24], both involving spontaneously breathing patients, have shown that an inspira-
tory fall in CVP by ≥ 1 mmHg is highly predictive of a fluid responsive cardiac
index (positive predictive value 77%/84% and negative predictive value
81%/93%).
Although the CVP in the supine patient is a poor index of circulating volume
postural changes in CVP may be a more reliable indicator of intravascular volume
status [4]. Measurement of postural changes in CVP seems, however, unlikely to
106 T. Smith, R. M. Grounds, and A. Rhodes
become a widely adopted clinical tool within the context of the acutely or critically
ill patient in the ICU.
Utility of CVP as a Measure of Circulatory Filling
Central venous pressure can without doubt be affected by the intravascular vol-
ume. Approximately two thirds of the intravascular volume is contained in the
venous system and the total intravascular volume will affect the mean venous
pressure. Only a proportion of the total intravascular volume exerts any distend-
ing force on the vasculature [25] thereby causing a positive pressure within the
vasculature; this volume cannot be measured in the intact human and will vary
with the vascular tone which is therefore also an important determinant of the
CVP. The volume of blood in the central veins will also be affected by the distribu-
tion of the venous blood volume through the venous system: peripheral venocon-
striction and the effects of the muscle pump will redistribute volume from the
peripheral veins to the central veins and so increase CVP whereas peripheral
vasodilatation and upright posture will redistribute volume to the peripheral
venous compartment and decrease the CVP. Furthermore, the CVP depends not
only on the volume of blood in the central venous system but on the compliance of
that system. With so many factors other than intravascular volume affecting the
CVP one might expect that CVP would be a relatively inexact measure of intravas-
cular volume particularly in the intact organism where feedback mechanisms will
compensate for a decreased intravascular volume by stimulating vasoconstric-
tion. This expectation is borne out in clinical studies where not only has CVP been
shown to correlate poorly with blood volumes measured by indicator dilution but
the change in CVP after fluid resuscitation of shocked patients also correlated
poorly with the measured change in blood volume [26, 27]. CVP has also been
found to correlate poorly with the volume of fluid administered during ENT
surgery in spite of a progressive decrease in hematocrit during surgery suggesting
intravascular volume expansion [28].
Clinical Outcomes and CVP monitoring
Considering the paucity of data to support CVP as a useful physiological monitor
one would not expect CVP monitoring to have a significant positive effect on
outcome. There are relatively few studies that examine this issue particularly in
the critically ill, presumably because CVP monitoring has become an almost
routine part of ICU care.
Fluid administration targeted by CVP monitoring during hip surgery shortened
the time before patients were medically fit for discharge [29]. However, similar
results were obtained using Doppler flow monitoring to guide fluid administration
and it might be suggested that similar results in both groups could have been
achieved by simply giving larger volumes of fluid without additional monitoring.
In another study, fluid administration aiming to keep the CVP greater than 5
mmHg during renal transplant surgery resulted in a greater frequency of graft
Central Venous Pressure: Uses and Limitations 107
function within the first three postoperative days than in a control group without
CVP monitoring [30]. Whilst these studies probably demonstrate an important use
of CVP monitoring in detecting low circulating volumes in surgical patients which
when detected can be appropriately managed and thus lead to improved outcome
it is doubtful what bearing they have in critically ill patients where more usually
the CVP is relatively high and the question is whether fluid or vasoactive drugs
should be the next intervention.
In some circumstances CVP monitoring may provide prognostic information.
A CVP of > 15 mmHg after cardiac surgery is a significant predictor of poor
outcome [31].
Of more relevance to ICU medicine, the decrease in cardiac output in response
to an increase in PEEP (from 0 to 30 mmHg) correlates with the initial level of CVP
and patients with an initial CVP of ≤ 10 mmHg experience a greater fall in cardiac
index than patients with CVP >10 mmHg (–30% +/– 9 vs. –8% +/– 7) [32].
Maintaining a CVP of >10 mmHg may therefore be desirable in the ventilated
patient. Surprisingly the inspiratory decrease in CVP appears unable to predict the
cardiovascular response to PEEP in a similar way [33].
When considering the utility of CVP monitoring it is appropriate to make the
analysis in the context of other possible modalities of monitoring available to
measure similar physiological variables. The most common alternative to CVP
monitoring as an index of cardiac preload and volume status is pulmonary artery
pressure monitoring using a pulmonary artery catheter (PAC). The use of PACs
has been associated with greater morbidity and cost than the use of central venous
catheters and a number of studies have suggested that in many cases they do not
offer any advantages over CVP monitoring, particularly in low risk surgical patients
[34] and may in fact worsen outcome increasing both the complication rate and
time spent intubated after cardiac surgery [35]. An examination of the utility of
PACs as an alternative to central venous catheters is outside the scope of this
chapter but it is to be hoped that a clear answer to this question will be given by the
large multicenter study currently underway.
Perhaps the most powerful studies indicating the usefulness of CVP monitoring,
or lack thereof, in critical care are those involving goal directed therapy. One such
study in septic patients showed no difference in outcome between patients with
CVP or PAC monitoring where therapy was directed towards achieving normal
values of measured variables; however, in those patients where therapy was di-
rected to achieving supraphysiological values for cardiac index and oxygen delivery
an improved outcome was seen [36]. Clearly such goal directed therapy requires
monitoring other than simple CVP monitoring. Similarly, early goal directed
therapy of septic patients in the emergency department resulted in significantly
improved outcomes when therapy was directed at improving mixed venous satu-
rations rather than at normalizing the CVP, mean arterial pressure (MAP) and
urine output [37].
108 T. Smith, R. M. Grounds, and A. Rhodes
Conclusion
The two clinical studies on surgical patients [29, 30] confirm the potential utility
of CVP monitoring in some patient groups. As a decrease in CVP is a relatively late
sign of intravascular volume depletion in a patient with intact vasoconstrictor
reflexes it is possible that in the patient groups in these two studies there is a
significant risk of severe hypovolemia which would, if not detected by CVP moni-
toring, remain untreated causing increased morbidity. It may, however, be argued
that CVP may be a better measure of volume status in anesthetized patients whose
vasoconstrictor reflexes are pharmacologically impaired by the anesthetic drugs.
There is no convincing evidence that CVP monitoring improves outcome in the
critically ill patient, particularly when other variables are being assessed. Addition-
ally, it is clear from studies examining goal directed therapy that targeting fluid
therapy to normalizing the CVP in a critically ill patient is not an optimal treatment
strategy.
There is no doubt that there is a significant morbidity and possibly even
mortality associated with obtaining central venous access; central cannulation
having a complication rate of up to 6% even when performed by experienced staff
[38]. This risk may outweigh the risk of giving large volumes of fluid without central
pressure monitoring in the general surgical population. However, the majority of
critically ill patients require central venous access for the administration of drugs
or potassium and there appears to be some potential advantage in measuring
central venous oxygen saturation at least during the early stages of treatment for
which central access is also required. If central venous access is to be obtained then
it would seem appropriate to monitor the CVP. As long as this variable is consid-
ered in the context of the whole clinical picture and other monitored and laboratory
variables and the underlying pathophysiology taken into account then it is unlikely
that CVP monitoring will lead to a worsened outcome and there are some situations
such as a large occult blood loss or extreme vasodilatation where a change in CVP
may provide an early warning of the problem.
The role for CVP as a monitor for use in the cardiovascular optimization of
critically ill patients remains important largely because most critically ill patients
will require central venous access for other reasons and so monitoring the CVP
becomes essentially a risk free procedure as the risks are associated with obtaining
access rather than the monitoring process itself. However, as a monitor it has
significant weaknesses and with the increasing availability of other less invasive
and apparently better measures of preload and circulatory filling the importance
of CVP monitoring is likely to decline in this context, at least within the critical care
setting, although it may be some time before other preload monitors are available
on general wards in our hospitals.
References
Bạn đang xem 24] - UPDATE IN INTENSIVE CARE AND EMERGENCY MEDICINE PART 3 POTX