| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Anesthesia and Critical Care, Heart Institute, INCOR, Hospital das Clinicas, University of Sao Paulo, SP, Brazil; and Department of Anesthesia and Critical Care, Béclère Hospital-University Paris XI, Paris, France.
Address correspondence and reprint requests to Frédéric Michard, MD, PhD, Department of Anesthesia and Critical Care, Hopital Antoine Béclère, 157 rue de la porte de Trivaux, 92141, Clamart, France. Address e-mail to michard.frederic{at}free.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
PP) has been shown to be a predictor of fluid responsiveness. Until now, PP has had to be calculated offline (from a computer recording or a paper printing of the arterial pressure curve), or to be derived from specific cardiac output monitors, limiting the widespread use of this parameter. Recently, a method has been developed for the automatic calculation and real-time monitoring of PP using standard bedside monitors. Whether this method is to predict reliable predictor of fluid responsiveness remains to be determined.
METHODS: We conducted a prospective clinical study in 59 mechanically ventilated patients in the postoperative period of cardiac surgery. Patients studied were considered at low risk for complications related to fluid administration (pulmonary artery occlusion pressure <20 mm Hg, left ventricular ejection fraction 
40%). All patients were instrumented with an arterial line and a pulmonary artery catheter. Cardiac filling pressures and cardiac output were measured before and after intravascular fluid administration (20 mL/kg of lactated Ringer’s solution over 20 min), whereas PP was automatically calculated and continuously monitored.
RESULTS: Fluid administration increased cardiac output by at least 15% in 39 patients (66% = responders). Before fluid administration, responders and nonresponders were comparable with regard to right atrial and pulmonary artery occlusion pressures. In contrast, PP was significantly greater in responders than in nonresponders (17% ± 3% vs 9% ± 2%, P < 0.001). The PP cut-off value of 12% allowed identification of responders with a sensitivity of 97% and a specificity of 95%.
CONCLUSION: Automatic real-time monitoring of PP is possible using a standard bedside monitor and was found to be a reliable method to predict fluid responsiveness after cardiac surgery. Additional studies are needed to determine if this technique can be used to avoid the complications of fluid administration in high-risk patients.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although the use of the pulmonary artery catheter is a subject of controversy,2,3 cardiac filling pressures are still widely used to determine fluid therapy in patients undergoing cardiac surgery. However, many clinical studies have shown that cardiac filling pressures are of little value to predict ventricular filling volume, cardiac performance, or the hemodynamic effects of intravascular value expansion.4,5
Mechanical ventilation induces cyclic variations in cardiac preload that are reflected in cyclic changes in aortic blood flow and arterial pulse pressure within the timeframe of a few heart beats.6 The arterial pulse pressure variation induced by mechanical ventilation (PP) has been shown to be useful to discriminate between responder and nonresponder patients to volume loading.7–16
However, until now, PP has had to be calculated offline, from a computer recording or a paper printing of the arterial pressure curve, or to be obtained from specific cardiac output monitors (PiCCO, Pulsion Medical Systems AG, Germany; or PulseCO, LiDCO, United Kingdom). This considerably limits the widespread use of this variable in clinical practice. Recently, a method has been developed for the automatic calculation and real-time monitoring of PP using a regular bedside monitor. Whether this method allows for reliable prediction of fluid responsiveness remains to be determined.
We designed the present study to investigate whether this new method would be useful to predict fluid responsiveness during the immediate postoperative period of cardiac surgery.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Patients
We studied 59 mechanically ventilated patients. This group comprised 25 men and 34 women, aged between 39 and 80 yr (mean age: 61 ± 10 yr). Inclusion criteria were as follows: 1) immediate postoperative period of cardiac surgery, 2) instrumentation with indwelling radial and pulmonary artery catheters, and 3) clinical requirement for a rapid intravascular fluid challenge according to the attending physician’s decision. Because our aim was to test the predictive value of PP in a "real life" situation, we did not predefine the criteria used by the attending physician to determine fluid therapy. Patients were excluded if they had arrhythmias, spontaneous breathing activity, a pulmonary artery occlusion pressure 20 mm Hg, or a left ventricular ejection fraction <40% on the preoperative echocardiography Doppler examination.
Hemodynamic Measurements
Patients were studied while supine, and zero pressure was defined to be at the midaxillary line. Right atrial pressure and pulmonary artery occlusion pressure were recorded throughout the respiratory cycle and measured at end-expiration. Cardiac output was calculated as the mean of five measurements obtained by injecting 10 mL of dextrose solution randomly during the respiratory cycle. Cardiac index was calculated as cardiac output divided by body surface area.
Automatic Calculation and Online Monitoring of Arterial Pulse Pressure Variation
All patients were instrumented with a radial arterial line (20 G) and a capnographic sensor was connected to the respiratory circuit (Fig. 1). We used a multiparameter bedside monitor (DX 2020, Dixtal, Sao Paulo, SP, Brazil) to continuously and simultaneously record the electrocardiogram, the pulse oximetry signal, the right atrial and pulmonary artery pressure curves, the radial arterial pressure curve, and the capnographic signal (Capnostat Mainstream CO2 sensor, Respironics, Murrysville, PA). A specific software was uploaded into the monitor, allowing the recognition of respiratory cycles (from the analysis of the capnographic signal) and the automatic calculation of PP over each respiratory cycle (Fig. 1). Briefly, systolic and diastolic arterial blood pressures are measured on a beat-to-beat basis and pulse pressure is calculated as the difference between systolic and diastolic pressure. Maximum and minimum values for pulse pressure (PPmax and PPmin, respectively) for each mechanical breath are determined and PP is calculated as previously described17:
|
|
|
As shown in Figure 1, the mean value of PP is calculated over three consecutive floating periods of 10 respiratory cycles (from cycles 1 to 10, 2 to 11, and 3 to 12) and the median value of this triple determination is displayed on the bedside monitor. After a new respiratory cycle, the PP value is updated from the analysis of cycles 2 to 11, 3 to 12, and 4 to 13.
Study Protocol
All patients were mechanically ventilated in a volume-controlled mode with a tidal volume of 8 mL/kg and a positive end-expiratory pressure of 5 cm H2O. Spontaneous breathing activity was excluded by clinical examination and the visual inspection of the airway pressure curve on the ventilator and of the capnographic signal on the bedside monitor. Measurements were performed in duplicate, first, before intravascular volume expansion and then after volume expansion using 20 mL/kg of lactated Ringer’s solution over 20 min. Ventilator settings and dosages of inotropic and vasopressive drugs were held constant. A single fluid challenge was studied per patient (the first) during the immediate postoperative period.
Statistical Analysis
The effects of intravascular volume expansion on hemodynamic variables were assessed using a nonparametric Wilcoxon’s ranked sum test. Patients were divided into two groups according to the percent increase in cardiac index in response to intravascular volume expansion. In accord with previous studies,4 we assumed that a 15% change in cardiac index was needed for clinical significance. Therefore, patients with a cardiac index increase induced by intravascular volume expansion 15% and <15% were classified as responders and nonresponders, respectively. The comparison of hemodynamic variables before intravascular volume expansion in responder and nonresponder patients was assessed using a nonparametric Mann-Whitney U-test. Results were expressed as mean values ± sd. Receiver operating characteristics (ROC) curves were built for each parameter. The area under the ROC curves (±se) were calculated and compared. Linear correlations were tested using the Spearman rank method. A P value <0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The attending physician’s decision of intravascular fluid administration was related to tachycardia (heart rate >100 bpm, n = 37), and/or low cardiac output (cardiac index <2.5 L • min–1 • m–2, n = 16), and/or hypotension (mean arterial blood pressure <70 mm Hg, n = 14). Intravascular volume expansion induced an increase in right atrial pressure, pulmonary artery occlusion pressure, cardiac index, and mean arterial blood pressure, and a significant decrease in heart rate and PP (Table 1). Before intravascular fluid administration, PP was significantly and positively correlated (r = 0.76, P < 0.0001) with the percent increase in cardiac index in response to intravascular fluid administration. Conversely, right atrial pressure and pulmonary artery occlusion pressure measured before intravascular volume expansion were not correlated with intravascular volume expansion-induced changes in cardiac index (r = 0.169, P = 0.20; and r = 0.194, P = 0.14, respectively).
|
Thirty-nine patients were responders (cardiac index increase 15%) and 20 patients were nonresponders. Responders and nonresponders were comparable in terms of preoperative left ventricular ejection fraction (62% ± 6% vs 62% ± 8%, P = 0.78). Before intravascular fluid administration, responders and nonresponders were also comparable with regard to heart rate, mean arterial blood pressure, right atrial pressure, and pulmonary artery occlusion pressure (Table 2). Cardiac index was significantly lower in responders than in nonresponders (Table 2), but a significant overlap was observed between groups (Fig. 2). In contrast, PP was significantly greater in responders than in nonresponders (Table 2). The area under the ROC curves (±se) were as follows: 0.98 ± 0.01 for PP, 0.74 ± 0.07 for cardiac index, 0.63 ± 0.07 for pulmonary artery occlusion pressure, and 0.58 ± 0.08 for right atrial pressure (Fig. 3). The area for PP was significantly greater than the area for cardiac index (P = 0.001), pulmonary artery occlusion pressure (P < 0.001), and right atrial pressure (P < 0.001). The threshold PP value of 12% allowed discrimination between responder and nonresponder patients with a sensitivity of 97% and a specificity of 95%.
|
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
PP is possible using a regular bedside monitor and allows one to predict which patients may benefit from fluid therapy in the postoperative period of cardiac surgery. A critical analysis4 of studies designed to examine fluid responsiveness has shown that when the decision of intravascular fluid administration is based on clinical evaluation or on the measurement of cardiac filling pressures, a significant number of patients do not respond to fluid administration with an improvement in hemodynamics. In the present study, one-third of the patient population did not respond to intravascular fluid administration with a clinically relevant increase in cardiac output. This finding emphasizes the limited value of clinical criteria (e.g., hypotension and tachycardia), although widely used to guide a decision to administer fluid, to discriminate between patients who are on the steep portion of the Frank-Starling curve (and hence will turn intravascular fluid administration into a significant increase in cardiac output) and patients who are on the flat part of the curve and will not respond to intravascular fluid administration. This result also underscores the need for better predictors of fluid responsiveness to prompt the decision of intravascular fluid administration in patients identified as responders, and to prevent excessive intravascular and extravascular volume in patients identified as nonresponders.
In cardiac surgery and septic shock patients, the offline calculation of PP has been shown to identify responders to intravascular fluid administration with a sensitivity ranging from 85% to 100%9,15 and a specificity between 87% and 100%,10,15 the best discriminative threshold value for PP ranging between 9.4%14 and 17%15 (mean 12%). Therefore, our findings are in line with these studies because our PP cut-off value of 12% allowed discrimination between responders and nonresponders with a sensitivity of 97% and a specificity of 95%.
Two cardiac output monitors currently available on the market (PiCCOplus from Pulsion Medical Systems, Germany, and LiDCOplus from LiDCO, United Kingdom) calculate and automatically display the arterial pulse pressure variation. These devices are not connected nor synchronized with the ventilator or any kind of respiratory signal, the pulse pressure variation being calculated over an arbitrary period of several seconds. This time window may include several respiratory cycles such that the pulse pressure variation displayed on these monitors may differ from PP, the real pulse pressure variation induced by a single mechanical breath. As far as we know, the PulseCO-derived pulse pressure variation has not been evaluated, whereas a single study has evaluated the clinical value of the PiCCO derived-pulse pressure variation. In patients undergoing off-pump coronary artery bypass grafting, Hofer et al.13 showed that the pulse pressure variation calculated by the PiCCO over a 30-s period decreases with intravascular volume expansion and is able to discriminate between responder and nonresponder patients with a sensitivity and a specificity of 72%.
Our study was not designed to compare the clinical value of the pulse pressure variation calculated by cardiac output monitors to PP obtained with our method. Therefore, our study does not demonstrate the superiority of our method over methods currently available on the PiCCO or on the PulseCO monitors. However, it is important to note that most patients undergoing cardiac surgery are not instrumented with these specific cardiac output monitors. Conversely, they are usually instrumented with an arterial catheter (for continuous arterial blood pressure monitoring and multiple blood samples) and may benefit from the automatic calculation and online monitoring of PP from their regular bedside monitor. This new approach should greatly facilitate and may significantly increase the clinical use of PP in the near future.
Over the last few years, other variables than PP have been proposed to predict fluid responsiveness, such as the respiratory variation in aortic blood flow,18,19 the pulse contour stroke volume variation,20,21 the preejection period variation,8,15 or the vena cava diameter variation.10,22 Although all these variables are useful to predict fluid responsiveness, their assessment requires a specific, sophisticated, and expensive device (e.g., an echocardiography Doppler apparatus) not necessarily available or used in the immediate postoperative period of cardiac surgery. Moreover, none of the more sophisticated or complex parameters mentioned earlier was shown to be more accurate than PP. Actually, several studies have shown that PP is a better predictor of fluid responsiveness than pulse contour stroke volume variation14 or preejection period variation.8 Therefore, PP seems to be not only the most simple but also the most accurate tool to guide intravascular fluid therapy.
Because all patients were studied immediately after the return from the operating room, they were still sedated and passively ventilated. Spontaneous breathing is characterized by irregularity of tidal volume and respiratory frequency.23 As a consequence, in spontaneously breathing patients, PP may vary from one respiratory cycle to another and cannot be used to predict fluid responsiveness.16 Similarly, PP cannot be interpreted in patients with cardiac arrhythmias because the beat-by-beat variation in pulse pressure no longer reflects the effects of mechanical ventilation on cardiac loading conditions.24 Patients with a pulmonary artery occlusion pressure 20 mm Hg, or a left ventricular ejection fraction <40%, were not enrolled in the study to minimize the risk of pulmonary edema during intravascular fluid administration. In this respect, our findings cannot be extrapolated to this specific population, though dynamic predictors of fluid responsiveness have also been shown to be reliable variables in patients with left ventricular systolic dysfunction.21
In conclusion, our study shows that the automatic calculation and online monitoring of PP is possible on a regular bedside monitor and, in contrast to the measurement of cardiac filling pressures, can be used to predict the hemodynamic effects of intravascular fluid administration during the immediate postoperative period of cardiac surgery. Because we studied a population of patients at low risk for complications related to fluid administration, further studies are needed to determine whether this technique can be used to avoid the complications of fluid administration in higher risk patients.
|
Footnotes |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Pestel, K. Fukui, V. Hartwich, P. M. Schumacher, A. Vogt, L. B. Hiltebrand, A. Kurz, Y. Fujita, D. Inderbitzin, and D. Leibundgut Automatic Algorithm for Monitoring Systolic Pressure Variation and Difference in Pulse Pressure Anesth. Analg., June 1, 2009; 108(6): 1823 - 1829. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Fowler, N. K. J. Adhikari, D. C. Scales, W. L. Lee, and G. D. Rubenfeld Update in Critical Care 2008 Am. J. Respir. Crit. Care Med., May 1, 2009; 179(9): 743 - 758. [Full Text] [PDF]
|
|
|
|
|
|
A. Perel Automated Assessment of Fluid Responsiveness in Mechanically Ventilated Patients Anesth. Analg., April 1, 2008; 106(4): 1031 - 1033. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
