Shock is a syndrome of cardiovascular system failure resulting in inadequate tissue perfusion. Hypotension is a common but not universal feature. Shock is often a feature of patients suffering sepsis, pneumonia, multiple trauma or multiple organ failure. Consequently, it is commonly seen in general intensive care units.

The clinical features vary depending on the cause and the physiological response. Two patterns are typically recognized, although there is a continuum from one to the other:

Hyperdynamic shock: Warm, pink, vasodilated, a tachycardia, with high cardiac output and hypotension. (typical of septic shock, see p. 33)

Hypodynamic shock: Cold, grey, sweaty, vasoconstricted, peripherally shut down with low cardiac output, blood pressure may be relatively well maintained (typical of cardiac shock, see p. 108).

Other features of shock may include an increased or decreased core temperature, hypoventilation or hyperventilation, renal and hepatic dysfunction, disseminated intravascular coagulation, and altered mental status. (See Systemic inflammatory response syndrome, p. 326 and Multiorgan dysfunction syndrome, p. 327.) The young fit patient with an adequate cardiac reserve may well be able to initially compensate for the pathophysiological disturbances leading to shock. Consequently, shock occurring in an otherwise fit young person often represents a greater physiological disturbance or a later stage of illness as compared with the situation in an elderly patient, in whom conditions such as fixed cardiac output, peripheral and coronary artery disease and reduced organ function reduce the ‘physiological reserve’.

The aetiology of shock is frequently multifactorial. Typical causes are listed in Table 4.1. Although all causes of shock are seen on the ICU, the commonest in practice is septic shock (see Septic shock, p. 331).

TABLE 4.1 Typical causes of shock^*

* Note, more than one cause may be present in an individual patient.

Apart from the mechanical causes of shock, in which relief of the mechanical obstruction may take priority, the principles of managing shock states are similar in all patients regardless of aetiology. They include:

treatment of the underlying pathology

optimization of circulating blood volume

optimization of cardiac output

optimization of blood pressure

optimization of oxygen delivery

support for any organ failure.

You should understand the factors that influence tissue perfusion and oxygen delivery, oxygen consumption and cardiac output.

Oxygen delivery is defined as the total amount of oxygen delivered to the tissues (body) per minute. It depends on the cardiac output (CO) and the oxygen content of arterial blood, as shown:

Therefore, ignoring dissolved oxygen, which is insignificant at atmospheric pressure, typical figures are:

Oxygen consumption is the total amount of oxygen consumed by the tissues (body) per minute. It can be calculated from the difference in oxygen content of arterial and mixed venous blood (SO₂).

If cardiac index is used in the above calculations (see below), then oxygen delivery and oxygen consumption can also be expressed as an index relative to body surface area. Typical normal values are shown in Table 4.2.

TABLE 4.2 Typical adult values for oxygen delivery and oxygen consumption

Under normal circumstances, oxygen consumption by the tissues is only about 25% of the oxygen delivered. This provides a large margin for safety, so that if oxygen requirements increase, for example in exercise, more oxygen can be extracted and utilized. In disease states however, while oxygen requirements may be raised, the ability of the tissues to extract and utilize oxygen may be impaired. Under these circumstances, utilization of oxygen by the tissues may be limited by the available supply. Oxygen delivery may be increased by manipulation of cardiac output, improvement in oxygen saturation and transfusion (see below). Oxygen extraction cannot be significantly influenced except by general improvement in the patient’s condition.

From the formula for oxygen delivery above, it follows that oxygen delivery can be improved by:

Ensuring adequate arterial oxygen saturation.

Optimizing haemoglobin (see below).

Optimizing cardiac output (see Optimizing haemodynamic status, p. 78).

If oxygen delivery remains critical, then tracheal intubation and assisted ventilation with paralysis (muscle relaxants) may help to reduce the muscle utilization of oxygen and reduce oxygen requirements.

Since haemoglobin (Hb) carries oxygen to the tissues, one might assume that raising a patient’s Hb to 15 g / dL might provide optimal oxygen delivery. However, clinical studies have not confirmed that raising the haemoglobin concentration improves outcome and indeed, the opposite is likely to be true. Evidence from the Canadian Multicentre Clinical Trials Group suggests that restrictive transfusion strategies produce the best outcome in critically ill patients and that the optimal level in most cases is a Hb of 8–10 g / dL. (See Indications for blood transfusion, p. 246.) This study excluded patients, however, with significant coronary artery and vascular disease, who may benefit from a higher Hb. If in doubt about optimal haemoglobin for an individual patient, seek senior advice.

A guide to the balance between oxygen delivery and oxygen consumption can be obtained by the measurement of mixed venous oxygen saturation. Traditionally mixed venous oxygen was measured in blood from the pulmonary artery via a pulmonary artery catheter, either intermittently (by blood gas measurement) or continuously (using an oximetric catheter.) More recently, measurement of oxygen saturation of blood from a central catheter in the superior vena cava (S_cO₂) has been shown to correlate with the true SO₂ sufficiently closely to be used almost interchangeably in clinical practice.

Resuscitation guided by SO₂ or S_cO₂ has been shown to improve outcome by reducing the severity of organ failure and the duration of intensive care. Normal SO₂ is 55–75%. Values below this imply inadequate oxygen delivery. Values above this imply either supranormal oxygen delivery or reduced oxygen consumption. Conditions that may result in impaired oxygen consumption include sepsis, metabolic poisoning and widespread cellular death.

In a healthy heart there is little change in stroke volume with fluctuations in heart rate occurring within the physiological range (70–160 beats/min). Therefore, as heart rate increases, CO increases. The elderly, those with pre-existing heart disease, and critically ill patients, however tolerate a much narrower range of heart rates, and values outside 100–120 beats/min may significantly compromise CO.

At low heart rates, SV may be maintained but CO falls as a function of heart rate. At high rates, inadequate filling leads to a fall in SV and a subsequent reduction in CO. Tachycardias associated with abnormalities of cardiac rhythm (e.g. atrial fibrillation) further reduce ventricular filling and CO. Tachycardias also lead to increased myocardial oxygen consumption, while simultaneously reducing the time for diastolic perfusion of the ventricles. In patients with ischaemic heart disease, this may produce significant myocardial ischaemia, which may further compromise CO.

The stroke volume is the volume of blood ejected with each heartbeat, and is the difference between the volume of the full ventricle (end diastolic volume) and the volume of the ventricle after ejection of blood is completed (end systolic volume). Traditionally, SV has been thought of as being determined by preload, contractility and afterload.

Preload

This is defined as the ventricular wall tension at the end of diastole. In simple terms preload refers to the degree of ventricular filling. According to the Frank–Starling law of the heart, the greater the degree of ventricular filling, the greater the force of myocardial contraction and thus SV. Above a certain point, however, the ventricle becomes over-stretched and further filling may result in a fall in SV. Heart failure and pulmonary oedema may then develop (see Fig. 4.1).

Fig. 4.1 Diagrammatic representation of Frank–Starling Curves. (a) Increasing LVEDV leads to increased SV. (b) Effect of increased contractility, e.g. as a result of inotropes.

Preload is therefore a function of the volume status of the patient. It also depends upon adequate ventricular relaxation in order to allow ventricular filling to take place. Ventricular relaxation is an active process. In critically ill patients, failure of ventricular relaxation (diastolic failure) may result in the ventricle becoming stiff and non-compliant. When this occurs, the ventricle cannot fill normally, regardless of the patient’s volume status and cardiac output is reduced.

A more recent approach to understanding the interdependency of preload contractility and afterload and the effects thereof, on cardiac output and stroke volume, is to consider pressure–volume–flow loops of the left ventricle (see Fig. 4.2).

Fig. 4.2 (A) Pressure–volume loops of the left ventricle. (B) Effect of increased end diastolic volume (preload) (C) Effect of reduced contractility (S) and diastolic dysfunction (D).

End diastolic volume, which represents ventricular filling, is a function of venous return (pressure) and the diastolic compliance. Increased venous pressure leads to increased ventricular filling and this additional filling is greatest where diastolic compliance is optimal. Where diastolic dysfunction or failure occurs, the diastolic compliance is reduced (with a steeper diastolic compliance curve) and higher venous pressures are required to achieve adequate ventricular filling. Diastolic dysfunction may be seen in hypoxia, myocardial ischaemia, metabolic derangement or as a consequence of mechanical compromise such as pericardial effusion or tamponade.

The end systolic point describes the relationship between the end systolic volume (the volume of blood remaining in the ventricle at the end of systole) and the ejection systolic pressure. This point is determined by a combination of contractility and outflow resistance (afterload). The end systolic point moves upwards and to the right if the ejection systolic pressure is increased (e.g. increased afterload), and downwards to the left if the ejection systolic pressure is reduced (e.g. decreased afterload). Thus under conditions of vasodilation, (e.g. sepsis) there is reduced afterload, lower ejection systolic pressure and reduced end systolic volume. Conversely, vasoconstriction (increased afterload) leads to increased ejection systolic pressure, and increased end systolic volume.

Assuming contractility is unchanged, the ejection systolic point at the end of any heart beat will fall along a curve. Changes in contractility will shift the position of this curve (Fig. 4.2). Increased contractility for example as a result of inotropic drugs, shifts the curve upwards and to the left, so that the same ejection systolic pressure is associated with an increased ejection fraction, greater stroke volume and reduced end systolic volume. Reduced contractility, for example resulting from intrinsic heart disease, or from the myocardial depressant effects of acidosis, hypoxia or sepsis shifts the curve to the right and flattens it. The same systolic blood pressure is associated with a reduced ejection fraction and stroke volume and a much greater end systolic volume.

Arterial cannulation allows beat-to-beat measurement of arterial blood pressure and easy serial blood gas sampling. Significant respiratory variation in the amplitude of the arterial pressure wave (‘respiratory swing’) is characteristic of hypovolaemia. This pulse pressure variability (which relates to stroke volume variability caused by changes in venous return) can be formally measured by modern monitoring systems and is described as a percentage. Pulse variability less than 10% implies adequate filling and has been used as an end point for volume resuscitation. (See Arterial cannulation, p. 372.)

Central venous catheterization provides a route of delivery for drugs and fluids and enables measurement of right heart filling pressures (CVP). Central venous access can be achieved via the jugular, subclavian, brachial or femoral routes. (See Central venous cannulation, p. 376, and Optimizing filling status, p. 376.)

Pulmonary artery (PA) catheterization has for a number of years been the gold standard cardiovascular monitoring tool in ICU. This technique enables the measurement of pulmonary artery pressure, pulmonary artery occlusion pressure (PAOP) and CO, and also allows many other haemodynamic variables to be calculated or derived. Typical values are given in Table 4.4.

TABLE 4.4 Normal values of common haemodynamic variables derived from PA catheterization

These ‘normal values’ provide a guide only. They may not be achievable or appropriate for all critically ill patients (See Goal directed therapy, p. 78).

The value of pulmonary artery catheters has recently been questioned and the technique has been the subject of a number of major multicentre trials leading to a critical re-evaluation of the role of PA catheterization. Recent trials failed to show benefit or harm in a mixed adult ICU population with the suggestion that it should be reserved for the more complicated case where specific questions about the dynamic variables are required to be answered. The use of PA catheterization has fallen significantly with the introduction of alternative forms of monitoring. Relatively non-invasive systems for continuous cardiac output monitoring are available based on transthoracic bioimpedance, oesophageal Doppler, pulse contour and pulse power analysis.

Currently, systems based on pulse contour analysis are perhaps the most likely to enter widespread use. In general terms these require venous access (peripheral or central) and an arterial line with a sensor either built in or attached. To calibrate the system an indicator is injected into the venous catheter and is detected by the arterial line producing a standard dilutional CO measurement. From this the systems are able to provide continuous CO, SV and systemic vascular resistance by analysis of the pulse waveform. To maintain accuracy they must be calibrated every 8–12 h or whenever there is a significant change in cardiovascular status. A popular device using this technique is the Picco system, which combines pulse contour analysis with intermittent calibration by thermodilution. The thermodilution calibration curves can also be used to estimate thermal volumes of distribution in the chest, providing a reasonably well validated measure of global end diastolic volume (a surrogate for filling) and extra vascular lung water, useful in resuscitation and the management of acute lung injury.

A related approach is an analysis of the area under the curve of the arterial waveform. This is known as ‘pulse power analysis’ and has some theoretical advantages over pulse contour analysis. In particular, it does not require proximal arterial cannulation and should in theory be slightly more robust under conditions of damping. It is used commercially in the LiDCO system. This device combines pulse power analysis with calibration using lithium dilution every 12–24 h. An injection of a very low dose of lithium into a vein (not necessarily a central vein) is accompanied by measurement of a dilation curve in an artery. Blood is pumped from the arterial line across an electrode calibrated to measure monovalent cations. In theory, the only such ion whose concentration varies in the time course of the measurement is lithium. Injection of other charged substances at the same time invalidate the measurement. In practice, the only common confounding factor is recent injection of atracurium (a positively charged quaternary nitrogen compound).

A Doppler ultrasound probe is placed in the oesophagus and directed to obtain a signal from the descending aorta. The signal obtained is displayed on the screen and indicates peak velocity and flow time. By making a number of assumptions about the nature of flow in the aorta, the cross-sectional area of the aorta (estimated from body surface area and age) and the percentage of CO passing down the thoracic aorta, SV and CO can be estimated. Trends in values and response to changes in therapy are more useful than absolute values. It is particularly useful for assessing response to fluid challenges but is rather operator-dependent.

Transthoracic echocardiography is minimally invasive and can be repeated at short intervals to assess cardiac function and response to therapy. The key information available from echocardiography is shown in Table 4.5.

TABLE 4.5 Key information available from echocardiography

There is increasing availability of bedside, transthoracic and to a lesser extent transoesophageal, echocardiography in intensive care, for use by suitably trained intensive care staff, as opposed to cardiologists or sonographers. Abbreviated training packages have been developed for this purpose, which enable individuals to undertake focused examinations (e.g. FATE). All these packages accept the limitations of examinations by non-specialists and emphasize the need for formal referral to echocardiography where there is doubt about the findings. Nevertheless it seems likely that echocardiography will increasingly become an extension of the traditional clinical examination of patients in ICU. All patients will be likely to receive a focused echocardiography examination on admission and at intervals and the findings used to guide haemodynamic therapy.

The availability of so many measured haemodynamic variables (above) led to attempts to improve the outcome of critically ill patients by manipulation of the variables to achieve ‘standard goals’. Shoemaker, for example, compared measured variables in trauma survivors and non-survivors and suggested that outcome could be improved by manipulating haemodynamics to achieve supranormal values of cardiac index (4.5 L / min / m²), oxygen delivery index (650 mL / min / m²) and oxygen consumption index (165 mL/min/m²). There is little evidence, however, that this approach improves outcome in the general population of critically ill patients.

The difficulty with population-based targets of this sort is that they risk over- or under-treating individual patients. The more modern approach therefore is to optimize cardiovascular performance of each individual patient by titration of fluids and vasoactive drugs to achieve the ‘optimal response’ for that individual.

Avoid aiming to achieve absolute numbers for CO and other variables. Use ‘normal values’ only as a guide and think in terms of achieving adequate haemodynamic performance for the individual patient. A rational approach is to optimize fluid (filling) status first and then to add an inotrope or vasoconstrictor as required. Figure 4.3 provides a simple algorithm for optimizing haemodynamic status and the management of shock regardless of the underlying cause.

Fig. 4.3 Optimization of the haemodynamic system.

There is a tendency to refer to all vasoactive drugs as inotropes. This is not only incorrect but can lead to confusion when deciding which agent to choose in any given circumstance. By classifying the available agents according to their receptor pharmacology and actions, a rational approach to their use can be achieved. Table 4.6 shows the effects of agonists at various receptors.

TABLE 4.6 Effects of agonists on cardiovascular receptors

On the basis of activity at these receptors, drugs can be classified as inotropes (directly increase cardiac contractility), vasodilators, vasopressors, or a combination of these, e.g. inodilator or inopressor. Classifying agents in this way enables rational choices to be made in selecting agents for use. (See Optimizing cardiac output and Optimizing perfusion pressure below.)