly to a minimum. In contrast,
when arteriolar tone is high, the blood fl ow velocity falls more
slowly. The area under the pulsatile amplitude of the velocity
waveform, and the height of the pulse velocity wave, may be used
to estimate the proportion of blood fl ow stored in arterial seg-
ments during the peak of the pulse cycle and released when pres-
sure falls during the cycle.
One of the major problems with the currently used Doppler
indices is that they were initially developed for use in peripheral
vascular examination of large - diameter arteries such as the
femoral, dorsalis pedis, and brachial arteries. Indices such as the
PI and RI focus on the systolic component of the velocity profi le.
The traditional Doppler indices of hemodynamics (i.e. the RI and
PI) provide limited data regarding arteriolar tone when applied
to the cerebral circulation. Both the RI, defi ned as:

velocity velocity velocity
systolic diastolic systolic
−
()()

and the PI, defi ned as:

velocity velocity velocity
systolic diastolic mean
−
()
()

the minimal increases in tidal volume and respiratory rate associ-

ated with labor contractions may be of importance. Labor itself
has been shown to be associated with decreases in mean middle
cerebral artery fl ow velocity.
Hemoconcentration and hemodilution are also important and
attention should be paid to the hematocrit level in studies where
blood loss or volume infusion may have altered the hemoglobin
content of the blood during the study period.
The segmental nature of the vasospasm seen in some condi-
tions, notably pre - eclampsia, is of concern as well, since the same
segment of artery can show completely different velocity profi les
depending on its state of contraction. Thus, if the region of vessel
being insonated is apt to change, its diameter velocity readings
may be inaccurate, particularly if some indication of downstream
vascular condition is being extrapolated. In this regard, the M1
portion of the middle cerebral artery has been shown to be
unlikely to change diameter [40] , since it is well supported by
alveolar tissue in its bony canal. The angle of insonation is critical
since the velocity is related to the cos of the angle of insonation
( q ). If q is less than 10 ° the error involved is almost negligible and
quite acceptable for most purposes. Because of the anatomy of
the bony canal through which the M1 portion of the MCA runs,
the angle of insonation very rarely exceeds 10 ° . This ensures that,
in almost all cases, once the optimum signal is obtained the angle
of insonation is less than 10 ° .
The effect of maternal cigarette smoking on middle cerebral
artery blood fl ow velocities during normal pregnancy was
described by Irion et al. [41] . They found that the systolic, dia-
stolic, and mean velocities of the middle cerebral artery, detected
in both the left lateral decubitus and sitting positions, were sig-
nifi cantly higher at 18 and 26 weeks gestation in women who

smoked cigarettes. They determined that the number of cigarettes
smoked positively correlated with increased middle cerebral
artery velocities. This factor must be taken into account when
studying women known to smoke, and is an important con-
founding factor in some of the earlier studies published. Posture
should be taken into account when studying pregnant, and in
particular, pre - eclamptic pregnant women. A change from lying
to sitting has been shown to signifi cantly increase both systolic
and diastolic velocities in the middle cerebral artery in such
patients.
Another important variable that must be taken into account
when studying pregnant women is the gestational age. As preg-
nancy advances there is a reduction in middle cerebral artery
velocity which should be controlled for when comparing women
of different gestational ages.
Cerebral perfusion pressure
Under normal conditions, the arterioles in the cerebrovascular
system are responsible for about 80% of the vascular resistance.
Because arterioles have active smooth muscle tone, they do not
behave simply like tubes of variable dimension. Smooth muscle
tone in the arterioles reduces their diameter when systolic pres-
sure is transmitted into them via the arteries. In addition, this
Non-Invasive Monitoring
213
eclamptics routine use of this modality is not recommended until
further research has confi rmed the fi ndings. However, in those
cases of refractory seizure activity unresponsive to conventional
therapy, TCD may offer another diagnostic option. In those cases
where CPP is shown to be signifi cantly elevated, drug therapy can
be tailored to lowering the CPP (i.e. with labetalol) versus those

rare cases where there is a low CPP and presumably cerebral
ischemia from underperfusion, a cerebral vasodilator such as
nimodipine can be used.
Conclusion
Non - invasive techniques of monitoring will become increasingly
utilized as an alternative to the invasive techniques currently
practiced in most intensive care units. This technology, however,
requires expertise in the application and interpretation of data.
Even correctly interpreted data are of unknown utility and strin-
gent evaluation is necessary before this (often) expensive technol-
ogy is incorporated into routine clinical practice.
References
1 Flachskampf FA , Hoffmann R , Verlande M , Schneider W , Ameling
W , Hanrath P . Initial experience with a multiplane transoesophageal
echo - transducer: assessment of diagnostic potential . Eur Heart J 1992 ;
13 ( 9 ): 120 – 126 .
2 Krebs W , Klues HG , Steinert S et al. Left ventricular volume calcula-
tions using a multiplanar transoesophageal echoprobe; in vitro vali-
dation and comparison with biplane angiography . Eur Heart J 1996 ;
17 ( 8 ): 1279 – 1288 .
3 Flachskampf FA . The standard TEE examination: procedure, safety,
typical cross - sections and anatomic correlations, and statistical analy-
sis . Semin Cardiothorac Vasc Anesth 2006 ; 10 ( 1 ): 49 – 56 .
4 Lee LC , Black IW , Hopkins A , Walsh WF . Transoesophageal echocar-
diography in heart disease – old technologies, new tricks . Aust N Z J
Med 1992 ; 22 ( 5 Suppl ): 527 – 531 .
5 Mesa A , Jessurun C , Hernandez A et al. Left ventricular diastolic
function in normal human pregnancy . Circulation 1999 ; 99 ( 4 ):
511 – 517 .
6 Valensise H , Novelli GP , Vasapollo B et al. Maternal cardiac systolic

and diastolic function: relationship with uteroplacental resistances. A
Doppler and echocardiographic longitudinal study . Ultrasound Obstet
Gynecol 2000 ; 15 ( 6 ): 487 – 497 .
7 Mone SM , Sanders SP , Colan SD . Control mechanisms for physio-
logical hypertrophy of pregnancy . Circulation 1996 ; 94 ( 4 ): 667 – 672 .
8 Geva T , Mauer MB , Striker L , Kirshon B , Pivarnik JM . Effects of
physiologic load of pregnancy on left ventricular contractility and
remodelling . Am Heart J 1997 ; 133 : 53 – 59 .
9 Poppas A , Shroff SG , Korcarz CE et al. Serial assessment of the car-
diovascular system in normal pregnancy. Role of arterial compliance
and pulsatile arterial load . Circulation 1997 ; 95 ( 10 ): 2407 – 2415 .
10 Robson SC , Hunter S , Boys RJ , Dunlop W . Serial changes in pulmo-
nary haemodynamics during human pregnancy: a non - invasive study
using Doppler echocardiography . Clin Sci (Colch) 1991 : 80 ( 2 ):
113 – 117 .
are signifi cantly infl uenced by the systolic velocity which refl ects
large - caliber arterial constriction. These indices were originally
developed using older technology and larger diameter arteries
(femoral artery and aorta). The typical waveform shape from
such arteries has a tall peaked systolic component, a steep dia-
stolic slope, and a low/non - existent diastolic component. The
smaller diameter arteries that are now easily visualized with
modern equipment provide completely different waveforms from
those seen in the larger diameter, higher velocity, and higher
resistance vessels. Using indices that focus on the systolic velocity
tends to ignore aspects of waveform shape peculiar to lower resis-
tance vascular beds. Specifi cally, the typical waveform seen in low
resistance, low velocity, smaller diameter arteries has a low sys-
tolic velocity, fl atter diastolic downslope, and a proportionately
higher diastolic velocity, than that seen in high - resistance, high -

velocity arteries.
A further defi ciency of the current cerebral Doppler assessment
techniques is that they fail to take into account the systemic arte-
rial pressure, a vital component of the cerebral perfusion pres-
sure. In 1986 Aaslid et al. [43] validated a Doppler method of
estimating CPP. They measured velocity in the middle cerebral
artery (Doppler ultrasound), and intraventricular pressure and
radial arterial blood pressure (direct strain gauge transducers) in
10 patients undergoing a supratentorial shunt procedure. They
estimated CPP using the following ratio: (mean fl ow velocity)/
(pulsatile amplitude of fl ow velocity) multiplied by the arterial
blood pressure. To increase the accuracy, Fourier analysis was
used and only the amplitude of the fi rst harmonic of the pulsatil-
ity in both fl ow - velocity and arterial blood pressure recordings
were used. They expressed their calculations as:

CPP ABP=×
V
V
0
1
1

where V
0
is the mean and V
1
is the amplitude of the fi rst harmonic
of the velocity waveform, and ABP
1

is the fi rst harmonic of the
arterial pressure wave. Their experimental results confi rmed the
validity of the method. The standard deviation between estimated
CPP (CPP
e
) and measured CPP (CPP
m
) was 8.2 mmHg at a CPP
of 40 mmHg, and the mean deviation was only 1 mmHg.
Belfort et al. [44] have adapted the method of Aaslid et al. [43]
by altering the formula to refl ect the area under the pulsatile
amplitude of the fl ow velocity and arterial blood pressure wave-
forms rather than the fi rst harmonic. Their equation, using areas
under pulsatile amplitudes, is as follows [44] :

CPP
Velocity
Velocity Velocity
BP BP
mean
mean diastolic
mean di
=
−
×−
aastolic
()

Recently Belfort et al. [45] have suggested that elevated CPP,
rather than decreased CBF, is the key determinant of cerebral

injury in pre - eclampsia/eclampsia. Since this technology is still in
its infancy as a non - invasive monitoring tool in severe pre -
Chapter 15
214
29 Hojnik M , George J , Ziporen L , Shoenfeld Y . Heart valve involvement
(Libman – Sacks endocarditis) in the antiphospholipid syndrome .
Circulation 1996 ; 93 ( 8 ): 1579 – 1587 .
30 Elkayam U , Ostrzega E , Shotan A , Hehra A . Cardiovascular problems
in pregnant women with the Marfan syndrome . Ann Intern Med 1995 ;
123 ( 2 ): 117 – 122 .
31 Ecknauer E , Schmidlin D , Jenni R , Schmid ER . Emergency repair of
incidentally diagnosed ascending aortic aneurysm immediately after
caesarean section . Br J Anaesth 1999 ; 83 ( 2 ): 343 – 345 .
32 Madu EC , Shala B , Baugh D . Crack - cocaine - associated aortic dissec-
tion in early pregnancy – a case report . Angiology 1999 ; 50 ( 2 ):
163 – 168 .
33 Singer M , Clarke J , Bennet ED . Continuous hemodynamic monitor-
ing by esophageal Doppler . Crit Care Med 1989 ; 17 : 447 – 452 .
34 Penny JA , Anthony J , Shennan AH , de Swiet M , Singer M . A com-
parison of hemodynamic data derived by pulmonary artery fl otation
catheter and the esophageal Doppler monitor in preeclampsia . Am J
Obstet Gynecol 2000 ; 183 : 658 – 661 .
35 Aaslid R , Huber P , Nornes H . Evaluation of cerebrovascular spasm
with transcranial Doppler ultrasound . J Neurosurg 1984 ; 60 ( 10 ):
37 – 41 .
36 Polito A , Ricci Z , Di Chiara L , Giorni C , Iacoella C , Sanders SP ,
Picardo S . Cerebral blood fl ow during cardiopulmonary bypass in
pediatric cardiac surgery: the role of transcranial Doppler – a sys-
tematic review of the literature . Cardiovasc Ultrasound 2006 ; 4 : 47 .
37 Belfort MA , Tooke - Miller C , Allen JC et al. Changes in fl ow velocity,

resistance indices, and cerebral perfusion pressure in the maternal
middle cerebral artery distribution during normal pregnancy . Acta
Obstet Gynecol Scand 2001 ; 80 : 104 – 112 .
38 Riskin - Mashiah S , Belfort MA . Preeclampsia is associated with global
cerebral hemodynamic changes . J Soc Gynecol Investig 2005 ; 12 ( 4 ):
253 – 256 .
39 Belfort MA , Tooke - Miller C , Allen JC Jr , Varner MA , Grunewald C ,
Nisell H , Herd JA . Pregnant women with chronic hypertension and
superimposed pre - eclampsia have high cerebral perfusion pressure .
BJOG 2001 ; 108 ( 11 ): 1141 – 1147 .
4 0 G i b o H , C a r v e r C C , R h o t o n A L , J r . , L e n k e y C , M i t c h e l l R J .
Microsurgical anatomy of the middle cerebral artery . J Neurosurg
1981 ; 54 ( 2 ): 151 – 169 .
41 Irion O , Moutquin JM , Williams K , Forest JC . Reference values and
infl uence of smoking on maternal middle cerebral artery blood fl ow .
Am J Obstet Gynecol 1996 ; 174 : 367A .
42 Dewey RC , Pieper HP , Hunt WE . Experimental cerebral hemody-
namics. Vasomotor tone, critical closing pressure, and vascular bed
resistance . J Neurosurg 1974 ; 41 ( 5 ): 597 – 606 .
43 Aaslid R , Lundar T , Lindegaard KF , Nornes H . Estimation of cerebral
perfusion pressure from arterial blood pressure and transcranial
Doppler recordings . In: Miller JD , Teasdale GM , Rowan JO , Galbraith
SL , Mendelow AD , eds. Intracranial Pressure VI . Berlin, Heidelberg :
Springer - Verlag , 1986 : 226 – 229 .
44 Belfort MA , Tooke - Miller C , Varner M et al. Evaluation of a non -
invasive transcranial Doppler and blood pressure bases method for
the assessment of cerebral perfusion pressure in pregnant women .
Hypertens Pregnancy 2000 ; 19 ( 3 ): 331 – 340 .
45 Belfort MA , Varner MC , Dizon - Townson DS , Grunewald C , Nisell
H . Cerebral perfusion pressure, and not cerebral blood fl ow, may be

the critical determinant of intracranial injury in preeclampsia: a new
hypothesis . Am J Obstet Gynecol 2002 ; 187 : 626 – 634 .


11 Belfort MA , Rokey R , Saade GR , Moise KJ Jr . Rapid echocardio-
graphic assessment of left and right heart hemodynamics in critically
ill obstetric patients . Am J Obstet Gynecol 1994 ; 171 ( 4 ): 884 – 892 .
12 Rokey R , Belfort MA , Saade GR . Quantitative echocardiographic
assessment of left ventricular function in critically ill obstetric
patients: a comparative study . Am J Obstet Gynecol 1995 ; 173 ( 4 ):
1148 – 1152 .
13 Belfort MA , Mares A , Saade GR , Wen T , Rokey R . Two - dimensional
echocardiography and Doppler ultrasound in managing obstetric
patients . Obstet Gynecol 1997 ; 90 ( 3 ): 326 – 330 .
14 Bosio PM , McKenna PJ , Conroy R , O ’ Herlihy C . Maternal central
hemodynamics in hypertensive discorders of pregnancy . Obstet
Gynecol 1999 ; 94 ( 6 ): 978 – 984 .
15 Borghi C , Esposti DD , Immordino V et al. Relationship of systemic
hemodynamics, left ventricular structure and function, and plasma
natriuretic peptide concentrations during pregnancy complicated by
preeclampsia . Am J Obstet Gynecol 2000 ; 183 ( 1 ): 140 – 147 .
16 Pearson GD , Veille JC , Rahimtoola S et al. Peripartum cardiomyopa-
thy: National Heart, Lung, and Blood Institute and Offi ce of Rare
Diseases (National Institutes of Health) workshop recommendations
and review . JAMA 2000 ; 283 ( 9 ): 1183 – 1188 .
17 De Souza JL , de Carvalho Frimm C , Nastari L , Mady C . Left ventricu-
lar function after a new pregnancy in patients with peripartum car-
diomyopathy . J Card Fail 2001 ; 7 ( 1 ): 30 – 35 .
18 Sutton MS , Cole P , Plappert M , Saltzman D , Goldhaber S . Effects of
subsequent pregnancy on left ventricular function in peripartum car-

diomyopathy . Am Heart J 1991 ; 121 ( 6 Pt 1 ): 1776 – 1778 .
19 Demakis JG , Rahimtoola SH , Sutton GC et al. Natural course of
peripartum cardiomyopathy . Circulation 1971 ; 44 : 1053 – 1061 .
20 Lampert M , Weinert L , Hibbard J , Korcarz C , Lindheimer M , Lang
RM . Contractile reserve in patients with peripartum cardiomyopathy
and recovered left ventricular function . Am J Obstet Gynecol 1997 ;
176 ( 1 Pt 1 ): 189 – 195 .
21 Gultekin F , Baskin E , Gokalp A , Dogan K . A pregnant woman with
Ebstein ’ s anomaly. A case report . Mater Med Pol 1994 ; 26 ( 4 ):
149 – 151 .
22 Ben Farhat M , Gamra H , Betbout F et al. Percutaneous balloon
mitral commissurotomy during pregnancy . Heart 1997 ; 77 ( 6 ):
564 – 567 .
23 Martinez Reding J , Cordero A , Kuri J , Martinez Rios MA , Salazar E .
Treatment of severe mitral stenosis with percutaneous balloon val-
votomy in pregnant patients . Clin Cardiol 1998 ; 21 ( 9 ): 659 – 663 .
24 Niwa K , Perloff JK , Kaplan S , Child JS , Miner PD . Eisenmenger syn-
drome in adults: ventricular septal defect, truncus arteriosus, univen-
tricular heart . J Am Coll Cardiol 1999 ; 34 ( 1 ): 223 – 232 .
25 Wilansky S , Phan B , Adam K . Doppler echocardiography as a predic-
tor of pregnancy outcome in the presence of aortic stenosis: A case
report . J Am Soc Echocardiogr 1999 ; 12 : 324 – 325 .
26 Barbosa PJ , Lopes AA , Feitosa GS et al. Prognostic factors of reheu-
matic mitral stenosis during pregnancy and puerperium . Arq Bras
Cardiol 2000 ; 75 ( 3 ): 215 – 224 .
27 Mangione JA , Lourenco RM , dos Santos ES et al. Long - term follow -
up of pregnant women after percutaneous mitral valvuloplasty .
Catheter Cardiovasc Interv 2000 ; 50 ( 4 ): 413 – 417 .
28 Gleason CB , Stoddard MF , Wagner SG , Longaker RA , Pierangeli S ,
Harris EN . A comparison of cardiac valvular involvement in the

primary antiphospholipid syndrome versus anticardiolipin - negative
systemic lupus erythematosus . Am Heart J 1993 ; 125 ( 4 ): 1123 – 1129 .
215
Critical Care Obstetrics, 5th edition. Edited by M. Belfort, G. Saade,
M. Foley, J. Phelan and G. Dildy. © 2010 Blackwell Publishing Ltd.
16
Pulmonary Artery Catheterization
Steven L. Clark
1
& Gary A. Dildy III
2


1
Women ’ s and Children ’ s Clinical Services, Hospital Corporation of America, Nashville, TN, USA

2
Maternal - Fetal Medicine, Mountain Star Division, Hospital Corporation of America, Salt Lake City, UT and Department of
Obstetrics and Gynecology, LSU Health Sciences Center, School of Medicine in New Orleans, New Orleans, LA, USA
Introduction
Following its introduction into clinical medicine three decades
ago, the pulmonary artery catheter was shown to play an impor-
tant role in the management of critically ill patients in a number
of specialties, including obstetrics [1 – 6] . Several early prospective
trials demonstrated the benefi ts of pulmonary artery catheteriza-
tion in select critically ill patients. Such benefi ts include a reduc-
tion in operative morbidity and mortality in certain complicated
surgical patients and a signifi cant mortality reduction in patients
in shock in whom catheter - obtained parameters led to changes
in therapy [7,8] . In one study, management recommendations

changed as a direct result of knowledge obtained by pulmonary
artery catheter placement in 56% of patients admitted to an
intensive care unit [9] . In patients with major burn injuries,
survival is predicted by early response to pulmonary artery
catheter - guided resuscitation [10] . This technique, however, was
not without its critics [11] . In a non - randomized observational
study, Califf and colleagues [12] demonstrated increased mortal-
ity and cost associated with pulmonary artery catheterization,
and suggested that a randomized trial aimed at better patient
selection was needed. An subsequent randomized controlled trial
(n = 201) of the pulmonary artery catheter in critically ill patients
concluded that its use is not associated with increased mortality
[13] .
In response to concerns of increased morbidity and mortality
associated with the pulmonary catheter in observational studies,
the National Heart, Lung, and Blood Institute (NHLBI) and the
US Food and Drug Administration (FDA) conducted the
Pulmonary Artery Catheterization and Clinical Outcomes work-
shop in 1997 to develop recommendations to improve pulmo-
nary artery catheter utility and safety [14] . They concluded that
a “ need exists for collaborative education of physicians and
nurses in performing, obtaining, and interpreting information
from the use of pulmonary artery catheters. This effort should be
led by professional societies, in collaboration with federal agen-
cies, with the purpose of developing and disseminating standard-
ized educational programs. ” Areas given high priority for clinical
trials were pulmonary artery catheter use in persistent/refractory
congestive heart failure, acute respiratory distress syndrome,
severe sepsis and septic shock, and low - risk coronary artery
bypass graft surgery.

Since this conference, several investigators have attempted to
better defi ne benefi ts and risks of pulmonary artery catheteriza-
tion both in general categories of critical illness, and in specifi c
subsets of critically ill patients. Most studies that used broad and
non - specifi c patient inclusion criteria (such as “ critically ill
patients ” or “ high - risk surgical patients ” ) have, not surprisingly,
generally detected neither benefi cial nor detrimental effects of
pulmonary artery catheterization on mortality rates [15 – 17] and
the use of pulmonary artery catheterization has decreased in the
United States over the past decade [18] . On the other hand,
studies directed at specifi c subsets of critically ill patients have
proven much more informative. It would appear, for example,
that such monitoring techniques are not typically associated with
improved survival in patients with acute lung injury and acute
respiratory distress syndrome [19,20] . On the other hand survival
benefi t has been demonstrated in patients with severe trauma or
illness, those admitted in severe shock and in older trauma
patients [21,22] . Another study demonstrating lack of benefi t of
pulmonary artery catheterization in patients with severe septic
shock does not address the question of whether patients so
managed before late or end - stage disease may benefi t from the
information provided by these techniques [23] . Interpretation of
such data is further compounded by the general lack of uniform,
evidence - based management protocols for most patients in
whom pulmonary artery catheters are utilized. No diagnostic
testing modality can improve outcomes in any disease in the
absence of effective therapy [15,24] . Thus, at present, the pulmo-
nary artery catheter should be viewed neither as a panacea for all
seriously ill patients, nor as a technique lacking diagnostic value
Chapter 16

216
punctured at the junction of the two clavicular heads, and the
needle is directed with constant aspiration toward the ipsilateral
nipple at an angle approximately 30 ° superior to the plane of the
skin. Free fl ow of venous blood confi rms the position of the
internal jugular vein. Next, the needle is withdrawn and the vein
once again entered with a 16 - gauge needle and syringe. Then a
guidewire is placed through the needle and into the jugular vein.
This placement is perhaps the most crucial part of the entire
procedure, and it is vital that the guidewire passes freely without
any resistance whatsoever. Free passage confi rms entrance into
the vein.
Next, the needle is removed with the guidewire left in place.
The incision is widened with a scalpel, and the introducer sheath/
vein dilator apparatus is introduced over the guidewire. During
introduction of the introducer sheath/vein dilator, it is crucial
that the proximal tip of the guidewire be visible at all times, to
avoid inadvertent loss of the guidewire into the central venous
system. The introducer sheath/vein dilator apparatus is advanced
with a slight turning motion along the guidewire. In general, the
point of entry into the vein is felt clearly by a sudden decrease in
resistance. The sheath apparatus then is advanced to the hilt. The
conscious patient is instructed to hold her breath to prevent nega-
tive intrathoracic pressure and air embolism, and the guidewire
and trocar are quickly removed with the sheath left in place.
Occasionally, portable real - time sonography may be helpful in
guiding central venous cannulation [28,29] .
Most current introducer systems contain an accessory port,
which attaches to the proximal end of the introducer sheath and
includes a one - way valve that prevents air introduction into the

central venous system during removal of the guidewire and
trocar. To keep the line open, the sheath then is infused with a
crystalloid solution containing 1 unit of heparin per milliliter and
secured in place with suture.
Insertion of the c atheter
Phase two involves the actual placement of the pulmonary artery
catheter (Figure 16.1 ). Careful attention must be paid to main-
taining sterile technique as the catheter is removed from the
package. The distal and proximal ports are fl ushed to assure
patency. The balloon then is tested with 1 mL of air. When the
catheter has been attached to the physiologic monitor and the air
in any patient. In a recent review article focusing on the use of
this technique in pregnant patients, Fujitani and Baldisseri [25]
concluded “ Invasive monitoring remains useful when the patho-
physiology of critically ill obstetric patients cannot be explained
by non - invasive monitoring, and the patient fails to respond
to conservative medical management; invasive hemodynamic
monitoring may be helpul to guide management. ” As emphasized
by Harvey et al, future studies will need to be adequately powered
and focus on specifi c patient subsets receiving targeted therapies
in order to better defi ne the proper role of this technique in the
management of critically ill patients [26] .
This chapter provides an overview of placement techniques
and complications; indications for the use of this diagnostic tool
in the obstetric patient are examined in more detail in the ensuing
chapters.
Catheter p lacement
The procedure for catheter placement involves two phases. The
initial phase of pulmonary artery catheterization is establishing
venous access with a large - bore sheath. Access is most commonly

obtained via the internal jugular or subclavian veins; however,
under certain circumstances (e.g. where access to the neck or
thoracic region is diffi cult or in a patient with a coagulopathy
where bleeding from a major artery could be hazardous), peri-
pheral veins – including cephalic or femoral – can be used [27] .
Insertion of the introducer sheath via the right internal jugular
vein is described here.
Insertion of the s heath
To catheterize the internal jugular vein, the patient is placed
supine in a mild Trendelenburg position with the head turned to
the left. The landmark for insertion is the junction of the clavicu-
lar and sternal heads of the sternocleidomastoid muscle. When
this junction is indistinct, its identifi cation can be facilitated by
having the patient raise her head slightly. When the landmark has
been identifi ed, 1% lidocaine is infi ltrated into the skin and
superfi cial subcutaneous tissue.
The internal jugular vein is entered fi rst with a fi nder needle,
consisting of a 21 - gauge needle on a 10 - mL syringe. The skin is
RV-paceport lumen hub
(facing infusion)
PA distal
lumen hub
Proximal injectate hub
Thermistor connector
Balloon
inflation valve
RV port
@ 19
cm
Thermistor

Balloon
PA distal lumen
Proximal injectate
port @ 30
cm
Edwards
Figure 16.1 Pulmonary artery catheter.
(Reproduced by permission from American Edwards
Laboratories.)
Pulmonary Artery Catheterization
217
the pulmonary vasculature where the balloon diameter exceeds
that of the corresponding pulmonary arterial branch. At this
point, a wedge tracing is observed. If the balloon is defl ated, the
tracing should return to a pulmonary artery pattern.
Following catheter placement, it is essential that healthcare
personnel skilled in the interpretation of these waveforms con-
tinuously monitor the waveforms for evidence of catheter migra-
tion (spontaneous advancement), which may lead to pulmonary
infarction. This may be manifest by the appearance of a spontane-
ous “ wedge ” tracing at the distal port, rather than the pulmonary
artery waveform, which should be continuously manifest on the
display monitor. Alternately, the appearance of a pulmonary
artery waveform in the central venous pressure port will alert the
attendant to distal catheter migration and the need for adjust-
ment [30] . Komadina et al. described disturbingly high interob-
server variability in the interpretation of waveform tracings,
although agreement on numerical wedge pressure readings was
high [31] . In a similar manner, Iberti et al. reported a wide varia-
tion in the understanding of pulmonary artery catheter wave-

forms and techniques among critical care nurses using this device
[32] . It would appear that graphic recording at end - expiration is
the most reliable means of measuring hemodynamic pressures
[33] . Clearly, continuous training and credentialing programs are
essential for healthcare providers utilizing these techniques.
Recently described digital output volumetric pulmonary artery
catheters have been shown to reduce interoperator interpretation
variability and to improve consistency of treatment decisions.
[34] Normal ranges for hemodynamic parameters in term
pregnancy have been described, and are useful in assessing and
managing the pregnant woman requiring invasive monitoring
techniques [35,36] .
completely fl ushed from the system, minute movements in the
catheter tip should produce corresponding oscillations on the
monitor. The catheter tip is introduced through the sheath and
advanced approximately 20 cm. At this point, the balloon is
infl ated and the catheter advanced through the introducer sheath
into the central venous system. Occasionally, portable real - time
sonography may be helpful in guiding central venous cannulation
[29] .
Waveforms and c atheter p lacement
Once within the superior vena cava, the balloon on the tip of the
catheter will advance with the fl ow of blood into the heart.
Characteristic waveforms and pressures are observed (Figure
16.2 ). Entrance into the right ventricle is signaled by a high
spiking waveform with diastolic pressures near zero. This is the
time of maximum potential complications during catheter place-
ment, because most arrhythmias occur as the catheter tip impinges
on the interventricular septum. For this reason, the catheter must
be advanced rapidly through the right ventricle and into the

pulmonary artery. If premature ventricular contractions occur
during this process and the catheter does not advance promptly
out of the right ventricle, the balloon should be defl ated and the
catheter withdrawn to the right atrium.
As soon as the catheter enters the pulmonary artery, the wave-
form has two notable characteristics. First, and most important,
is the rise in diastolic pressure from that seen in the right ven-
tricle. Second, a notching of the peak systolic waveform often is
seen and represents closure of the pulmonic valve. After entrance
into the pulmonary artery has been confi rmed (in most pregnant
women, this occurs between 40 and 45 cm of catheter length), the
catheter is advanced farther until the tip reaches a point within
30
20
10
0
RA RV
30
20
10
0
R A PA PAO P
Figure 16.2 Pulmonary artery catheter placement.
Catheter tip position, corresponding waveforms, and
normal pressure ranges are demonstrated. (Reproduced by
permission from American Edwards Laboratories.)
Chapter 16
218
access. Such events include pneumothorax and insertion site
infection and occur in 1 – 5% of patients undergoing this proce-

dure [50 – 52] . Potential complications of pulmonary artery cath-
eterization per se include air embolism, thromboembolism,
pulmonary infarction, catheter - related sepsis, direct trauma to
the heart or pulmonary artery, postganglionic Horner ’ s syn-
drome, and catheter entrapment [53 – 58] . Such complications
occur in 1% or less of patients. More recently, a pressure release
balloon has been described to limit overinfl ation and potentially
reduce the risk of vessel rupture [59] . Arrhythmias, consisting
of transient premature ventricular contractions, occur during
catheter insertion in 30 – 50% of patients and are generally of
no clinical consequence.
The remaining complications can be minimized or eliminated
by careful attention to proper insertion maintenance and removal
techniques [37] ). In patients with right - to - left shunts, the use of
this catheter is hazardous; when its placement is deemed manda-
tory, the use of carbon dioxide instead of air for balloon infl ation
may minimize the risk of systemic air embolism [60] . A Food and
Drug Administration task force has summarized recommenda-
tions regarding methods to minimize complications of central
venous catheterization procedures [61] . A recent study suggested
that with proper attention to aseptic technique of placement and
catheter maintainance, a pulmonary artery catheter may be left
in place for up to 7 days before replacement becomes mandatory
[62] .
Numerous studies have documented the frequent discrepancy
between measurements of pulmonary capillary wedge pressure
and central venous pressure during pregnancy [4,63 – 65] . In such
circumstances, clinical use of the central venous pressure would
be misleading. Both techniques entail the risks of obtaining
central venous access, the principal source of complication for

either procedure. For these reasons, in a modern perinatal inten-
sive care unit, central venous monitoring is uncommonly
utilized. Where proper equipment and personnel exist, the vast
amount of additional information obtainable by pulmonary
artery catheterization often outweighs the slight potential increase
in risk attributable to catheter placement itself. When the hemo-
dynamic status of the critically ill pregnant woman is unclear,
pulmonary artery catheterization is nearly always preferable.
Non - i nvasive t echniques
Despite the small risks associated with properly managed pulmo-
nary artery catheterization, the search continues for non - invasive
methods of central hemodynamic assessment of the critically ill
patient. Such techniques generally focus on sonographic or bio-
impedance techniques to estimate cardiac output, and have been
described in both pregnant and non - pregnant patients [66 – 70] .
In addition, investigation continues into techniques to allow non -
invasive central pressure determination [71] . These techniques
appear to be useful in a research setting or in patients requiring
only a single evaluation of hemodynamics in order to classify
Caution also is advised during pulmonary artery catheter
removal; techniques to avoid complications have been described
[37] .
Cardiac o utput d etermination
Once in place, cardiac output is obtained with the use of a cardiac
output computer connected to a terminal on the pulmonary artery
catheter. This instrument derives cardiac output from thermodi-
lution curves created by the injection of cold or room - temperature
saline into the proximal central venous port of the catheter. The
resultant fl ow - related temperature changes detected at the distal
thermistor are converted into cardiac output by the computer and

correlate well in pregnant women with those obtained by the more
precise, but clinically cumbersome, oxygen extraction (Fick) tech-
nique [35] . Nevertheless, it should be emphasized that cardiac
output determinations are of most value in following trends in
individual patients; caution is advised in relying on absolute
cardiac output values, and sound clinical judgment is essential in
data interpretation [38] . One study suggests that the thermodilu-
tion technique may overestimate cardiac output, especially with
very low values [39] . In addition, meticulous attention must be
paid to technique if reliable information regarding cardiac output
is to be obtained. The exact injectate temperature must be known,
the proximal injectate port must have advanced beyond the intro-
ducer sheath, and the introducer sheath sidearm must be closed
[40] . If the central venous port line becomes non - functional,
room - temperature thermodilution cardiac outputs can be used
with saline injection into the sideport, with the understanding that
a slight overestimation of cardiac output will occur [41] . Additional
issues that affect the validity of cardiac output measurements
include the rate of injection, the timing of injection during the
respiratory cycle, the position of the patient, and the presence of
other, concurrent infusions [42] . More recently, techniques have
been evaluated for continuous cardiac output measurement, both
by thermodilution and with the use of a special fl ow - directed
Doppler pulmonary artery catheter [43,44] . Penny et al. [45]
demonstrated that esophageal Doppler monitoring consistently
underestimates cardiac output in patients with pre - eclampsia by
approximately 40%, compared to direct measurements with
pulmonary artery catheters.
With appropriate modifi cation of technique, right ventricular
ejection fraction measurements also may be obtained with the

pulmonary artery catheter [46,47] . Specially designed fi beroptic
catheters allow continuous assessment of mixed venous oxygen
saturation in critically ill patients. Newer techniques for continu-
ous thermodilution measurement compare well with conven-
tional methods [48,49] .
Complications
Most complications encountered in patients undergoing pulmo-
nary artery catheterization are a result of obtaining central venous