| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
 
|
|
|
||
|
|
© 2001 International Anesthesia Research Society
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Neurosurgical patients undergoing either craniotomy or complex spine surgery are subject to wide variations in blood volume and vascular tone. The ratio of these variables yields a pressure that is traditionally mea-sured at the superior vena cava and referred to as "central venous pressure" (CVP). We have investigated an alternative to CVP by measuring peripheral venous pressure (PVP), which, in parallel animal studies, correlates highly with changes in absolute blood volume (r = 0.997). We tested the hypothesis that PVP trends parallel CVP trends and that their relationship is independent of patient position. We also tested and confirmed the hypothesis, during planned circulatory arrest, that PVP approximates mean systemic pressure (circulatory arrest pressure), which reflects volume status independent of cardiac function. PVP was compared with CVP across 1026 paired measurements in 15 patients undergoing either craniotomy (supine, n = 8) or complex spine surgery (prone, n = 7). Repeated-measures analysis of variance indicated a highly significant relationship between PVP and CVP (P < 0.001), with a Pearson correlation coefficient of 0.82. The correlation was best in cases with significant blood loss (estimated blood loss >1000 mL; r = 0.885) or hemodynamic instability (standard deviation of CVP > 2; r = 0.923).
Implications: In patients undergoing either elective craniotomy or complex spine surgery, peripheral venous pressure (PVP) trends correlated with central venous pressure (CVP) trends with a mean offset of 3 mm Hg (PVP > CVP). PVP trends provided equivalent physiological information to CVP trends in this subset of patients, especially during periods of hemodynamic instability. In addition, measurements made during a planned circulatory arrest support the hypothesis that PVP approximates mean systemic pressure (systemic arrest pressure), which is a direct index of patient volume status independent of cardiac or respiratory activity.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Neurosurgical patients undergoing either craniotomy or complex spine surgery are at risk for significant blood loss and hemodynamic changes while under anesthesia. A portion of that risk is associated with the surgery itself, and the remainder is because of the effects of anesthetics, mechanical ventilation, and positional changes on cardiac function, relative blood volume, and vascular tone. For patients having procedures with a likelihood of significant blood loss, or for patients in whom hemodynamic changes are especially threatening because of coincident cardiovascular disease, central venous pressure (CVP) monitoring has been widely used since the introduction of percutaneous jugular vein cannulation in 1969 (1). CVP is a direct measure of pressure, not volume. However, pressure is assumed to reflect the ratio of blood volume to vascular compliance, and it is a way to assess "relative blood volume" in patients who are either critically ill or undergoing major surgery (2). In this capacity, and in the normal range of hemodynamic values, CVP appears to have more utility as a trend monitor than for titration to an absolute end point (3,4).
The measurement of CVP requires the insertion and maintenance of a catheter in the thorax. Typically, the catheter is positioned in the superior vena cava from an insertion site in an external or internal jugular vein, subclavian vein, or, less often, from a femoral or antecubital vein (5,6). A small but potentially serious risk accompanies every CVP placement (7–11), in addition to adding time and monetary costs to surgery.
This study specifically addresses an alternative to the CVP as a physiological volume monitor. Peripheral venous pressure (PVP) reflects an "upstream" venous variable that is coupled to CVP by a continuous column of blood, analogous to the fluid continuity that exists between a pulmonary artery occlusion catheter and the left atrium. We tested the hypothesis that PVP values are easily obtained, free from local occlusion artifact, and closely correlated with CVP trends in a variety of major neurosurgical procedures, patient positions, and degrees of hemodynamic instability. We also tested the hypothesis that PVP reflects mean systemic pressure (circulatory arrest pressure), which is the unique pressure to which all vascular beds converge in the absence of cardiac function. Because mean systemic pressure is defined as the simple ratio of blood volume to vascular compliance, it is a uniquely useful variable for volume assessment, but has not been accessible to clinical measurement techniques during intact circulatory function.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
IRB approval was obtained for an observational study without the requirement for written consent to evaluate hemodynamic data from patients undergoing either craniotomy or complex spine surgery using existing central and peripheral venous lines. Criteria for inclusion included presentation for either craniotomy or complex spine instrumentation surgery, acceptable access for central venous catheter placement, and acceptable access for customary peripheral IV placement. One patient was excluded from the study on the basis of inadequate peripheral IV access.
Hemodynamic data were recorded beginning with placement and calibration of CVP and PVP catheters and positioning of the patient for surgery, and ending at the conclusion of surgery. Real-time waveforms were displayed throughout the case for both CVP and PVP, and numeric pressure values were retained by the same monitoring system ( GE Marquette 9500 Patient Monitor with Tram 350 Module ; GE Marquette, Milwaukee, WI). Central venous catheters (either 7F 2-lumen or 12F 3-lumen, 16 cm., Arrow International, Reading, PA) were placed through either the left or right internal jugular approach. Peripheral IV catheters (either 18-gauge or 20-gauge, 1.16 in., B-D AngiocathTM , Becton Dickinson Infusion Therapy Systems, Inc., Sandy, UT) were placed, or used in situ, from dorsal hand or distal forearm veins. The PVP catheter site was maintained at midthorax height throughout each case. After flushing and room-air zero calibration ( Becton-Dickinson Transducer Sets ; Becton-Dickinson Critical Care Systems Pte. Ltd., Singapore), the transducer sets were mechanically flushed with saline and also maintained at midthorax level throughout surgery. Continuity of the PVP catheter with the downstream venous system was demonstrated at the beginning of each case by observing coincident pressure changes in the PVP waveform during circumferential, proximal arm occlusion. All hemodynamic data were recorded to the nearest 1 mm Hg at 5-min intervals throughout the recording period.
No alterations were made in the usual anesthetic technique for the cases. Anesthesia was induced with IV sodium thiopental or propofol, and maintained by nitrous oxide/oxygen with isoflurane; or by IV propofol and remifentanil infusion with nitrous oxide/oxygen and small-dose isoflurane. Muscle relaxation was maintained by intermittent IV doses of either pancuronium or rocuronium, except in two spine surgery cases (4 and 10) where electromyography was used to monitor spinal cord integrity and no muscle relaxant was administered after an intubating dose of succinylcholine. All patients were intubated and mechanically ventilated throughout surgery. Arterial pressure was monitored with a standard 20-gauge radial arterial line throughout surgery. Decisions regarding IV fluid administration, blood or blood product administration, and treatment with vasoactive medications were made according to the usual judgement of each staff anesthesiologist without regard to PVP readings.
One adult patient who underwent repair of a giant basilar artery aneurysm under deep hypothermic circulatory arrest was studied. Simultaneous measurements of radial arterial pressure (ART), pulmonary arterial pressure (PAP), CVP, and PVP were made throughout the case. After systemic heparinization and cannulation of the right femoral artery and vein, the patient was placed on cardiopulmonary bypass (CPB) and cooled to a target temperature of 18°C with a chilled bypass circuit. After temperature stabilization, CPB was abruptly stopped and the femoral venous cannula connecting the patient to the venous reservoir was simultaneously occluded. Real-time continuous pressure measurements of ART, PAP, PVP, and CVP were recorded before, during, and after circulatory arrest.
The hemodynamic data are reported as single values (in mm Hg, referenced to midthorax level), rounded to the nearest 1 mm Hg unless otherwise specified. Time intervals are reported as minutes after the initiation of recording. All venous pressures were measured at 5-min intervals throughout the surgical procedure. To test for a significant relationship, PVP and CVP trends were compared by repeated-measures analysis of variance. PVP and CVP trends were also compared by simple regression analysis to obtain Pearson correlation coefficients, and reanalyzed according to significant estimated blood loss (defined as >1000 mL), and hemodynamic instability (defined as SD of CVP >2). Scatter plots of CVP versus PVP and mean arterial pressure (MAP) versus PVP were made using all 1026 respective paired measurements. Because PVP was expected to have an overall physiological offset more than simultaneous CVP measurements (and, therefore, a physiological "bias"), and because our hypothesis concerned trend correlation rather than equivalence of absolute values, we applied statistical tests of trend correlation rather than of bias.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Table 1 summarizes the demographic and hemodynamic features of each case. Patient ages ranged from 16 to 76 yr, and ASA physical status ranged from II-IV. Estimated surgical blood loss ranged from 200 mL to 8500 mL. All 15 surgeries were considered "major," either craniotomy for tumor or vascular malformation, or complex spinal instrumentation surgery. Patient positioning was supine for craniotomy (n = 8), or prone for spine surgery (n = 6). Two cases (repair of scoliosis, 10; complex spine surgery, 15) were conducted in two immediately sequential phases during the same operation: lateral (Case 10), or supine (Case 15) for anterior release, then prone for posterior spinal instrumentation. Hemodynamic monitoring continued throughout both phases and positions.
|
PVP was easily obtained in all 15 patients, in supine, lateral, and prone positions. Figure 1 displays a 24-s recording of simultaneous PVP and CVP in the same patient. Panel A shows uninterrupted waveforms, and panel B demonstrates a proximal arm occlusion and release that was used in each case to test for the continuity of measured PVP with the downstream venous system. The coincidence of an abrupt pressure change in the PVP waveform when the arm was occluded or released at a site far removed from the PVP catheter site demonstrates this continuity. The CVP tracing reveals higher amplitude respiratory variations, as well as characteristic a, c, and v-wave excursions. In no case was "wall-artifact" encountered, or developed, after initial calibration and positioning of the recorded limb. Panel C demonstrates the range of PVP–CVP gradient among three separate cases and reveals the relatively increased degree of respiratory variation seen in CVP tracings compared with PVP tracings.
|
Simultaneous PVP and CVP tracings for each of the 15 cases are displayed in Figure 2. In most cases, an offset (PVP > CVP) is evident, but the degree of offset tended to remain relatively constant throughout each case. Periods of hemodynamic instability unmasked a tight dynamic correlation between the two variables. In contrast, cases with little hemodynamic (CVP) instability (for example, 5 and 8) also revealed little change in PVP, and the correlation coefficient was predictably small in those cases. Repeated-measures analysis of variance revealed a significant relationship between PVP and CVP (P < 0.001). Table 2 displays statistical tests of correlation for individual cases, as well as the overall correlation from 1026 paired mea-surements (r = 0.82). Tests of correlation for cases with large estimated blood loss (>1000 mL; r = 0.885), and large CVP variability (SD CVP > 2; r = 0.923) revealed essentially the same result as is visually evident from Figure 2: where CVP trends changed rapidly, so did PVP trends, and in synchrony with each other. The mean CVP value in these patients was 10 mm Hg and the mean PVP value was 13 mm Hg, with an overall difference of 3 mm Hg. Our data revealed no consistent relationship between absolute PVP values, or the PVP–CVP gradient, and catheter site (data not shown) or position (Table 2). Similarly, no consistent relationship was evident between type of surgery, anesthetic technique, or patient position and the PVP–CVP gradient.
|
|
We counted the individual sampling intervals that demonstrated either a movement of PVP and CVP in opposite directions (7 out of 921), or movement of one variable by >2 mm Hg without a corresponding movement in the other variable (6 out of 921). The cumulative frequency of either of these events was 13/921 (1.4%).
Figure 3A reveals the relationship between CVP and PVP in a scatter plot that includes all 1026 data points. Because of the large number of measurements, each point on the graph represents at least one, but may represent more than one, paired measurement. In contrast, the scatter plot of MAP versus PVP in Figure 3B reveals no consistent relationship between the two variables. This is also graphically evident in each case where MAP trends are included along with PVP and CVP trends (representative example in Fig. 3C).
|
Figure 4 displays data before CPB (A), immediately before and during (B), and after (C, D) circulatory arrest. Compared with each of the other hemodynamic variables (ART, PAP, and CVP), PVP revealed a more stable waveform, and was the only pressure that remained constant through the arrest transition. After arrest, ART and PAP decreased, and CVP increased to converge on an essentially unchanging PVP. That pattern defined PVP as equivalent to mean systemic pressure for the case. It is not clear why the arterial pressure plateaued briefly before converging to PVP, but this deviation from a simple monoexponential decay is characteristic of other circulatory arrests in our institution (data not shown). Panel 4E displays hemodynamic data from a separate patient with intact circulatory function. During each diastolic runoff, ART decreased and CVP trended upward (arrows), but PVP remained stable, revealing the same tendency toward pressure convergence to PVP as during circulatory arrest, but interrupted by the next systole.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
It is not surprising that PVP and CVP trends are linked, given that the two sites of measurement are part of the same venous continuum. The demonstration of that relationship in surgical patients raises several questions, however. First, why has the coupling of PVP to CVP not been noticed and taken advantage of before? The expectation that PVP and CVP are related is not new. Since Holt’s (12) demonstration in 1943 that the pressure in the antecubital vein tracks intrathoracic pressures in spontaneously breathing patients, several studies have attempted to define that coupling as a diagnostic tool (13). With the availability of cardiac catheterization techniques in 1952, Pedersen (14) measured the pressure gradients between the antecubital vein and the right atrium in 18 spontaneously breathing adults. He found that the gradient varied between 0.7 and 5.8 cm H2O, with a mean value of 2.6 cm H2O. He also noted that the respiratory and cardiac excursions evident in the right atrium were largely absent in the periphery. Those observations are in general accord with our own, and precede ours by a half-century.
Each of the early studies that examined PVP-CVP gradients revealed a significant variability between subjects. Given that variability, and the primitive techniques for hemodynamic measurement in that era, it is not surprising that PVP measurement never matured into a clinically useful tool. The claim by Sykes (15), in 1963, that measurement of CVP was a useful adjunct to volume resuscitation was criticized for being technically unwieldy, and never widely adopted. Even in 1970, when Eustace (16) explored the relationship between PVP and CVP in surgical patients, the use of fluid manometers, widely separated measurement points, and an inability to separate mean pressure values from maximum and minimum manometric excursions handicapped any effort to define a dynamic relationship between PVP and CVP.
More recently, we have demonstrated that PVP correlates to blood volume changes in an animal model (r = 0.997 for PVP compared with hemorrhage volume; unpublished data). This result is consistent with the observations of another group (17–20), which used a retrograde venous balloon-occlusion catheter in anesthetized dogs during either hemorrhage or volume loading to examine the relationship between "postcapillary occlusion pressure" and volume status. The result is also consistent with reports of single-point comparisons between PVP and CVP during surgery (21), and in the postanesthesia care unit (22). For this clinical study, we chose to use standard peripheral IV catheters. Despite its simplicity, the technique of measuring PVP from existing or standard arm vein catheters is apparently free from vessel wall or occlusion artifact, and lends easy accessibility to the technique.
A second question raised by our results is whether the measurement of PVP reveals local (peripheral vein), rather than systemic, physiological information. On an empirical level, the correlation between PVP and CVP argues that PVP reflects a systemic phenomenon. At that same empiric level, we chose to study a wide variety of patient and arm positions, surgeries, blood loss and hemodynamic profiles, and catheter sites. If the measurement of PVP was prejudiced by local venous tone and position variables, it is unlikely that such a robust coupling to CVP trends would have been demonstrated. We also know from clinical experience that, in the absence of extravasation or obstruction by clotting, IV catheters continue to flow unimpeded into the central circulation throughout surgery. This implies that fluid continuity is maintained between PVP and CVP sites despite changes in venous geometry that may occur with repositioning, and despite any venous valves that may intervene between the PVP site and the central circulation. Such valves are, by definition, open during steady state venous flow, and should therefore not disrupt fluid continuity between the two sites.
At a physiological level, the PVP reflects "downstream" pressures in the right atrium analogous to the pulmonary artery occlusion catheter reflecting pressure phenomenon in the left atrium. Because the PVP site is postcapillary, vessel occlusion is not required to eliminate arterial pressure influences. The isolation of the PVP waveform, and PVP steady-state pressures, from arterial pressure changes is demonstrated in Figures 1 and 3 B and C, respectively.
Our interest in PVP goes beyond its potential role as a substitute for CVP monitoring in selected patients. The initial physiological rationale for our study was to measure vascular pressure at a point in the systemic circulation close to mean systemic pressure (Pms), which is a static, circulatory arrest pressure first defined by Starling and others (23), and refined by Guyton et al. (24) as part of an analysis of venous return (see Fig. 5). During controlled circulatory arrest in humans, all vascular pressures equalize to approximately 15 mm Hg (unpublished observations from planned adult circulatory arrest anesthetics at our institution), which is coincident with the observed arrest pressure noted by Levy (25) in a dog model. This value is simply the ratio of blood volume to vascular compliance. It is precisely this ratio that is useful in defining "relative blood volume." With steady-state cardiac activity, arterial pressures are increased, and CVP depressed, according to the ratio of arterial-to-venous compliances. CVP reflects, then, the operation of cardiac activity on Pms, whereas PVP, as an "upstream" venous pressure, is by definition closer to Pms and less influenced by cardiac activity.
|
The variable offset between patients may reflect either technical variables, such as catheter tip height differences, or physiological variables, such as differences in venous flow and resistance. The gradient between PVP and CVP was remarkably constant within subjects, however, and the small intersubject variability should not detract from the use of PVP as a trend monitor, in exact analogy to the use of CVP as a trend monitor. Throughout the 1026 measurement intervals, the frequency of single point PVP and CVP divergence was only 1.4%, suggesting that PVP trend information could have functioned as an equivalent physiological monitor for these cases. This observation is consistent with the data from parallel animal studies indicating that PVP was at least as good as CVP in estimating volume changes (unpublished data).
In summary, these results are consistent with the results of a parallel animal study that demonstrate a high degree of correlation between PVP and changes in absolute blood volume. The results are also consistent with the hypothesis that PVP reflects mean systemic pressure; and therefore PVP measurement may provide a method of estimating mean systemic pressure during normal circulatory function. Together with the observation of a strong correlation between PVP trends and CVP trends in both craniotomy and complex spine surgery, PVP appears to warrant further investigation as a clinical volume monitor.
|
Acknowledgments |
|---|
Supported by the Division of Anesthesiology and Critical Care Medicine, The Cleveland Clinic.
We thank David Nelson for statistical consultation, Carrie Ceremuga for assistance with clinical study coordination, and David DeFily for helpful discussions.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- English ICW, Frew RH, Pigott JF, Zaki M. Percutaneous catheterization of the internal jugular vein. Anaesthesia 1969; 24: 521–31.[Web of Science][Medline]
- Shippy CR, Appel PL, Shoemaker WC. Reliability of clinical monitoring to assess blood volume in critically ill patients. Crit Care Med 1984; 12: 107–12.[Web of Science][Medline]
- Kelman GR. Interpretation of CVP measurements. Anaesthesia 1971; 26: 209–15.[Web of Science][Medline]
- Mark JB. Central venous pressure monitoring: clinical insights beyond the numbers. J Cardiothorac Vasc Anesth 1991; 5: 163–73.[Medline]
- Reich DL, Moskowitz DM. Complications of cardiovascular monitoring. In: Benumof JL, Saidman LJ, eds. Anesthesia and perioperative complications. 2nd edition. St. Louis: Mosby, 1999: 31–6.
- Saunders DL. On the dangers of monitoring. Or, primum non nocere revisited. Anaesthesia 1997; 52: 399–400.[Web of Science][Medline]
- Lewis CA, Allen TE, Burke DR, et al. Quality improvement guidelines for central venous access. J Vasc Interv Radiol 1997; 8: 475–9.[Web of Science][Medline]
- Cook TL, Ducker CW. Tension pneumothorax following internal jugular cannulation and general anesthesia. Anesthesiology 1976; 45: 554–5.[Web of Science][Medline]
-
Sznajder JI, Zveibil FR, Bitterman H, et al. Central vein catheterization: failure and complication rates by three percutaneous approaches. Arch Intern Med 1986; 146: 259–61.
[Abstract/Free Full Text] -
Royster RL, Johnston WE, Gravlee GP, et al. Arrhythmias during venous cannulation prior to pulmonary artery catheter insertion. Anesth Analg 1985; 64: 1214–6.
[Abstract/Free Full Text] -
Connors AF Jr, Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT investigators. J Am Med Assoc 1996; 276: 889–97.
[Abstract/Free Full Text] -
Holt JP. The effect of positive and negative intra-thoracic pressure on peripheral venous pressure in man. Am J Physiol 1943; 139: 208–11.
[Free Full Text] -
Landis EM, Hortenstine JC. Functional significance of venous blood pressure. Physiol Rev 1950; 30: 1–32.
[Free Full Text] - Pedersen A. The pressure in the central veins and its bearing on peripheral venous pressure measurement. Acta Med Scand 1952; 142: 829–37.
- Sykes MK. Venous pressure as a clinical indication of adequacy of transfusion. Ann R Coll Surg 1963; 33: 185–97.
- Eustace BR. A comparison between peripheral and central venous pressure measurement under clinical circumstances. Injury 1970; 2: 14–8.[Medline]
- Sheldon CA, Balik E, Dhanalal K, et al. Peripheral postcapillary venous pressure–a new hemodynamic monitoring parameter. Surgery 1982; 92: 663–9.[Web of Science][Medline]
- Sheldon CA, Cerra FB, Bohnhoff N, et al. Peripheral postcapillary venous pressure: a new, more sensitive monitor of effective blood volume during hemorrhagic shock and resuscitation. Surgery 1983; 94: 399–406.[Web of Science][Medline]
- Wolf WM, Synder CL, Porter J, et al. Cuff-occluded rate of rise of peripheral venous pressure: a new, highly sensitive technique for monitoring blood volume status during hemorrhage and resuscitation. Surgery 1987; 101: 304–9.[Web of Science][Medline]
- Synder CL, Saltzman D, Happe J, et al. Peripheral venous monitoring with acute blood volume alteration: cuff-occluded rate of rise of peripheral venous pressure. Crit Care Med 1990; 18: 1142–5.[Web of Science][Medline]
- Amar D, Melendez JA, Zhang H, et al. Peripheral venous pressure predicts intraoperative central venous pressure. Anesthesiology 1998; 89(3A): A946.
- Amar D, Melendez JA, Zhang H, et al. Peripheral venous pressure predicts central venous pressure in the post-anesthesia care unit. Anesthesiology 1999; 91(3A): A545.
- Starling EH. Some points in the pathology of heart disease. Lancet 1897; 1: 652–5.
-
Guyton AC, Lindsey AW, Abernathy B, Richardson T. Venous return at various right atrial pressures and the normal venous return curve. Am J Physiol 1957; 189: 609–15.
[Abstract/Free Full Text] -
Levy MN. The cardiac and vascular actors that determine systemic blood flow. Circ Res 1979; 44: 739–47.
[Free Full Text]
This article has been cited by other articles:
|
|
 |
|
  L. Baty, P. Russo, and J. D. Tobias Measurement of Central Venous Pressure From a Peripheral Intravenous Catheter Following Cardiopulmonary Bypass in Infants and Children With Congenital Heart Disease J Intensive Care Med, March 1, 2008; 23(2): 136 - 142. [Abstract] [PDF]
|
|
|
|
 |
|
 T. N. Weingarten, J. Sprung, and J. R. Munis Peripheral Venous Pressure as a Measure of Venous Compliance During Pheochromocytoma Resection Anesth. Analg., October 1, 2004; 99(4): 1035 - 1037. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
