Septic shock is the most common cause of morbidity and mortality in critically ill patients. Relative and absolute hypovolaemia due to vasodilatation, increased microvascular permeability and capillary leakage may facilitate development of organ failure. Fluid administration is essential for restoring and optimising cardiac output (CO) and organ perfusion. Early and aggressive fluid resuscitation is one of the cornerstones of management in septic shock patients,1 and previous studies have shown that about 50% of severe sepsis and septic shock patients respond beneficially to fluid loading (fluid responsiveness).2–10 In contrast, in about half of the patients, fluid administration can induce deleterious pulmonary oedema, compromising microvascular perfusion and oxygen delivery. Several studies on sepsis have shown that positive cumulative fluid balance is associated with a high mortality rate.11–13 Hence, an accurate and reliable technique to guide fluid resuscitation is required. Currently, none of the routinely used static variables of cardiac preload, such as central venous pressure (CVP),3,7,10,14–20 pulmonary artery occlusion pressure3,14–17,20 or global end-diastolic volume index,18,19 reliably predict fluid responsiveness. On the contrary, dynamic indicators derived from the arterial pressure waveform such as stroke volume variation (SVV)5,16–23 and pulse pressure variation (PPV)3,9,17–20,22,23 can be used to assess fluid responsiveness in mechanically ventilated patients.
Recently, several new software applications and algorithms have been developed to automatically and continuously calculate these indices. A new device for the noncalibrated pulse contour analysis (FloTrac/Vigileo; Edwards Lifescience, Irvine, California, USA) can be used for automatic and continuous CO monitoring.24 The screen of this device continuously displays the SVV, which is the percentage variation of stroke volume (SV) over a floating period. Previous studies have shown that the SVV obtained by FloTrac/Vigileo could be successfully used for predicting fluid responsiveness in surgical patients,17,18,20,22,23 with the suggestion that this approach could improve postoperative outcome.25 A recently developed automated and continuous measurement of variation in arterial pulse pressure (automated PPV) uses an IntelliVue MP monitor (Philips Medical Systems, Boeblingen, Germany).26,27 Cannesson et al.16 found that there was strong agreement between the automated and manual PPV in terms of fluid responsiveness, particularly in coronary artery bypass graft (CABG) patients. Measurements obtained by one specific device or method cannot be applied to indices from other devices and may result in differences in the threshold value of fluid responsiveness. Therefore, the specific validation of an automated index is needed before clinical use. The aim of this study was to assess and compare the ability of the two novel algorithms of the SVV, obtained by FloTrac/Vigileo, and of the automated PPV, obtained by an IntelliVue MP monitor to predict fluid responsiveness in mechanically ventilated septic shock patients.
This prospective study was conducted in the 10-bed Medical Intensive Care Unit of Songklanagarind Hospital, Prince of Songkla University, Thailand. It was approved by the Institutional Ethics Committee (Chairperson Associate Professor Verapon Chandeying, on 23 February 2009) and informed written consent was obtained from the next of kin of each patient.
The inclusion criteria were as follows: having septic shock as defined by the International Sepsis Definitions Conference,28 being on mechanical ventilation with a tidal volume of at least 8 ml kg−1 without any spontaneous breathing efforts as detected by continuous airway pressure and flow monitoring on the ventilator screen, and having a clinical requirement for a rapid volume challenge according to the attending physician. The physician's decision was based on the presence of clinical signs of acute circulatory failure [mean arterial pressure (MAP) <65 mmHg, urine output <0.5 ml kg−1 h−1, tachycardia and skin mottling] and on the absence of any contraindication to a fluid challenge, including life-threatening hypoxaemia and/or evidence of volume overload by physical examination or chest radiograph. Patients were excluded if their chest radiographs were compatible with acute pulmonary oedema, cardiac arrhythmia (atrial fibrillation or frequent premature beats), severe peripheral vascular disease and left ventricular ejection fraction less than 40%.
Stroke volume variation obtained by FloTrac/Vigileo
A FloTrac sensor kit was connected to the arterial line and coupled to the Vigileo monitor (software version 03.01). The system was zeroed to ambient pressure at the phlebostatic axis and measurement initiated. The FloTrac analysed the arterial pressure waveform 100 times per second over 20 s. The SV was based on the contribution of pulse pressure relative to SV, which is the proportion of pulse pressure to the standard deviation of arterial pulse pressure (APsd). The device calculated SV as APsd × Khi (χ), where χ compensates for differences in vascular compliance and resistance derived from a multivariate regression model. This system allowed for real-time adjustments to changing vascular tone with a recalculation of χ every minute. The CO was calculated from SV × heart rate (HR) and SVV used the following equation: SVV (%) = (SVmax−SVmin)/SVmean. SVmax, SVmin and SVmean were determined by this system over a window of 20 s. The haemodynamic data were set to display continuously in 1-min intervals on the Vigileo monitor. The mean values of the three determinations before and after completion of volume loading were recorded for statistical analysis.
Automated pulse pressure variation by IntelliVue MP monitor
The automated PPV was displayed in real-time as a percentage by a Philips IntelliVue MP70 monitor (Philips Medical Systems). The algorithm is commercially available and has been described by Aboy et al.26 It allows PPV determination from the arterial pressure waveform alone with no need for airway pressure acquisition. This method is based on an automatic detection algorithm, kernel smoothing, and rank-order filters involving seven consecutive steps (beat minimal detection, beat maximum detection, beat mean calculation, pulse amplitude pressure, envelope estimation, pulse amplitude pressure estimation and PPV estimation). PPV was calculated and averaged over four cycles of 8 s. The mean values of the three determinations before and within 3 min after completion of volume loading were recorded for statistical analysis.
The fluid challenge was performed with 500 ml of 6% hydroxyethyl starch (130/0.4; Voluven; Fresenius Kabi, Bad Homburg, Germany) over 30 min. Variables including HR, SBP and DBP, MAP, CVP, cardiac index (CI), stroke volume index (SVI), systemic vascular resistance index (SVRI) and oxygen delivery index (DO2I) were recorded before and within 3 min after volume administration. The dynamic arterial elastance (Eadyn) defined as the PPV to SVV ratio was calculated before and after volume loading.29 Cardiac arrhythmias were absent during haemodynamic readings. The ventilator setting and dosage of vasopressors were all kept constant during the study period.
Continuous variables were assessed for normal distribution (Kolmogorov–Smirnov test for normality). All data were presented as mean ± SD unless otherwise specified. The comparisons of haemodynamic data before and after volume expansion were assessed using the paired Student's t-test and the comparisons between responders and nonresponders were assessed using the two-sample Student's t-test. Responders were defined as patients with an increase in SVI of at least 15% after fluid administration.2,21 Receiver operating characteristic (ROC) curves were generated for SVV and PPV by varying the discriminative threshold, and areas under the ROC curves (AUC) with 95% confidence interval were calculated and compared.30 The threshold values for SVV and PPV were determined by considering the values that yielded the greatest sensitivity and specificity for predicting fluid responsiveness. The correlation between SVV and PPV and between the baseline values of each variable and the SVI response to volume infusion was determined using Pearson's linear correlation coefficient. A P value less than 0.05 was considered statistically significant. All data were analysed using SPSS version 11.5 (SPSS, Chicago, Illinois, USA).
A calculated sample size of 42 patients was considered necessary to detect a 5% (SD = 5%) difference in SVV and PPV between fluid responders and nonresponders (∝ = 0.05 and power = 0.9).
Forty-two patients were enrolled in this study. Thirty-eight of them were sedated with morphine and midazolam and only three received vecuronium during the study period. All received norepinephrine, which was supplemented with dopamine in six patients. Baseline patient characteristics are shown in Table 1. Thirty-four (80.9%) patients were studied within 24 h after onset of septic shock. The arterial line was inserted in either the radial artery (83.3%) or the femoral artery (16.7%).
The correlation between baseline SVV and PPV was strong, (r = 0.96; P < 0.001; Fig. 1). Twenty-four (57.1%) patients were classified as responders to fluid administration. The haemodynamic variables in responders and nonresponders are shown in Table 2. Volume infusion produced an increase in MAP, CVP, SVI, CI and DO2I. In contrast, the intravenous fluid significantly decreased HR, SVRI, SVV and PPV. Both baseline SVV and PPV were significantly higher in responders than in nonresponders, 15.5 ± 4.5 vs. 8.2 ± 3.3% and 16.4 ± 5.2 vs. 8.3 ± 3.5, respectively (P < 0.001 for both). Other baseline haemodynamic data were not different between responders and nonresponders. The amount of fluid resuscitation before enrollment in the study and within 24 h after study was not different between responders and nonresponders 2748.8 ± 785.1 vs. 2810.3 ± 1193.4 ml (P = 0.84) and 5416 ± 1438.5 vs. 4911.3 ± 2081.3 ml (P = 0.38, respectively). The AUC (95% confidence interval) of SVV and PPV was 0.92 (0.832–1.00) and 0.916 (0.829–1.00), respectively. The comparison of the ROC curves showed that SVV and PPV predicted fluid responsiveness similarly (Fig. 2). There was no difference between the AUC of SVV and PPV (P = 0.73). The optimal threshold value for discrimination between responders and nonresponders was 10% for SVV (sensitivity 91.7% and specificity 83.3%) and 12% for PPV (sensitivity 83.3% and specificity 83.3%).
When the 35 patients with radial artery catheterisation were selected, the results were statistically unchanged. The AUC (95% confidence interval) of SVV and PPV was 0.94 (0.862–1.00) and 0.933 (0.849–1.00), respectively. The optimal threshold value for SVV and PPV was 10% (sensitivity 90.9% and specificity 84.6%) and 12% (sensitivity 81.2% and specificity 84.6%), respectively.
There was a statistically significant positive linear correlation between baseline SVV and PPV with percentage changes in SVI by volume loading [r = 0.59 (P < 0.001) and r = 0.57 (P < 0.001), respectively].
In the 25 patients with hypotension and preserved preload dependence, defined as the presence of SVV at least 10%, fluid administration induced at least 15% increase in MAP in 12 (MAP responder). The baseline Eadyn was not different between MAP responders and MAP nonresponders (1.02 ± 0.13 vs. 1.08 ± 0.14, respectively).
This study demonstrated that both the SVV measurement by FloTrac/Vigileo and the automated PPV obtained by IntelliVue MP70 can be used to predict fluid responsiveness in mechanically ventilated septic shock patients. This is the first study to validate fluid responsiveness in this group using novel continuous automated devices and the recently released version (third generation) of FloTrac/Vigileo software.
Our results are in agreement with a recent study that evaluated and compared the SVV obtained by FloTrac/Vigileo and automated PPV, assessed with an IntelliVue monitor, during major abdominal surgery.22 Derichard et al.22 found that these two devices performed similarly in terms of fluid responsiveness in haemodynamically unstable surgical patients. Previous studies have investigated the ability of the SVV obtained by FloTrac/Vigileo to predict fluid responsiveness in surgical patients. The results of our study are in accordance with recent studies affirming the value of SVV obtained by this device as an accurate predictor of fluid responsiveness, with the AUC range from 0.824 to 0.95 and optimal threshold value between 9 and 12%.17,18,20,22,23
Automated PPV measurement correlates strongly with that obtained manually in predicting fluid responsiveness in pre-CABG patients, with best threshold values of 10% and 12%, respectively,16 and Derichard et al.22 also reported that automated PPV obtained by an IntelliVue monitor correlated with manual PPV. Their optimal threshold value was 13% for both PPV methods. Our results agree that the automated and continuous PPV from the IntelliVue MP monitor is able to predict fluid responsiveness, using the algorithm proposed by Aboy et al.
Our study showed that the discrimination and optimal threshold values of each variable were not statistically dependent on the site of arterial catheterisation. Previous studies have shown that CO measurements via radial and femoral arteries using FloTrac/Vigileo were comparable.31,32 This supports the use of the proprietary algorithm of the FloTrac/Vigileo to allow CO or SVV monitoring irrespective of the signal detection site.
SVV and PPV are shown to be good predictor of fluid responsiveness in various patient groups and devices.16–23 However, the validation of these indices has not yet been confirmed in a large multicentre study. Nevertheless, previous studies have shown that SVV25 or PPV33-guided fluid management can reduce postoperative complications. Further study is, thus, clearly needed to establish whether goal-directed fluid management based on SVV and PPV will result in improved outcomes in septic shock patients.
Dynamic indices of fluid responsiveness cannot be used in the settings of cardiac arrhythmias, spontaneous breathing activity, ventilation with low tidal volumes, right and left ventricular dysfunction and open chest condition. Potential users should be aware of these limitations before use in clinical practice.
Recently, Monge Garcia et al.29 found that baseline Eadyn was significantly different between MAP responders and MAP nonresponders, and that a baseline Eadyn more than 0.89 predicted a MAP increase after fluid loading. We failed to find a difference in baseline Eadyn between MAP responders and nonresponders, and it is possible that different methods for PPV measurement may be responsible for these conflicting results. Monge Garcia et al. measured PPV manually, in contrast to automated PPV in our study. Clearly more evaluation is needed before Eadyn can be used in a clinical setting.
Some limitations in this study must be considered. First, validation studies on the accuracy of CO obtained by FloTrac/Vigileo have yielded conflicting results. Recent studies have demonstrated that this device does not accurately monitor CO in septic shock when compared with the pulse induced continuous cardiac output system.34,35 However, the earlier software versions for this device were used in these studies. De Backer et al.36 found that the third generation software of FloTrac/Vigileo is more accurate, precise, and less affected by total systemic vascular resistance than the second generation software and is better suited to septic shock. FloTrac/Vigileo is able to track changes in SV and CO induced by volume loading and has been used to evaluate fluid responsiveness in numerous previous studies.17,18,20,22,23 Second, a cut-off point of at least a 15% increase in SVI after fluid administration was used to classify a patient as a responder, which could affect the threshold value of SVV and PPV. Different responder definitions influence the thresholds obtained by the ROC analysis.
In conclusion, SVV measurement by FloTrac/Vigileo and automated PPV obtained by the IntelliVue MP70 demonstrated comparable performance in terms of predicting fluid responsiveness in passively ventilated patients with septic shock, with a tidal volume of at least 8 ml kg−1 and a regular cardiac rhythm. Further study of the impact on optimisation of management by these devices and the corresponding outcome among septic shock patients is necessary.
The authors would like to thank Dr Alan Geater of the Epidemiology Unit, Faculty of Medicine, Prince of Songkla University, for his assistance with the statistical components of the article.
The present study was supported by a research grant of the Faculty of Medicine, Prince of Songkla University.
None of the authors has any conflict of interest.
1. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock
. N Engl J Med
2. Tavernier B, Makhotine O, Lebuffe G, et al. Systolic pressure variation as a guide to fluid
therapy in patients with sepsis-induced hypotension. Anesthesiology
3. Michard F, Boussat S, Chemla D, et al. Relation between respiratory changes in arterial pulse pressure and fluid
responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med
4. Feissel M, Michard F, Mangin I, et al. Respiratory changes in aortic blood velocity as an indicator of fluid
responsiveness in ventilated patients with septic shock
5. Marx G, Cope T, McCrossan L, et al. Assessing fluid
responsiveness by stroke volume
variation in mechanically ventilated patients with severe sepsis. Eur J Anaesthesiol
6. Vieillard-Baron A, Chergui K, Rabiller A, et al. Superior vena caval collapsibility as a gauge of volume status in ventilated septic patients. Intensive Care Med
7. Barbier C, Loubieres Y, Schmit C, et al. Respiratory changes in inferior vena cava diameter are helpful in predicting fluid
responsiveness in ventilated septic patients. Intensive Care Med
8. Feissel M, Michard F, Faller JP, Teboul JL. The respiratory variation in inferior vena cava diameter as a guide to fluid
therapy. Intensive Care Med
9. Feissel M, Teboul JL, Merlani P, et al. Plethysmographic dynamic indices predict fluid
responsiveness in septic ventilated patients. Intensive Care Med
10. Durairaj L, Schmidt GA. Fluid
therapy in resuscitated sepsis: less is more. Chest
11. Alsous F, Khamiees M, DeGirolamo A, et al. Negative fluid
balance predicts survival in patients with septic shock
: a retrospective pilot study. Chest
12. Vincent JL, Sakr Y, Sprung CL, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med
13. Murphy CV, Schramm GE, Doherty JA, et al. The importance of fluid
management in acute lung injury secondary to septic shock
14. Kumar A, Anel R, Bunnell E, et al. Pulmonary artery occlusion pressure and central venous pressure fail to predict ventricular filling volume, cardiac performance, or the response to volume infusion in normal subjects. Crit Care Med
15. Osman D, Ridel C, Ray P, et al. Cardiac filling pressures are not appropriate to predict hemodynamic response to volume challenge. Crit Care Med
16. Cannesson M, Slieker J, Desebbe O, et al. The ability of a novel algorithm for automatic estimation of the respiratory variations in arterial pulse pressure to monitor fluid
responsiveness in the operating room. Anesth Analg
17. Biais M, Nouette-Gaulain K, Cottenceau V, et al. Uncalibrated pulse contour-derived stroke volume
variation predicts fluid
responsiveness in mechanically ventilated patients undergoing liver transplantation. Br J Anaesth
18. Hofer CK, Senn A, Weibel L, Zollinger A. Assessment of stroke volume
variation for prediction of fluid
responsiveness using the modified FloTrac and PiCCOplus system. Crit Care
19. Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid
responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med
20. Cannesson M, Musard H, Desebbe O, et al. The ability of stroke volume
variations obtained with Vigileo/FloTrac system to monitor fluid
responsiveness in mechanically ventilated patients. Anesth Analg
21. Reuter DA, Felbinger TW, Kilger E, et al. Optimizing fluid
therapy in mechanically ventilated patients after cardiac surgery by on-line monitoring
of left ventricular stroke volume
variations. Comparison with aortic systolic pressure variations. Br J Anaesth
22. Derichard A, Robin E, Tavernier B, et al. Automated pulse pressure and stroke volume
variations from radial artery: evaluation during major abdominal surgery. Br J Anaesth
23. Biais M, Bernard O, Ha JC, et al. Abilities of pulse pressure variations and stroke volume
variations to predict fluid
responsiveness in prone position during scoliosis surgery. Br J Anaesth
24. Manecke GR. Edwards FloTrac sensor and Vigileo monitor: easy, accurate, reliable cardiac output
assessment using the arterial pulse wave. Expert Rev Med Devices
25. Benes J, Chytra I, Altmann P, et al. Intraoperative fluid
optimization using stroke volume
variation in high risk surgical patients: results of prospective randomized study. Crit Care
26. Aboy M, McNames J, Thong T, et al. A novel algorithm to estimate the pulse pressure variation index deltaPP. IEEE Trans Biomed Eng
27. Austin D, Staats C, Aboy M. A novel approach to pulse pressure variation estimation. Conf Proc IEEE Eng Med Biol Soc
28. Levy MM, Fink MP, Marshall JC, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med
29. Monge Garcia MI, Gil Cano A, Gracia Romero M. Dynamic arterial elastance to predict arterial pressure response to volume loading in preload-dependent patients. Crit Care
30. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology
31. Hofer CK, Button D, Weibel L, et al. Uncalibrated radial and femoral arterial pressure waveform analysis for continuous cardiac output
measurement: an evaluation in cardiac surgery patients. J Cardiothorac Vasc Anesth
32. Manecke GR Jr, Auger WR. Cardiac output
determination from the arterial pressure wave: clinical testing of a novel algorithm that does not require calibration. J Cardiothorac Vasc Anesth
33. Lopes MR, Oliveira MA, Pereira VO, et al. Goal-directed fluid
management based on pulse pressure variation monitoring
during high-risk surgery: a pilot randomized controlled trial. Crit Care
34. Sakka SG, Kozieras J, Thuemer O, van Hout N. Measurement of cardiac output
: a comparison between transpulmonary thermodilution and uncalibrated pulse contour analysis. Br J Anaesth
35. Monnet X, Anguel N, Naudin B, et al. Arterial pressure-based cardiac output
in septic patients: different accuracy of pulse contour and uncalibrated pressure waveform devices. Crit Care
36. De Backer D, Marx G, Tan A, et al. Arterial pressure-based cardiac output monitoring
: a multicenter validation of the third-generation software in septic patients. Intensive Care Med