Secondary Logo

Journal Logo

Editorial

Predicting fluid responsiveness

Does it answer the right question?

van Eijk, Lucas T.; Servaas, Sjoerd; Slagt, Cor; Malagon, Ignacio

Author Information
European Journal of Anaesthesiology: May 2021 - Volume 38 - Issue 5 - p 449-451
doi: 10.1097/EJA.0000000000001455
  • Free

This Editorial accompanies the following original article: Kim E-H, Lee J-H, Jang Y-E, et al. Prediction of fluid responsiveness using lung recruitment manoeuvre in paediatric patients receiving lung-protective ventilation: a prospective observational study. Eur J Anaesthesiol 2021; 38:452–458.

‘Prediction is very difficult, especially if it's about the future’

Niels Bohr

Few subjects have been as thoroughly studied as fluid administration to critically ill patients or those undergoing surgery and yet the pathophysiology of fluid administration is not totally elucidated. Over time the importance of prevention of fluid overload has emerged, with emphasis on restricting fluid administration to the occasions where objective measures indicate that a fluid bolus will result in an increase in cardiac output, a condition termed fluid responsiveness. By restriction of fluids to cases where patients are predicted to be fluid responsive, fluid overload may be prevented, as it increases mortality.1–3 In this issue, Kim et al.4 describe whether a lung recruitment manoeuvre (inspiratory hold at 25 cm H2O for 20 s) would augment the fluid responsiveness predictability of several dynamic indices in mechanically ventilated children following cardiac surgery. The authors found that a lung recruitment manoeuvre could not improve the ability of the plethysmographic variability index (PVI) to predict fluid responsiveness. Although a negative study with regard to their hypothesis, their study once again identifies PVI as being a reliable predictor of fluid responsiveness in children. It is still unclear from this study and many others investigating fluid responsiveness how the identification of fluid responsiveness should be managed.

Numerous ways of predicting fluid responsiveness have been studied in adults. Static parameters, which are based on a single observation in time (e.g. blood pressure, central venous pressure and pulmonary artery occlusion pressure), generally do not allow adequate prediction of fluid responsiveness. However, some dynamic parameters that reflect variations over time in response to mechanical ventilation or passive leg raising, such as pulse pressure variation (PPV) and stroke volume variation, may be useful in adults.5,6 In children, however, prediction of fluid responsiveness can be even more challenging, most notably due to differences in lung, vascular and cardiac compliance compared with adults. In a recent consensus conference evaluating haemodynamic monitoring of critically ill children,7 the use of ultrasound/Doppler technology to measure cardiac output was strongly recommended. There was no clear recommendation on what method of monitoring should be used to track fluid responsiveness.

A meta-analysis concluded that plethysmographic waveform amplitude (ΔPOP) and PVI were equally effective for predicting fluid responsiveness in ventilated adult patients.8 ΔPOP and PVI were able to predict changes in cardiac index, stroke volume or stroke volume index. A responder was defined by a 15% change in any of the previous variables. Unfortunately, the method used to measure cardiac output was not uniform (thermodilution, transthoracic echocardiography, oesophageal Doppler and arterial pulse wave form). An error of 13% when measuring cardiac output with a thermodilution method is acceptable. This figure is relevant in itself. When Bland and Altman proposed in 1986 to use bias and precision (mean difference and standard deviation of the mean) to compare a new method of clinical measurement with the gold standard, they solved a problem and created a new one. When confronted with bias and precision in a new method of clinical measurement, a clinician must decide if these values are acceptable. To help in the decision making, the concept of percentage error was proposed.9 The consensus is that a percentage error of 30% is acceptable. A meta-analysis assessing the bias and precision of oesophageal Doppler in the paediatric population found an average bias of 10% and precision of 30% (no percentage error was documented).10

Further limitations to the use of noninvasive PPV include the presence of arrhythmias, the use of vasoconstrictors, the presence of vasoplegia and cardiac anomalies, such as tricuspid insufficiency. Low cardiac output and the use of spontaneous ventilation negatively affect its accuracy. Even the contacting force applied by the device may affect the ability to predict fluid responsiveness.11

A meta-analysis reviewed the literature in which prediction of fluid responsiveness in children was investigated.12 The authors concluded that respiratory variation in aortic blood flow peak velocity was the only variable shown to predict fluid responsiveness in children. The evidence for dynamic variables based on plethysmography was inconclusive.

Being fluid responsive does not need treatment per se

Not all patients who are fluid responders necessarily require volume expansion. During normal physiological conditions, humans are fluid responders and this condition should be cherished. The fact we normally function on the steeper part of the Frank–Starling curve illustrates the reserve capacity of dealing with fluid overload. This reserve capacity is limited, as by nature humans have not evolved to deal with hypervolaemia, which is mostly an iatrogenic condition. This curve is superimposed with the Marik–Phillips curve of extravascular lung water (EVLW) as a function of preload, illustrating the increase in EVLW in response to the increased cardiac filling pressures and transmitted hydrostatic pressures (Fig. 1).13 In other words, by fluid loading we accept an increase in EVLW and tissue oedema. As neither curve is linear, this is most notably the case in patients who function on the plateau of the Frank–Starling curve. Therefore, a point on the steeper part of the Frank–Starling curve should be pursued unless this would result in compromised tissue perfusion.

Fig. 1
Fig. 1:
Superimposition of the Frank–Starling and Marik–Phillips curves demonstrating the effects of increasing preload on stroke volume and extravascular lung water (EVLW)

Although not a primary endpoint, Kim et al. did not find any differences in vasoactive and inotropic support between the responders and nonresponders. Significant reductions in blood pressure and increases in central venous pressure during the lung recruitment manoeuvre were observed in both responder and nonresponder groups. Studies in the paediatric population using this technology with clinical outcome measures as primary endpoint are lacking. Those performed in adults are inconclusive.14

Choosing the time to evaluate the effect of fluid loading

Classically, fluid loading should be performed over a short period of time (10 to 15 min) to be able to evaluate its immediate effect. An increase in stroke volume of more than 10 to 15% is considered an adequate response to fluids.15 Typically, this evaluation takes place directly after administration of the desired fluid bolus. These actions are well described and universally adopted. However, one could consider this a rather primitive way of assessing fluid responsiveness. For the patient, of more relevance is what will happen in the period after fluid loading. In current literature, it is not well defined how a relapse after an initial successful fluid bolus should be evaluated and managed. This is even more relevant in the light of the short intravascular half-life of crystalloids, which is often ignored in clinical management. In healthy volunteers, the intravascular half-life of crystalloids is estimated to be around 30 min and is probably comparable in children.16 In addition, the extravascular transit time may vary with the clinical condition of the patient. For example, intravascular half-life is increased during mechanical ventilation and it may be decreased during severe inflammatory states, such as sepsis, pancreatitis, acute respiratory distress syndrome and burns.16 Interestingly, these conditions shift the Marik–Phillips curve to the left (Fig. 1), characterised by an increase in EVLW at equal filling pressures. For these patients, fluid management based on prediction of fluid responsiveness is not sufficient. Unfortunately, clinical trials comparing fluids with alternative treatments, such as low-threshold administration of inotropic drugs, are lacking. With regard to the evaluation of fluid responsiveness, at least the effect over time should be taken into account in clinical management and future research.

Controlling the endothelial barrier

From a pathological point of view, questioning whether there is an absolute volume deficit or a shift of water across the endothelial barrier is a hypothesis that deserves exploring. Ideally, only a deficit should be treated with fluids and a shift by controlling endothelial permeability. Interestingly, fluid loading may promote capillary leakage by damaging the glycocalyx.17 Major surgery, cardiopulmonary bypass and major bleeding cause a systemic inflammatory reaction which is associated with increased capillary permeability. Controlling the endothelial barrier will form a major leap forward in the clinical management of inflammation-induced low-cardiac output states. Microvascular permeability is controlled by the endothelial angiopoietin/vascular endothelial growth factor/Tie2 system. Recently, an angiopoietin-1 mimetic called vasculotide was shown to reduce vascular leakage in a murine model of cardiopulmonary bypass.18 Although a promising result, studies in humans are eagerly awaited.

Fluid management and its monitoring have gained increasing attention in the past decades. Clinicians are more aware than ever of the possible negative effects of fluid overload. We emphasise that being fluid responsive does not need treatment per se, as this may result in an increase in EVLW and tissue oedema. In addition, in fluid responsive patients, fluid loading may not increase cardiac output to a relevant extent in the long run, as the half-life of fluids is short, and may be even substantially shorter in patients with systemic inflammation. Attention has been focused on prediction of fluid responsiveness, as Kim et al. have eloquently done. Improvements in software may reduce the percentage error and improve the trending abilities of the monitors used, as demonstrated with other technologies.19 If we accept the limitations and uncertainty imposed by the monitoring devices, trials with clinical outcome measures as primary endpoints in the paediatric population should be the next step.

Acknowledgements relating to this article

Assistance with the Editorial: none.

Financial support and sponsorship: none.

Conflicts of interest: none.

Comment from the editor: this article was checked and accepted by the Editors but was not sent for external peer-review.

References

1. Rosenberg AL, Dechert RE, Park PK, et al. Review of a large clinical series: association of cumulative fluid balance on outcome in acute lung injury: a retrospective review of the ARDSnet tidal volume study cohort. J Intensive Care Med 2009; 24:35–46.
2. Murphy CV, Schramm GE, Doherty JA, et al. The importance of fluid management in acute lung injury secondary to septic shock. Chest 2009; 136:102–109.
3. Boyd JH, Forbes J, Nakada TA, et al. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality. Crit Care Med 2011; 39:259–265.
4. Kim E-H, Lee J-H, Jang Y-E, et al. Prediction of fluid responsiveness using lung recruitment manoeuvre in paediatric patients receiving lung-protective ventilation: a prospective observational study. Eur J Anaesthesiol 2021; 38:452–458.
5. Bendjelid K, Romand JA. Fluid responsiveness in mechanically ventilated patients: a review of indices used in intensive care. Intensive Care Med 2003; 29:352–360.
6. 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; 18:587–597.
7. Singh Y, Villaescusa JU, da Cruz EM, et al. Recommendations for hemodynamic monitoring for critically ill children-expert consensus statement issued by the cardiovascular dynamics section of the European Society of Paediatric and Neonatal Intensive Care (ESPNIC). Crit Care 2020; 24:620–633.
8. Sandroni C, Cavallaro F, Marano C, et al. Accuracy of plethysmographic indices as predictors of fluid responsiveness in mechanically ventilated adults: a systematic review and meta-analysis. Intensive Care Med 2012; 38:1429–1437.
9. Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999; 15:85–91.
10. Chew MS, Poelaert J. Accuracy and repeatability of pediatric cardiac output measurement using Doppler: 20-year review of the literature. Intensive Care Med 2003; 29:1889–1894.
11. Park J, Yang S, Lee JH, et al. The importance of sensor contacting force for predicting fluid responsiveness in children using respiratory variations in pulse oximetry plethysmographic waveform. J Clin Monit Comput 2019; 33:393–401.
12. Gan H, Cannesson M, Chandler JR, Ansermino JM. Predicting fluid responsiveness in children: a systematic review. Anesth Analg 2013; 117:1380–1392.
13. Marik PE, Lemson J. Fluid responsiveness: an evolution of our understanding. Br J Anaesth 2014; 112:617–620.
14. Deng QW, Tan WC, Zhao BC, et al. Is goal-directed fluid therapy based on dynamic variables alone sufficient to improve clinical outcomes among patients undergoing surgery? A meta-analysis. Crit Care 2018; 22:298–314.
15. Jalil BA, Cavallazzi R. Predicting fluid responsiveness: a review of literature and a guide for the clinician. Am J Emerg Med 2018; 36:2093–2102.
16. Hahn RG, Lyons G. The half-life of infusion fluids: an educational review. Eur J Anaesthesiol 2016; 33:475–482.
17. Chappell D, Bruegger D, Potzel J, et al. Hypervolemia increases release of atrial natriuretic peptide and shedding of the endothelial glycocalyx. Crit Care 2014; 18:538–545.
18. Dekker NAM, van Meurs M, van Leeuwen ALI, et al. Vasculotide, an angiopoietin-1 mimetic, reduces pulmonary vascular leakage and preserves microcirculatory perfusion during cardiopulmonary bypass in rats. Br J Anaesth 2018; 121:1041–1051.
19. Slagt C, Malagon I, Groeneveld AB. Systematic review of uncalibrated arterial pressure waveform analysis to determine cardiac output and stroke volume variation. Br J Anaesth 2014; 112:626–637.
Copyright © 2021 European Society of Anaesthesiology and Intensive Care. Unauthorized reproduction of this article is prohibited.