229
Clinica Chimica Acta, 99 (1979) 229-234 0 Elsevier/North-Holland Biomedical Press
CCA 1177
HALOTHANE INTERFERENCE WITH pOz MEASUREMENTS A METHOD OF INHIBITING ITS EFFECTS
AND
A.G.W. NORDEN and F.V. FLYNN * Department of Chemical Pathology, University College Hospital, Gower Street, London WClE 6AlJ (U.K.) (Received April 25th, 1979)
Summary
The oxygen electrode of a blood-gas analyser (Instrumentation Laboratories Inc. Model 313) responded to halothane, the peak of the response being reached some 30-90 min after exposure. The effect was reduced about tenfold by placing a silicone-rubber membrane over the polypropylene membrane of the oxygen electrode.
Introduction
It has been established that halothane may cause two types of error in the polarographic measurement of oxygen. The first is an electrode response in the absence of oxygen [l-5]; the second is a change in the sensitivity to oxygen [ 1,4]. In one report these errors were not differentiated [ 61. The magnitude of the effects reported has been very variable. The concentrations of halothane used in anaesthesia [7] have sometimes caused overestimation of pOz measurements by amounts that are clinically significant [6]. We have therefore examined the response to halothane of a commonly-used blood-gas analyser, the Instrumentation Laboratories Model 313 instrument (IL 313), and devised a simple electrode modification to render the interference with p02 measurements insignificant. Material and methods Electrode
The oxygen electrode used in this study had been used for about three years in the routine laboratory; the membrane and electrolyte had been changed at * To whom correspondence should be addressed.
230
approximately monthly intervals. A Bryans 27000 chart-recorder was attached to the IL 313 to provide a continuous record of the pOz measurement. Polypropylene membranes (IL 19017) were used unless otherwise stated; they were cut from the same strip of pofypropylene. The membrane and electrolyte were changed whenever the effects of a new halothane-containing specimen were to be studied; an exception was made when the maximum effect of the halothane was equivalent to less than 1 kPa pOz, when up to three experiments were performed without renewing the membrane or electrolyte. Unless the membrane and electrolyte were changed, results were confounded by the effects of previous halothane exposure. For certain experiments (Fig. 3), the polypropylene membrane fitted to the ABL-1 blood-gas analyser (Radiometer Ltd., No. D604) was substituted. For others (Fig. 4) a silicone-rubber membrane, as used on the TL 313 carbon dioxide electrode (IL 56280), was mounted tautly over the polypropylene membrane and fixed with the same O-ring. Before fitting, the silicone membrane was soaked overnight in 0.9% sodium chloride (saline) and a drop of the same solution was used to fill the potential space between the membranes. Standard
gases
The composition of gas mixtures used for calibrating the IL 313 was nominally (a) 12% oxygen, 5% carbon dioxide and 83% nitrogen, and (b) 10% carbon dioxide and 90% nitrogen. Halothane
solutions
Unless otherwise stated, solutions were prepared using halothane with 0.01% thymol (‘Fluoth~e’, ICI. Ltd.) without further purification. Solutions in saline were usually prepared by dilution of a saturated solution in saline at
H&thaw
0
30
60
5-o
Timetmmi Fig. 1. Time-course of the IL oxygen electrode response to halothane. At A, oxygen-free carbon dioxide in nitrogen was passed into the electrode compartment. At B, halothane 4 mmol/l in saline was introduced and kept in contact with the electrode for 19 min. At C. the compartment was again filled with oxygen-free gas; subsequently the ordinate scale was expanded ten-fold and thereafter the electrode response should be read against the scale on the right.
231
U
10 Halothane Concentration
(m!)
20
Halothane
ExposureTime (min)
Fig. 2. Dependence of the IL oxygen electrode response on halothane concentration. Each of a series of halothane solutions was kept in contact with the electrode for a specified time and the peak of the apparent oxygen tension when the electrode compartment was subsequently filled with oxygen-free gas was halothane then measured. 3----O: halothane solutions in saline, exposwe time 1.1 min; X -X: halothane solutions in saline, exposure time 19 min; solutions in saline, exposure time 10 min; O----O: o------o: halothane solutions in blood, exposure time 20.4 min; made UP by introducing six samples for 3.4 min each. Fig. 3. Dependence of the peak response of the IL oxygen electrode on the time of contact with halothane and on the source of the polypropylene membrane. Halothane in saline was kept in contact with the oxygen electrode for varying times, followed by oxygen-free carbon dioxide in nitrogen, and the peak o---0: peak responses with IL membrane, halothane 1 mm&l; response subsequently measured. X-X: peak responses with ABL-1 membrane A, halothane 1 mmol/l; O----O: peak responses with 0: peak responses with IL membrane. halothane 9.5 ABL-1 membrane B, halothane 1 mmol/l; qmmol/l.
room temperature. For some experiments halothane vapour was passed through saline and the resulting solution used. In all cases the saline used was equilibrated with room air at 20-25°C. Saline solutions of halothane were standardised by ultraviolet spectrophotometry against known concentrations prepared by adding 2% (v/v) halothane in acetone to saline; absorbance was measured at 210 nm. Solutions of halothane were kept in stoppered tubes or glass syringes for a maximum of 6 h. Results Preliminary experiments using a range of 1 to 15 mmol/l halothane solutions demonstrated that (a) electrode responses did not begin until about 10 min
232
,!
,
0
,L
30
,
60
,wth SillconeMembrane
90
Tms(min)
Fig. 4. Reduction of the hdothane response of the IL oxygen electrode by a silicone-rubber membrane. At A. oxygen-free carbon dioxide in nitrogen was introduced into the electrode compartment. Starting at B. six samples of blood containing 2.0 mmol/l halothane were introduced. each being kept in contact with the electrode for 3.4 min. At C, the electrode compartment was again filled with oxygen-free gas. The upper curve shows the response of the unmodified electrode and the lower line the response of the electrode with a silicone-rubber membrane superimposed on the polypropylene membrane.
after first exposure to halothane, (b) the maximum shift in the zero oxygen response occurred about 1 h after initial exposure and (c) the maximum shift in the zero oxygen response occurred earlier with higher concentrations of halothane, the delay being 35 min with a 15 mmolfl solution and up to 90 min with a 1 mmol/l solution. The effects of the following variables on electrode response were therefore investigated: (a) time after exposure to halothane solution, (b) the concentration of halothane, and (c) the time of exposure to halothane. The results are shown in Figs. 1, 2 and 3 respectively. Fig. 3 also shows the responses with different polypropylene membranes. Control experiments showed that (a) over the course of 8 h the shift of the baseline when no samples were introduced to the electrode was equivalent to less than 0.3 kPa pOs, (b) over 3 h there was less than a 5% change in the sensitivity to 12% oxygen, and (c) over 2 h there was no systematic change in the zero oxygen response or in the response to 12% oxygen when, 0.01% (w/v) thymol, 10 mmol/l acetone, 100 mmol/l ethanol or saline were introduced. The responses to 1 mmol/l halothane in blood or in saline were indistinguishable under standardised conditions, with either the IL or ABL-1 membranes in use. Maximum responses with the ABL-1 membrane occurred in 20-30 min as compared with 60-90 min with the IL membrane when 1 mmoI/l solutions were examined. Halothane 2.5 mmol/l in saline gave the same response regardless of whether the solution was prepared from halothane vapour or a saturated solution in saline. No change in the sensitivity of the electrode to oxygen was observed in any of our experiments; for instance, the sensitivity to 12% oxygen remained unaItered over the course of 5 h following a 10 min exposure of the electrode to
233
halothane in concentrations up to 9.5 mmol/l. The sample-line tubing of the IL 313 was found to reduce the ultraviolet absorbance of l-15 mmol/l halothane solutions by small but consistent amounts (less than 5%) when containing solutions for 1 h. This effect was probably due to uptake of halothane by the plastic. It is unlikely that this phenomenon would have significantly affected the results reported here, because the dwell-time within the tubing in our experiments was less than 10 sec. Occluding the silicone-rubber membrane of the carbon dioxide electrode, which is directly opposite the oxygen electrode in the IL 313, did not change the response to 2.25 mmol/l halothane, in saline or blood, after an exposure time of 19 min. Changing the membrane and the electrolyte at the time of maximum response to 2.25 mmol/l halothane, to which the electrode had been exposed for 19 min, abolished the halothane effect. Changing the membrane and electrolyte in the absence of halothane and oxygen altered the zero oxygen response by less than the equivalent of 0.15 kPap0,. When a silicone-rubber membrane was placed over the polypropylene membrane of the IL 313, the response to l-5.5 mmol/l halothane was reduced more than ten-fold; the results of one experiment are shown in Fig. 4. Even after 20 samples of blood containing 1 mmol/l halothane were analysed sequentially, each with an electrode exposure time of 3.4 min, the maximum response attributable to halothane was equivalent to less than 0.1 kPapO*. The effect on the response time of the electrode to oxygen of superimposing a silicone-rubber membrane on the polypropylene membrane was investigated; the response time was never more than doubled. No systematic errors resulted from using the silicone membrane at 3 kPa or 20 kPa oxygen tension in blood with a 3.4 min sampling time. Discussion In 1971 Severinghaus et al. [l] showed that halothane may be electrochemitally reduced at oxygen electrodes. This reduction is believed to be the principal cause of the additional current produced at platinum [l] or gold [ 31 electrodes in the presence of halothane. A rapid response to halothane was found with an in vivo electrode by Dent and Netter [3], but we have found delays comparable to those observed by Severinghaus et al. [ 11, Bates et al. [ 21 and Douglas et al. [ 61. In a clinical laboratory, the time-course of the responses we have observed (Fig. 1) means that for a number of hours after analysis of halothane-containing blood, the measurements will be variably displaced. Although one may attempt to compensate for this by adjustment of the machine zero point, the baseline will remain unstable. The observation that changing the source of the polypropylene membrane markedly alters the response time suggests that the delayed response is partly, if not wholly, due to the membrane. Using the solubility data of Steward et al. [8], we have calculated the approximate responses to 1 mmol/l halothane found by previous ,workers. Expressed as though due to oxygen, these were zero [4], 5.3 kPa [2], 5.9 kPa [6], 18.8 kPa [3] and from zero to 20 kPa depending on the polarizing voltage [ 51. We found a peak response of 0.27 kPa. The magnitude of the response to
234
halothane is a steep function of the polarizing voltage used; this is nominally 0.6 volts (IL 313 Operator’s Manual, 1972). This is rather less than that used with the ABL-1 oxygen electrode which is 0.630 volts (ABL-1 M~ual). This difference in operating conditions and the greater response which we found when an ABL-1 membrane was employed (Fig. 3), may account for the larger effect seen by Douglas et al. [6]. The response to halothane as a function of concentration has variously been found to be linear from 0.1 to 1 mmol/l [ 31, saturating at about 1 mmoI/l [ 21 or only signific~t above 0.5 mmol/l with a trans~ut~eous electrode (41. We have found an exponential response except at long exposure times (Fig. 2). Comparison in this manner is approximate, since the integrated response over time, rather than the peak response, can be expected to be related more closely to the concentration of halothane. Many plastics are known to swell in contact with halothane (“Fluothane” Reference Manual, 1976, I.C.I. Ltd.) and thus halothane may progressively facilitate its own passage through the membranes thereby accounting for an exponential rise in electrode current. Previous authors have suggested that low polarizing voltages [ 1,4], the use of polytetrafluoroethylene membranes [ 31 or abrasion of the cathode surface [l] may be useful in reducing the response to halothane. However, McHugh et al. results with variation of the polarizing voltage and [ 5] reported inconsistent . discouraging results when using various different membrane coatings. An observation that halothane is avidly taken up by silicone-rubber led us to conceive the idea of superimposing a silicone-rubber membrane on the polypropylene membrane of the oxygen electrode to trap any halothane and thus prevent its interfering effects. This modification produced a marked and selective reduction of the response of the IL 313 electrode to halothane. It should be effective with other oxygen electrodes where anaesthetic concentrations of halothane apparently cause gross overestimation of oxygen tension [ 61. References 1 Severinghaus, J.W., Weiskopf, R.B., Nishimura, M. and Bradiey, A.F. (1971) J. Appl. Physiol. 31.640642 2 Bates, M.L., Feingold. A. and Gold, M.I. (1975) Am. J. Clin. Pathol. 64. 448451 3 Dent, J.G. and Netter, K.J. (1976) Br. J. Anaesth. 48.195-197 4 Stosseck, K. (1977) Anaesthesiat 26.453455 5 McHugh, R.D.. Epstein, R.M. and Longnecker, D.E. (1979) Anesthesiology 50.4749 6 Douglas, I.H.S., McKenzie, P.J., Ledingham, I.McA. and Smith, G. (1978) Lancet ii, 1370-1371 7 Duncan. W.A.M. and Raventds, J. (1959) Br. J. Anaesth. 31.302-315 8 Steward, A., Allott, P.R., Cowfes, A.L. and Mapleson, W.W. (1973) Br. J. Anaesth. 45, 282-293
Clinica Chimica Acta, 99 (1979) 229-234 0 Elsevier/North-Holland Biomedical Press
CCA 1177
HALOTHANE INTERFERENCE WITH pOz MEASUREMENTS A METHOD OF INHIBITING ITS EFFECTS
AND
A.G.W. NORDEN and F.V. FLYNN * Department of Chemical Pathology, University College Hospital, Gower Street, London WClE 6AlJ (U.K.) (Received April 25th, 1979)
Summary
The oxygen electrode of a blood-gas analyser (Instrumentation Laboratories Inc. Model 313) responded to halothane, the peak of the response being reached some 30-90 min after exposure. The effect was reduced about tenfold by placing a silicone-rubber membrane over the polypropylene membrane of the oxygen electrode.
Introduction
It has been established that halothane may cause two types of error in the polarographic measurement of oxygen. The first is an electrode response in the absence of oxygen [l-5]; the second is a change in the sensitivity to oxygen [ 1,4]. In one report these errors were not differentiated [ 61. The magnitude of the effects reported has been very variable. The concentrations of halothane used in anaesthesia [7] have sometimes caused overestimation of pOz measurements by amounts that are clinically significant [6]. We have therefore examined the response to halothane of a commonly-used blood-gas analyser, the Instrumentation Laboratories Model 313 instrument (IL 313), and devised a simple electrode modification to render the interference with p02 measurements insignificant. Material and methods Electrode
The oxygen electrode used in this study had been used for about three years in the routine laboratory; the membrane and electrolyte had been changed at * To whom correspondence should be addressed.
230
approximately monthly intervals. A Bryans 27000 chart-recorder was attached to the IL 313 to provide a continuous record of the pOz measurement. Polypropylene membranes (IL 19017) were used unless otherwise stated; they were cut from the same strip of pofypropylene. The membrane and electrolyte were changed whenever the effects of a new halothane-containing specimen were to be studied; an exception was made when the maximum effect of the halothane was equivalent to less than 1 kPa pOz, when up to three experiments were performed without renewing the membrane or electrolyte. Unless the membrane and electrolyte were changed, results were confounded by the effects of previous halothane exposure. For certain experiments (Fig. 3), the polypropylene membrane fitted to the ABL-1 blood-gas analyser (Radiometer Ltd., No. D604) was substituted. For others (Fig. 4) a silicone-rubber membrane, as used on the TL 313 carbon dioxide electrode (IL 56280), was mounted tautly over the polypropylene membrane and fixed with the same O-ring. Before fitting, the silicone membrane was soaked overnight in 0.9% sodium chloride (saline) and a drop of the same solution was used to fill the potential space between the membranes. Standard
gases
The composition of gas mixtures used for calibrating the IL 313 was nominally (a) 12% oxygen, 5% carbon dioxide and 83% nitrogen, and (b) 10% carbon dioxide and 90% nitrogen. Halothane
solutions
Unless otherwise stated, solutions were prepared using halothane with 0.01% thymol (‘Fluoth~e’, ICI. Ltd.) without further purification. Solutions in saline were usually prepared by dilution of a saturated solution in saline at
H&thaw
0
30
60
5-o
Timetmmi Fig. 1. Time-course of the IL oxygen electrode response to halothane. At A, oxygen-free carbon dioxide in nitrogen was passed into the electrode compartment. At B, halothane 4 mmol/l in saline was introduced and kept in contact with the electrode for 19 min. At C. the compartment was again filled with oxygen-free gas; subsequently the ordinate scale was expanded ten-fold and thereafter the electrode response should be read against the scale on the right.
231
U
10 Halothane Concentration
(m!)
20
Halothane
ExposureTime (min)
Fig. 2. Dependence of the IL oxygen electrode response on halothane concentration. Each of a series of halothane solutions was kept in contact with the electrode for a specified time and the peak of the apparent oxygen tension when the electrode compartment was subsequently filled with oxygen-free gas was halothane then measured. 3----O: halothane solutions in saline, exposwe time 1.1 min; X -X: halothane solutions in saline, exposure time 19 min; solutions in saline, exposure time 10 min; O----O: o------o: halothane solutions in blood, exposure time 20.4 min; made UP by introducing six samples for 3.4 min each. Fig. 3. Dependence of the peak response of the IL oxygen electrode on the time of contact with halothane and on the source of the polypropylene membrane. Halothane in saline was kept in contact with the oxygen electrode for varying times, followed by oxygen-free carbon dioxide in nitrogen, and the peak o---0: peak responses with IL membrane, halothane 1 mm&l; response subsequently measured. X-X: peak responses with ABL-1 membrane A, halothane 1 mmol/l; O----O: peak responses with 0: peak responses with IL membrane. halothane 9.5 ABL-1 membrane B, halothane 1 mmol/l; qmmol/l.
room temperature. For some experiments halothane vapour was passed through saline and the resulting solution used. In all cases the saline used was equilibrated with room air at 20-25°C. Saline solutions of halothane were standardised by ultraviolet spectrophotometry against known concentrations prepared by adding 2% (v/v) halothane in acetone to saline; absorbance was measured at 210 nm. Solutions of halothane were kept in stoppered tubes or glass syringes for a maximum of 6 h. Results Preliminary experiments using a range of 1 to 15 mmol/l halothane solutions demonstrated that (a) electrode responses did not begin until about 10 min
232
,!
,
0
,L
30
,
60
,wth SillconeMembrane
90
Tms(min)
Fig. 4. Reduction of the hdothane response of the IL oxygen electrode by a silicone-rubber membrane. At A. oxygen-free carbon dioxide in nitrogen was introduced into the electrode compartment. Starting at B. six samples of blood containing 2.0 mmol/l halothane were introduced. each being kept in contact with the electrode for 3.4 min. At C, the electrode compartment was again filled with oxygen-free gas. The upper curve shows the response of the unmodified electrode and the lower line the response of the electrode with a silicone-rubber membrane superimposed on the polypropylene membrane.
after first exposure to halothane, (b) the maximum shift in the zero oxygen response occurred about 1 h after initial exposure and (c) the maximum shift in the zero oxygen response occurred earlier with higher concentrations of halothane, the delay being 35 min with a 15 mmolfl solution and up to 90 min with a 1 mmol/l solution. The effects of the following variables on electrode response were therefore investigated: (a) time after exposure to halothane solution, (b) the concentration of halothane, and (c) the time of exposure to halothane. The results are shown in Figs. 1, 2 and 3 respectively. Fig. 3 also shows the responses with different polypropylene membranes. Control experiments showed that (a) over the course of 8 h the shift of the baseline when no samples were introduced to the electrode was equivalent to less than 0.3 kPa pOs, (b) over 3 h there was less than a 5% change in the sensitivity to 12% oxygen, and (c) over 2 h there was no systematic change in the zero oxygen response or in the response to 12% oxygen when, 0.01% (w/v) thymol, 10 mmol/l acetone, 100 mmol/l ethanol or saline were introduced. The responses to 1 mmol/l halothane in blood or in saline were indistinguishable under standardised conditions, with either the IL or ABL-1 membranes in use. Maximum responses with the ABL-1 membrane occurred in 20-30 min as compared with 60-90 min with the IL membrane when 1 mmoI/l solutions were examined. Halothane 2.5 mmol/l in saline gave the same response regardless of whether the solution was prepared from halothane vapour or a saturated solution in saline. No change in the sensitivity of the electrode to oxygen was observed in any of our experiments; for instance, the sensitivity to 12% oxygen remained unaItered over the course of 5 h following a 10 min exposure of the electrode to
233
halothane in concentrations up to 9.5 mmol/l. The sample-line tubing of the IL 313 was found to reduce the ultraviolet absorbance of l-15 mmol/l halothane solutions by small but consistent amounts (less than 5%) when containing solutions for 1 h. This effect was probably due to uptake of halothane by the plastic. It is unlikely that this phenomenon would have significantly affected the results reported here, because the dwell-time within the tubing in our experiments was less than 10 sec. Occluding the silicone-rubber membrane of the carbon dioxide electrode, which is directly opposite the oxygen electrode in the IL 313, did not change the response to 2.25 mmol/l halothane, in saline or blood, after an exposure time of 19 min. Changing the membrane and the electrolyte at the time of maximum response to 2.25 mmol/l halothane, to which the electrode had been exposed for 19 min, abolished the halothane effect. Changing the membrane and electrolyte in the absence of halothane and oxygen altered the zero oxygen response by less than the equivalent of 0.15 kPap0,. When a silicone-rubber membrane was placed over the polypropylene membrane of the IL 313, the response to l-5.5 mmol/l halothane was reduced more than ten-fold; the results of one experiment are shown in Fig. 4. Even after 20 samples of blood containing 1 mmol/l halothane were analysed sequentially, each with an electrode exposure time of 3.4 min, the maximum response attributable to halothane was equivalent to less than 0.1 kPapO*. The effect on the response time of the electrode to oxygen of superimposing a silicone-rubber membrane on the polypropylene membrane was investigated; the response time was never more than doubled. No systematic errors resulted from using the silicone membrane at 3 kPa or 20 kPa oxygen tension in blood with a 3.4 min sampling time. Discussion In 1971 Severinghaus et al. [l] showed that halothane may be electrochemitally reduced at oxygen electrodes. This reduction is believed to be the principal cause of the additional current produced at platinum [l] or gold [ 31 electrodes in the presence of halothane. A rapid response to halothane was found with an in vivo electrode by Dent and Netter [3], but we have found delays comparable to those observed by Severinghaus et al. [ 11, Bates et al. [ 21 and Douglas et al. [ 61. In a clinical laboratory, the time-course of the responses we have observed (Fig. 1) means that for a number of hours after analysis of halothane-containing blood, the measurements will be variably displaced. Although one may attempt to compensate for this by adjustment of the machine zero point, the baseline will remain unstable. The observation that changing the source of the polypropylene membrane markedly alters the response time suggests that the delayed response is partly, if not wholly, due to the membrane. Using the solubility data of Steward et al. [8], we have calculated the approximate responses to 1 mmol/l halothane found by previous ,workers. Expressed as though due to oxygen, these were zero [4], 5.3 kPa [2], 5.9 kPa [6], 18.8 kPa [3] and from zero to 20 kPa depending on the polarizing voltage [ 51. We found a peak response of 0.27 kPa. The magnitude of the response to
234
halothane is a steep function of the polarizing voltage used; this is nominally 0.6 volts (IL 313 Operator’s Manual, 1972). This is rather less than that used with the ABL-1 oxygen electrode which is 0.630 volts (ABL-1 M~ual). This difference in operating conditions and the greater response which we found when an ABL-1 membrane was employed (Fig. 3), may account for the larger effect seen by Douglas et al. [6]. The response to halothane as a function of concentration has variously been found to be linear from 0.1 to 1 mmol/l [ 31, saturating at about 1 mmoI/l [ 21 or only signific~t above 0.5 mmol/l with a trans~ut~eous electrode (41. We have found an exponential response except at long exposure times (Fig. 2). Comparison in this manner is approximate, since the integrated response over time, rather than the peak response, can be expected to be related more closely to the concentration of halothane. Many plastics are known to swell in contact with halothane (“Fluothane” Reference Manual, 1976, I.C.I. Ltd.) and thus halothane may progressively facilitate its own passage through the membranes thereby accounting for an exponential rise in electrode current. Previous authors have suggested that low polarizing voltages [ 1,4], the use of polytetrafluoroethylene membranes [ 31 or abrasion of the cathode surface [l] may be useful in reducing the response to halothane. However, McHugh et al. results with variation of the polarizing voltage and [ 5] reported inconsistent . discouraging results when using various different membrane coatings. An observation that halothane is avidly taken up by silicone-rubber led us to conceive the idea of superimposing a silicone-rubber membrane on the polypropylene membrane of the oxygen electrode to trap any halothane and thus prevent its interfering effects. This modification produced a marked and selective reduction of the response of the IL 313 electrode to halothane. It should be effective with other oxygen electrodes where anaesthetic concentrations of halothane apparently cause gross overestimation of oxygen tension [ 61. References 1 Severinghaus, J.W., Weiskopf, R.B., Nishimura, M. and Bradiey, A.F. (1971) J. Appl. Physiol. 31.640642 2 Bates, M.L., Feingold. A. and Gold, M.I. (1975) Am. J. Clin. Pathol. 64. 448451 3 Dent, J.G. and Netter, K.J. (1976) Br. J. Anaesth. 48.195-197 4 Stosseck, K. (1977) Anaesthesiat 26.453455 5 McHugh, R.D.. Epstein, R.M. and Longnecker, D.E. (1979) Anesthesiology 50.4749 6 Douglas, I.H.S., McKenzie, P.J., Ledingham, I.McA. and Smith, G. (1978) Lancet ii, 1370-1371 7 Duncan. W.A.M. and Raventds, J. (1959) Br. J. Anaesth. 31.302-315 8 Steward, A., Allott, P.R., Cowfes, A.L. and Mapleson, W.W. (1973) Br. J. Anaesth. 45, 282-293
