ACID-BASE CHANGES IN ARTERIAL BLOOD ASSOCIATED WITH SPONTANEOUS AND CONTROLLED VENTILATION DURING ANAESTHESIA

Brit. J. Anaesth. (1965), 37, 492

ACID-BASE CHANGES LN ARTERIAL BLOOD ASSOCIATED WITH SPONTANEOUS AND CONTROLLED VENTILATION DURING ANAESTHESIA BY

R. A. MILLAR AND B. E. MARSHALL

With the technical assistance of D. L. POWELL New Addenbrooke's Hospital, Cambridge

The present study was initiated primarily to assess changes in the whole blood non-respiratory component of acid-base balance occurring over several hours of neurosurgical anaesthesia, based on techniques employing nitrous oxide, oxygen and halothane. In this report are also included normal acid-base values, changes associated with induced hypothermia for neurosurgery, and measurements made during anaesthesia for surgery outside the abdomen and thorax, with spontaneous and controlled ventilation.

METHOD

Neurosurgical patients Measurements of arterial pH, Pco2, and (whole blood) standard bicarbonate were made in 78 patients, aged 16 to 73 years, before and during anaesthesia for neurosurgical procedures. Premedication was as follows: in 13 patients, morphine 10 mg or pethidine 50 mg, with promethazine 25 or 50 mg or atropine 0-6 mg; in 8 patients, papaveretum 20 mg with hyoscdne 0-4 mg; in 3 patients, promethazine 50 mg; in 17 patients, atropine only;

the remaining 37 patients were not premedicated. After local infiltration with 2 per cent procaine, a Riley needle was inserted percutaneously into the brachial or radial arteries, usually before induction of anaesthesia. An arterial sample (2 ml) was withdrawn and anaesthesia was induced with intravenous thiopentone (2-5 per cent) in amounts from 200 to 400 mg, with suxamethonium 60 mg. A cuffed, armoured latex endotracheal tube was inserted following pertracheal injection of 2-3 ml 4 per cent lignocaine hydrochloride. Anaesthesia was continued with halothane in nitrous oxide and oxygen (concentrations of which were 30 per cent or higher); light surgical anaesthesia with spontaneous respiration was maintained during positioning of the patient, intravenous cannulation, attachment of percutaneous e.c.g. needle electrodes, head shaving, and other essential manipulations. A slow intravenous infusion of 5 per cent dextrose in water was begun. During this period of spontaneous respiration, using a Magill semi-open circuit, one or more arterial samples (2 ml) were withdrawn before transfer of the patient from the anaesthetic to the operating room.

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

SUMMARY

Arterial pH, Pa>2, and whole blood standard bicarbonate were measured in surgical patients before, during, and after spontaneous or controlled pulmonary ventilation with nitrous oxide, oxygen and halothane. Relative increases in arterial Pco2 were accompanied by significant reductions in standard bicarbonate, indicative of a non-respiratory acidaemia. This effect may be attributable in part to differences between the log Pco2/pH equilibration lines for blood and for the whole organism. Small reductions in standard bicarbonate occurred over several hours of pulmonary overventilation in neurosurgical patients during induced hypothermia, and in the early postoperative period.

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION

Orthopaedic patients.

Similar arterial acid-base measurements were made in 32 patients anaesthetized for orthopaedic surgery. Pulmonary ventilation was spontaneous in 16 patients and controlled in the same number. Light narcotic premedication was given, with atropine 0-6 mg. Normal conscious subjects.

Arterial pH, Pco2 and standard bicarbonate were measured in 51 subjects attending outpatient clinics; blood was withdrawn slowly from the brachial or radial arteries, using a technique described previously (Marshall, 1964), the subject

lying supine. The patients were free from known respiratory or cardiovascular disease. Acid-base measurements.

Arterial blood was withdrawn into 2-ml syringes containing a stainless steel mixing ring, and with the deadspace filled with heparin (Evans Medical Ltd.) 5,000 units/ml. Arterial pH, Pco*, and standard bicarbonate were measured by the interpolation technique of Siggaard-Andersen et al. (1960). The following Radiometer equipment was used: pH meter 27; microelectrode E 5021; tonometer AMT 1. Two complete sets of this apparatus were available within the same laboratory. Thermostat temperature was maintained at 38±0-5°C, except when temperature variations were being studied. BOC calibrated gas cylinders were used, containing approximately 4 and 8 per cent carbon dioxide in oxygen. The carbon dioxide concentrations were checked repeatedly with a microanalyzer (Scholander, 1947); when discrepancies (in approximately one in four cylinders) were detected between the stated and measured gas concentrations, the values obtained in our laboratory were used. Radiometer buffer, pH 7-381, was used only when first unsealed, in order to cross-check buffer solution made to the same formula in our laboratory. Stock buffer solutions were made every 3 months, and stored at 3°C; the supply was drawn upon weekly, for daily use in the laboratory. Between whole blood pH measurements, the electrode was rinsed with "old" buffer solution, and the correct "fresh" buffer reading was rechecked; the electrode was then flushed with distilled water. All pH readings were made in duplicate; if the readings differed by more than 0015 pH, they were rejected and the measurements were repeated. Syringes were held in the vertical position, nozzle upwards, during the period of sampling and reading of actual arterial pH. Small air bubbles inevitably introduced by sampling were immediately expelled. Measurement of pH was made when a steady reading was maintained; this took longer when arterial pH was high. Reliable readings could not be obtained unless the suction connection at the rear of the microelectrode was disconnected during measurement (in most recent microelectrodes this is automatic).

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

In 30 patients intravenous infusions of up to 90 g lyophilized urea in 10 per cent invert sugar solution (Urevert, Travenol Lab. Inc.) were begun at about the time of transfer from the anaesthetic to the operating room. In 5 patients, respiration remained spontaneous throughout the subsequent surgical procedures. In the remainder, mechanical ventilation was begun on transfer to the operating theatre; a Bird ventilator Marks 4 and 8 was used, usually with a Boyle Mark 3 circle carbon dioxide absorber, but occasionally with a Ruben valve alone. Ventilation volumes varied up to 30 l./min (measured with a Wright respirometer during the expiratory phase). Tubocurarine (Tubarine) in total amounts up to 30 mg, was used in the majority of patients. In 5 neurosurgical patients the lungs were ventilated mechanically immediately after endotracheal intubation, and throughout the remaining period of anaesthesia. Muscle relaxation was reversed, following surgery, with intravenous atropine 1-2 mg and neostigmine 2-5 mg. Body temperature was lowered by surface cooling with ice in 9 patients; oesophageal and rectal temperatures were measured with thermocouples. In hypothermic patients, oesophageal temperature was used for correction of acid-base measurements when these were made at 38°C. Arterial samples were withdrawn usually at intervals of 1 hour or less during anaesthesia in all patients; a few patients were also studied postoperatively. Acid-base measurements made at 38°C were corrected to the patient's rectal temperature at the time of sampling.

493

494

BRITISH JOURNAL OF ANAESTHESIA

RESULTS

Under the conditions specified under "Method", the mean difference within members of 200 duplicate pairs of whole blood pH determinations was 0004. The mean pH difference within 200 duplicate pairs of equilibrations was 0-O05 for the approximately 4 per cent carbon dioxide concentration, and 0-006 for the approximately 8 per cent carbon dioxide concentration. Twenty determinations of standard bicarbonate were carried out in sequence on a sample of human arterial blood, which was immersed in iced water between measurements. The standard deviation about the mean value (23-3 mM/1.) was ±0-24 mM/1. Although the latter measurements were made over a period of nearly 5 hours, there was no reduction in standard bicarbonate with time, confirming that storage in iced water is effective for several hours in preventing changes in the metabolic component caused by glycolysis. Since a small fall in body temperature sometimes occurred during neurosurgical operations, it was considered whether discrepancies between the actual body temperature at which a blood sample was removed, and the temperature of equilibration (38°C) of the same sample, influenced the measured value of standard bicarbonate. An attempt was made to determine this indirectly; measurements of standard bicarbonate were made in triplicate on samples of blood withdrawn from 9 patients, the equilibrations and pH determinations being carried out at both 33°C and 38°C, using one micro-Astrup equipment. The 38°C nomogram was used for both temperatures, with appropriate temperature adjust-

ment to pH meter and buffer values, and for water vapour pressure. The difference between the averaged values for the two temperatures was established for the 9 patients individually; the mean overall difference, 013 mM/1. (SD±0-17), was insignificantly different from zero by Student's t-test. A similar comparison on anothertblood sample, involving six determinations at both 38°C and 30°C, using two micro-Astrup equipments, gave mean standard bicarbonate values of 26-6 (SD±0-30) and 26-3 mM/1. (SD±0-13) respectively. These results suggest that negligible "errors" in standard bicarbonate are likely to be introduced by equilibration at 38°C of blood samples whose temperature is substantially lower at time of withdrawal. Replotting of the previous measurements obtained by equilibration at 30°C, on a nomogram adjusted for the correct pK at 30°C, gave "standard bicarbonate" values averaging 23-9 mM/1. (SD±013), indicating that in accord with the definition of standard bicarbonate the equilibrated pH values should always be referred to the 38°C nomogram (Siggaard-Andersen, 1964), in spite of differences either in the withdrawal temperature of sequential blood samples, or in the temperature of equilibration and pH measurement. Normal acid-base values.

Table I shows the arterial pH, Pco2, and standard bicarbonate values, with standard deviations, which were measured in 51 conscious, recumbent subjects free from known respiratory or cardiovascular disease. In contrast to arterial P02, which is age dependent (Marshall and Millar, 1965), mean arterial pH, Pco2, and standard bicarbonate were almost identical in the age groups 20-40 and 41-71 years. Spontaneous respiration in anaesthetized prior to neurosurgery.

patients

The effects of induction of anaesthesia, and a period of spontaneous respiration, were assessed in a group of 30 consecutive and unselected patients requiring major neurosurgical procedures. Since a few individuals showed minor unsuspected acidbase disturbances (usually metabolic alkalosis), the values before induction of anaesthesia are not representative of the normal population. Average arterial pH, Pco2, and standard bicarbonate, were

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Whole blood equilibrations were carried out in duplicate at each carbon dioxide concentration and were repeated if the pH readings differed by more than 0015. The equilibrating chambers were cleaned regularly, Kleenex tissue being preferred for this purpose. At the start and close of each measurement period, buffer solution was placed in the equilibrating chambers, which were shaken for several minutes, then dried thoroughly. No corrections for the temperature difference of approximately 0-4°C between the pH electrode and the equilibrating chambers of the Radiometer apparatus have been made to the measurements presented.

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION

495

sample withdrawal; the fall in standard bicarbonate

TABLE I

ArterialpH, Pcov and standard bicarbonate (mean ±SD) averaged 0-93 mM/1. (range +0-5 to -2-0), in 51 subjects free of knovm respiratory or cardiovascular associated with a mean rise in arterial Pco2 of 130 disease.

pH Total

n=51

Age 21 to 40 n = 27 Age 41 to 71 n=24

7-411

±0013 7-409 ±0016 7-412 ±0010

Standard Pco, bicarbonate mm Hg (mM/1.) 37-2 ±1-6 37-2

±1-7 37-1

±1-5

23-3 ±0-90 23-3 ±0-70 23-3 ±1-1

mm Hg (range 7-22). In these groups, the data did not demonstrate greater increases in arterial Pco2 when morphine, pethidine, or papaveretum was administered prior to spontaneous respiration of nitrous oxide, oxygen and halothane. Spontaneous respiration during neurosurgery.

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Because of the unsatisfactory neurosurgical operating conditions existing when respiration was allowed to continue unassisted during halothane TABLE II Changes {mean ±SE) in arterial pH, Pcov and standard anaesthesia, extended studies of spontaneous bicarbonate in 30 patients after 15 to 60 minutes of anaes- respiration in neurosurgery were limited to 5 thesia with nitrous oxide, oxygen and halothane with patients who were anaesthetized for 4 to 5 hours; spontaneous respiration, prior to neurosurgery. a non-rebreathing system was used. Standard bicarbonate was reduced by a mean of 1-7 mM/1. Standard Pco, bicarbonate after 1 hour of spontaneous respiration, associated mm Hg mM/1. with a small rise in arterial Pa>2 to an average of 47 -0089 -0-82 mm Hg. In these 5 patients the mean difference + 121 ±0032 ±015 between the standard bicarbonate values measured ±1-2 after 1 hour and after 3 hours of spontaneous respectively 7-428 (range 7-333-7-500), 36-4 mm respiration was only 0-2 mM/1., although Pco2 had Hg (range 24-18), and 24-0 mM/L (range 22-0-28-4). increased to 55 mm Hg over this period. However, Following induction of anaesthesia and 15 to at 3 hours, standard bicarbonate was reduced by a 60-minute periods of spontaneous respiration of further 0-6 mM/1. in the 2 patients not given urea, halothane (0-5-1-5 per cent) in nitrous oxide and and increased by 0-7 mM/1. in the 3 patients to oxygen (30 per cent), standard bicarbonate showed whom urea was administered. These findings are an average fall of 0-82 mM/1. in the 30 patients inconclusive in showing whether continued periods (table II). This was a highly significant change of spontaneous respiration lead to a progressive from the pre-anaesthetic level (SE of mean dif- reduction in standard bicarbonate, although the ference ±0-15, P < 0-001). Negligible changes changes produced appear to be small and may be occurred in 8 patients, and in the remainder the fall influenced by administration of intravenous urea. in standard bicarbonate ranged from 0-2 to 3-3 mM/1. There was an accompanying respiratory Mechanical overventilation following earlier acidosis, the increase in arterial Pcc>2 over prespontaneous respiration. anaesthetic levels averaging 12-1 mm Hg (range In most of the neurosurgical patients studied, +0-3 to 27); this change was highly significant mechanical pulmonary overventilation was started (P< 0-001). At these variable times, the magnitude following earlier periods of spontaneous respiration. of the decrease in standard bicarbonate was not Since the minute ventilation volumes induced were significantly correlated with the degree of increase in the range 12 to 30 l./min, a pronounced respirain arterial Pco2- The associated fall in arterial pH tory alkalosis frequently resulted. Table III shows was to 7-339, also a highly significant reduction the changes in standard bicarbonate after 1,2, and from the pre-anaesthetic level (SE of mean dif- 3 hours of pulmonary ventilation, in comparison ference ±0032, P<001) (table II). In this group with the values measured both before induction of no intravenous urea had been given. anaesthesia and prior to starting mechanical ventiThe changes were similar in 7 patients to whom lation; in this group of patients no urea was small amounts of urea had been given at the time of administered at any time. Arterial Pco2 averaged

BRITISH JOURNAL OF ANAESTHESIA

496 TABLE III

Changes in standard bicarbonate associated with pulmonary overventilation in neurosurgical patients, related to the levels before induction of anaesthesia, and to those measured at the end of the period of spontaneous respiration which preceded overventilation. No urea was administered. P is the probability that the mean changes in standard bicarbonate differ significantly from zero. NS=not significant.

1 hour

2 hours

3 hours

n Change in standard bicarbonate (mM/1.)

-11

-1-5

-1-7

SE P

±0-25 <0001

±0-36

±0-41

n SE P

12

20

16

-0-59 ±0-32

-0-26 ±0-25 NS

Mean arterial Pco t (mmHg)

< 0-001

NS

14

14

7

<001 8 -11

±0-42 <005

[ J

Comparison with preanaesthetic values Comparison with values before start of overventilation

1 1

J

12

TABLE IV

Changes in standard bicarbonate associated with pulmonary overventilation in neurosurgical patients to whom urea had been administered intravenously before the 1 hour sample. The changes relate to the standard bicarbonate levels before induction of anaesthesia, and to those measured at the end of the period of spon~ taneous respiration which preceded overventilation. P is the probability that the mean changes in standard bicarbonate differ significantly from zero. NS=not significant.

1 hour n

Change in standard bicarbonate (mM/1.)

SE P

22

-1-2

±0-26

<0001

n

Change in standard bicarbonate (mM/1.) SE P

Mean arterial Pco, (mm Hg)

25

+ 0 07 ±0-17 NS

13

37-9 before anaesthesia, and 47-2 mm Hg during spontaneous respiration preceding overventilation; the mean levels during overventilation are shown in the table. It is apparent that standard bicarbonate remained significantly lower in the course of pulmonary overventilation than before induction of anaesthesia, and that there was a further progressive fall in standard bicarbonate during overventilation which became significant after 3 hours. Table IV shows the changes measured in patients to whom intravenous infusions of 30 per cent urea were administered within the first 30 minutes of

2 hours 19

-0-83

±0-16

< 0-001 21

+ 0-37 ±0-26 NS

12

3 hours 11

-0-64

±0-20 <0-01 11

+ 0-38 ±0-44 NS

~)

Comparison with

f

preanaesthetic J values "I Comparison with values V before start of J overventilation

11

overventilation. The fall in standard bicarbonate already present during the period of spontaneous respiration persisted during mechanical ventilation, but there was a gradual reversion toward preanaesthetic levels. Average arterial PcO2 during overventilation is shown in the table; the value before anaesthesia was 35-8, while during the period of spontaneous respiration preceding overventilation Pco2 averaged 47-9 mm Hg. The changes from pre-anaesthetic standard bicarbonate levels after 3 hours of mechanical ventilation in the patients from the urea and nonurea groups were respectively —0-64 and —1-7

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Change in standard bicarbonate (mM/1.)

17

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION

periods of pulmonary overventilation, toward the end of neurosurgical procedures. Arterial Pco2 increased, over the transitional period, from an average of 16 mm Hg just prior to discontinuing mechanical ventilation, to 37 mm Hg at the time when spontaneous respiration was resumed or shortly thereafter. Accompanying these rises in Pco2 there was a highly significant reduction in standard bicarbonate averaging 0-76 mM/1. (SE of mean difference ±0-22, P<001). The individual values are shown in table V. TABLE V

The differences, in 17 patients, between standard bicarbonate levels during pulmonary overventilation and as spontaneous respiration was resumed toward the end of operation. The relevant arterial Pco, data are also shown. Difference between standard bicarbonate (mM/1.) during ovcrventiianon and on resumption of spontaneous respiration -0-9 -1-7 -1-8

Arterial Pco, (mmHg)

During overventilfltion

-1-9 -0-5 -0-5 -0-5 -10

+ 0-5 +0-3 -0-8 -1-5

-20

In 17 patients, an assessment was made of the change in standard bicarbonate associated with the resumption of spontaneous respiration following

+ 19

13 16 12 25 16 22 14 17 14 11 16 13 17 16 23 8 17

00

Change to spontaneous respirationfollowing pulmonary overventilation.

Difference at start of spontaneous respiration

+ 1-3 -0-7 -1-2

+21

+ 18

+27 +48

+ 25 + 19 + 15

+35

+ 10 + 17 +27

+6

+20

+ 15

+9 +25

TABLE VI

Effects of an inducedrisein arterial Pco, on standard bicarbonate values, in 7 patients previously overventilated with nitrous oxide, oxygen and halo thane. Overventilation

pH

7-785 7-771 7-521 7-734 7-782 7-685 7-697

Pco, (mm Hg) 10 10 16 10 10 11 15

Standard bicarbonate (mM/1.) 23 0 24-5 22 0 21 0 20-2 21-1 24-5

CO, ventilation

.Standard bicarbonate (mM/1.) 21-5 221

21-5 20-2 18-7 19-6 22-8

Pco, (mm Hg)

PH

53 91 82 65 37 86 57

7-300 7-231 7185 7-232 7-340 7165 7-267

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

mM/1.; these values differed significantly (SE of difference between means ±0-37, P<0-05). This evidence indicates that, under the conditions of study, urea administration limited the progression of non-respiratory aridaemia, or permitted its reversal, when a period of mechanical pulmonary overventilation followed spontaneous respiration. However, although no modifications to the treatment given to these two groups were made because of urea administration, it is not possible to exclude factors related to the neurosurgical pathology, which might both have determined the indications for urea infusion and have affected the changes measured. The average amounts of blood infused were 2-0 and 1-3 standard units, and the proportion of patients transfused was 96 per cent and 65 per cent, in the urea and non-urea groups respectively; this is indicative of a greater number of patients with more major intracranial pathology in the urea group. The rate of blood administration did not exceed 2 units per hour in any patient. Extreme respiratory alkalosis was frequently induced during mechanical overventilation in these neurosurgical patients; a comparison of the preanaesthetic values for standard bicarbonate with those accompanying arterial pH levels above 7-70 during anaesthesia showed a mean reduction of 0-51 mM/1. (SD ±0-84) in 9 patients to whom urea had been administered, and of 1-3 mM/1. (SD ± 1-1) in 7 patients who had not been given urea.

497

498

BRITISH JOURNAL OF ANAESTHESIA

Hises in arterial Pa>2 induced in paralyzed, mechani- Acid-base state before anaesthesia and immediately cally overventilated patients. following surgery.

In 14 patients from this neurosurgical group, the standard bicarbonate levels prior to induction of anaesthesia were compared with those when spontaneous respiration appeared to be adequately restored after surgery had terminated. The individual changes, together with relevant values for arterial Pcx»2, are shown in Table VII. The reduction in standard bicarbonate at the end of operation, mean 1-6 mM/1., was highly significant (SE of mean difference ±0-37, P < 0001). Mechanical ventilation after intubation in neurosurgical patients.

In view of the previous findings, which point to changes in pulmonary ventilation as the main Differences between standard bicarbonate before anaesthesia, and during spontaneous respiration at the end of surgery.factor causing reductions in standard bicarbonate The pre-anaesthetic and end-operation arterial Pcot levels during anaesthesia, 5 patients anaesthetized for are also shovm. major neurosurgical procedures were ventilated mechanically immediately following endotracheal Standard bicarbonate intubation. Table VIII shows the individual (mM/1.) arterial Pco2 and standard bicarbonate values prior Difference to anaesthesia, and after periods of mechanical between Arterial Pco, (mm Hg) pulmonary ventilation lasting for 20-45 minutes and End-operation before neurosurgery. The average time, 34 minutes, end-operation (spontaneous is by chance identical to that for the spontaneous values Pre-anaesthetic respiration) respiration group of table II. In these 5 patients -10 41 35 arterial Pco2 was moderately reduced and there 32 43 -3-5 -21 41 49 was no accompanying non-respiratory acidaemia 21 35 + 1-2 (table VIII). In 3 of the same patients mechanical 34 33 -2-3 16 pulmonary ventilation was continued, during major 41 +0-5 38 13 -0-4 neurosurgical procedures, for 4 hours. At this time 34 38 -1-7 standard bicarbonate was reduced by 0-9, 10, and -20 38 35 1-4 mM/1. respectively, the associated levels of 36 -1-1 45 -11 40 32 arterial Pco2 being 34-5, 13-5, and 18-8 tnm Hg. 34 -2-8 30 -20 37 52 Postoperative acid-base changes in neurosurgical 39 -3-6 64 patients were similar to those measured in the TABLE VII

TABLE VIII

Arterial Pco% and standard bicarbonate before anaesthesia, and after periods of mechanical ventilation with nitrous oxide, oxygen and halo thane, started immediately after endotracheal intubation, for which suxamethonium was used in patients 1 and 4, and tubocurarine in the remainder. Pco, (mm Hg)

Patient No. 1 2 3 4 5

During

Before anaesthesia 32-5 41-5 33-5

ventilation 24-4 23-4 28-0

380

180

35 0

28-5

Standard bicarbonate (mM/1.) Before anaesthesia 25 0 23-6 231

24-2 25-5

During ventilation 25 0 23-7 23-2 24-7 25-7

anaesthesia (minutes) 45 20 40 30 35

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Following periods of mechanical overventilation in 7 fully curarized patients, arterial Pco2 was increased from a level averaging about 12 mm Hg, by addition of carbon dioxide to the inspired gases (without changing the oxygen concentration or minute volume). The changes in individual patients are shown in table VI. After 5 to 15 minutes, arterial Pcc>2 had increased to a mean of 66 mm Hg, and there was a highly significant reduction in standard bicarbonate averaging 1-5 mM/1. (SE of mean difference ± 0 1 9 , P<0-001). Subsequent lowering of P2 toward normal resulted in a rise in the level of standard bicarbonate.

499

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION orthopaedic group (see bdow); representative data from a patient anaesthetized for posterior fossa exploration are shown in table IX. The relative non-respiratory acidosis which developed gradually during anaesthesia, and more markedly as spontaneous respiration was resumed at the end of operation, was already being reversed 3 hours postoperatively. It is of interest that the changes in arterial Pc»2 at the close of anaesthesia and surgery were small, the upper level not exceeding 40 mm Hg, and that the patient was shivering.

end of periods of anaesthesia and surgery lasting from 35 to 180 minutes (mean 71); there was also a highly significant reduction in standard bicarbonate of 1 -4 mM/1., which became gradually less marked over the early postoperative period. Arterial Pco2 was still significantly raised, and standard bicarbonate reduced, 1 hour postoperatively by comparison with the pre-anaesthetic levels. After 3 hours the changes were regressing but still evident.

TABLE XA

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Controlled ventilation during anaesthesia for orthopaedic surgery. In contrast to the above group, the 16 patients TABLE IX Standard bicarbonate and arterial Pcot levels during and ventilated mechanically showed small, statistically after neurosurgery in a patient ventilated mechanically, in insignificant reductions in standard bicarbonate at the sitting position, for posterior foua exploration. the end of operation, in association with an average lowering of arterial Pco2 from 36 mm Hg before Standard bicarbonate Pco, anaesthesia to 22 mm Hg after periods of anaes(mm Hg) (mM/1.) thesia lasting from 40 to 240 minutes (mean 84) 35 0 Pre-anaesthetic. 25-5 (table XB). However, in relation to the postAfter 4 hours mechanical venti24-6 34-5 operative rises in arterial Pcc>2 above pre-anaesthetic lation in sitting position. 400 After 5 minutes spontaneous levels, standard bicarbonate was then reduced 23-5 respiration of 100% oxygen. significantly. The average values remained lower 13 minutes later; after 5 minutes for 3 hours postoperatively, although the reductions 22-6 36-5 b r e a t h i n g air; awake and shivering (rectal temp. 36-4°C). were insignificant statistically after 1 and 3 hours. 3 hours later in ward; breathing 23-8 32-7 In patients breathing spontaneously (table XA) or air. ventilated mechanically during operation (table XB) acid-base values were restored virtually to the Spontaneous respiration during anaesthesia for pre-operative levels with 24 hours. orthopaedic surgery. Acid-base measurements were made in a further Hypothermia. series of patients anaesthetized for orthopaedic Values for standard bicarbonate measured before surgery (table XA). The 16 patients who were anaesthesia at an oesophageal temperature of 37 ± allowed to breathe spontaneously throughout 0-5°C, and during hypothermia to approximately operation showed a mean rise in arterial Pcc>2, from 31 °C (30-5-32°C) in 9 patients, are shown in 37 mm Hg before anaesthesia, to 54 mm Hg at the table XI. In 5 cases respiration remained spon-

Pre-anaesthetic levels (mean jzSE)of arterialpH,Pcot, and standard bicarbonate, with the changes(mean difference ±>Sis) at the end of anaesthesia and during the postoperative period. The patients breathed spontaneously during anaesthesia. The P values refer to the probability of the mean differences differing significantly from zero. NS=not significant. End-operation Pre-anaesthetic change 15 minutes pH

7-416

±0 007

Pco, (mm Hg)

371 ±0-44

Standard bicarbonate (mM/1.)

23-7 ±0-58

-0-106

±0016 P<001 + 16-9 ±2-2 P<001 -1-36 ±0-25 P<001

- 0 078 ±0011 P<001 + 6-8 ±1-7 P<001 -1-29

±0-22 P<005

Postoperative changes 1 hour - 0 039 ±0012 P<001 + 4-3 ±1-4 P<001 -0-55 ±0-17 P<005

3 hours -0031 ±0015 NS + 1-6 ±0-94 NS -0-36 ±0-27 NS

1st day -0001 ±0-009 NS +0-70 ±0-40 NS + 0-22 ±0-15 NS

500

BRITISH JOURNAL OF ANAESTHESIA TABLE XB

Pre-anaesthetic levels {mean ± SE) of arterial pH, Pco „ and standard bicarbonate, with the changes {mean difference ± SE) at the end of anaesthesia and during the postoperative period. The patients were ventilated mechanically during anaesthesia. The P values refer to the probability that the mean differences from control levels are significantly different from zero. NS=not significant.

Postoperative changes

End-operation Pre-anaesthetic change

15 minutes

1 hour

3 hours

1st day

7-434 ±0-006

+ 0-101 ±0-025 P<001

-0061 ±0006 P<001

-0043 ±0-009 P<001

- 0 024 ±0-006 P<001

+ 0008 ±0005 NS

Pco, (mm Hg)

36-4 ±0-53

-14-9 ±1-9 P<001

+4-6 ±1-0 P<001

+3-8 ±0-70 P<001

+2-0 ±0-75 P<005

+ 0-60 ±0-60 NS

Standard bicarbonate (mM/1.)

23-9 ±0-44

-0-20 ±0-25 NS

-11 ±0-35 P<001

±0-33 NS

-0-37

-0-44 ±0-26 NS

+0-14 ±0-33 NS

TABLE XI

Standard bicarbonate before anaesthesia and during hypothermia to about 31°C. Respiration was spontaneous in 5 and controlled in 4 patients at the lower temperature; the corresponding arterial Pco, levels are shown, together with the duration of anaesthesia. Duration of anaesthesia intervening between

Standard 1bicarbonate (mM/1)

Spontaneous

Controlled

37±0-5°C

31°C (approx.)

24 0 22-8 23-8 23-6

22 0 20-5 20-5 22 0

220

201

22 0 23-5 25 0 22-5

22 0 21-9

taneous down to temperatures of 32°C or lower; subsequently, every patient was ventilated while a steady level of hypothermia was maintained during surgery. For the measurements in table XI, standard bicarbonate was derived by equilibration and measurement at 38°C. The data show that a mild non-respiratory acidaemia is likely to accompany this degree of hypothermia in neurosurgical anaesthesia; however, with the exception of 2 patients who were ventilated mechanically from the start, the data are confounded with effects of concurrent or preceding spontaneous respiration per se, as has

240 221

J|

31

\_J m i l l 3L \^t

(mm Hg) measurements at approx. 31°C (minutes) 32 29 38 42 32

53 60 97 100 105

10 17 18 19

135 150 140 145

already been demonstrated. The group is small, but it is of interest that in the 5 patients still breathing spontaneously the average fall in standard bicarbonate at 31°C, 2-2 mM/1., was associated with a mean arterial Pco2 level of 35 mm Hg, while during mechanical ventilation in the other 4 patients standard bicarbonate was reduced by only 0-75 mM/1. at the lower Pcc>2 of 16 mm Hg. When the standard bicarbonate values before the start of mechanical ventilation, during hypothermia, are compared with those in the first sample withdrawn after ventilation has begun, as shown for 7 patients

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

PH

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION in table XII, it is apparent that an increase in pulmonary ventilation was associated with a significant rise in standard bicarbonate (mean difference 0-99 mM/1., SE ±0-27, P<001). The standard bicarbonate levels tended to fall subsequently as mechanical ventilation was continued. TABLE XII

Standard bicarbonate before and after instituting mechanical ventilation during hypothermia (pesophageal temp. 29-2-

31-3°C). Duration of

Spontaneous respiration 20-5 20-5

ventilation (min)

220

21-9 20-2 22 0 22-3

Controlled ventilation 220

12

22-2 22-5 21-9

22

220

23 0 22-7

30 70 45 25 65

At all temperatures down to 28°C the maximum reductions in standard bicarbonate measured in 9 patients during anaesthesia ranged from 1-5 to 3-7, mean 2-4 mM/1. In 2 patients, 15 measurements of standard bicarbonate were made at both 38°C and at the actual body temperature at the time of withdrawal of each sample (covering the range 28-4 to 36-5°C). The mean difference between the values measured at the two temperatures was 005 mM/1. (SE ±0-44), which was insignificantly different from zero. The use of two separate equilibration units, pH electrodes and meters may have contributed to the variations encountered between individual samples, but the acid-base changes associated with hypothermia were reflected similarly whether the measurements of standard bicarbonate were made at 38°C or at true body temperature (without changes to the 38°C nomogram, but with temperature corrections for pH meter, buffer, and water vapour pressure). DISCUSSION

The whole blood standard bicarbonate level is usually considered to remain constant when arterial Pa>2 changes (Astrup et al., 1960). It has been shown in dogs, however, that standard bicarbonate (or base excess) falls when arterial Pc»2

rises (Morris and Millar, 1962; Siggaard-Andersen, 1962). Reference to the literature establishes that this was described previously by Shaw and Messer (1932); in a study on dogs, these workers used the method of Van Slyke and Neill to measure the carbon dioxide content of blood samples withdrawn during acute respiratory acidosis and alkalosis. Correction of the content to that appropriate to Pco2 40 mm Hg showed that the carbon dioxide combining capacity of the blood was reduced during respiratory addosis. The basis of this effect appeared to be a migration of bicarbonate ions, when these were formed during carbon dioxide breathing, from the well-buffered blood to the less-buffered tissues; the changes were completed within 5 minutes. Shaw and Messer (1930) had calculated previously that the carbon dioxide combining capacity of the tissue fluid was only 78 per cent of that of whole blood. In respiratory acidosis, therefore, bicarbonate leaves the blood and is replaced by chloride or other anions, a process which is similar to that occurring when bicarbonate is formed in the red cells. The rapid change, in the reverse direction, when the excess carbon dioxide is eliminated, suggests that the nonrespiratory acidaemia caused by this loss of bicarbonate is an intrinsic accompaniment to the acute respiratory acidosis. The effect should be distinguished from the metabolic acidosis, associated with release of catecholamines, which has been measured during and following severe respiratory acidosis (Morris and Millar, 1962). Shaw and Messer implied in 1932, therefore, that the log Pcoa/pH equilibration line for the whole body must differ from that for blood. This has been confirmed in recent studies (Siggaard-Andersen, 1962; Morris and Millar, 1962; Cunningham, Lloyd and Michel, 1962; Norman and Linden, 1965). From his own experiments, SiggaardAndersen (1962) established an appropriate whole body log Pco2/pH equilibration line, the slope of which lay intermediate between the (relatively steeper) line for blood and the (flatter) line for plasma. The line for the whole organism corresponded to that for blood with a haemoglobin concentration of 5-7 g/100 ml. The reduction in base excess, per litre of blood per mm Hg rise in Pco2, was approximately 0-068 mM (roughly equivalent to a 0-052 mM fall in standard bicarbonate). Applying this data to the measurements made

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Standard bicarbonate (mM/1.)

501

502

amounts during the initial period of spontaneous respiration, and seldom exceeded 250 ml in any patient during the surgical procedures. In our studies, a relative non-respiratory alkalaemia could not be induced by pulmonary overventilation when pre-existing arterial Pco2 and standard bicarbonate levels were near-normal. Shaw and Messer (1932), in their experiments on dogs, found that changes in the carbon dioxide capadty of whole blood were small and variable when arterial Pc02 was lowered markedly; they concluded that at tensions below normal the carbon dioxide absorption curves of tissue fluids and blood were similar. This conclusion is not challenged decisively by the experiments of SiggaardAndersen (1962), although recent studies by Norman and Linden (1965) imply that measurement of standard bicarbonate during respiratory alkalosis falsely demonstrates an assodated nonrespiratory alkalaemia (due to the different log Pcoe/pH equilibration lines for the whole organism and blood). This has not been demonstrable in the present studies. The non-respiratory addaemia which occurred in patients breathing spontaneously after induction of anaesthesia, was increased subsequently when arterial PcO2 was lowered by gross pulmonary overventilation. However, the average reduction in standard bicarbonate noted in a group of neurosurgical patients, by 1-7 mM/1. from the preanaesthetic level after 3 hours of overventilation, indicates that extreme lowering of arterial Pcos in anaesthetized patients produces little disturbance in the whole blood non-respiratory component of add-base balance. This agrees with previous studies (Papadopoulos and Keats, 1959; Markello, Cutter and King, 1963). Robinson (1961), who measured the standard bicarbonate of separated plasma, reported a mild metabolic addosis during passive overventilation in consdous or anaesthetized subjects. The changes in the non-respiratory component during respiratory alkalosis in our neurosurgical patients partly reflect the responses to the preceding period of spontaneous respiration after induction of anaesthesia. When mechanical ventilation was started immediately following endotracheal intubation, there was little change in standard bicarbonate during the subsequent 45 minutes of moderate respiratory alkalosis, at Pco2

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

in the present report, for example in the neurosurgical patients breathing spontaneously with moderate respiratory addosis, perhaps 0-63 mM of the 0-82 mM/1. reduction in standard bicarbonate could be attributable solely to the difference between the whole organism and blood equilibration lines. In this case, therefore, as with the other measurements reported here, a portion of the reductions in standard bicarbonate which accompanied acute rises in arterial Pco2 cannot be attributed to a true metabolic addosis, which implies production offixedadds. The present study shows that a relative nonrespiratory addaemia occurs under the following conditions related to general anaesthesia: during spontaneous respiration associated with rises in arterial PC02; accompanying subsequent pulmonary overventilation during neurosurgery; during hypothermia by surface cooling; when spontaneous respiration is resumed after periods of mechanical overventilation; and in the early postoperative period. Holaday, Ma and Papper (1957) described a metabolic addosis in association with respiratory depression in anaesthetized patients. They measured an average fall in buffer base of 3-7 m.equiv/1., with a mean rise in arterial Pco2 of 18 mm Hg. This degree of non-respiratory addosis was more pronounced than in the present studies, and may have been influenced by the surgical procedures being undertaken and their duration. Since reductions in standard bicarbonate occurred whether arterial Pcx>2 was raised during respiratory depression or by passive ventilation with carbon dioxide, active muscular effort was not an essential factor in causing the non-respiratory addaemia. The changes in standard bicarbonate measured here reflect realistic clinical conditions, and could have been influenced by a number of uncontrolled factors. Requirements of transfused blood were small, and replacement was immediate. Variations in arterial pressure were commonly induced in the neurosurgical patients; the levels maintained were consistently below 110 mm Hg, frequently about 90 mm Hg, and rarely below 60 mm Hg. At least 30 per cent oxygen was administered at all times; this was probably adequate to prevent relative hypoxaemia in the great majority of the patients studied. Slow intravenous infusions of 5 per cent dextrose in water were given in insignificant

BRITISH JOURNAL OF ANAESTHESIA

503

ACID-BASE CHANGES ASSOCIATED WITH VENTILATION

standard bicarbonate which occurred consistently at this time was associated with rises in arterial Pc02, these were not necessarily to above the preanaesthetic level. In conscious subjects, Brown et al. (1949) noted a fall in whole blood carbon dioxide capacity shortly after the resumption of spontaneous respiration following prolonged pulmonary overventilation. In close agreement with Hobsley (1963), nonrespiratory acidosis in the early postoperative period appears to be commonly associated with relative increases in arterial Pco2 occurring at this time, or consequent to previously existing respiratory acidosis during anaesthesia. ACKNOWLEDGMENTS

The co-operation of the neurosurgical staff, notably Mr. W. S. Lewin and Mr. J. Gleave, and of the orthopaedic surgeons, is gratefully acknowledged. These studies were partly supported by a grant for equipment and technical assistance from the Medical Research Council. Dr. Marshall held an Elmore Studentship of the University of Cambridge. REFERENCES

Astrup, P., Jorgensen, K., Siggaard-Andersen, O., and Engel, K. (I960). The acid-base metabolism: a new approach. Lancet) 1,1035. Brown, E. B. jr., Campbell, S., Elam, J. O., Gollan, F., Hemingway, A., and Visscher, M. B. (1949). Electrolyte changes with chronic passive hyperventilation in man. J. appl. Physiol., 1,848. Cunningham, D. J. C , Lloyd, B. B., and Michel, C. C. (1962). Acid-base changes in the blood during hypercapnia and hypocapnia in normal man. J. Physiol. (Lond.), 161,26P. Cutter, J. A., and King, B. D. (1961). Spontaneous readjustments in acid-base balance at the termination of prolonged hyperventilation. Anesthesiology, 22,130. Dobkin, A. B. (1959). The effect of Fluothane on acidbase balance. Artesthesiology, 20,10. Hobsley, M. (1963). Respiratory disturbances caused by general surgical operations. Arm. roy. Coll. Surg. Engl, 33,105. Holaday, D. A., Ma, D., and Papper, E. M. (1957). The immediate effects of respiratory depression on acidbase balance in anesthetized man J. din. Invest., 36,1121. Holmdahl, M. H., and Payne, J. P. (1960). Acid-base changes under halothane, nitrous oxide and oxygen anaesthesia during spontaneous respiration. Ada anaesth. scand., 4,173.

Markello, R., Cutter, J. A., and King, B. D. (1963). Hyperventilation studies during nitrous oxidenarcotic-relaxant anesthesia. Artesthesiology, 24, 225. Marshall, B. E. (1964). A method of arterial puncture. Lancet, 2,40. Millar, R. A. (1965). Factors influencing the occurrence of post-operative hypoxaemia. Anaesthesia. (In press.)

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

levels of 18 to 29 mm Hg. The findings were similar in the orthopaedic patients mechanically ventilated to the same degree. The prevention, or delayed onset, of nonrespiratory acidaemia by early pulmonary overventilation excludes a direct metabolic action of halothane (or nitrous oxide) as an important factor in the changes measured. Previous reports have not shown any metabolic disturbance during controlled halothane anaesthesia (Dobkin, 1959), although they have also not demonstrated any fall in standard bicarbonate during spontaneous respiration of nitrous oxide, oxygen and halothane mixtures (Holmdahl and Payne, 1960). It has been reported that certain hyperosmotic solutions (sodium chloride, mannitol), but not urea, induce a metabolic addosis (Winters et al., 1964). In our studies, intravenous urea appeared to oppose the non-respiratory acidaemia of respiratory depression and/or pulmonary overventilation. No adverse influence on neurosurgical operating conditions could be ascribed to the early period of relative respiratory depression following induction of anaesthesia, provided that pulmonary overventilation was started 10-15 minutes before the start of the neurosurgical procedure. Harmful sequelae were not noted as a result of maintaining arterial Pco2 below 15 mm Hg in many patients, but there was also no convincing evidence that neurosurgical operating conditions were improved thereby over those existing at Pco2 levels of 25-30 mm Hg. The maximum reductions in standard bicarbonate accompanying surface cooling to 30cC, averaging 2-4 mM/1., were relatively small, and partly include effects associated with previous spontaneous respiration. Similar results, in the dog, were described by Severinghaus, Stupfel and Bradley (1957). The present data suggest that in the cooled patient mechanical overventilation from the start of anaesthesia minimizes changes in the non-respiratory component of acid-base balance. In 9 of 17 neurosurgical patients studied, arterial Pco2 was lower on resuming spontaneous respiration after surgery than in the conscious state before anaesthesia, confirming that the initiation of natural respiration at the end of surgical operations depends on many factors other than arterial Pco2 (Cutter and King, 1961; Utting and Gray, 1962). It was interesting, also, that while the fall in

504

BRITISH JOURNAL OF ANAESTHESIA

Downloaded from http://bja.oxfordjournals.org/ at Harvard University on July 6, 2015

Moms, M. E., and Millar, R. A. (1962). Blood pH/plasma LES MODIFICATIONS DE L'EQULLIBRE catecholamine relationships: respiratory acidosis. ACIDO-BASIQUE ASSOCIEES A LA VENTILABrit.J. Anacsth., 34,672. TION SPONTANEE ET CONTROLEE PENDANT L'ANESTHESIE Norman, J., and Linden, R. J. (1965). Hyperventilatdon and acid-base balance. Brit. J. Anaesth., 37, 290. SOMMAIRE Papadopoulos, C. N., and Keats, A. S. (1959). The Le pH artiriel, la Pco, et le bicarbonate standard du metabolic acidosis of hyperventilation produced by sang total ont ixi mesures chez des malades chirurgicontrolled respiration. Anesthesiology, 20, 156. caux avant, pendant et apres la ventilation pulmonaire Robinson, J. S. (1961). Some biochemical effects of spontan6e ou contr61ee avec de l'oxyde nitreux, de l'oxygene et de l'halothane. Les augmentations relatives passive hyperventilation. Brit. J. Anaesth., 33, 69. de la Pco, artirielle s'accompagnaient de inductions Scholander, P. F. (1947). Analyzer for accurate estima- significatives du bicarbonate standard, indices d'une tion of respiratory gases in one-half cubic centimeter acidose non respiratoire. Cet effet peut gtre attribuable samples. J. biol. Chan., 167,235. en partie aux differences entre les lignes d'equilibration Severinghaus, J. W., Stupfel, M. A., and Bradley, A. F. log PcOj/pH pour le sang et pour tout l'organisme. (1957). Alveolar dead space and arterial to end-tidal De petites reductions du bicarbonate standard se sont carbon dioxide differences during hypothermia in produites pendant plusieurs heures de surventilation pulmonaire chez les malades chirurgicaux et pendant dog and man. J. appl. Physiol., 10,349. l'hypothermie induite. Les resultats suggerent que les Shaw, L. A., and Messer, A. C. (1930). The carbon modifications de la composante non respiratoire de dioxide capacity of the body and the rate at which l'equilibre acido-basique peuvent etre diminuees en the body comes into equilibrium with changes in the instituant une ventilation pulmonaire efficiente imalveolar carbon dioxide tension. Amer. J. Phytiol., mddiatement apres l'induction de l'anesthesie. 93,422. (1932). The transfer of bicarbonate between VERANDERUNGEN DES SAURE-BASENthe blood and tissues caused by alterations of the carbon dioxide concentration in the lungs. Amer. J. HAUSHALTES WAHREND DER NARKOSE MIT SPONTANER UND KONTROLLIERTER Physiol., 100,122. ATMUNG Siggaard-Andersen, O. (1962). Acute experimental acidbase disturbances in dogs: an investigation of the ZUSAMMENFASSUNG acid-base and electrolyte content of blood and urine. Vor, wahrend und nach spontaner oder kiinstiicher Scand.J. din. Lab. Invest., Suppl. 66. Atmung mit Lachgas, Sauerstoff und Halothan bei (1964). The Acid-base Statiu of the Blood. Copen- chirurgischen Patienten wurde arterielle pH-, Pco,hagen: Munksgaard (in U.S.A., Williams and und Standard-Bikarbonatbestimmungen im Vollblut Wilkins). vorgenommen. Relative Anstiege des arteriellen Pco, Engel, K., Jorgensen, K., and Astrup, P. (1960). waren gefolgt von einer signifikantcn Verringerung des A micromethod for determination of pH, carbon Standardbikarbonats, was fur eine nichtrespiratorische dioxide tension, base excess and standard bicarbo- Azidose spricht. Diese Wirkung kann teilweise dem nate in capillary blood. Scand. J. din. Lab. Invest., Unterschied zwischen dem Logarithmus der Pco,/pHGleichgewichtsgeraden fiir das Blut und den Gesamt12,172. organismus zugeschrieben werden. Geringe Abfalle des Utting, J. E., and Gray, T. C. (1962). The initiation of Standardbikarbonats traten wahrend mehrcrer Stunden respiration after anaesthesia accompanied by passive nach Hyperventilation bei neuro-chirurgischen pulmonary hyperventilation. Brit. J. Anaesth., 34, Patienten und wahrend einer induzierten Hypotomie 785. auf. Die Ergebnisse sprechen dafur, daB die VeranderWinters, R. W., Scaglione, P. R., Nahas, G. G., and ungen der nicht respiratorischen Komponente des Verosky, M. (1964). The mechanism of acidosis Saure-Basen-Gleichgewichtes durch Hcrstellung einer produced by hyperosmotic infusions. J. din. ausreichenden Beatmung unmittelbar nach Einleitung der Narkosc gering gehalten werden konnen. Invest., 43, 647.