Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from

IO0

POSTGRAD. MED. J., 1961, 37, 10 BLOOD GAS MEASUREMENTS IN CLINICAL PRACTICE E. J. M. CAMPBELL, M.D., M.R.C.P. Assistant, The Professorial Medical Unit, The Middlesex Hospital, London, W.I

UNTIL recent years methods of blood gas measure- not be drawn by suction because this is liable to ment have been difficult and they were therefore introduce bubbles of air. The period of sampling not widely available. Their possible value in should be between one and two minutes and the routine clinical practice was consequently largely patient should be relaxed and breathing naturally unknown and unexplored, and many clinicians to ensure that the sample reflects steady-state con- have forgotten their basic knowledge of oxygen and ditions as far as possible. Temporary fluctuations carbon dioxide physiology. in the breathing may otherwise cause the state of The sequence is now in reverse. There are easy the arterial blood gases to be unrepresentative. methods which could be widely available. After withdrawal, the site of puncture should be Clinicians are learning their value in practice and compressed sufficiently firmly to obliterate the in the process are relearning respiratoryphysiology. distal pulse for at least five minutes. The aim of this article is to catalyse this chain Discomfort or complications following com- reaction by indicating the most practicable methods petent arterial sampling are unrecorded. I have and by describing the use of the measurements in been told of damage to the median nerve in the practice. course of a ' search' for the brachial artery by an There is now no need for physicians to rely inexperienced person, and I have known an alarm- entirely on clinical judgment in assessing, for ing haematoma to develop in a confused patientcopyright. example, methods of treating respiratory de- who was allowed to flex the elbow repeatedly while pression in cases of poisoning. They would be an indwelling arterial needle was in place. There unhappy without measurements of blood sugar in can be few clinical procedures that are free from diabetic coma, urea in renal failure or sodium in such hazards. I have not heard or read of any adrenal failure. They are now entitled to require other serious complications following arterial punc- knowledge of the Pco2 in respiratory failure. ture in the total experience of several workers to over Blood Collection for Gas Analysis amounting 25,000 punctures. and of Arterial Blood This presents the first difficulty which has dis- Storage Transport http://pmj.bmj.com/ tinguished blood gas measurements from other Samples biochemical investigations. The exchanges of The overriding consideration in the care of oxygen and carbon dioxide between blood and blood samples is that they be kept out of contact tissues prevent samples taken from the antecubital with the air. Measurements of Po2, Pco2 and pH veins in the usual manner from having much value. must usually be made within a few minutes of Venous blood can be used for the estimation of collection so the blood is most conveniently kept carbon dioxide content (see below) and even, in the syringe. Neither cooling nor the addition with special precautions, for the estimation of of fluoride is satisfactory in preventing the meta- on October 3, 2021 by guest. Protected carbon dioxide tension (Pco2) and pH (Brooks bolism of the cells. Early analysis of blood kept and Wynn, I959), but arterial blood is preferable in the syringe is the best arrangement. for these and, of course, essential for meaningful These considerations do not apply to blood to be estimates of oxygen. Arterial puncture continues analysed for carbon dioxide content which is not to be viewed askance partly because of alleged seriously affected by the metabolism of the cells difficulty and partly because of fear of damage. nor by minor degrees of contact with the air. Like all clinical procedures, arterial puncture is Blood for carbon dioxide content can be trans- best learned by example and after half a dozen ferred to tubes (by 'filling from the bottom'), successful attempts one is almost as good as the which should be completely filled and then stop- most expert. A syringe lubricated by heparin pered. Blood for oxygen content or saturation (5,000 units/ml.) should be used with heparin presents an intermediate problem. Analysis can be filling the deadspace of nozzle and needle. Mercury deferred up to an hour or so (depending on the is unnecessary. The blood should be allowed to accuracy required) if the blood is cooled. Such fill the syringe under its own pressure; it should blood is preferably and conveniently kept in the Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from January 1961 CAMPBELL: Blood Gas Measurements in Clinical Practice II syringe. If it is transferred to another container routine laboratories; although with improve- greater care is required than for carbon dioxide ments in pH meters this statement is becoming content and the transfer should be anaerobic. less true. With the development ofmore stable and reliable Methods pH meters two further methods of estimating Under this heading I propose to review the blood Pco2 have been introduced. First, the practicability of the various available methods interpolation method of Astrup (1956) and, bearing in mind cost, technical skill, time and secondly, the titration electrode method of reliability. Severinghaus and Bradley (1958). These have brought Pco2 measurements within the reach of Blood Oxygen routine laboratories, although the skill required is (i) Content and Saturation. For accuracy, still above that demanded by most routine chemical reliability and cheapness the Van Slyke method procedures. (Van Slyke and Neill, I924) remains supreme. The Rebreathing Method for Estimating Mixed Unfortunately, it requires more skill and time Venous and Arterial Carbon Dioxide Tension. than most routine biochemical techniques, so other Collier's (1956) introduction of the rebreathing methods should be considered if the number of method for estimating blood carbon dioxide tension analyses is large or the analyst's time is precious. is the greatest practical advance in clinical respira- For large numbers one of the many spectrophoto- tory physiology of the last 20 or 30 years. This metric or reflectrometric methods is probably best method abolishes all analytical problems of Pco2 and can be set up at reasonable cost (Refsum and estimation. Using a rapid continuous carbon Hisdal, 1958; Verel, Saynor and Kesteven, 1960). dioxide gas analyser, the method is extremely (2) Tension. Until recently the only practicable simple. Without this instrument it is still, given method for measuring blood oxygen tension was slightly greater understanding, possible to apply the microtonometric or ' bubble ' method of Riley the method using simple and cheap equiprient (Riley, Campbell and Shepard, I957). The (Campbell and Howell, 1960). The whole pro- apparatus is inexpensive, but each analysis re- cedure, including preparation and analysis occupiescopyright. quires about 40 minutes of skill and concentration. io minutes. It should soon be replaced by the polarographic methods (e.g. Severinghaus and Bradley, 1958; Practical Conclusions Bishop, I960), which promise to be sufficiently reliable and for routine All general hospitals should today be able to inexpensive, rapid use. provide facilities for the following methods: External Oximetry (I) Arterial oxygen saturation by the Van Slyke Spectrophotometric techniques for the estima- method. tion of the arterial saturation in vivo (2) Mixed venous carbon dioxide tension by the oxygen http://pmj.bmj.com/ (Nilsson, 1960) are unfortunately not very accurate rebreathing method. with respect to absolute values, but can be used (3) Blood carbon dioxide content by the Van to follow changes (see Campbell and Dickinson, Slyke method. 1960, p. I20). The instruments are still moderately With these and an understanding of the under- and lying physiology the clinician can evaluate any expensive temperamental. disturbance of blood gas regulation. With par- Carbon Dioxide ticular reference to respiratory disease, the of a in (I) Content. As in the case of oxygen, the Van possession recording spirometer addition on October 3, 2021 by guest. Protected Slyke method is the most accurate and the to the above enables the vast majority of respira- cheapest. The analysis for carbon dioxide content tory problems to be adequately assessed. is less laborious than that for oxygen, so the Supplementation of these four essentials should demands on time and skill are not so great. Never- be governed by the particular interests to be theless, automatic chemical analysers promise to served. replace the Van Slyke method in busy laboratories. (2) Tension. Until recently there have been only Physiological Principles Underlying Clinical two practicable methods for the measurement of Interpretation blood carbon dioxide tension. First, the ' direct' Like most laboratory measurements, values for microtonometric bubble method of Riley, and blood gas concentrations cannot be interpreted secondly, the 'indirect' method in which Pco2 clinically without appreciation of the biological is calculated from the pH and carbon dioxide factors affecting them. An adequate exposition content. Both methods require time, skill and of the physiology of the blood gases is far beyond attention far above that generally available in the scope of this article, but there are certain im- Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from 12 POSTGRADUATE MEDICAL JOURNAL January 1961

u o50so /

030I0 ' a.20 10

0 20 40 60 80 100 120 140 OXYGEN PRESSURE mm.H9

FIG. i.-Oxygen dissociation curve of blood. copyright. portant principles which deserve mention because increase the oxygen tension difference between the of their practical significance. alveolar air and the pulmonary capillary blood and thus to reduce arterial oxygen tension. When this Oxygen falls below 60 mm. Hg, the saturation falls rapidly. The relationship between oxygen saturation and Hence, in a subject whose resting saturation is tension is given by the dissociation curve (Fig. i). over go%, exercise can be used to stress respiratory

The normal arterial tension is about go mm. Hg, function and the amount of exercise (as judged by http://pmj.bmj.com/ giving a saturation of about 97%. The shape of the the oxygen uptake) required to cause the arterial dissociation curve is such that the saturation of saturation to fall below 85-90% provides a useful arterial blood does not fall below 90% until the method of detecting and assessing many types of tension falls below 60 mm. Hg, whereas further disorder. In lung diseases it is an excellent sub- reduction in tension causes increasingly severe stitute for measurements of diffusing capacity to desaturation. The practical implications of this which it is in some ways superior and which it relationship are: firstly, that respiratory function can be used to calculate (Shepard, 1958). has to be seriously disordered before measure- on October 3, 2021 by guest. Protected ments of saturation become unequivocally ab- Carbon Dioxide normal; secondly, when the saturation falls below Interpretation of changes in blood carbon di- 90% it becomes so sensitive to small changes in oxide of two relation- tension that variations in breathing and bodily depends upon appreciation activity cause the saturation to vary considerably, ships: making isolated measurements unreliable indica- (I) Arterial carbon dioxide tension is governed tions of the average level. These facts partly by the ratio of carbon dioxide production to explain the eagerness with which methods of con- effective, i.e. alveolar, ventilation: tinuous or tension determinations Arterial Pco, = frequent oxygen CO2 are being developed. productiontion barometric Exercise increases oxygen consumption, in- alveolara polrventilation X pressure creases the flow of blood through the lungs and If alveolar ventilation is halved, arterial Pco2 is causes the blood entering the lungs to have a doubled and vice versa. lower saturation. These changes all tend to (2) The proportion of carbon dioxide in the Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from January 1961 CAMPBELL: Blood Gras Measurements in Clinical Practice 13 0 3 ~~G0IZ 0 5 ... ?· ?,.-,· atl... : ...... -.,,v... 45 '-I.·..:',-;.' .'.':.~. · ·'. ....,· 4mfiH. 02.!'.!"'.1,' .'' j. . ...

'~l .. ".

', ' ·.·-ru · .' " , '. .' 'Z,'.

· . ' ·~~~~· .:'"i., . ·· '"· .,.pH .UNIS :.'/.'

,.4 69 . 70.. 7 7 7.3 7.4. :7.· . 76 .A779'7..pHNTU ' , .:".~'..~%il '.", copyright. FIG. 2.-Bicarbonate/pH diagram. The numbers refer to the cases described in the text. form of dissolved carbon dioxide, the proportion Arterial Oxygen Measurements in the form of bicarbonate and the pH are inter- (None of the following remarks can necessarily be dependent. This interdependence is expressed in applied to cases with intracardiac shunts which the Henderson-Hasselbalch equation: present special problems.) It is still uncertain how = bicarbonate concentration valuable arterial measurements are or H= p l oxygen pH pk' d- log. CO2 concentration would be outside units. special cardiopulmonary http://pmj.bmj.com/ As the concentration of dissolved carbon dioxide They have been essential to the elucidation of the is a constant function of the partial pressure of functional disturbance in many diseases, but carbon dioxide, the equation can be written: evidence is still lacking about their value in the bicarbonate concentration of the individual This view pHp=- pk'' +-+ log.l, management patient. 0.03 x Pco2 can be elaborated under three headings. This equation enables the third variable to be (I) Diagnosis. Cases in which a measurement calculated from knowledge of the other two. This of arterial oxygen saturation is essential to the is valuable technically and also important clinically diagnosis are very uncommon. The use of such on October 3, 2021 by guest. Protected because it shows that the full effects of a disturb- manoeuvres as exercise and/or varying the oxygen ance of any one of the three depends upon the concentration of the inspired air to supplement other two. This is illustrated by Cases 3, 7, 8 and clinical and radiological evidence will enable a io described later. diagnosis to be reached on nearly all occasions. There are several ways of expressing the Cases I and 2 described below may seem to be Henderson-Hasselbalch relationship graphically, exceptions to this general rule, but, in fact, the ofwhich one ofthe most popular is the bicarbonate: diagnosis in both was strongly suspected and pH diagram (Fig. 2). Its use is indicated by the could have been made by other methods. The case histories. diagnosis in the patient with polycythaemia (Case 5) depended on the estimation of Pco2 rather than Clinical Value on the oxygen saturation. Some overall judgments are required before (2) Prognosis. Evidence is still scanty that describing the illustrative cases if these are to be measurements of arterial oxygenation add much seen in true perspective. to the assessment of prognosis in either acute or Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from I14 POSTGRADUATE MEDICAL JOURNAL January I96 chronic respiratory disease (Cotes, I960). Such CASE I.-Differentiation of Cardiac and Respiratory evidence must await the introduction of Cause of Breathlessness. A woman of 25 had had severe adequate kyphoscoliosis since childhood. In the preceding few methods for assessing the variability of arterial months she had become increasingly breathless and oxygenation. blue on exertion. The heart was enlarged and there Treatment. If anoxic anoxia is was a widely split second sound in the pulmonary area. (3) suspected The arterial saturation at rest breathing air was 72%, during an acute illness, the inspired air can and and while breathing 00o% oxygen it was 86%. The should be enriched with oxygen. If anaemic or Pco, was normal. stagnant anoxia is suspected during an acute illness Comment the inspired air should be Ioo% oxygen. These These figures illustrate the difficulty of inter- decisions can be made clinically without recourse pretation of changes in saturation in this range. to measurements of blood oxygenation. Para- The patient's colour changed unequivocally from doxically, the oxygen concentration to give a blue to -pink on receiving oxygen, an observation patient is better judged from the arterial or mixed which had led to the belief that her anoxia was venous Pco2 (see Case 4 below). In patients pulmonary in origin. In fact, she had a true suffering from chronic anoxia the value of oxygen venous-arterial shunt and the findings are explic- can be adequately judged on clinical grounds, able as follows: firstly, at a saturation of 72% particularly by observing the effects on exercise cyanosis is easily seen, whereas at 86% it may be tolerance. undetectable. Secondly, the data breathing air suggested that there was a shunt equal to 50 to 70% Blood Carbon Dioxide Measurements of the cardiac output. This implies that 30 to 50% Measurements of blood carbon dioxide content of the blood traversed the lungs, where, while are, of course, essential to the assessment of breathing Ioo% oxygen, it could take up z ml./ metabolic (i.e. non-respiratory) disorders of acid Ioo ml. extra oxygen in solution. This would be base metabolism (e.g. Cases 8, 0o and i ) and are sufficient to account for an increase in arterial for the full assessment of dis- saturation of about Io%. The remaining 4% required respiratory would be attributable to the

turbances Cases and readily overcoming copyright. (e.g. 3 7). 'distribution' effect of her chest deformity, Estimations of arterial or mixed venous Pco2 which, while breathing air, probably caused the are of great clinical value and provide information blood flowing through parts of her lungs to be which often cannot be obtained by other methods. inadequately oxygenated. The management of Cases 6, 9 and i was par- None of this difficulty would have arisen if 30% ticularly helped by having such facilities immedi- oxygen had been used instead of Ioo% (see Case 2). ately and continuously available. Subsequently this patient was found to have an There is no substitute for measurements atrial septal defect and progressive pulmonary of Pco2 when the adequacy of ventilation is in hypertension, causing a ' reversed ' shunt. doubt. Anyone who attempts to forecast the http://pmj.bmj.com/ adequacy of ventilation on the basis of clinical CASE 2.-Acute Cerebral Disease and a Localized evidence finds that he is very often wrong. Further- Pulmonary Lesion. A woman of 53 developed an acute there are other measurements neurological illness with evidence of a focal cerebral more, although lesion and meningeal irritation. The cerebro-spinal which may help, none of them is as simple as the fluid contained many polymorphs but no organisms. rebreathing method for estimating Pco2. The differential diagnosis included meningitis, brain In metabolic disturbances of ion abscess and cerebral tumour. X-ray of the chest hydrogen regu- showed a lobulated shadow in the left lower lobe with lation the Pco2 is valuable in combination with the prominent vascular connection. Arterial blood was on October 3, 2021 by guest. Protected blood carbon dioxide content in providing an sampled while the patient was breathing 30% oxygen estimate of the pH. This topic is further discussed and was found to have an oxygen saturation of 85%. elsewhere in this issue (Ashby and Campbell, Comment 196I). Arterial unsaturation while breathing 30% oxygen can only be due to anatomical venous- Illustrative Cases arterial shunt. (Desaturation due to generalized It has been easy to recollect patients in whom lung disease is corrected by 30% oxygen.) This measurements of Pco2 were of great value and finding provided the lynch-pin of the diagnosis of radically affected the management. It has been pulmonary arterio-venous fistula, a condition difficult to think of cases in which measurements of known to predispose to meningo-encephalitis. In- arterial oxygen saturation were important, although vestigation for cerebral tumour was therefore not many could be found in which the measurement pursued. The patient recovered. was interesting and added to rather than changed CASE 3.-Combined Respiratory and Metabolic Dis- the picture. order in Emergency Surgery. A neurotic woman of 38 Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from January I96i CAMPBELL: Blood Gas Measurements in Clinical Practice I5 was admitted with suspected peritonitis of four days' may well have been attributable to under- duration and physical signs of congestion or pneumonia of the lower lobes. There was clinical evidence of ventilation. dehydration, sodium depletion, hyperventilation and arterial desaturation. The arterial blood (Fig. 2, CASE 4.-Management of Anoxia in Respiratory point 31) Pco2 was 26 mm.Hg; plasma carbon dioxide Failure. A man of 55 with severe chronic obstructive content was I9.i mM./l.; pH 7.40; oxygen saturation lung disease was admitted with pneumonia. He was was 87%. slightly confused and there was central cyanosis. The Exploratory laparotomy showed retroperitoneal arterial Pco2 was 8I mm.Hg and the oxygen saturation haemorrhages and cedema but no ruptured viscus. The was 44%. He was given 34% oxygen to breathe which operation, which took one hour, had just finished increased the saturation to 82.5% but caused him to when the heart stopped. After cardiac massage through become semi-comatose due to underventilation and a thoracotomy, the heart started beating normally. aggravation of the respiratory acidosis-the Pco2 rose Arterial blood taken at this time had a pH of 7.00 to 99 mm.Hg. Reduction of the inspired oxygen con- and a Pco2 of 90 mm.Hg (Fig. 2, point 32). Shortly centration to 28% improved his mental state and the afterwards the heart stopped again and she died. Pco2 fell to 86.5 mm.Hg. with restoration of some Autopsy showed the intestine to contain large quan- anoxic drive to his ventilation as the saturation fell tities of broken glass which she had apparently to 66%. The inspired oxygen concentration was sub- swallowed. sequently gradually increased without ill effect. Comment Comment The pre-operative findings for pH, Pco2 and This case illustrates the paradox that knowledge carbon dioxide content were similar to those in the of the Pco2 is more helpful than knowledge of the patient with salicylate poisoning (Case io) in that oxygen saturation in managing the treatment of there was a combined metabolic acidosis and anoxia in such a patient. He clearly had to receive respiratory alkalosis with a normal pH. The oxygen. The question was: how much? The metabolic acidosis was probably in part due to the answer is: as much as can be tolerated without electrolyte disturbance caused by her abdominal allowing severe respiratory depression as indicated condition, but may have been partly due to the by a serious increase in the Pco2. respiratory alkalosis. That the respiratory alkalosis was not a of the metabolic CASE 5.-Polycythcemia. An obese man of 46 was copyright. simply consequence found to have polycythaemia (Hb. i8 g./loo ml.). His acidosis, but was probably due to the anoxia caused dusky colour was attributed to a combination of the by the pneumonia, is indicated by the high normal polycythaemia and a colourful racial origin. After a pH. time, however, the presence of undue cyanosis was appreciated and the arterial oxygen saturation was The combination of a low arterial Pco2 and a found to be about 50%. A cardiac cause was excluded low arterial oxygen saturation is the pattern of and he was referred for pulmonary function studies as blood gas changes seen in conditions causing mal- a suspected case of chronic lung disease. His venti- distribution of pulmonary blood flow. The data latory capacity as judged by spirometry was found to indicate that over 40% ofthe blood flowingthrough be normal but the arterial Pco2 was 85 mm.Hg. this patient's lungs was ' shunted ', i.e. exchanged Comment http://pmj.bmj.com/ no oxygen or carbon dioxide. The consequent The combination of underventilation in the arterial hypoxia causes hyperventilation which presence of normal ventilatory capacity is diag- blows off carbon dioxide from the relatively nor- nostic ofthe syndrome of primary underventilation mal parts of the lungs, but cannot correct the (Burwell, Robin, Whaley and Bickelman, 1956), hypoxia. which is usually accompanied by polycythamia The and obesity. (I am indebted to Dr. J. B. L. Howell combination of a high alveolar ventilation for the description of this case.) and maldistribution of pulmonary blood flow on October 3, 2021 by guest. Protected almost certainly means that this patient required a total ventilation. This CASE 6.-Failure to Breathe Post-operatively. A very large she obviously woman of 33 was operated on for a subphrenic abscess. did not receive during the operation because an The exposure was difficult and the anaesthetist main- arterial Pco2 of 90 mm. Hg can only be reached by tained controlled ventilation with the patient fully fairly prolonged underventilation. The effects of paralysed by curare. After operation the patient would such a acidosis on the not breathe. The possibility of carbon dioxide retention respiratory pH was greatly was considered and the artificial ventilation given aggravated by the pre-existing metabolic acidosis more vigorously but the blood pressure fell. The (see Fig. 2). mixed venous Pco2 was estimated by the rebreathing It be method and found to be about I50 mm.Hg. Over- may suggested that this patient presented ventilation was therefore given with greater confidence a hopeless case, but if such patients are to be saved and the Pco2 maintained at about 70 mm.Hg. The greater attention must be paid to respiratory func- patient began to breathe and recovered completely. tion and less absolute reliance placed on clinical Comment judgment. The anaesthetist thought he had over- The patient's toxaemia and the various drugs ventilated this patient. In fact, the cardiac arrest used in the course of anaesthesia had made any Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from 16 POSTGRADUATE MEDICAL JOURNAL January I96I reliance on 'signs of carbon dioxide retention' measured so that the effects of this drug could be The anaesthetist did not know documented. The Pco2 was found to be 32 mm.Hg. unjustifiable. No drug was given and the patient woke up eight whether the patient was being over- or under- hours later. ventilated. Comment Anybody who measures Pco2 in 'complicated' Now that easy methods of Pco2 estimation are cases who are ' not doing well' after operation can available, comparisons of methods of treating report similar experiences (see also Cases 3 and 7). respiratory depression should become more con- CASE 7.-Distress After Thoracotomy, Cardiac By- vincing. pass and Aortic Surgery. A man of 21 had his aortic stenosis relieved with the aid of hypothermia and CASE io.-Salicylate Poisoning. A woman of 52 cardiac by-pass. In the immediate post-operative swallowed 25 g. of aspirin and was admitted to hos- period his condition was satisfactory, but after three pital two days later with oliguria. There was obvious hours it began to deteriorate. He was distressed and hyperventilation. The plasma bicarbonate concentra- breathless with rapid shallow breathing and had an tion was 13 mEq/l. and the arterial Pco2 17 mm.Hg, increasing tachycardia. The mixed venous Pco2 was giving a calculated pH of 7.46. The blood urea was estimated by the rebreathing method and found to be 255 mg./Ioo ml. and the plasma salicylate concentration 68 mm.Hg, corresponding to an arterial Pco2 of was 55 mg./xoo ml. A diuresis began shortly after about 62. Assisted ventilation was not tolerated so admission. She was treated conservatively with water tracheostomy was performed. His condition imme- and electrolyte replenishment governed by her urinary diately improved and his Pco2 fell to about 40. losses and weight changes. Her blood urea, bicarbonate, Pco2 and pH were little changed two days later, but Comment subsequently she recovered rapidly. An arterial Pco2 of 62 mm. Hg is not dangerous, Comment but the plasma bicarbonate concentration was later At the time of the measurements this patient reported to have been 22 mEq/l. at the time of his still had the respiratory alkalosis characteristic of deterioration. This is apparently a' normal' value, the early stages of salicylate intoxication (Singer, but when considered in conjunction with the Pco2 I954). She also was in severe metabolic acidosis, (Fig. 2, point 7) it can be seen to represent a which have been due either to the significant metabolic acidosis, being well below may salicylatecopyright. the normal range of the bicarbonate buffer line. poisoning or the renal failure. The resultant was a Thus the patient had both a respiratory and a slight alkalemia. Correction of the respiratory metabolic acidosis a severe acidaemia alkalosis by the administration of carbon dioxide giving (pH or by reducing ventilation would have caused a 7. 9) and improvement in ventilation was therefore fairly severe acidaemia; correction of the metabolic doubly important. acidosis by the administration of alkali would have caused a severe alkasemia. Either of these CASE 8.-Resistant ' Heart Failure '. A woman of 60 steps had gross cedema presumed to be due to ischaemic might have been unfortunate in such an ill patient. heart disease the and ECG were This case, therefore, illustrates the of although history importance http://pmj.bmj.com/ equivocal. Initially she responded well to mercurial a full evaluation of acid-base disturbances. (This diuretics, but oedema recurred and the plasma 'alkali case has been described elsewhere: reserve' was found to be 33 mEq./l. Although the fully Campbell plasma potassium concentration was normal she was and MacLaurin, I958.) thought to be suffering from metabolic alkalosis secondary to potassium depletion caused by overuse CASE II.-Renal Asthma. This case, described else- of the mercurial diuretics. The arterial pH was found where (Ashby and Campbell, I96I) (p. 43), illustrates to be 7.31, implying a Pco2 of about 65 mm.Hg (Fig. 2, many points, the most important lesson for routine point 8). A review of the case showed that she was, practice probably being the value of a method simple on October 3, 2021 by guest. Protected in fact, suffering from severe obstructive lung disease. of estimating Pco2 in supporting a clinical impression Comment which, by itself, carried little conviction. The diag- Patients who 'cor almost nosis of bronchial asthma had the blessing of six develop pulmonale' doctors who had seen the patient previously. The always have carbon dioxide retention. This possibility that it was wrong and the patient was really empirical observation, whatever its explanation overbreathing was not taken sufficiently seriously to (Campbell and Short, I960), can provide a useful justify spending time on arterial blood analyses in the aid to in 'heart failure' of uncertain middle of a busy day. In fact, the first reading of a diagnosis very low mixed venous Pco2 was disbelieved, but the cause. simplicity of the method allowed it to be checked in ten minutes. CASE 9.-Barbiturate Poisoning. An unconscious woman of 33 was admitted one Sunday evening. She Summary had swallowed an apparently large amount of cyclo- (i) Practical problems and methods of blood gas barbitone early in the day. Her respiration rate was are discussed. 8/min. and she was considered by most observers to be analysis very blue. Preparations were made for the administra- (2) Some physiological principles underlying tion of a stimulant drug and the arterial Pco2 was clinical interpretation are emphasized. Postgrad Med J: first published as 10.1136/pgmj.37.423.10 on 1 January 1961. Downloaded from January 1961 CAMPBELL: Blood Gas Measurements in Clinical Practice (3) Eleven illustrative cases are described. J. D. N. Nabarro, Mr. T. Holmes-Sellors and Mr. L. P. Le Quesne for permission to describe Acknowledgments these cases and to Mr. M. Hobsley, F.R.C.S., and I am grateful to Professor A. Kekwick, Dr. Dr. J. B. L. Howell for reading the manuscript. REFERENCES ASHBY, R. R., and CAMPBELL, E. J. M. (I96I): A Case of Renal Asthma, Postgrad. med. J., 37, 43. ASTRUP, P. (1956): A Simple Electrometric Technique for the Determination of Carbon Dioxide Tension in Blood and Plasma, Total Content of Carbon Dioxide in Plasma, and Bicarbonate Content in ' Separated ' Plasma at a Fixed Carbon Dioxide Tension (40 mm.Hg), Scand. . . din. Lab. Invest., 8, 33. BISHOP, J. M. (1960):-Measurement of Blood Oxygen Tension, Proc. roy. Soc. Med., 53, 177. BROOKS, D., and WYNN, V. (1959): Use of Venous Blood for pH and Carbon Dioxide Studies, Lancet, i, 227. BURWELL, C. S., ROBIN, E. D., WHALEY, R. D., and BICKELMAN, A. G. (1956): Extreme Obesity Associated with Alveolar Hypoventilation-a Pickwickian Syndrome, Amer. J. Med., 21, 8 1. CAMPBELL, E. J. M., and DICKINSON, C. J. (I960): Clinical Physiology, p. 120. Oxford: Blackwells. , and HOWELL, J. B. L. (1960): Simple Rapid Methods of Estimating Arterial and Mixed Venous Pco2, Brit. med..., i, 458. , and MACLAURIN, R. E. (1958): Acute Renal Failure in Salicylate Poisoning, Ibid., i, 503. , and SHORT, D. S. (1960): The Cause of CEdema in ' Cor Pulmonale ', Lancet, i, 1I84. COLLIER, C. R. (1956): Determination of Mixed Venous Carbon Dioxide Tensions by Rebreathing, Y. appl. Physiol., 9, 25. COTES, J. E. (I960): Respiratory Function and Portable Oxygen Therapy in Chronic Non-specific Lung Disease in Relation to Prognosis, Thorax, 15, 244. NILSSON, N. J. (1960): Oximetry, Physiol. Rev., 40, I. REFSUM, H. E., and HISDAL, B. (1958): Construction and Use of a Simple Reflectometer for Determination of Haemo- globin Oxygen Saturation in Blood, Scand. . clin. Lab. Invest., 10, 439. RILEY, R. L., CAMPBELL, E. J. M., and SHEPARD, R. H. (1957): A Bubble Method for Estimation of Pco2 and Poe in Whole Blood, J. appl. Physiol., II, 245. SEVERINGHAUS, J. W., and BRADLEY, A. F. (1958): Electrodes for Blood Po2 and Pco2 Determination, Ibid., 13, 515. SHEPARD, R. H. (1958): Effect of Pulmonary Diffusing Capacity on Exercise Tolerance, Ibid., 12, 487. SINGER, R. B. (1954): The Acid-base Disturbance in Salicylate Intoxication, Medicine (Baltimore), 33, I. VAN SLYKE, D. D., and NEILL, J. M. (1924): The Determination of Gases in Blood and other Solutions by Vacuumcopyright. Extraction and Manometric Measurement (I), J. biol. Chem., 6I, 523. VEREL, D., SAYNOR, R., and KESTEVEN, A. B. (1960): A Spectrophotometric Method of Estimating Blood Oxygen Using the Unicam SP 600, J. clin. Path., 13, 361. EXTRA BIBLIOGRAPHY DAVENPORT, H. W. (1958): The ABC of Acid-base Chemistry, 4th ed. Chicago: University of Chicago Press. MOLLER, B. (I959): The Hydrogen Ion Concentration in Arterial Blood, Acta med. scand., 165, supp. 348. WOOLMER, R. F. (I959): pH and Blood Gas Measurement. A Ciba Foundation London: & A. Churchill. Symposium. J. http://pmj.bmj.com/ on October 3, 2021 by guest. Protected