Despite major advances in critical care and surgical techniques the incidence of morbidity and mortality due to sepsis or septic shock has not decreased substantially. A recent epidemiology survey estimated there are 751 000 cases per year of severe sepsis in the USA with a mortality of 28.6% . Sepsis and septic shock are associated with both a relative and an absolute intravascular volume deficit . A variety of mediators and inflammatory cascading reactions that occur in sepsis induce increased microvascular permeability, which in turn results in interstitial fluid accumulation and tissue oedema .
Sepsis and septic shock are, therefore, characterized by a reduction of the cardiac preload and cardiac output resulting in arterial hypotension associated with impaired tissue perfusion and organ oxygenation causing organ hypoperfusion and organ dysfunction. In this clinical situation adequate fluid resuscitation aiming at restoring and maintaining circulating plasma volume is fundamental in order to improve organ perfusion and nutritive microcirculatory flow. Recently, it has been shown that early goal-directed fluid resuscitation in patients with severe sepsis and septic shock provides significant benefits related to outcome . With this respect the evaluation of the actual degree of hypovolaemia is very important. One dynamic parameter of the response to fluid resuscitation may be the variation of the stroke volume. It has been described as a functional preload parameter to predict fluid responsiveness in the perioperative situation . The increase of intrathoracic pressure during inspiration in the mechanically ventilated patient leads to a temporary reduction of cardiac preload, and in consequence to a temporarily reduced stroke volume. This phenomenon can often be observed in hypovolaemic patients as undulating variations in the arterial pressure curve trace .
Our hypothesis was that stroke volume variation (SVV) during mechanical ventilation of the lungs would allow prediction and assessment of changes in cardiac index (CI) in response to fluid loading in patients with severe sepsis.
The study was approved by the institutional Ethics Committee and informed consent was obtained from the next of kin of the patients. Ten patients with severe sepsis or septic shock were studied. Severe sepsis and septic shock were defined using the criteria of the American College of Chest Physicians/Society of Critical Care Consensus Conference Guideline . Severity of illness was characterized by the Sequential Organ Failure Assessment (SOFA) Score . No patient had a history of congestive heart failure. Patients with a cardiac rhythm other than sinus rhythm were not included. All patients received antibiotic therapy based on microbiological results.
All patients were receiving continuous infusions of analgesic and sedating agents, intubated and received controlled mechanical ventilation of the lungs throughout the study period. PaO2 was kept between 9 and 14 kPa and PaCO2 was maintained between 4.6 and 7.0 kPa. The lung ventilation was pressure controlled to reach a tidal volume between 6-8 mL kg−1. No changes in the ventilator setting were made during the study period. All patients required continuous infusions of a vasopressor agent (norepinephrine).
A pulmonary artery catheter (139HF75, 7.5F®; Edwards Lifesciences Ltd, Newbury, UK), for the measurement of central venous and pulmonary artery occlusion pressure (PAOP), was placed in all patients. A thermistor tipped arterial catheter (PV2015L, 5F Pulsiocath®; Pulsion Medical Systems, Munich, Germany) was inserted into the femoral artery of all patients. This catheter allows additional measurement of cardiac output by transpulmonary thermodilution and of derived variables including intrathoracic blood volume index (ITBVI), continuous cardiac output by pulse-contour analysis and SVV. The correct position of the catheters was confirmed by chest radiography. All transducers were positioned at the mid-axillary line and zeroed to atmospheric pressure.
The arterial thermodilution catheter was connected to a monitor for pulse-contour analysis (PiCCO®; Pulsion Medical Systems, Munich, Germany). Prior to the start of the study, three consecutive measurements of cardiac output by transpulmonary thermodilution were performed by injecting iced glucose 5% 15 mL randomly throughout the respiratory circle into the atrial port of the pulmonary artery catheter to calibrate the pulse-contour cardiac output. Heart rate, systolic, diastolic and mean arterial pressure, pulmonary arterial pressure and central venous pressure (CVP) were registered continuously, as were pulse-contour cardiac output, pulse-contour stroke volume and SVV. End-diastolic PAOP was determined at end-expiration from a tracing provided by the haemodynamic monitor and averaged from three consecutive cycles. Cardiac output and intrathoracic blood volume (ITBV) were determined by transpulmonary thermodilution technique, using an average of three determinations obtained by injection of iced glucose 5% 15 mL randomly throughout the respiratory cycle.
Calculation of SVV
The SVV expresses the variation of beat-to-beat stroke volume around the mean during the respiratory cycle and is calculated by the following equation:
The algorithm used during this investigation utilises a continuously sliding time window of 30 s to calculate the mean SV (SVmean). The time window is divided in four 7.5 s periods; within each 7.5 s, the highest (SVmax) and the lowest (SVmin) values of stroke volume were used to calculate SVV.
Calculation of ITBV
For calculation of ITBV, mean transit time and exponential decay time of the injected indicator between the injection site (right atrium) and detection site (abdominal aorta) and thermodilution cardiac output were registered and analysed. All raw data were indexed using the respective body surface area to calculate CI and ITBVI.
During periods of haemodynamic stability patients were initially monitored for 15 min to establish baseline values. Subsequently, all patients received hydroxyethylstarch 10% 500 mL (MW 200 kD, 0.5 degree of substitution, B Braun, Melsungen, Germany) over 30 min under continuous monitoring of the cardiac filling pressures. Fluid loading was indicated if PAOP < 14 mmHg or urine output < 0.5 mL kg−1h−1. Measurements were performed at T1: t = 0 min; T2: t = 3 min; T3: t = 6 min. Measurements after fluid loading were performed at T4: t = 45 min (immediately after fluid loading); T5: t = 48 min: T6: t = 51 min. No changes in vasopressor therapy were made during the period of measurements. Ventilator settings were not altered.
All haemodynamic data are presented as mean ± standard deviation (SD). Data were statistically analysed using SPSS for Windows® (SPSS, Chicago, Ill, USA: release 10.07). After verifying normal distribution of the data (skewness <1.0), we determined whether the haemodynamic parameters had changed in relation to fluid loading using paired t-tests with Bonferroni's adjustment. The relation between each preload parameter measured at baseline (T1) and the changes of CI in response to fluid loading was tested using Pearson product moment correlation. Finally, the relation between changes of preload parameters and the changes of CI in response to fluid loading were tested using Pearson product moment correlation. For all comparisons, statistical significance was assumed if P < 0.05.
Sources of sepsis, patient characteristics data and SOFA scores are given in Table 1. After fluid loading SVV decreased significantly, whereas CVP, PAOP, ITBVI and CI increased significantly (Table 2). Baseline values of SVV, ITBVI, CVP and PAOP were compared to changes of CI (Fig. 1). Baseline values of SVV and ITBVI were significantly correlated to CI changes in response to fluid loading (r = 0.64; P = 0.02 and r = −0.73, 0.009, respectively) whereas neither baseline values of CVP (r = 0.41; P = 0.18) nor PAOP (r = 0.23; P = 0.26) revealed a significant correlation to changes in CI. After fluid loading changes in CI (Fig. 2) were significantly correlated to percentage changes in SVV (r = −0.65; P < 0.001) and changes in ITBVI (r = 0.52; P = 0.002), whereas changes in CI revealed no significant correlation to changes in CVP (r = 0.28; P = 0.07) nor to changes in PAOP (r = 0.29; P = 0.06).
Our data indicate that measurement of SVV allows prediction and monitoring of fluid responsiveness in ventilated patients with severe sepsis. Thus, SVV might improve haemodynamic treatment in these patients.
Until today, adequate monitoring of fluid resuscitation in the critically ill, especially in septic patients, remains a challenge. In this respect the usefulness of traditional parameters, e.g. PAOP and CVP have been questioned [9,10] because they assume a predictable relationship between measurements of cardiac filling pressures and actual cardiac volumes. However, this is not always the case in septic patients who have abnormal compliance of the left or right ventricle . Recently, a new haemodynamic-monitoring device using pulse-contour analysis has been developed (PiCCO, Pulsion Medical Systems, Munich, Germany). This method requires transpulmonary thermodilution for calibration and, therefore, the placement of central venous and arterial catheters only. The technique calculates left ventricular stroke volume from the impedance characteristics of the pulse pressure (PP) waveform using a complex algorithm .
Our study demonstrates that it may be possible to predict changes in CI to fluid loading using SVV in septic patients. The greater the SVV the more cardiac output can be expected to increase in response to fluid loading. These results are consistent with the findings of Reuter and colleagues,  and Berkenstadt and colleagues , who showed, in surgical patients perioperatively, the greater ability of the dynamic parameter SVV to predict and to monitor cardiac responsiveness to fluid loading in comparison to the static indicators of cardiac preload CVP and PAOP. In their studies neither CVP nor PAOP could predict responsiveness to fluid loading nor were changes of CI associated with changes of CVP or PAOP. In contrast, other investigators have suggested the use of PAOP  or CVP  to identify cardiac responsiveness to fluid loading. We were surprised to find that in our study ITBVI although being a static parameter, turned out to be a predictor of cardiac output response to fluid loading.
The effects of positive pressure ventilation-induced changes in left ventricular output have been assessed using other functional parameters. The systolic pressure variation (SPV), which is the difference between maximal and minimal systolic blood pressure values during one mechanical breath, has been shown to be a sensitive indicator of preload experimentally and clinically [15,16]. The delta down component of SPV, which is the decrease in systolic pressure from an apnoeic baseline, has been shown to be an accurate predictor of the response of CI to fluid loading in septic patients with hypotension . Changes in left ventricular stroke volume can be easily monitored as beat-to-beat changes in arterial PP since the only other determinants of PP, arterial resistance, and compliance cannot change enough to alter PP during a single breath . Michard and colleagues have shown that respiratory changes in PP (PP = maximal PP at inspiration (PPmax) minus minimal PP at expiration (PPmin) and calculated as: PP (%) = 100 (PPmax − PPmin)/(PPmax + PPmin)/2) predict the effect of fluid loading on CI in patients in ventilated septic patients . They compared PP and SPV to CVP and PAOP and demonstrated that PP and SPV are superior to CVP and PAOP in their ability to predict cardiac output response to fluid loading. Furthermore, the authors could show a slight but significant advantage of PP over SPV, which might be explained by the fact that measurement of left ventricular stroke volume by PP is not influenced by the intrathoracic pressure-induced changes of systolic arterial pressure. The value of dynamic parameters for prediction of fluid responsiveness and monitoring fluid administration has been emphasized in a recent review .
In our study fluid loading induced a significant reduction in SVV. At the same time reduced SVV was associated with an increased CI. SVV has previously been described as a measure to quantify responsiveness to preload changes perioperatively . In patients after elective cardiac surgery changes in SVV were correlated significantly with changes in stroke volume index and with changes in SPV . Using receiver operating characteristic curves, the area under the curve was statistically greater for SVV and SPV than for CVP. Thus, we consider SVV to be a dynamic preload measurement, which can be used as a guide to predict fluid responsiveness and to permit titration of fluid loading in various patient populations.
In accordance with our data, Brock and colleagues recently found in postoperative patients that ITBVI reflects the preload dependency of CI more accurately than CVP or PAOP . In ventilated patients with acute respiratory failure  and sepsis  it has been suggested that ITBVI is a better indicator of the cardiac preload than traditional pulmonary artery catheter derived CVP and PAOP. Similar to our results changes of CI have been associated with changes of ITBVI in both studies, but not with changes of cardiac filling pressures. In this context it is important to note that a linear relationship between ITBV and global end-diastolic volume (GEDV) has been found, where ITBV = 1.25 × GEDV . Thus, measurement of ITBVI via transpulmonary technique is based on a constant relationship between GEDV and ITBV.
There are some limitations of preload assessment using SVV: the analysis of the pulse-contour derived stroke volume is not possible in patients with dysrhythmias . As a rise in tidal volume may influence the magnitude of the respiratory variation in stroke volume , the patient must be fully sedated. Thereby, it is assured that the SVV is affected only by the intermittent positive ventilation and not by the patient's respiratory efforts. In patients with severe sepsis or septic shock, who are generally sedated, this does not appear to be a major limitation . For the measurement of SVV, cannulation of the femoral artery is required. While the need for arterial cannulation is obvious in ventilated septic patients, there are still some concerns about the infection risk associated with the femoral site, despite the fact that a similar rate of infection between the radial and femoral sites for arterial cannulation has been demonstrated in a series of 4932 patients . In general, according to a recent review arterial cannulation appears to be safe procedure as major complications after insertion of an arterial line occurred in less than 1% of cases .
We conclude that measuring SVV may be a useful way of guiding fluid therapy in ventilated patients with severe sepsis as it allows estimation of preload and prediction of CI changes in response to fluid loading.
This study was supported by a grant from Pulsion Medical Systems, Munich, Germany. It was presented in part to the Anaesthetic Research Society, 11-12 April, 2002 (Br J Anaesth 2002; 89: 194P).
1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med
2. Imm A, Carlson RW. Fluid resuscitation in circulatory shock. Crit Care Clin
3. Groeneveld AB, Bronsveld W, Thijs LG. Hemodynamic determinants of mortality in human septic shock. Surgery
4. Rivers E, Nguyen B, Havstad S, et al.
Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med
5. Berkenstadt H, Margalit N, Hadani M, et al.
Stroke volume variation as a predictor of fluid responsiveness in patients undergoing brain surgery. Anesth Analg
6. Reuter DA, Felbinger TW, Schmidt C, et al.
Stroke volume variations for assessment of cardiac responsiveness to volume loading in mechanically ventilated patients after cardiac surgery. Intensive Care Med
7. Bone RC, Balk RA, Cerra FB, et al.
Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest
8. Vincent JL, Moreno R, Takala J, et al.
The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med
9. Dalen JE, Bone RC. Is it time to pull the pulmonary artery catheter? JAMA
10. Weil MH. The assault on the Swan-Ganz catheter: a case history of constrained technology, constrained bedside clinicians, and constrained monetary expenditures. Chest
11. McLean AS. Echocardiography assessment of left ventricular function in the critically ill. Anaesth Intensive Care
12. Wesseling K, De Wit B, Weber J, Smith N. A simple device for the continuous measurement of cardiac output. Adv Cardiovasc Physiol
13. Packman MI, Rackow EC. Optimum left heart filling pressure during fluid resuscitation of patients with hypovolemic and septic shock. Crit Care Med
14. Magder S. More respect for the CVP. Intensive Care Med
15. Perel A, Pizov R, Cotev S. Systolic blood pressure variation is a sensitive indicator of hypovolemia in ventilated dogs subjected to graded hemorrhage. Anesthesiology
16. Coriat P, Vrillon M, Perel A, et al.
A comparison of systolic blood pressure variations and echocardiographic estimates of end-diastolic left ventricular size in patients after aortic surgery. Anesth Analg
17. Tavernier B, Makhotine O, Lebuffe G, Dupont J, Scherpereel P. Systolic pressure variation as a guide to fluid therapy in patients with sepsis-induced hypotension. Anesthesiology
18. Pinsky MR. Functional hemodynamic monitoring. Intensive Care Med
19. Michard F, Boussat S, Chemla D, et al.
Relation between respiratory changes in arterial pulse pressure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med
20. Michard F, Teboul JL. Predicting fluid responsiveness in ICU patients: a critical analysis of the evidence. Chest
21. Reuter DA, Felbinger TW, Kilger E, Schmidt C, Lamm P, Goetz AE. Optimizing fluid therapy in mechanically ventilated patients after cardiac surgery by on-line monitoring of left ventricular stroke volume variations. Comparison with aortic systolic pressure variations. Br J Anaesth
22. Brock H, Gabriel C, Bibl D, Necek S. Monitoring intra-vascular volumes for postoperative volume therapy. Eur J Anaesthesiol
23. Lichtwarck-Aschoff M, Zeravik J, Pfeiffer UJ. Intrathoracic blood volume accurately reflects circulatory volume status in critically ill patients with mechanical ventilation. Intensive Care Med
24. Sakka SG, Bredle DL, Reinhart K, Meier-Hellmann A. Comparison between intrathoracic blood volume and cardiac filling pressures in the early phase of hemodynamic instability of patients with sepsis or septic shock. J Crit Care
25. Sakka SG, Rühl CC, Pfeiffer UJ, et al.
Assessment of cardiac preload and extravascular lung water by single transpulmonary thermodilution. Intensive Care med
26. Michard F, Teboul JL. Respiratory changes in arterial pressure in mechanically ventilated patients. In: Vincent J, ed. Yearbook of Intensive Care and Emergency Medicine.
Berlin, Germany: Springer, 2000: 696-704.
27. Szold A, Pizov R, Segal E, Perel A. The effect of tidal volume and intravascular volume state on systolic pressure variation in ventilated dogs. Intensive Care Med
28. Frezza EE, Mezghebe H. Indications and complications of arterial catheter use in surgical or medical intensive care units: analysis of 4932 patients. Am Surg
29. Scheer B, Perel A, Pfeiffer UJ. Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care