Changes of central venous oxygen saturation define fluid responsiveness in patients with septic shock: A prospective observational study

Central venous oxygen saturation changes as a reliable predictor of the change of CI in septic shock: To explore potential influencing factors

PURPOSE

Assessing fluid responsiveness relying on central venous oxygen saturation (ScvO2) yields varied outcomes across several studies. This study aimed to determine the ability of the change in ScvO2 (ΔScvO2) to detect fluid responsiveness in ventilated septic shock patients and potential influencing factors.

METHODS

In this prospective, single-center study, all patients conducted from February 2023 to January 2024 received fluid challenge. Oxygen consumption was measured by indirect calorimetry, and fluid responsiveness was defined as an increase of cardiac index (CI) ≥ 10% measured by transthoracic echocardiography. Multivariate linear regression analysis was conducted to evaluate the impact of oxygen consumption, arterial oxygen saturation, CI, and hemoglobin on ScvO2 and its change before and after fluid challenge.

RESULTS

Among 49 patients (31 men, aged (59 ± 18) years), 27 were responders. The patients had an acute physiology and chronic health evaluation II score of 24 ± 8, a sequential organ failure assessment score of 11 ± 4, and a blood lactate level of (3.2 ± 3.1) mmol/L at enrollment. After the fluid challenge, the ΔScvO2 (mmHg) in the responders was greater than that in the non-responders (4 ± 6 vs. 1 ± 3, p = 0.019). Multivariate linear regression analysis suggested that CI was the only independent influencing factor of ScvO2, with R2 = 0.063, p = 0.008. After the fluid challenge, the change in CI became the only contributing factor to ΔScvO2 (R2 = 0.245, p < 0.001). ΔScvO2 had a good discriminatory ability for the responders and non-responders with a threshold of 4.4% (area under the curve = 0.732, p = 0.006).

CONCLUSION

ΔScvO2 served as a reliable surrogate marker for ΔCI and could be utilized to assess fluid responsiveness, given that the change of CI was the sole contributing factor to the ΔScvO2. In stable hemoglobin conditions, the absolute value of ScvO2 could serve as a monitoring indicator for adequate oxygen delivery independent of oxygen consumption.

Variation in central venous oxygen saturation to evaluate fluid responsiveness: a systematic review and meta-analysis

BACKGROUND

Since oxygen content and oxygen consumption typically remain unchanged within a short period, variation in central venous oxygen saturation (ΔScvO₂) during fluid challenge can theoretically track the changes in cardiac output (CO). We conducted this meta-analysis to systematically assess the diagnostic performance of ΔScvO₂ during a fluid challenge for fluid responsiveness in mechanically ventilated patients receiving volume expansion.

METHODS

Electronic databases were systematically searched to identify relevant studies published before October 24, 2022. As the cutoff value of ΔScvO₂ was expected to vary across the included studies, we estimated the area under the hierarchical summary receiver operating characteristic curve (AUHSROC) as the primary measure of diagnostic accuracy. The optimal threshold of ΔScvO₂ and the corresponding 95% confidential interval (CI) were also estimated.

RESULTS

This meta-analysis included 5 observational studies comprising 240 participants, of whom 133 (55%) were fluid responders. Overall, the ΔScvO₂ during the fluid challenge exhibited excellent performance for defining fluid responsiveness in mechanically ventilated patients receiving volume expansion, with an AUHSROC of 0.86 (95% CI 0.83-0.89), a pooled sensitivity of 0.78 (95% CI 0.69-0.85), a pooled specificity of 0.84 (95% CI 0.72-0.91), and a pooled diagnostic odds ratio of 17.7 (95% CI 5.9-53.2). The distribution of the cutoff values was nearly conically symmetrical and concentered between 3 and 5%; the mean and median cutoff values were 4% (95% CI 3-5%) and 4% (95% CI not estimable), respectively.

CONCLUSIONS

In mechanically ventilated patients receiving volume expansion, the ΔScvO2 during the fluid challenge is a reliable indicator of fluid responsiveness. Clinical trial registration PROSPERO, https://www.crd.york.ac.uk/prospero/ , registry number: CRD42022370192.

Fluid challenge in critically ill patients receiving haemodynamic monitoring: a systematic review and comparison of two decades

Introduction

Fluid challenges are widely adopted in critically ill patients to reverse haemodynamic instability. We reviewed the literature to appraise fluid challenge characteristics in intensive care unit (ICU) patients receiving haemodynamic monitoring and considered two decades: 2000–2010 and 2011–2021.

Methods

We assessed research studies and collected data regarding study setting, patient population, fluid challenge characteristics, and monitoring. MEDLINE, Embase, and Cochrane search engines were used. A fluid challenge was defined as an infusion of a definite quantity of fluid (expressed as a volume in mL or ml/kg) in a fixed time (expressed in minutes), whose outcome was defined as a change in predefined haemodynamic variables above a predetermined threshold.

Results

We included 124 studies, 32 (25.8%) published in 2000–2010 and 92 (74.2%) in 2011–2021, overall enrolling 6,086 patients, who presented sepsis/septic shock in 50.6% of cases. The fluid challenge usually consisted of 500 mL (76.6%) of crystalloids (56.6%) infused with a rate of 25 mL/min. Fluid responsiveness was usually defined by a cardiac output/index (CO/CI) increase ≥ 15% (70.9%). The infusion time was quicker (15 min vs 30 min), and crystalloids were more frequent in the 2011–2021 compared to the 2000–2010 period.

Conclusions

In the literature, fluid challenges are usually performed by infusing 500 mL of crystalloids bolus in less than 20 min. A positive fluid challenge response, reported in 52% of ICU patients, is generally defined by a CO/CI increase ≥ 15%. Compared to the 2000–2010 decade, in 2011–2021 the infusion time of the fluid challenge was shorter, and crystalloids were more frequently used.

Effects of rapid fluid infusion on hemoglobin concentration: a systematic review and meta-analysis

Background

Rapid fluid administration may decrease hemoglobin concentration (Hb) by a diluting effect, which could limit the increase in oxygen delivery (DO₂) expected with a positive response to fluid challenge in critically ill patients. Our aim was to quantify the decrease in Hb after rapid fluid administration.

Methods

Our protocol was registered in PROSPERO (CRD42020165146). We searched PubMed, the Cochrane Database, and Embase from inception until February 15, 2022. We selected studies that reported Hb before and after rapid fluid administration (bolus fluid given over less than 120 min) with crystalloids and/or colloids in adults. Exclusion criteria were studies that included bleeding patients, or used transfusions or extracorporeal circulation procedures. Studies were divided according to whether they involved non-acutely ill or acutely ill (surgical/trauma, sepsis, circulatory shock or severe hypovolemia, and mixed conditions) subjects. The mean Hb difference and, where reported, the DO₂ difference before and after fluid administration were extracted. Meta-analyses were conducted to assess differences in Hb before and after rapid fluid administration in all subjects and across subgroups. Random-effect models, meta-regressions and subgroup analyses were performed for meta-analyses. Risk of bias was assessed using the Cochrane Risk of Bias Assessment Tool. Inconsistency among trial results was assessed using the I² statistic.

Results

Sixty-five studies met our inclusion criteria (40 in non-acutely ill and 25 in acutely ill subjects), with a total of 2794 participants. Risk of bias was assessed as “low” for randomized controlled trials (RCTs) and ‘low to moderate’ for non-RCTs. Across 63 studies suitable for meta-analysis, the Hb decreased significantly by a mean of 1.33 g/dL [95% CI − 1.45 to − 1.12; p < 0.001; I² = 96.88] after fluid administration: in non-acutely ill subjects, the mean decrease was 1.56 g/dL [95% CI − 1.69 to − 1.42; p < 0.001; I² = 96.71] and in acutely ill patients 0.84 g/dL [95% CI − 1.03 to − 0.64; p = 0.033; I² = 92.91]. The decrease in Hb was less marked in patients with sepsis than in other acutely ill patients. The DO₂ decreased significantly in fluid non-responders with a significant decrease in Hb.

Conclusions

Hb decreased consistently after rapid fluid administration with moderate certainty of evidence. This effect may limit the positive effects of fluid challenges on DO₂ and thus on tissue oxygenation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13054-022-04191-x.

Do ScvO2 variations induced by passive leg raising predict fluid responsiveness? A prospective study

OBJECTIVE

The present study investigates whether ScvO2 variations induced by passive leg raising (PLR) are able to predict fluid responsiveness (FR) in mechanically ventilated patients.

DESIGN

A monocentric prospective clinical study.

SETTING

An intensive care division in a tertiary hospital.

PATIENTS

The inclusion criteria were elective postoperative cardiac surgery patients who were over 18 years old, deeply sedated, mechanically ventilated and needed volume expansion (VE). Fluid responders (R) were defined as patients who increased their left ventricular outflow tract velocity time integral (VTI) ≥15% after VE.

INTERVENTION

In patients included in this study, continuous ScvO2  monitoring (CeVOX device, Pulsion Medical Systems) and VTI (transthoracic echocardiography) were measured simultaneously before and during a PLR test and before and after VE (with 500 ml of saline).

MEASUREMENTS AND MAIN RESULTS

Thirty-three consecutive patients were included in this study. In 15 patients with a positive PLR test (increase in VTI ≥15%), ScvO2 increased during PLR by 9 ± 4%. In the 18 patients with a negative PLR test, ScvO2 did not significantly change during PLR. VE increased ScvO2 by 9 ± 6% and 2 ± 4% in responders and nonresponders, respectively. If ScvO2 increased by >4% during the PLR test, then a positive VTI response (≥15%) was diagnosed with a sensitivity of 93% (68-99%) and a specificity of 94% (63-99%) (Area under the receiver operating characteristic curve 0.92 ± 0.58, p < 0.05). Moreover, ScvO2 variations were able to distinguish responders to VE from nonresponders to VE with a sensitivity of 87% (68-99%) and a specificity of 89% (63-99%) (Area under the receiver operating characteristic curve 0.89 ± 0.07, p < 0.05).

CONCLUSIONS

ScvO2 variation induced by PLR is a reliable, minimally invasive parameter for predicting FR at the postoperative cardiac surgery bedside of mechanically ventilated, critically ill patients.

Changes in perfusion can detect changes in the cardiac index in patients with septic shock

Objective

The perfusion index (PI) is usually used to assess peripheral perfusion, which can be influenced by the cardiac index (CI). CI monitoring is often needed during the treatment of patients with shock. We investigated the relationship between changes in the PI (ΔPI) and changes in the CI (ΔCI) in patients with septic shock.

Methods

This retrospective study included patients with septic shock who underwent pulse-induced continuous cardiac output monitoring. We measured the CI and PI before and after fluid loading during the first 6 hours of intensive care unit admission. Fluid responsiveness was defined as a ≥10% ΔCI after fluid loading. Other hemodynamic and oxygen-derived parameters were also collected at the exact time of each CI measurement.

Results

Fifty-five patients were included in the study (29 fluid responders, 26 fluid non-responders). In the univariate analysis, ΔPI was positively correlated with ΔCI. In the multivariable analysis, ΔPI was independently associated with ΔCI. The receiver operating characteristic curve showed that ΔPI was an appropriate marker with which to discriminate a CI increase with an area under the curve of 0.776.

Conclusion

This study showed a positive correlation between ΔPI and ΔCI during the early treatment phase of septic shock.

The ability of pulse oximetry-derived peripheral perfusion index to detect fluid responsiveness in patients with septic shock

BACKGROUND

Fluid challenge test is a widely used method for the detection of fluid responsiveness in acute circulatory failure. However, detection of the patient's response to the fluid challenge requires monitoring of cardiac output which is not feasible in many settings. We investigated whether the changes in the pulse oximetry-derived peripheral perfusion index (PPI), as a non-invasive surrogate of cardiac output, can detect fluid responsiveness using the fluid challenge test or not.

METHODS

We prospectively enrolled 58 patients with septic shock on norepinephrine infusion. Fluid challenge test, using 200 mL crystalloid solution, was performed in all study subjects. All patients received an additional 300 mL crystalloid infusion to confirm fluid responsiveness. Velocity time integral (VTI) (using transthoracic echocardiography), and PPI were measured at the baseline, after 200 mL fluid challenge, and after completion of 500 mL crystalloids. Fluid responsiveness was defined by 10% increase in the VTI after completion of the 500 mL. The predictive ability of ∆PPI [Calculated as (PPI after 200 mL - baseline PPI)/baseline PPI] to detect fluid responders was obtained using the receiver operating characteristic curve.

RESULTS

Forty-two patients (74%) were fluid responders; in whom, the mean arterial pressure, the central venous pressure, the VTI, and the PPI increased after fluid administration compared to the baseline values. ∆PPI showed moderate ability to detect fluid responders [area under receiver operating characteristic curve (95% confidence interval) 0.82 (0.70-0.91), sensitivity 76%, specificity 80%, positive predictive value 92%, negative predictive value 54%, cutoff value ≥ 5%]. There was a significant correlation between ∆PPI and ∆VTI induced by the fluid challenge.

CONCLUSION

∆PPI showed moderate ability to detect fluid responsiveness in patients with septic shock on norepinephrine infusion. Increased PPI after 200 mL crystalloid challenge can detect fluid responsiveness with a positive predictive value of 92%; however, failure of the PPI to increase does not exclude fluid responsiveness.

CLINICAL TRIAL IDENTIFIER

NCT03805321. Date of registration: 15 January 2019. Clinical trial registration URL: https://clinicaltrials.gov/ct2/show/NCT03805321?term=ahmed+hasanin&rank=9 .

Contemporary hemodynamic monitoring, fluid responsiveness, volume optimization, and endpoints of resuscitation: an AAST critical care committee clinical consensus

This article, on hemodynamic monitoring, fluid responsiveness, volume assessment, and endpoints of resuscitation, is part of a compendium of guidelines provided by the AAST (American Association for the Surgery of Trauma) critical care committee. The intention of these guidelines is to inform practitioners with practical clinical guidance. To do this effectively and contemporarily, expert consensus via the critical care committee was obtained. Strict guideline methodology such a GRADE (Grading of Recommendations Assessment, Development and Evaluation) was purposefully NOT used so as not to limit the possible clinical guidance. The critical care committee foresees this methodology as practically valuable to the bedside clinician.

N-terminal pro-B-type natriuretic peptide for predicting fluid challenge in patients with septic shock

Background

The aim of this study is to examine whether plasma N-terminal pro-B-type natriuretic peptide (NT-proBNP) concentration could predict fluid responsiveness in septic shock patients following fluid challenge (FC).

Methods

We reviewed prospectively collected data from 79 septic shock patients who received invasive cardiac output (CO) monitoring following a 500 mL FC. Haemodynamics were recorded, and blood sampling for NT-proBNP values was performed. Patients were divided into responders and non-responders according to fluid responsiveness, which was defined as cardiac index (CI) increase ≥10% induced by FC. The NT-proBNP and the CI changes were analysed using Pearson correlation. The area under the curve (AUC) for NT-proBNP was used to test its ability to distinguish responders and non-responders. Subgroup analyses were also explored.

Results

Among 79 patients, there were 55 responders. High NT-proBNP values were common in the study cohort. Baseline NT-proBNP values were comparable between responders and non-responders. In general, NT-proBNP values were not significantly correlated with CI changes after FC (r=-0.104, P=0.361). Similarly, the NT-proBNP baseline values could not identify responders to FC with an AUC of 0.508 (95% confidence interval, 0.369-0.647). This result was further confirmed in the subgroup analyses.

Conclusions

Baseline NT-proBNP concentration value may not serve as an indicator of fluid responsiveness in patients with septic shock and should not be an indicator to withhold fluid loading.

Effects of time delay and body temperature on measurements of central venous oxygen saturation, venous-arterial blood carbon dioxide partial pressures difference, venous-arterial blood carbon dioxide partial pressures difference/arterial-venous oxygen difference ratio and lactate

BACKGROUND

Central venous oxygen saturation (ScvO₂), venous-arterial blood carbon dioxide partial pressures difference (Pv-aCO₂), venous-arterial blood carbon dioxide partial pressures difference/arterial-venous oxygen difference ratio (Pv-aCO₂/Ca-vO₂) and lactate are important parameters employed during shock resuscitation. We designed this study to confirm the effects of time delay and body temperature on measurements of these four parameters.

METHODS

Arterial and central venous blood samples were simultaneously drawn by plastic syringes via indwelling intra-arterial and central venous catheters from critically ill patients. Blood gas analyses were performed on both samples and repeated after 10, 20, 30, 40, 50 and 60 min. Patients were divided into a control group and a high temperature group according to whether the body temperature was greater than 38 °C.

RESULTS

A total of 30 critically ill patients were enrolled. There was a trend of increasing values for ScvO₂, Pv-aCO₂, Pv-aCO₂/Ca-vO₂ and lactate over time (P < 0.001). The ScvO₂ differences were all lower in high temperature group after 10, 20, 30, 40, 50 and 60 min when compared to the corresponding differences in the control group (P < 0.05). The differences in lactate values were slightly higher in the high temperature group, relative to the control group after 20, 30, 40, 50 and 60 min (P < 0.05).

CONCLUSIONS

Measurements of ScvO₂, Pv-aCO₂, lactate and Pv-aCO₂/Ca-vO₂ were affected by time delay or body temperature. We recommend that arterial and central venous blood gas samples be analyzed quickly within 10 min, especially for patients with body temperature <38 °C.

TRIAL REGISTRATION

ChiCTR, ChiCTR1800014484 . Registered 16 January 2018.