Secondary Logo

Journal Logo


Diagnostic accuracy of inferior vena caval respiratory variation in detecting fluid unresponsiveness

A systematic review and meta-analysis

Das, Saurabh K.; Choupoo, Nang S.; Pradhan, Debasis; Saikia, Priyam; Monnet, Xavier

Author Information
European Journal of Anaesthesiology: November 2018 - Volume 35 - Issue 11 - p 831-839
doi: 10.1097/EJA.0000000000000841



Fluid unresponsiveness is an abnormal state of the heart that operates on the flat portion of the Frank–Starling curve, as does a failing heart. In these circumstances, fluid administration may dramatically increase cardiac filling pressures and may lead to pulmonary oedema and impairment of oxygenation.1 Although fluid resuscitation is the first step in the management of the majority of acute circulatory failures, 40 to 70% of patients do not respond to it with a significant increase in cardiac output (CO).2,3 During fluid resuscitation, several boluses may be given, and it becomes important to identify when the physiological limit might be exceeded. Those who are unlikely to benefit or even made worse by fluid challenge should be adequately assessed before performing volume expansion. Well validated reliable tools are needed to divide patients into fluid responders and nonresponders.

Various static and dynamic indices have been evaluated to guide the clinician during fluid resuscitation in a patient with acute circulatory instability. Although dynamic indices of preload are recommended for predicting fluid responsiveness, their accuracy is debated.2 Among the dynamic indices, ultrasonographic measurement of respiratory variation of inferior vena cava diameter (rvIVC) enjoys a unique place because of its ease of use, its noninvasiveness, short learning curve, wide availability and general applicability as part of the sonographic assessment of shock.4 However, several studies conducted both in spontaneously breathing and mechanically ventilated patients found that rvIVC lacked consistent accuracy in the diagnosis of fluid responsiveness.5–19 One meta-analysis of eight studies and 235 patients reported that rvIVC had pooled sensitivity of 0.76 [95% confidence interval (CI) 0.61 to 0.86] and specificity of 0.86 (95% CI 0.69 to 0.95).20 A recent meta-analysis that analysed 433 patients, reported pooled sensitivity and specificity of 0.63 (95% CI 0.56 to 0.69) and 0.73 (95% CI 0.67 to 0.78), respectively.21 These two meta-analyses pooled both spontaneously breathing and mechanically ventilated patients despite there being two distinct respiratory physiologies. Most of the studies to date have considered the usefulness of rvIVC in predicting fluid responsiveness, but the majority of patients who presented with acute circulatory failure were found to be fluid unresponsive.21

We conducted this review to evaluate the accuracy of rvIVC in identifying patients unlikely to benefit from fluid administration. Spontaneously breathing and mechanically ventilated patients each have their own meta-analysis, excluding children and pregnant women, and several new studies are included.22–30


The method of review was decided before beginning the data search. The protocol was registered at an international prospective register of systematic reviews (identifier CRD 42017068028).31

Amendments were made to the original protocol to include several subgroup and sensitivity analyses. We undertook and reported this systematic review according to the preferred reporting items for systematic reviews and meta-analyses guidelines (

Patient, problem or population, intervention, comparison, control or comparator, outcome statement

The PICO statement was as following:

  1. P – patients, problem or population: patients with acute circulatory failure in whom the effect of volume expansion on CO is unknown and needs to be predicted.
  2. I – index test: rvIVC, that is collapsibility or distensibility.
  3. C – comparison or gold standard: fluid responsiveness was defined as a significant increase in CO, cardiac index (CI) or any other cardiovascular variable during volume expansion.
  4. O – outcomes assessment: ability of rvIVC to predict fluid nonresponsiveness.

Data search

All publications that reported diagnostic accuracy of rvIVC in the prediction of response to fluid bolus were reviewed up to 11 June 2017. We searched MEDLINE through PubMed, EMBASE through Ovid and the Cochrane Library. To find publications in languages other than English, we searched KoreaMed and LILCAS. We searched WHO Clinical Trial Registry ( to retrieve unpublished work and ongoing studies. New links displayed beside the online version of the published abstracts were followed and retrieved. The bibliographies of retrieved articles were searched and checked for additional studies. Where there was confusion regarding the data, the authors of the articles were contacted. We translated all languages to English by Google Translate ( The details of the search strategy are shown in the Appendix,

Selection of the studies

Any cross-sectional or cohort study that used rvIVC to assess the effect of a fluid bolus was included in the qualitative analysis on the basis of following criteria.

  1. Studies that verified the result of the index test (rvIVC) with that of a reference test.
  2. Studies that evaluated the accuracy of rvIVC in living adult humans. Studies on cadavers, neonates and pregnant women were excluded.
  3. Studies were included in the meta-analysis if there were data enabling the construction of 2 × 2 tables, or if they could be calculated or could be obtained from authors and met the above cited criteria.

The full inclusion and exclusion criteria are presented in the Appendix,

Data extraction

A data extraction form was used to extract equivalent information from each study and it included year of publication, country of origin, study design, sample size, number of fluid responders and nonresponders, reference test, threshold for reference and index test, observed average collapsibility or distensibility index in fluid responders and nonresponders, sensitivity, specificity, Area Under the Receiver Operating Characteristic curve (AUC), breathing status, ventilator settings in ventilated patients and cardiac rhythm. Two authors (SKD and NSC) independently reviewed the articles, assessed quality and ascertained the criteria for inclusion in the pooled data analysis. A third reviewer (DP) blinded to the primary reviewers decision checked the article selection, data extraction and risk of bias assessment. Any disagreement was resolved by consensus. Articles from the same author were carefully investigated to avoid duplication of studies included for analysis.

Risk of bias assessment

Quality of study reports was assessed with QUADAS 2 tool.32 Cochrane recommends QUADAS 2 for use in diagnostic test reviews.33 The detailed method used for risk of bias assessment is presented in the Appendix,

Data synthesis

All the original studies assessed fluid responsiveness instead of fluid unresponsiveness. To assess fluid unresponsiveness, we rearranged the data in 2 × 2 tables expressing true positives, false positives, false negatives and true negatives. We tabulated original specificity as sensitivity and sensitivity as specificity. In the primary studies in which data were not expressed in a contingency table, we constructed 2 × 2 tables from available data whenever possible. If a study reported multiple thresholds, we took the optimum threshold and corresponding sensitivity and specificity.

Data analysis

A bivariate model was used to estimate pooled sensitivity and specificity as this preserves the two-dimensional nature of diagnostic accuracy data. It analysed the logarithmically (logit) transformed sensitivity {log[sens/ (1 sens)]} and specificity {log[spec/(1 spec)]} of each study in a single model. A random effects approach was used for both sensitivity and specificity.34 All calculations and plots were computed with software RStudio; R version 3.2.5 (2016-04-14) and R package ‘mada’ , Bristol Medical School: Population Health Sciences ,University of Bristol, Senate House, Tyndall Avenue, Bristol, BS8 1TH, UK.35

To investigate the source of study heterogeneity, we conducted subgroup analyses. We compared studies with pulse contour analysis (PCA), transpulmonary thermodilution (TPTD) and transthoracic echocardiography (TTE) against studies with other methods of CO measurement, making a hypothesis that the reliability of rvIVC test is better with PCA, TPTD and TTE than other methods. We compared studies in which volume expansion was with colloids with those using crystalloids, testing the hypothesis that the reliability of rvIVC is better when volume expansion is performed with colloids. Subgroup analysis was conducted among studies with different definitions of fluid responsiveness, such as 10 vs. 15% increase in CO or CI or stroke volume (SV) or SV index (SVI) or velocity time integral (VTI) of the left ventricular outflow tract. We also performed sensitivity analysis by excluding studies with high risk of bias.


The review process and selection of included studies are presented in Fig. 1. Emails were sent to authors of four studies for clarification of data, but no additional data were received. Our search returned a total of 158 233 publications and 23 articles analysing 1574 patients who were selected for qualitative analysis. Most of the studies were from ICU, two were from emergency departments and one was from an operating theatre. The most common indication for fluid resuscitation was septic shock. In 12 studies, the reference test was CO, SV or VTI as measured by echocardiography.6,7,9,10,12,16,18–20,22,29,30 CO was measured by PCA in seven studies,11,17,23–27 by TPTD in one study14 and impedance cardiography in another study.5 Fluid response was defined by a fluid induced increase in CO, CI, SV, SVI or VTI by more than 15% in 14 studies6,9–12,14,16–19,22–24,26 and 10% in seven studies.5,7,15,25,27,29,30 Two studies used increases in SBP or mean artery pressure by 10 mmHg to define fluid responsiveness.8,28 Nine studies used colloid,10,12,14,18,23,24,26,27,30 eight used crystalloid,6–8,15,16,18,22,25 three used passive leg raising (PLR)5,11,29 and one used both colloid and crystalloid for fluid bolus.17 The quantity of fluid bolus was not uniform across the studies, ranging from 200 to 1000 ml (Tables 1 and 2).

Fig. 1:
The review process and selection of included studies.
Table 1:
Characteristics of included studies of spontaneously breathing patients
Table 2:
Characteristics of included studies that assess fluid responsiveness in mechanically ventilated patients

Twenty studies including 761 patients that fulfilled the inclusion criteria were pooled to summarise the diagnostic accuracy of rvIVC in predicting fluid unresponsiveness. Three studies could not be included in the quantitative analysis due to inability to obtain or calculate contingency data.5,28,29 Out of 761 patients, 330 were spontaneously breathing and 431 were mechanically ventilated. A total of 47% patients were fluid nonresponsive.

Nine studies encompassing 694 spontaneously breathing patients showed that sensitivity and specificity of rvIVC in predicting fluid unresponsiveness were 0.67 to 1 and 0.31 to 1, respectively.5–8,19,22,26,28,30 The pooled sensitivity and specificity of seven studies and 330 patients were 0.80 (95% CI 0.68 to 0.89) and 0.79 (95% CI 0.60 to 0.90).6–9,22,26,28 The AUC was 0.857 (Fig. 2 and Appendix,

Fig. 2:
SROC curve of spontaneously breathing patients using bivariate model. Ellipse denotes confidence region, Δ denotes data, ○ denotes summary estimate. SROC, summary receiver operating characteristics.

Fourteen studies containing 853 mechanically ventilated patients reported that rvIVC had a sensitivity of 0.53 to 1 and a specificity of 0.38 to 1 to predict fluid unresponsiveness.9–12,14–18,23–25,27,29 The pooled sensitivity and specificity of 13 studies involving 431 patients were 0.79 (95% CI 0.67 to 0.86) and 0.70 (95% CI 0.63 to 0.76), respectively.9–12,14–18,23–25,27 The AUC was 0.75 (Fig. 3 and Appendix,

Fig. 3:
SROC curve of mechanically ventilated patients using bivariate model. Ellipse denotes confidence region, Δ denotes data, ○ denotes summary estimate. SROC, summary receiver operating characteristics.

All studies but one reported the optimum distensibility index14 ranging from 12 to 46%. Similarly, the optimum collapsibility index ranged from 13 to 50%. Both distensibility and collapsibility index when measured before fluid resuscitation were significantly different between fluid responders and nonresponders. The pooled mean difference of observed collapsibility index in fluid responders and nonresponders in spontaneously breathing patients was 9% (95% CI 5 to 12), P less than 0.00001. Similarly, pooled mean difference of observed distensibility index in fluid responders and nonresponders in mechanically ventilated patients was 8% (95% CI 7 to 9), P less than 0.00001.

Risk assessment of bias was presented in Table 3. Nine studies had an unclear risk of bias due to lack of appropriate exclusion criteria for patient selection. Seven studies were at high risk of bias due to inappropriate reference test, flow and timing.

Table 3:
Applicability and risk assessment of bias using QUADAS 2 tool

Subgroup and sensitivity analysis are presented in the Appendix, In spontaneously breathing patients, subgroup analysis showed lower accuracy of rvIVC in studies in which PCA, TPTD or echocardiography were not used as gold standard. In mechanically ventilated patients, results of subgroup and sensitivity analysis did not differ from that of the primary analysis.


The meta-analysis to find the diagnostic accuracy of rvIVC in predicting fluid unresponsiveness involved 20 studies with 761 adult patients. The extent of rvIVC (distensibility or collapsibility) when assessed prior to fluid administration differed significantly between fluid responders and nonresponders. Summary estimation of sensitivity and specificity showed that rvIVC was moderately accurate in predicting fluid unresponsiveness in both spontaneously breathing and mechanically ventilated patients. The optimum thresholds of rvIVC in mechanically ventilated and spontaneously breathing patients were 12 to 46% and 13 to 50%, respectively.

Respiration induces cyclical variation in right atrial and central venous pressure (CVP) due to transmission of pressure changes from the pleural space. Depending on the compliance of the vessel, these changes can also affect the dimensions of the vena cava. Positive pressure breathing in mechanically ventilated patients increases the size of inferior vena cava, whereas spontaneous negative pressure breathing reduces its size. These variations are popularly known as distensibility and collapsibility. When the vena cava is under filled, the compliance is greatest. Accordingly, large respiratory variation in vena cava dimensions suggests reduced intravenous volume and predicts fluid responsiveness.36 But it is not very clear whether decreased or absence of respiratory variation also indicates adequacy of intravascular fluid volume. Fortunately, the test developed to detect fluid responsiveness can also serve to detect volume unresponsiveness.1 Therefore, we analysed the result of the studies that primarily intended to assess fluid responsiveness and estimated the accuracy of rvIVC to predict fluid unresponsiveness.

There were three meta-analyses that tested the efficacy of rvIVC to predict fluid responsiveness.20,21,36 The first two reported almost similar efficacy as reported by the current study.20,36 However, their findings were limited by very small numbers of patients, including in each only 116 and 137 mechanically ventilated, and 40 and 99 spontaneously breathing patients.20,36 The most recent meta-analysis which analysed 158 spontaneously breathing and 275 mechanically ventilated patients reported sensitivity of rvIVC in the prediction of fluid responsiveness at 0.67 (95% CI 0.58 to 0.75) and 0.68 (95% CI 0.60 to 0.76) and specificity at 0.52 (95% CI 0.42 to 0.62) and 0.77 (95% CI 0.68 to 0.84) in mechanically ventilated and spontaneously breathing patients, respectively.21 The reported accuracy was lower than the current study that analysed a total of 761 patients.

A recent systematic review reported the diagnostic accuracy of various static and dynamic variables.36 According to this review, sensitivity and specificity of CVP, SV variation (SVV), pulse pressure variation (PPV) and PLR were 0.62 (95% CI 0.54 to 0.69) and 0.76 (95% CI 0.60 to 0.87), 0.79 (95% CI 0.67 to 0.87) and 0.84 (95% CI 0.74 to 0.90), 0.84 (95% CI 0.75 to 0.90) and 0.84 (95% CI 0.77 to 0.90), 0.92 (95% CI 0.82 to 0. 97) and 0.92 (95% CI 0.86 to 0.98), respectively. The reported sensitivity and specificity of rvIVC in the mechanically ventilated in this review were 0.77 (95% CI 0.44 to 0.94) and 0.85 (95% CI 0.49 to 0.97). The review also reported that sensitivity and specificity of rvIVC in spontaneously breathing patients were 0.31 to 0.70 and 0.80 to 0.97.36 This finding suggests that the diagnostic accuracy of rvIVC is higher than static variables like CVP, lower than PLR and equivalent to other dynamic variables like SVV and PPV.

We chose to assess accuracy of rvIVC in the prediction of fluid unresponsiveness. Responsiveness and unresponsiveness are two sides of a coin and our findings can also be used to inform on the accuracy of rvIVC in predicting fluid responsiveness. We extensively searched all major databases without any language restriction. We also attempted to find unpublished reports. This review included more studies and patients than earlier reviews. Combining spontaneously breathing and mechanically ventilated patients introduced both clinical and statistical heterogeneity in earlier studies. We included only nonpregnant adults and analysed spontaneously breathing and mechanically ventilated patients separately. We consider this an improvement on earlier studies in terms of methodology and precision.

In spontaneously breathing patients, IVC collapsibility reflects the gradient between abdominal and intrathoracic pressure. The IVC may collapse despite adequate volume status as is seen when there is increased respiratory effort or diaphragmatic motion.37 Earlier meta-analyses of limited sample size reported that rvIVC was less accurate in predicting fluid responsiveness in spontaneously breathing patient than in those mechanically ventilated.20,21 But our analysis found equivalent accuracy of rvIVC in spontaneously breathing and mechanically ventilated patients. This is the first meta-analysis to find that the diagnostic accuracy of rvIVC was also good in spontaneously breathing patients. Nevertheless, this result should be analysed cautiously. The diagnostic accuracy of rvIVC may be impaired in various clinical scenarios. First, physiologically, it is clear that if the strength of the inspiratory effort is significantly irregular and varies from one cycle to the other, it is impossible to define any threshold of rvIVC that could predict fluid responsiveness. In breathless patients with strong inspiratory effort, the changes in the intrathoracic and intra-abdominal pressures are much larger than in normal breathing. We cannot exclude the possibility that strong inspiratory effort increases the intra-abdominal pressure and squeezes the IVC even in the presence of fluid nonresponsiveness. By the same argument, the diagnostic accuracy must be lower in patients with significant expiratory exertion, especially in shortness of breath associated with expiratory dyspnoea. In the presence of intra-abdominal hypertension, the physiology of the variation of IVC volume might be profoundly altered, although this has not been thoroughly investigated. Therefore the findings of this study should be used in the appropriate clinical context and should be incorporated in clinical practice in conjunction with other clinical measures to assess fluid status. Our finding in spontaneously breathing patients needs to be validated with more studies with a higher sample size.

This study has also some other limitations. The meta-analysis could not include three studies due to inability to construct a 2 × 2 table.5,28,29 Out of these three, two had a large sample size.28,29 One study in spontaneously breathing patients had reported accuracy similar to our finding.28 Another study in mechanically ventilated patients showed lower accuracy in comparison with the present meta-analysis. This may be due to a lower diagnostic threshold, lack of uniformity in fluid challenge and the severely ill patient cohort.29 Most of studies that were included in this review incorporated patients with systemic vasodilatation such as septic shock. It is therefore difficult to extrapolate the result of this meta-analysis to other settings to detect fluid unresponsiveness. We did not test the diagnostic accuracy of the respiratory changes in the superior vena cava; this should be evaluated in further reviews.

The current systematic review and meta-analysis found that the extent of inferior vena caval respiratory variation differed significantly in patients depending upon their response to fluid administration. In particular, decreased inferior vena caval respiratory variation has moderate accuracy in predicting fluid unresponsiveness both in spontaneously breathing and mechanically ventilated patients.

Acknowledgements relating to this article

Assistance with the study: none.

Financial support and sponsorship: none.

Conflicts of interest: none.

Presentation: none.


1. Teboul J-L, Monnet X. Detecting volume responsiveness and unresponsiveness in intensive care unit patients: two different problems, only one solution. Crit Care 2009; 13:175.
2. Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest 2002; 121:2000–2008.
3. Cecconi M, Hofer C, Teboul JL, et al. Fluid challenges in intensive care: the FENICE study: a global inception cohort study. Intensive Care Med 2015; 41:1529–1537.
4. Brennan JM, Ronan A, Goonewardena S, et al. Handcarried ultrasound measurement of the inferior vena cava for assessment of intravascular volume status in the outpatient hemodialysis clinic. Clin J Am So Nephrol 2006; 1:749–753.
5. Corl K, Napoli AM, Gardiner F. Bedside sonographic measurement of the inferior vena caval index is a poor predictor of fluid responsiveness in emergency department patients. Emerg Med Australs 2012; 24:534–539.
6. Lanspa MJ, Grissom CK, Hirshberg EL, et al. Applying dynamic parameters to predict hemodynamic response to volume expansion in spontaneously breathing patients with septic shock. Shock 2013; 39:155–160.
7. Airapetian N, Maizel J, Alyamani O, et al. Does inferior vena cava respiratory variability predict fluid responsiveness in spontaneously breathing patients? Crit Care 2015; 19:400.
8. De Valk S, Olgers TJ, olman M, et al. The caval index: an adequate noninvasive ultrasound parameter to predict fluid responsiveness in the emergency department? BMC Anesthesiol 2014; 14:114.
9. Barbier C, Loubiéres Y, Schmit C, et al. Respiratory changes in inferior vena cava diameter are helpful in predicting fluid responsiveness in ventilated septic patients. Intensive Care Med 2004; 30:1740–1746.
10. Feissel M, Michard F, Faller JP, et al. The respiratory variation in inferior vena cava diameter as a guide to fluid therapy. Intensive Care Med 2004; 30:1834–1837.
11. Varas JL, Diaz CM, Blancas R, et al. Inferior vena cava distensibility index predicting fluid responsiveness in ventilated patients. Intensive Care Med Exp 2015; 3:A600.
12. Charbonneau H, Riu B, Faron M, et al. Predicting preload responsiveness using simultaneous recording of inferior and superior vena cavae diameters. Crit Care 2014; 18:473.
13. Juhl-Olsen P, Frederiksen CA, Sloth E. Ultrasound assessment of inferior vena cava collapsibility is not a valid measure of preload changes during triggered positive ventilation: a controlled cross-over study. Ultraschall Med 2012; 33:1529.
14. Moretti R, Pizzi B. Inferior vena cava distensibility as a predictor of fluid responsiveness in patients with subarachnoid hemorrhage. Neurocrit Care 2010; 13:3–9.
15. Machare-Delgado E, Decaro M, Marik PE. Inferior vena cava variation compared to pulse contour analysis as predictors of fluid responsiveness: a prospective cohort study. J Intensive Care Med 2011; 26:116–124.
16. De Oliveira OH, Freitas FG, Ladeira RT, et al. Comaparison between respiratory changes in the inferior vena cava diameter and pulse pressure variation to predict fluid responsiveness in postoperative patients. J Crit Care 2016; 34:46–49.
17. Theerawit P, Morasert T, Sutherasan Y. Inferior vena cava diameter variation compared with pulse pressure variation as pressure variation as predictors of fluid responsiveness in patients with sepsis. J Crit Care 2016; 36:246–251.
18. Sobczyk D, Nycz K, Andruszkiewicz P, et al. Ultrasonographic caval indices do not significantly contribute to predicting fluid responsiveness immediately after coronary artery bypass grafting when compared to passive leg raising. Cardiovasc Ultrasound 2016; 14:23.
19. Muller L, Bobbia X, Toumi M, et al. AzuRea Group: Respiratory variations of inferior vena cava diameter to predict fluid responsiveness in spontaneously breathing patients with acute circulatory failure: need for a cautious use. Crit Care 2012; 16:R188.
20. Zhang Z, Xu X, Ye S, et al. Ultrasonographic measurement of the respiratory variation in the inferior vena cava diameter is predictive of fluid responsiveness in critically ill patients: systematic review and meta-analysis. Ultrasound Med Biol 2014; 5:845–853.
21. Long E, Oakley E, Duke T, et al. Paediatric Research in Emergency Department International Collaborative (PREDICT): Does respiratory variation in inferior vena cava diameter predict fluid responsiveness: a systematic review and meta-analysis. Shock 2017; 47:550–559.
22. Hasanin AM, Lotfy AM, El-aal IRA, et al. Sensitivity and specificity of ultrasound measured dimension of IVC and IJV compared to central venous pressure in predicting fluid responsive. Med J Cairo Univ 2016; 84:251–257.
23. Zhang X, Fenq J, Zhu P, et al. Ultrasonographic measurement of the inferior vena cava variation as predictor fluid responsiveness in patients undergoing anaesthesia surgery. J Surg Res 2016; 204:118–122.
24. Weihua Z, Linjun W, Xiaohong W, et al. Measurement of brachial artery velocity variation and inferior vena cava variability to estimate fluid responsiveness. Chin Care Med 2016; 28:713–717.
25. Lu N, Xi X, Jiang L, et al. Exploring the best predictors of fluid responsiveness in patients with septic shock. Am J Emerg Med 2017; 35:1258–1261.
26. Zhao J, Wang G. Inferior vena cava collapsibility index is a valuable and noninvasive index for elevated general heart end-diastolic volume index estimation in septic shock patients. Med Sci Monit 2016; 22:3843–3848.
27. Wu J, Zhang X, Wang Z, et al. Clinical evaluation of transthoracic echocardiography predicting fluid responsiveness in septic shock. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue 2014; 26:36–40.
28. Sawe HR, Haeffele C, Mfinanga JA, et al. Predicting fluid responsiveness using bedside ultrasound measurements of the inferior vena cava and physician gestalt in the Emergency Department of an Urban Public Hospital in sub-Saharan Africa. PLoS One 2016; 11:e0162772.
29. Vignon P, Repessé X, Béot E, et al. Comparison of echocardiographic indices used to predict fluid responsiveness in ventilated patients. Am J Respir Crit Care Med 2017; 195:1022–1032.
30. Preau S, Bortolotti P, Colling D, et al. Diagnostic accuracy of the inferior vena cava collapsibility to predict fluid responsiveness in spontaneously breathing patients with sepsis and acute circulatory failure. Crit Care Med 2016; 45:e290–e2297.
31. Das SK, Choupoo NS, Pradhan D, Saikia P, Monnet X. Diagnostic accuracy of inferior vena caval respiratory variation to detect fluid unresponsiveness: a systematic review and meta-analysis. PROSPERO: International prospective register of systematic reviews. Available from
32. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011; 155:529–536.
33. The Cochrane Collaboration, Deeks JJ, Wisniewski S, Davenport C. Deeks JJ, Bossuyt PM, Gatsonic C. Chapter 4: guide to the contents of a Cochrane diagnostic test accuracy protocol. Cochrane handbook for systematic reviews of diagnostic test accuracy version 1.0.0 2013.
34. Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity produces informative summary measures diagnostic reviews. J Clin Epidemiol 2005; 58:982–990.
35. RStudio Team: RStudio: integrated development for R. Boston, MA: RStudio, Inc.; 2015.
36. Bentzer P, Griesdale DE, Boyd J, et al. Will this hemodynamically unstable patient respond to a bolus of intravenous fluids? JAMA 2016; 316:1298–1309.
37. Gignon L, Roger C, Bastide S, et al. Influence of diaphragmatic motion on inferior vena cava diameter respiratory variation in healthy volunteers. Anesthesiology 2016; 124:1338–1346.

Supplemental Digital Content

© 2018 European Society of Anaesthesiology