BJA Advance Access originally published online on November 22, 2006
British Journal of Anaesthesia 2007 98(1):45-52; doi:10.1093/bja/ael310
Non-invasive metabolic monitoring of patients under anaesthesia by continuous indirect calorimetry—an in vivo trial of a new method
1 Department of Electrical and Computer Systems Engineering, Faculty of Engineering, Monash University Victoria, Australia
2 Department of Anaesthesia, The Austin Hospital Heidelberg, Victoria 3084, Australia
3 Department of Anaesthesia and Perioperative Medicine The Alfred, Melbourne, Victoria 3004, Australia
*Corresponding author: Department of Electrical & Computer System Engineering, Clayton Campus, Building 72, Monash University, VIC 3800, Australia. E-mail: Christopher.Stuart-Andrews{at}eng.monash.edu.au
Accepted for publication September 28, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Background. Oxygen uptake is an important form of metabolic monitoring for patients under anaesthesia. In critically ill patients oxygen uptake has been shown to provide valuable clinical information in directed therapy and acts as a useful monitor of cardiovascular dysfunction. A new method of continuous real time monitoring of metabolic gas exchange was tested in patients during anaesthesia.
Methods. Using a standard anaesthetic machine with attached semi-closed circle absorber system, oxygen uptake was measured continuously throughout surgery in 30 patients undergoing cardiopulmonary bypass surgery and compared with paired measurements made with the reverse Fick method. The method is an indirect calorimetry technique which uses fresh gas rotameters for control, regulation and measurement of the gas flows into the system, with continuous sampling of mixed exhaust gas.
Results. When compared with the reverse Fick method the oxygen uptake showed a mean difference (and SD) of 20.7 ml min–1 or 12.1% (25.3 ml min–1) pre-bypass and 13.9 ml min–1 or 8.1% (27.0 ml min–1) post-bypass. This bias is consistent with previous studies comparing oxygen uptake measured at the mouth against oxygen uptake by reverse Fick, which have shown a difference of approximately 10–15% accounted for by the consumption of oxygen by lung tissue.
Conclusions. As the method allows continuous measurement of gas exchange and can be adapted to a modern anaesthetic workstation it is an attractive method for use in clinical setting.
Keywords: measurement techniques, Fick principle; monitoring, oxygen; metabolism, oxygen consumption
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Measurement of oxygen uptake (
) is an important form of metabolic monitoring for patients during anaesthesia and critical care.1–4 It provides an indicator of a patient's metabolic status and oxygen delivery to the tissues, and of cardio-respiratory function. As with other vital physiological parameters, such as systemic blood pressure, continuous automated measurement of permits early intervention by the physician when deterioration unexpectedly occurs.
A simple, continuous and non-invasive method of measurement of would assist anaesthetists in monitoring patient stability intra-operatively. However, the measurement of has traditionally been uncommon in the operating room. This is largely because of the difficulties of accurate and reproducible non-invasive measurement of gas exchange, particularly at a high 
and in the presence of inhalational anaesthetic agents. Invasive techniques, such as the reverse Fick method, using a pulmonary artery catheter to measure mixed venous oxygen content and cardiac output, are rarely felt justified. For these reasons, measurement of is not considered a part of common clinical anaesthetic practice.
Biro5 proposed a method for continuous measurement in a breathing system, which was based on a formula by Foldes and colleagues.6 The Biro equation, which utilized continuous measurement of inspired O2 concentration to calculate , was tested in a clinical study against the reverse Fick method by Leonard and colleagues who found poor agreement, with the Biro method underestimating the reverse Fick by 40–50%.2 This error was all the more marked in view of the expected effect of lung tissue uptake of O2, which should cause a bias in the opposite direction. This finding was subsequently confirmed using a lung gas exchange simulator by Stuart-Andrews and colleagues7 who demonstrated a large mean bias in the calculated relative to a precisely simulated value.
We describe an approach which uses the fresh gas rotameters for control, regulation and measurement of the gas flows into a standard semi-closed breathing circuit, with continuous sampling of mixed exhaust gas instead. This method permits continuous real time monitoring of . The system has been previously tested in vitro using a lung gas exchange simulator,8 and shown to have excellent accuracy and precision, across a wide range of simulated values for , , fresh gas flows and ventilatory settings. In this study, the accuracy and precision of the method was tested under clinical conditions against the reverse Fick method, in anaesthetized patients.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Theory
The method presented here is based on mass balance principles for O2 and CO2 within a breathing system attached to a patient breathing an air/O2 mixture. For any given setting for fresh gas O2 rotameter flow rate (
) and air rotameter flow rate (
), 
is calculated from the following equation:
 | (1) |
and 
are mixed exhaust gas concentrations of O2 and CO2 respectively.
Equation (1) allows continuous monitoring of from a measurement of mixed exhaust gas oxygen and CO2 concentrations. We have termed this continuous indirect calorimetry. The derivation of the equation is given in the Appendix, but it should be noted that the presence of other gases in the inspired mixture is readily incorporated and their uptake can also be measured simultaneously. This includes CO2 elimination although accurate measurement of 
is prevented in a circle absorber system by its uptake by soda lime.
Experimental protocol
Thirty patients undergoing cardiopulmonary bypass graft surgery at The Alfred hospital, Melbourne, Australia were recruited to the study. Ethical approval was obtained from the institution's human research ethics committee and informed written consent was obtained from each patient at the time of surgical admission.
Before surgery patients were cannulated in accordance with routine anaesthetic management with a peripheral arterial line and pulmonary artery catheter (Edwards, Irvine, CA, USA). After a period of 1–2 min of pre-oxygenation, anaesthesia was induced with a mixture of fentanyl, a benzodiazepine, propofol and a neuromuscular blocker. Maintenance of anaesthesia was achieved using an infusion of propofol titrated according to the depth of anaesthesia with the assistance of bispectral index monitoring. After endotracheal intubation, controlled ventilation was initiated using a 7900 series ventilator (GE Healthcare, Helsinki, Finland) with tidal volumes of approximately 7–10 ml kg–1 at a rate of 9–12 bpm using a standard circle absorber breathing system. The fresh gas mixture was set to 5 litre min–1 (3 litre min–1 air and 2 litre min–1 O2) giving a fresh gas O2 concentration of slightly more than 50%, although this was able to be increased by the anaesthetist if felt clinically indicated.
During surgery, simultaneous paired blood samples were drawn from the arterial line and the distal lumen of the pulmonary artery catheter. Blood samples were analysed immediately at point of contact for oxygen saturation, partial pressure and haemoglobin content on the operating suite's blood gas analyser (Rapidlab 1265, Bayer Diagnostics, Sudbury, UK). During the measurement period a set of five cardiac output measurements were made by thermodilution using a 10 ml bolus of room temperature saline and the results averaged. Results were excluded if they were found to lie more than ±10% outside of the mean value. Oxygen content of both the arterial and mixed venous samples were obtained by calculation using Equation (2) below. Values for saturation and partial pressure made by the blood gas analyser were corrected to 37°C before calculation of oxygen content was made
 | (2) |
is the measured percentage saturation of oxygen and 
is the partial pressure of oxygen (mm Hg).
Using the cardiac output measurement in conjunction with the calculated arteriovenous oxygen content difference, oxygen uptake (
) was determined by the reverse Fick method.
 | (3) |
is the calculated arterial oxygen content of the sample, 
is the mixed venous oxygen content and 
is the cardiac output measured by thermodilution. Heart rate, mean arterial pressure, pulmonary capillary wedge pressure, end-tidal CO2 concentration, room and patient temperature measured by nasopharyngeal probe were also noted at this time.
During this period measurements of oxygen uptake by continuous indirect calorimetry (
) were averaged over a 5 min period and corrected to STPD (standard temperature and pressure dry). This process was repeated at two discrete points during the operation, once pre-bypass and once post-bypass. The first measurement was made approximately 30 min post-sternotomy while the final measurement was made post-bypass immediately after sternal closure.
Measurement system
A standard anaesthetic delivery system was used consisting of an Excel 210SE anaesthetic machine (GE Healthcare, Helsinki, Finland) and semi-closed circle absorber system (Fig. 1). Fresh gas air was obtained from an ‘E’ sized cylinder of medical air attached to the anaesthetic machine, instead of using the hospital's wall supply, as this was found to vary significantly in pressure over a period of time, causing variation in fresh gas air flow rates. Cylinder air supply pressure however was monitored throughout the experimentation period and a correction for any change in supply pressure was made against a previously calibrated flow–pressure curve (obtained from the elapsed time taken to fill a 1 litre dry gas syringe). The observed variance of the bottle supply pressure was 0.5 kPa, giving very stable flow measurement. The oxygen supply was connected directly into the wall. There was only a small variation in oxygen supply pressure but in any case this change did not result in a change in oxygen flow across the rotameters as oxygen supply pressure to the anaesthetic machine is regulated in a two-stage process resulting in a constant oxygen flow.
|
Values for fresh gas flow were set visually on the flow control valves and the read value entered on the computer. The rotameters used were standard design bobbin rotameters incorporated into the anaesthesia workstation and had been initially calibrated by the manufacturer for individual gas species at 20°C and 101.3 kPa. To ensure their linearity over a range of delivered flows we performed a further calibration using a 1 litre dry gas syringe. From this a correction (<1%) was applied to this value in software using the calculated calibration curve.
Gas concentrations were measured by side-stream sampling by a Capnomac Ultima rapid gas analyser (GE Healthcare, Helsinki, Finland). Analogue data from the analyser was downloaded to a desktop computer via an analogue to a digital converter card. The rapid gas analyser measured O2 concentration paramagnetically, with an observed SD under steady-state conditions of 0.1%. CO2 was measured by infrared spectroscopy, and N2 concentration calculated by subtraction of all other gases from 100%. Analogue data were downloaded from the analyser and pressure transducers to the computer every 100 ms via the analogue-digital converter card (12-bit Burr-Brown, AZ, USA). This sampling rate provided more than adequate precision of measurement of raw data, given that stable concentrations of mixed gases were being sampled. Computations were made in real time using Borland C++ operating on a personal desktop computer. Gas concentration samples were averaged and reported updates every 15 s.
Periodic (half hourly) recalibration of the system was performed by brief sampling of fresh gas, as follows. The fresh gas O2 concentration calculated from the rotameter flow of air and oxygen was compared with an average value measured by the analyser over a period of 15 s at the beginning of the measurement process and again after each automatic recalibration (or ‘zero’) of the analyser which occurred every half hour. The measured difference (
FO2) was used to correct all subsequent 
measurements until the next recalibration. This effectively calibrated the O2 concentrations measured by the analyser against the rotameter settings. The gas analyser continuously sampled exhaust gas at all other times, between recalibrations.
This system had been previously validated against a laboratory benchtop lung simulator.8 The system was checked for leaks using a static pneumatic compression manoeuvre, and found to lose not more than 10 ml min–1 under average operating circuit pressures. Before each operation a dynamic system calibration was performed by ventilation of a pair of silicone bags of suitable compliance in place of the patient in the circuit in order to determine the value of any significant zero offset for in the system. This was generally found to be approximately –10 ml min–1. This value was used to correct all measurements made during the operation.
The measurement of agreement between simultaneous paired measurements of measured by continuous indirect calorimetry using Equation (1) () and was done by the method of Bland and Altman,9 calculating mean difference (bias being –) and the SD of the difference and limits of agreement [bias ± 2 SD]. The Pearson correlation coefficient r and its statistical significance were also determined.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Data were collected from all 30 patients whose average age was 69 yr. Twenty-two patients were undergoing coronary artery bypass graft surgery, four were undergoing a valve replacement whilst the remaining four patients were undergoing both bypass graft surgery and valve replacement. Further patient characteristics and type of surgery are shown in Table 1. Perioperative haemodynamic data showed that cardiac output was higher post-bypass, accompanied by a lower haemoglobin concentration attributable to haemodilution during cardiopulmonary bypass (Table 2). Mean
pre-bypass was 139 ml min–1 and rose to 160 ml min–1 post-bypass (P=0.02 on an unpaired t-test) (Table 3). These values are quite typical of elderly anaesthetized patients, who were all under mild to moderate hypothermia, as is typical anaesthetic practice in the pre-bypass period. Despite the higher Fick, O2 delivery was lower, because of the anaemia post-bypass.
|
|
|
Comparison of
against 
from Equation (1) are shown for both pre- and post-bypass results in Table 3. Results are reported as mean value {SD [median (inter-quartile range)]}. Overall the mean bias was found to be 17.2 ml min–1 (10.5% of the mean), with a SE of 2.4 ml min–1 and a SD of 26.1 ml min–1. This gives a 95% confidence limit for the mean bias of ±4.9 ml min–1. Figure 2 shows a Bland and Altman plot of these data, which distinguishes pre- and post-bypass measurements. The mean bias (and SD) pre-bypass was 20.7 ml min–1 (25.3 ml min–1) while post-bypass it was 13.9 ml min–1 (27.0 ml min–1). Correlation between the two methods was good (r=0.72, P<0.01 overall) given the relatively narrow range of oxygen uptake among the patient population.10
|
The continuous nature of the method is demonstrated by data collected from one patient in the study. Figure 3 shows continuous measurement of
in a 71-yr-old female. Her ECG was noted to change from a normal sinus rhythm at 70 min–1 to a rapid atrial fibrillation (AF) at approximately 120 min–1, accompanied by a decrease in systemic arterial pressure before the administration of intravenous metaraminol. The patient was hastened onto cardiopulmonary bypass, the rapid AF continuing until cardioplegia was delivered. Figure 3 shows a sudden and sustained increase in (accompanied by a proportional increase in exhaust gas CO2 concentration indicating a change in as well) at this time.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Non-invasive measurement of
can be made in a variety of different ways;11–15 in a fully closed breathing system, a patient's oxygen uptake can be determined from the flow rate of oxygen required to maintain a constant volume, pressure and oxygen concentration inside the breathing circuit, but such approaches are not practical for routine use.16 Measurement of gas exchange inside a semi-closed breathing circuit has generally been performed by measurement of gas concentration and flow both into and out of the circuit. The change in total flow between these two points can be determined from the measured or calculated concentrations of an insoluble marker gas such as nitrogen (N2) in fresh gas and mixed exhaust gas (the Haldane transformation) when the total inflow rate of N2 is known.17 At these points, thorough mixing will ensure stable gas concentrations in the face of tidal fluctuations in concentration at the mouthpiece, which improves accuracy and precision of the measurement. Indirect calorimetric methods such as this can be automated to provide ongoing measurement,18 but need repeated switching of gas sampling to achieve this, which limits the frequency of measurement and precludes truly continuous gas exchange monitoring.
The continuous method tested here is an adaptation of the Haldane transformation and requires the presence of some N2 in the exhaust gas flow in order to compute Equation (1). The results of this study demonstrate that the accuracy and precision of the Haldane approach of measuring gas exchange are retained, while permitting continuous monitoring to occur. Measurement of gas uptake from the breathing circuit after maximal gas mixing improves measurement stability. Furthermore, the continuous sampling allowed data averaging and smoothing to be done while maintaining the frequency of CIC calculation at four times per minute.
The ‘reverse Fick’ method is a commonly used physiological standard against which calorimetric methods have been compared but it has its limitations.1 While measurement of cardiac output by thermodilution is traditionally assumed to be accurate to within ±10% of the true value,19 poorer agreement than this has been demonstrated against gold standards such as indwelling flow probes in animals.20 Overestimation of low cardiac output values in humans has also been shown.21–24 During a complete ‘reverse Fick’ calculation, propagation of random measurement errors throughout the calculation process has been shown previously.1 Conservative estimates of error in measured input variables produce errors of almost 20% in calculated . This has prompted some commentators to suggest that indirect calorimetry is a more accurate technique.25 This is interesting, as there are relatively few examples in clinical measurement where a less invasive approach proves to be more reliable than its invasive alternative.
These concerns about reproducibility and precision of individual measurements with the reverse Fick method may explain much of the random scatter in agreement with continuous calorimetry we encountered. Similar levels of scatter in agreement between indirect calorimetry and the reverse Fick method were found by Walsh and colleagues26 and Marson and colleagues3 in intensive care patients. Previous laboratory validation of our measurement system using a lung gas uptake simulator showed only a –1.3 ml min–1 mean bias with a SD of 6.5 ml min–1 in agreement with the target simulated .8
Nevertheless, despite this, the reverse Fick method does provide a useful standard to assess the presence of bias in a comparator method, provided a large enough series of data is collected to reduce the standard error of the mean, and allowance is made for the presence of lung tissue O2 uptake, which is not measured by the reverse Fick approach. Several previous studies have demonstrated this to be in the order of 10–15% of measured at the mouth, in patients undergoing cardiac surgery.2 3 27–29 The mean bias of 10.5% found in this study is consistent with this. Because its calculation shares no input variables with Equation (1), the reverse Fick method avoids the potential for artifactual correlation with calorimetry because of mathematical coupling.1
The method is designed to provide continuous monitoring, which has been demonstrated in a previous in vitro study.8 This study was designed to validate its accuracy under clinical conditions against an independent standard. However, the continuous nature of the method is demonstrated by data in Figure 3, collected from one patient in the study during the onset of rapid AF. The likely acute reduction in cardiac output suggested by the decrease in blood pressure accompanying the AF would be expected to cause a transient decrease in . Therefore, the sudden and sustained increase in we found was an interesting observation; clinical and laboratory studies have shown that AF reduces myocardial efficiency while increasing myocardial O2 consumption.30 31 This anecdote demonstrates that routine measurement of metabolic gas exchange may provide deeper insights into the physiological changes occurring in patients during surgery than are currently available to us.
While providing continuous monitoring of , the method tested here has the disadvantage of a delayed response time in its measurement with recirculating systems such as circle absorber systems, which have relatively high circuit volumes. This is attributable to the wash through time required for a step change in a patient's to be measured as a change in exhaust gas concentration at the distal end of the circuit. This is also seen with changes to the fresh gas concentration or flow rates, which require time for washthrough. We minimized this delay by using a relatively high fresh gas flow rate, having previously demonstrated in vitro that our system would still retain sufficient precision of measurement at higher flow rates, despite the consequent reduction in the fresh gas to exhaust gas O2 concentration gradient.8 This delay is minimal with partial rebreathing systems, such as the Mapleson D (Bain circuit), or with non-rebreathing systems, such as are used with the Manley ventilator. The disadvantage of these high flow systems is poorer economy of gas usage. In addition we found that fresh gas oxygen concentrations up to 80% still provided sufficient accuracy in measurements.
The advantage of the method lies in its adaptability. We have demonstrated the accuracy and practicability of the method by setting and reading fresh gas bobbin flowmeters manually. However, better exploitation of the technique might use electronic flowmeters, such as are found as standard equipment on modern anaesthetic delivery workstations. This avoids the need to manually enter the selected fresh gas flows into software, as was necessary on our prototype measurement system.
The purpose of the method described here, to provide continuous monitoring, is shared with the earlier study of Biro. There are obvious similarities between Equation (1) and the equation proposed by Biro,5 which also used nominated fresh gas flow settings, but used measurement of inspired O2 concentration. His equation was derived from an early model of an anaesthetic breathing system presented by Foldes.6 This simplistic model contained erroneous assumptions as to the mass balance within a breathing system, ignoring the loss of gas to exhaust in a semi-closed breathing system, and did not take into account the uptake or elimination by the patient of other gases in the gas mixture.7 8 For these reasons Biro's equation has been demonstrated by previous studies2 7 to be inaccurate.
Our results demonstrate that where O2 concentration is more correctly monitored in mixed exhaust gas, accurate monitoring is achievable with minimal change to existing anaesthetic equipment design. This may allow measurement of to become a modality which clinicians could expect to see incorporated into routine anaesthetic monitoring in the future.
|
Appendix |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
Derivation of Equation (1).
Based on mass balance principles, in a patient receiving an O2–air mixture, attached to a breathing system with total fresh gas flow 
, the total flow of mixed exhaust gas at the point of gas concentration sampling (
) is
 | (A1) |
, the flow of CO2 in mixed exhaust gas, is the same as if there is no soda lime present and if there are no leaks in the system.
If 
is the fractional mixed exhaust concentration of CO2
 | (A2) |
Substituting in (A1) and transposing
 | (A3) |
Now, if 
is the total fresh gas flow of O2 (from both air and O2 rotameters) and is its mixed exhaust concentration
 | (A4) |
Substituting from (3)
 | (A5) |
 | (A6) |
Given that the fresh gas air rotameter delivers 20.93% O2, and that 
is the sum of the fresh gas O2 rotameter flow rate () and the air rotameter flow rate (
)
 | (A7) |
|
Acknowledgments |
|---|
This project was supported financially by a research grant provided by GE Healthcare and the Australian Society of Anaesthetists.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixReferences |
|---|
1 Walsh T. Recent advances in gas exchange measurement in intensive care patients. Br J Anaesth 2003; 91:120–31
2 Leonard IE, Weitkamp B, Jones K, Aittomaki J, Myles PS. Measurement of systemic oxygen uptake during low-flow anaesthesia with a standard technique vs. a novel method. Anaesthesia 2002; 57:654–8[CrossRef][Web of Science][Medline]
3 Marson F, Martins MA, Coletto FA, Campos AD, Basile-Filho A. Correlation between oxygen consumption calculated using Fick's method and measured with indirect calorimetry in critically ill patients. Arq Bras Cardiol 2003; 82:77–81
4 Durham R, Neunaber K, Mazuski J, Shapiro M, Baue A. The use of oxygen consumption and delivery as endpoints for resuscitation in critically ill patients. J Trauma 1996; 41:32–9[Web of Science][Medline]
5 Biro P. A formula to calculate oxygen uptake during low flow anesthesia based on FIO2 measurement. J Clin Monit Comput 1998; 14:141–4[CrossRef][Web of Science][Medline]
6 Foldes FF, Ceravolo AJ, Carpenter SL. The administration of nitrous oxide-oxygen anesthesia in closed systems. Ann Surg 1952; 136:978–81[Web of Science][Medline]
7 Stuart-Andrews C, Peyton P, Robinson GJB, Lithgow B. Accuracy of the Foldes-Biro equation for measurement of oxygen uptake during anaesthesia: a laboratory simulation. Anaesthesia 2004; 59:541–4[CrossRef][Web of Science][Medline]
8 Stuart-Andrews C, Peyton P, Robinson G, Lithgow B. Continuous indirect calorimetry—a laboratory simulation of a new method for non-invasive metabolic monitoring of patients under anaesthesia or in critical care. Physiol Meas 2006; 27:155–64[CrossRef][Web of Science][Medline]
9 Bland J and Altman D. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307–10[CrossRef][Web of Science][Medline]
10 Guyatt G, Walter S, Shannon H, Cook D, Jaenschke R, Heddle N. Basic statistics for clinicians: 4. Correlation and regression. CMAJ 1995; 152:497–504[Abstract]
11 Makita K, Nunn JF, Royston B. Evaluation of metabolic measuring instruments for use in critically ill patients. Crit Care Med 1990; 18:638–44[Web of Science][Medline]
12 Behrendt W, Weiland C, Kalff J, Giani G. Continuous measurement of oxygen uptake. Evaluation of the Engstrom metabolic computer and clinical experiences. Acta Anaesthesiol Scand 1987; 31:10–4[Web of Science][Medline]
13 Takala J, Keinanen O, Vaisanen P, Kari A. Measurement of gas exchange in intensive care: laboratory and clinical validation of a new device. Crit Care Med 1998; 17:1041–7
14 Peyton PJ, Robinson GJB, Terry D, Junor PA. A new method for measurement of gas exchange during anaesthesia using an extractable marker gas. Physiol Meas 2004; 25:1–14[CrossRef][Web of Science][Medline]
15 Robinson G, Peyton P, Terry D, Malekzadeh S, Thompson B. Continuous measurement of gas uptake and elimination in anesthetized patients using an extractable marker gas. J Appl Physiol 2004; 97:960–6
16 Barton F and Nunn JF. Totally closed circuit nitrous oxide/oxygen anaesthesia. Br J Anaesth 1975; 47:350–7
17 Lumb A. Nunn's Applied Respiratory Physiology2000. 5th edn. Oxford Butterworth-Heinemann pp. 299
18 Vartuli GM, Burfoot RB, Robinson GJB, Peyton PJ, Junor PA. Development and in vitro validation of a device for measuring non-shunt cardiac output by nitrous oxide throughflow. Med Biol Eng Comput 2002; 40:415–22[CrossRef][Web of Science][Medline]
19 Nishikawa T and Dohi S. Errors in the measurement of cardiac output by thermodilution. Can J Anaesth 1993; 40:142–53[Web of Science][Medline]
20 Bajorat J, Hofmockel R, Vagts DA, Pohl B, Beck C, Noeldge-Schomburg G. Comparison of invasive and less-invasive techniques of cardiac output measurement under different haemodynamic conditions in a pig model. Eur J Anaesth 2006; 23:23–30[CrossRef][Web of Science][Medline]
21 Doi M, Morita K, Ikeda K. Frequently repeated Fick cardiac output measurements during anesthesia. J Clin Monit 1990; 6:107–12[Web of Science][Medline]
22 Tournadre J, Chassard D, Muchada R. Overestimation of low cardiac output measured by thermodilution. Br J Anaesth 1997; 79:514–6
23 Van Grondelle A, Ditchey R, Groves B, Wagner W, Reeves J. Thermodilution method overestimates low cardiac output in humans. Am J Physiol 1983; 245:H690–2
24 Runciman WB, Ilsley AH, Roberts JG. Thermodilution cardiac output—a systematic error. Anaesth Intensive Care 1981; 9:135–9[Web of Science][Medline]
25 deBoisblanc B, McClarity E, Lord K. Oxygen consumption in the intensive care unit: indirect calorimetry is the way to go, but where? Crit Care Med 1998; 26:1153–4[CrossRef][Web of Science][Medline]
26 Walsh T, Hopton P, Lee A. A comparison between the Fick method and indirect calorimetry for determining oxygen consumption in patients with fulminant hepatic failure. Crit Care Med 1998; 26:1200–7[CrossRef][Web of Science][Medline]
27 Peyton PJ and Robinson GJB. Measured pulmonary oxygen consumption: difference between systemic oxygen uptake measured by the reverse Fick method and indirect calorimetry in cardiac surgery. Anaesthesia 2005; 60:146–50[CrossRef][Web of Science][Medline]
28 Keinanen O and Takala J. Calculated versus measured oxygen consumption during and after cardiac surgery. Is it possible to estimate lung oxygen consumption? Acta Anaesthesiol Scand 1997; 41:803–9[Web of Science][Medline]
29 Bizouarn P, Blanloeil Y, Pinaud M. Comparison between oxygen consumption calculated by Fick's principle using a continuous thermodilution technique and measured by indirect calorimetry. Br J Anaesth 1995; 75:719–23
30 Kumon K, Kuwabara M, Hirata T, et al. Continuous measurement of coronary sinus oxygen saturation after cardiac surgery. Crit Care Med 1987; 15:595–7[Web of Science][Medline]
31 Friedman HS, Scorza J, McGuinn R, Shaughnessy E. The effects of atrial fibrillation on myocardial blood flow and energetics. Proc Soc Exp Biol Med 1985; 180:1–8[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||