Airway	If threatened
Breathing	All respiratory arrests
Respiratory rate	<5 breaths/min
Respiratory rate	>36 breaths/min
Circulation	All cardiac arrests
Pulse rate	<40 beats/min
Pulse rate	>140 beats/min
Systolic blood pressure	<90 mmHg
Neurology	Sudden fall in level of consciousness (fall in Glasgow Coma Scale of >2 points)
	Repeated or prolonged seizures
Other	Any patient who does not fit the criteria above, but about whom you are seriously worried

From Hillman K, Chen J, Cretikos M et al. Introduction of the medical emergency team (MET) system: a cluster-randomized controlled trial. Lancet 2005; 365:2091–2097, with permission.

In some of the most seriously ill patients, the precise underlying diagnosis is initially unclear but in all cases, the immediate objective is to preserve life and prevent, reverse or minimize damage to vital organs such as the lungs, brain, kidneys and liver. This involves a rapid assessment of the physiological derangement followed by prompt institution of measures to support cardiovascular and respiratory function in order to restore perfusion of vital organs, improve delivery of oxygen to the tissues and encourage the removal of carbon dioxide and other waste products of metabolism (following the ABC approach: Airway, Breathing, Circulation, see Fig. 16.25, below). The patient’s condition and response to treatment should be closely monitored throughout. The underlying diagnosis may only become clear as the results of investigations become available, a more detailed history is obtained and a more thorough physical examination is performed. In practice resuscitation, assessment and diagnosis usually proceed in parallel.

Critically ill patients require multidisciplinary care with:

Intensive skilled nursing care (usually 1 : 1 or 1 : 2 nurse/patient ratio in the UK).

Specialized physiotherapy including mobilization and rehabilitation.

Management of pain and distress with judicious administration of analgesics and sedatives (see p. 893).

Constant reassurance and support (critically ill patients easily become disorientated and delirium is common.

H₂-receptor antagonists or proton pump inhibitors in selected cases to prevent stress-induced ulceration.

Compression stockings (full-length and graduated), pneumatic compression devices and subcutaneous low-molecular-weight heparin to prevent venous thrombosis.

Care of the mouth, prevention of constipation and of pressure sores.

Nutritional support (see p. 222). Protein energy malnutrition is common in critically ill patients and is associated with muscle wasting, weakness, delayed mobilization, difficulty weaning from ventilation, immune compromise and impaired wound healing. There is also an association between malnutrition and increased mortality. It is therefore recommended that nutritional support should be instituted as soon as is practicable in those unable to meet their nutritional needs orally, ideally within 1–2 days of the acute episode. Enteral nutrition, which is less expensive, preserves gut mucosal integrity, is more physiological and is associated with fewer complications, is preferred. Recently, the value of early feeding has been questioned, apart from giving small amounts to ensure gut viability. Parenteral nutrition is sometimes indicated at a later stage for those unable to tolerate or absorb enteral nutrition and should be initiated without delay, at least within 3 days. Persistent attempts at enteral nutrition in those with gastrointestinal intolerance leads to underfeeding and malnutrition.

Critically ill patients commonly require intravenous insulin infusions, often in high doses, to combat hyperglycaemia and insulin resistance (see p. 1006). Although the use of intensive insulin therapy to achieve ‘tight glycaemic control’ (blood glucose level between 4.4 and 6.1 mmol/L) was shown to improve outcome (at least when combined with aggressive nutritional support), more recent studies have failed to confirm this finding and have indicated that this approach is associated with an unacceptably high incidence of hypoglycaemia, and possibly an increase in mortality. Current recommendations suggest that blood glucose levels should be maintained at <8–10 mmol/L.

Discharge of patients from intensive care should normally be planned in advance and should ideally take place during normal working hours. Planned discharge often involves a period in a ‘step-down’ intermediate care area. Premature or unplanned discharge, especially during the night, has been associated with higher hospital mortality rates. A summary including ‘points to review’ should be included in the clinical notes and there should be a detailed handover to the receiving team (medical and nursing). The intensive care team should continue to review the patient, who might deteriorate following discharge, on the ward and should be available at all times for advice on further management (e.g. tracheostomy care, nutritional support). In this way, deterioration and readmission to intensive care (which is associated with a particularly poor outcome) or even cardiorespiratory arrest might be avoided.

This chapter concentrates on cardiovascular and respiratory problems. Many patients also have failure of other organs such as the kidney and liver; treatment of these is dealt with in more detail in the relevant chapters.

FURTHER READING

Fullerton JN, Perkins GD. Who to admit to intensive care? Clin Med 2011; 11:601–604.

FURTHER READING

Casaer MP, Mesotten O, Hermans G et al. Early versus late parenteral nutrition in critically ill adults. N Engl J Med 2011; 365:506–517.

The NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009; 360:1293–1297.

Ziegler TR. Parenteral nutrition in the critically ill patient. N Engl J Med 2009; 361:1088–1097.

Applied cardiorespiratory physiology

Oxygen delivery and consumption

Oxygen delivery (DO₂) (Fig. 16.1) is defined as the total amount of oxygen delivered to the tissues per unit time. It is dependent on the volume of blood flowing through the microcirculation per minute (i.e. the total cardiac output, ) and the amount of oxygen contained in that blood (i.e. the arterial oxygen content, C_aO₂). Oxygen is transported both in combination with haemoglobin and dissolved in plasma. The amount combined with haemoglobin is determined by the oxygen capacity of haemoglobin (usually taken as 1.34 mL of oxygen per gram of haemoglobin) and its percentage saturation with oxygen (SO₂), while the volume dissolved in plasma depends on the partial pressure of oxygen (PO₂). Except when hyperbaric oxygen is administered, the amount of dissolved oxygen in plasma is insignificant.

Figure 16.1 Tissue oxygen delivery and consumption in a normal 70 kg person breathing air. Oxygen delivery (DO₂) = cardiac output × (haemoglobin concentration × oxygen saturation (S_aO₂) × 1.34). In normal adults, oxygen delivery is roughly 1000 mL/min, of which 250 mL is taken up by tissues. Mixed venous blood is thus 75% saturated with oxygen. , mixed venous oxygen content; , mixed venous oxygen saturation; CaO₂, arterial oxygen content.

Clinically, however, the utility of this global concept of oxygen delivery is limited because it fails to account for changes in the relative flow to individual organs and the distribution of flow through the microcirculation (i.e. the efficiency with which oxygen delivery is matched to the metabolic requirements of individual tissues or cells). Furthermore, some organs (such as the heart) have high oxygen requirements relative to their blood flow and will receive insufficient oxygen, even if the overall oxygen delivery is apparently adequate. Lastly, microcirculatory flow is influenced by blood viscosity.

Oxygenation of the blood

Oxyhaemoglobin dissociation curve

The saturation of haemoglobin with oxygen is determined by the partial pressure of oxygen (PO₂) in the blood, the relationship between the two being described by the oxyhaemoglobin dissociation curve (Fig. 16.2). The sigmoid shape of this curve is significant for a number of reasons:

Modest falls in the partial pressure of oxygen in the arterial blood (P_aO₂) may be tolerated (since oxygen content is relatively unaffected) provided that the percentage saturation remains above 92%.

Increasing the P_aO₂ to above normal has only a minimal effect on oxygen content unless hyperbaric oxygen is administered (when the amount of oxygen in solution in plasma becomes significant).

Once on the steep ‘slippery slope’ of the curve (percentage saturation below about 90%), a small decrease in P_aO₂ can cause large falls in oxygen content, whereas increasing P_aO₂ only slightly, e.g. by administering 28% oxygen to a patient with chronic obstructive pulmonary disease (COPD), can lead to a useful increase in oxygen saturation and content.

Figure 16.2 The oxyhaemoglobin dissociation curve. a, arterial point; v, venous point; x, arteriovenous oxygen content difference. HbO₂ (%) is the percentage saturation of haemoglobin with oxygen. The curve will move to the right in the presence of acidosis (metabolic or respiratory), pyrexia or an increased red cell 2,3-DPG concentration. For a given arteriovenous oxygen content difference, the mixed venous PO₂ will then be higher. Furthermore, if the mixed venous PO₂ is unchanged, the arteriovenous oxygen content difference increases and more oxygen is off-loaded to the tissues (see p. 374). P₅₀ (the PO₂ at which haemoglobin is half saturated with O₂) is a useful index of these shifts – the higher the P₅₀ (i.e. shift to the right), the lower the affinity of haemoglobin for O₂.

The P_aO₂ is in turn influenced by the alveolar oxygen tension (P_AO₂), the efficiency of pulmonary gas exchange, and the partial pressure of oxygen in mixed venous blood .

Alveolar oxygen tension (P_AO₂)

The partial pressures of inspired gases are shown in Figure 16.3. By the time the inspired gases reach the alveoli they are fully saturated with water vapour at body temperature (37°C), which has a partial pressure of 6.3 kPa (47 mmHg) and contain CO₂ at a partial pressure of approximately 5.3 kPa (40 mmHg); the P_AO₂ is thereby reduced to approximately 13.4 kPa (100 mmHg).

Figure 16.3 The composition of inspired and alveolar gas (partial pressures in kPa).

The clinician can influence P_AO₂ by administering oxygen or by increasing the barometric pressure.

Pulmonary gas exchange

In normal subjects there is a small alveolar-arterial oxygen difference (P_A–aO₂). This is due to:

a small (0.133 kPa, 1 mmHg) pressure gradient across the alveolar membrane

a small amount of blood (2% of total cardiac output) bypassing the lungs via the bronchial and thebesian veins

a small degree of ventilation/perfusion mismatch.

Pathologically, there are three possible causes of an increased P_A–aO₂ difference:

Diffusion defect. This is not a major cause of hypoxaemia even in conditions such as lung fibrosis, in which the alveolar-capillary membrane is considerably thickened. Carbon dioxide is also not affected, as it is more soluble than oxygen.

Right-to-left shunts. In certain congenital cardiac lesions or when a segment of lung is completely collapsed, a proportion of venous blood passes to the left side of the heart without taking part in gas exchange, causing arterial hypoxaemia. This hypoxaemia cannot be corrected by administering oxygen to increase the P_AO₂, because blood leaving normal alveoli is already fully saturated; further increases in PO₂ will not, therefore, significantly affect its oxygen content. On the other hand, because of the shape of the carbon dioxide dissociation curve (Fig. 16.4), the high PCO₂ of the shunted blood can be compensated for by over-ventilating patent alveoli, thus lowering the CO₂ content of the effluent blood. Indeed, many patients with acute right-to-left shunts hyperventilate in response to the hypoxia and/or to stimulation of mechanoreceptors in the lung, so that their P_aCO₂ is normal or low.

Ventilation/perfusion mismatch (see p. 796). Diseases of the lung parenchyma (e.g. pulmonary oedema, acute lung injury) result in mismatch, producing an increase in alveolar deadspace and hypoxaemia. The increased deadspace can be compensated for by increasing overall ventilation. In contrast to the hypoxia resulting from a true right-to-left shunt, that due to areas of low can be partially corrected by administering oxygen and thereby increasing the P_AO₂ even in poorly ventilated areas of lung.

Figure 16.4 The carbon dioxide dissociation curve. Note that in the physiological range the curve is essentially linear.

Mixed venous oxygen tension and saturation ()

The is the partial pressure of oxygen in pulmonary arterial blood that has been thoroughly mixed during its passage through the right heart. Assuming P_aO₂ remains constant, and will fall if more oxygen has to be extracted from each unit volume of blood arriving at the tissues. A low therefore indicates either that oxygen delivery has fallen or that tissue oxygen requirements have increased without a compensatory rise in cardiac output. If falls, the effect of a given degree of pulmonary shunting on arterial oxygenation will be exacerbated. Thus, worsening arterial hypoxaemia does not necessarily indicate a deterioration in pulmonary function but might instead reflect a fall in cardiac output and/or a rise in oxygen consumption.

Conversely, a rise in and may reflect impaired tissue oxygen extraction (due to microcirculatory dysfunction) and/or reduced oxygen utilization (e.g. due to mitochondrial dysfunction) as seen in severe sepsis (see below).

Monitoring the oxygen saturation in central venous (), rather than pulmonary artery blood is less invasive and has been shown to be a valuable guide to the resuscitation of critically ill patients (see p. 891).

Preload

This is defined as the tension of the myocardial fibres at the end of diastole, just before the onset of ventricular contraction, and is therefore related to the degree of stretch of the fibres. As the end-diastolic volume of the ventricle increases, tension in the myocardial fibres is increased and stroke volume rises (Fig. 16.6). Myocardial oxygen consumption () increases only slightly with an increase in preload (produced, for example, by a ‘fluid challenge’, see below) and this is therefore the most efficient way of improving cardiac output.

Figure 16.6 The Frank–Starling relationship: as preload is increased, stroke volume rises. If the ventricle is overstretched, stroke volume will fall (x). In myocardial failure, the curve is depressed and flattened. Increasing contractility, e.g. due to sympathetic stimulation, shifts the curve upwards and to the left (z).

Myocardial contractility

This refers to the ability of the heart to perform work, independent of changes in preload and afterload. The state of myocardial contractility determines the response of the ventricles to changes in preload and afterload. Contractility is often reduced in critically ill patients, as a result of either pre-existing myocardial damage (e.g. ischaemic heart disease), or the acute disease process itself (e.g. sepsis). Changes in myocardial contractility alter the slope and position of the Starling curve; worsening ventricular performance is manifested as a depressed, flattened curve (Fig. 16.6 and Fig. 14.5). Inotropic drugs can be used to increase myocardial contractility (see below).

Afterload

This is defined as the myocardial wall tension developed during systolic ejection. In the case of the left ventricle, the resistance imposed by the aortic valve, the peripheral vascular resistance and the elasticity of the major blood vessels are the major determinants of afterload. Ventricular wall tension will also be increased by ventricular dilatation, an increase in intraventricular pressure or a reduction in ventricular wall thickness.

Decreasing the afterload (exercise, sepsis, vasodilator agents) can increase the stroke volume achieved at a given preload (Fig. 16.7), while reducing . The reduction in wall tension also leads to an increase in coronary blood flow, thereby improving the myocardial oxygen supply/demand ratio. Excessive reductions in afterload will cause hypotension.

Figure 16.7 The effect of changes in afterload on the ventricular function curve. At any given preload, decreasing afterload increases the stroke volume.

Increasing the afterload (increased sympathetic activity, vasoconstrictor agents), on the other hand, can cause a fall in stroke volume and is a potent cause of increased . Right ventricular afterload is normally negligible because the resistance of the pulmonary circulation is very low but is increased in pulmonary hypertension.

Monitoring critically ill patients

As well as allowing immediate recognition of changes in the patient’s condition, monitoring can also be used to establish or confirm a diagnosis, to gauge the severity of the condition, to follow the evolution of the illness, to guide interventions and to assess the response to treatment. Invasive monitoring is generally indicated in the more seriously ill patients and in those who fail to respond to initial treatment. These techniques are, however, associated with a significant risk of complications, as well as additional costs and patient discomfort and should therefore only be used when the potential benefits outweigh the dangers. Likewise, invasive devices should be removed as soon as possible.

Assessment of tissue perfusion

Pale, cold skin, delayed capillary refill and the absence of visible veins in the hands and feet indicate poor perfusion. Although peripheral skin temperature measurements can help clinical evaluation, the earliest compensatory response to hypovolaemia or a low cardiac output, and the last to resolve after resuscitation is vasoconstriction in the splanchnic region.

Metabolic acidosis with raised lactate concentration suggests that tissue perfusion is sufficiently compromised to cause cellular hypoxia and anaerobic glycolysis. Persistent, severe lactic acidosis is associated with a very poor prognosis. In many critically ill patients, especially those with sepsis, however, lactic acidosis can also be caused by metabolic disorders unrelated to tissue hypoxia and can be exacerbated by reduced clearance owing to hepatic or renal dysfunction as well as the administration of adrenaline (epinephrine).

Urinary flow is a sensitive indicator of renal perfusion and haemodynamic performance.

Blood pressure

Alterations in blood pressure are often interpreted as reflecting changes in cardiac output. However, if there is vasoconstriction with a high peripheral resistance, the blood pressure may be normal, even when the cardiac output is reduced. Conversely, the vasodilated patient may be hypotensive, despite a very high cardiac output.

Hypotension jeopardizes perfusion of vital organs. The adequacy of blood pressure in an individual patient must always be assessed in relation to the premorbid value. Blood pressure is traditionally measured using a sphygmomanometer but if rapid alterations are anticipated, continuous monitoring using an intra-arterial cannula is indicated (Practical Box 16.1; Fig. 16.8).

Practical Box 16.1

Radial artery cannulation

Technique

1. The procedure is explained to the patient and, if possible, consent obtained.

2. The arm is supported, with the wrist extended, by an assistant. (Gloves should be worn.)

3. The skin should be cleaned with chlorhexidine.

4. The radial artery is palpated where it arches over the head of the radius.

5. In conscious patients, local anaesthetic is injected to raise a weal over the artery, taking care not to puncture the vessel or obscure its pulsation.

6. A small skin incision is made over the proposed puncture site.

7. A small parallel-sided cannula (20 gauge for adults, 22 gauge for children) is used in order to allow blood flow to continue past the cannula.

8. The cannula is inserted over the point of maximal pulsation and advanced in line with the direction of the vessel at an angle of approximately 30°.

9. ‘Flashback’ of blood into the cannula indicates that the radial artery has been punctured.

10. To ensure that the shoulder of the cannula enters the vessel, the needle and cannula are lowered and advanced a few millimetres into the vessel.

11. The cannula is threaded off the needle into the vessel and the needle withdrawn.

12. The cannula is connected to a non-compliant manometer line filled with saline. This is then connected via a transducer and continuous flush device to a monitor, which records the arterial pressure.

Complications

Thrombosis

Loss of arterial pulsation

Distal ischaemia, e.g. digital necrosis (rare)

Infection

Accidental injection of drugs – can produce vascular occlusion

Disconnection – rapid blood loss.

Figure 16.8 Percutaneous cannulation of the radial artery.

Central venous pressure (CVP)

This provides a fairly simple, but approximate method of gauging the adequacy of a patient’s circulating volume and the contractile state of the myocardium. The absolute value of the CVP is not as useful as its response to a fluid challenge (the infusion of 100–200 mL of fluid over a few minutes) (Fig. 16.9). The hypovolaemic patient will initially respond to transfusion with little or no change in CVP, together with some improvement in cardiovascular function (falling heart rate, rising blood pressure, increased peripheral temperature and urine output). As the normovolaemic state is approached, the CVP usually rises slightly and reaches a plateau, while other cardiovascular values begin to stabilize. At this stage, volume replacement should be slowed, or even stopped, in order to avoid overtransfusion (indicated by an abrupt and sustained rise in CVP, often accompanied by some deterioration in the patient’s condition). In cardiac failure, the venous pressure is usually high; the patient will not improve in response to volume replacement, which will cause a further, sometimes dramatic, rise in CVP.

Figure 16.9 The effects on the central venous pressure (CVP) of a rapid administration of a ‘fluid challenge’ to patients with a CVP within the normal range.

(From Sykes MK. Venous pressure as a clinical indication of adequacy of transfusion. Annals of Royal College of Surgeons of England 1963; 33:185–197.)

Central venous catheters are usually inserted via a percutaneous puncture of the subclavian or internal jugular vein using a guidewire technique (Practical Box 16.2; Figs 16.10, 16.11). The guidewire techniques can also be used in conjunction with a vein dilator for inserting multilumen catheters, double lumen cannulae for haemofiltration or pulmonary artery catheter introducers. The routine use of ultrasound to guide central venous cannulation reduces complication rates.

Practical Box 16.2

Internal jugular vein cannulation

Technique

1. The procedure is explained to the patient and, if possible, consent obtained.

2. The patient is placed head-down to distend the central veins (this facilitates cannulation and minimizes the risk of air embolism but may exacerbate respiratory distress and is dangerous in those with raised intracranial pressure).

3. The skin is cleaned with an antiseptic solution such as chlorhexidine. Sterile precautions are taken throughout the procedure.

4. Local anaesthetic (1% plain lidocaine) is injected intradermally to raise a weal at the apex of a triangle formed by the two heads of sternomastoid with the clavicle at its base.

5. A small incision is made through the weal.

6. The cannula or needle is inserted through the incision and directed laterally downwards and backwards in the direction of the nipple until the vein is punctured just beneath the skin and deep to the lateral head of sternomastoid.

Ultrasound-guided puncture is recommended to reduce the incidence of complications.

7. Check that venous blood is easily aspirated.

8. The cannula is threaded off the needle into the vein or the guidewire is passed through the needle (see Fig. 16.11).

9. The CVP manometer line is connected to a manometer/transducer.

10. A chest X-ray should be taken to verify that the tip of the catheter is in the superior vena cava and to exclude pneumothorax.

Possible complications

Haemorrhage

Accidental arterial puncture (carotid or subclavian)

Pneumothorax

Damage to thoracic duct on left

Air embolism

Thrombosis

Catheter-related sepsis.

Figure 16.10 Cannulation of the right internal jugular vein.

Figure 16.11 Seldinger technique – insertion of a catheter over guidewire. (1) Puncture vessel; (2) advance guidewire; (3) remove needle; (4) dilate vessel; (5) advance catheter over guidewire; (6) remove guidewire; (7) catheter in situ.

The CVP should be read intermittently using a manometer system or continuously using a transducer and bedside monitor. It is essential that the pressure recorded always be related to the level of the right atrium. Various landmarks are advocated (e.g. sternal notch with the patient supine, sternal angle or mid-axilla when the patient is at 45°), but which is chosen is largely immaterial provided it is used consistently in an individual patient. Pressure measurements should be obtained at end-expiration.

The following are common pitfalls in interpreting central venous pressure readings:

Blocked catheter. This results in a sustained high reading, with a damped or absent waveform, which often does not correlate with clinical assessment.

Transducer wrongly positioned. Failure to level the system is a common cause of erroneous readings.

Catheter tip in right ventricle. If the catheter is advanced too far, an unexpectedly high pressure with pronounced oscillations is recorded. This is easily recognized when the waveform is displayed.

FURTHER READING

Ortega R, Song M, Hansen CJ et al. Ultrasound-guided internal jugular vein cannulation. Videos in Medicine. N Engl J Med 2010; 362:e57.

Left atrial pressure

In uncomplicated cases, careful interpretation of the CVP provides a reasonable guide to the filling pressures of both sides of the heart. In many critically ill patients, however, there is a disparity in function between the two ventricles. Most commonly, left ventricular performance is worse, so that the left ventricular function curve is displaced downward and to the right (Fig. 16.12). High right ventricular filling pressures, with normal or low left atrial pressures, are less common but occur with right ventricular dysfunction and in cases where the pulmonary vascular resistance (i.e. right ventricular afterload) is raised, such as in acute respiratory failure and pulmonary embolism.

Figure 16.12 Left ventricular (LV) and right ventricular (RV) function curves in a patient with left ventricular dysfunction. Since the stroke volume of the two ventricles must be the same (except perhaps for a few beats during a period of circulatory adjustment), left atrial pressure (LAP) must be higher than right atrial pressure (RAP). Moreover, an increase in stroke volume (x) produced by expanding the circulatory volume may be associated with a small rise in RAP (y) but a marked increase in LAP (z).

Pulmonary artery pressure

A ‘balloon flotation catheter’ enables reliable catheterization of the pulmonary artery. These ‘Swan–Ganz’ catheters can be inserted centrally (Fig. 16.10) or through the femoral vein, or via a vein in the antecubital fossa. Passage of the catheter from the major veins, through the chambers of the heart, into the pulmonary artery and into the wedged position is monitored and guided by the pressure waveforms recorded from the distal lumen (Practical Box 16.3; Fig. 16.13). A chest X-ray should always be obtained to check the final position of the catheter. In difficult cases, screening with an image intensifier may be required.

Practical Box 16.3

Passage of a pulmonary artery balloon flotation catheter through the chambers of the heart into the ‘wedged’ position

Consent should be obtained if possible from the patient after explanation of the procedure.

Note: (a), (b), (c), (d) refer to Figure 16.13.

1. A balloon flotation catheter is inserted through a large vein (see text).

2. Once in the thorax, respiratory oscillations are seen. The catheter should be advanced further towards the lower superior vena cava/right atrium (a), where pressure oscillations become more pronounced. The balloon should then be inflated and the catheter advanced.

3. When the catheter is in the right ventricle (b), there is no dicrotic notch and the diastolic pressure is close to zero. The patient should be returned to the horizontal, or slightly head-up, position before advancing the catheter further.

4. When the catheter reaches the pulmonary artery (c) a dicrotic notch appears and there is elevation of the diastolic pressure. The catheter should be advanced further with the balloon inflated.

5. Reappearance of a venous waveform indicates that the catheter is ‘wedged’. The balloon is deflated to obtain the pulmonary artery pressure. The balloon is inflated intermittently to obtain the pulmonary artery occlusion (also known as pulmonary artery, or capillary, ‘wedge’) pressure (d).

Figure 16.13 Passage of pulmonary artery balloon flotation catheter through the chambers of the heart into the ‘wedged’ position to measure the pulmonary artery occlusion pressure. (See Practical Box 16.3.)

Once in place, the balloon is deflated and the pulmonary artery mean, systolic and end-diastolic pressures (PAEDP) can be recorded. The pulmonary artery occlusion pressure (PAOP, previously referred to as the pulmonary artery or capillary ‘wedge’ pressure) is measured by reinflating the balloon, thereby propelling the catheter distally until it impacts in a medium-sized pulmonary artery. In this position there is a continuous column of fluid between the distal lumen of the catheter and the left atrium, so that PAOP is usually a reasonable reflection of left atrial pressure.

The technique is generally safe – the majority of complications such as ‘knotting’, valve trauma and pulmonary artery rupture (which can be fatal) are related to user inexperience. Pulmonary artery catheters should preferably be removed within 72 h, since the incidence of complications, especially infection, then increases progressively

Cardiac output

Cardiac output can be continuously monitored using a modified pulmonary artery catheter which transmits low heat energy into the surrounding blood and constructs a ‘thermodilution curve’. These catheters also optically measure and continuously display .

In general, pulmonary artery catheters enable the clinician to optimize cardiac output and oxygen delivery, while minimizing the risk of volume overload. They can also be used to guide the rational use of inotropes and vasoactive agents and are particularly helpful in patients with pulmonary hypertension. There is, however, a considerable body of evidence to suggest that the unselective use of this monitoring device in the absence of evidence-based haemodynamic goals does not lead to improved outcomes and less invasive techniques are increasingly preferred.

Less invasive techniques for assessing cardiac function and guiding volume replacement

Arterial pressure variation as a guide to hypovolaemia

Systolic arterial pressure decreases during the inspiratory phase of intermittent positive pressure ventilation (p. 894). The magnitude of this cyclical variability has been shown to correlate more closely with hypovolaemia than other monitored variables, including CVP. Systolic pressure (or pulse pressure) variation during mechanical ventilation can therefore be used as a simple and reliable guide to the adequacy of the circulatory volume. The response to fluid loading can also easily be predicted by observing the changes in pulse pressure during passive leg raising.

Oesophageal Doppler

Stroke volume, cardiac output and myocardial function can be assessed non-invasively using Doppler ultrasonography. A probe is passed into the oesophagus to continuously monitor velocity waveforms from the descending aorta (Fig. 16.14). Although reasonable estimates of stroke volume, and hence cardiac output can be obtained, the technique is best used for trend analysis rather than for making absolute measurements. Oesophageal Doppler probes can be inserted quickly and easily and are particularly valuable for perioperative optimization of the circulating volume and cardiac performance in the unconscious patient. They are contraindicated in patients with oropharyngeal/oesophageal pathology.

Figure 16.14 Doppler ultrasonography. Velocity waveform traces are obtained using oesophageal Doppler ultrasonography.

Pulse contour analysis

Lithium dilution/pulse contour analysis does not require pulmonary artery catheterization or instrumentation of the oesophagus and is suitable for use in conscious patients. A bolus of lithium chloride is administered via a central venous catheter and the change in arterial plasma lithium concentration is detected by a lithium-sensitive electrode. This sensor can be connected to an existing arterial cannula via a three-way tap. A small battery-powered peristaltic pump is used to create a constant blood flow through the sensor and over the electrode tip. The cardiac output determined in this way can be used to calibrate an arterial pressure waveform (‘pulse contour’) analysis programme that will continuously monitor changes in cardiac output. Devices that use uncalibrated pulse contour analysis to estimate cardiac output are also available. As with pulse pressure variation, stroke volume variation using these devices can accurately predict fluid replacements.

Echocardiography

Echocardiography is being used increasingly often to provide immediate diagnostic information about cardiac structure and function (myocardial contractility, ventricular filling) in the critically ill patient. Transoesophageal echocardiography (TOE) is preferred because of its superior image clarity (Fig. 16.15).

Figure 16.15 Aortic dissection (transoesophageal echocardiography, TOE). (a) Mid-oesophageal, long axis view showing Type A aortic dissection. (b) Short axis view of descending aorta showing intimal flap with false and true lumen.

(From Hinds CJ, Watson JD. Intensive Care: A Concise Textbook, 3rd edn. Edinburgh: Saunders; 2008. Courtesy of Dr C. Rathwell.)

If there is disagreement between clinical signs and a monitored variable, it should be assumed that the monitor is incorrect until all sources of potential error have been checked and eliminated. Changes and trends in monitored variables are more informative than a single reading.

Disturbances of acid–base balance

The physiology of acid–base control is discussed on page 660. Acid–base disturbances can be described in relation to the diagram illustrated in Figure 13.13, p. 663 (which shows P_aCO₂ plotted against arterial [H⁺]).

Both acidosis and alkalosis can occur, each of which are either metabolic (primarily affecting the bicarbonate component of the system) or respiratory (primarily affecting P_aCO₂). Compensatory changes may also be apparent. In clinical practice, arterial [H⁺] values outside the range 18–126 nmol/L (pH 6.9–7.7) are rarely encountered.

Blood gas and acid–base values (normal ranges) are shown in Table 16.2. (For blood gas analysis, see p. 891.)

Table 16.2 Arterial blood gas and acid–base values (normal ranges)

H⁺	35–45 nmol/L	pH 7.35–7.45
PO₂ (breathing room air)	10.6–13.3 kPa	(80–100 mmHg)
PCO₂	4.8–6.1 kPa	(36–46 mmHg)
Base deficit	±2.5
Plasma HCO₃^–	22–26 mmol/L
O₂ saturation	95–100%

Respiratory acidosis. This is caused by retention of carbon dioxide. The P_aCO₂ and [H⁺] rise. A chronically raised P_aCO₂ is compensated by renal retention of bicarbonate, and the [H⁺] returns towards normal. A constant arterial bicarbonate concentration is then usually established within 2–5 days. This represents a primary respiratory acidosis with a compensatory metabolic alkalosis (see p. 666). Common causes of respiratory acidosis include ventilatory failure and COPD (type II respiratory failure where there is a high P_aCO₂ and a low P_aO₂, see p. 814).

Respiratory alkalosis. In this case, the reverse occurs and there is a fall in P_aCO₂ and [H⁺], often with a small reduction in bicarbonate concentration. If hypocarbia persists, some degree of renal compensation may occur, producing a metabolic acidosis, although in practice this is unusual. A respiratory alkalosis may be produced, intentionally or unintentionally, when patients are mechanically ventilated; it may also be seen with hypoxaemic (type I) respiratory failure (see Ch. 15, p. 817), spontaneous hyperventilation and in those living at high altitudes.

Metabolic acidosis (p. 664). This may be due to excessive acid production, most commonly lactate and H⁺ (lactic acidosis) as a consequence of anaerobic metabolism during an episode of shock or following cardiac arrest. A metabolic acidosis may also develop in chronic renal failure or in diabetic ketoacidosis. It can also follow the loss of bicarbonate from the gut or from the kidney in renal tubular acidosis. Respiratory compensation for a metabolic acidosis is usually slightly delayed because the blood–brain barrier initially prevents the respiratory centre from sensing the increased blood [H⁺]. Following this short delay, however, the patient hyperventilates and ‘blows off’ carbon dioxide to produce a compensatory respiratory alkalosis. There is a limit to this respiratory compensation, since in practice values for P_aCO₂ less than about 1.4 kPa (11 mmHg) are rarely achieved. Spontaneous respiratory compensation cannot occur if the patient’s ventilation is controlled or if the respiratory centre is depressed, for example by drugs or head injury.

Metabolic alkalosis. This can be caused by loss of acid, for example from the stomach with nasogastric suction, or in high intestinal obstruction, or excessive administration of absorbable alkali. Overzealous treatment with intravenous sodium bicarbonate is sometimes implicated. Respiratory compensation for a metabolic alkalosis is often slight, and it is rare to encounter a P_aCO₂ above 6.5 kPa (50 mmHg), even with severe alkalosis.

Shock, sepsis and acute disturbances of haemodynamic function

Shock is the term used to describe acute circulatory failure with inadequate or inappropriately distributed tissue perfusion resulting in generalized cellular hypoxia and/or an inability of the cells to utilize oxygen.

Causes of shock

Abnormalities of tissue perfusion can result from:

failure of the heart to act as an effective pump

mechanical impediments to forward flow

loss of circulatory volume

abnormalities of the peripheral circulation.

The causes of shock are shown in Table 16.3. Often shock can result from a combination of these factors (e.g. in sepsis, distributive shock is frequently complicated by hypovolaemia and myocardial depression).

Table 16.3 Causes of shock

Hypovolaemic	Obstructive
Exogenous losses (e.g. haemorrhage, burns) Endogenous losses Cardiogenic (e.g. ischaemic myocardial injury)	Obstruction to outflow (e.g. pulmonary embolus) Restricted cardiac filling (e.g. cardiac tamponade, tension pneumothorax) Distributive (e.g. sepsis, anaphylaxis) Vascular dilatation Sequestration Arteriovenous shunting Maldistribution of flow Myocardial depression