- Email: [email protected]
Exploring the best predictors of fluid responsiveness in patients with septic shock
Accepted Manuscript Exploring the best predictors of fluid responsiveness in patients with septic shock
Nianfang Lu, Xiuming Xi, Li Jiang, Degang Yang, Kai Yin PII: DOI: Reference:
S0735-6757(17)30219-X doi: 10.1016/j.ajem.2017.03.052 YAJEM 56570
To appear in: Received date: Revised date: Accepted date:
9 January 2017 15 February 2017 21 March 2017
Please cite this article as: Nianfang Lu, Xiuming Xi, Li Jiang, Degang Yang, Kai Yin , Exploring the best predictors of fluid responsiveness in patients with septic shock. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Yajem(2017), doi: 10.1016/j.ajem.2017.03.052
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Exploring the Best Predictors of Fluid Responsiveness in Patients with Septic Shock Nianfang Lu 1, Xiuming Xi 1*, Li Jiang1, Degang Yang 2, Kai Yin 3 1
Department of Critical Care Medicine, Fuxing Hospital, Capital Medical University,
2
SC RI PT
Beijing, China, 100038
Department of Spinal and neural functional reconstruction, China Rehabilitation Research Center, School of Rehabilitation Medicine, Capital Medical University,
Department of Critical Care Medicine, Beijing Electric Power Hospital, Beijing, China, 100073
ED
*Correspondence to: Xiuming Xi
MA
3
NU
Beijing, China, 100068
PT
Email:[email protected]
CE
Tel: +86-010-13801244610
AC
Acknowledgements
This study was funded by the Ministry of science and technology of the People’s Republic of China (grant number 2012BAI11B05) and Beijing Municipal Science and Technology Commission (BSTC)(grant number D101100050010058).
ACCEPTED MANUSCRIPT Exploring the Best Predictors of Fluid Responsiveness in Patients with Septic Shock
AC
CE
PT
ED
MA
NU
SC RI PT
Running title：Peripheral artery peak velocity and fluid status
ACCEPTED MANUSCRIPT
Abstract Objective: To evaluate respiratory variations in carotid and brachial peak velocity and other hemodynamic parameters to predict responsiveness to fluid challenge.
SC RI PT
Methods: A prospective observational study was performed on mechanically ventilated patients with septic shock. Outcomes included the measurements of central venous pressure, intrathoracic blood volume index, stroke volume variation (SVV), pleth variability index(PVI), and ultrasound assessments of respiratory variations in
NU
inferior vena cava diameter (ΔIVC), carotid Doppler peak velocity (ΔCDPV), and
MA
brachial artery peak velocity (ΔVpeak brach).
Results: All patients received 200ml normal saline challenge. There were 27
ED
responders and 22 non-responders. Responders had higher SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach measurements. In addition, all these measurements had
PT
statistically significant linear correlations with changes in cardiac index (CI).When
CE
responders were defined by ΔCI ≥ 10%, receiver operating characteristics (ROC)
AC
curve analysis showed that fluid responsiveness could be predicted:11.5% optimal cut-off 1evels of SVV with sensitivity of 75% and specificity of 85%, 15.5% optimal cut-off 1evels of PVI with sensitivity of 65% and
specificity of 80%, 20.5% optimal
cut-off 1evels of ΔIVC with sensitivity of 67% and specificity of 77%, 13% optimal cut-off 1evels of ΔCDPV with sensitivity of 78%％ and specificity of 90%, 11.7% optimal cut-off 1evels of ΔVpeak brach with sensitivity of 70% and specificity of 80%.
ACCEPTED MANUSCRIPT
Conclusion: Ultrasound assessment of ΔIVC and ΔVpeak brach, especially ΔCDPV, could predict fluid responsiveness and might be recommended as a continuous and noninvasive method to monitor functional hemodynamic parameter in mechanically
SC RI PT
ventilated patients with septic shock.
Keywords: septic shock, ventilation, Doppler ultrasound, inferior vena cava diameter,
AC
CE
PT
ED
MA
NU
carotid Doppler peak velocity, brachial artery peak velocity
ACCEPTED MANUSCRIPT Introduction Septic shock is a serious infectious condition characterized by low blood pressure and multiple organ damage. One of the traditional recommendations is to administer intravenous fluids as the first step to improve blood pressure[1, 2]. However, studies
SC RI PT
have shown that not every patient benefits from aggressive intravenous hydration[3, 4]. Only about 40% of hypotensive patients with sepsis respond to fluid infusion with improvement in blood pressure and outcomes[5, 6]. Those who do not respond to fluid infusion are liable to develop high intravascular pressure, pulmonary edema, heart
NU
failure with a high associated mortality[7-9]. Therefore, it is crucial to develop a
MA
hemodynamically-guided approach to evaluate volume status and to identify patients who are likely to benefit from fluid administration.
ED
Previous studies have shown that certain parameters may correlate with volume status. The traditional static parameters, such as central venous pressure (CVP), pulmonary
PT
wedge pressure, and intrathoracic blood volume index (ITBVI), have been shown not to
CE
correlate with patient volume status[10, 11]. Hemodynamic parameters, such as stroke
AC
volume variation (SVV) and pleth variability index (PVI) may better predict fluid responsiveness. However, assessments of these parameters require invasive procedures and special monitoring equipment, which limits their clinical application [12]. In recent years, ultrasound has been proposed as a tool to help guide fluid resuscitation
[13, 14]
. According to the Frank-Starling curve, when patients are in the
low volume status, the cardiac preload is low and the curve is in the rising phase,therefore intrathoracic pressure fluctuations by breathing could have a greater
ACCEPTED MANUSCRIPT impact on cardiac stroke volume(SV)[15, 16]. The variation of SV may be assessed by variation of arterial blood peak velocity on the Doppler ultrasound. At last, it leads the higher variation of SV and arterial blood peak velocity. Studies have shown that respiratory variation in aortic blood peak velocity had high sensitivity and specificity
SC RI PT
to predict fluid responsiveness[17-19]. However, measurements of aortic blood flow velocity require transesophageal ultrasound which is an invasive procedure. Measurements of femoral artery blood flow are frequently affected by changes in intra-abdominal pressure. Measures of carotid or brachial artery flow were recently [18-21]
. Both these peripheral arteries are
NU
shown to predict fluid responsiveness
MA
relatively superficial large vessels which can provide easy ultrasound evaluation and high-quality images. However, assessment of respiratory variation in artery peak
ED
velocity in these two arteries in ventilated patients with septic shock has not been studied.
PT
In the current study, we measured the respiratory variation in arterial blood peak
CE
velocity in carotid and brachial arteries and compared their use against that of other
AC
static and hemodynamic parameters for predicting fluid responsiveness in ventilated patients with septic shock. Clinical application of these measures is discussed.
Materials and Methods Study design and patient selection A prospective observation study was performed in the Intensive Care Unit in our hospital between January 2012 and December 2015. Study protocol was approved by
ACCEPTED MANUSCRIPT the Institutional Ethics Committee. Written informed consent was obtained from every patient’s health care proxy. Inclusion criteria were: 1) age ≥ 18 years; 2) patients who met the diagnostic criteria for septic shock, which was defined as systolic blood pressure (SBP) <90 mmHg, or
SC RI PT
mean arterial pressure (MAP) <70 mmHg, or SBP decreases 40 mmHg or less than two standard deviations below normal for age in the absence of other causes of hypotension[1]; 3) mechanical ventilation was prescribed and administered by clinical physicians. Ventilator settings followed the hospital written protocols. Exclusion
NU
criteria were: 1) any contraindication to fluid resuscitation, such as congestive heart failure or evidence of fluid overload; 2) pregnant women; 3) patients with neurogenic
MA
shock, cerebrovascular accident, or traumatic brain injury; 4) conditions which could affect abdominal ultrasound, such as abdominal compartment syndrome, flatulence,
ED
and patients who had undergone upper abdominal surgery; 5) arrhythmia; 6)
CE
PT
peripheral vascular disease or stenosis.
Study protocol and outcome measurements
AC
Patients’ baseline characteristics, including gender, age, body mass index, source of infection, SOFA (Sequential Organ Failure Assessment) and APACHE (Acute Physiology and Chronic Health Evaluation) scores, were recorded. All patients received fluid challenge with a rapid infusion of 200 mL of normal saline administered via a central venous line within 10 minutes[22]. CO was monitored by PiCCO (PiCCO Plus, Pulsion Medical Systems, Munich, Germany).
ACCEPTED MANUSCRIPT Cardiac index was calculated as(cardiac output)/(body surface area). Patients who showed an increase in cardiac index of ≥ 10% were categorized as responders; those who showed < 10% increase in cardiac index were categorized as non-responders group.
SC RI PT
Central venous pressure (CVP) was monitored via a central venous catheter (ARROW, Arrow international, INC. New Jersey, USA); intrathoracic blood volume index (ITBVI) and stroke volume variation (SVV) were assessed using a PiCCO system. Pleth variability index (PVI) was monitored by pleth variability index
NU
machine (Masimo, Radical-7, USA) and was calculated from respiratory variations in
MA
pulse oximeter.
Inferior vena cava was evaluated by a subcostal long axis view with a 4 MHz
ED
frequency ultrasound probe (Sonosite, WA, USA). A time-motion record of the IVC diameter was generated by M-mode imaging at 2cm from the right atrium. Maximum
PT
and minimum diameters of the IVC were recorded within one respiratory cycle and
CE
were repeated three times. Respiratory variation in inferior vena cava diameter
AC
(ΔIVC) was calculated as ΔIVC = (IVCmax - IVCmin) / IVCmin×100%. Carotid artery was identified by a 12 MHz frequency ultrasound probe (Sonosite, WA, USA) transversely placed at the inferolateral border of the thyroid cartilage. Then, 2 cm below the carotid artery bifurcation, probe was turned 90 degrees to show longitudinal view of the carotid artery. Pulsed Doppler analysis was performed at the center of the vessel, with an angulation of no more than 60 degrees. Maximum and minimum peak systolic velocities were recorded in a single respiratory cycle;
ACCEPTED MANUSCRIPT measurements were repeated three times. Respiratory variation in carotid Doppler peak velocity (ΔCDPV) was calculated as 2 × (CDPVmax - CDPVmin) / (CDPVmax + CDPVmin) ×100%. Brachial artery was examined with pulsed Doppler analysis (Sonosite, WA, USA) at
SC RI PT
the elbow fossa when the patients were in supine positions. Maximum and minimum peak velocity in a single respiratory cycle was recorded and repeated three times. Respiratory variations in brachial artery peak velocity (ΔVpeakbrach) was calculated as (maxVpeak brach -
minVpeak
brach) / [(maxVpeak brach +
brach)/2] ×
NU
100%.
minVpeak
before and after fluid challenge.
MA
All these hemodynamic parameters were measured by certified ultrasound technicians
ED
All patients were administered mechanical ventilation (VT 8–10 ml/kg, PEEP 5-12 cmH2O), antibiotics, vasoactive agents, sedative and analgesic medications, as
CE
PT
determined by the treating physicians according to each patient's situation.
AC
Statistical analysis
Numerical data are presented as mean ± standard deviation, when appropriate, and analyzed using student t-test. Correlations were assessed on Pearson correlation analysis. Predictive value of the measured parameters for volume resuscitation was evaluated on receiver operating characteristic (ROC) curve analysis, and presented as area under curve (AUC) and 95% Confidence Intervals (CI). Statistical analyses were
ACCEPTED MANUSCRIPT performed with SPSS version 17.0 (SPSS, IBM, United States). P< 0.05 was considered statistically significant.
Results
SC RI PT
A total of 49 patients were enrolled in the study, of which 27 patients were categorized as responders on fluid challenge (responder group) and 22 as non-responders (non-responder group). The baseline characteristics are summarized in Table 1. There were no statistically significant differences in demographic and
NU
clinical variables. Most common source of infection was respiratory infection.
MA
Before fluid challenge, patients in the responder group had higher SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach as compared to that in patients in the non-responder
ED
group (Table 2). After fluid challenge, there were no statistically significant between-group differences in these hemodynamic parameters.
PT
A positive correlation of SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach measured
CE
prior to fluid challenge with change in cardiac index after fluid challenge was
AC
observed (Table 3).
ΔCDPV was found to be the best predictor of fluid responsiveness value for fluid responsiveness. A cut-off value of 13% was associated with 78% sensitivity and 90% specificity; AUC 0.910) (Table 4, Figure 1).
Discussion
ACCEPTED MANUSCRIPT In the present study, we compared the predictive values of several measures for fluid responsiveness in ventilated patients with septic shock. Respiratory variation in carotid Doppler peak velocity (ΔCDPV) showed the best predictive value for fluid responsiveness in this patient population.
SC RI PT
Intravenous fluid infusion was recommended as a key element in the treatment of patients with septic shock. Fluid infusion increases the preload on the heart, augments cardiac output, and improves tissue perfusion. However, not every patient with septic shock benefits from fluid resuscitation[3, 4]. Previous studies have recommended the
NU
use of CVP, PAWP, ITBVI, SVV, and PVI as measures to predict fluid
MA
responsiveness. However, these measurements either require invasive procedures or could not provide adequate accuracy in predicting fluid responsiveness.
ED
CVP was relatively easy to monitor. However, we did not observe any significant difference in CVP between responders and non-responders. In addition, there was no
PT
correlation between CVP and change in cardiac index. ROC analysis also showed a
CE
relatively low sensitivity and specificity of CVP to predict fluid responsiveness.
AC
These results are consistent with those reported elsewhere[23-25]. ITBVI reflects cardiac end-diastolic volume. It is not affected by changes in intrathoracic pressure, myocardial compliance, mechanical ventilation, and use of medications. In the current study, ROC analysis showed ITBVI has a sensitivity and specificity slightly higher than those of CVP, but in this study there was no correlation between ITBVI and change in cardiac index. In addition, it is a static
ACCEPTED MANUSCRIPT parameter and could not be used to indicate fluid responsiveness in patients with shock[26].
The present study showed a statistically significant difference in SVV between responders and non-responders. ROC analysis also showed a relatively high
SC RI PT
sensitivity and specificity of SVV to predict responsiveness to fluid challenge. However, SVV measurement is an invasive procedure with high costs[27,
28]
.
Moreover, PiCCO cannot monitor cardiac diastolic function and right heart function.
NU
All of these limited its clinical applications.
PVI measurement is a continuous calculation of respiratory variation in the pulse
MA
oximeter waveform amplitude. It has been shown to predict fluid responsiveness in
ED
patients in shock, and our results are consistent with previous research conclusions[29]. However, PVI does not indicate cardiac function which limits its clinical application.
PT
Bedside ultrasound examination is increasingly being used in most hospitals and
CE
intensive care units. It is relatively easy to learn with low medical expenses. In the current study, we compared the predictive values of measurements at IVC, carotid
AC
artery, and brachial artery by bedside ultrasound. The study showed that delta CDPV had greater predictive value than delta Vpeak brach and delta IVC. The last two are limited in that the brachial artery is more distally located and the inferior vena cava is affected by intra-abdominal pressures. Limitations of the current study include its single-centre scope and a relatively small sample size. We only studied ventilated patients. A few patients on mechanical ventilation received muscle relaxants, whereas most patients with septic shock had
ACCEPTED MANUSCRIPT spontaneous respiration. In addition, patients might receive different antibiotics, vasoactive agents, sedative and analgesic medications as determined by their treating physicians. These could cause potential bias in our study result. Future studies with a larger sample size of patients with spontaneous respiration are required.
SC RI PT
In summary, our study showed that bedside ultrasound could be used to predict the response of septic shock patients to fluid challenge and to distinguish fluid responders from non-responders. Respiratory carotid artery peak velocity variation had a higher predictive value, sensitivity and specificity as compared to that of measurements in
AC
CE
PT
ED
MA
NU
the IVC and brachial artery.
ACCEPTED MANUSCRIPT References [1] Dellinger R P, Levy M M, Rhodes A, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012[J]. Intensive Care Med, 2013,39(2):165-228.
SC RI PT
[2] Dellinger R P. The Surviving Sepsis Campaign: 2013 and beyond[J]. Chin Med J (Engl), 2013,126(10):1803-1805.
[3] Vandervelden S, Malbrain M L. Initial resuscitation from severe sepsis: one size does not fit all[J]. Anaesthesiol Intensive Ther, 2015,47 Spec No:s44-s55.
NU
[4] Groeneveld A B. Fluids in septic shock: too much of a good thing?[J]. Crit Care,
MA
2010,14(1):101.
[5] Thiel S W, Kollef M H, Isakow W. Non-invasive stroke volume measurement
ED
and passive leg raising predict volume responsiveness in medical ICU patients: an observational cohort study[J]. Crit Care, 2009,13(4):R111.
PT
[6] Preau S, Saulnier F, Dewavrin F, et al. Passive leg raising is predictive of fluid
CE
responsiveness in spontaneously breathing patients with severe sepsis or acute
AC
pancreatitis[J]. Crit Care Med, 2010,38(3):819-825. [7] Smith J S. Current recommendations for diagnosis and management of sepsis and septic shock[J]. JAAPA, 2013,26(10):42-45. [8] Mallat J, Meddour M, Durville E, et al. Decrease in pulse pressure and stroke volume variations
after mini-fluid
challenge accurately predicts
fluid
responsivenessdagger[J]. Br J Anaesth, 2015,115(3):449-456. [9] Lammi M R, Aiello B, Burg G T, et al. Response to fluid boluses in the fluid and
ACCEPTED MANUSCRIPT catheter treatment trial[J]. Chest, 2015,148(4):919-926. [10] Bindels A J, van der Hoeven J G, Graafland A D, et al. Relationships between volume and pressure measurements and stroke volume in critically ill patients[J]. Crit Care, 2000,4(3):193-199.
SC RI PT
[11] Hu B, Xiang H, Liang H, et al. Assessment effect of central venous pressure in fluid resuscitation in the patients with shock: a multi-center retrospective research[J]. Chin Med J (Engl), 2013,126(10):1844-1849.
[12] Pinsky M R. Functional haemodynamic monitoring[J]. Curr Opin Crit Care,
NU
2014,20(3):288-293.
MA
[13] Schefold J C, Storm C, Bercker S, et al. Inferior vena cava diameter correlates with invasive hemodynamic measures in mechanically ventilated intensive care
ED
unit patients with sepsis[J]. J Emerg Med, 2010,38(5):632-637. [14] Muller L, Bobbia X, Toumi M, et al. Respiratory variations of inferior vena cava
PT
diameter to predict fluid responsiveness in spontaneously breathing patients with
CE
acute circulatory failure: need for a cautious use[J]. Crit Care, 2012,16(5):R188.
AC
[15] Hofer C K, Cannesson M. Monitoring fluid responsiveness[J]. Acta Anaesthesiol Taiwan, 2011,49(2):59-65. [16] Liu N, Gu Q, Yu J F. [The influence of positive end-expiratory pressure on stroke volume variation for the accuracy of evaluating volume][J]. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue, 2012,24(7):419-422. [17] Feissel M, Michard F, Mangin I, et al. Respiratory changes in aortic blood velocity as an indicator of fluid responsiveness in ventilated patients with septic
ACCEPTED MANUSCRIPT shock[J]. Chest, 2001,119(3):867-873. [18] Ibarra-Estrada M A, Lopez-Pulgarin J A, Mijangos-Mendez J C, et al. Respiratory variation in carotid peak systolic velocity predicts volume responsiveness in mechanically ventilated patients with septic shock: a
SC RI PT
prospective cohort study[J]. Crit Ultrasound J, 2015,7(1):29. [19] Song Y, Kwak Y L, Song J W, et al. Respirophasic carotid artery peak velocity variation as a predictor of fluid responsiveness in mechanically ventilated patients with coronary artery disease[J]. Br J Anaesth, 2014,113(1):61-66.
NU
[20] Monge G M, Gil C A, Diaz M J. Brachial artery peak velocity variation to predict
MA
fluid responsiveness in mechanically ventilated patients[J]. Crit Care, 2009,13(5):R142.
ED
[21] Zhu W, Wan L, Wan X, et al. [Measurement of brachial artery velocity variation and inferior vena cava variability to estimate fluid responsiveness][J]. Zhonghua
PT
Wei Zhong Bing Ji Jiu Yi Xue, 2016,28(8):713-717.
CE
[22] Wujtewicz M. Fluid use in adult intensive care[J]. Anaesthesiol Intensive Ther,
AC
2012,44(2):92-95.
[23] Osman D, Ridel C, Ray P, et al. Cardiac filling pressures are not appropriate to predict hemodynamic response to
volume challenge[J]. Crit Care Med,
2007,35(1):64-68. [24] Magder S. Understanding central venous pressure: not a preload index?[J]. Curr Opin Crit Care, 2015,21(5):369-375. [25] Marik P E, Baram M, Vahid B. Does central venous pressure predict fluid
ACCEPTED MANUSCRIPT responsiveness? A systematic review of the literature and the tale of seven mares[J]. Chest, 2008,134(1):172-178. [26] Proulx F, Lemson J, Choker G, et al. Hemodynamic monitoring by transpulmonary thermodilution and pulse contour analysis in critically ill
SC RI PT
children[J]. Pediatr Crit Care Med, 2011,12(4):459-466. [27] Angappan S, Parida S, Vasudevan A, et al. The comparison of stroke volume variation with central venous pressure in predicting fluid responsiveness in septic patients
with
acute
circulatory failure[J].
J
Crit
Care
Med,
NU
2015,19(7):394-400.
Indian
MA
[28] Slagt C, Malagon I, Groeneveld A B. Systematic review of uncalibrated arterial pressure waveform analysis to determine cardiac output and stroke volume
ED
variation[J]. Br J Anaesth, 2014,112(4):626-637. [29] Sandroni C, Cavallaro F, Marano C, et al. Accuracy of plethysmographic indices
review
CE
systematic
PT
as predictors of fluid responsiveness in mechanically ventilated adults: a
AC
2012,38(9):1429-1437.
and
meta-analysis[J].
Intensive
Care
Med,
ACCEPTED MANUSCRIPT
Table 1. Baseline characteristics of study participants Responder group (N = Non-responder group (N = 22)
SC RI PT
27) 55.7 ± 12.6
55.0 ± 12.8
Gender, male / female, N
19 / 8
14 / 8
Body mass index, kg/m2, mean ± SD
24.6 ± 9.3
APACHE II score, mean ± SD
26.5 ± 10.0
MA
NU
Age, year, mean ± SD
27.2 ± 10.5
18.3 ± 7.2
18.6 ± 7.5
Respiratory tract
18 (66.7%)
15 (68.2%)
Urinary tract
4 (14.8%)
3 (13.6%)
Gastrointestinal
2 (7.4%)
1 (4.5%)
Hematogenous
1 (3.7%)
1 (4.5%)
2 (7.4%)
2 (9.1%)
ED
SOFA score, mean ± SD
25.3 ± 9.5
AC
CE
PT
Sources of infection, N (%)
Others
SOFA: Sequential Organ Failure Assessment APACHE: Acute Physiology and Chronic Health Evaluation
ACCEPTED MANUSCRIPT Table 2. Comparison of hemodynamic parameters between responder and non-responder groups After fluid challenge
Responder group
Non-responder group
Responder group
Non-respond er group
CVP (mmHg)
7.3 ± 3.2
8.0 ± 3.6
9.8 ± 3.8
10.6 ± 3.9
ITBVI (mL/m2)
880.2 ± 185.3
841.2 ± 190.0
SC RI PT
Before fluid challenge
Hemodynamic parameters
932.3
± 928.3 ± 202.5
210.8
13.5 ± 2.3
9.0 ± 3.1 *
11.0 ± 4.0
8.9 ± 3.0
PVI (%)
16.3 ± 3.1
12.5 ± 3.5 *
15.9 ± 3.3
11.8 ± 3.0
ΔIVC (%)
23.3 ± 5.2
16.5 ± 3.8 *
16.3 ± 4.2
14.2 ± 2.3
ΔCDPV (%)
15.2 ± 3.2
10.2 ± 2.5 *
12.0 ± 2.5
10.0 ± 3.0
ΔVpeak brach (%)
14.6 ± 3.4
9.5 ± 2.5 *
11.5 ± 2.8
8.8 ± 2.1
PT
ED
MA
NU
SVV (%)
* P< 0.05
CE
Data expressed as mean ± standard deviation
AC
CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory variations in brachial artery peak velocity.
ACCEPTED MANUSCRIPT Table 3. Correlation between hemodynamic parameters and changes in cardiac index on fluid challenge P value
CVP
-0.331
0.13
ITBVI
-0.347
0.15
SVV
0.843
PVI
0.811
ΔIVC
0.805
ΔCDPV
0.852
0.01
ΔVpeak brach
0.803
0.04
SC RI PT
Hemodynamic parameters r value prior to fluid challenge
0.02
ED
MA
NU
0.03
0.03
PT
CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke
CE
volume variation; PVI, pleth variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity;
AC
ΔVpeak brach, respiratory variations in brachial artery peak velocity.
ACCEPTED MANUSCRIPT Table 4. ROCanalysis for prediction of fluid responsiveness 95% Confidence P value Interval
Threshol d
Specifi Sensitivit city y (%) (%)
CVP
0.675
0.506 - 0.844
0.058
6.5
65
70
ITBVI
0.664
0.493 - 0.835
0.076
871
55
65
SVV
0.848
0.726 - 0.969
0.000
11.5
75
85
PVI
0.816
0.686 - 0.946
0.001
15.5
65
80
ΔIVC
0.805
0.671 - 0.939
0.001
20.5
67
77
ΔCDPV
0.910
0.817 - 1.0
< 0.001
13.0
78
90
ΔVpeak
0.761
0.604 – 0.918
0.005
11.7
70
80
ED
MA
NU
SC RI PT
Hemodynami AUC c parameters
brach
PT
AUC, Area under the curve; CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth variability index;
CE
ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory variations in
AC
brachial artery peak velocity.
ACCEPTED MANUSCRIPT Figure Legends: Figure 1. ROC curve analysis for different hemodynamic parameters for predicting fluid responsiveness in ventilated patients with septic shock. When responders were defined byΔCI ≥ 10%, 11.5% optimal cut-off 1evels of SVV predicted fluid responsiveness with a sensitivity of 75% and specificity of 85%; 15.5% optimal cut-off 1evels of PVI predicted fluid responsiveness with a sensitivity of 65% and
SC RI PT
specificity of 80%. 20.5% optimal cut-off 1evels of ΔIVC predicted fluid responsiveness with a sensitivity of 67% and specificity of 77%. 13% optimal cut-off 1evels of ΔCDPV predicted fluid responsiveness with a sensitivity of 78% and specificity of 90%. 11.7% optimal cut-off 1evels of ΔVpeak brachpredicted fluid
NU
responsiveness with a sensitivity of 70% and specificity of 80%.
MA
ROC, receiver operating characteristics; CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth
ED
variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory
AC
CE
PT
variations in brachial artery peak velocity.
SC RI PT
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
MA
NU
Fig. 1
Nianfang Lu, Xiuming Xi, Li Jiang, Degang Yang, Kai Yin PII: DOI: Reference:
S0735-6757(17)30219-X doi: 10.1016/j.ajem.2017.03.052 YAJEM 56570
To appear in: Received date: Revised date: Accepted date:
9 January 2017 15 February 2017 21 March 2017
Please cite this article as: Nianfang Lu, Xiuming Xi, Li Jiang, Degang Yang, Kai Yin , Exploring the best predictors of fluid responsiveness in patients with septic shock. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Yajem(2017), doi: 10.1016/j.ajem.2017.03.052
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Exploring the Best Predictors of Fluid Responsiveness in Patients with Septic Shock Nianfang Lu 1, Xiuming Xi 1*, Li Jiang1, Degang Yang 2, Kai Yin 3 1
Department of Critical Care Medicine, Fuxing Hospital, Capital Medical University,
2
SC RI PT
Beijing, China, 100038
Department of Spinal and neural functional reconstruction, China Rehabilitation Research Center, School of Rehabilitation Medicine, Capital Medical University,
Department of Critical Care Medicine, Beijing Electric Power Hospital, Beijing, China, 100073
ED
*Correspondence to: Xiuming Xi
MA
3
NU
Beijing, China, 100068
PT
Email:[email protected]
CE
Tel: +86-010-13801244610
AC
Acknowledgements
This study was funded by the Ministry of science and technology of the People’s Republic of China (grant number 2012BAI11B05) and Beijing Municipal Science and Technology Commission (BSTC)(grant number D101100050010058).
ACCEPTED MANUSCRIPT Exploring the Best Predictors of Fluid Responsiveness in Patients with Septic Shock
AC
CE
PT
ED
MA
NU
SC RI PT
Running title：Peripheral artery peak velocity and fluid status
ACCEPTED MANUSCRIPT
Abstract Objective: To evaluate respiratory variations in carotid and brachial peak velocity and other hemodynamic parameters to predict responsiveness to fluid challenge.
SC RI PT
Methods: A prospective observational study was performed on mechanically ventilated patients with septic shock. Outcomes included the measurements of central venous pressure, intrathoracic blood volume index, stroke volume variation (SVV), pleth variability index(PVI), and ultrasound assessments of respiratory variations in
NU
inferior vena cava diameter (ΔIVC), carotid Doppler peak velocity (ΔCDPV), and
MA
brachial artery peak velocity (ΔVpeak brach).
Results: All patients received 200ml normal saline challenge. There were 27
ED
responders and 22 non-responders. Responders had higher SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach measurements. In addition, all these measurements had
PT
statistically significant linear correlations with changes in cardiac index (CI).When
CE
responders were defined by ΔCI ≥ 10%, receiver operating characteristics (ROC)
AC
curve analysis showed that fluid responsiveness could be predicted:11.5% optimal cut-off 1evels of SVV with sensitivity of 75% and specificity of 85%, 15.5% optimal cut-off 1evels of PVI with sensitivity of 65% and
specificity of 80%, 20.5% optimal
cut-off 1evels of ΔIVC with sensitivity of 67% and specificity of 77%, 13% optimal cut-off 1evels of ΔCDPV with sensitivity of 78%％ and specificity of 90%, 11.7% optimal cut-off 1evels of ΔVpeak brach with sensitivity of 70% and specificity of 80%.
ACCEPTED MANUSCRIPT
Conclusion: Ultrasound assessment of ΔIVC and ΔVpeak brach, especially ΔCDPV, could predict fluid responsiveness and might be recommended as a continuous and noninvasive method to monitor functional hemodynamic parameter in mechanically
SC RI PT
ventilated patients with septic shock.
Keywords: septic shock, ventilation, Doppler ultrasound, inferior vena cava diameter,
AC
CE
PT
ED
MA
NU
carotid Doppler peak velocity, brachial artery peak velocity
ACCEPTED MANUSCRIPT Introduction Septic shock is a serious infectious condition characterized by low blood pressure and multiple organ damage. One of the traditional recommendations is to administer intravenous fluids as the first step to improve blood pressure[1, 2]. However, studies
SC RI PT
have shown that not every patient benefits from aggressive intravenous hydration[3, 4]. Only about 40% of hypotensive patients with sepsis respond to fluid infusion with improvement in blood pressure and outcomes[5, 6]. Those who do not respond to fluid infusion are liable to develop high intravascular pressure, pulmonary edema, heart
NU
failure with a high associated mortality[7-9]. Therefore, it is crucial to develop a
MA
hemodynamically-guided approach to evaluate volume status and to identify patients who are likely to benefit from fluid administration.
ED
Previous studies have shown that certain parameters may correlate with volume status. The traditional static parameters, such as central venous pressure (CVP), pulmonary
PT
wedge pressure, and intrathoracic blood volume index (ITBVI), have been shown not to
CE
correlate with patient volume status[10, 11]. Hemodynamic parameters, such as stroke
AC
volume variation (SVV) and pleth variability index (PVI) may better predict fluid responsiveness. However, assessments of these parameters require invasive procedures and special monitoring equipment, which limits their clinical application [12]. In recent years, ultrasound has been proposed as a tool to help guide fluid resuscitation
[13, 14]
. According to the Frank-Starling curve, when patients are in the
low volume status, the cardiac preload is low and the curve is in the rising phase,therefore intrathoracic pressure fluctuations by breathing could have a greater
ACCEPTED MANUSCRIPT impact on cardiac stroke volume(SV)[15, 16]. The variation of SV may be assessed by variation of arterial blood peak velocity on the Doppler ultrasound. At last, it leads the higher variation of SV and arterial blood peak velocity. Studies have shown that respiratory variation in aortic blood peak velocity had high sensitivity and specificity
SC RI PT
to predict fluid responsiveness[17-19]. However, measurements of aortic blood flow velocity require transesophageal ultrasound which is an invasive procedure. Measurements of femoral artery blood flow are frequently affected by changes in intra-abdominal pressure. Measures of carotid or brachial artery flow were recently [18-21]
. Both these peripheral arteries are
NU
shown to predict fluid responsiveness
MA
relatively superficial large vessels which can provide easy ultrasound evaluation and high-quality images. However, assessment of respiratory variation in artery peak
ED
velocity in these two arteries in ventilated patients with septic shock has not been studied.
PT
In the current study, we measured the respiratory variation in arterial blood peak
CE
velocity in carotid and brachial arteries and compared their use against that of other
AC
static and hemodynamic parameters for predicting fluid responsiveness in ventilated patients with septic shock. Clinical application of these measures is discussed.
Materials and Methods Study design and patient selection A prospective observation study was performed in the Intensive Care Unit in our hospital between January 2012 and December 2015. Study protocol was approved by
ACCEPTED MANUSCRIPT the Institutional Ethics Committee. Written informed consent was obtained from every patient’s health care proxy. Inclusion criteria were: 1) age ≥ 18 years; 2) patients who met the diagnostic criteria for septic shock, which was defined as systolic blood pressure (SBP) <90 mmHg, or
SC RI PT
mean arterial pressure (MAP) <70 mmHg, or SBP decreases 40 mmHg or less than two standard deviations below normal for age in the absence of other causes of hypotension[1]; 3) mechanical ventilation was prescribed and administered by clinical physicians. Ventilator settings followed the hospital written protocols. Exclusion
NU
criteria were: 1) any contraindication to fluid resuscitation, such as congestive heart failure or evidence of fluid overload; 2) pregnant women; 3) patients with neurogenic
MA
shock, cerebrovascular accident, or traumatic brain injury; 4) conditions which could affect abdominal ultrasound, such as abdominal compartment syndrome, flatulence,
ED
and patients who had undergone upper abdominal surgery; 5) arrhythmia; 6)
CE
PT
peripheral vascular disease or stenosis.
Study protocol and outcome measurements
AC
Patients’ baseline characteristics, including gender, age, body mass index, source of infection, SOFA (Sequential Organ Failure Assessment) and APACHE (Acute Physiology and Chronic Health Evaluation) scores, were recorded. All patients received fluid challenge with a rapid infusion of 200 mL of normal saline administered via a central venous line within 10 minutes[22]. CO was monitored by PiCCO (PiCCO Plus, Pulsion Medical Systems, Munich, Germany).
ACCEPTED MANUSCRIPT Cardiac index was calculated as(cardiac output)/(body surface area). Patients who showed an increase in cardiac index of ≥ 10% were categorized as responders; those who showed < 10% increase in cardiac index were categorized as non-responders group.
SC RI PT
Central venous pressure (CVP) was monitored via a central venous catheter (ARROW, Arrow international, INC. New Jersey, USA); intrathoracic blood volume index (ITBVI) and stroke volume variation (SVV) were assessed using a PiCCO system. Pleth variability index (PVI) was monitored by pleth variability index
NU
machine (Masimo, Radical-7, USA) and was calculated from respiratory variations in
MA
pulse oximeter.
Inferior vena cava was evaluated by a subcostal long axis view with a 4 MHz
ED
frequency ultrasound probe (Sonosite, WA, USA). A time-motion record of the IVC diameter was generated by M-mode imaging at 2cm from the right atrium. Maximum
PT
and minimum diameters of the IVC were recorded within one respiratory cycle and
CE
were repeated three times. Respiratory variation in inferior vena cava diameter
AC
(ΔIVC) was calculated as ΔIVC = (IVCmax - IVCmin) / IVCmin×100%. Carotid artery was identified by a 12 MHz frequency ultrasound probe (Sonosite, WA, USA) transversely placed at the inferolateral border of the thyroid cartilage. Then, 2 cm below the carotid artery bifurcation, probe was turned 90 degrees to show longitudinal view of the carotid artery. Pulsed Doppler analysis was performed at the center of the vessel, with an angulation of no more than 60 degrees. Maximum and minimum peak systolic velocities were recorded in a single respiratory cycle;
ACCEPTED MANUSCRIPT measurements were repeated three times. Respiratory variation in carotid Doppler peak velocity (ΔCDPV) was calculated as 2 × (CDPVmax - CDPVmin) / (CDPVmax + CDPVmin) ×100%. Brachial artery was examined with pulsed Doppler analysis (Sonosite, WA, USA) at
SC RI PT
the elbow fossa when the patients were in supine positions. Maximum and minimum peak velocity in a single respiratory cycle was recorded and repeated three times. Respiratory variations in brachial artery peak velocity (ΔVpeakbrach) was calculated as (maxVpeak brach -
minVpeak
brach) / [(maxVpeak brach +
brach)/2] ×
NU
100%.
minVpeak
before and after fluid challenge.
MA
All these hemodynamic parameters were measured by certified ultrasound technicians
ED
All patients were administered mechanical ventilation (VT 8–10 ml/kg, PEEP 5-12 cmH2O), antibiotics, vasoactive agents, sedative and analgesic medications, as
CE
PT
determined by the treating physicians according to each patient's situation.
AC
Statistical analysis
Numerical data are presented as mean ± standard deviation, when appropriate, and analyzed using student t-test. Correlations were assessed on Pearson correlation analysis. Predictive value of the measured parameters for volume resuscitation was evaluated on receiver operating characteristic (ROC) curve analysis, and presented as area under curve (AUC) and 95% Confidence Intervals (CI). Statistical analyses were
ACCEPTED MANUSCRIPT performed with SPSS version 17.0 (SPSS, IBM, United States). P< 0.05 was considered statistically significant.
Results
SC RI PT
A total of 49 patients were enrolled in the study, of which 27 patients were categorized as responders on fluid challenge (responder group) and 22 as non-responders (non-responder group). The baseline characteristics are summarized in Table 1. There were no statistically significant differences in demographic and
NU
clinical variables. Most common source of infection was respiratory infection.
MA
Before fluid challenge, patients in the responder group had higher SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach as compared to that in patients in the non-responder
ED
group (Table 2). After fluid challenge, there were no statistically significant between-group differences in these hemodynamic parameters.
PT
A positive correlation of SVV, PVI, ΔIVC, ΔCDPV, and ΔVpeak brach measured
CE
prior to fluid challenge with change in cardiac index after fluid challenge was
AC
observed (Table 3).
ΔCDPV was found to be the best predictor of fluid responsiveness value for fluid responsiveness. A cut-off value of 13% was associated with 78% sensitivity and 90% specificity; AUC 0.910) (Table 4, Figure 1).
Discussion
ACCEPTED MANUSCRIPT In the present study, we compared the predictive values of several measures for fluid responsiveness in ventilated patients with septic shock. Respiratory variation in carotid Doppler peak velocity (ΔCDPV) showed the best predictive value for fluid responsiveness in this patient population.
SC RI PT
Intravenous fluid infusion was recommended as a key element in the treatment of patients with septic shock. Fluid infusion increases the preload on the heart, augments cardiac output, and improves tissue perfusion. However, not every patient with septic shock benefits from fluid resuscitation[3, 4]. Previous studies have recommended the
NU
use of CVP, PAWP, ITBVI, SVV, and PVI as measures to predict fluid
MA
responsiveness. However, these measurements either require invasive procedures or could not provide adequate accuracy in predicting fluid responsiveness.
ED
CVP was relatively easy to monitor. However, we did not observe any significant difference in CVP between responders and non-responders. In addition, there was no
PT
correlation between CVP and change in cardiac index. ROC analysis also showed a
CE
relatively low sensitivity and specificity of CVP to predict fluid responsiveness.
AC
These results are consistent with those reported elsewhere[23-25]. ITBVI reflects cardiac end-diastolic volume. It is not affected by changes in intrathoracic pressure, myocardial compliance, mechanical ventilation, and use of medications. In the current study, ROC analysis showed ITBVI has a sensitivity and specificity slightly higher than those of CVP, but in this study there was no correlation between ITBVI and change in cardiac index. In addition, it is a static
ACCEPTED MANUSCRIPT parameter and could not be used to indicate fluid responsiveness in patients with shock[26].
The present study showed a statistically significant difference in SVV between responders and non-responders. ROC analysis also showed a relatively high
SC RI PT
sensitivity and specificity of SVV to predict responsiveness to fluid challenge. However, SVV measurement is an invasive procedure with high costs[27,
28]
.
Moreover, PiCCO cannot monitor cardiac diastolic function and right heart function.
NU
All of these limited its clinical applications.
PVI measurement is a continuous calculation of respiratory variation in the pulse
MA
oximeter waveform amplitude. It has been shown to predict fluid responsiveness in
ED
patients in shock, and our results are consistent with previous research conclusions[29]. However, PVI does not indicate cardiac function which limits its clinical application.
PT
Bedside ultrasound examination is increasingly being used in most hospitals and
CE
intensive care units. It is relatively easy to learn with low medical expenses. In the current study, we compared the predictive values of measurements at IVC, carotid
AC
artery, and brachial artery by bedside ultrasound. The study showed that delta CDPV had greater predictive value than delta Vpeak brach and delta IVC. The last two are limited in that the brachial artery is more distally located and the inferior vena cava is affected by intra-abdominal pressures. Limitations of the current study include its single-centre scope and a relatively small sample size. We only studied ventilated patients. A few patients on mechanical ventilation received muscle relaxants, whereas most patients with septic shock had
ACCEPTED MANUSCRIPT spontaneous respiration. In addition, patients might receive different antibiotics, vasoactive agents, sedative and analgesic medications as determined by their treating physicians. These could cause potential bias in our study result. Future studies with a larger sample size of patients with spontaneous respiration are required.
SC RI PT
In summary, our study showed that bedside ultrasound could be used to predict the response of septic shock patients to fluid challenge and to distinguish fluid responders from non-responders. Respiratory carotid artery peak velocity variation had a higher predictive value, sensitivity and specificity as compared to that of measurements in
AC
CE
PT
ED
MA
NU
the IVC and brachial artery.
ACCEPTED MANUSCRIPT References [1] Dellinger R P, Levy M M, Rhodes A, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012[J]. Intensive Care Med, 2013,39(2):165-228.
SC RI PT
[2] Dellinger R P. The Surviving Sepsis Campaign: 2013 and beyond[J]. Chin Med J (Engl), 2013,126(10):1803-1805.
[3] Vandervelden S, Malbrain M L. Initial resuscitation from severe sepsis: one size does not fit all[J]. Anaesthesiol Intensive Ther, 2015,47 Spec No:s44-s55.
NU
[4] Groeneveld A B. Fluids in septic shock: too much of a good thing?[J]. Crit Care,
MA
2010,14(1):101.
[5] Thiel S W, Kollef M H, Isakow W. Non-invasive stroke volume measurement
ED
and passive leg raising predict volume responsiveness in medical ICU patients: an observational cohort study[J]. Crit Care, 2009,13(4):R111.
PT
[6] Preau S, Saulnier F, Dewavrin F, et al. Passive leg raising is predictive of fluid
CE
responsiveness in spontaneously breathing patients with severe sepsis or acute
AC
pancreatitis[J]. Crit Care Med, 2010,38(3):819-825. [7] Smith J S. Current recommendations for diagnosis and management of sepsis and septic shock[J]. JAAPA, 2013,26(10):42-45. [8] Mallat J, Meddour M, Durville E, et al. Decrease in pulse pressure and stroke volume variations
after mini-fluid
challenge accurately predicts
fluid
responsivenessdagger[J]. Br J Anaesth, 2015,115(3):449-456. [9] Lammi M R, Aiello B, Burg G T, et al. Response to fluid boluses in the fluid and
ACCEPTED MANUSCRIPT catheter treatment trial[J]. Chest, 2015,148(4):919-926. [10] Bindels A J, van der Hoeven J G, Graafland A D, et al. Relationships between volume and pressure measurements and stroke volume in critically ill patients[J]. Crit Care, 2000,4(3):193-199.
SC RI PT
[11] Hu B, Xiang H, Liang H, et al. Assessment effect of central venous pressure in fluid resuscitation in the patients with shock: a multi-center retrospective research[J]. Chin Med J (Engl), 2013,126(10):1844-1849.
[12] Pinsky M R. Functional haemodynamic monitoring[J]. Curr Opin Crit Care,
NU
2014,20(3):288-293.
MA
[13] Schefold J C, Storm C, Bercker S, et al. Inferior vena cava diameter correlates with invasive hemodynamic measures in mechanically ventilated intensive care
ED
unit patients with sepsis[J]. J Emerg Med, 2010,38(5):632-637. [14] Muller L, Bobbia X, Toumi M, et al. Respiratory variations of inferior vena cava
PT
diameter to predict fluid responsiveness in spontaneously breathing patients with
CE
acute circulatory failure: need for a cautious use[J]. Crit Care, 2012,16(5):R188.
AC
[15] Hofer C K, Cannesson M. Monitoring fluid responsiveness[J]. Acta Anaesthesiol Taiwan, 2011,49(2):59-65. [16] Liu N, Gu Q, Yu J F. [The influence of positive end-expiratory pressure on stroke volume variation for the accuracy of evaluating volume][J]. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue, 2012,24(7):419-422. [17] Feissel M, Michard F, Mangin I, et al. Respiratory changes in aortic blood velocity as an indicator of fluid responsiveness in ventilated patients with septic
ACCEPTED MANUSCRIPT shock[J]. Chest, 2001,119(3):867-873. [18] Ibarra-Estrada M A, Lopez-Pulgarin J A, Mijangos-Mendez J C, et al. Respiratory variation in carotid peak systolic velocity predicts volume responsiveness in mechanically ventilated patients with septic shock: a
SC RI PT
prospective cohort study[J]. Crit Ultrasound J, 2015,7(1):29. [19] Song Y, Kwak Y L, Song J W, et al. Respirophasic carotid artery peak velocity variation as a predictor of fluid responsiveness in mechanically ventilated patients with coronary artery disease[J]. Br J Anaesth, 2014,113(1):61-66.
NU
[20] Monge G M, Gil C A, Diaz M J. Brachial artery peak velocity variation to predict
MA
fluid responsiveness in mechanically ventilated patients[J]. Crit Care, 2009,13(5):R142.
ED
[21] Zhu W, Wan L, Wan X, et al. [Measurement of brachial artery velocity variation and inferior vena cava variability to estimate fluid responsiveness][J]. Zhonghua
PT
Wei Zhong Bing Ji Jiu Yi Xue, 2016,28(8):713-717.
CE
[22] Wujtewicz M. Fluid use in adult intensive care[J]. Anaesthesiol Intensive Ther,
AC
2012,44(2):92-95.
[23] Osman D, Ridel C, Ray P, et al. Cardiac filling pressures are not appropriate to predict hemodynamic response to
volume challenge[J]. Crit Care Med,
2007,35(1):64-68. [24] Magder S. Understanding central venous pressure: not a preload index?[J]. Curr Opin Crit Care, 2015,21(5):369-375. [25] Marik P E, Baram M, Vahid B. Does central venous pressure predict fluid
ACCEPTED MANUSCRIPT responsiveness? A systematic review of the literature and the tale of seven mares[J]. Chest, 2008,134(1):172-178. [26] Proulx F, Lemson J, Choker G, et al. Hemodynamic monitoring by transpulmonary thermodilution and pulse contour analysis in critically ill
SC RI PT
children[J]. Pediatr Crit Care Med, 2011,12(4):459-466. [27] Angappan S, Parida S, Vasudevan A, et al. The comparison of stroke volume variation with central venous pressure in predicting fluid responsiveness in septic patients
with
acute
circulatory failure[J].
J
Crit
Care
Med,
NU
2015,19(7):394-400.
Indian
MA
[28] Slagt C, Malagon I, Groeneveld A B. Systematic review of uncalibrated arterial pressure waveform analysis to determine cardiac output and stroke volume
ED
variation[J]. Br J Anaesth, 2014,112(4):626-637. [29] Sandroni C, Cavallaro F, Marano C, et al. Accuracy of plethysmographic indices
review
CE
systematic
PT
as predictors of fluid responsiveness in mechanically ventilated adults: a
AC
2012,38(9):1429-1437.
and
meta-analysis[J].
Intensive
Care
Med,
ACCEPTED MANUSCRIPT
Table 1. Baseline characteristics of study participants Responder group (N = Non-responder group (N = 22)
SC RI PT
27) 55.7 ± 12.6
55.0 ± 12.8
Gender, male / female, N
19 / 8
14 / 8
Body mass index, kg/m2, mean ± SD
24.6 ± 9.3
APACHE II score, mean ± SD
26.5 ± 10.0
MA
NU
Age, year, mean ± SD
27.2 ± 10.5
18.3 ± 7.2
18.6 ± 7.5
Respiratory tract
18 (66.7%)
15 (68.2%)
Urinary tract
4 (14.8%)
3 (13.6%)
Gastrointestinal
2 (7.4%)
1 (4.5%)
Hematogenous
1 (3.7%)
1 (4.5%)
2 (7.4%)
2 (9.1%)
ED
SOFA score, mean ± SD
25.3 ± 9.5
AC
CE
PT
Sources of infection, N (%)
Others
SOFA: Sequential Organ Failure Assessment APACHE: Acute Physiology and Chronic Health Evaluation
ACCEPTED MANUSCRIPT Table 2. Comparison of hemodynamic parameters between responder and non-responder groups After fluid challenge
Responder group
Non-responder group
Responder group
Non-respond er group
CVP (mmHg)
7.3 ± 3.2
8.0 ± 3.6
9.8 ± 3.8
10.6 ± 3.9
ITBVI (mL/m2)
880.2 ± 185.3
841.2 ± 190.0
SC RI PT
Before fluid challenge
Hemodynamic parameters
932.3
± 928.3 ± 202.5
210.8
13.5 ± 2.3
9.0 ± 3.1 *
11.0 ± 4.0
8.9 ± 3.0
PVI (%)
16.3 ± 3.1
12.5 ± 3.5 *
15.9 ± 3.3
11.8 ± 3.0
ΔIVC (%)
23.3 ± 5.2
16.5 ± 3.8 *
16.3 ± 4.2
14.2 ± 2.3
ΔCDPV (%)
15.2 ± 3.2
10.2 ± 2.5 *
12.0 ± 2.5
10.0 ± 3.0
ΔVpeak brach (%)
14.6 ± 3.4
9.5 ± 2.5 *
11.5 ± 2.8
8.8 ± 2.1
PT
ED
MA
NU
SVV (%)
* P< 0.05
CE
Data expressed as mean ± standard deviation
AC
CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory variations in brachial artery peak velocity.
ACCEPTED MANUSCRIPT Table 3. Correlation between hemodynamic parameters and changes in cardiac index on fluid challenge P value
CVP
-0.331
0.13
ITBVI
-0.347
0.15
SVV
0.843
PVI
0.811
ΔIVC
0.805
ΔCDPV
0.852
0.01
ΔVpeak brach
0.803
0.04
SC RI PT
Hemodynamic parameters r value prior to fluid challenge
0.02
ED
MA
NU
0.03
0.03
PT
CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke
CE
volume variation; PVI, pleth variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity;
AC
ΔVpeak brach, respiratory variations in brachial artery peak velocity.
ACCEPTED MANUSCRIPT Table 4. ROCanalysis for prediction of fluid responsiveness 95% Confidence P value Interval
Threshol d
Specifi Sensitivit city y (%) (%)
CVP
0.675
0.506 - 0.844
0.058
6.5
65
70
ITBVI
0.664
0.493 - 0.835
0.076
871
55
65
SVV
0.848
0.726 - 0.969
0.000
11.5
75
85
PVI
0.816
0.686 - 0.946
0.001
15.5
65
80
ΔIVC
0.805
0.671 - 0.939
0.001
20.5
67
77
ΔCDPV
0.910
0.817 - 1.0
< 0.001
13.0
78
90
ΔVpeak
0.761
0.604 – 0.918
0.005
11.7
70
80
ED
MA
NU
SC RI PT
Hemodynami AUC c parameters
brach
PT
AUC, Area under the curve; CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth variability index;
CE
ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory variations in
AC
brachial artery peak velocity.
ACCEPTED MANUSCRIPT Figure Legends: Figure 1. ROC curve analysis for different hemodynamic parameters for predicting fluid responsiveness in ventilated patients with septic shock. When responders were defined byΔCI ≥ 10%, 11.5% optimal cut-off 1evels of SVV predicted fluid responsiveness with a sensitivity of 75% and specificity of 85%; 15.5% optimal cut-off 1evels of PVI predicted fluid responsiveness with a sensitivity of 65% and
SC RI PT
specificity of 80%. 20.5% optimal cut-off 1evels of ΔIVC predicted fluid responsiveness with a sensitivity of 67% and specificity of 77%. 13% optimal cut-off 1evels of ΔCDPV predicted fluid responsiveness with a sensitivity of 78% and specificity of 90%. 11.7% optimal cut-off 1evels of ΔVpeak brachpredicted fluid
NU
responsiveness with a sensitivity of 70% and specificity of 80%.
MA
ROC, receiver operating characteristics; CVP, central venous pressure; ITBVI, intrathoracic blood volume index; SVV, stroke volume variation; PVI, pleth
ED
variability index; ΔIVC, respiratory variation in inferior vena cava diameter; ΔCDPV, respiratory variation in carotid Doppler peak velocity; ΔVpeak brach, respiratory
AC
CE
PT
variations in brachial artery peak velocity.
SC RI PT
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
MA
NU
Fig. 1