Weaning from artificial respiration: value of continuous monitoring of mixed venous oxygen saturation

Weaning from mechanical ventilation is particularly difficult in patients with combined cardiac and respiratory failure. Continuous monitoring of mixed venous blood oxygen saturation (SvO2) redefines weaning in terms of tissue oxygenation. A stable SvO2 greater than 60% during weaning is a reliable index of weanability. However, further studies are required to establish a tolerance threshold for SvO2 during weaning. In the limited experience reported here, an immediate and abrupt fall in SvO2, when the patient started to breathe spontaneously was invariably associated with difficulties in weaning. In some patients, other signs of left ventricular dysfunction rapidly ensued, with a fall in cardiac index. Weaning remained possible if the treatment was capable of increasing cardiac output and normalizing SvO2. If, during spontaneous breathing, SvO2 remained stable in the 50-55% range, with no significant decrease in cardiac output, abrupt and unpredictable drops of SvO2 under 40% range occurred. Such falls always preceded signs of tissue hypoxia, leading to a resumption of controlled mechanical ventilation. However, further studies are required to fully delineate the role of SvO2 in the fine tuning of inotropic support and ventilatory assistance in the difficult weaning of patients recovering from cardio-respiratory failure.     

What is the role of monitoring of mixed venous blood saturation in cardiac surgery?

The equipment available for mixed venous blood saturation (Svo2) monitoring is now accurate. SvO2 is not a direct measure of cardiac output, because it depends on the balance between oxygen delivery (TaO2) and consumption (VO2). As haemoglobin affinity for oxygen increases during cardio-pulmonary bypass (CPB), the optimal level of SvO2 after CPB should be above 65-70%. There is a critical level of TaO2 below which VO2 is dependent on TaO2. Below this level, SvO2 has no clinical value as it no longer depends on TaO2. Similarly, SvO2 has no clinical value during lactic acidosis. When these limitations are taken into account, SvO2 monitoring is useful for the interpretation of intra- and post-operative haemodynamic alterations occurring during cardiac surgery. It is particularly indicated in patients with preoperative NYHA class III or IV congestive heart failure.     

Venous oxygen saturation during normovolaemic haemodilution in the pig

BACKGROUND: Hypovolaemia may be considered to represent a volume-restricted cardiac output (CO), but CO varies inversely with the haemoglobin concentration (Hb) and a maximal mixed venous oxygen saturation (SvO2) may be a better target for volume administration than a maximal CO. METHODS: In 10 anaesthetized pigs, volume loading with 6% hydroxyethyl starch was performed to obtain a maximal SvO2 followed by normovolaemic haemodilution with 6% hydroxyethyl starch. RESULTS: Volume loading increased SvO2 from 55.0+/-5.2% to 64.8+/-9.0% (mean+/-SD) associated with an increase in CO (2.3+/-0.4 to 3.5+/-0.9 l/min) and central venous oxygen saturation (ScvO2; 68.2+/-9.3% to 79.4+/-7.2%; P<0.05). Heart rate (HR), mean arterial (MAP), central venous (CVP), pulmonary arterial mean (PAMP), and occlusion pressures (PAOP) increased as well (P<0.05). In contrast, during progressive haemodilution, SvO2 and ScvO2 remained statistically unchanged until the haemoglobin concentration had decreased from 5.5+/-0.4 to 2.9+/-0.2 mM, while CO and HR increased at a haemoglobin value of 4.4+/-0.4 and 4.0+/-0.4 mM and CVP and PAOP decreased at a haemoglobin of 4.0+/-0.4 and 2.9+/-0.2 mM, respectively (P<0.05) leaving MAP unaffected. CONCLUSION: This study found that volume loading increased cardiac output and mixed and central venous oxygen saturations in parallel, but during normovolaemic haemodilution an increase in cardiac output left mixed and central venous oxygen saturations statistically unchanged until haemoglobin concentration was reduced by approximately 50%. Accordingly, volume therapy should be directed to maintain a high venous oxygen saturation rather than a change in cardiac output.     

Mixed venous oxygen saturation during mobilization after cardiac surgery: are reflectance oximetry catheters reliable?

BACKGROUND: Oximetry catheters immediately reflect changes in mixed venous oxygen saturation (SvO2). We have used the Baxter 2-SAT system to register changes in SvO2 during early mobilizations after cardiac surgery. To assess catheter reliability, readings were compared to blood gases. METHODS: A total of 352 paired catheter and bench haemoximetry measurements were obtained at the expected highest and lowest levels of SvO2 during the mobilization procedures. The agreement between methods was explored by a Bland-Altman plot. The influence of haemoglobin (Hgb), pH, cardiac output (CO), posture, catheter identity and catheter calibration on agreement was assessed through analysis of covariance. RESULTS: Data included a substantial number of low SvO2 values, 95 paired means of SvO2 < or = 50% and 37 paired means < or = 40%. Mean oxygen saturation difference between catheter and haemoximeter readings was -1.6 +/- 5.7% (SD). Agreement between the methods depended upon the level of SvO2. At SvO2 of 65%, the two methods were virtually identical. Below 65%, the catheters increasingly underestimated the corresponding haemoximetric values by 1.5% for every 10% reduction in SvO2. Agreement was to some degree dependent on individual calibrations and catheter identity, but to a lesser extent on Hgb, CO and posture. CONCLUSION: The two methods are interchangeable for most clinical purposes. Catheter readings are, however, substantially lower than the corresponding haemoximetric measurements at low SvO2 values. Careful interpretation of the absolute values resulting from catheter measurements is recommended, especially when SvO2 readings are low.     

Continuous venous oximetry in surgical patients.

A prospective study was performed to evaluate the efficacy of continuous venous oximetry to supplement traditional hemodynamic monitoring in 39 critically ill surgical patients. There was no statistically significant difference in SvO2 between the continuous in vivo values and in vitro values (0.694 +/- 0.095 vs. 0.698 +/- 0.108). There was no statistically significant correlation between continuously measured SvO2 and PaO2 (r = 0.09, p greater than 0.5), SaO2 (r = 0.08, p greater than 0.5), or oxygen consumption (r = 0.46, p greater than 0.5). There was a slight but statistically significant correlation between continuously measured SvO2 and cardiac output (r = 0.40, p less than 0.025) and oxygen delivery (r = 0.49, p less than 0.005). There was a highly significant correlation between continuously measured SvO2 and oxygen utilization coefficient (r = -0.96, p less than 0.001). Continuously measured SvO2 is a reliable predictor of SvO2 measured intermittently by in vitro methods. In critically ill surgical patients, SvO2 does not correlate highly with the individual determinants of oxygen transport but rather correlates with the oxygen utilization coefficient and therefore reflects the overall balance between oxygen consumption and delivery.     

Continuous venous oximetry for hemodynamic and oxygen transport stability post cardiac surgery.

Under the diagnostic-related group (DRG) reimbursement system, hospitals are looking to decrease costs related to unnecessary laboratory measurements. To assess the efficacy of continuous SvO2 as the only means to monitor the balance of the oxygen transport of the stable postoperative cardiac patient in the ICU, we studied 26 adult patients undergoing cardiac surgery with an uneventful postoperative course. All subjects had an Opticath fiberoptic PA catheter inserted for 29.6. +/- 11.0 hours (range 16-66) and spent an average of 42.4 +/- 17.5 hours in the Intensive Care Unit (range 20-87). Cardiac output, and Hemoglobin/Hematocrit were determined serially every 2 hours during the first 6 postoperative hours and 4 hours respectively according to our ICU practice. Arterial blood gases were determined freely in relation to changes in the hemodynamic and respiratory status. No clinical decisions were undertaken on the basis of SvO2. Retrospectively it was determined whether basing decisions on the SvO2 would have reduced the number of unnecessary cardiac outputs, ABGs and Hgb/Hcts. Using the SvO2 as potential indicator of hemodynamic and oxygen transport stability it could significantly reduce the number of determinations per patient, ie, cardiac output (11.7 +/- 4.2 vs 2.1 +/- 0.3, p less than 0.05), ABGs (11.3 +/- 2.8 vs 2.8 +/- 0.4, p less than 0.05) and Hgb/Hcts (5.7 +/- 1.3 vs 2.0 +/- 0.0, p less than 0.05). The use of SvO2 would save the hospital $84.5 +/- 27.5 (range 31.5 +/- 140.9) per stable patient in the ICU and a total of 220.4 +/- 69.9 minutes (range 90-300) of ICU nursing time.(ABSTRACT TRUNCATED AT 250 WORDS)     

Routine SvO2 measurement after CABG surgery with a surgically introduced pulmonary artery catheter.

OBJECTIVE: It has been argued that the poor correlation between cardiac output and mixed venous oxygen saturation (SvO2) reduces the value of SvO2. Routine use of Swan Ganz catheters is also controversial in cardiac surgery. Here our clinical experience with a simplified method for routine hemodynamic monitoring and the short-term prognostic value of SvO2 after CABG surgery is presented. METHOD: Peroperatively an epidural catheter is routinely introduced through the outflow tract of the right ventricle into the pulmonary artery for monitoring of pressure and blood sampling. Clinical data were retrospectively retrieved from the records and related to SvO2 routinely obtained on admission to the ICU after 488 CABG procedures. RESULTS: Average SvO2 on arrival to ICU was 67+/-7%. The SvO2 value of 55% represented a cut off point below which a high incidence of complications were found. Outcome after 456 procedures with SvO2 > or = 55% compared with 32 procedures with SvO2 < 55%: mortality 0 vs. 9.4% (P = 0.0003), perioperative myocardial infarction 6.2 vs. 29% (P < 0.0001), ventilator treatment 8.9+/-10.1 vs. 25.7+/-54.9 h (P = 0.0074), ICU stay 1.4+/-1.2 vs. 2.1+/-1.7 days (P = 0.0010). CONCLUSIONS: SvO2 was of prognostic value and due to its specificity it seems particularly useful for telling which patients are unlikely to develop cardiorespiratory problems. Thus, this simple method for hemodynamic monitoring could contribute to cost containment as it seems that we can safely reserve Swan Ganz catheters for high-risk patients.     

Mixed venous oximetry

We now have the technology through reflectance spectrophotometry to evaluate and display continuously mixed venous oxygen saturation SvO2 through use of a modified pulmonary artery catheter. Adding this method of assessing the balance of oxygen supply and demand to our standard armamentarium of hemodynamic monitoring may improve our ability to diagnose and treat cardiovascular aberrations at an earlier stage than was previously possible. Through analysis of the Fick equation, it can be seen that SvO2 depends upon the cardiac output, the arterial oxygen saturation, the hemoglobin level, and the rate of oxygen consumption. These are, in turn, affected by a great number of factors (see Fig 8). As seen in the variety of patient care examples cited above, the usefulness of SvO2 monitoring continues to grow. It appears that there are no intrinsic risks associated with SvO2 monitoring beyond those of customary PA monitoring. This new technology provides us with online information not previously available, at an associated cost that needs to be further examined.     

Post-operative myocardial dysfunction does not affect the physiological response to early mobilization after coronary artery bypass grafting

BACKGROUND: An acute increase in oxygen demand can be compensated for either by increased cardiac index (CI) or increased oxygen extraction, resulting in reduced mixed venous oxygen saturation (SvO2). We tested the hypothesis that post-operative cardiac dysfunction may explain why oxygen extraction alone is increased during early mobilization after cardiac surgery. METHODS: Twenty patients with a pre-operative ejection fraction > 50% were included in an open prospective observational study comparing the changes in SvO2 and hemodynamics during mobilizations immediately prior to surgery and on the first post-operative morning. RESULTS: Mobilization induced an absolute reduction in SvO2 of 17.7 +/- 7.4% pre- and 19.0 +/- 5.5% post-operatively (NS). ANOVA for a series of measurements throughout the mobilization sequence identified no different effect on SvO2 between pre- and post-operative mobilizations (P = 0.567). The SvO2 level was reduced post-operatively resulting in a SvO2 during standing exercise of 55% before and 49% after the surgery (P < 0.01). Mobilization increased the heart rate (HR) and decreased the stroke volume index (SVI), leaving CI unchanged. This response was similar pre- and post-operatively (NS). Compared with pre-operative measurements, CI and HR increased post-operatively while SVI remained unchanged despite elevated cardiac filling pressures and reduced systemic vascular resistance. The left ventricular stroke work index was reduced, indicating reduced myocardial performance. CONCLUSION: Myocardial function was reduced on the first morning after coronary artery bypass grafting (CABG), but during post-operative mobilization this reduction did not significantly influence the changes in CI or SvO2.     

Mixed venous oxygen saturation as a predictor of cardiac output in the postoperative cardiac surgical patient

Mixed venous oxygen saturation (SvO2) was measured continuously with a fiberoptic pulmonary artery catheter in 25 patients during the first 24 hours after cardiac surgery and was compared with the thermodilution cardiac index (CI). The mean correlation coefficient between SvO2 and CI was 0.05 +/- 0.42, and was not significantly different from zero. Although the mean correlation coefficient between the change in SvO2 and the change in CI was significant (p less than .05), the magnitude of the coefficient (0.19 +/- 0.44) indicates poor predictive value. The correlation did not improve when adjusted for multiple clinical variables, and the SvO2 was not predictive of a CI less than 2 L/min/m2, a level of cardiac performance that might require intervention. In conclusion, SvO2 was not predictive of CI postoperatively in the cardiac surgical patient.     

Cardiac output monitoring during off-pump coronary artery bypass grafting

OBJECTIVE: To evaluate and compare monitors of cardiac output during repositioning and stabilization of the heart for off-pump coronary artery bypass (OPCAB) surgery. DESIGN: Prospective, observational, clinical study. SETTING: University teaching hospital. PARTICIPANTS: Consecutive patients scheduled to undergo elective OPCAB (n = 19). INTERVENTIONS: Monitoring, induction, and anesthesia followed a routine protocol for coronary artery bypass patients. This included the use of transesophageal echocardiography (TEE) and pulmonary artery catheter placement. MEASUREMENTS AND MAIN RESULTS: After positioning and stabilization for OPCAB surgery, the changes in descending aortic flow velocity (VTI) times heart rate (HR) and the mixed venous oxygen saturation (SvO(2)) could be used to predict the changes in thermodilution cardiac output (TDCO) using the following model: deltaTDCO((calc))=-13.15+0.35(deltaVTI*HR)+0.61(deltaSvO(2)) where Delta indicates the percentage change from baseline values. The changes in mean arterial pressure, mean pulmonary artery pressure, and continuous cardiac output did not correlate with the changes in TDCO. CONCLUSION: The use of the VTI*HR, as determined by TEE, in addition to the SvO(2) can strengthen clinical decision making during repositioning and stabilization of the heart during OPCAB. Changes in the VTI*HR and SvO(2) can be used as surrogate markers for changes in CO during OPCAB surgery.     

Continuous monitoring of mixed venous oxygen saturation in anesthesia in pulmonary surgery

The multiplicity of potential causes of variations in mixed venous oxygen saturation (SvO2) during one lung ventilation (OLV), including a constant ventilation/perfusion mismatch, explains that it has been suggested as a routine monitoring procedure. To assess its usefulness, 12 adults undergoing OLV were monitored during surgery with an Oximetrix pulmonary catheter, placed on the side opposite to the surgical field under fluoroscopic control. Seventy two complete sets of haemodynamic measurements were obtained at 6 different times during surgery. We studied the ability of changes in SvO2 to predict changes in arterial oxygen saturation (SaO2), cardiac output (CO), and venous admixture (VA) by calculating sensitivities (Se), specificities (Sp) and predictive values with regard to these variables. There were no complications due to the protocol. However left-sided catheter placement failed in four cases. Correlation between optical and measured SvO2 was very strong (r = 0.94; p less than 0.001). SvO2, oxygen consumption (VO2) and the rate of oxygen extraction remained constant throughout the procedure, even when CO, mean arterial pressure, VA, SaO2 and PaO2 varied. Clamping the pulmonary artery returned VA, SaO2 and PaO2 values to those found before OLV, but produced a significant decrease in CO. SvO2 had low Se and Sp for changes in other variables (CO: 76 +/- 7, 48 +/- 9; PaO2: 79 +/- 6, 59 +/- 9; VA: 54 +/- 7, 48 +/- 7 respectively). In this type of surgery, alterations in variables related to oxygen are probably balanced by haemodynamic changes. In fact, according to Fick's formula, SvO2 is almost completely determined by SaO2 and CO, when VO2 and haemoglobin remain stable.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison between intermittent and continuous measurement of cardiac output after acute normovolemic hemodilution in pigs

Continuous cardiac output (CO) and mixed venous oxygen saturation (SvO2) determined through the pulmonary artery catheter may be helpful in monitoring hemodynamic conditions in critically ill patients. This study aimed to evaluate CO and SvO2 in a model of acute normovolemic hemodilution (ANH), analyzing the accuracy of the continuous versus intermittent method for CO and SvO2 measurement in pigs. Twenty-three pigs were enrolled to three groups: control, ANH with 6% hydroxyethyl starch (HES), or ANH with lactated Ringer's (LR) solution. After hemodilution, we showed that SvO2 was reduced in both groups, mainly in LR animals (P < 0.05). Regarding the evaluation of CO, we showed an increase in both groups submitted to ANH (P < 0.05). Through Bland-Altman analysis, we showed that the continuous CO catheter presented lower values than the intermittent method after hemodilution, mainly with HES (P < 0.001), and there was no difference in the measurement of SvO2. The ANH promoted a decrease in SvO2 and an increase in CO values, mainly in animals submitted to hemodilution with HES. The use of continuous and intermittent (laboratory) measurement of SvO2 showed clinical applicability and good agreement, an effect not reproduced by the CO measurement. New studies are needed to further investigate the agreement between the continuous and intermittent methods for the measurement of CO in adverse hemodynamic conditions such as ANH.     

Continuous In-Line Monitoring of Oxygen Delivery to Control Artificial Heart Output

In order to evaluate the pump output control based on the oxygen delivery to peripheral tissues, arterial and mixed venous hemoglobin content ([Hb]) and oxygen saturation (SO2) were continuously monitored in three biventricular bypass animals (3-, 6-, and 40-day experiments) with fibrillating ventricles. The specially developed oxygen sensors were mounted in the outflow ports of the artificial hearts to measure [Hb] and SO2. One animal was exercised on the treadmill at 2.0 mile/h for 15 min with pump flows fixed to deliver oxygen of (a) above 13 cc/min/kg, (b) 10, and (c) 9. In (a), the mixed venous saturation (SvO2) dropped to approximately 25% with no increase in the blood lactate level. In (b) and (c), the SvO2 decreased to approximately 10-15% with increase in blood lactate levels from 4 to 10-30 mg/dl. Also, the recovery of the SvO2 in these groups following the termination of the exercise was slower in comparison to (a). The lower limit of the SvO2 level that would create oxygen debt situation in the peripheral tissues was approximately 25-30% for the exercise of 2.0 mile/h. The SvO2 reflects changes in respiratory status, pump output, hemoglobin level, and metabolism, and is thus a useful indicator to diagnose quickly the circulatory status as well as possibly to control the artificial heart output.     

Blood gas analysis of mixed venous blood during normoxic acute isovolemic hemodilution in pigs

Mixed venous oxygen saturation of hemoglobin (SvO2) and mixed venous oxygen tension (PvO2) may reflect the overall balance between oxygen consumption and delivery. Because of the potential value of monitoring SvO2 and PvO2 as indications of the state of tissue oxygenation, the aim of this study was to determine, during normoxic acute isovolemic hemodilution in pigs, the critical PvO2, critical SvO2, and critical oxygen extraction ratio (ER) at which oxygen uptake starts to decline during further induced hemodilution. During stepwise induced isovolemic hemodilution, a gradual decline in SvO2 and PvO2 was observed in all animals. The mean +/- SD of the critical PvO2 of six animals was 32.3 +/- 3.1 mm Hg. The mean +/- SD of the critical SvO2 was 44.2% +/- 7.9%. The ER increased gradually. At an ER of 0.57 +/- 0.08, oxygen uptake started to decline. A significant correlation was found between changes in SvO2 and changes in ER. These degrees of hemodilution were accompanied by an increase in cardiac index, pulmonary wedge pressure, heart rate, and left ventricular stroke work index. Only a slight decrease in systemic vascular resistance was observed. We conclude that measurements of PvO2 and SvO2 can be used as indicators of the critical point of hemodilution and that the SvO2 during hemodilution reflects the overall balance between oxygen uptake and oxygen delivery, confirmed by the strong correlation found between SvO2 and oxygen extraction ratio.     

Goal-directed therapy in anesthesia: any clinical impact or just a fashion?

Goal-directed therapy (GDT) describes the protocolized use of cardiac output and related parameters as end-points for fluid and/or inotropic therapy administration. Identifying the patient who will benefit from it has implications throughout perioperative management. The fundamental principle behind GDT is optimizing tissue perfusion by manipulating heart rate, stroke volume, hemoglobin and arterial oxygen saturation to improve oxygen delivery by using fluids, inotropes, red blood cells and supplementary oxygen. Although cardiac output and SvO2 were previously measured using the pulmonary artery catheter, a number of less invasive methods are now available. For intraoperative GDT, the esophageal Doppler-derived Flow Time correct (FTc) is the parameter used most frequently, although other parameters such as stroke volume obtained from Vigileo, PICCO and/or LiDCO, mixed and/or central venous oxygen saturation (SvO2/ScvO2), oxygen delivery and global end diastolic volume (PiCCO system) may be applied in daily clinical practice. The correct target to be followed during the intraoperative period must be clearly established. Most parameters depend primarily on O2 consumption and are not reliable or useful during anesthesia. To date, the quantity and the type of fluids to administer during major elective surgery remain an object of continuing debate. In conclusion, in terms of evidence-based medicine, GDT during anesthesia has a clinical impact when performed using an FTc-based fluids algorithm protocol. In contrast, GDT can be considered unreliable if confusing targets such as SvO2 or ScvO2 higher than 70% during anesthesia are followed.     

Profound hypoxemia during treatment of low cardiac output after cardiopulmonary bypass

PURPOSE: To illustrate the multiple causes of hypoxemia to be considered following cardiopulmonary bypass and how therapy given to improve oxygen delivery may have contributed to a decrease in arterial oxygen saturation to life-threatening levels. CLINICAL FEATURES: A 61 yr old man with severe mitral regurgitation and chronic obstructive lung disease underwent surgery for mitral valve repair. A pulmonary artery catheter with the capacity to measure cardiac output and mixed venous oxygen saturation (SvO2) continuously was used. Two unsuccessful attempts were made to repair the valve which was finally replaced, requiring cardiopulmonary bypass of 317 min. Dobutamine 5 micrograms.kg-1.min-1 and sodium nitroprusside 1 microgram.kg-1.min-1 were used to increase cardiac output. Soon after, the SvO2 decreased progressively from 55 to 39%. The patient became cyanotic with a PaO2 of 39 mmHg. Sodium nitroprusside was stopped and amrinone 100 mg bolus followed by 10 micrograms.kg-1.min-1 was given in addition to adding PEEP to the ventilation. With these measures PaO2 could be maintained of safe levels but PEEP and high inspired oxygen concentrations were needed postoperatively until the trachea could be extubated on the third postoperative day. CONCLUSION: The profound hypoxemia in this case was likely due to a combination of intra- and extrapulmonary shunt, both augmented by sodium nitroprusside. The desaturation of mixed venous blood amplified the effect of these shunts in decreasing arterial oxygen saturation. The interaction of these factors are analyzed in this report.     

Continuous monitoring of mixed venous oxygen saturation during aortofemoral bypass grafting

Measurement of mixed venous oxygen saturation (SvO2) may be helpful in the care of critically ill patients. Serial determinations of SvO2 give an index of the relationship between oxygen delivery and tissue oxygen consumption. Continuous monitoring of SvO2 is now readily available with the Shaw Oximetrix pulmonary artery catheter (Oximetrix Inc., Mountain View, CA). This system has provided useful information in the high risk cardiac surgery patient. Continuous monitoring of mixed venous saturation may be helpful in high risk or critically ill general and peripheral vascular surgery patients both in the intensive care unit and in the operating room. The following clinical report is presented to illustrate the usefulness of continuous SvO2 monitoring in a high risk vascular surgery patient.     

Detection of venous air embolism by continuous mixed venous oximetry in dogs

Continuous mixed venous oxygen saturation (SvO 2) was evaluated as a monitor of venous air embolism in a canine model. Nineteen dogs were anesthetized, paralyzed, and mechanically ventilated. Invasive monitoring included SvO 2, systemic and pulmonary artery blood pressures, and thermodilution cardiac outputs. Air boluses of 0.25 and 0.5 ml/kg were injected in six dogs and 1 ml/kg in all. All 1 ml/kg emboli were detected by greater than or equal to 5% decreases in the SvO 2. The SvO 2 decreased from 82 +/- 8% to 72 +/- 11% (mean +/- SD), an average decrease of 9 +/- 5% (p = 0.004). Time to the SvO 2 nadir was 2.6 +/- 2.5 min. Of the 0.5 and 0.25 ml/kg emboli, 50% and 17% were detected, respectively. Cardiac output decreased from 2.9 +/- 0.8 to 2.1 +/- 0.8 L/min after the 1 ml/kg emboli (p = 0.02). The 1 ml/kg emboli increased pulmonary artery pressures and decreased systemic blood pressure in 100% and 75% of animals, respectively. Peak changes in pulmonary artery pressure occurred at 1.2 +/- 0.8 min. In the present study, time to maximum change was greater for SvO 2 than for pulmonary artery pressure changes. Use of fiberoptic pulmonary artery catheters for continuous measurement of SvO 2 can add a new diagnostic modality to venous air embolism detection in patients who require a pulmonary artery catheter for other medical indications.     

Factors influencing mixed venous oxygen saturation in intensive care

Changes in mixed venous blood oxygen saturation (SvO2) were studied in 2 groups of patients. Group I patients (n = 10) were all hypoxaemic, suffering from acute respiratory failure, requiring that FIO2 be maintained at 1 throughout the study; respiratory and haemodynamic conditions were improved using PEEP and cardiovascular support. On the other hand, Group II patients (n = 13) were non-hypoxaemic patients with circulatory shock in whom FIO2 was gradually increased, and the haemodynamic status was improved using positive inotropic drugs (dopamine, dobutamine, adrenaline, amrinone). All 23 patients had a Swan-Ganz catheter set up for monitoring; all the usual haemodynamic and respiratory parameters were measured. Haematocrit values were kept at the same level throughout the study. Haemodynamic parameters were measured each time a new therapeutic procedure was carried out. No close relationship between SvO2 changes and changes in cardiac index or O2 consumption were found. However, a close relationship existed between changes in SvO2 and changes in O2 extraction (EAO2): SvO2 = -EAO2 + 102 (Group I; r = 0.90, n = 54); SvO2 = -1.2 EAO2 + 103 (Group II; r = 0.93, n = 66). A strong relationship was also found between changes in SvO2 and in FIO2 in each patient of Group II. In the complicated physiological set-up of an intensive care patient, SvO2 reflects oxygen extraction. A fall in SvO2 is related to an altered oxygen demand: oxygen supply ratio. In the most seriously ill patients, there is no relationship between changes in SvO2 and cardiac index.