key: cord-0017749-23z1qe49
authors: Berg, Jonas M.; Nielsen, Dorthe V.; Abromaitiene, Vijoleta; Hjørnet, Niels E.; Vistisen, Simon T.
title: Changes in arterial blood pressure characteristics following an extrasystolic beat or a fast 50 ml fluid challenge do not predict fluid responsiveness during cardiac surgery
date: 2021-05-26
journal: J Clin Monit Comput
DOI: 10.1007/s10877-021-00722-z
sha: bb9c465014939b4de6315514664c45c889b108b0
doc_id: 17749
cord_uid: 23z1qe49

ABSTRACT: Prediction of fluid responsiveness is essential in perioperative goal directed therapy, but dynamic tests of fluid responsiveness are not applicable during open-chest surgery. We hypothesised that two methods could predict fluid responsiveness during cardiac surgery based on their ability to alter preload and thereby induce changes in arterial blood pressure characteristics: (1) the change caused by extrasystolic beats and (2) the change caused by a fast infusion of 50 ml crystalloid (micro-fluid challenge). Arterial blood pressure and electrocardiogram waveforms were collected during surgical preparation of the left internal mammary artery in patients undergoing coronary artery bypass surgery. Patients received a fluid challenge (5 ml/kg ideal body weight). The first 50 ml were infused in 10 s and comprised the micro-fluid challenge. Predictor variables were defined as post-ectopic beat changes (compared with sinus beats preceding ectopy) in arterial blood pressure characteristics, such as pulse pressure and systolic pressure, or micro-fluid challenge induced changes in the same blood pressure characteristics. Patients were considered fluid responsive if stroke volume index increased by 15% or more after the full fluid challenge. Diagnostic accuracy was calculated by the area under the receiver operating characteristics curve (AUC). Fifty-six patients were included for statistical analysis. Thirty-one had extrasystoles. The maximal AUC was found for the extrasystolic change in pulse pressure and was 0.70 (CI [0.35 to 1.00]). The micro-fluid challenge method generally produced lower AUC point estimates. Extrasystoles did not predict fluid responsiveness with convincing accuracy in patients undergoing cardiac surgery and changes in arterial waveform indices following a micro-fluid challenge could not predict fluid responsiveness. Given a low number of fluid responders and inherently reduced statistical power, our data does not support firm conclusions about the utility of the extrasystolic method. CLINICAL TRIAL REGISTRATION: Unique identifier: NCT02903316. https://clinicaltrials.gov/ct2/show/NCT02903316?cond=NCT02903316&rank=1. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s10877-021-00722-z.

Haemodynamic instability is often treated with fluids, but fluids have side effects and should only be considered in fluid responsive patients [1, 2] , i.e. patients whose stroke volume will significantly increase with fluid infusion.

Static indicators for fluid responsiveness prediction (e.g. central venous pressure) have all proved unreliable [3] [4] [5] . Some protocols for major surgery dictate systematic fluid challenges of 100-500 ml to guide and titrate fluid administration [6, 7] . Such a fluid strategy entails a risk of overhydration, which is associated with increased postoperative complications [8] . In contrast, ventilator-induced dynamic variables such as pulse pressure variation (PPV) and stroke volume variation (SVV), and passive leg raising (PLR) [4, [9] [10] [11] have consistently proven reliable to predict fluid responsiveness. Common to these techniques is that preload fluctuations are utilized to predict fluid responsiveness.

However, during open-chest surgery, sternotomy and pericardiotomy during heart surgery changes the complex physiology of ventilator-induced heart-lung interactions by alterations to the interplay between preload, afterload and aortic compliance rendering dynamic preload variables less reliable [12, 13] .

Thus, ventilator-induced indicators are less reliable in this setting and the classical PLR manoeuvre is hardly applicable during open-chest surgery [9, [13] [14] [15] [16] [17] , although some studies altering the positioning of the patient have been performed at the end of cardiac surgery after chest closure [18, 19] .

Therefore, a reliable technique for prediction of fluid responsiveness is missing for cardiac surgical patients during the open-chest condition. Still, the convincing results of PLR and dynamic variables in other settings indicate that the way forward is to look for alternative preload altering mechanisms.

Extrasystoles comprise a preload altering mechanism. The heartbeat following the extrasystolic beat, i.e. the post-ectopic beat, has a prolonged RR-interval (filling time) which results in increased preload compared with preceding sinus beats [20, 21] . Recent experimental and clinical studies have showed promising results using extrasystoles for fluid responsiveness prediction [22] [23] [24] but a study during general anaesthesia did not provide similar results [25] . Yet, the approach has not been investigated in general anaesthesia with thermodilution cardiac output as outcome. Furthermore, recently a micro-fluid challenge (MFC, 50 ml infused within 10 s) showed interesting results regarding fluid responsiveness prediction in the intensive care unit (ICU) when assessed by changes in echocardiographic measurements [26] .

Accordingly, the present study aimed to validate the predictive value of both the extrasystoles and the micro-fluid challenge methods in open cardiac surgery with closed pericardium. We hypothesised that these two different methods could predict fluid responsiveness in anaesthetized patients undergoing coronary artery bypass graft surgery.

This prospective single-center study was approved by the Danish Data Agency (1-16-02-316-16) , registered at clinicaltrials.gov (NCT02903316) prior to initiation and conducted at Aarhus University Hospital. The study is reported in alignment with STARD guidelines [27] .

Patients scheduled for elective on-pump coronary artery bypass graft surgery without additional surgical procedures were screened for exclusion criteria: ejection fraction < 35%, haemodialysis, major cognitive dysfunction, age < 18, pregnancy and cardiac arrhythmia, e.g. atrial fibrillation.

Patients were equipped with plethysmography, electrocardiogram and radial arterial blood pressure monitoring. Furthermore, patients were catheterized with a pulmonary artery catheter (Swan-Ganz CCOmbo CCO/SvO 2 , 744F75, Edwards Lifesciences, Irvine, California, US) for monitoring of continuous cardiac output (CO) and mixed venous oxygen saturation (SvO 2 ).

Patients were anaesthetized with intravenous infusion of propofol and sufentanil and controlled mechanically ventilated. Figure 1 shows the study timeline. Patients were observed for extrasystoles during preparation of the left internal mammary artery. Observation began after sternotomy and insertion of the sternal retractor and continued until approximately 10-15 min prior to pericardial opening estimated by the surgeon. Following the observation period, a predetermined volume of 5 ml/kg (ideal body weight) fluid (Ringer's acetate, B. Braun, Melsungen, Germany) was manually infused with a 50-ml syringe in the central venous catheter. Ideal body weight was calculated from height and sex; men receiving 5 ml per cm above 100 cm and women 5 ml per cm above 105 cm. The first 50 ml of the total volume were infused within 10 s (the MFC; described below). A timing marker was added to the acquired waveform data simultaneously with infusion start. The rest of the total fluid challenge was subsequently infused within approximately 5 min. Three consecutive continuous cardiac index (CCI) and SvO 2 measurements were read off the monitor trend data before and approximately 5-7 min after completion of the fluid challenge allowing for delay in response time for CCI [28] . Three consecutive CCI measurements were obtained from the STAT display on the Vigilance II monitor, Edwards Lifesciences, using the Swan-Ganz's continuous thermodilution method. Waveforms for arterial blood pressure, pulmonary artery pressure, plethysmography, electrocardiogram and central venous pressure were extracted from the patient's monitor (Philips MP70, Philips Healthcare) by commercially available software (ixTrend, iXellence GmbH, Wildau, Germany) for offline analysis. All waveforms were sampled at 125 Hz except the electrocardiogram, which was sampled at 500 Hz. During surgery, patients were excluded if (1) there were other haemodynamically significant changes during the study period e.g. changed infusion rates or boluses of sedatives, analgesics, inotropes, or vasopressor drugs or major changes of the bed position, (2) technical problems occurred e.g. placing of the pulmonary artery catheter (3) the attending anaesthesiologist deemed the scheduled fluid challenge inappropriate for any reason.

R spike detection was done semi-automatically in Kubios HRV (Kubios Oy, Finland) with visual inspection and correction. RR intervals were subsequently analysed in Matlab. Extrasystoles with at least 10 preceding sinus beats were included. Additionally, extrasystolic beats should at least have a 20% shortening of the RR interval (coupling interval) to induce a significant variation in cardiac preload compared to baseline RR intervals [22] . Potential eligible extrasystoles were visually examined by two investigators, who verified that the simultaneous arterial waveform was not damped or presented with other artefacts. Differences between the postectopic beat and the median of the 10 preceding sinus beats were calculated for pulse pressure (∆PP ES ), systolic blood pressure (∆SBP ES ), maximal pressure upstroke slope (∆dP/ dt ES ) and pre-ejection period (∆PEP ES = time from R spike in the electrocardiogram to arterial upstroke onset in arterial blood pressure). If multiple extrasystoles were eligible in the observation period, an average of the post-ectopic changes was calculated. Changes in PP, SBP and dP/dt were all calculated as relative changes whereas the ∆PEP (decrease) was calculated as absolute change (in ms) to omit influence of variation in vascular transit time across patients [29] . Further details of signal processing and detection of PEP have been described previously [22] . Stroke volume index (SVI) was calculated before and after the fluid infusion by dividing CCI with heart rate. Heart rate was manually verified from the electrocardiogram and pressure curves in combination.

It was tested if a micro-fluid challenge could induce changes in PP, SBP, dP/dt and PEP that could be used for fluid responsiveness prediction. ∆PP MFC , ∆SBP MFC , ∆dP/dt MFC and ∆PEP MFC were calculated by subtracting the upper quartile of the 20 s following the fluid infusion from the corresponding median of the preceding 30 s to allow for capturing a peak change in the variables.

All patients without excluding factors were used to test the micro-fluid challenge. From this cohort, all patients with extrasystoles present before fluid infusion were identified. Patients were classified as fluid responders if they had a 15% increase or more in SVI following the 5 ml/kg fluid challenge. Sample size for the extrasystole method (primary outcome) was calculated using previous data [23] , assuming equal numbers of fluid responders and non-responders. Using a significance level of 0.05 and power of 0.9, we needed 30 patients with extrasystoles. Prior to initiating the study, the expected occurrence of extrasystoles was determined retrospectively inspecting ectopic activity as captured by the Philips ® monitoring system from 59 cardiac surgeries. Approximately 70% of patients had at least one extrasystole, which was similar to a previous study [30] . Estimating conservatively, we therefore planned to include 60 patients.

In addition to using ∆SVI as the outcome variable, we also secondarily in retrospect analysed our data with ∆SvO 2 as outcome because SvO 2 may reflect changes in CO more rapidly during metabolic stable conditions. Patients were classified as fluid responders if ∆SvO 2 > 0 to catch subtle changes in tissue perfusion.

The researcher analysing the arterial waveforms and extracting predictors (STV) was blinded for the outcome, ∆SVI, whereas the researcher (JMB) calculating the ∆SVI was blinded for predictor indices. Data are reported as area under the receiver operating characteristics (ROC) curve (AUC) together with confidence intervals (95% CI) using deLong method. Optimal sensitivity and specificity were found using Youden Index. Spearman Rank Correlation ρ is reported. Demographic and clinical variables are presented as mean ± standard deviation (SD) except ventilation parameters, which are presented as median (interquartile range). Wilcoxon Mann-Whitney U Test was used to compare responders with non-responders. Categorical data was compared with Fisher's Exact Test due to low number of observations. All statistical calculations and figures were done with StataMP (version 13, StataCorp., College Station, TX, USA).

Patient inclusion is shown in Fig. 2 . Demographics and baseline cardio-respiratory characteristics are shown in Table 1 for the extrasystolic cohort (n = 31) and in Table 2 for the micro-fluid challenge cohort (n = 52). In both cohorts, fluid responders are compared with non-responders.

Out of 56 eligible patients, 31 had extrasystoles (55%). Responders and non-responders had similar background characteristics. Three patients out of the 31 (10%) were fluid responders. Patients received 385 (SD: 47) ml infusion. Post-ectopic changes in haemodynamic indices are plotted against ∆SVI in Fig. 3 and corresponding ROC curves can be seen in Fig. 4 .

The post-ectopic change in PP, SBP, dP/dt and PEP predicted fluid responsiveness with AUCs between 0.62 and 0.70. ∆PP ES predicted fluid responsiveness with an AUC of 0.70 (CI [0.35 to 1]). The optimal threshold for ∆PP ES was 25% resulting in sensitivity of 67% and specificity of 86% (ρ = 0.23, p = 0.21). ∆SBP ES had an AUC of 0.65 (CI [0.23 to 1]). ∆dP/dt ES and ∆PEP ES had lower AUCs (see detailed data in Fig. 4 and Table 3 ).

The micro-fluid challenge cohort was based on the 56 eligible patients but four of these datasets were uncertain regarding exact indication of start of fluid infusion in the curve data and were thus excluded. Out of the 52 remaining patients, 29 were in the extrasystolic cohort as well. Seven of the 52 patients (13%) were fluid responders. The haemodynamic changes following the micro-fluid challenge are plotted against ∆SVI in Fig. 5 . All micro-fluid challenge measures had ROC curves following the identity line (Fig. 6 ). All indices included 0.5 in their confidence interval. Optimal cut-off values and corresponding sensitivity and specificity are presented in Table 3 for indices with AUCs above 0.5.

AUCs for baseline static variables (central venous pressure, mean arterial pressure and SvO 2 ) are presented in Table 3 . SvO 2 rose modestly (0.6%) but significantly following the fluid infusion in the entire group (p < 0.0122, n = 54). There was no significant difference in ∆SvO 2 between fluid responders compared with non-responders defined by ∆SVI (p < 0.183).

ROC curves for all measures of the extrasystole method and micro-fluid challenge using ∆SvO 2 as the outcome variable did not show better classification (for detailed classification data see supplementary material).

The post-ectopic changes in four blood pressure characteristics did not provide a clinically satisfactory prediction of fluid responsiveness. The best predictive indicator was ∆PP ES with AUC of 0.70 (0.35 to 1). All other indices tested (SBP ES , dP/dt ES and PEP ES ) had similar findings with lower AUCs. Although we did not find any significant results Table 2 Background characteristics for micro-fluid challenge cohort Data are presented as mean ± standard deviation (SD) except ventilation parameters, which are presented as median (interquartile range) HR heart rate, SBP systolic blood pressure, MAP mean arterial pressure, CCI continuous cardiac index, LVEF left ventricle ejection fraction, FiO 2 fraction of inspiratory oxygen *p < 0.05 responders compared to non-responders All participants (n = 52) Responders (n = 7) Non-responders (n = 45) p-value for the extrasystole method, all AUC point estimates were slightly above 0.5. This may support the physiological concept, but in a specific patient during cardiac surgery, the extrasystolic method does not seem applicable. Still, only three patients in the extrasystolic cohort (10%) responded with at least 15% increase in SVI following the fluid challenge. This proportion was not anticipated and not in accordance with the assumptions for our power calculation. The micro-fluid challenge combined with velocity time integral measures was a reliable method to predict fluid responsiveness in an ICU study [26] . We wanted to make the micro-fluid challenge method less operator dependent and more applicable in the operating theatre by analysing arterial blood pressure characteristics as predictors instead of using echocardiography. None of our four arterial blood pressure derived indices revealed a predictive value as all of them had AUC point estimates around 0.5 or less.

Despite the low number of responders, our AUC estimates for the extrasystole method shows similar predictive values as reported in a comparable general anaesthesia study [25] but inferior value compared with previous studies in the post-operative cardio-thoracic ICU and a more general ICU, where sensitivities and specificities around 80% were found for ∆PEP ES , and ∆SBP ES [23, 24] . All our patients had low heart rates during anaesthesia (mean: 58 beats/min), likely related to anaesthesia and possibly the prevalent use of β-blockers before surgery. This is similar to a comparable general anaesthesia study [25] also associated with mediocre classification, but contrasted by the heart rates generally encountered in the ICU, where the method has been more reliable. Prolongation of the RR interval might entail relatively less preload increase at the post-ectopic beat due to the relatively long passive filling time already present for the regular sinus beats.

The micro-fluid challenge method did not show any predictive value, which was surprising considering Wu et al.'s results [26] . Wu et al. studied another patient population and used velocity time integrals to assess the micro-fluid challenge induced fluid response as opposed to our use of arterial blood pressure characteristics.

A limitation for our study is the low proportion of fluid responders, which is not in alignment with previous studies. During open thoracic surgery a study found 37% fluid responders [31] and for a heterogeneous surgical population, the proportion was even higher [32] . Several reasons might explain the low proportion of fluid responders. First, our patients received less fluid (385 ± 47 ml) compared with other fluid responsiveness studies [32] . However, this amount should be enough for increasing preload [6] as goal directed therapy studies often use 250 ml as their intervention [33] . Another explanation could be related to protocol timing with the open-chest condition induced immediately prior to our observation window. Open-chest results in decreased central venous pressure [13, 34] , thus increasing the venous return flow, which might reduce preload responsiveness probably due to a rightward shift on the Frank-Starling relation between ventricular preload and stroke volume [11, 12] . Indeed, we observed a higher response rate in similar patients in the subsequent post-operative setting [23] . This, together with the fact that our patients were not extremely hypovolemic might cause the low number of fluid responders. Additionally, the low fluid response rate may be related to the observational nature of the study: The attending anaesthesiologist changed infusion rates or gave boluses of vasodilating agents in some excluded patients (Propofol, Sevoflurane or Sufentanil), in some cases due to a blood pressure increase before aortic cannulation. Although speculative, these few cases may have been fluid responders if not excluded during infusion. In addition, we cannot exclude that due to the nature of cardiac procedures and high dose opioid anaesthesia, SVI in itself might decrease over observation time.

Finally, the use of Swan-Ganz for CO measurements might partly explain the low proportion of fluid responders. Swan-Ganz is the gold standard when precise CO measurements are needed. However, when using the continuous CO modality, Haller et al. showed that following an intervention, the detection of the change in CO is delayed [28] . The proprietary algorithm reducing thermal noise in the vigilance II system might have stabilised our CO measurements and delayed the detection of the fluid response. The delay confirmed by Edwards Tech support should be 6-10 min following an intervention. Our data collection was designed to account for this, and only if the response time was longer than 10 min, would a partial and not the entire fluid response have been detected. Alternatively, if a (positive) stroke volume/cardiac output response was very transient (i.e. few minutes), we would deem our study design to be problematic but in a previous study, the response to much slower fluid infusions lasted and was captured by the CCO method [29] . We therefore speculate that clinical rather than technical circumstances gave rise to a limited number of fluid responders.

Using ∆SvO 2 as the outcome variable instead of ∆SVI in our analyses showed similar classification results and thus did not indicate better prediction with either of the two fluid responsiveness monitoring concepts under investigation. This, supports our interpretation of a low predictive value of the methods during cardiac surgery and also an actual, albeit unexpected, low number of fluid responders.

Considering the basis of our power calculation, we calculated in retrospect the power of the present study based on the extrasystolic results for pulse pressure (with ROC area at 0.70). The power is 0.21. We also did a new power calculation based on this data set (AUC = 0.70, alpha = 0.05, power = 0.8 and a distribution of responders/non-responders similar to this study). A study would then need to include approximately 180 patients to show that the ROC area is indeed different from 0.5. Given the observed distribution of responders and non-responders (3 and 28), the minimal detectable AUC could also be calculated, and it is 0.927 (and 0.813 for the micro-fluid challenge method). If an acceptable AUC range is defined between 0.8 and 0.9, then, strictly speaking, the present study can only conclude that the micro-fluid challenge based methods are probably not acceptable. As a consequence, firm conclusions for the extrasystole method are not possible to draw from this statistical point of view. Yet, this statistical aspect should also be seen in the context of the remaining results and the clinical expectations to a fluid responsiveness test. All of the variables (morphologic Fig. 6 Receiver operating characteristic (ROC) curves for micro-fluid challenge indices (PP pulse pressure, SBP systolic pulse pressure, ∆dP/dt maximal pressure upstroke slope and PEP pre-ejection period). Area under the ROC-curve (AUC) are presented below each panel extrasystolic and micro-fluid challenge features) under investigation were producing a point estimate for the ROC area far from acceptable for clinical application. Also, as opposed to our existing extrasystole studies, more patients with e.g. an SBP ES beyond 5% were non-responders in this study compared with the previous studies. This strongly indicates that the previously suggested threshold of e.g. 5% for SBP ES in other clinical contexts is not optimal in this context of cardiac surgery-despite the low statistical power to make overall firm conclusions about the extrasystole method.

In conclusion, post-extrasystolic and micro-fluid challenge induced changes in the arterial blood pressure curve had low predictive values for predicting fluid responsiveness. Given a low number of fluid responders and inherently reduced statistical power, our data does not support firm conclusions about the utility of the extrasystolic method. Still, our point estimates, supports the evidence that dynamic preload indicators are inappropriate to predict fluid responsiveness in the open-chest condition. Thus, systematic administration and evaluation of a fluid challenge remains the best way to assess-but not predict-fluid responsiveness in patients undergoing cardiac surgery.

The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s10877-021-00722-z.

Author contributions JMB: Designed and planned the study, collected data, made statistical analysis, wrote the first manuscript draft and approved the final version. DVN: Designed and planned the study, collected data, and approved the final version. VA: Helped designing the study, collected data and approved the final version. NEH: Helped designing the study, collected data and approved the final version. STV Designed and planned the study, waveform data analysis, statistical analysis and approved the final version.

Funding JMB was financially supported by grants from Lippmann foundation, Etatsraad C. G. Filtenborg and hustru Marie Filtenborg foundation, and Helga and Peter Korning's foundation. STV was financially supported by a grant from the Independent Research Fund Denmark (DFF -4183-00540).

Data availability Raw data sets are not available due to lack of patient consent and lack of approval from the Danish Data Agency. Statistic codes used to process data can be requested from the corresponding author upon reasonable request.

The authors declare that they have no conflict of interest.

Ethical approval All procedures performed in this study were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments. The Danish regional ethics committee classified the study as observational because the fluid infusion was considered a systematic framework of clinical practice.

Informed consent Written informed consent for the use of data was obtained before surgery for all individual participants included in the study.

Perioperative fluid therapy: a statement from the international Fluid Optimization Group

Wet, dry or something else?

Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence

Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature

Does the central venous pressure predict fluid responsiveness? An updated meta-analysis and a plea for some common sense

What is a fluid challenge

Optimisation of perioperative cardiovascular management to improve surgical outcome II (OPTIMISE II) trial: study protocol for a multicentre international trial of cardiac output-guided fluid therapy with low-dose inotrope infusion compared with usual care in patients undergoing major elective gastrointestinal surgery

Perioperative cardiovascular monitoring of high-risk patients: a consensus of 12

Passive leg raising: five rules, not a drop of fluid! Crit Care

Diagnostic accuracy of passive leg raising for prediction of fluid responsiveness in adults: systematic review and meta-analysis of clinical studies

Changes in arterial pressure during mechanical ventilation

Limitations of arterial pulse pressure variation and left ventricular stroke volume variation in estimating cardiac pre-load during open heart surgery

Dynamic preload indicators fail to predict fluid responsiveness in open-chest conditions

Passive leg raising for predicting fluid responsiveness: a systematic review and meta-analysis

Pulse pressure variations to predict fluid responsiveness: influence of tidal volume

Dynamic indices do not predict volume responsiveness in routine clinical practice

Prediction of fluid responsiveness: an update

Prediction of fluid responsiveness in patients during cardiac surgery

End-of-procedure volume responsiveness defined by the passive leg raise test is not associated with acute kidney injury after cardiopulmonary bypass

Experimental studies on myocardial contractility and hemodynamics in extrasystoles

The influence of left ventricular filling on postextrasystolic potentiation in the dog heart

Variations in the pre-ejection period induced by ventricular extra systoles may be feasible to predict fluid responsiveness

Using extra systoles to predict fluid responsiveness in cardiothoracic critical care patients

Extrasystoles for fluid responsiveness prediction in critically ill patients

Using extra systoles and the micro-fluid challenge to predict fluid responsiveness during cardiac surgery

A 10-second fluid challenge guided by transthoracic echocardiography can predict fluid responsiveness

STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies

Evaluation of a new continuous thermodilution cardiac output monitor in critically ill patients: a prospective criterion standard study

Automated pre-ejection period variation indexed to tidal volume predicts fluid responsiveness after cardiac surgery

Prevalence and temporal distribution of extrasystoles in septic ICU patients: the feasibility of predicting fluid responsiveness using extrasystoles

Stroke volume variation and pulse pressure variation are not useful for predicting fluid responsiveness in thoracic surgery

Fluid challenge during anesthesia: a systematic review and meta-analysis

Effect of a perioperative, cardiac output-guided hemodynamic therapy algorithm on outcomes following major gastrointestinal surgery: a randomized clinical trial and systematic review

Prediction of volume response under open-chest conditions during coronary artery bypass surgery