Secondary Logo

Journal Logo

Fluids

Tidal volume challenge to predict fluid responsiveness in the operating room

An observational study

Messina, Antonio; Montagnini, Claudia; Cammarota, Gianmaria; De Rosa, Silvia; Giuliani, Fabiana; Muratore, Lara; Della Corte, Francesco; Navalesi, Paolo; Cecconi, Maurizio

Author Information
European Journal of Anaesthesiology: August 2019 - Volume 36 - Issue 8 - p 583-591
doi: 10.1097/EJA.0000000000000998

Abstract

Introduction

Dedicated algorithms and protocols of anaesthetic care regarding fluid therapy are key factors to prevent peri-operative hypovolaemia or hypervolaemia, which are both known to increase morbidity and length of hospital stay.1–3 Fluid responsiveness [i.e. the increase in stroke volume (SV) after a fluid challenge] is limited to about 50% of critically ill or surgical patients.4–8 For this reason, fluid challenge administration should be based on predictors of fluid responsiveness.9 Static indexes, such as central venous pressure and pulmonary wedge pressure, are unsuited for this purpose,10 but the dynamic indexes, such as pulse pressure variation (PPV) and stroke volume variation (SVV), reliably predict the effect of fluid challenge administration during controlled mechanical ventilation when a tidal volume (VT) of at least 8 ml kg−1 is used.11–13

Use of an intra-operative lung-protective ventilation strategy (VT of <8 ml kg−1 of predicted body weight, PBW) is associated with a better outcome14 and is now suggested as standard practice in the operating room.15 Such small VTs limit the assessment of fluid responsiveness in surgical patients by means of dynamic indexes. To overcome this VT-related limitation of PPV and SVV, the prediction of fluid responsiveness can be achieved by applying functional haemodynamic tests16 aimed at increasing venous return and enhancing right ventricle preload dependence.10 Among these, the interruption of positive pressure ventilation (the so-called end-expiratory occlusion test, EEOT),17 initially proposed in critically ill patients,17–19 has been tested in two studies of elective surgical patients, with conflicting results. In patients ventilated with a mean VT of 8.2 ml kg−1, the EEOT was unable to reliably predict fluid responsiveness,20 while the opposite was demonstrated in patients ventilated with a mean VT of 6.8 ml kg−1.7

More recently, in 20 critically ill patients with acute circulatory failure, Myatra et al.21 successfully tested the hypothesis that fluid responsiveness could be reliably predicted by evaluating PPV and SVV changes after a ‘VT challenge’ (VTC), defined as a 1-min increase of VT from 6 to 8 ml kg−1 PBW.21

We hypothesised that the baseline reliability of the dynamic indexes in elective surgical patients undergoing protective ventilation would be enhanced by the use of functional haemodynamic tests. Therefore, we designed this study to assess the sensitivity and specificity of PPV and SVV changes after a VTC in predicting fluid responsiveness in a population of neurosurgical patients ventilated with 6 ml kg−1 PBW. We also the reliability of VTC and EEOT performed at both 6 and 8 ml kg−1 PBW.

Materials and methods

Patients

A prospective study was conducted in the neurosurgery operating rooms at the University Hospital ‘Maggiore della Carità’ in Novara, Italy. The protocol was designed in accordance with the principles outlined in the Declaration of Helsinki; the study was approved by the local institutional ethics committee (Comitato Etico Interaziendale; Corso Mazzini n. 18, 28100 Novara, Italy; protocol number 192/17; approval date 15 December 2017; Chairperson Prof. Gian Carlo Avanzi) and registered (ACTRN12618000351213). Informed consent was obtained from all the participants.

All elective supine neurosurgical patients (age >18 years) requiring invasive arterial monitoring and showing a reduction of more than 20% in systolic arterial pressure (SAP) with respect to the values observed before anaesthesia induction were considered eligible. The exclusion criteria were: any recurrent cardiac arrhythmia; reduced ventricular systolic function - left (ejection fraction <40%), right (systolic peak velocity of tricuspid annular motion <0.17 m s−1); BMI more than 30; intra-operative use of vasopressors or inotropes before or during VTC and EEOT applications; chronic lung disease; pre-operative use of beta blocking agents; and clinical or radiological signs of intracranial hypertension.

Perioperative management

All patients received standard intra-operative monitoring, including heart rate, peripheral oxygen saturation, continuous electrocardiography and noninvasive blood pressure monitoring. After pre-oxygenation, general anaesthesia was induced with propofol, remifentanil and cisatracurium besilate (0.15 to 0.2 mg kg−1), and maintained with propofol (1.5 to 3.0 mg kg−1 h−1) or sevoflurane (1to 2%) along with remifentanil (0.1 to 0.5 μg kg−1 min−1). All patients received an intermittent bolus of cisatracurium 0.15 mg kg−1 every 40 to 50 min to guarantee a complete neuromuscular blockade throughout the intervention. Anaesthetic administration was targeted to maintain a bispectral index (BIS monitor; Medtronic, Brooklyn Park, Minnesota, USA) of 40 to 60 throughout the surgical procedure.22 All patients received lactated Ringer's solution at 4 ml kg−1 per hour as a maintenance fluid infusion during surgery and were ventilated at baseline in volume-control mode with the following settings: VT of 6 ml kg−1 PBW and positive end-expiratory pressure set between 3 and 6 cmH2O (FLOW-I C40 ventilator; Maquet Critical Care, Sweden) to achieve and maintain a peripheral oxygen saturation of 96% and an end-tidal carbon dioxide concentration between 30 and 35 mmHg. The PBW (kg) was calculated as follows: X + 0.91[height (cm) - 152.4]; (X = 50 for men and 45.5 for women). After induction of general anaesthesia, invasive blood pressure monitoring was obtained by inserting a 20-G cannula into the radial artery. The pressure signal was then connected to both the operating room monitor (Mindray BeneView T8; Soma Technology, Inc., Bloomfield, Connecticut, USA) and to the MostCare device (Vyetech Health, Padua, Italy), by means of the manufacture's Y cable. A square-wave test was used in all patients to exclude under or overdamping of the pressure signal.23

Haemodynamic monitoring and tests

The MostCare works with a sampling rate of 1.000 points (Pt) per second, analysing both the systolic and the diastolic part of arterial waveform signal and calculates SV as the ratio between the area under the systolic component of the curve and the systemic vascular impedance by analysing the profile of the ‘points of instability’. These points are generated by the mechanical interaction (i.e. pressure/time changes) between forward (due to cardiac systole) and backward pressure waves (coming from the peripheral vessels) and define the specific profile of each arterial waveform, which is analysed by (MostCare) for the calculation of the vascular impedance.24,25 Arterial pressures (systolic, diastolic, mean, dicrotic) and PPV are directly measured from arterial pressure waveform, while SVV is calculated by analysing the changes in SV over time. All the indexed values, including SV index, SVI and cardiac index, CI, are calculated using the patient's anthropometric measurements. All the haemodynamic variables recorded during each set of measurements were averaged according to the default time-setting (30 s) of MostCare and imported into a dedicated EXCEL (Microsoft, Redwood, Mississippi, USA) spreadsheet for further analysis.

Study protocol

Measurements were started during a period of haemodynamic stability (defined as changes in mean arterial pressure less than 10% over 5 min7,26).

The study protocol (see Fig. 1) was, an initial recording of measurements (step1) and then the first EEOT test (EEOT6a); second, after 1 min, a set of measurements was recorded (step 2); third, the VTC was applied by increasing the VT up to 8 ml kg−1 PBW and PPV and SVV changes were recorded (ΔPPVVTC and ΔSVVVTC, respectively, calculated as percentage of variations between the values of PPV and SVV recorded at step 2 and 1 min after the VT increase); fourth, after 1 min, another set of measurements was recorded (step 3). Following this, a second EEOT (EEOT8) was performed. Fifth, the VT was reduced back to 6 ml kg−1 PBW and after 1 min, a set of measurements was recorded (step 4); sixth, a third EEOT was performed (EEOT6b); seventh, after 1 min, a set of measurements was recorded (baseline) and the FC of 250 ml of Ringer's solution was infused over 10 min. Only the haemodynamic data obtained from the first fluid challenge administered to each enrolled patient were used for the analysis. Positive end-expiratory pressure was kept constant during the study period. Each EEOT was performed by interrupting mechanical ventilation for 30 s, using the software functionality ‘expiratory hold’ on the FLOW-I C40. For the safety of the patient, the interruption of the protocol was at discretion of the attending anaesthetist.

Fig. 1
Fig. 1:
Study protocol (see text for further explanations). EEOTs were performed by interrupting mechanical ventilation for 30 s. The VTC was performed by increasing the tidal volume from 6 to 8 ml kg−1 PBW for 1 min (green line). EEOT, end-expiratory occlusion test; FC, fluid challenge; PBW, predicted body weight; VTC, tidal volume challenge.

Statistical analysis

A receiver operating characteristic (ROC) curve approach was used to evaluate the reliability of VTC and EEOT tests in predicting fluid responsiveness. ROC [95% confidence interval (95% CI)] curves were constructed for PPV and SVV values before fluid challenge administration; for the percentage change in PPV and SVV between 6 and 8 ml kg−1 PBW ventilation after 1 min of VTC application; and for the percentage change in SVI and CI after each EEOT application (ΔSVI and ΔCI, respectively). The haemodynamic data recorded at the end of each test were compared with baseline values from the minute before the application. A patient was considered fluid responsive if the SVI increased at least 10% after fluid challenge administration.7,8 Cut-off values were chosen with the highest Youden index and, finally, statistically significant ROC curves (P < 0.05) were compared using the De Long test.27 Considering the possibility of an overlap between responders and nonresponders, we determined a grey zone for ΔPPVVTC, ΔSVVVTC, ΔSVI and ΔCI, considering a low cut-off value including 90% of negative fluid challenge responses, and a high cut-off value predicting positive fluid challenge in 90% of cases.7,28

The sample size of the study was calculated by means of the comparison of the areas under the ROC curves test (AUC). For this purpose, we predicted an AUC of at least 0.75, which is the threshold for considering a diagnostic test as accurate,29 and compared it with the null hypothesis (AUC = 0.50; no discriminating power). Accordingly, a sample size of 38 patients was calculated (type I error of 5% and type II error of 20%).

Normal distribution was evaluated by means of the d’Agostino-Pearson test and, accordingly, data are expressed as median with interquartile [IQR] range or mean (SD). Changes in continuous variables after VTC were compared using a paired t test or Wilcoxon signed rank sum test, while an independent t test or Mann–Whitney U test was used for subgroup comparisons, as appropriate. For dichotomous or categorical variables, a Chi-square test for comparison of proportions were applied.

Statistical analyses were conducted using GraphPad PRISM V6 (GraphPad Software Inc., San Diego, California, USA) and Medcalc (Software 8.1.1.0; Mariakerke, Belgium). For all comparisons, we considered significant P values less than 0.05.

Results

From January 2018 to May 2018, 65 consecutive neurosurgical patients were considered eligible for inclusion. However, 22 were excluded before and three after the enrolment (see Fig. 2). Finally, 40 patients were analysed. The study protocol was never interrupted by the attending anaesthetist and no adverse effects were reported after EEOT or VTC use. The study protocol was applied within the first hour after skin incision in all the enrolled patients. Fluid challenge administration induced an increase in SVI at least 10% in 21 patients (52.5%, fluid challenge responders). Demographic characteristics, comorbidities, surgical procedures, risk scores and ventilatory variables were comparable between responders and nonresponders and are reported in Table 1.

Fig. 2
Fig. 2:
Flow of patients in the study. FC, fluid challenge.
Table 1
Table 1:
Patients’ characteristics at enrolment

The measurements from each step of the study protocol are reported in Table 2. The haemodynamic values of responders and nonresponders before fluid challenge administration were comparable, except for SVV, which was significantly higher in responders as compared to nonresponders. Fluid challenge administration significantly increased CI, SVI and SAP in responders, while it reduced PPV and SVV. In nonresponders, fluid challenge administration reduced CI and SVI and did not affect any of the other considered variables.

Table 2
Table 2:
Haemodynamic variables at each step of the protocol

PPV and SVV values recorded at baseline, before fluid challenge administration were poor at discriminating fluid responsiveness (see Table 3).

Table 3
Table 3:
Reliability of dynamic indexes, tidal volume challenge and end-expiratory occlusion tests in predicting fluid responsiveness

Effect of VTC on pulse pressure variation and stroke volume variation

VTC application increased both PPV and SVV in responders (from 6.3% [4.1 to 7.5] to 10.3% [7.6 to 12.7], P < 0.0001, and from 7.3% [5.3 to 9.4] to 10.8% [7.8 to 12.8], P < 0.0001, for PPV and SVV, respectively), but not in nonresponders (see Table 2 and Supplemental Figure 1 in the Supplemental Digital Content, http://links.lww.com/EJA/A201). These changes reliably predicted fluid responsiveness (see Table 3 and Fig. 3).

Fig. 3
Fig. 3:
Receiver operating characteristic curves of pulse pressure variation and stroke volume variation variations after tidal volume challenge application [ΔPPVVTC (red line) and ΔSVVVTC (blue line), left] and after the EEOT performed at 8 ml kg−1 of PBW (EEOT8) [ΔCI (green line) and ΔSVI (purple line), right]. The ROC curves of PPV (black line) and SVV (grey line) at baseline (before fluid challenge) are also reported in each figure. Both VTC and EEOT8 performed better than baseline PPV and SVV in predicting fluid responsiveness (see Table 4). EEOT, end-expiratory occlusion test; PBW, predicted body weight; PPV, pulse pressure variation; SVV, stroke volume variation; VTC, tidal volume challenge.

After VTC application, ΔPPVVTC predicted fluid responsiveness with a sensitivity of 94.7% (95% CI 73.9 to 99.8) and a specificity of 76.1% (95% CI 52.8 to 91.7) for a PPVVTC increase cut-off of 13.3%, while ΔSVVVTC predicted fluid responsiveness with a sensitivity of 78.9% (95% CI 54.4 to 93.9) and a specificity of 95.2% (95% CI 76.1 to 99.8) for a SVVVTC increase cut-off of 12.1%.

Effect of end-expiratory occlusion test on stroke volume index and cardiac index

EEOT6 did not predict fluid responsiveness when performed at 6 ml kg−1 PBW ventilation. On the contrary, the changes in CI and SVI after EEO8 reliably predicted fluid responsiveness (see Table 3 and Fig. 3).

After EEO8, ΔCI predicted fluid responsiveness with a sensitivity of 89.4% (95% CI 66.8 to 98.7) and a specificity of 85.7% (95% CI 63.6 to 96.5), with a CI increase cut-off of 3.6%, while ΔSVI predicted fluid responsiveness with a sensitivity of 89.4% (95% CI 66.8 to 98.7) and a specificity of 85.7% (95% CI 63.6 to 96.9), with a SVI increase cut-off of 4.7%.

Receiver operating characteristic comparisons

The comparisons with EEOT6 were not performed, as the ROC curve for the test was not significant. The AUCs of VTC and EEOT8 were all significantly greater than the AUCs of baseline PPV and SVV (P < 0.001 for all the comparisons). The AUCs of VTC and EEOT8 were not different (see Table 4).

Table 4
Table 4:
Receiver operating characteristic curve comparisons

Discussion

The main findings of our study conducted in elective neurosurgical patients are first, the use of a VTC increases both PPV and SVV in fluid responders and these changes accurately predict fluid responsiveness; second, the changes in SVI and CI after an EEOT are reliable only when the test is performed at 8 ml kg−1 PBW ventilation; third, the sensitivity and specificity of VTC and EEOT8 in revealing preload dependency are not significantly different; and fourth, baseline PPV and SVV in elective neurosurgical patients undergoing protective ventilation do not predict fluid responsiveness.

The reliability of dynamic indexes in predicting fluid responsiveness is affected by several clinical variables,18,30–34 which are unfortunately present in the vast majority of critically ill patients.11,12 Moreover, in the operating room, PPV values ranging between 9 and 13% are poorly predictive of fluid responsiveness,28 and require additional haemodynamic tests to assess preload dependence and to avoid inappropriate fluid administration.8,21

Our results confirm that a protective ventilation strategy precludes the use of baseline PPV and SVV in the assessment of the volume status, even if all other validity criteria are respected. Moreover, the percentage of fluid responsive patients was about 50% (consistent with previous findings in elective surgical patients7,8,35,36), suggesting that functional haemodynamic tests should be used in patients undergoing protective ventilation in the operating room, to enhance the predictive value of PPV and SVV.

Recently, Myatra et al.21 successfully tested the hypothesis in critically ill patients that increasing the VT from 6 to 8 ml kg−1 for only 1 min would correct baseline PPV and SVV to values discriminating responders from nonresponders. In fact, as protective ventilation causes false-negative values of the dynamic indexes,37 raising VT and intrathoracic pressure should increase PPV and SVV to a different extent in responders and nonresponders.

In our study, the VTC increased both PPV and SVV only in fluid responders, but to a smaller extent than the previous results of Myatra et al.21

Moreover, the thresholds of ΔPPVVTC and ΔSVVVTC identified by the ROC curve analysis are different in our study compared with the study by Myatra et al.21 This finding could be explained by the different extent of the haemodynamic effect of VTC application in a population of elective supine surgical, not obese patients as compared to the population of Myatra et al.,21 showing a median compliance of the respiratory system of about 28 ml cmH2O−1 in critically ill patients with acute respiratory disease.21 Chest wall and respiratory system compliances affect the transmission of the intrathoracic pressure to the pleural and atrial pressure and, in turn, the dynamic indexes.38 In fact, only about one-third of the applied airway pressure is transmitted to the pericardium and the vena cava, and this effect is enhanced by a reduced chest wall compliance,38 which is present in about 30% of critically ill patients.39

EEOT did not predict fluid responsiveness when performed during a 6 ml kg−1 PBW ventilation, as previously reported in the literature.21 It showed a sensitivity and specificity comparable to the VTC when performed during a 8 ml kg−1 PBW ventilation with a cut-off of about 5% of increase in CI and SVI. The increase in right ventricle preload and, in turn, in CI and SVI after the EEOT manoeuvre is related to the changes in intrathoracic pressure during mechanical ventilation. The effect on venous return of a lung protective ventilatory strategy could be insufficient and the consequent changes in CI and SVI after the occlusion manoeuvre too small to discriminate between responders and nonresponders.

As suggested by Myatra et al.,21 the VTC could potentially be used in resource-limited settings, as PPV measurement can be achieved without a dedicated cardiac output device. Nevertheless, 27.5% of patients were included in the grey zone analysis of ΔPPVVTC, suggesting caution in the sole use of this parameter. A simultaneous increase of PPV more than 15% and SVI more than 6% after VTC and EEOT8 would identify the vast majority of responders. However, the assessment of these haemodynamic changes requires both a ventilator equipped with the end-expiratory occlusion function, which are not widely available in operating rooms, and a continuous cardiac output monitoring.

Some limitations of this work should be acknowledged. First, fluid responsiveness and fluid challenge assessment are closely related. Changing the amount of fluid infused, the rate of administration and the thresholds to define the response would affect the number of fluid responders and, as a consequence, the ROC curve analysis. We chose a 250-ml fluid challenge administered over 10 min, as proposed during intra-operative goal-directed fluid therapy and now suggested by international statements,2,40 and SVI at least 10% as target, while Myatra et al.21 adopted 7 ml kg−1 bolus in 10 min and CI more than 15%.22 This discrepancy would limit the comparability of the studies. Second, EEOT has been proposed as a 15-s hold in critically ill patients.17–19 We used a 30-s interruption as previously described in surgical patients,7 which corresponds to the MostCare default time-setting modality for averaging haemodynamic data. Although the haemodynamic effect of EEOT could change between the two time-points, the best thresholds for both ΔCI and ΔSVI (about 4%) are consistent with those reported in surgical and critically ill patients for a 30-s or a 15-s EEOT (4% to 5%) and the ROCs are comparable.7,18,21 Thirdly, the accuracy of MostCare is highly dependent on the quality of the signal obtained from the arterial catheter.25 Artefacts related to under/overdamping and to the transmission of the signal may influence the reliability of the device and some expertise is needed to correctly recognise these erroneous patterns.23 For this reason, the reliability of the MostCare remains operator-dependent and the centre involved in the study is highly trained in the use of this device. Fourthly, at least two exclusion criteria for this study (i.e. the chronic use of beta-blocking agents and obesity) would limit the external validation of the results in high-risk surgical patients, requiring further specific investigations in this subgroup of patients. Finally, we used a 1-min delay between the EEOTs and the VTC to guarantee the return to the steady state before the application of the subsequent test in the protocol and to minimise the risk of bias due to a carry-over effect, which, cannot be completely excluded.

Conclusion

In neurosurgical patients undergoing protective ventilation, the baseline values of dynamic indexes are unsuitable in assessing preload dependence. In this setting, the changes in PPV and SVV obtained after VTC are reliable and comparable to the changes in CI and SVI obtained after EEOT performed at 8 ml kg−1 in predicting fluid responsiveness. These tests should be considered as adjuctive, well tolerated and useful methods to guide intra-operative fluid therapy.

Acknowledgements relating to this article

Assistance with the article: the authors acknowledge Dr. Victoria Bennett for the English editing.

Financial support and sponsorship: none.

Conflict of interest interests: PN's research laboratory has received equipment and grants from Maquet Critical Care, Draeger and Intersurgical S.P.A. He also received honoraria/speaking fees from Maquet Critical Care, Draeger, Breas, Philips, Resmed, Hillrom, MSD and Novartis. Dr. Navalesi contributed to the development of the helmet Next, whose license for patent belongs to Intersurgical S.P.A., and received royalties for that invention. Prof. Cecconi is a consultant for Edwards Lifesciences, LiDCO and Cheetah Medical. AM received travel expenses for scientific meetings from Vygon.

Presentation of preliminary data: 31st ESICM, Paris, France, 20 to 24 October 2018.

References

1. Thacker JK, Mountford WK, Ernst FR, et al. Perioperative fluid utilization variability and association with outcomes: considerations for enhanced recovery efforts in sample US surgical populations. Ann Surg 2016; 263:502–510.
2. Navarro LH, Bloomstone JA, Auler JO Jr, et al. Perioperative fluid therapy: a statement from the international Fluid Optimization Group. Periop Med 2015; 4:3.
3. Loftus TJ, Stelton S, Efaw BW, Bloomstone J. A system-wide enhanced recovery program focusing on two key process steps reduces complications and readmissions in patients undergoing bowel surgery. J Healthcare Qual 2017; 39:129–135.
4. MacDonald N, Ahmad T, Mohr O, et al. Dynamic preload markers to predict fluid responsiveness during and after major gastrointestinal surgery: an observational substudy of the OPTIMISE trial. Br J Anaesth 2014; 114:598–604.
5. Messina A, Longhini F, Coppo C, et al. Use of the fluid challenge in critically ill adult patients: a systematic review. Anesth Analg 2017; 125:1532–1543.
6. 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 2009; 37:2642–2647.
7. Biais M, Larghi M, Henriot J, et al. End-expiratory occlusion test predicts fluid responsiveness in patients with protective ventilation in the operating room. Anesth Analg 2017; 125:1889–1895.
8. Biais M, de Courson H, Lanchon R, et al. Mini-fluid challenge of 100 ml of crystalloid predicts fluid responsiveness in the operating room. Anesthesiology 2017; 127:450–456.
9. Cecconi M, Parsons AK, Rhodes A. What is a fluid challenge? Curr Opin Crit Care 2011; 17:290–295.
10. Pinsky MR. Heart lung interactions during mechanical ventilation. Curr Opin Crit Care 2012; 18:256–260.
11. Biais M, Ehrmann S, Mari A, et al. Clinical relevance of pulse pressure variations for predicting fluid responsiveness in mechanically ventilated intensive care unit patients: the grey zone approach. Crit Care 2014; 18:587.
12. Mahjoub Y, Lejeune V, Muller L, et al. Evaluation of pulse pressure variation validity criteria in critically ill patients: a prospective observational multicentre point-prevalence study. Br J Anaesth 2014; 112:681–685.
13. De Backer D, Heenen S, Piagnerelli M, et al. Pulse pressure variations to predict fluid responsiveness: influence of tidal volume. Intensive Care Med 2005; 31:517–523.
14. Futier E, Constantin JM, Paugam-Burtz C, et al. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med 2013; 369:428–437.
15. Eikermann M, Kurth T. Apply protective mechanical ventilation in the operating room in an individualized approach to perioperative respiratory care. Anesthesiology 2015; 123:12–14.
16. Pinsky MR, Payen D. Functional hemodynamic monitoring. Crit Care 2005; 9:566–572.
17. Monnet X, Osman D, Ridel C, et al. Predicting volume responsiveness by using the end-expiratory occlusion in mechanically ventilated intensive care unit patients. Crit Care Med 2009; 37:951–956.
18. Monnet X, Bleibtreu A, Ferre A, et al. Passive leg-raising and end-expiratory occlusion tests perform better than pulse pressure variation in patients with low respiratory system compliance. Crit Care Med 2012; 40:152–157.
19. Silva S, Jozwiak M, Teboul JL, et al. End-expiratory occlusion test predicts preload responsiveness independently of positive end-expiratory pressure during acute respiratory distress syndrome. Crit Care Med 2013; 41:1692–1701.
20. Guinot PG, Godart J, de Broca B, et al. End-expiratory occlusion manoeuvre does not accurately predict fluid responsiveness in the operating theatre. Br J Anaesth 2014; 112:1050–1054.
21. Myatra SN, Prabu NR, Divatia JV, et al. The changes in pulse pressure variation or stroke volume variation after a ‘tidal volume challenge’ reliably predict fluid responsiveness during low tidal volume ventilation. Crit Care Med 2017; 45:415–421.
22. Myles PS, Leslie K, McNeil J, et al. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet 2004; 363:1757–1763.
23. Romagnoli S, Ricci Z, Quattrone D, et al. Accuracy of invasive arterial pressure monitoring in cardiovascular patients: an observational study. Crit Care 2014; 18:644.
24. Scolletta S, Bodson L, Donadello K, et al. Assessment of left ventricular function by pulse wave analysis in critically ill patients. Intensive Care Med 2013; 39:1025–1033.
25. Scolletta S, Franchi F, Romagnoli S, et al. Comparison between Doppler-echocardiography and uncalibrated pulse contour method for cardiac output measurement: a multicenter observational study. Crit Care Med 2016; 44:1370–1379.
26. Biais M, Lanchon R, Sesay M, et al. Changes in stroke volume induced by lung recruitment maneuver predict fluid responsiveness in mechanically ventilated patients in the operating room. Anesthesiology 2017; 126:260–267.
27. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988; 44:837–845.
28. Cannesson M, Le Manach Y, Hofer CK, et al. Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a ‘gray zone’ approach. Anesthesiology 2011; 115:231–241.
29. Ray P, Le Manach Y, Riou B, Houle TT. Statistical evaluation of a biomarker. Anesthesiology 2010; 112:1023–1040.
30. Monnet X, Rienzo M, Osman D, et al. Passive leg raising predicts fluid responsiveness in the critically ill. Crit Care Med 2006; 34:1402–1407.
31. De Backer D, Taccone FS, Holsten R, et al. Influence of respiratory rate on stroke volume variation in mechanically ventilated patients. Anesthesiology 2009; 110:1092–1097.
32. Mahjoub Y, Pila C, Friggeri A, et al. Assessing fluid responsiveness in critically ill patients: false-positive pulse pressure variation is detected by Doppler echocardiographic evaluation of the right ventricle. Crit Care Med 2009; 37:2570–2575.
33. Soubrier S, Saulnier F, Hubert H, et al. Can dynamic indicators help the prediction of fluid responsiveness in spontaneously breathing critically ill patients? Intensive Care Med 2007; 33:1117–1124.
34. Messina A, Colombo D, Cammarota G, et al. Patient-ventilator asynchrony affects pulse pressure variation prediction of fluid responsiveness. J Crit Care 2015; 30:1067–1071.
35. 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 2010; 104:407–413.
36. Biais M, Stecken L, Ottolenghi L, et al. The ability of pulse pressure variations obtained with CNAP device to predict fluid responsiveness in the operating room. Anesth Analg 2011; 113:523–528.
37. Monnet X, Marik PE, Teboul JL. Prediction of fluid responsiveness: an update. Ann Intensive Care 2016; 6:111.
38. Lansdorp B, Hofhuizen C, van Lavieren M, et al. Mechanical ventilation-induced intrathoracic pressure distribution and heart-lung interactions. Crit Care Med 2014; 42:1983–1990.
39. Gattinoni L, Chiumello D, Carlesso E, Valenza F. Bench-to-bedside review: chest wall elastance in acute lung injury/acute respiratory distress syndrome patients. Crit Care 2004; 8:350–355.
40. Thiele RH, Raghunathan K, Brudney CS, et al. Correction to: American Society for Enhanced Recovery (ASER) and Perioperative Quality Initiative (POQI) joint consensus statement on perioperative fluid management within an enhanced recovery pathway for colorectal surgery. Periop Med 2018; 7:5.
© 2019 European Society of Anaesthesiology