Abstract
Conflicting results have been found regarding correlations between right atrial pressure (RAP) and inferior vena cava (IVC) diameter in mechanically ventilated patients. This finding could be related to an increase in intra-abdominal pressure (IAP). This study was designed to clarify whether variations in IVC flow rate caused by positive pressure ventilation are associated with changes in the retrohepatic IVC cross-section (ΔIVC) during major changes in volume status and IAP. Nine pigs were anesthetized, mechanically ventilated and equipped. IAP was set at 0, 15 and 30 mmHg during two conditions, i.e. normovolemia and hypovolemia, generated by blood removal to obtain a mean arterial pressure value lower than 60 mmHg. At each IAP increment, cardiac output, IVC flow and surface area were respectively assessed by flowmeters and transesophageal echocardiography. At normal IAP, even in presence of respiratory changes in IVC flows, no ΔIVC were observed during the two conditions. At high IAP, neither ΔIVC nor modulations of IVC flow were observed whatever the volemic status. The majority of animals with an IVC area of less than 0.65 cm2 showed evidence of IAP greater than RAP values. Negative RAP–IAP pressure gradients were found to occur with an IVC area of less than 0.65 cm2, suggesting that IVC dimensions determined using standard ultrasound techniques may indicate the direction of the RAP–IAP gradient. The clinical relevance of the present findings is that volume status should not be estimated from retrohepatic IVC dimensions in cases of high IAP.
Export citation and abstract BibTeX RIS
Abbreviations
| CO | cardiac output |
| ΔIVC | respiratory changes in the retrohepatic IVC area |
| IAH | intra-abdominal hypertension |
| IAP | intra-abdominal pressure |
| IVC | inferior vena cava |
| L | volume loading |
| MAP | mean arterial pressure |
| RAP | right atrial pressure |
| ROC | receiver operating characteristic |
| VT | tidal volume |
Introduction
Acute elevation of intra-abdominal pressure (IAP) is frequently encountered in critically ill patients who have suffered from 'abdominal' events such as pancreatitis, retroperitoneal hemorrhage (Malbrain et al 2004), and/or edematous bowel secondary to excessive peri-operative fluid resuscitation (Nisanevich et al 2005). Moreover, adverse effects of intra-abdominal hypertension (IAH) on venous return and cardiovascular function in the case of high-volume resuscitation have already been documented (Balogh et al 2003).
Although a great deal of research has been undertaken to elucidate the physiology of the blood circulation, the functioning of the large conduit veins during IAH remains poorly understood (Tyberg 1992). Previous experimental studies in our laboratory have evaluated the relationship between acute IAH and inferior vena cava (IVC) flow and their impact on dynamic indexes of fluid responsiveness in pigs subjected to hemorrhagic shock (Duperret et al 2007, Vivier et al 2006). In these studies, we demonstrated that moderate IAH led to displacement of blood volume toward the thoracic compartment whereas severe IAH results in impaired venous return (Vivier et al 2006). We also demonstrated that IVC respiratory flow fluctuations decline more rapidly with a stepwise increase in IAP under hypovolemia than under normovolemia (Duperret et al 2007). Our findings were in agreement with the concept of the abdominal vascular zones' condition described by Takata et al (1990, 1992), where ventilation related IVC flow modulations are dependent on the right atrial pressure–IAP gradient (RAP–IAP gradient). In short, we demonstrated that an IAP value higher than RAP was the main determinant of venous return and the occurrence of the starling resistor (Duperret et al 2007, Vivier et al 2006).
Conflicting results have been found regarding correlations between RAP and IVC size in mechanically ventilated patients (Bendjelid et al 2002, Jue et al 1992), even if respiratory change in IVC size (ΔIVC) determined by echocardiography is increasingly used to assess preload dependence (Barbier et al 2004, Feissel et al 2004, Michard 2005).
Indeed, Jue et al report a low correlation between IVC diameter and RAP in mechanically ventilated patients (Jue et al 1992), while Bendjelid and colleagues demonstrated that the magnitude of this correlation depends on the echocardiographic method used to measure IVC diameter in these kinds of patients. Since RAP–IAP gradients could affect the retrohepatic IVC size–RAP relationship, this study was designed to clarify whether variations in IVC flow rate caused by positive pressure ventilation are associated with changes in IVC cross-section during major changes in volume status (RAP) and/or IAP.
Methods
Approval from the Ethics Committee for Animal Research of the Claude Bernard University (Lyon, France) was obtained prior to initiating the study. Handling of animals was done according to the guidelines described in the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, 1996). The protocol was designed to study the impact of IAP, hypo- and normovolemia, RAP–IAP gradient and tidal volume positive pressure mechanical ventilation variables on the IVC flow rate and cross-sectional area.
Nine domestic pigs (weight range: 25–30 kg) were studied. Animals were premedicated with ketamine (15 mg kg−1) and anesthetized by propofol injection (1 mg kg−1) followed by continuous infusion of propofol (100 µg kg−1 min−1) and sufentanil (1 µg kg−1 h−1). After tracheal intubation, pigs were mechanically ventilated (Servo ventilator 900 C-Siemens-Elema AB, Solna, Sweden) in a volume-controlled mode with a FiO2 of 0.4, a respiratory rate of 18 breaths min−1, an inspiratory:expiratory ratio of 1:2, an end-expiratory pressure of 0 cmH2O and a tidal volume set to maintain the end-expiratory partial pressure of CO2 within the normal range. This tidal volume was kept constant during the experiment.
General procedure
A fluid-filled catheter was inserted into the right carotid artery to monitor arterial pressure. Another catheter was placed into the right internal jugular vein for fluid and drug administration, and for RAP measurement. We considered its placement adequate when respiratory changes in RAP curves matched with IVC flows. Vascular pressures were measured using calibrated pressure transducers (Honeywell, Zürich, Switzerland) positioned at the level of the left atrium. An 8 cm air-filled latex cylindrical balloon (Marquat, Ref C48 Sde Guenard, Boissy-St-Leger, France) was positioned in the peritoneal cavity via a stab wound to measure abdominal pressure. After median sternotomy and longitudinal pericardiotomy, ultrasound transit-time flow probes were placed around the aortic root and IVC (14 mm A series; Transonic System, Ithaca, NY, USA). The IVC probe was positioned around the vessel just below the junction with the right atrium. We used a probe system with a probe body and a space to receive in a secure but detachable fashion the compliant venous vessel. The pericardium was then partially closed and suspended as a pericardial cradle. Thoracic drains were externally inserted into the pleural space and positioned under water to prevent air from entering the thorax. Pleural pressure was recorded with another air-filled balloon placed in the mediastinal pleural space before closing the chest (Marquat, Ref C48 Sde Guenard, Boissy-St-Leger, France). Using a syringe, the two balloons (mediastinal and abdominal) were filled with air to a volume of 0.5 to 1 ml. The abdomen was then closed and banded with a Velcro belt fixed by three inextensible belts. A large inflatable balloon was placed in between these belts to increase IAP in a progressive and incremental manner. Tracheal pressure was measured and the VT calculated by digital integration of a flow signal measured by a pneumotachograph (Gould Godart, model 17212), the two connected to the endotracheal tube. All pressure and flow signals were recorded with a multi-channel recording system (MP 100; Biopac System, Santa Barbara, CA, USA) and assessed with a Starter system for PC/Windows (AcqKnowledge software, Biopac Systems, Santa Barbara, CA, USA). The data acquired online were stored on a laptop computer for subsequent analysis.
Echographic measurements
Echographic measurements were stored on video tape and retrospectively analyzed by two independent investigators. Transesophageal echocardiography was performed to record the retrohepatic IVC dimensions (Sonos 1500, Philips Medical System, Andover, MA, USA), proximal to the site of hepatic venous efflux. Analyses of static IVC area dimensions at end-expiration and -inspiration and respiratory changes in IVC area (ΔIVC) were performed offline. ΔIVC has been defined as the IVC area at end-inspiration minus the IVC area at end-expiration indexed on the mean of the two measurements. IVC diameters were only measured at baseline IAP, since with increasing IAP, IVC shape changed from a circular to a triangular or elliptic shape (results not analyzed). In two animals, IVC pulsed Doppler was also recorded to compare IVC diameter measurements to IVC Doppler flows at baseline. The present procedure has been added to prove that the venous flow measured by the flow probe technique was matched to the venous flow measured by the Doppler technique.
Experimental protocol for bleeding and reperfusion
In a similar experimental protocol (Vivier et al 2006), a 60 min stabilization period following instrumentation demonstrated perfect stability; a 15 min stabilization period was observed after surgical preparation in the present experimentation. Under steady-state anesthesia and normal IAP (IAP0 mmHg), circulatory, respiratory and echographic variables were recorded. Then, IAP was gradually increased to 15 mmHg (IAP15) and 30 mmHg (IAP30) respectively and new data were recorded at each steady state. Each step lasted 15 min and during the last 5 min all variables were recorded. At an IAP of 30 mmHg, 500 ml of Ringer solution was infused and new data recorded after fluid loading (IAP 30L, normovolemic condition). The balloon was then deflated to decrease IAP to its baseline level. Hypovolemia was subsequently created by blood removal to obtain a mean arterial pressure (MAP) less than 60 mmHg. As most often the arterial pressure was partially restored, a new withdrawal of blood was necessary to stabilize the MAP at 60 mm Hg. Usually, this entailed removing 30% of total blood volume within 5–10 min. After 15 min of stability, the same protocol was realized (hypovolemic condition) increasing IAP0 to IAP15 and then to IAP30, before proceeding to subsequent fluid loading (equal to the blood volume previously removed, IAP 30L, hypovolemic condition).
Statistics
For the statistical analysis, Stata Statistical Software, Release 8.0® (Stata Corporation, College Station, TX, USA), was used. All values are shown as mean ± SD. 3 × 2 × 2 omnibus analysis of variance (ANOVA) or 3 × 2 ANOVA for repeated measures was performed as indicated. A receiver operating characteristic (ROC) curve was also generated to estimate the prediction of a negative RAP–IAP gradient by static IVC area measurements and the area under the ROC curve (±SE) was calculated. All tests were two-tailed and a p value less than 0.05 was considered statistically significant. For further details, please see the online data supplement available at stacks.iop.org/PM/33/615.
Results
Effect of alterations in IAP on circulatory and respiratory variables under normovolemia, hemorrhage (table 1) and fluid loading (figure 1)
Increasing IAP to 15 and 30 mmHg during hemorrhage induces marked changes in MAP, RAP–IAP gradient and cardiac output (CO; figure 1). As shown in figure 1, blood restitution at high IAP was followed by an increase in MAP, RAP–IAP gradient and CO under both volemic conditions, implying that fluid loading increased ventricular preload. At constant VT (13 ± 1 ml kg−1), mean airway and pleural pressures tracked the progressive increase in IAP, reflecting an alteration in thoraco-pulmonary compliance.
Figure 1. Pressure and flow variables as a function of incremental increase in IAP and volume loading for each group of animals: normovolemic (open circles) and hypovolemic (closed circles) (n = 9). Measurements were made during end-expiration. Post-hoc test with Bonferroni correction: *p < 0.05/***p < 0.001 versus IAP (0 mmHg) value within groups. ∓p < 0.05/∓∓∓p < 0.001 versus normovolemic animals. ¥p < 0.05 IAP30 versus 30L. MAP, mean arterial pressure; RAP–IAP, right atrial pressure–intra-abdominal pressure gradient; CO, cardiac output; L: volume loading. Values are means ± SD. Measurements were averaged over five consecutive respiratory cycles. For further details, please see the online data supplement available at stacks.iop.org/PM/33/615.
Download figure:
Standard imageTable 1. Effects of alterations in IAP and blood volume on circulatory and respiratory parameters (n = 9).
| IAP (mmHg) | 0 | 15 | 30 |
|---|---|---|---|
| HR (min–1) | |||
| Normovolemic | 105 ± 28 | 107 ± 27 | 110 ± 19 |
| Hypovolemic | 142 ± 28* | 135 ± 27* | 145 ± 20* |
| RAPm (mmHg) | |||
| Normovolemic | 9.7 ± 3.5 | 16.4 ± 2.9# | 17.6 ± 5.8# |
| Hypovolemic | 7.2 ± 3.6* | 12.4 ± 3.3#* | 13.1 ± 3.4#* |
| RAPm-tm (mmHg) | |||
| Normovolemic | 7.8 ± 3.7 | 12.0 ± 4.8# | 11.0 ± 5.6# |
| Hypovolemic | 4.9 ± 4.3* | 7.0 ± 4.2* | 6.4 ± 4.3* |
| Peak airway pressure (cmH2O) | |||
| Normovolemic | 27.7 ± 3.5 | 43.9 ± 6.9# | 58.3 ± 8.0#° |
| Hypovolemic | 29.5 ± 3.3 | 43.1 ± 8.0# | 56.5 ± 9.3#° |
| Mean airway pressure (cmH2O) | |||
| Normovolemic | 8.5 ± 1.5 | 13.3 ± 2.7# | 17.4 ± 3.4#° |
| Hypovolemic | 9.3 ± 2.0 | 13.3 ± 2.7# | 17.0 ± 3.9#° |
| ΔPleural pressure (mmHg) | |||
| Normovolemic | 4.4 ± 2.3 | 10.1 ± 7.0# | 18.1 ± 10.7#° |
| Hypovolemic | 5.9 ± 4.4 | 10.5 ± 6.4# | 17.7 ± 11.3#° |
| ΔIAP (mmHg) | |||
| Normovolemic | 1.8 ± 0.7 | 4.6 ± 2.2# | 5.9 ± 3.4# |
| Hypovolemic | 1.5 ± 1.0 | 5.3 ± 2.8# | 5.7 ± 4.1# |
Abbreviations: ΔIAP = (maximal inspiratory IAP – minimal expiratory IAP); Δpleural pressure = (maximal inspiratory pleural pressure – minimal expiratory pleural pressure); HR = heart rate; RAPm = mean right atrial pressure; RAPm-tm = transmural mean right atrial pressure; # p < 0.05 versus IAP0; ° p < 0.05 versus IAP15; * p < 0.05 versus normovolemic.
Effects of gradual increase in IAP and decrease in volemia on IVC size and its respiratory variations (figures 2 and 3)
At IAP15 and IAP30, and both during inspiration and expiration, retrohepatic IVC areas were reduced. End-expiratory areas were smaller under hypovolemia than under normovolemia, at IAP0 and IAP15: 1.37 ± 0.43 versus 1.78 ± 0.35 cm2 and 0.33 ± 0.29 versus 1.08 ± 0.73 cm2, respectively (p < 0.05). At IAP30, the difference in vessel areas between the two volemic states did not reach statistical significance. IVC area reduction was influenced by the combined effects of IAH and volemia. Indeed, a mean IVC area of less than 1 cm2 was observed at IAP15 under hypovolemia while the same value was experienced at IAP30 under normovolemia. Absolute and relative respiratory variations in IVC area were limited, whatever the IAP level or the volemic status.
Effect of gradual increase in IAP and decrease in volemia on respiratory changes in IVC flow (figures 2–4)
IVC respiratory flow fluctuations were dependent on both volemia and IAP. A clear difference was observed between the effect of a breath cycle at IAP15 and at IAP30. Indeed, in normovolemic animals, at IAP0 and IAP15, IVC flow fluctuated within the breath, while no fluctuation was observed at IAP30 (figures 2 and 3). In contrast, following hemorrhage, IVC flow significantly fluctuated within the breath only at IAP0, which infers that IVC flow modulations related to ventilation decrease more rapidly with the progressive increase in IAP during hypovolemia (figure 3). As we were not able to explore echographically the shape of the juxta-diaphragmatic portion of the IVC where the flowmeter was placed, we also measured IVC venous flow at the level of the retrohepatic IVC echo measurement using pulsed Doppler in two pigs. At this retrohepatic point, IVC flow varied similarly to that of the ultrasound transit-time flow probe (figure 4).
Figure 2. Recordings of airway pressure (Paw) and IVC flow, over three breath cycles at baseline IAP (top) and IAP 30 mmHg (bottom) and the corresponding expiratory IVC areas (right) in a normovolemic animal. The absence of IVC flow fluctuations at IAP30 suggests the presence of a 'vascular waterfall' zone condition in presence of a static collapsed IVC area. Note the increase in airway pressure in presence of elevated IAP and the high frequency changes of IVC flow related to heart beats. Also, see that IVC vessel is surrounded by the liver.
Download figure:
Standard image
Figure 3. Effects of cyclic ventilation, gradual increase in IAP and volume loading on IVC flow and area in normovolemic and hypovolemic animals (n = 9). The white and shaded boxes represent expiratory and inspiratory periods, respectively. Post-hoc test with Bonferroni correction: *p < 0.05/***p < 0.001 inspiratory versus expiratory values. ∓p < 0.05 versus normovolemic animals. See the absence of respiratory change in IVC areas in presence of changes in IVC flows. IAP30L: IAP30 after fluid loading. For further details, please see the online data supplement available at stacks.iop.org/PM/33/615.
Download figure:
Standard image
Figure 4. Compiled recordings (from top to bottom) of airway pressure (Paw, white curve), M-mode IVC on 2D-echographic short axis view, M-mode IVC ECG, IVC flows (IVC flow by the ultrasound transit-time probe, white curve; IVC flow by Doppler (deferred recording), gray curve) and the ECG of the Doppler flow, synchronized during one breath cycle in a normovolemic animal at IAP0. First, note the lack of marked IVC shape modification during cardiac beats in presence of high frequency changes of IVC flow. Second, observe the marked change in IVC flow (measured by ultrasound transit-time probes and Doppler) with mechanical breath in presence of minor changes in IVC diameter (arrows).
Download figure:
Standard imageEffects of mechanical breath on blood volume shifted and sequestered in the splanchnic compartment at baseline IAP0 (figure 5)
The transient thoracic blood volume pooled upstream in the splanchnic compartment during each mechanical insufflation was also considered. The blood volume shifted and sequestered was calculated based on the fact that in pigs with low IAP, positive pressure ventilation causes an increase in RAP and a decrease in the driving pressure of venous return with a subsequent fall in IVC flow. We calculated tidal changes in the total blood volume within the splanchnic/hepatic vascular compartment over consecutive breaths by the addition of the successive beat to beat volume decreases (integration of the flow signal), beginning with the maximal volume until any new increase was reached (figure 5). We found that the blood volume shifted in hypovolemic pigs was quite similar to that shifted in normovolemic pigs, 25 ± 12 versus 29 ± 9 ml, respectively, (figure 5).
Figure 5. Example of calculation of blood volume pooled upstream in the splanchnic compartment during insufflation in a normovolemic animal. Calculation was based on the fact that mechanical breath decreases IVC flow and shifts blood volume toward the abdominal compartment. Over five consecutive breaths, the total volume stored (gray column) was calculated by adding the successive beat to beat volume decreases (white column), beginning with the maximal volume, until any subsequent increase was reached. X-axis = time line.
Download figure:
Standard imageImpact of the RAP–IAP gradient on IVC area and flow (figure 6)
We simultaneously plotted IVC echographic area against the RAP–IAP gradient for all animals (figure 6). An IVC area smaller than 0.65 cm2 (best threshold value) predicted a negative gradient between RAP and IAP with a sensitivity and specificity of 84% and 84%, respectively. The area under the ROC curve to predict a negative gradient was 0.92 (95% CI 0.88–0.96). This threshold value of 0.65 cm2 corresponded to a 62% decrease in baseline mean IVC area. A similar exponential relationship was found between IVC echographic areas and RAP–IAP gradients, whatever the circulatory state or the inhalation–exhalation periods (figure 7).
Figure 6. Bivariate scatter-plot of inspiratory and expiratory IVC area and flow values and the RAP–IAP gradient for all animals (normovolemic and hypovolemic) presenting different abdominal pressure values (n = 197 for IVC area). Horizontal dotted line: best IVC area threshold value (see the text). Vertical dotted line: 0 mmHg RAP–IAP gradient. IVC: inferior vena cava. RAP–IAP: right atrial pressure–intra-abdominal pressure gradient (n = 9).
Download figure:
Standard image
Figure 7. The correlation between IVC area and RAP–IAP gradient throughout various circulatory states and respiratory phases. Horizontal dotted line: best IVC area threshold value (see the text). Vertical dotted line: 0 mmHg RAP–IAP gradient. IVC: inferior vena cava. RAP–IAP: right atrial pressure–intra-abdominal pressure gradient.
Download figure:
Standard imageDiscussion
In the present model of sedated, mechanically ventilated pigs, the results enabled us to make two assertions. First, the presence of IVC respiratory flow variations are not linked to changes in shape of the retrohepatic portion of this vessel. Second, the majority of animals exhibiting an inspiratory and expiratory IVC area of less than 0.65 cm2 have IAP values greater than RAP, suggesting that measuring IVC dimensions could be a means to predict a negative RAP–IAP gradient and the probability of IAH.
Using an IVC transit-time flow probe and transesophageal echography, we demonstrated that the RAP–IAP gradient affects retrohepatic IVC size and flow fluctuations (figure 6). However, the finding of a lack of respiratory change in IVC area, especially during hypovolemia, in presence of positive pressure-induced IVC flow fluctuations, is the key point of the present paper, particularly at IAP0 where IVC shape is not influenced by surrounding pressure (figure 3) (Takata et al (1992)). Hence, in contrast to what has been demonstrated in critically ill septic patients, our experiment in this porcine model of hemorrhagic shock indicates that lack of retrohepatic IVC respiratory shape oscillations does not consistently indicate a normovolemic state.
This unexpected phenomenon of respiratory modulation of IVC flow at low IAP in absence of respiratory variations of vessel shape can be attributed to low venous compliance of the retrohepatic IVC. Indeed, the respiratory-induced IVC flow fluctuations indicate that a given amount of blood has been held back in the abdominal compartment during inspiration, and then released into the thoracic compartment at the onset of expiration (van den Berg et al 2002). This abdominal blood volume change should have occurred in more compliant vascular zones than that of the retrohepatic IVC. This result was further confirmed by the lack of IVC shape change with cardiac beats in presence of high frequency changes of IVC flow (related to heart beats), and hence the low compliance of this vessel segment (figure 4).
As in previous works reporting respiratory-induced IVC diameter changes in hypovolemic mechanically ventilated patients (Barbier et al 2004, Feissel et al 2004), we computed the amount of blood volume redistributed toward the splanchnic compartment at the expense of the thoracic compartment during mechanical insufflations (van den Berg et al 2002) in normo- and hypovolemic animals submitted to low IAP (figure 5). This volume was found to be quite similar in both groups. This result confirms that the change in IVC flow was the same for a same backward pressure change induced by the insufflation, whatever the volemic status. Incidentally and according to the pressure–flow relationship of the IVC, this means that IVC conductance (1/resistance) was the same in the two volemic states. However, some differences between the two publications cited above (Barbier et al 2004, Feissel et al 2004) and our experiment must be noted. Firstly, their measurements were carried out in septic patients, whereas our results were acquired in an animal model of hemorrhagic hypovolemia; it could therefore be hypothesized that the venous capacitance was quite different in the two situations. Secondly, in order to avoid any effects of IAP on IVC shape, and the effect of ultrasound beam deviation on IVC measurement, we chose to measure the cross-sectional area of the retrohepatic IVC instead of the diameter (Bendjelid et al 2002), and to perform transesophageal- instead of transthoracic-echography. Thirdly, in our experimental study, we also used the transesophageal echographic approach to avoid the IVC compression by the probe during forced positive pressure ventilation. If a transthoracic subcostal view is used, problems could arise in measuring absolute superior–inferior and anterior–posterior distances, as the ultrasound probe simultaneously compresses the abdomen to diaphragm and creates an abdominal wall shift (Kocis et al 1997).
Another finding of this work is that the majority of animals exhibiting an IVC area of less than 0.65 cm2 show evidence of IAP values greater than RAP values (figure 6). The present result suggests that IVC dimensions determined using echography may indicate the direction of the RAP–IAP gradient with the clinical relevance that a negative gradient implies that RAP should not be estimated from IVC size. This finding could explain the conflicting results found regarding correlations between RAP and IVC size in previous studies (Bendjelid et al 2002, Jue et al 1992), as IAP is another determinant of this relationship. Thus, in cases of high IAP, RAP should not be estimated from IVC size. Moreover, IVC measurement could also confound the noninvasive assessment of pulmonary arterial pressure using Doppler-echocardiography (Bendjelid et al 2002, Jue et al 1992).
This study has some limitations. As already discussed above, we investigated the retrohepatic IVC and could not ascertain that vessel collapse does not occur caudally to this point as already demonstrated by Lambert et al in pigs (Lambert et al 2007) and Jardin et al in humans (Jardin et al 1982). Indeed, as shown in figure 2, we investigated an IVC segment surrounded by the liver, a fact that could have reduced IVC compliance. We chose this location of the vessel for two reasons: first, this site was the IVC portion investigated in critically ill patients (Barbier et al 2004, Feissel et al 2004) and second, we were not able to explore the juxta-diaphragmatic position of the vein as the flowmeter was placed here. However, since we also measured IVC venous flow at the level of the retrohepatic IVC echographic area measurement using pulsed Doppler in two pigs, we can guarantee that at this point, IVC flow varied similarly to that of the ultrasound transit-time flow probe (figure 4).
Conclusions
In contrast with previous studies, our results show that variations in IVC flow rate caused by positive pressure ventilation are not associated with a change in the retrohepatic IVC cross-section and are unaffected by volume status or IAP. Negative RAP–IAP pressure gradients were found to occur with an IVC area of less than 0.65 cm2, suggesting that IVC dimensions determined using standard ultrasound techniques may indicate the direction of the RAP–IAP gradient. The clinical relevance of the present findings is that volume status should not be estimated from retrohepatic IVC dimensions in cases of high IAP.
Acknowledgment
The authors are grateful for the translation support provided by Dr Dominique Vala, Cardiovascular Surgery Division, Geneva University Hospitals (CA).