9. pH AND TEMPERATURE

If a patient's central temperature is not near 37°C and measurement of pH and PCO2 are made on electrodes at 37°C problems of interpretation of the results exist. The readings of pH and PCO2 differ from those which would be obtained from electrodes held at the temperature of the patient.

A second set of problems are related to the physiological meaning of pH and PCO2 readings at patient temperatures other than 37°C even if the readings are made at or corrected for such a temperature.

A third set of problems are those related to gradients from one part of the body to another. If this is important, it will be most critical during extracorporeal cooling or warming if rapid changes in temperature are attempted, e.g. 1°C/minute change in oesophageal temperature.

I think that at all temperatures from which survival can be reasonably expected (10 to 44°C), optimal pH is near 7.4 and optimal PCO2 near 40mmHg at the temperature of the patient.

In Appendix 9, there is further discussion of various "optimal" desired pH and PCO2 states advocated by different authorities. It is probably not possible to resolve the differences because the criteria of "optimal" state are often not comparable. For example, mortality of surgery, physiological manifestations and biochemical phenomena have been used. The practical consequences to be discussed are relevant whichever level of PCO2 or pH one decides to aim at:

9.1. CO2 Production when Body Temperature is low.

In the steady state of low body temperature the metabolic rate is low and so CO2 production is low. The alveolar ventilation will have to match this low production if the PaCO2 is to be kept near 40mmHg. This applies to both artificial and spontaneous respiration.

In accidental hypothermia which occurs associated with exposure to cold climatic conditions, often in patients who are already debilitated or intoxicated, there may be little or no respiratory movement when first seen in the Emergency Department. It is tempting to institute intermittent positive pressure respiration in the presence of such a clinical picture. This may result in a very low PaCO2, e.g. less than 10mmHg.

In clinically induced hypothermia the same problem can occur, if as is usual, artificial respiration is used for the anaesthetic required to enable the body to be cooled. Ordinary respiratory rates and tidal volumes will result in very low PaCO2's.

Although I know of no studies specifically examining the effects of low PCO2's in hypothermia, when respiration is controlled, it is reasonable to assume very low PCO2's would not be beneficial. Cerebral vasoconstriction and a shift of the oxygen dissociation curve are the main theoretical consequences which would have to be sought in any such studies. The extreme shift in the oxygen dissociation curve could lead to tissue hypoxia resulting in the release of acidic products of metabolism and causing hypoxic tissue damage.

To keep the PCO2 from falling the following solutions can be applied together with monitoring of end tidal CO2:

• Increasing the inspired oxygen partial pressure but allowing the respiration to remain spontaneous, e.g. endotracheal intubation and connecting the tube to some type of anaesthetic circuit to which only oxygen is added.
• Artificial ventilation with 5% CO2 in oxygen.
• Artificial ventilation with a closed anaesthetic circuit without soda lime absorption and with a low fresh gas flow.

9.2 CO2 Balance during the Cooling Process.

During the cooling process the balance of CO2 also has to be taken into account. If cooling is rapid (e.g. 1°C every 3-5 minutes), the quantity of CO2 which has to be dissolved in body tissues to prevent the PCO2 falling is great and even if the lungs or the oxygenator are ventilated with a CO2 containing mixture the PCO2 in arterial blood and/or mixed venous blood may fall.

To maintain the PCO2 while the temperature is being altered the following recommendations should be applied:

a) Surface Cooling:

i) Administer 5% CO2 with hyperventilation during cooling. Check the PaCO2 during and after stability has been achieved. Correct PaCO2 for temperature or set an electrode system at the lower temperature. Adjust inspired CO2 to keep PaCO2 near 40mmHg at the temperature of the patient. (End tidal ie alveolar PCO2 would be a rapid guide in achieving the desired PCO2. If a Swan-Ganz catheter is in place the mixed venous PCO2 may be more useful than the arterial)

ii) Use a low flow of gas into the anaesthetic circuit or allow spontaneous respiration during the cooling process. Again it will be necessary to check and adjust the PaCO2.

b) Extracorporeal Cooling:

Limit rate of cooling to not greater than 1°C/3 min. This is important to ensure eveness of cooling as well as for maintenance of the desired pH and PCO2 state. Administer 20% CO2 into the oxygenator during cooling. Check the PCO2 of the mixed venous blood (i.e. from the venous reservoir of the extracorporeal pump). Arterial blood would not be satisfactory as there is usually a very large gradient for CO2 between the arterial blood and the other tissues during cooling.

(Note: The arterial blood is the same as the blood leaving the pump whereas the venous reservoir blood PCO2 would be closer to that of the tissues during a period of rapid uptake of CO2).

During rewarming the liberation of CO2 increases and will require adjustment of alveolar ventilation and/or stopping of CO2 administration to the extracorporeal circuit to keep the PCO2 at the desired level. (See Appendix A.9, below).

A.9. Temperature and pH

Partial pressure of a gas is that pressure which would be present if the particular gas alone occupied the space containing the gas mixture. The total pressure of a gas mixture is equal to the sum of the separate partial pressure of each component of the mixture. The partial pressure of a gas in a liquid is defined as the partial pressure of that gas which would be present in a gas mixture in equilibrium with the liquid.

Solubility of a gas in liquids increases as the temperature decreases, i.e. a larger quantity of gas will dissolve at any particular partial pressure if the temperature falls or when there is no gaseous phase in contact with the liquid, the partial pressure falls when the temperature falls.

When a patient's central temperature deviates much from 37°C interpretation in the literature of pH and PCO2 data is very confused. Reasons for confusion:

1) Electrode systems of measuring physiological pH and PCO2 are usually held at 37°C.

2) Solubility of CO2 varies with temperature. Changes in temperature when blood is placed in an electrode will alter the PaCO2 and, therefore, the pH independently of any change in pH which would occur in a buffer solution which did not contain CO2.

3) pH of the standard buffers vary with temperature. If the temperature of electrodes are held near to the patient's central temperature a correction factor has to be applied to the standard. Such correction factors for the standards can be justifiable on theoretical grounds and the conventions of standard buffer solutions (Vogel, 1961).

PCO2 electrodes are calibrated against gas mixtures; the partial pressure of CO2 in the gas mixture is independent of temperature.

4) There is no concensus of the "normal" or optimal arterial or mixed venous pH or PCO2 in humans at temperatures other than 37°C. (Williams & Marshall, 1982).

5) There is uncertainty in the questions:

a) Whether changes in temperature per se produce changes in pH by, for example, increased lactic acid production.

b) Whether incidental changes, e.g. in PCO2 may produce circulatory states which result in tissue hypoxia and therefore, increased metabolic acid production.

c) Whether any such process can be modified by manipulation of pH by adding inorganic acid (HCl) or CO2.

In the literature it is often impossible to decide how the measurements were made. Controls are rare when various clinical regimens have been reported. In vitro measurements are often not accompanied by a clear indication of the details of the patient and the electrode temperatures.

The Rosenthal factor (1948) corrects blood pH for the difference between the temperature of the patient and that of the measuring electrode. It is applied to in vitro measurements where no gas can escape from the sample before or during the measurement.

The factor is 0.0147 per C°. The formula for its use is :

Blood pHt1 = Blood pHt2 - 0.0147(t1-t2) where t1 and t2 are temperature readings.

Example: Blood is taken from a patient with a central temperature of 30°C. The pH is measured in an electrode set at 37°C. To ascertain what the pH is at 30°C apply the formular pH at 30°C = pH reading at 37°C - 0.0147 (30-37).

Edmark (1959) and Brooks (1964) point out that, if the blood is tonometered with gas at PCO2 40mmHg and the pH measured at the same temperature as the tonometer, the pH does not alter with the temperature. In other words if the PCO2 is kept at 40mmHg, pH of blood which was 7.4 at 37°C, will be 7.4 at other temperatures or that the alteration of pH of blood with temperature is mainly, if not only, due to changes in PCO2.

I am surprisied that the pH does not alter with temperature when it is the partial pressure of carbon dioxide which is kept constant at 40mmHg. As the solubility of CO2 (and all gases) increases as the temperature falls the quantity of CO2 dissolved will be greater at lower temperatures. For the ratios between dissolved CO2, carbonic acid and the products of ionisation H⁺ and HCO3^- to be such that the pH happens to end up at 7.4 if the PCO2 is 40mmHg seems too neat to be fortuitous.

It seems to me that:

a) There must be some important physico-chemical reason for this, or,

b) there are special properties of protein buffering mechanism to account for it, or,

c) that all authors have overlooked some important factors.

In the practical and clinical application of pH and PCO2 manipulation in hypothermia, the papers of Edmark (1959), Carson et al (1962), Brooks (1964), Burton (1964), McConnell et al (1975), Rahn et al (1975), Norwood et al (1974), Becker et al (1981), White (1981) and Williams et al (1982) are relevant.

Edmark formed the impression that at low temperatures the optimal pH should be lower than 7.4 measured at the temperature of the patient. This he achieved by adding HCl during extracorporeal cooling. In his subjectgs it appears that artificial respiration was continued. It is not stated if CO2 was added to the pump but he says that pH would have risen to 7.53 if HCl was not added. The rate of cooling was 12°C in 20 minutes. He seems to have compared cardiac rhythm and ease of defibrillation at 25°C when the pH was 7.53 with no CO2 or HCl added and at 25°C when HCl was added to produce a pH of 7.15 without adding CO2. The comparison was clinical and without any statistics. He did not examine 25°C with pH 7.1 by adding more CO2.

Brooks although obviously aware of Edmark's work and giving a full discussion of in vitro changes in pH and PCO2 with temperature elected in some (he does not state the proportion) patients to ventilate the lungs normally (i.e. minute volumes suitable for normal metabolism at 37°C) while using Drew's technique for extra-corporeal cooling. The Drew technique uses the patient's lungs but replaces the right and left ventricles with separate roller pumps. He added NaHCO2 in amounts required to keep the pH from falling during the period of hypothermia. In some cases he added CO2 to the respiratory circuit to keep the end tidal PCO2 at 40mmHg. It is not possible to deduce from this paper what, if any, differences there were in the end result with different methods of manipulation of the pH and PCO2. It appears that at least when CO2 was not added a metabolic acidosis occurred.

Burton (1964) reported on satisfactory clinical results on 69 cases of induced hypothermia to around 15°C when the PCO2 (end tidal and arterial) were kept near 40mmHg. This was not a controlled study but before attempting to keep the PCO2 near 40mmHg the end tidal CO2 had been noticed in other subjects to be unmeasurably low and may have contributed to a state of metabolic acidosis which presented post-operatively.

Carson and Morris added CO2 to the extracorporeal circuit to produce a pH of the level recommended by Edmark. They claim that this method prevented the occurrence of metabolic acidosis. After cessation of cardiopulmonary bypass non-respiratory acidosis did occur in cases which were "unsuccessful", presumably due to low cardiac output. It is, of course, possible that the metabolic disturbance might have caused or aggravated the low cardiac output. These authors noted that at low temperatures when CO2 was administered, respiration continued and general anaesthesia had to be administered as the patients appeared to be awakening. In this paper the low temperatures reported are those of the pumpoxygenator during the period of cardiopulmonary bypass and oesophageal during the rest of the operation. 25°C was the lowest temperature mentioned in the context of clinical hypothermia but as the records show these to be pump temperatures one must presume that the oesophageal and naso-pharyngeal temperatures were higher.

Our experience has been in cardiac surgery using deep (20°C) and mild (30°C) hypothermia using extracorporeal circulation. As well as this we operated on some twenty patients with cerebral aneurysms who were cooled to 15°C for operation.

The circumstances of the neurosurgical procedures enabled us to make some observations. We used extra-thoracic cooling by way of the femoral artery and both femoral veins and partial bypass until or if ventricular fibrillation occurred. As the chest was not open it was feasible to use spontaneous respiration with halothane anaesthesia. The oxygenator was ventilated with oxygen to which was added halothane and up to 20% CO2. After some trials and errors we limited the rate of cooling of the nasopharynx to 1°C/3 minutes. If cooling was faster, PCO2 in the mixed venous blood often fell below 40mmHg at the nasopharyngeal temperature of the patient and respiration ceased. If cooling was stopped but perfusion continued respiration would recommence after some minutes. Blood taken from the venous reservoir when respiration recommenced was invariably near 40mmHg at the nasopharyngeal temperature of the patient.

The explanation of this is that the temperature fell and the solubility of CO2 increased so that it was difficult to supply the quantity of CO2 required to keep the tissue PCO2 near 40mmHg even if the arterial blood had a high PCO2. At a constant low temperature equilibrium could occur so that adding 5% CO2 into the pump oxygenator was sufficient but during the cooling phase the PCO2 of the tissues fell if the rate of cooling was too fast and/or insufficient CO2 was added to the pump. It is possible that other workers cooled patients too quickly and/or that CO2 administered to the oxygenator was insufficient for the PCO2 of the tissues to be near 40mmHg during parts of the periods of hypothermia.

We also found, as did Carson et al, that anaesthesia was necessary at low temperatures. In our initial cases we turned off the halothane to the oxygenators when the temperature fell below 30°C, but we soon found that it was necessary to continue to administer halothane. One patient appeared to be arousing 15°C. Halothane administration rendered the patient "quiet". Spontaneous respiration continued at 15°C at a rate of about 4/minute but with quite large tidal volumes provided again the PCO2 was at or above 40mmHg. Metabolic acidosis was not a feature of these cases. Base excess measurements were made at 37°C. If we added NaHCO2 to the priming solution of the pump so that the base excess of the prime was zero, no alteration of base excess occurred in the patient during or after cooling. If we did not "zero" the base excess in the pump, then the base excess dropped suddenly when bypass started. It remained unchanged while the temperature was low and then corrected slowly during and after rewarming.

Our conclusion was that if tissue PCO2 is kept near 40mmHg at all temperatures, metabolic acidosis does not occur, and that some complex CNS functions continue as at normal temperature. So that we think that the optimal pH is 7.4 and PCO2 40mmHg at all temperatures.

In contrast, Rhan et al, (1975) propose that at all temperatures the CO2 content of the blood should be the same, i.e. the optimal pH and PCO2 would be the same as those which occur if blood is warmed or cooled in a closed system and the pH and PCO2 measured at the altered temperature. If the pH and PCO2 measuring system is maintained at 37°C then the desired pH would be 7.4 and desired PCO2 40mmHg.

In the clinical situation of induced hypothermia, pulmonary ventilation would be similar to that imposed at normal temperature as the increased solubility and decreased production would at least partially compensate for each other. Monitoring would be done on arterial blood if extracorporeal circulation is not being used or the mixed venous blood from the venous reservoir if the circulation is extracorporeal.

The justification for these recommendations are:

a) The neutral point of water changes with temperature, i.e. pH 6.8 at 37°C (which corresponds to normal intracellular pH) or pH 7.5 at 0°C.

b) In cold blooded animals pH and PCO2 alter with temperature in a similar way. This is therefore suggested as being optimal for human hypothermia.

In contrast, hibernating animals which presumably are evolutionarily closer to humans than poikilothermic animals, respond differently to low temperatures. Hibernating animals keep the PCO2 near 40 torr or higher, while the temperature is low. A summary of the published findings for the hibernating animals is given by Goodrich (1973).

The argument for optimal pH management at low temperatures by analogy with poikilothermic animals is not conclusive. One could just as well argue that pH optimisation should follow the hibernating animals or some other set of conditions.

Becker et al, (1981) have advocated blood pH higher than 7.4 at low temperature. This is achieved by lowering PCO2 by ventilation of the lungs or extracorporeal oxygenator. They use small groups of animals and show quite large differences in circulatory dynamics and ease of resuscitation. The comparison is between animals in which the pH is kept at 7.4 and others where the pH is above 7.4 (both measured at 37°C). The claim is higher cardiac output and easier cardiac rhythm managements when pH is greater than 7.4. Similar claims were made by Edmark and Carson and Morris when the pH was deliberately lowered below 7.4 on cooling. Carson and Morris also claimed that the low pH due to CO2 administration was associated with low production of non-respiratory acids; the same is claimed by Becker et al for the technique of using a high pH.

It is apparent that there is great confusion in the literature which cannot be resolved by the methods being used at present (i.e. using small groups of animals and extrapolating results to human experiences where much larger groups are used and results between different groups using different pH management are not obvious). There are several types of arguments being used. I will try to classify these:

a) Evolutionary or comparative considerations. These compare what cold blooded or hibernating animals do biochemically when the temperatures are below 37°C. Any conclusions from these must be applied to large numbers of humans and produce a very obvious difference before acceptance.

Animal experiments are done on a small (e.g. six or twelve) numbers of animals (McConnell (1979), Becker (1975) and Norwood et al (1979)). The results show marked superiority of one technique. If such results were transferable to clinical practice, very obvious differences in clinical outcome would be apparent. Clinical outcome for coronary artery grafting and congenital heart repairs do not appear to be much influenced by such details of perfusion technique as the pH and PCO2 of the patient.

b) Arguments based on the clinical outcome when various pH and PCO2 manoeuvres are used at low temperatures. Criteria are such things as rhythm changes and stability, ease of discontinuation of extracorporeal circulation and alterations in contractile activity of the myocardium.

c) Arguments of physiological response during periods of low temperature.

i) Peripheral circulation (presence or absence of vasoconstriction or dilatation).

ii) Rates and eveness of cooling.

iii) Occurrence of ventricular or arterial arrythmias or fibrillation and ease of management of these.

iv) Central nervous system activity as shown by respiratory muscle contraction or states of arousal.

The following is a recapitulation of the main regimens recommended for pH control when body temperature is less than 37°C:

		Electrodes at 37°C		Electrodes at core temp.
		pH	PCO2	pH	PCO2
Edmark 1959				<7.4
Carson and Morris 1962				<7.4	> 40 mmHg
This author*		< 7.4	> 40	7.4	40mmHg
White 1981, Rahn et al. 1975, McConnell et al 1975		7.4	40	>7.4	< 40mmHg
Becker (1981)		> 7.4	< 40	>> 7.4	<< 40mmHg

*Said to be commonest clinical practice by Becker et al 1981 and Williams and Marshall, 1982.

The mortality of open heart surgery for congenital and acquired disease is less than 3% (2.7% of 15,399 cases. Lancet 1982). Studies show large differences in small groups of animals subject to alterations in pH management during hypothermia. Such differences in pH management have been applied in the clinical situation by different groups. The clinical outcome of such applications has not been shown to be very different. It is hard to imagine a major clinical cardiovascular or cerebral alteration which is obvious in 6 animals not being apparent in the massive clinical use during coronary artery grafting. One can only conclude that the animals results are not applicable to the clinical situation. The reasons for not being able to apply them may range from species difference to poor design or execution of the studies.

At present, I think that it is not possible from the available studies to conclude that any of the possible set of aimed at pH or PCO2 conditions appears to have been established as being superior to any other. I think that it is unlikely that it will be shown that any difference between the possibilities of pH management could produce a very large difference in the clinical results.