Category Archives: Intensive care medicine

Overall patient management

These critically ill patients require multidisciplinary care with:
• Intensive skilled nursing care (patient/nurse ratio 1:1).
• Regular physiotherapy.
• Careful management of pain and distress with analgesics and sedation as necessary.
• Constant reassurance and support. Critically ill patients easily become disorientated and psychologically disturbed.
• Nutritional support. Enteral nutrition should always be used if possible.
• H2-receptor antagonists or sucralfate (to prevent stress-induced ulceration). They are generally used, but are probably unnecessary in the fed patient.
• TED stockings and subcutaneous heparin to prevent venous thrombosis.
• Care of the mouth, prevention of constipation and of pressure sores.

RESULTS, COSTS AND PATIENT SELECTION

For many critically ill patients, intensive care is undoubtedly life-saving and resumption of a normal life-style is to be expected.
In the most seriously ill patients, however, immediate mortality rates are high, a significant number die soon after discharge from the intensive care unit, and the quality of life for some of those who do survive may be poor. Moreover, intensive care is expensive, particularly for those with the worst prognosis.
Inappropriate use of intensive care facilities has other implications. The patient may experience unnecessary suffering and loss of dignity, while relatives may also have to endure considerable emotional pressures. In some cases treatment may simply prolong the process of dying, or sustain life of dubious quality, and in others the risks of interventions may outweigh the potential benefits. Both for a humane approach to the management of critically ill patients and to ensure that limited resources are used appropriately, it is therefore important to avoid admitting patients who cannot benefit from intensive care and to limit further aggressive therapy when the prognosis is clearly hopeless.
Currently decisions to limit therapy, or not to resuscitate in the event of cardiorespiratory arrest, are made jointly by the medical staff of the unit, the primary physician or surgeon and the nurses, normally in consultation with the patient’s family.

SCORING SYSTEMS

A variety of scoring systems have been developed that canbe used to evaluate the severity of a patient’s illness. These  have included an assessment of the severity of the acute disturbance of physiological function (acute physiology, age, chronic health evaluation-APACHE) and a measure of the therapeutic effort expended on a patient (therapeutic intervention scoring system-TISS). Other systems have been designed for particular categories of patient (e.g. the injury severity score for trauma victims). The APACHE score is widely applicable and has been extensively validated. It can accurately quantify the severity of illness and predict the overall mortality for large groups of critically ill patients, and is therefore useful when auditing a unit’s clinical activity, for comparing results nationally or internationally and as a means of characterizing groups of patients in clinical studies. Although the APACHE methodology can also be used to estimate individual risks of mortality, no scoring system has yet been devised that can predict with certainty the outcome in an individual patient; they must not, therefore, be used in isolation as a basis for limiting or discontinuing treatment.

Further reading

Barton R & Cerra FB (1989) The hypermetabolism multiple organ failure syndrome. Chest 96, 1153-1160. Forrester JS, Ganz W, Diamond G, McHugh T, Chonette DW & Swan HJC (1972) Thermodilution cardiac output determination with a single flow-directed catheter. American Heart Journal 83, 306-311. Hinds CJ & Watson JD (1994) Intensive Care: A Concise Textbook. London: Bailliere Tindall. Jennett B (1982) Brain death. Intensive Care Medicine 8, 1-3.
Knaus WA, Wagner DP, Draper EA, Zimmerman JE, Bergner M, Bastos PG et al. (1991) The APACHE III Prognostic System. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 100, 1619-1636.
Parrillo JE, Parker MM, Natanson C, Suffredini AF, Danner RL, Cunnion RE & Ognibene FP (1990) Septic shock in humans. Advances in the understanding of pathogenesis, cardiovascular dysfunction and therapy. Annals of Internal Medicine 113, 227-242.
Ridley S, Jackson R, Findlay J & Wallace P (1990) Longterm survival after intensive care. British Medical Journa1301,
1127-1130. Stauffer JL, Olson DE & Petty TL (1981) Complications and consequences of endotracheal intubation and tracheostomy.
A prospective study of 150 critically ill adult patients. American Journal of Medicine 70, 65- 76.

Wiener-Kronish JP, Gropper MA & Matthay MA (1990) The adult respiratory distress syndrome: definition and prognosis, pathogenesis and treatment. British Journal of Anaesthesia 65, 107-129.

Brain death means ‘the irreversible loss of the capacity for consciousness combined with the irreversible loss of the capacity to breathe’. Both these are essentially functions of the brain stem. Death, if thought of in this way, can arise either from causes outside the brain (i.e. respiratory and cardiac arrest) or from causes within the head. With the advent of artificial ventilation it became possible to support such a dead patient temporarily, although in all cases cardiovascular failure eventually supervenes and progresses to asystole.
Before considering a diagnosis of brain death it is essential that certain preconditions and exclusions are fulfilled.

Preconditions

• The patient must be in apnoeic coma (i.e. unresponsive and on a ventilator, with no spontaneous respiratory efforts).
• Irremediable structural brain damage due to a disorder that can cause brain stem death must have been diagnosed with certainty (e.g. head injury, intracranial haemorrhage).

Exclusions

• The possibility that unresponsive apnoea is the result of poisons, sedative drugs or neuromuscular blocking agents must be excluded.
• Hypothermia must be excluded as a cause of coma. The central body temperature should be more than 35°C.
• There must be no significant metabolic or endocrine disturbance that could produce or contribute to coma. There should be no profound abnormality of the plasma electrolytes, acid-base balance, or blood glucose levels.

DIAGNOSTIC TESTS FOR THE CONFIRMATION OF BRAIN DEATH

All brain stem reflexes are absent in brain death. The following tests should not be performed in the presence of seizures or abnormal postures.
• The pupils should be fixed and unresponsive to bright light. Both direct and consensual light reflexes should be absent. The size of the pupils is irrelevant, although most often they will be dilated.
• Corneal reflexes should be absent.
• Oculocephalic reflexes should be absent, i.e. when the head is rotated from side to side, the eyes move with the head and therefore remain stationary relative to the orbit. In a comatose patient whose brain stem is intact, the eyes will rotate relative to the orbit (i.e. doll’s eye movements will be present).
• There are no vestibulo-ocular reflexes on caloric testing.
• There should be no motor responses within the cranial nerve territory to painful stimuli applied centrally or peripherally. Spinal reflexes may be present.
• There must be no gag or cough reflex in response to pharyngeal, laryngeal or tracheal stimulation.
• Spontaneous respiration should be absent. The patient should be ventilated with 5% CO2 in 95% O2 for 10 min and then disconnected from the ventilator for a further 10 min. Oxygenation is maintained by insufflation with 100% oxygen at high flow rates via a catheter placed in the endotracheal tube. The patient is observed for any signs of spontaneous respiratory eforts. A blood gas sample should be obtained during this period to ensure that the PaC02 is sufficiently high to stimulate spontaneous respiration (>6.7 kPa [50 mmHg]).
The examination should be performed (and repeated after a few hours) by two doctors of senior status a minimum of 6 hours after the onset of coma or, if due to cardiac arrest, at least 24 hours after restoration of an adequate circulation.
It is not necessary to perform confirmatory tests such as EEG and carotid angiography, as these may be misleading.
In suitable cases, and provided the patient was carrying a donor card and/or the consent of relatives has been obtained, the organs of those in whom brain stem death has been established may be used for transplantation. In all cases in the UK the coroner’s consent must be obtained.

Definition and causes

This syndrome was originally described in 1967 as acute respiratory distress in adults characterized by severe dyspnoea, tachypnoea, cyanosis refractory to oxygen therapy, a reduction in lung compliance and diffuse alveolar infiltrates seen on the chest X-ray. ARDS can therefore be defined as diffuse pulmonary infiltrates, refractory hypoxaernia, stiff lungs and respiratory distress. A PAWP less than 16 mmHg is often included in the definition in an attempt to exclude cardiogenic pulmonary oedema. ARDS can occur as a non-specific reaction of the lungs to a wide variety of insults, including shock (especially septic shock), sepsis, fat embolism, trauma, burns, pancreatitis,  cardiopulmonary bypass, lung contusion, inhalation of smoke or toxic gases, Goodpasture’s syndrome, amniotic fluid embolism and aspiration pneumonia. By far the commonest predisposing factor is sepsis and 20- 40% of patients with severe sepsis will develop ARDS. Pneumonia is a common complication of ARDS.

PATHOPHYSIOLOGY

ARDS can be considered as the earliest manifestation of a generalized inflammatory reaction and is therefore usually associated with the development of MOF.

Non-cardiogenic pulmonary oedema

This is the cardinal feature of ARDS and is the first and clinically most evident sign of a generalized increase in vascular permeability caused by the microcirculatory changes and release of inflammatory mediators described previously. The pulmonary epithelium is also damaged in the early stages of ARDS, reducing surfactant production and lowering the threshold for alveolar flooding.

Pulmonary hypertension

This is a common feature of ARDS. Initially, mechanical obstruction of the pulmonary circulation may occur as a result of vascular compression by interstitial oedema and subsequently oedema of the vessel wall itself. Later, constriction of the pulmonary vasculature may develop in response to increased autonomic nervous activity and circulating substances such as catecholarnines, 5-hydroxytryptamine, thrornboxane, FDPs, complement and activated leucocytes. Those vessels supplying alveoli with low oxygen tensions constrict (the ‘hypoxic vasoconstrictor response’), diverting pulmonary blood flow to better oxygenated areas of lung, thus limiting the degree of shunt.

Haemorrhagic intra-alveolar exudate

This is rich in platelets, fibrin, fibrinogen and clotting factors; fibrin and fibronectin are deposited along the alveolar ducts with the incorporation of cellular debris. This exudate may inactivate surfactant and stimulate inflammation, as well as promoting hyaline membrane formation.

Fibrosis

Within 7 days of the onset of ARDS, formation of a new epithelial lining is underway and activated fibroblasts accumulate in the interstitial spaces. Subsequently, interstitial fibrosis progresses, with loss of elastic tissue and obliteration of the lung vasculature, together with lung destruction and emphysema.

Physiological changes

Shunt and dead space increase, compliance falls and there is evidence of airflow limitation. Although the lungs in ARDS are diffusely injured, the pulmonary lesions, when identified as densities on a CT scan, are predominantly located in dependent regions. This is probably explained by the effects of gravity on the distribution of extravascular lung water and areas of lung collapse.

CLINICAL PRESENTATION

The first sign of the development of ARDS is often an unexplained tachypnoea, followed by increasing hypoxaemia, dyspnoea and laboured breathing. Fine crackles are heard throughout both lung fields. Later, the chest X-ray shows bilateral, diffuse shadowing with an alveolar pattern and air bronchograms that may then progress to the picture of complete ‘white-out’.

MANAGEMENT

This is based on treatment of the underlying condition (e.g. eradication of sepsis) and supportive measures (such as mechanical ventilation). Pulmonary oedema formation should be limited by minimizing left ventricular filling pressure with fluid restriction, the use of diuretics and, if these measures fail to prevent fluid overload, by haemofiltration. The aim should be to achieve a consistently negative fluid balance. If possible plasma oncotic pressure should be maintained by administering colloidal solutions with a long half-life. In patients with ARDS, however, colloids are unlikely to be retained within the vascular compartment; once they enter the interstitial space, the transvascular oncotic gradient is lost and the main determinants of interstitial oedema formation become the microvascular hydrostatic pressure and lymphatic drainage. There is therefore some controversy concerning the relative merits of colloids or crystalloids for volume replacement in patients likely to develop ARDS, or in whom the condition is established. Cardiovascular support and the reduction of oxygen requirements are also important. The administration of high-dose steroids to patients with established ARDS does not appear to improve outcome and current evidence suggests that prophylactic administration to those at risk of developing ARDS is of no value. Moreover, there is a suggestion that steroids may have an adverse effect on the prognosis of ARDS and their use is no longer recommended.

Inhaled nitric oxide

This vasodilator, when inhaled, can improve V/Q matching and oxygenation by increasing perfusion of ventilated lung units, as well as reducing pulmonary hypertension. Its role in the management of ARDS has yet to be established.

Prostacyclin

Although this agent reduces pulmonary and systemic vascular resistance, and consequently improves cardiac output, its use may be complicated by hypotension and a deterioration in gas exchange. Outcome does not seem to be improved.

PROGNOSIS

Despite the treatment outlined, the mortality from established severe ARDS remains high at more than 50% overall. Prognosis is, however, very dependent on aetiology;
when ARDS occurs in association with septic shock mortality rates may be as high as 90%, whereas in ARDS associated with fat embolism around 90% may survive. Approximately 40% of uncomplicated cases die, but the mortality rises with increasing age and failure of other organs such as kidneys and liver. Many of those dying with ARDS now do so as a result of MOF and haemodynamic instability rather than impaired gas exchange.

The respiratory muscles eventually become weak and uncoordinated as they perform no work during conventional mechanical ventilation. Moreover, there is usually some persisting abnormality of lung function. Thus, in patients who have been artificially ventilated for any length of time, spontaneous respiration usually has to be resumed gradually.

Critical illness neuropathy

This recently recognized acquired polyneuropathy has most often been described in association with persistent sepsis and MOF. It is characterized by a primary axonal degeneration involving both motor and sensory nerves. Clinically the initial manifestation is often difficulty in weaning the patient from respiratory support. There is muscle wasting, the limbs are weak and flaccid and deep tendon reflexes are reduced or absent. Cranial nerves are relatively spared.
Nerve conduction studies confirm axonal damage. The cerebrospinal fluid (CSF) protein concentration is normal or minimally elevated. These findings differentiate critical illness neuropathy from Cuillain-Barre syndrome in which nerve conduction studies show evidence of demyelination and CSF protein is usually high.
The cause of critical illness neuropathy is not known and there is no specific treatment. With resolution of the underlying critical illness, complete recovery can be anticipated between 1 and 6 months, although weaning from respiratory support and rehabilitation are likely to be prolonged.

Criteria for weaning patients from artificial ventilation
Clinical assessment is of paramount importance when deciding whether a patient can be weaned from the ventilator. The patient’s conscious level, psychological state, metabolic function, the effects of drugs, cardiovascular performance and mechanical factors must all be taken into account. Objective criteria are based on an assessment of pulmonary gas exchange (blood gas analysis), lung mechanics and muscular strength.

Techniques for weaning

Patients who have received artificial ventilation for less than 24 hours, e.g. elective IPPV after major surgery, can usually resume spontaneous respiration immediately and no weaning process is required. This procedure can also be adopted for those who have been ventilated for longer periods but who clearly fulfil the objective criteria for weaning.
The traditional method of weaning in difficult cases is to allow the patient to breathe entirely spontaneously for a short time, following which [PPV is reinstituted. The periods of spontaneous breathing are gradually increased and the periods of IPPV are reduced. Initially it is usually advisable to ventilate the patient throughout the night. This method can be stressful and tiring both for patients and staff, although some patients do not tolerate IMV (see below) and the traditional method of weaning may then be necessary. SIMV can be used to provide a smoother, more controlled method of weaning; it may also enable weaning to commence at an earlier stage than is possible using the conventional method. There is no evidence, however, that SIMV enables patients who could not be weaned using conventional methods to resume spontaneous respiration, and in some cases the weaning process may be unnecessarily prolonged.
The application of CPAP can prevent the alveolar collapse, hypoxaemia and fall in compliance that might otherwis  occur when patients start to breathe spontaneously. It is therefore often used during weaning with [MV and in spontaneously breathing patients prior to extubation, particularly when they were previously receiving  IPPV with PEEP.

Extubation

his should not be considered until patients can cough, swallow, protect their own airway and are sufficiently alert to be cooperative. Patients are assessed on their ability to breathe spontaneously via the endotracheal tube over a period of time. In those who have undergone prolonged artificial ventilation, this period may need to be 24-48 hours, or even longer, while patients ventilated for less than 12-24 hours can often be extubated within 10- 15 min. During this ‘trial of spontaneous respiration’ the patient should be closely observed for any signs of respiratory distress.

Types and causes

The respiratory system consists of a gas exchanging organ(the lungs) and a ventilatory pump (respiratory  muscleslthorax) either or both of which can fail and precipitate respiratory failure.
Respiratory failure occurs when pulmonary gas exchange is sufficiently impaired to cause hypoxaemia with or without hypercarbia. In practical terms respiratory failure is present when the Pa02 is <8 kPa (60 mmHg) or the Pac02 is >7 kPa (55 mmHg).
It can be divided into:
• Type I respiratory failure, in which the Pa02 is low and the Pac02 is normal or low
• Type II respiratory failure, in which the Pa02 is low and the Pac02 is high
TYPE I or ‘acute hypoxaemic’ respiratory failure ·occurs with diseases that damage lung tissue, with hypoxaemia due to right-to-left shunts or \rlo. mismatch. Common causes include pulmonary oedema, pneumonia, ARDS and, in the chronic situation, pulmonary fibrosing alveolitis.
TYPE II or ‘ventilatory failure’ occurs when alveolar ventilation is insufficient to excrete the volume of carbon dioxide being produced by tissue metabolism. Inadequate  alveolar ventilation is due to reduced ventilatory effort, inability to overcome an increased resistance to ventilation, failure to compensate for an increase in dead space and/or carbon dioxide production, or a combination of these factors. The most common cause is chronic bronchitis and emphysema. Other causes include chest-wall deformities, respiratory muscle weakness (e.g. Guillain- Barre syndrome) and depression of the respiratory centre. Deterioration in the mechanical properties of the lungs and/or chest wall increases the work of breathing and the oxygen consumption/carbon dioxide production of the respiratory muscles. The concept that respiratory muscle fatigue (either acute or chronic) is an important factor in the pathogenesis of respiratory failure is controversial.

MONITORING

A clinical assessment of respiratory distress should be made on the following criteria:
• The use of accessory muscles of respiration
• Tachypnoea
• Tachycardia
• Sweating
• Pulsus paradoxus
• Inability to speak
• Signs of carbon dioxide retention
• Asynchronous respiration (a discrepancy in the rate of movement of the abdominal and thoracic compartments)
• Paradoxical respiration (abdominal and thoracic compartments move in opposite directions)
• Respiratory alternans (breath-to-breath alteration in the relative contribution of intercostal/accessory muscles and the diaphragm)
This can be supplemented by measuring tidal volume and vital capacity. Blood gas analysis should be performed to guide oxygen therapy and to provide an objective assessment of respiratory function.
The most sensitive clinical indicator of increasing respiratory difficulty is a rising respiratory rate. Tidal volume is a less sensitive indicator. Minute ventilation rises initially in acute respiratory failure and falls precipitously only at a late stage when the patient is exhausted. Vital capacity is often a better guide to deterioration and is particularly useful in patients with respiratory inadequacy due to neuromuscular problems, e.g. the Guillain-Barre syndrome, in which the vital capacity decreases as weakness increases.

Pulse oximetry

Lightweight oximeters which measure the changing amount of light transmitted through pulsating arterial blood and provide a continuous, non-invasive assessment of 5a02 can be applied to an ear lobe or finger. These devices are reliable, easy to use and do not require calibration, although it is important to appreciate that pulse oximetry is not a very sensitive guide to changes in oxygenation.

Blood gas analysis

Automation of measurements can give a false impression of reliability and accuracy and may lead to an uncritical acceptance of the results. Errors can result. from m~nction of the analyser or incorrect sampling techniques. Care must be taken over the following.
1 The sample should be analysed immediately or the srringe should be immersed in iced water (the end having first been sealed with a plastic cap) to prevent the continuing metabolism of white cells causing a reduction in P02 and a rise in Pco;
2 The sample must be adequately anticoagulated to prevent clot formation within the analyser. However, excessive dilution of the blood with heparin, which is acidic, will significantly reduce its pH. Heparin (1000 i.u. ml-‘) should just fill the dead space of the syringe, i.e. approximately 0.1 ml. This will adequately anti oagulate a 2 ml sample.
3 Airalmost inevitably enters the sample. The gas tensions within these air bubbles will equilibrate with  hose in the blood, thereby lowering the Pco2, and, usually, raising the P02 of the sample. However, provided the bubbles are ejected immediately by inverting the syringe and expelling the air that rises to the top of the sample, their effect is insignificant. Normal values of blood gas analysis are shown. The interpretation of the results of blood gas analysis can be considered in two separate parts:
• Disturbances of acid-base balance
• Alterations in oxygenation
Interpretation of results requires a knowledge of the history, the age of the patient, the inspired oxygen conc~ntration and any other relevant treatment (e.g. the adrninistration of sodium bicarbonate, and the ventilator settings for those on mechanical ventilation).
Disturbances of acid-base balance The physiology of acid-base control is discu~sed ~n p. 514. Acid-base disturbances can be descnbed in relation to the diagram illustrated, which shows Paco2 plotted against arterial [H+).
Both acidosis and alkalosis can occur, each of which may be either metabolic (primarily affecting the. bicarbonate component of the system) or respiratory (primarily affecting Pco2). Compensatory changes may also be apparent. In clinical practice, arterial [H+) values outside the range 18-126 nmol litre-I (pH 6.9-7.7) are very rarely encountered.

RESPIRATORY ACIDOSIS. This is caused by retention of carbon dioxide. The Paco2 and [H+) rise. A chronically raised PaC02is compensated by renal retention of bicarbonate and the [H+) returns towards normal. A constant arterial bicarbonate concentration is then usually established within 5 days. This represents a primary respiratory acidosis with a compensatory metabolic alkalosis. Common causes of respiratory acidosis include ventilatory failure and chronic bronchitis and emphysema (type II respiratory failure where there is a high Pac02 and a low P.o2).
RESPIRATORY ALKALOSIS. In this case the reverse occurs and there is a fall in Pac02 and [H+), often with a small reduction in bicarbonate concentration. If hypocarbia persists, some degree of renal compensati?n may occur, producing a metabolic acidosis, although in practice this is unusual. A respiratory alkalosis is often produced, intentionally or unintentionally, when patients are artificially ventilated; it may also be seen with hypoxaemic (type I) respiratory failure , sp~ntaneous hyperventilation and in those living at high altitudes.
METABOLIC ACIDOSIS. This may be due to excessive acid production, most commonly lactic acid during a.n episode of shock or following cardiac arrest. A metabolic acidosis may also develop in chronic renal failure and following the loss of large amounts of alkali, e.g. from the gut or from the kidney in renal tubular. acidosis.

Respiratory compensation for a metabolic ~cldosls. IS usually slightly delayed because the blood-brain .barner initially prevents the respiratory centre from sensmg 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 values for Pac02 less than about 1.4 kPa (11 mmHg) are, in practice, never achieved. It should also be noted that respiratory compensation cannot occur if the patient:s ventilation is controlled or if their respiratory centre IS depressed, for example by drugs or head injury.
METABOLIC ALKALOSIS. This can be caused by loss of acid, e.g. from the stomach with nasogastric suction or in high intestinal obstruction, or excessive admi~ist~ation of absorbable alkali. Overzealous treatment With intravenous sodium bicarbonate is frequently implicated. Respiratory compensation for a metabolic alkalosis is often slight and it is rare to encounter a Paco2 >6.5 kPa (50 mmHg , even with severe alkalosis.

Alterations in oxygenation

When interpreting the Pa02 it is important to remember that it is the oxygen content of the arterial blood that matters and that this is determined by the percentage saturation of haemoglobin with oxygen. The relationship between the latter and the P02 is determined by the oxyhaemoglobin dissociation curve. In general, if the saturation is greater than 90%, oxygenation can be considered to be adequate. It must be remembered, however, that on the steep portion of the oxygen dissociation curve small falls in Pa02 will cause significant reductions in oxygen content. Pa02 is also influenced by factors other than pulmonary function, including alterations in Pv02 caused by changes in the metabolic rate and/or cardiac output.

MANAGEMENT

Conventional management of patients with respiratory failure includes the administration of supplemental oxygen, the control of secretions, the treatment of pulmonary infection, the control of airways obstruction and measures to limit pulmonary oedema. Correction of abnormalities which may lead to respiratory muscle weakness, such as hypophosphataemia and malnutrition, is also important.

Oxygen therapy

Methods of oxygen administration

Oxygen is initially given via a face mask. In the majority of patients (except patients with chronic bronchitis and chronically elevated Paco2) the concentration of oxygen given is not important and oxygen can therefore be given by a simple face mask or nasal cannula. With these devices the inspired oxygen concentration varies from about 35 to 55%, with oxygen flow rates of between 6 and lO litres min -I. Nasal cannulae are often preferred because they are less claustrophobic and do not interfere with feeding or speaking, but they can cause ulceration of the nasal or pharyngeal mucosa. should be compared with the fixed performance mask shown, with which the oxygen concentration can be controlled. It is vital to use this latter type of mask in patients with chronic bronchitis and emphysema with chronic type II failure.
Oxygen toxicity Experimentally, mammalian lungs have been shown to be damaged by continuous exposure to high concentrations of oxygen; oxygen toxicity in humans is less well proven. Nevertheless, it is reasonable to assume that high concentrations of oxygen might damage the lungs and so the lowest inspired oxygen concentration compatible with adequate arterial oxygenation should be used. Long-term administration of 50% oxygen or less, or of lOO%oxygen
for less than 24 hours, is probably safe. Dangerous hypoxia should never be tolerated through a fear of oxygen toxicity.

Respiratory support

If, despite the above measures, the patient continues to deteriorate or fails to improve, the institution of some form of respiratory support should be considered. Some of the techniques of respiratory support currently available are shown. Negative-pressure ventilation is now seldom used, but is occasionally employed for long-term ventilation of patients with chronic respiratory failure due to neuromuscular disease or skeletal deformity. The patient’s body is enclosed in an airtight ‘tank’ within which a negative pressure is created intermittently by a separate pump. Cuirass ventilators encase only the thorax.

INTERMITTENT POSITIVE-PRESSURE VENTILATION (IPPV)

IPPV has a number of important advantages over negative- pressure ventilation. In particular, the airway is secured and protected and secretions can be aspirated more easily. In addition, IPPV can be used more successfully in those with diseases involving the lung parenchyma. Furthermore, access to and movement of the patient is relatively unrestricted. A number of refinements and modifications of IPPV have been developed, including IPPV with positive endexpiratory pressure (PEEP), intermittent mandatory ventilation (IMV), and low volume pressure limited inverseratio ventilation. Other techniques include highfrequency jet ventilation (HFJV) and extracorporeal respiratory assistance. These will be discussed later.
The rational use of IPPV depends on a clear understanding of its potential beneficial effects, as well as its dangers.

Beneficial effects

These include:
IMPROVED CARBON DIOXIDE ELIMINATION. By adjusting the volume of ventilation, the PaC02can be returned to within normal limits.
RELIEF FROM EXHAUSTION. Artificial ventilation removes the work of breathing and relieves the extreme exhaustion that may be present in patients with respiratory failure. In some cases, if ventilation is not instituted, this exhaustion may culminate in respiratory arrest.
EFFECTS ON OXYGENATION. In those with severe pulmonary parenchymal disease, the lungs may be very stiff and the work of breathing is therefore greatly increased. Under these circumstances the institution of IPPV may significantly reduce total body oxygen consumption; consequently Pv02′ and thus Pao2, may improve. Because ventilated patients are connected to a leak-free circuit it is possible to administer high concentrations of oxygen (up to 100%) accurately and to apply a positive end-expiratory pressure. In selected cases the latter may reduce shunting and increase Pao2 .

Indications

ACUTE RESPIRATORY FAILURE with signs of severe respiratory distress (e.g. respiratory rate >40 min -1, inability to speak, patient exhausted) persisting despite maximal therapy. Confusion, restlessness, agitation, a decreased conscious level, a rising PaC02 and extreme hypoxaemia are further indications. Care should be taken before ventilating patients with chronic lung disease as patients previously severely incapacitated will be difficult to wean off the ventilator and also relapse early. The most important criteria are the patient’s previous exercise tolerance and ability to lead an independent existence.
ACUTE VENTILATORY FAILURE due, for example, to myasthenia gravis or Guillain-Barre syndrome. Artificial ventilation should be instituted when the vital capacity has fallen to 10–15 ml kg-I. This will avoid complications such as atelectasis and infection as well as preventing respiratory arrest. The tidal volume and respiratory rate are relatively insensitive in the above conditions and change late in the course of the disease. A high Pac02 (particularly if rising) is an indication for urgent artificial ventilation.
OTHER INDICATIONS include:
• Prophylactic postoperative ventilation in poor risk patients
• Head injury-to avoid hypoxia and hypercarbia which increase cerebral blood flow and intracranial pressure, hyperventilation to reduce intracranial pressure
• Trauma-chest injury and lung contusion
• Severe left ventricular failure with pulmonary oedema
• Coma with breathing difficulties, e.g. following drug overdose

Institution

IPPV requires endotracheal intubation. If the patient is conscious the procedure must be fully explained before anaesthesia is induced. Intubating patients in severe respiratory failure is an extremely hazardous undertaking and should only be performed by experienced staff. In extreme emergencies it may be preferable to ventilate the patient by hand using an oropharyngeal airway, a face mask and a self-inflating bag until experienced help arrives. The patient is usually hypoxic and hypercarbic, with increased sympathetic activity, and the stimulus of laryngoscopy and intubation can precipitate dangerous arrhythmias and even cardiac arrest. If possible, therefore, the ECG and oxygen saturation should be monitored, and the patient preoxygenated with 100% oxygen before intubation. In some deeply comatose patients, no sedation will be required, but in the majority of patients a shortacting intravenous anaesthetic agent followed by muscle relaxation will be necessary.
The complications of endotracheal intubation are given.
Endotracheal tubes can now safely be left in place forseveral weeks and tracheostomy is therefore less often performed. Tracheostomy may be required for the longterm control of excessive bronchial secretions, particularlyin those with a reduced conscious level, and/or to  maintain an airway and protect the lungs in those with impaired pharyngeal and laryngeal reflexes. A life-threatening obstruction of the upper respiratory tract that cannot be bypassed with an endotracheal tube should have a cricothyroidotomy, which is safer, quicker and easier to perform. Tracheostomy performed under these circumstances can be extremely hazardous. Other indications are head and neck injuries, including burns to the face and upper airway. Tracheostomy has a mortality rate of up to 3%. Complications of tracheostomy are shown.
Minitracheostomy involves inserting a small diameter tube percutaneously into the trachea via the cricothyroid membrane using a guide wire. It can be performed under local anaesthesia. This technique facilitates the clearance of copious secretions in those who are unable to cough effectively but can protect their airway.

Dangers of IPPY

General dangers include:
COMPLICATIONS of endotracheal intubation or tracheostomy.
DISCONNECTION OR MECHANICAL FAILURE. These are unusual but dangerous. A method of manual ventilation, e.g. a self-inflating bag, and oxygen must always be available by the bedside.
BAROTRAUMA. Overdistension of the lungs during IPPV can rupture alveoli and cause air to dissect centrally along the perivascular sheaths. This pulmonary interstitial air can sometimes be seen on chest X-ray as linear or circular perivascular collections or subpleural blebs. Other complications are pneumothorax, pneumomedi astinum, pneumoperitoneum and subcutaneous emphysema. Intra-abdominal air originating from the alveoli is probably always associated with pneumomediastinum.
The incidence of barotrauma is greatest in those patients who require high inflation pressures, with or without a positive end-expiratory pressure, and the risk of pneumothorax is increased in those with destructive lung disease (e.g. staphylococcal pneumonia,
emphysema), asthma or fractured ribs. A tension pneumothorax can be rapiclly fatal in ventilated patients with respiratory failure. Suggestive signs include the development or worsening of hypoxia,
fighting the ventilator, an unexplained increase in inflation pressure, as well as hypotension and tachycardia, sometimes accompanied by a rising CVP. Examination may reveal unequal chest expansion, mediastinal shift (deviated trachea, displaced apex beat) and a hyperresonant hemithorax. Although, traditionally, breath sounds are diminished over the pneumothorax,
this sign can be extremely misleading in ventilated patients. If there is time, the diagnosis can be confirmed by chest X-ray.
Other dangers of IPPV include:
RESPIRATORY COMPLICATIONS. IPPV is frequently complicated by a deterioration in gas exchange due to VIO mismatch and collapse of peripheral alveoli. The latter can largely be prevented by using large tidal volumes (l0-15 rnl kg-I) and reducing the respiratory  rate (usually to 10-12 min-I) to avoid hypocarbia or by the application of PEEP (see below). Secondary pulmonary infection is a common complication of IPPV and high inflation pressure may contribute to ‘ventilation induced’ lung injury.
~ARDIOVASCULARCOMPLICATIONS. The intermittent application of positive pressure to the lungs and thoracic wall impedes venous return and distends alveoli, thereby ‘stretching’ the pulmonary capillaries and causing a rise in pulmonary vascular resistance. Both these mechanisms can produce a fall in cardiac output. In normal subjects, the fall in cardiac output is prevented by constriction of capacitance vessels, which restores venous return. Hypovolaemia, pre-existing pulmonary hypertension, right ventricular failure and autonomic dysfunction (as may be present in those with Guillain-Barre syndrome, acute spinal cord injury or diabetes) will exacerbate the haemodynamic disturbance. Expansion of the circulating volume, on the other hand, can often restore cardiac output. In patients with heart failure, cardiac output and blood pressure are usually unaffected, or even increased by positive pressure ventilation. Therefore, IPPV should be used without hesitation in patients with cardiogenic pulmonary oedema who have severe respiratory distress and exhaustion.
GASTROINTESTINAL COMPLICATIONS. Initially, many artificially ventilated patients will develop abdominal distension associated with an ileus. The cause is unknown, although the use of non-depolarizing neuromuscular blocking agents and opiates may in part be responsible.
SALT AND WATER RETENTION. IPPV, particularly with PEEP, causes increased ADH secretion and possibly a reduction in circulating levels of atrial natriuretic peptide. Combined with a fall in cardiac output and a reduction in renal cortical blood flow, these can cause salt and water retention. This fluid retention is often particularly noticeable in the lungs.

Positive end-expiratory pressure (PEEP)

A positive airway pressure can be maintained at a chosen level throughout expiration by attaching a threshold resistor valve to the expiratory limb of the circuit. PEEP should be considered if it proves impossible to achieve adequate oxygenation of arterial blood (more than 90% saturation) using conventional positive-pressure ventilation without raising the inspired oxygen concentration to potentially dangerous levels (conventionally 50%). PEEP is not, however, a panacea for all patients who are hypoxic and, indeed, it may often be detrimental, not least because the use of levels of PEEP in excess of 5 cmH20 is associated with an increased risk of barotrauma. Most recommend that pressures in excess of 15- 20 cmH20 should not be exceeded. The primary effect of PEEP is to re-expand underventilated lung units thereby reducing shunt and increasing the Pa02.
Unfortunately, the inevitable rise in mean intrathoracic pressure that follows the application of PEEP may further impede venous return, increase pulmonary vascular resistance and thus reduce cardiac output. This effect is probably least when the lungs are stiff. The fall in cardiac output can be ameliorated by expanding the circulating  volume, although in some cases inotropic support may be required. Thus, although arterial oxygenation is often improved by the application of PEEP, a simultaneous fallin cardiac output can lead to a reduction in total oxygen  delivery.

OTHER TECHNIQUES FOR RESPIRATORY SUPPORT

Continuous positive airway pressure (CPAP)

The application of CPAP achieves for the spontaneously breathing patient what PEEP does for the ventilated patient. Oxygen and air are delivered under pressure via an endotracheal tube or via a tightly fitted face mask. Not only can it improve oxygenation but the lungs become less stiff, breathing becomes easier and vital capacity
improves.

Intermittent mandatory ventilation

This technique allows the patient to breathe spontaneously between the ‘mandatory’ tidal volumes delivered by the ventilator. It is important that these mandatorybreaths are timed to coincide with the patient’s own inspiratory effort (synchronized IMV:SIMV). SIMV can  be used with or without PEEP or CPAP. It was originally introduced as a technique for weaning patients from artificial ventilation but is now used extensively as an alternative to conventional IPPV. Spontaneous respiration may be assisted during SIMV by applying a constant preset airway pressure at the start of inspiration (‘pressure support’). The level of pressure support can be reduced as the patient improves.

High-frequency jet ventilation

Adequate oxygenation -and CO2 elimination can be achieved by injecting gas into the trachea at rates of up to several thousand breaths per minute. In clinical practice rates of between 60 and 300 breaths per minute are usually employed.
Potential advantages of HFJV are largely related to the low peak airway pressures; for example the risk of barotrauma is reduced. Moreover, HFJV can be used to ventilate patients with large air leaks due, for example, to a bronchopleural fistula or lung lacerations. The place of HFJV in the management of patients with acute respiratory failure is less clear.
Low volume, pressure limited inverse-ratio mechanical ventilation. A constant preset inspiratory pressure is delivered for a prescribed time, generating low tidal volumes and reducing peak inspiratory pressure. Respiratory rate is increased in order to achieve adequate carbon dioxide removal. When combined with a prolonged inspiratory  time and low level PEEP this technique may provide optimal oxygenation whilst minimising the high peak airway pressures which are thought to exacerbate pulmonary damage. Hypercarbia is inevitable but should be accepted (‘permissive hypercarbia’).

Extracorporeal respiratory assistance

Recently there has been renewed interest in the use of extracorporeal gas exchange to reduce ventilation requirements and ‘rest’ the lungs. Carbon dioxide is removed using low flow veno-venous bypass through a membrane lung. The combination of normal pulmonary perfusion and minimum ventilation may reduce barotrauma and provide optimal conditions for lung healing. The indications for the use of this demanding technique are at present unclear.

Acute renal failure is a common and serious complication of critical illness which adversely affects the prognosis. The importance of preventing renal failure by rapid and effective resuscitation, as well as the avoidance of nephrotoxic drugs (especially NSAIDs), cannot be overemphasized. Shock and sepsis are the commonest causes of acute renal failure in the critically ill but it remains important to diagnose the cause of renal dysfunction and exclude reversible pathology, especially obstruction.
Oliguria is usually the first indication of renal impairment and should prompt immediate attempts to optimize cardiovascular function, particularly by expanding the circulating volume. Low-dose dopamine may be used to enhance renal blood flow. If these measures fail to reverse oliguria some recommend administration of diuretics such as frusemide or mannitol.
If oliguria persists, it is important to reduce crystalloid intake and review drug doses. Dialysis is indicated for fluid overload, electrolyte disturbances (especially hyperkalaemia), acidosis and, to a lesser extent, uraemia. Intermittent haemodialysis has a number of disadvantages in the critically ill. In particular it is frequently complicated by hypotension and it may be difficult to remove sufficient volumes of fluid. Peritoneal dialysis is also frequently unsatisfactory in these patients and is contraindicated in those who have undergone intra-abdominal surgery. The use of continuous haemofiltration, usually with dialysis, is therefore preferred.

Diuretics increase salt and water excretion by the kidneys, thereby decreasing ventricular filling pressure (pre-load). This is the major form of therapy in sodium retention with fluid overload.

Vasodilator therapy

In selected cases, after-load reduction may be used to increase stroke volume and decrease myocardial oxygen requirements by reducing the systolic ventricular wall tension. Vasodilatation also decreases heart size and the diastolic ventricular wall tension so that coronary blood flow is improved. The relative magnitude of the falls in pre-load and after-load depends on the pre-existing haemodynamic disturbance, concurrent volume replacement and the agent selected.
Vasodilator therapy is most beneficial in patients with cardiac failure in whom the ventricular function curve is flat and falls in pre-load have only a limited effect on stroke volume. This form of treatment may therefore sometimes be useful in cardiogenic shock and in the management of patients with pulmonary oedema associated with low cardiac output. Vasodilators may also be valuable in shocked patients who remain vasoconstricted and oliguric despite restoration of an adequate blood pressure.
Such therapy is potentially dangerous and should be guided by continuous haemodynamic monitoring, including pulmonary artery catheterization or direct measurement of left atrial pressure. The circulating volume must be adequate before treatment is started. Falls in pre-load should be prevented, except in those with cardiac failure, in order to avoid serious reductions in cardiac output and blood pressure. If diastolic pressure is allowed to fall, coronary blood flow may be jeopardized and, particularly if a reflex tachycardia develops in response to the hypotension, myocardial ischaemia may be precipitated.
HYDRALAZINE. This predominantly affects arterial resistance vessels. It therefore reduces after-load and blood pressure, while cardiac output and heart rate usually increase. Hydralazine is usually given as an intravenous bolus to control acute increases in blood pressure, particularly after cardiac surgery.

A-Adrenergic antagonists

These predominantly dilate arterioles and therefore mainly influence after-load.
Phenoxybenzamine is unsuitable for use in the critically ill because of its slow onset (1-2 hours to maximum effect) and prolonged duration of action (2-3 days). Phentolamine is very potent with a rapid onset and short duration of action (15-20 min). It can be used for short-term control of blood pressure in a hypertensive crisis, but can produce a marked tachycardia. Vasodilators acting directly on the vessel wall These agents are those most commonly used to achieve vasodilatation in the critically ill.
SODIUM NITROPRUSSIDE (SNP). This dilates arterioles and venous capacitance vessels, as well as the pulmonary vasculature by donating nitric oxide. SNP therefore reduces the after-load and pre-load of both ventricles and can improve cardiac output and the myocardial oxygen supply-demand ratio. It has been suggested that SNP can exacerbate myocardial ischaemia by producing a ‘steal’ phenomenon in the coronary circulation. The effects of SNP are rapid in onset and spontaneously reversible within a few minutes of discontinuing the infusion.
A large overdose of SNP can cause cyanide poisoning, with intracellular hypoxia caused by inhibition of cytochrome oxidase, the terminal enzyme of the respiratory chain. This is manifested as a metabolic acidosis and a fall in the arteriovenous oxygen content difference.
NITROGLYCERINE (NTG) AND ISOSORBIDE DINITRATE (I SDN). These are both predominantly venodilators. They can therefore cause marked reductions in pre-load, which may be associated with falls in cardiac output and compensatory vasoconstriction. For the reasons discussed above, they are of most value in those with cardiac failure in whom pre-load reduction may reduce ventricular wall tension and improve coronary perfusion without adversely affecting cardiac performance.
Furthermore, these agents may reverse myocardial ischaemia by increasing and redistributing coronary blood flow. They are therefore often used in preference to SNP in patients with cardiac failure and/or myocardial ischaemia. Both NTG and ISDN reduce pulmonary vascular resistance by donating nitric oxide, an effect that can occasionally be exploited in patients with a low cardiac output secondary to pulmonary hypertension. Mechanical support of the myocardium
Intra-aortic balloon counterpulsation (IABep) is the most widely used technique for mechanical support of the failing myocardium. It is discussed.

Adjunctive therapy in shock

Attempts have been made to identify agents that would prevent the release, or inhibit the effects, of the various mediators released in shock. For example, non-steroidal anti-inflammatory drugs (NSAIDs) (which inhibit cyclooxygenase) have been used to limit prostaglandin production, naloxone has been used to block the effects of endogenous opioid peptides, PAF antagonists are available and monoclonal antibodies to some of the cytokines or their receptors, as well as endotoxin itself, have been developed and investigated. Recently there has been considerable interest in the ability of NO synthase inhibitors to reverse the vasodilatation associated with some forms of circulatory shock. Other approaches have included administration of prostacyclin and removal of mediators by plasma exchange/haemofiltration. In the future naturally occurring cytokine antagonists or their soluble receptors may prove useful. At present, however, the role of these various adjunctive therapies in clinical practice remains unclear.
In animal studies very large doses of steroids have been shown to reduce mortality in septic shock, but clinical trials in humans have shown that steroids are of no benefit and their administration to such patients is no longer recommended.

BLOOD. This is conventionally given for haemorrhagic shock as soon as it is available. In extreme emergencies, uncross matched group 0 negative blood can be used, but an emergency crossmatch can be performed in about 30 min and is as safe as the standard procedure. Donor blood is often separated into its various components for storage, necessitating the transfusion of packed red cells to maintain haemoglobin and plasma, or a plasma substitute, for volume replacement.
Complications of blood transfusion are discussed. Special problems arise when large volumes of stored blood are transfused rapidly. These include:
TEMPERATURE CHANGES. Bank blood is stored at 4°C and transfusion may result in hypothermia, peripheral venoconstriction (which slows the rate of the infusion) and arrhythmias. Some therefore recommend that if possible blood should be warmed prior to the transfusion.
COAGULOPATHY. Stored blood has essentially no effective platelets and is deficient in clotting factors. Large transfusions can therefore produce a coagulation defect. This may need to be treated by replacing clotting factors with fresh frozen plasma and the administration of platelet concentrates.
METABOLIC ACIDOSIS/ALKALOSIS. Stored blood is now preserved in citrate/phosphate/dextrose (CPD) solution, which is less acidic than the acid/citrate/dextrose (ACD) solution used previously. Metabolic acidosis attributable solely to blood transfusion is rare and in any case rarely requires correction. A metabolic alkalosis often develops 24-48 hours after a large blood transfusion, probably mainly due to metabolism of the citrate; this will be exacerbated if the preceding acidosis has been corrected with intravenous sodium bicarbonate.
HYPOCALCAEMIA. Stored blood is anticoagulated with citrate, which binds calcium ions. This can reduce total body ionized calcium levels and cause myocardial depression. This is uncommon in practice, but if necessary can be corrected by administering 10 ml of 10% calcium chloride intravenously. Routine treatment with calcium is not recommended.
INCREASED OXYGEN AFFINITY. In stored blood, the red cell 2,3-DPG content is reduced, so that the oxyhaemoglobin dissociation curve is shifted to the left. The oxygen affinity of haemoglobin is therefore increased and oxygen delivery is impaired. This effect is less marked with blood stored in CPD. Red cell levels of 2,3-DPG are substantially restored within 12 hours of transfusion.
HYPERKALAEMIA. Plasma potassium levels rise progressively as blood is stored. However, hyperkalaemia is rarely a problem as prewarming of the blood increases red cell metabolism-the sodium pump becomes active and potassium levels fall.
MICROEMBOLISM. Microaggregates in stored blood may be filtered out by the pulmonary capillaries. This process is thought by some to contribute to ARDS.
RED CELL CONCENTRATES. Nutrient additive solutions, i.e. saline, adenine, glucose and mannitol (SAGM), are now available which allow red cell storage in the absence of plasma.
Because of the complications of blood transfusion, in particular the risk of disease transmission, as well as their expense, the use of crystalloid solutions, plasma, plasma substitutes and oxygen-carrying solutions for volume replacement is assuming greater importance.
CRYSTALLOID SOLUTIONS. Although crystalloid solutions, e.g. saline, are cheap, convenient to use and free of side-effects, the administration of large volumes of these fluids to critically ill patients should, in general, be avoided. They are rapidly lost from the circulation into the extravascular spaces and volumes of crystalloid two to four times that of colloid are required to achieve an equivalent haemodynamic response. Although volume replacement with predominantly crystalloid solutions is advocated by some for the uncomplicated, previously healthy patient with traumatic or perioperative hypovolaemia, a more reasonable approach is to use crystalloids initially but to use colloids in addition if there is continued need for volume replacement in excess of about 1 litre.
COLLOIDAL SOLUTIONS. These produce a greater, and more sustained increase in plasma volume, with associated improvements in cardiovascular function and oxygen transport. They also increase colloid osmotic pressure.
Human albumin solution (HAS) is a natural colloid and is not generally used for routine volume replacement, particularly if volume losses are continuing since other cheaper solutions are equally effective in the short term. Some recommend administration of HAS at a later stage in those who are hypoalbuminaemic.
Dextrans are polymolecular polysaccharides in either 5% dextrose or normal saline. They are commercially available as low molecular weight dextran (dextran 40; mol. wt 40000) and dextran 70, and have a powerful osmotic effect. They interfere with cross matching and have a small rate of allergic reactions (0.07-1.1%). Normally a dose of 1.5 g dextran per kilogram of body weight should not be exceeded because of the risk of renal damage. In practice dextrans are rarely used in the UK because of the availability of other agents.
Polygelatin solutions (Haemaccel, Gelofusine) have an average molecular weight of 35 000, which is iso-osmotic with plasma. They are cheap. Large volumes can be administered, since coagulation defects do not occur and renal function is not impaired. However, because theyreadily cross the glomerular basement membrane, their
half-life in the circulation  is approximately 4 hours and they can promote an osmotic diuresis. Allergic reactions occur in up to 10% of cases. These solutions are particularly useful during the acute phase of resuscitation, especially when volume losses are continuing but in many patients colloids with a longer half-life will be required later to achieve haemodynamic stability. Hydroxyethyl starch (HES) has a mean molecular weight of approximately 450000 and a half-life of about 6 hours. Volume expansion is equivalent to, or slightly greater than, the volume infused. The incidence of allergic reactions is approximately 0.1%. Although more expensive than gelatins, HES is a valuable volume expander.
OXYGEN-CARRYING BLOOD SUBSTITUTES are being developed, e.g. fluorocarbon emulsions. Haemoglobin solutions also have potential as oxygendelivering resuscitation fluids, but their use is limited by their short intravascular retention time and their high affinity for oxygen.

Myocardial contractility and inotropic agents

Myocardial contractility can be impaired by hypoxaemia and hypocalcaemia, as well as by some drugs (e.g. (3- blockers, antiarrhythmics and sedatives). Severe lactic acidosis can depress myocardial contractility and may limit the response to inotropes. Attempted correction of acidosis with intravenous sodium bicarbonate, however, generates additional carbon dioxide which diffuses across cell membranes producing or exacerbating intracellular acidosis. Other disadvantages of bicarbonate therapy include sodium overload and a left shift of the oxyhaemoglobin dissociation curve. Also ionized calcium levels may be reduced and, combined with the fall in intracellular pH, may be responsible for impairing myocardial performance. Treatment of lactic acidosis should therefore concentrate on correcting the cause, while acidosis may be most safely controlled by hyperventilation. Bicarbonate should only be administered to correct extreme and persistent metabolic acidosis.
When a patient remains hypotensive despite adequate volume replacement, and perfusion of vital organs is jeopardized, pressor agents may be administered to improve cardiac output and blood pressure. In some cases inotropic agents are given to redistribute blood flow (e.g. dopamine can be used to increase renal perfusionsee below). There is evidence that survival of patients with septic or traumatic shock (as well as following major surgery and in those with ARDS) is associated with supranormal values for cardiac output, oxygen delivery and oxygen consumption. Some now advocate that in the most severely ill patients, and in those who fail to respond to simple measures, treatment should be directed at increasing these variables until they equal or exceed the median values found in survivors.
It must be remembered, however, that all inotropes increase myocardial oxygen consumption, particularly if a tachycardia develops, and that this can lead to an imbalance between myocardial oxygen supply and demand, with the development or extension of ischaemic areas. For this reason such agents should be used with caution, particularly in cardiogenic shock following myocardial infarction and those known to have ischaemic heart disease.
All inotropic agents should be administered via a large central vein, and their effects carefully monitored. Some of the currently available inotropes are considered here.

Adrenaline

Adrenaline stimulates both 0′- and {3-adrenergic receptors, but at low doses {3 effects seem to predominate. This produces a tachycardia, with an increase in cardiac index and a fall in peripheral resistance. At higher doses, 0′- mediated vasoconstriction develops. If this produces a useful increase in perfusion pressure, urine output may increase and renal failure may be avoided. However, as the dose is further increased, cardiac output may actually fall, accompanied by marked vasoconstriction, tachycardia and a metabolic acidosis. A reduction in renal blood flow then occurs, with oliguria and a risk of acute renal failure. Prolonged high-dose administration may eventually cause peripheral gangrene. For these reasons the minimum effective dose should be used for as short a time as possible. The addition of low-dose dopamine to the regimen may help to preserve renal function. Despite its disadvantages, adrenaline remains a useful potent inotrope and is used when other agents have failed. When haemodynamic monitoring is not available adrenaline is probably the agent of choice in septic shock, and should probably be combined with lowdose dopamine.

Noradrenaline

This is predominantly an a-adrenergic agonist. It can be of value in those with severe hypotension associated with a low systemic resistance, for example in septic shock. There is a risk of producing excessive vasoconstriction with impaired organ perfusion and increased after-load. oradrenaline administration must therefore be accompanied by full haemodynamic monitoring, including determination of cardiac output and calculation of the peripheral resistance.

Isoprenaline

This f3-adrenergic stimulant has both inotropic andchronotropic effects. It reduces peripheral resistance by  dilating skin and muscle blood vessels and diverts flow away from vital organs such as the kidneys. The increase in cardiac output produced by isoprenaline is mainly due to the tachycardia, and this, together with the development of arrhythmias, seriously limits its value. There are now few indications for isoprenaline in the criticially ill adult.

Dopamine

This is a natural precursor of noradrenaline which acts on D1 and D2 dopamine receptors, as well as a- and 13- adrenergic receptors. Its main action (at a dose of 3-10 p,g kg:’ min-I) is on f3l-adrenoreceptors on cardiac muscle, increasing cardiac contractility without increasing rate. The dosage, however, is critical.
IN LOW DOSES (1-3 p,g kg” min-‘) dopaminergic vasodilatory receptors in the renal, mesenteric, cerebral and coronary circulations are activated. D1 receptors are located on postsynaptic membranes and mediate vasodilatory effects whilst D2 receptors are presynaptic and potentiate these vasodilatory effects by preventing the release of noradrenaline. This results in an increase in renal plasma flow and glomerular filtration rate with an improved urinary output. It also increases hepatic blood flow.
MEDIUM DOSES (3-10 p,g kg'” min-I) activation of f3l-adrenoreceptors occurs with an increase in heart rate, myocardial contractility and cardiac output. IN HIGH DOSES (>20 p,g kg-I min-‘) dopamine increases noradrenaline and therefore activates aladrenergic receptors leading to vasoconstriction. This causes an increase in after-load and raises the ventricular filling pressures.

Dopexamine

This analogue of dopamine is a vasodilator (f32-agonist) that does not stimulate a-receptors. It is a positive inotrope and increases renal blood flow, but in septic shock may exacerbate hypotension by further reducing systemic vascular resistance. It is likely to be most useful in those with low cardiac output and peripheral vasoconstriction.

Dobutamine

Dobutamine is closely related to dopamine with predominant 131activity and less a constricting activity, but equal positive ‘inotropic’ effect.
• It has no specific effect on the renal vasculature although urine output often increases as cardiac output and blood pressure improve.
• It reduces systemic resistance and improves cardiac performance, thereby decreasing both after-load and ventricular filling pressures.
• It produces a greater improvement in cardiac output than dopamine for a given increase in myocardial oxygen consumption.
For these reasons, dobutamine is probably the agent of choice in patients with cardiogenic shock and cardiac failure.

Enoximone

This agent, active both orally and intravenously, is a phosphodiesterase inhibitor with inotropic and vasodilator properties. Enoximone may, however, cause profound vasodilatation and precipitate or worsen hypotension due to profound vasodilatation. It is sometimes useful in the management of acute cardiac failure.

Summary

Many still consider dopamine to be the inotrope of choice in most critically ill patients, largely because of its effects on splanchnic blood flow, although others favour dopexamine as a means of increasing cardiac output and organ blood flow. Dobutamine is equally popular and is particularly indicated in patients in whom the vasoconstriction caused by dopamine could be dangerous (i.e. patients with cardiac disease and septic patients with fluid overload or myocardial failure). The combination of dobutamine and noradrenaline is currently popular for the management of patients who are shocked with a low systemic resistance. Dobutamine is given to achieve an optimalcardiac output, while noradrenaline is used to restore anadequat e blood pressure by reducing vasodilatation. However this combination can only be used safely when guided by full haemodynamic monitoring. Adrenaline, because of its potency, remains a useful agent in those patients unresponsive to other measures, particularly after cardiac surgery, and is a cheap, effective
agent for the management of septic shock.

In uncomplicated cases careful interpretation of the CVP is an adequate guide to the filling pressures of both sides of the heart. In many critically ill patients, however, this is not the case and there is a disparity in function between the two ventricles. Most commonly, left ventricular performance is worst, so that the left ventricular function curve is displaced downward and to the right. This situation is encountered in some patients with clinically significant ischaemic heart disease and has also been reported in major trauma, sepsis, peritonitis, hepatic failure, valvular heart disease and after cardiac surgery. High right ventricular filling pressures, with normal or low left atrial pressures, are less common but may occur in right ventricular ischaemia and in situations where the pulmonary vascular resistance (i.e. right ventricular afterload) is raised, such as acute respiratory failure and pulmonary embolism.

If there is a disparity in ventricular function after cardiac surgery, then the left atrium can be cannulated directly. If the thorax is not open, however, some other means of determining left ventricular filling pressure must be devised.

Pulmonary artery pressure

A ‘balloon flotation catheter’ enables prompt and reliableatheterization of the pulmonary artery, without the need for screening, and minimizes the incidence of arrhythmias.
These ‘Swan-Ganz’ catheters can be inserted centrally 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 wedge position is monitored and guided by the pressure waveforms recorded from the distal lumen. A chest X-ray should always be obtained to check the final position of the catheter. 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 known as pulmonary artery wedge pressure PAWP) 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 reflection of left atrial pressure. The technique is generally safe-the majority of complications are related to user inexperience. Pulmonary artery catheters should preferably be removed within 72 hours, since the incidence of complications then increases progressively.

Cardiac output

The only quantitatively accurate methods for measuring cardiac output are invasive. Of these, the thermodilution technique is most commonly used clinically. This uses a modified pulmonary artery catheter with a lumen opening in the right atrium and a thermistor located a few centimetres from its tip. A known volume (usually 10 rnl) of ice-cold 5% dextrose is injected as a bolus into the right atrium. This mixes with, and cools, the blood passing through the heart and the transient fall in temperature is continuously recorded by the thermistor in the pulmonary artery. The cardiac output is computed from the total amount of indicator (i.e. cold) injected, divided by the average concentration, i.e. the amount of cooling, and the time taken to pass the thermistor.

MANAGEMENT

Delays in making the diagnosis and initiating treatment, as well as inadequate resuscitation, contribute to the development of MOF and must be avoided. A patent airway must be maintained and oxygen is given. If necessary, an oropharyngeal airway or an endotracheal tube is inserted. The latter has the advantage of preventing aspiration of gastric contents. Very rarely emergency tracheostomy is indicated. The underlying cause of shock should be corrected, e.g. haemorrhage should be controlled or infection eradicated. In patients with septic shock, every effort must be made to identify the source of infection and isolate the causative organism. As well as a thorough history and clinical examination, X-rays, ultrasonography and CT scanning may be required to locate the origin of the infection. Appropriate samples (urine, sputum, cerebrospinal fluid, pus drained from abscesses) should be sent to the laboratory for microscopy, culture and sensitivities. Several blood cultures should be performed and ‘blind’ antibiotic therapy should be commenced. If an organism is isolated, the therapy can be adjusted appropriately. The choice of antibiotic depends on the source of infection-whether this was acquired in hospital or in the community. Abscesses must be drained and infected indwelling catheters removed.
Whatever the aetiology of the haemodynamic abnormality, tissue blood flow must be restored by achieving and maintaining an adequate cardiac output, as well as ensuring that arterial blood pressure is sufficient to maintain perfusion of vital organs.

Pre-load and volume replacement

Optimizing pre-load is the most efficient way of increasing cardiac output. Volume replacement is obviously essential in hypovolaemic shock but is also required in anaphylactic and septic shock because of vasodilatation, sequestration of blood and loss of circulating volume due to capillary leak.
In mechanical shock, high filling pressures may be required to maintain an adequate stroke volume. Even in cardiogenic shock, careful volume expansion may, on occasions, lead to a useful increase in cardiac output. On the other hand, patients with severe cardiac failure, in whom ventricular filling pressures may be markedly elevated, often benefit from measures to reduce pre-load (and after-load) such as the administration of diuretics and vasodilators.
The circulating volume must be replaced quickly (in minutes not hours) to reduce tissue damage and prevent acute renal failure. Fluid is administered via wide-bore intravenous cannulae to allow large volumes to be given quickly and the effect is continuously monitored. Care must be taken to prevent volume overload, which leads to cardiac dilatation, a reduction in stroke volume, and a rise in left atrial pressure with a risk of pulmonary oedema. Pulmonary oedema is more likely in very ill patients because of a low colloid osmotic pressure (usually due to a low serum albumin) and disruption of the alveolar-capillary membrane (e.g. in ARDS). The development of pulmonary oedema can also be influenced by other unquantifiable factors, such as the hydrostatic and oncotic pressures within the interstitial spaces. Since the pulmonary lymphatics remove excess fluid, pulmonary oedema will only occur when this mechanism is overwhelmed or impaired. Left ventricular filling pressures should therefore not be allowed to rise to more than 15-18 mmHg in the critically ill. In general, however, many more patients are undertransfused rather than overtransfused.

Although many clinical features are common to all types of shock there are certain important respects in which they differ:

Hypovolaemic shock

1 Inadequate tissue perfusion:
(a) Skin-cold, pale, blue, slow capillary refill
(b) Kidneys-oliguria, anuria
(c) Brain-confusion and restlessness
2 Increased sympathetic tone:
(a) Tachycardia, narrowed pulse pressure
(b) Sweating
(c) Blood pressure-may be maintained initially (despite up to a 25% reduction in circulating volume if the patient is young and fit), but later hypotension supervenes
3 Metabolic acidosis and tachypnoea
Additional clinical features may occur III the following types of shock.

Cardiogenic shock

1 Signs of myocardial failure, e.g. raised jugular venous pressure Ovp), pulsus alternans, ‘gallop’ rhythm, basal crackles, pulmonary oedema

Mechanical shock

1 Elevated JVP
2 Pulsus paradoxus and muffled heart sounds in cardiac tamponade
3 Kussmaul’s sign OVP rises on inspiration) in cardiac tamponade
4 Signs of pulmonary embolism (if present)

Anaphylactic shock

1 Signs of profound vasodilatation:
(a) Warm peripheries
(b) Low blood pressure
2 Erythema, urticaria, angio-oedema, pallor, cyanosis
3 Bronchospasm, rhinitis
4 Oedema of the face, pharynx and larynx
5 Pulmonary oedema
6 Hypovolaemia due to capillary leak
7 Nausea, vomiting, abdominal cramps, diarrhoea

Septic shock

1 Pyrexia and rigors, or hypothermia (unusual)
2 Nausea, vomiting
3 Vasodilatation, warm peripheries
4 Bounding pulse
5 Rapid capillary refill
6 Hypotension
7 Occasionally signs of cutaneous vasoconstriction
8 Other signs:
(a) Jaundice
(b) Coma (rare)
(c) Coagulopathy
The diagnosis of septicaemia is easily missed. In the elderly, the classical signs may not be present and, for example, mild confusion, tachycardia and tachypnoea may be the only clues, sometimes associated with unexplained hypotension, a reduction in urine output, a rising plasma creatinine and glucose intolerance.

MONITORING 

Invasive monitoring is unnecessary in straightforward cases, such as a fit young man with moderate traumatic haemorrhage, but will be required in the more seriously ill patients and in those who fail to respond to initial treatment (see later). Clinical assessment must never be neglected.

Clinical indices of tissue perfusion

Pale, cold skin, delayed capillary refill and the absence of visible veins in the hands and feet indicate poor perfusion. Skin temperature measurements can help clinical evaluation as vasoconstriction is an early compensatory response.
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. The absolute level of blood pressure is also important, since hypotension may jeopardize 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 with a sphygmomanometer, but automated instruments using a microphone to detect Korotkoffs sounds or continuous monitoring with an intra-arterial cannula, usually in the radial artery.

Central venous pressure (CVP)

This provides a fairly simple method of assessing the adequacy of a patient’s circulating volume and the contractile state of the myocardium. The absolute value of the cVP is not as important as its response to a fluid challenge (the infusion of 100-200 ml of fluid over 1-3 min). 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 and increased peripheral temperature). As the normovolaemic state is approached, the CVP usually rises slightly and stabilizes, while other cardiovascular values normalize. 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 respond to volume replacement, which will cause a further, sometimes dramatic, rise in CVP. The CVP may be read intermittently using a manometer system or continuously using a transducer connected to an oscilloscope, similar to that used for intra-arterial monitoring.
Common pitfalls in interpreting CVP results are:
BLOCKED CATHETER. This results in a sustained high reading, with a damped waveform which often does not correlate with clinical assessment.
MANOMETER OR TRANSDUCER WRONGLY r-o s- ITIONED. Failure to level the CVP after changing the patient’s position is a common cause of erroneous readings.
INCORRECT CALIBRATION  If an electronic transducer and oscilloscope are used, the system should be zeroed and calibrated prior to use.
ONE OR MORE INFUSIONS IN PROGRESS THROUGH THE CVP CATHETER. The CVP catheter maybe used for other infusions and the pressure measured intermittently. A falsely high reading will result if these fluids continue to be administered by an infusion pump while the pressure is recorded.
CATHETER TIP IN RIGHT VENTRICLE. If the catheter is advanced too far, an unexpectedly high pressure is recorded.
The catheter should be positioned in the superior vena cava. It is usually inserted via a percutaneous puncture of a subclavian or internal jugular vein.