Objectives: The low tidal volume ventilatory technique poses limitations on the application of dynamic indices, like pulse pressure and stroke volume variations, in determining fluid responsiveness. This study investigated whether shifts in pulse pressure variation and stroke volume variation following a tidal volume challenge could predict fluid responsiveness in liver transplant recipients facing low tidal volume. Materials and Methods: We conducted a prospective intervention study in liver transplant recipients ventilated with low tidal volume. Patients received tidal volume challenge (transiently increasing the tidal volume from 6 to 8 mL/kg predicted body weight) at a single time frame, when the patient developed hypotension before administration of fluid bolus or any vasopressor agents, and changes in pulse pressure variation and stroke volume variation were noted. Patients were classified as fluid responders or nonresponders based on whether fluid challenge resulted or did not result in increased stroke volume index of >10%.
Results: Among the 28 included patients, 12 (42.9%) were fluid responders and 16 (57.1%) were fluid nonresponders. During the tidal volume challenge, mean changes in pulse pressure and stroke volume variations were 2.00 ± 0.60 and 2.33 ± 0.77 in responders and 1.06 ± 0.44 and 0.87 ± 0.62 in nonresponders, respectively. Change in pulse pressure variation with area under the curve was 0.87 (95% CI, 0.73-1.00; P = .001), with optimal cut-off value of 1.5 indicating sensitivity and specificity of 83.3% and 87.5%; change in stroke volume variation with area under the curve was 0.91 (95% CI: 0.79-1.00, P = < .001), with the same cut-off value of 1.5 indicating sensitivity and specificity of 83.3% and 87.5%.
Conclusions: Changes in pulse pressure and stroke volume variations after tidal volume challenge have good predictability of fluid responsiveness in liver transplant recipients ventilated with lung protective ventilation strategy.

Key words : Intraoperative fluid administration, Low tidal volume ventilation, Orthotopic liver transplantation, Pulse pressure variation, Stroke volume variation

Introduction

Hemodynamic changes such as decreased systemic vascular resistance, increased cardiac output, central functional hypovolemia, increased arterial compliance, and peripheral vasodilatation associated with end-stage liver disease can complicate fluid management in liver transplant recipients.^1,2 These changes are crucial to consider during decisions on when and how much fluid should be administered, as less fluid can result in organ hypoperfusion and injury while excessive fluid results in pulmonary complications and graft congestion.³ In systemic reviews, Morkane and colleagues and Carrier and colleagues recommended that moderately restrictive or replacement only strategy, especially during the dissection phase of liver transplant and hypervolemia, based on absence of fluid responsiveness, should be avoided.^1,2 As a result, perioperative fluid management of liver transplant recipients is crucial for limiting these complications and enhancing patient outcomes.⁴
Previous investigations have demonstrated that just 50% of critically ill or surgical patients exhibit a favorable response to fluid administration. Consequently, the administration of fluids should be predicated on parameters capable of forecasting fluid responsiveness.⁵ Dynamic indices, including pulse pressure variation (PPV) and stroke volume variation (SVV), serve as superior predictors of preload and fluid responsiveness compared with static indices such as central venous pressure, even among liver transplant recipients during controlled mechanical ventilation at a tidal volume (TV) of no less than 8 mL/kg.^6-8
People with severe liver conditions may show lung-related problems, such as hepato-pulmonary syndrome, porto-pulmonary hypertension, hepatic hydrothorax, and spontaneous bacterial empyema, regardless of any previous pulmonary ailments.⁹ Because improved clinical outcomes are correlated with a lung-protective ventilatory strategy (TV not exceeding 6 mL/kg of predicted body weight [PBW]), this approach has been adopted as a standard practice within the operating room, even for liver transplant recipients. However, this ap-proach can limit the usefulness of PPV and SVV during lung protective ventilatory strategies.¹⁰
To address this limitation, the tidal volume challenge (TVC) has been proposed as a method for anticipating fluid responsiveness during low TV ventilation. Variations in dynamic indices, induced by a transient augmentation of the TV from 6 to 8 mL/kg for a duration of 1 minute, have been noted to predict fluid responsiveness in patients subjected to lung protective ventilation, particularly in critically ill individuals and neurosurgical patients who are given low TV ventilation.^11-19 However, limited data are available on the usefulness of TVC in patients undergoing liver transplant. The primary objective of this study was to investigate whether changes in PPV and SVV after TVC could reliably predict fluid responsiveness in patients undergoing liver transplant who are ventilated with lung protective ventilation.

Materials and Methods

Study design and setting
We conducted a prospective interventional single-center, single-arm study of data from May 1, 2022, through May 31, 2023. The study was conducted at Mahatma Gandhi Medical College and Hospital, a tertiary-level university teaching hospital in Jaipur, Rajasthan, India. After approval by the institutional ethics committee (MGMC&H/IEC/JPR/2022/682), this study was registered (CTRI/2022/04/042099).
Inclusion criteria
All included patients provided written informed consent. Our inclusion criteria were patients diag-nosed with liver cirrhosis and undergoing living donor liver transplant (aged 18-60 y) who required invasive arterial and central line monitoring and subsequently exhibited hypotension (defined as a reduction in systolic arterial pressure [SAP] of ≥20% or a mean arterial pressure [MAP] falling below 65 mm Hg) relative to preanesthetic induction measurements after induction of anesthesia.
Degree of relationship of living donor to recipient
Living liver transplant donors were aged ≥18 years; connection between donors and recipients was either first- or second-degree relatives or spouses. Furthermore, ethical approval from the relevant committee was secured before the commencement of the transplant process.
Exclusion criteria
We excluded patients with recurrent cardiac arrhythmias, diminished left ventricular ejection fraction of <40%, a body mass index of >30 kg/m², restrictive pulmonary disorders, and moderate to severe pulmonary hypertension. We also excluded patients who received vasopressors or inotropes before or during TVC, patients who had emergence of novel intraoperative arrhythmias, and patients with heart rate-to-respiratory rate ratio of <3.6.
Perioperative management
Standard intraoperative monitoring protocols were used in all study patients. These included measure-ments of heart rate, peripheral oxygen saturation, continuous electrocardiography, and noninvasive blood pressure measurements, with baseline para-meters duly documented. Once the preoxy-genation phase was finished, induction for general anesthesia was started; agents for anesthesia were fentanyl, propofol, and rocuronium (1 mg/kg). The main-tenance phase involved the administration of the inhalational agent isoflurane, alongside intra-venous fentanyl for analgesic purposes and atracurium infusion as a muscle relaxant. Bispectral index of patients was closely observed during the operation, aiming to keep levels within the 40 to 60 range for the entire surgery. As maintenance fluid, a balanced salt solution (Plasmalyte) was provided at a rate of 2 mL/kg/h.
Each patient received ventilation in the volume-control mode, ensuring that the TV was set at 6 mL/kg of PBW, and the positive end-expiratory pressure was established at 5 cmH₂O to keep peripheral oxygen saturation at >96%. Predicted body weight (kg) was calculated as follows: X + 0.91[height (cm) - 152.4], where X equals 50 for male patients and 45.5 for female patients. Once anesthesia was induced, the central venous line and the arterial line were set up. The FloTrac system (Edwards Lifesciences) was connected to the patient for ongoing hemodynamic assessment.
Study protocol
Patients received the TVC assessment at a singular temporal point, at the period when the patient exhibited hypotension (a reduction in SAP exceeding 20% from the baseline and MAP falling below 65 mm Hg) preceding the administration of a fluid bolus or any vasopressor medication. The square-wave test was used to evaluate potential under-damping or overdamping of the pressure signal before the TVC procedure. The hemodynamic parameters, comprising pulse rate, SAP, diastolic arterial pressure, MAP, central venous pressure, stroke volume index, SVV, and PPV, were thoroughly documented before TVC was initiated.
Tidal volume challenge was performed by transiently increasing the TV to 8 mL/kg PBW for 1 minute, after which hemodynamic parameters were recorded. Changes in values of PPV and SVV (ΔPPV_6-8 = PPV₈ – PPV₆ and ΔSVV_6-8 = SVV₈ – SVV₆) were calculated. After completion of TVC, TV was reduced back to 6 mL/kg PBW, and hemodynamic parameters were again measured after 1 minute (baseline 2).
Next, fluid challenge was performed by infusing 250 mL of Plasmalyte solution over a period of 10 minutes and hemodynamic parameters were again measured. Based on whether the fluid challenge increased stroke volume index by more than 10% or not, patients were categorized as responders or as nonresponders. For analysis, we used data from each enrolled patient’s initial fluid challenge. This protocol could be adjusted at the discretion of the attending anesthetist for patient safety.
Statistical analyses
We used SPSS version 26 (IBM Corp), RStudio Team (2020) (RStudio: Integrated Development for R. RStudio), and Stata statistical suite (launch 14, 2015) (StataCorp) for statistical analyses. We presented continuous variables as mean ± SD or as median and interquartile range (IQR) and categorical data as frequencies (percentage).
We used the χ² test or the Fisher exact test to assess categorical data and the independent t test or the Mann-Whitney U test to assess continuous variables, including demographic characteristics and hemodynamic parameters between the responder and nonresponder cohorts, contingent on the distribution of the data. Furthermore, we used the paired t test or the Wilcoxon signed rank test to analyze continuous variables within the responder and nonresponder groups.
We used receiver operating characteristic (ROC) curves, accompanied by area under the curve (AUC) with 95% CI, to evaluate and compare the diagnostic efficacy of various parameters for detecting fluid responsiveness. This included parameters such as PPV at a TV of 6 mL/kg of PBW, PPV at a TV of 8 mL/kg PBW, ΔPPV_6-8 (change in PPV after increase in TV from 6 to 8 mL/kg PBW), SVV at 6 mL/kg PBW, SVV at 8 mL/kg PBW, and ΔSVV_6-8 (change in SVV after increase in TV from 6 to 8 mL/kg PBW).
We determined sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, positive predictive value, negative predictive value, and misclassification rate for diagnostic variables. We established the most appropriate threshold for each diagnostic variable through use of the Youden index (sensitivity + specificity - 1).
Sample size estimation
For sample size determination, we used area under the ROC curve, in accordance with Messina and colleagues.¹⁴ By projecting an area under the ROC curve for ΔPPV TVC = 0.94, with a null hypothesis of 0.50 and a ratio of sample size in the negative to positive groups set at 1, we calculated the conclusive sample size to be 28 (aggregated for both groups), ensuring an 80% statistical power and a significance level of 5%. Statistical assessments were undertaken at a significance benchmark of 5%, with P < .05 indicating statistical significance.

Results

Patient characteristics
From May 1, 2022, through May 31, 2023, 36 suc-cessive liver transplant recipients were evaluated for eligibility. Two patients were excluded before enrollment, and 6 patients were excluded after enrollment, ultimately leaving 28 individuals suitable for analyses (Figure 1). Among the 28 included patients, 12 (42.9%) were found to be fluid responders and 16 (57.1%) did not respond to fluid. General demographic and hemodynamic data among the included transplant recipients are listed in Table 1. Except for the fluid responders, who exhibited lower Model for End-Stage Liver Disease-sodium and Child Turcotte-Pugh scores compared with the nonresponder group, the other baseline characteristics were similar.
Hemodynamic parameters
Hemodynamic parameters in the responder and nonresponder groups at baseline 1 (TV 6 mL/kg), after TVC (in which TV increased from 6 to 8 mL/Kg), after TVC was lowered back to baseline 2 (ie, from 8 to 6 mL/kg), and after fluid challenge are listed in Table 2. Hemodynamic parameters were compared between and within the fluid responder and nonresponder groups. All hemodynamic parameters at baseline 1 and baseline 2 were comparable between the 2 groups.
Effect of changes in pulse pressure variation and stroke volume variation after tidal volume challenge on predicting fluid responsiveness
Fluid responders showed ΔPPV_6-8 of 2.00 ± 0.60 and ΔSVV_6-8 of 2.33 ± 0.77 compared with 1.06 ± 0.44 and 0.87 ± 0.62 in nonresponders. The responder group showed average percent increases of ΔPPV_6-8 of 25 ± 8.23% and ΔSVV_6-8 of 30.88 ± 12.41% compared with 15.07 ± 6.61% and 11.96 ± 7.94% in the nonresponder group. This result indicated that these parameters have the ability to classify between fluid responders and nonresponders (Table 3).
Receiver operating characteristic curve analysis
In ROC curve analysis (Figure 2), both change and percent increases showed ability to predict fluid responsiveness: ΔPPV_6–8 with AUC = 0.87 (95% CI, 0.73-1.00; P = .001) with optimal cut-off value of 1.5 indicating sensitivity and specificity of 83.3% and 87.5% and ΔSVV_6-8 with AUC = 0.91 (95% CI, 0.79-1.00; P ≤ .001) with the same cut-off value of 1.5 indicating sensitivity and specificity of 83.3% and 87.5%. Furthermore, percent ΔPPV_6-8 with AUC = 0.85 (95% CI, 0.72-0.99; P = .002) with optimal cut-off value of 17.4 indicated sensitivity and specificity of 75%, and percent ΔSVV_6-8 with AUC = 0.92 (95% CI, 0.82-1.00, P ≤ .001) with cut-off value of 21.1 indicated sensitivity and specificity of 75% and 87.5%.

Discussion

The main finding of our study was that changes in PPV and SVV after TVC (ΔPPV_6-8 and ΔSVV_6-8 with cut-of value 2.00 ± 0.60 and 2.33 ± 0.77, respectively) accurately predicted fluid responsiveness in liver transplant recipients ventilated with lung protective ventilatory strategies, whereas baseline PPV and SVV did not. Our findings also showed that, although percent ΔPPV_6-8 and ΔSVV_6-8 after TVC could predict the fluid responsiveness, the results required additional computations.
In patients undergoing liver transplant, precise evaluations of preload status and optimal intraope-rative fluid administration are especially pivotal in mitigating complications and enhancing patient outcomes. In our investigation, 42.8% of subjects de-monstrated fluid responsiveness, which is congruent with previous studies.^5,11,14,15 This finding implied that functional hemodynamic assessments should be conducted to augment the predictive efficacy of PPV and SVV in patients receiving protective pulmonary ventilation in the surgical suite.
The postulation of TVC was initially examined by Myatra and colleagues in 20 critically ill patients.¹¹ The investigators found that low TV during lung protective ventilation was inadequate to effectuate substantial alterations in intrathoracic pressure, resulting in a spurious negative value of dynamic indices such as PPV and SVV. Thus, augmenting TV for a brief duration from 6 mL/kg to 8 mL/kg, that is, the TVC, should elevate PPV and SVV to varying degrees in patients who respond or do not respond to fluid strategies.
The research by Myatra and colleagues¹¹ highlighted that a notable change in both PPV and SVV when analyzed with TVC reliably predicted fluid responsiveness, with threshold markers set at 3.5% and 2.5% and area under the ROC values reaching 0.99 and 0.97. However, this previous study only analyzed 20 patients, whereas more patients were examined in our investigation; the larger number of patients and/or the patient’s diverse pathophysiological conditions in chronic hepatic disease could account for the inconsistency in threshold value and specificity between the studies. Likewise, numerous investigations in the surgical environment have substantiated our conclusions. Previous work discovered that alterations in the PPV subsequent to augmenting TVC from 6 to 8 mL/kg of PBW precisely forecasted fluid responsiveness in patients undergoing neurosurgical procedures, robotic surgery administered in the Trendelenburg posture, and thoracoscopic interventions involving unilateral lung ventilation.^13-15,20
The ΔPPV_6-8 with TVC in our investigation con-sistently forecasted fluid reactivity and exhibiting sensitivity and specificity of 83.3% and 87.5% respectively. These results were inferior to the findings of Myatra and colleagues¹¹ who reported 94% sensitivity and 100% specificity. In a study of neurosurgical patients, Messina and colleagues¹⁴ showed sensitivity of 94.7% and specificity of 76.1% for ΔPPV_6-8 with TVC, which surpassed the results of our investigation. Moreover, the threshold values of ΔPPV_6-8 and ΔSVV_6-8, as ascertained through ROC curve analysis, showed considerable divergence in our study when compared with other research investigations.^13,15,20 This discrepancy may be elucidated by the varying degree of hemodynamic effects elicited by the TVC in a patient with chronic liver disease undergoing hepatic transplant.
In our investigation, we found that PPV at 8 mL/kg TV does not consistently forecast fluid responsiveness, whereas SVV at 8 mL/kg TV does, which aligns with earlier research conducted in liver transplant.^6,7,21 This is because the values of PPV and SVV in our study and other investigations involving liver transplant patients were near the gray zone area between 9 and 13, rendering them inconclusive in their ability to anticipate preload response.²² This observation further elucidates the significance of functional hemodynamic assessments like TVC to enhance the predictive value of PPV and SVV.
In settings with constrained resources and where sophisticated hemodynamic monitoring tools may not be available, assessments such as TVC may be utilized to differentiate between fluid responders and nonresponders during lung protective ventilation, as indicated by Myatra and colleagues¹¹ and modestly reinforced by the outcomes of our investigation.
As we advance toward goal-directed fluid therapy using dynamic parameters such as PPV and SVV rather than conventional static parameters for fluid and hemodynamic management, the utilization of TVC permits the implementation of a low tidal ventilation strategy in liver transplant recipients. Such implementation could facilitate the reduction of perioperative fluid imbalances and pulmonary complications. Future studies involving a large patient cohort with prolonged follow-up are essential to validate these results.
Strength of the study
As pulmonary protective ventilation methodology transforms into a crucial component of intraoperative management, our investigation showed an initial endeavor to use TVC assessments to differentiate between fluid responders and nonresponders among liver transplant recipients.
Limitations of the study
The use of PPV is subject to limitations in specific patient populations, including those with arrhythmias, spontaneous respiratory efforts, and pneumope-ritoneum, thereby rendering the TVC-PPV evaluation unreliable in these contexts. Patients with PPV in the gray zone require further studies to confirm our findings. The amount and duration of fluid admi-nistration and TV to be used in fluid challenge and TVC, respectively, also warrant further research. To monitor variations in SVV, a continuous cardiac output monitoring device is required, which may be difficult to obtain in a resource-limited setting. Moreover, the intervention in our study was limited to a specific temporal context; the intraoperative phase was inherently dynamic and prone to sudden variations. Consequently, the single-center design of our study with a restricted sample size also warrants additional investigation with a more extensive cohort.

Conclusions

The changes in PPV and SVV after TVC showed good predictability of fluid responsiveness in liver transplant recipients ventilated with lung protective ventilation strategy.

References:

1. Morkane CM, Sapisochin G, Mukhtar AM, et al; ERAS4OLT.org Working Group. Perioperative fluid management and outcomes in adult deceased donor liver transplantation: a systematic review of the literature and expert panel recommendations. Clin Transplant. 2022;36(10):e14651. doi:10.1111/ctr.14651
   CrossRef - PubMed
2. Carrier FM, Chassé M, Wang HT, Aslanian P, Bilodeau M, Turgeon AF. Effects of perioperative fluid management on postoperative outcomes in liver transplantation: a systematic review protocol. Syst Rev. 2018;7(1):180. doi:10.1186/s13643-018-0841-3
   CrossRef - PubMed
3. Kohli M, Garg N, Sindwani G, Tempe D, Pamecha V, Pasupuleti SS. Effect of positive cumulative fluid balance on postoperative complications after living donor liver transplantation: a retrospective analysis. Indian J Anaesth. 2021;65(5):383-389. doi:10.4103/ija.IJA_1457_20
   CrossRef - PubMed
4. Carrier FM, Chassé M, Wang HT, Aslanian P, Iorio S, Bilodeau M, Turgeon AF. Restrictive fluid management strategies and outcomes in liver transplantation: a systematic review. Can J Anaesth. 2020;67(1):109-127. doi:10.1007/s12630-019-01480-y
   CrossRef - PubMed
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(5):1532-1543. doi:10.1213/ANE.0000000000002103
   CrossRef - PubMed
6. Kim SH, Hwang GS, Kim SO, Kim YK. Is stroke volume variation a useful preload index in liver transplant recipients? A retrospective analysis. Int J Med Sci. 2013;10(6):751-757. doi:10.7150/ijms.6074
   CrossRef - PubMed
7. Su BC, Tsai YF, Cheng CW, Yu HP, Yang MW, Lee WC, Lin CC. Stroke volume variation derived by arterial pulse contour analysis is a good indicator for preload estimation during liver transplantation. Transplant Proc. 2012;44(2):429-432. doi:10.1016/j.transproceed.2011.12.037
   CrossRef - PubMed
8. Rudnick MR, De Marchi L, Plotkin JS. Hemodynamic monitoring during liver transplantation: a state of the art review. World J Hepatol. 2015;7(10):1302-1311. doi:10.4254/wjh.v7.i10.1302
   CrossRef - PubMed
9. Raevens S, Boret M, De Pauw M, Fallon MB, Van Vlierberghe H. Pulmonary abnormalities in liver disease: relevance to transplantation and outcome. Hepatology. 2021;74(3):1674-1686. doi:10.1002/hep.31770
   CrossRef - PubMed
10. Eikermann M, Kurth T. Apply protective mechanical ventilation in the operating room in an individualized approach to perioperative respiratory care. Anesthesiology. 2015;123(1):12-14. doi:10.1097/ALN.0000000000000707
    CrossRef - PubMed
11. Myatra SN, Prabu NR, Divatia JV, Monnet X, Kulkarni AP, Teboul JL. 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(3):415-421. doi:10.1097/CCM.0000000000002183
    CrossRef - PubMed
12. Yonis H, Bitker L, Aublanc M, et al. Change in cardiac output during Trendelenburg maneuver is a reliable predictor of fluid responsiveness in patients with acute respiratory distress syndrome in the prone position under protective ventilation. Crit Care. 2017;21(1):295. doi:10.1186/s13054-017-1881-0
    CrossRef - PubMed
13. Jun JH, Chung RK, Baik HJ, et al. The tidal volume challenge improves the reliability of dynamic preload indices during robot-assisted laparoscopic surgery in the Trendelenburg position with lung-protective ventilation. BMC Anesthesiol. 2019;19(1):142. doi:10.1186/s12871-019-0807-6
    CrossRef - PubMed
14. Messina A, Montagnini C, Cammarota G, et al. Tidal volume challenge to predict fluid responsiveness in the operating room: an observational study. Eur J Anaesthesiol. 2019;36(8):583-591. doi:10.1097/EJA.0000000000000998
    CrossRef - PubMed
15. Messina A, Montagnini C, Cammarota G, et al. Assessment of fluid responsiveness in prone neurosurgical patients undergoing protective ventilation: role of dynamic indices, tidal volume challenge, and end-expiratory occlusion test. Anesth Analg. 2020;130(3):752-761. doi:10.1213/ANE.0000000000004494
    CrossRef - PubMed
16. Shi R, Ayed S, Moretto F, et al. Tidal volume challenge to predict preload responsiveness in patients with acute respiratory distress syndrome under prone position. Crit Care. 2022;26(1):219. doi:10.1186/s13054-022-04087-w
    CrossRef - PubMed
17. Xu Y, Guo J, Wu Q, Chen J. Efficacy of using tidal volume challenge to improve the reliability of pulse pressure variation reduced in low tidal volume ventilated critically ill patients with decreased respiratory system compliance. BMC Anesthesiol. 2022;22(1):137. doi:10.1186/s12871-022-01676-8
    CrossRef - PubMed
18. Alvarado Sánchez JI, Caicedo Ruiz JD, Diaztagle Fernández JJ, et al. Predictors of fluid responsiveness in critically ill patients mechanically ventilated at low tidal volumes: systematic review and meta-analysis. Ann Intensive Care. 2021;11(1):28. doi:10.1186/s13613-021-00817-5
    CrossRef - PubMed
19. Wang X, Liu S, Gao J, Zhang Y, Huang T. Does tidal volume challenge improve the feasibility of pulse pressure variation in patients mechanically ventilated at low tidal volumes? a systematic review and meta-analysis. Crit Care. 2023;27(1):45. doi:10.1186/s13054-023-04336-6
    CrossRef - PubMed
20. Zhang Y, Ding Y, Zhang J, Huang T, Gao J. Tidal volume challenge-induced hemodynamic changes can predict fluid responsiveness during one-lung ventilation: an observational study. Front Med (Lausanne). 2023;10:1169912. doi:10.3389/fmed.2023.1169912
    CrossRef - PubMed
21. Gouvêa G, Diaz R, Auler L, Toledo R, Martinho JM. Evaluation of the pulse pressure variation index as a predictor of fluid responsiveness during orthotopic liver transplantation. Br J Anaesth. 2009;103(2):238-243. doi:10.1093/bja/aep123
    CrossRef - PubMed
22. 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(2):231-241. doi:10.1097/ALN.0b013e318225b80a
    CrossRef - PubMed