| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Deparment of Anesthesia & Pain Management, The Alfred Hospital, 2Department of Anesthesia, Austin Hospital, Heidelberg 3084 and 3Department of Allergy, Immunology, and Respiratory Medicine, The Alfred Hospital and Monash University, Melbourne, Australia 3181
Submitted 11 November 2003 ; accepted in final form 12 April 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
gas elimination; nitrogen balance; gas dilution
O2) and CO2 elimination (CO2), measurement of various markers of respiratory function such as diffusing capacity and shunt fraction (15), and measurement of pulmonary blood flow from CO2 or inert gas uptake (7, 12, 17, 21). Volumetric techniques employing a fully closed breathing system (2–4) have been commonly used for physiological studies, but these require careful manipulation for accurate measurement and are not suited for routine clinical use. Indirect calorimetry, which is also based on mass balance principles, has been found useful in the clinical setting. In indirect calorimetry, the flow rate into and out of a breathing system is estimated from the measured concentration of an insoluble marker gas, which is fed into the gas stream at a known flow rate (14). Nitrogen (N2) is often used for this purpose. With the assumption that it is neither taken up nor excreted by the lung, because it is relatively insoluble, exhaust gas flow can be related to fresh gas flow by the ratio of the their respective concentrations of N2 (the Haldane transformation). Continuous measurement of gas uptake and elimination is possible using this method, which has been widely used in a variety of clinical settings. A number of studies, however, have shown that zero N2 balance cannot always be assumed (8–11). Beatty et al. (4) showed that significant quantities of N2 are absorbed by exposed fat and viscera during open abdominal surgery and eliminated by the lungs. Furthermore, a relatively high inspired concentration of N2 is required to achieve precision of measurement (1, 23). These factors hinder the utility of the Haldane approach during anesthesia with O2 and nitrous oxide (N2O), where N2 is superfluous or may be contraindicated because its presence reduces the concentration of O2 or N2O that can be delivered to the patient.
We describe a method that permits continuous automated measurement of gas uptake and elimination for multiple gas species. The method uses CO2 as an extractable marker gas, which is added to both fresh gas and exhaust gas streams to permit total flow measurement at each point. This effectively allows any inspired concentration mixture to be administered unaltered to the patient's lungs and uses conventional clinical gas analyzers. The method was found to be accurate and precise during in vitro testing with a lung simulator by comparison to simultaneous paired measurements made with the Haldane transformation (16). We tested the method under clinical conditions by repeating this comparison in vivo in a series of patients anesthetized with N2O and isoflurane undergoing bowel surgery.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
For the purposes of this experiment, a fully automated measurement system was constructed to measure gas uptake and elimination simultaneously by both the extractable marker method and the Haldane transformation. Because significant error in the estimation of gas uptake with the Haldane transformation is expected in the presence of N2 uptake or elimination by the lungs, these values were corrected for predicted N2 elimination according to the formula of Beatty et al. (4), which was derived in a similar group of patients and adjusted for the inspired concentration of N2 in our study. The equations are given in the APPENDIX.
The system is shown schematically in Fig. 1. It consists of an anesthetic delivery unit, ventilator, and associated Mapleson D breathing system (Bain circuit), with 2 multigas analyzers (Capnomac Ultima, Datex-Ohmeda, Finland) and sample gas multiplexer. All fractional gas concentrations for O2, CO2, N2O, and volatile anesthetic agent were measured in fully mixed gas streams (fresh gas and mixed exhaust gas) to avoid inaccuracy due to tidal changes in gas concentrations. Gas mixing in exhaust gas was achieved by extending the breathing circuit with a length of mixing tubing proximal to sampling points. N2 concentration was calculated by subtraction of the sum of fractional concentrations of all other gases in the mixture from unity.
|
2 min. Sampling proximal to the addition of fresh gas marker CO2 was performed to allow measurement of gas uptake and elimination with the Haldane method. It is not necessary for the extractable marker gas method itself, with the exception that, where the patient's own CO2 elimination was to be measured, gas must still be sampled for CO2 analysis proximal to the addition of marker CO2 in the exhaust limb of the circuit.
A fresh gas flow of 100 ml·kg–1·min–1 was provided. This relatively high fresh gas flow rate reduces the time constant for washthrough of concentration changes in the exhaust limb of the system and improves the response time of the measurement system to real changes in gas uptake and elimination in the lungs. For the Haldane method, a known flow of N2 was added as medical air into the fresh gas flow, which was sufficient to achieve a fresh gas N2 concentration of 30%. For the extractable marker method, 350 ml/min of marker CO2 was added to both the fresh gas and exhaust gas flows, producing roughly a 5.0% concentration. This was chosen as it lay midway within the measurement range of the multigas analyzers for CO2. For both the extractable marker method and the Haldane method, marker gas flows (CO2 and N2) were monitored continuously from gas supply pressures (measured with solid-state pressure transducers) at a high-pressure gas source (cylinders of compressed medical air and CO2 in conjunction with variable pressure regulators, delivering pressures of 400–460 kPa) and the previously determined pressure-flow characteristics of a series of hollow-wire gas flow resistors (see APPENDIX) (6, 21). It was found during bench testing that, at this flow rate, full extraction of CO2 was ensured for 3–4 h before depletion of the soda lime on the fresh gas side became significant, using two standard 1-liter cannisters in series containing soda lime (Medisorb). For additional safety, a software routine was added that alarmed when the CO2 concentration in the fresh gas reaching the patient rose above 0.1% (3 times atmospheric concentration). At this point, the soda lime container could be readily replaced with a fresh one if measurements were continuing.
Despite full mixing of exhaust gas, it was found in patients that transient small changes in functional residual capacity of the lung, due to, for instance, adjustment of abdominal retractors, produced brief but significant real variations in exhaust gas flow (23). Because fresh gas flow rate remains steady, these variations caused excessive scatter in gas uptake measurements from the extractable marker method. These were not seen with the Haldane method because of the smoothing effect of mixing of marker gas N2 in the lung, which is intrinsic to that method. For the purposes of this study, this difficulty was solved by incorporating a second gas analyzer to measure exhaust marker CO2 concentration continuously and average it over two successive measurement cycles to obtain total exhaust flow. This effectively damped the response of the system to the measurement of transient variations in exhaust flow rate, producing a profile similar to that of the Haldane method (as evidenced by the variance of exhaust flow measurements by the 2 methods), thus enabling better comparison of the two methods. Use of separate analyzers for measurement of fresh and exhaust marker CO2 could potentially impair accuracy of uptake calculation. This was avoided by ensuring that the second analyzer was calibrated relative to the first with each measurement cycle by sampling CO2 from the same point in the system with each cycle and scaling the measured value from the second analyzer against the first.
The system was checked for gross leaks using a pneumostatic pressurization maneuver and then calibrated before each experiment by running the measurement system with fresh gas flows and concentrations similar to those to be delivered to the patient by ventilating a rubber bag of suitable compliance. Calculated gas uptakes should be zero during this process. Small adjustments of <1% were generally required in the measured exhaust gas flow rate to achieve this. The performance of the system was found to be stable throughout each experiment, as confirmed by a repeat calibration at the conclusion of each measurement period.
Patient protocol.
With approval from the local ethics committee for use of human subjects according to institutional guidelines and after informed consent was obtained, patients who were to undergo open colonic surgery at the Alfred Hospital, Melbourne, Australia were recruited to the study. Routine anesthetic monitoring was performed, including pulse oximetry and inline capnography. Anesthesia was induced with intravenous propofol and opioid and a neuromuscular blocker, the patient's trachea was intubated, and relaxant anesthesia was maintained with intravenous propofol and supplementary doses of neuromuscular blocker. Patients were ventilated with 70% O2-30% N2 initially until they were stabilized, and the measurement system was calibrated and primed with the fresh gas mixture. This was 30% O2-40% N2O-0.5% isoflurane-balance N2. Patients were then connected to the system, and measurements commenced. The depth of maintenance anesthesia was controlled by adjusting the propofol infusion rate as judged by the anesthesiologist. Measurements continued for up to 90 min or until conclusion of the surgery, whichever was sooner.
Statistical analysis. With each cycle, paired measurements were made by the extractable marker method and the Haldane method of uptake of O2, N2O, and isoflurane and CO2 elimination. The mean measured value by each method and the standard deviation of the difference between them were calculated according to the approach of Bland and Altman (5). The statistical significance of the measured mean difference (bias) was calculated by t-test. Two standard deviations on either side of the mean difference were expressed as the upper and lower limits of agreement between the two methods. The intraclass correlation coefficients (ICC) were also calculated (19). In addition, mean N2 flux was measured by the extractable marker method and compared with the prediction of the formula of Beatty et al. (4) to assess the appropriateness of the correction applied to the measurements made with the Haldane method.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
|
|
Figure 2 plots the difference between measurement of gas uptake and elimination by the extractable marker method (extraction method) and the Haldane transformation (Haldane method) vs. the average of the two methods for O2, N2O, isoflurane, and CO2. Figure 3 is the corresponding correlation plots.
Figure 4 shows a typical plot of changing N2O uptake (N2O) over time in one of our patients. The paired measurements using the two methods are plotted along with the uptake predicted by the formula of Severinghaus (18), adjusted for the inspired concentration of N2O in our study.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
The method can function with a simpler and more streamlined system than that described here, which was designed to compare the method with a known standard and had duplicate sampling. A single gas analyzer only is required to sample fresh gas and mixed exhaust gas in sequence during each cycle. Measurement of the patient's own CO2 production requires further sampling at a point proximal to the addition of marker CO2. However, it is important that most of the cycle time be apportioned to tracking exhaust flow rates because of the random fluctuations in mixed exhaust flow we showed to occur in patients undergoing surgical manipulation. For example, over a 2-min measurement cycle, 15 s of fresh gas sampling is required for accurate reproducible fresh gas flow measurement because of its stability. The remainder of the 2-min cycle would be spent measuring and averaging exhaust marker CO2 concentration.
Our results show that the extractable marker method gives good agreement with simultaneous continuous measurements of gas uptake and elimination with the Haldane approach. This was true for all gas species involved, reflecting stable performance of the method under clinical conditions across a wide range of inspired concentrations, from <1% for isoflurane to 40% for N2O. Further validation of accuracy and precision at a high inspired concentration may be warranted, with a measurement standard other than the Haldane method, allowing N2 to be dispensed with.
The very narrow limits of agreement for CO2 reflects the inherently superior precision of measurement of a gas that is not present in the inspired mixture (16). The limits of agreement were worst for N2O, a finding consistent with the results of our previous in vitro modeling of typical physiological gas exchange scenarios (16). Accurate and precise measurement of N2O is difficult to perform. This is because N2O is taken up at relatively low rates after some time has elapsed after commencement of anesthesia due to its lower solubility in blood and tissues (18). Because of this difficulty, clinical devices manufactured for the measurement of O2 and CO2 in the anesthetic setting have not been accompanied by devices to measure N2O as well, possibly because previous studies have shown that the Haldane transformation provides sufficient precision for measurement of O2 and CO2 if N2O is ignored or assumed (22).
With the use our approach, error in precision is minimized by ensuring that marker CO2 flow rates in both fresh gas and exhaust gas streams are measured from a common source, i.e., a common high-pressure marker gas CO2 supply, with similar gas flow resistors. Random errors or fluctuations in measured gas supply pressure (observed to be of the order of 1–2 kPa) largely cancel out in Eq. 10 for this reason (see APPENDIX).
The precision of the Haldane method has been investigated by previous workers using mathematical modeling. Precision is markedly worse where the fractional concentration of N2 is not directly measured (e.g., by mass spectrometry) but calculated by subtraction from the sum of concentrations of all other gases, as is necessary with conventional technologies for clinical gas concentration measurement (20). Furthermore, precision deteriorates exponentially as the inspired concentration of N2 falls (1). The acceptable lower limit for marker N2 concentration depends on the standard deviation of the measurement of N2 concentration in fresh and exhaust gas streams. We have found that, in our system, which used fully mixed gas streams without tidal variations in gas concentrations, acceptable scatter in measured uptake for all gases for clinical purposes is achieved with fresh gas N2 concentration of 30% (6, 17, 21).
The accuracy of methods relying on the Haldane transformation will suffer in situations, such as inhalational anesthesia, where marker gas mass balance might not be preserved during passage through the breathing system and patient lung. To avoid this potential source of inaccuracy, Heneghan et al. (13) used two different marker gas species (argon and N2) to independently measure fresh gas and exhaust gas flows, with acceptable precision for O2 and CO2 in patients anesthetized with an O2-N2O mixture. They did not attempt to measure N2O, and the use of N2 as a marker in the expired gas overlooked possible error due to N2 elimination in their patients. The extractable marker gas method needs to make no assumptions about the fate of the marker gas in the body.
Beatty et al. (4) measured significant levels of pulmonary elimination of N2 by the body during open abdominal surgery in patients ventilated with pure O2-N2O mixtures. This N2 was presumably absorbed from the atmosphere by exposed fat and viscera. Our data confirm their findings for the expected level of absorption and elimination by the lungs of atmospheric N2 during surgery. The mean measured value of N2 flux in our patients agreed closely with that predicted by Beatty et al. when adjustment was made for the presence of N2 in the inspired gas mixture in our study. It is important to note that where N2 absorption from atmosphere occurs, significant error in measured gas uptake by the Haldane method is expected if this is not taken into account. This is especially the case for N2O, which may be overestimated by 10% or more depending on the inspired N2 concentration (highest error at lower inspired N2 concentration). The extractable marker gas method confirms the importance of this correction. Other studies have suggested non-zero N2 balance in a variety of clinical and metabolic states (8–11, 24). Wilmore and Costill (24) found that this did not significantly influence measurement of O2 during exercise, primarily due to the high levels of O2 achieved. However, our data suggest that accuracy of measurement of lower levels of O2, such as occur under anesthesia, or uptake of other gases, such as N2O, would be more severely compromised.
In conclusion, we measured pulmonary gas uptake and elimination in patients undergoing surgery by means of an automated system that measures fresh gas and exhaust gas flows using CO2 as an extractable marker gas. Comparison with simultaneous measurements made with the Haldane transformation in patients anesthetized with N2O and isoflurane showed acceptable agreement for uptake or elimination of all gases studied. The technique permits continuous measurement of gas uptake and elimination with any inspired gas mixture.
|
APPENDIX |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
Where a marker gas M is fed into the gas stream, total flow rate (FT) can be calculated from the measured concentration of M after thorough mixing. Therefore, for the fresh gas flow
 | (1) |
 | (2) |
FM is the known flow rate of M in fresh gas flow, FFM is the fractional concentration of M in fresh gas flow, 
T is the total mixed exhaust gas flow rate, M is the known flow rate of M in mixed exhaust gas flow, and FM is the fractional concentration of M in mixed exhaust gas flow.
For gas G, uptake is calculated from
 | (3) |
G is uptake of G.
Substituting Eqs. 1 and 2 for these terms in Eq. 3
 | (4) |
is the fractional concentration of G in mixed exhaust gas.
For the extractable marker gas method, using CO2 as M
 | (5) |
For the Haldane transformation (14), 
is related to FT by the ratio of the the concentration of marker gas N2 in fresh gas and exhaust gas streams
 | (6) |
Substituting in Eq. 4
 | (7) |
Pressure-resistance method to measure gas flow.
Marker gas flow was monitored continuously from the pressure vs. flow relationships of the marker gas resistors. These were assumed to be linear over their narrow operating range, obeying an equation of the form
 | (8) |
Extractable marker method.
Because separate flows of marker gas CO2 are fed into fresh gas and exhaust gas streams, each CO2 resistor has its own equation of the form of Eq. 8. However, because identical resistors are used, these coefficients were found to be the same (to within 2 decimal places of precision)
 | (9) |
Accurate measurement of net gas balance in the system requires correction for loss of gas from the system upstream of the exhaust gas sampling point due to sidestream gas sampling by the multi-gas analyzer or to minor leaks (sampling). These were previously determined by volumetric measurement, and the simplifying assumption was made that all lost gas was mixed exhaust gas. Finally, substituting Eq. 5
|
| (10) |
Method using the Haldane transformation.
Given that air contains 79.07% N2 (plus argon etc.)
 | (11) |
Fair is is the known flow rate of air and PMair is the measured gas supply pressure of marker gas N2 in air.
No correction for sampled mixed exhaust gas is required, and the equivalent expression to Eq. 10 for the Haldane method becomes
 | (12) |
However, correction is required for the estimated N2 elimination by the patient (N2). Equation 6 is modified
 | (13) |
 | (14) |
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGMENTSREFERENCES |
|---|
O2 and CO2 measured at the mouth. J Appl Physiol 46: 1122–1126, 1979.
O2 in man. J Appl Physiol 35: 85–89, 1973.This article has been cited by other articles:
|
|
 |
|
  C. Stuart-Andrews, P. Peyton, G. Robinson, D. Terry, B. O'Connor, C. Van der Herten, and B. Lithgow Non-invasive metabolic monitoring of patients under anaesthesia by continuous indirect calorimetry--an in vivo trial of a new method Br. J. Anaesth., January 1, 2007; 98(1): 45 - 52. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
