ARTERIAL BLOOD GASES and ACID BASE BALANCE

Introduction
The role of hydrogen ion (H+ or proton) in living cells is as important as that of sodium or potassium. Its intracellular concentration has not been reliably measured in most mammalian tissues (except erythrocytes), but it is thought to be critically important. It is affected by such factors as its extracellular concentration and the trans-membrane balance of potassium ions. Hydrogen ions are constantly produced (about 15 mol day-1) by metabolism and additional amounts are ingested with food. There must be a balance between production and excretion. Some 80% (12 mol day-1) is eliminated through the lung as carbon dioxide and water, and the rest through the kidney (3 mol day-1). The human body has several built-in mechanisms to maintain the extracellular concentration of hydrogen ion within narrow limits, namely 35-45 nmol l-1 (pH 7.45 - 7.35). In blood, sustained concentrations above 100 nmol l-1 (pH< 7.0) or below 10 nmol l-1 (pH>8.0) are incompatible with life.

Regulatory mechanisms
There are two major regulatory mechanisms which tend to keep the hydrogen ion concentration at the physiological 'set point' of 40 nmol l-1 (pH 7.4): a fast mechanism operated via the lung and a slow one operated via the kidney. Finally, the liver plays a major role in recycling lactate to bicarbonate to replace that lost in the periphery by buffering H+.

Respiratory homeostatic mechanisms

The immediate determinant of hydrogen ion concentration in blood is the partial pressure of carbon dioxide (PaCO2), which is acutely dependent on pulmonary ventilation. In a mechanically ventilated subject, the relationship between minute ventilation and the partial pressure of carbon dioxide in arterial blood (PaCO2) is a curve close to a hyperbola, varying its shape with the metabolic consumption. Note that the same increase in minute ventilation produces a

much smaller fall in PaCO2 as levels decrease below normal. An excess of H+ in arterial blood stimulates both the peripheral chemoreceptors in the carotid bifurcation and the respiratory neurones in the brain stem. This causes an increase in ventilation that leads to a fall in PaCO2 and H+ in blood, closing the negative feedback loop of this regulation. It should be noted that chemoreceptors are also stimulated by an increase in PaCO2. The arterial chemoreceptors are independently stimulated by (PaO2) values below 8 kPa (60 mmHg); the three stimuli interact with each other in a multiplicative way.

Carbon dioxide and hydrogen ion are interrelated through the reaction of the former with water to form carbonic acid; the latter dissociates into hydrogen H+ and bicarbonate HCO3- ions, a major buffer system in blood. CO2 + H20 H2CO3 (catalysed by carbonic anhydrase) H2CO3 HCO3- + H+ Traditionally, hydrogen ion concentration has been measured in pH units, pH being the negative logarithm (base 10) of the H+ concentration. Its quantitative relationship has proved more convenient. pH = pKa + log (HCO3-)/(a x PaCO2) where pKa is 6.1, and a the CO2 solubility factor (0.03, if PaCO2 is measured in mmHg, 0.2 if in kPa) The latter is known as the Henderson-Hasselbalch equation. It has three variables, pH, (HCO3-) and PCO2. So, it cannot be represented by a single curve, but rather by a family of curves forming a 'surface'.

Various authors have represented this relationship in different ways: Davenport chose to plot pH against (HCO3-) at different PaCO2 values. Sigaard-Andersen plotted pH against PaCO2 at various (HCO3-) values. These plots were of great practical value before the era of electronic calculators to estimate the (HCO3-) and the amount of excess acid or base present in blood. Most blood-gas machines in clinical use now have a micro-computer incorporated that instantly calculates these derived variables.

The way each variable in the Henderson-Hasselbalch equation depends on each other is a predictable function of temperature, concentration of Hb (buffer) and Hb saturation. This is also taken into account by the computer in the blood-gas machine. Temperature is not measured but the blood sample is warmed (or cooled) to 370 C in the measuring cuvette. The print out of most blood-gas machines shows eight or more results, four from direct measurement (pH, PaCO2, PaO2 and Hb), and the others derived by calculation. The latter include base excess, bicarbonate, standard bicarbonate and Hb saturation. Base excess is a calculation of the amount of acid (HCl), or base (NaOH), that would need to be added to a litre of blood (in vitro) to titrate the pH back to 7.40 at a normal PaCO2 of 5.3 kPa (40 mmHg) and at 370 C. The base excess result is negative in acidosis (it is really a 'base deficit', since NaOH would be added) and positive in alkalosis (HCl added).

The standard bicarbonate is a calculation of the bicarbonate value if the blood were to be equilibrated with a PaCO2 of 5.3 kPa (40 mmHg). It is important to be aware that only PaCO2, pH and PaO2 are obtained by direct measurement.

Renal homeostatic mechanisms

One of the many functions of the kidney is to excrete hydrogen ions. The H+ in urine may be as high as 30,000 nmol l-1 (pH 4.5), 800 times that of plasma. At such high H+, the total quantity of H+ that can be excreted in the urine depends on its buffering power.

Two main buffer systems exist in urine: the phosphate system and the ammonia/ammonium system: H+ + HPO4- H2PO4- (pK = 6.8) NH3 + H+ NH4+ (pK = 5) Note that these only work as buffers in the region of their respective pKs. The renal synthesis of ammonia is subject to regulation. Normally about 30 mmol of ammonium are excreted per day; after three days of severe acidosis as much as 200 mmol may be excreted per day. The excretion of each H+ion into the urine leads to re-absorption of a Na+ ion, and it is dependent upon the local PCO2 and

availability of carbonic anhydrase. The K+ ion competes with H+ in urinary excretion; K+ depletion leads to urinary loss of H+ and alkalosis, and acidosis leads to K+ retention and hyperkalaemia. The kidney also compensates for alkalosis by excreting excess bicarbonate; it is more effective in metabolic alkalosis because the respiratory compensation leads to an increased PaCO2.

The role of the liver in acid base balance

About 1300 mmol of lactate and associated H+ (lactic acid) are produced in the tissues every day. H+ions are buffered locally by HCO3- and lactate is released into the general circulation. The liver is the main site of lactate uptake (70%) as well as the heart and kidneys. In the liver, lactate can be converted to glucose or oxidised, either reaction consuming H+ and generating HCO3- which is recirculated to the periphery to replace that which was lost. (A similar reaction accounts for the generation of HCO3- from the lactate in Lactated Ringers solution). It is clear that failure of the liver to take up lactate in proportion to its production could lead to depletion of bicarbonate and a metabolic acidosis. This is particularly likely to happen during hypovolaemic shock where diminished oxygen delivery to the tissues generatesexcess lactic acid (anaerobic glycolysis). The H+ are buffered by HCO3- as usual but results in excessive bicarbonate consumption and lactate formation. Although acidaemia increases the ability of the hepatocyte to take up lactate (and generate the necessary bicarbonate), reduced hepatic blood flow (due to hypovolaemia and cardiac depression from acidaemia) leads to a vicious cycle of lactic acidaemia and loss of bicarbonate.

Acid-base disturbances
Introduction

It is important to understand the time course of acid-base disturbances. If it is caused by an abrupt change such as acute respiratory failure, the renal homeostatic mechanisms only develop fully after 3 to 5 days. A sudden acid challenge, such as that following release of a tourniquet applied to the lower limb for 2 hours (e.g. orthopaedic operation under epidural anaesthesia), causes alterations in the values of the blood acid-base status that change rapidly with time as respiratory compensation occurs. Following this rapid respiratory phase over the course of a few minutes, a slower hepatic and

renal phase follows in the course of the next few hours. The four clinical situations described below usually progress in two phases. The initial phase is well understood and could easily be mimicked in vitro by adding or removing CO2 or acid to a sample of arterial blood in a test-tube and measuring the pH and PaCO2.

What are the differences between aemia and osis?

The terms acidaemia and alkalaemia refer only to the status of the blood, acid or alkaline in pH or H+concentration. Acidosis and alkalosis refer to pathological situations resulting from a positive or negative balance of protons, where there is a change in PaCO2 or HCO3- in an acid or alkaline direction. However, as these changes may be compensatory, they may not lead to acidaemia or alkalaemia. Consider a patient that develops an increased H+ of 60 (pH 7.2) i.e. they have an acidaemia. If the HCO3- concentration is also found to be low they have, by definition, a metabolic acidosis. The effect of the increase in H+ leads to respiratory centre stimulation that increases ventilation to reduce CO2 to develop a respiratory alkalosis. But, how do we know which change is primary and which is secondary? Did the patient develop the metabolic change first and then respiratory compensation or the other way round? The important point to remember is that compensation on its own will never be enough to bring H+ or pH back into the normal range (see later for actual example). So, it is easy to see in this example that the primary event must be the metabolic acidosis that has led to the acidaemia and a secondaryor compensatory respiratory alkalosis. Example: a diabetic with a metabolic acidosis and acidaemia with a low HCO3- and pH of < 7.1 or H+ > 60) responds with hyperventilation that results in a low PaCO2. The latter constitutes a respiratory alkalosis, but the patient is not alkalaemic as the change is secondary. To decide the primary and secondary changes: 1. look at the H+ (pH) to decide whether the patient is acidaemic or alkalaemic. 2. look at the HCO3- and CO2 levels to see whether they have a metabolic or respiratory acidosis or alkalosis 3. the osis that agrees with the aemia is the primary change

So, if the patient is acidaemic with a high PaCO2 then it is this primary respiratory acidosis which has led to the acidaemia. There will often be a higher than normal HCO3- (especially if it is a chronic

CO2increase) to match this change, in other words a secondary or compensatory respiratory alkalosis. It is not possible to quantitate precisely metabolic acidosis and alkalosis in the whole body. Clinical observation is the only reliable indicator of the severity of the situation. In the blood sample, the degree of metabolic acidaemia or alkalaemia is easily seen by looking at the base excess value: values below -12 tend to be associated with severe acidosis, needing urgent therapy, and values above +12 are usually associated with severe alkalosis (see later).

Respiratory acidosis
This is the commonest situation and follows retention of CO2 due to acute or chronic respiratory failure due to inadequate ventilation of the lungs.

in acute CO2 retention, blood (H+) rises by about 6 nmol l-1 for each 1 kPa (7 mm) rise in PaCO2(pH drops 0.1 unit per 10 mmHg CO2 rise). For each 1.5 kPa (10 mmHg) acute rise in PaCO2, HCO3- increases by 1 mmol l-1 due to reaction of CO2 with H20.

Example: acute PaCO2 retention occurs postoperatively due to excess opioid in an otherwise fit patient. PaCO2 rises acutely by 3 kPa (21 mmHg) to 7.5 kPa (61 mmHg), pH will then drop by 0.2 to 7.2 (H+increases by 18 nmol l-1 to 58). HCO3 will be 26 mmol l-1.

in chronic CO2 retention, renal mechanisms allow re-absorption of plasma (HCO3-) which increases by about 4 mmol l-1 for each 1.5 kPa (10 mmHg) CO2 rise. This is often sufficient to restore pH to near normal values.

Example: an elderly patient with chronic CO2 retention has a PaCO2 of 8 kPa (60 mmHg). Renal retention of HCO3- has been sufficient to raise the HCO3- by 8 mmol l-1 to 32 mmol l-1. H+ is 45 (pH is 7.3) i.e. nearly normal.

Metabolic acidosis
The second commonest situation may appear in a variety of disease states, such as:

Grossly uncompensated diabetes (ketoacidosis).

Lack of oxygen delivery to tissues, as in shock or severe hypoxia. Loss of intestinal alkaline secretions, as in severe diarrhoea or fistulae. Transplantation of the ureters into the ileum after total cystectomy. Acetazolamide therapy, which impairs bicarbonate re absorption in the kidney. Failure of normal homeostatic mechanisms, as in generalised renal or hepatic failure.

N.B. (see above) that if a patient develops a metabolic acidosis and acidaemia they will develop a compensatory respiratory alkalosis i.e. the PaCO2 will fall. It is important to have some idea of what degree of compensation is likely. In an experiment where healthy subjects were made acidaemic by infusion of dilute solutions of HCl it was found that the resulting lowered PaCO2 was related to theHCO3- concentration by the following formula PaCO2 (in mmHg) = 8.4 + (HCO3- * 1.3) e.g. if HCO3- is 10, the normal subject should be able to reduce PaCO2 to 21 mmHg (3 kPa). Anything much higher than this will mean that the patient is unable to compensate fully (e.g. has respiratory depression due to drugs or ventilatory inadequacy). It is important to make this assessment since the efficacy of exogenous administration of NaHCO3 in patients with metabolic acidosis and acidaemia relies on adequate ventilation to blow off the CO2 generated by the reaction of H+ and HCO3- (see later).

Respiratory alkalosis
This is a less common situation and may be due to spontaneous hyperventilation which occurs in certain patients for unknown reasons. Respiratory alkalosis occurs commonly in mechanically ventilated patients under anaesthesia or sedation in the ICU. Acutely, it causes a fall in H+ of the same proportion as for respiratory acidosis, 6 nmol l-1 H+ change for 1 kPa PaCO2 change).

Metabolic alkalosis

A rare situation, that sometimes presents clinically as tetany with normal plasma calcium levels. It is due to:

loss of acid secretions as in compulsive vomiting or in pyloric stenosis. excessive ingestion of alkali (over treatment of 'indigestion'), or intravenous administration of bicarbonate (frequently seem after successful resuscitation from cardiac arrest).

potassium depletion (extracellular alkalosis and intracellular acidosis).

Interpretation of Blood-gas Values

Blood gas results are nearly always obtained from arterial blood (see later). In very special circumstances mixed venous blood may also be analysed to compute oxygen consumption, cardiac output or other variables. The sample must always be heparinised by priming the dead space of the syringe (just the hub) with a 1:1000 solution of heparin (failure to do this results in damage to the machine taking hours of expensive labour to repair).

The interpretation of results includes examination of the acid-base status of the blood and its oxygencarrying capacity. It must be remembered that examination of the clinical state of the patient is the most important factor in any therapeutic decision. Arterial blood measurements are only a narrow observation window of a very complex and poorly understood homeostatic system. However, following a logical routine in the examination of the results increases the chances of a correct diagnosis.

The use of integral microcomputers in blood-gas machines will soon be extended to provide interpretation of the results as well, following the same logical steps as the clinician. Although SI units are now officially in use, it will be found that in the great majority of establishments blood gases are still reported in pH and mmHg.

Artefacts

The blood-gas result should match the clinical assessment of the patient; a PaO2 of 5 kPa (38 mmHg) in a conscious, non-narcotised patient without evidence of dyspnoea or central cyanosis is almost certainly artefactual. The commonest artefact is the sampling of venous blood instead of arterial, due to faulty technique of arterial puncture. Extreme values of pH or derived parameters are likely to result from mixing of the blood sample with some acidic or basic residue in the syringe. For example, if highly concentrated subcutaneous heparin solution is used to prevent sample clotting, a very acidic result may be obtained. Samples taken from indwelling arterial catheters attached to long plastic tubes should be preceded by withdrawal of 6-8 ml of blood into another syringe to remove the priming heparinised saline. Mixing of blood with saline will give an unexpectedly low PaCO2 and a low Hb (if measured). If the machine measures Hb, care should be taken to stir the sample well just prior to injection into the cuvette, to prevent the effect of sedimentation of red cells in the syringe. The Hb value measured enters many of the calculations.

The time elapsed between taking the sample and analysing it is of little importance in clinical practice. PaO2 and PaCO2 change only by about 2% of the original value if kept in a 2 ml syringe for 1 hour at room temperature. If stored in ice samples may be kept for up to 12 hours with little change in values; this precaution is only necessary for research purposes.

Diagnosis
The method used in interpretation is based on the above: Look at pH or hydrogen ion changes Normal arterial blood pH is 7.4 +- 0.05 (H+ = 40 +- 5); it can be below that value or above it. The pH is below 7.35 (H+ < 45): there is acidaemia. If the HCO3- is < 22 mmol l-1, metabolic acidosis is the likely cause of the acidaemia. The PaCO2 will be < 5 kPa or 40 mmHg. A reduction in HCO3- occurs due to excess H+ production and a failure of HCO3- recycling by the liver. As HCO3- falls and H+ rises the chemoreceptor stimulation causes an

increase in ventilation and a reduction in PaCO2. This respiratory alkalosis tends to limit the resulting acidemia. If the PaCO2 is > 5 kPa or 40 mmHg respiratory acidosis is the likely cause of the acidaemia. In a pure respiratory acidosis, the HCO3- is always raised above normal. The pH is above 7.45 ((H+) < 35): there is alkalaemia: If the HCO3- is > 26 mmol l-1, metabolic alkalosis is present and is the likely cause of the alkalaemia. The PaCO2 will be > 5.5 kPa or 40 mmHg as a secondary change limiting the effect on pH or H+ of the metabolic changes.

If the PaCO2 is < 4 kPa or 30 mmHg, respiratory alkalosis is present and is the likely cause of the alkalaemia. HCO3- will be lowered as the kidney attempts to excrete HCO3- to limit the alkalaemia i.e. a compensatory metabolic acidosis is present.

Oxygen changes

The partial pressure of oxygen in arterial blood (PaO2) falls with age. On average, it is approximately 13 kPa (100 mmHg) in the young adult, and declines steadily to about 10 kPa (75 mmHg) at the age of 80. There is hypoxia when PaO2 values are 2 kPa (15 mmHg) below the expected value for the patient's age. The peripheral arterial chemoreceptors in the aortic arch and in the bifurcation of the carotid arteries are the only known oxygen sensors in the body. They are progressively stimulated by PaO2 values below 8 kPa (60 mmHg), causing a reflex increase in ventilation. Hypoxic drive to ventilation, such as in pulmonary oedema or at altitude, usually leads to a fall in PaCO2. In acute respiratory failure, hypoxia may be accompanied by CO2 retention.

Treatment of Acid-base Disturbances

The cause of the acid-base disturbance must be treated and the normal respiratory and renal homeostatic mechanisms allowed to restore the balance of protons (H+). Only rarely is it indicated to

infuse intravenously an alkali or an acid. HCl infusions in severe alkalosis have been described in a few instances, but with doubtful benefit for the clinical status of the patient. On the other hand, sodium bicarbonate infusions have been widely employed in the past to treat severe metabolic acidosis; presently their indication has been much restricted as more deleterious side effects have been found. Its precise role in diabetic acidosis and cardiac arrest is hotly debated.

8.4 % (or molar) sodium bicarbonate (84 is the MW) carries a considerable load of sodium (1 mmol Na+ per mmol HCO3- per ml of solution). High CO2 pressures are generated on mixing with acidic blood which cause diffusion of CO2 into cells and intracellular acidosis. This is more likely to occur in mechanically ventilated patients or those in respiratory failure who have inefficient pulmonary CO2elimination (see above). When sodium bicarbonate is considered necessary to correct acidosis, it should be remembered that the figure for base excess refers to the deficit of bicarbonate in the extracellular fluid (ECF). Thus, it is necessary to calculate ECF in litres and multiply this value by the BE to arrive at a figure for sodium bicarbonate in mmol. In adults, about 1/3 to 1/5 of body weight is ECF, whilst in babies it is about 2/5. Usually only 1/2 the calculated amount is given initially, e.g. Weight = 60 kg BE = -12 ECF = 0.2 x 60 = 12 litres Thus, Initial dose of NaHCO3 needed = 0.5 * 12 * 12 = 72 mmol NB. As stated above, sodium bicarbonate must never be given if PaCO2 is raised or there is any evidence that the patient has respiratory impairment.

Arterial Puncture

Arterial blood for gas and acid-base balance measurements is usually sampled from the radial or femoral arteries. The volume required is only 0.5 ml so a 1 ml sample in a 2 ml syringe is sufficient. It should be taken with the smallest diameter needle allowing a rapid flow of blood into the syringe, i.e. 23G (blue hub), to minimise damage to the arterial wall (in some cases, an indwelling arterial catheter will be in place, so this should be used instead. Firstly, the dead space of the syringe must be filled with a 1:1000 solution of heparin and all the air removed; the needle must be firmly attached to the hub so air is not drawn in on suctioning. Introducing the needle perpendicular to the skin more easily punctures both radial and femoral artery.

In the femoral, the second and third fingers of the left hand should be kept over the course of the artery and the pulsation felt in both fingers which are kept some 1 cm apart. The needle is then inserted in between the two fingers applying a gentle constant pull to the plunger of the syringe; on entering the arterial lumen a sudden 'give' is usually felt and blood appears in the syringe.

In the radial artery the following alternative technique may be more effective: extend the wrist joint and start by feeling the arterial pulse with one finger some 2 cm proximal to the wrist, and mark the point where it is felt with a dot, using an ordinary pen. Repeat this procedure 2 or 3 times at steps 5 mm proximally to the last point; this way the course of the artery is identified by doing a mental 'curve fitting' exercise. Then, keeping the wrist extended insert the needle into the artery following the same procedure as for the femoral. There are specially made syringes for blood-gas sampling which are pre-heparinised and allow the blood to flow in by its own pressure, minimising the risk of accidental sampling of venous blood.

If it is suspected that the needle has entered a vein then the syringe may be disconnected from the needle and the absence of brisk blood flow through the latter confirms venous puncture.