Acute Respiratory Distress Syndrome (ARDS)

First described in Denver, Colorado, in 1967 by Ashbaugh. North American–European consensus group has changed the definition to ‘acute’ (as opposed to ‘adult’) respiratory distress syndrome, since the syndrome can occur in children. It is often the pulmonary component of the systemic inflammatory response syndrome (SIRS) and is characterized by severe hypoxia refractory to oxygen, low compliance, high airway pressure, bilateral diffuse alveolar infiltrates and microscopic atelectasis. It has an annual incidence of about 3.5:100 000.

Causes

Table 7.1 Causes of ARDS

Direct injury	Indirect injury
Pulmonary contusion	Septicaemia
Gastric aspiration	Major trauma
Fat and amniotic fluid embolus	Cardiopulmonary bypass
Infection	Massive blood transfusion
Cytotoxic drugs	Prolonged hypotension
Smoke inhalation	Hepatic and renal failure
Oxygen toxicity	Disseminated intravascular coagulation

Pathophysiology

Activated neutrophils adhere to endothelial cells and release inflammatory mediators, including oxygen-free radicals and proteases, to cause lung damage. Direct lung damage or endotoxins alone are sufficient to damage endothelial cells with cytokine release and an inflammatory cascade (Fig. 7.1).

Figure 7.1 Pathophysiology of systemic inflammatory response syndrome (SIRS).

Endothelial damage results in increased capillary permeability and formation of protein-rich alveolar exudate rich in neutrophils. Type I alveolar cells are damaged and type II cells proliferate. As the disease progresses, fibroblast infiltration and collagen proliferation cause microvascular obliteration and widespread fibrosis. Areas of lung involvement are not fixed but shift to dependent areas. Within areas of reduced lung volume, some alveoli remain open and capable of gas exchange, whereas others are filled with alveolar exudate.

IPPV may cause damage more through excess volume (‘volutrauma’) than through pressure itself. The significance of oxygen toxicity is controversial.

Treatment

Cardiovascular System

Inotropes

If shock persists despite adequate volume replacement and vital organ perfusion is jeopardized, inotropic drugs may be required to improve blood pressure and cardiac output (Table 7.2).

Table 7.2 Receptor actions of inotropes

Cardiogenic shock. Characterized by low cardiac output, high filling pressures and increased systemic vascular resistance (SVR). Inodilators (dobutamine, enoximone, milrinone, dopexamine) improve cardiac contractility and decrease SVR. Specific vasodilators (nitroprusside, GTN) may reduce afterload further, increasing stroke volume and decreasing cardiac work by decreasing systolic wall tension.

Septic shock. Characterized by high cardiac output (if hypovolaemia corrected) and decreased SVR. Vasoconstrictors (noradrenaline) reduce SVR. Dobutamine or adrenaline may be required to improve myocardial contractility.

Adrenergic receptor function

Arterioles and veins contain α₁ and α₂ receptors which cause vasoconstriction. Vessels in skeletal muscle are rich in β₂ receptors which cause vasodilation. Myocardium is rich in β₁ and β₂ receptors which are pro-chronotropic and pro-inotropic.

α₁ receptors activate phospholipase C which cleaves phosphoinositol 4,5-biphosphate (PIP₂) to diacylglycerol (DAG) and inositol triphosphate (IP₃). DAG activates protein kinase C to release arachidonic acid which increases cGMP synthesis (Fig. 7.2). Activation of β₁, β₂ and dopamine (D₁) receptors triggers the conversion of ATP to cAMP, activating protein kinase C and triggering protein phosphorylation which causes tachycardia and increases contractility (Fig. 7.3).

Figure 7.2 Activation of α₁ receptors.

Figure 7.3 Activation of β₁, β₂ receptors.

Pulmonary artery catheters

Although a recent meta-analysis concluded that PAC did not affect mortality, intensive care unit or hospital length of stay, analysis of the National Trauma Data Bank in 2006 (53 312 patients) has demonstrated improved outcomes when used for major trauma. Although there is limited evidence to show improved outcome with PAC use, the general consensus is that their use in appropriate patients by clinicians skilled in their insertion and data interpretation is of benefit.

Pressure changes during PAC insertion (Fig. 7.4)

Place catheter so that it does not wedge until at least 1.2 mL air is in the balloon (1.5 mL max.). Greatest degree of accuracy if tip of catheter is placed in West’s zone III. Zone I tends to measure airway pressure, particularly with hypovolaemia. Measure wedge from ‘a’ wave at end of expiration.

Figure 7.4 Pressure changes during pulmonary artery catheter insertion. RA, right atrium; RV, right ventricle; PA, pulmonary artery; PCW, pulmonary capillary wedge.

Complications

Table 7.3 Complications of pulmonary artery catheter insertion

Associated with insertion	Associated with catheter presence
Pneumothorax/haemothorax	Infection of catheter or site
Haematoma	Pulmonary thrombosis/infarct
Cardiac arrhythmias	Cardiac arrhythmias
Arterial puncture	Valve damage/endocarditis
Pulmonary artery perforation	Pulmonary artery erosion
Catheter knotting	Thrombocytopenia
Cardiac valve damage

Cardiac output monitoring

Invasive techniques

• Intermittent thermodilution. Change in temperature of blood following injection of iced saline. Calculated from the Stuart–Hamilton equation:

where:

CO = cardiac output

VI = volume of injectate

BT = blood temperature after injection

BT₀ = blood (pulmonary artery) temperature before injection

IT = injectate temperature.

• Continuous thermodilution. Change in temperature of blood warmed by heating coil on pulmonary artery catheter. Calculated from a modified Stuart–Hamilton equation.

• Indocyanine green dilution (historical)

• Fick calculated cardiac output

• Pulse contour cardiac output (PiCCO). Calculates cardiac output continuously from pulse contour analysis of the aortic waveform via an arterial catheter. Requires central venous access to perform a thermodilution cardiac output measurement for calibration (analysed by a thermistor present in the arterial catheter tip).

• Combined lithium dilution and pulse contour analysis (LiDCO). Uses lithium dilution curve for initial calibration followed by a pulse contour algorithm to determine beat-to-beat stroke volume from a mathematical analysis of the peripheral arterial waveform.

Non-invasive techniques

• Transthoracic/oesophageal Doppler ultrasonography. Measurement of aortic cross-sectional area and blood flow velocity allows calculation of cardiac output. Acceleration and peak velocity indicate myocardial performance, while flow time is related to circulating volume and peripheral resistance.

• Transthoracic impedance. Can be measured across externally applied electrodes. Small changes in impedance (<1Ω) occur with the cardiac cycle as the blood volume in the heart increases during diastole and decreases during systole. Rate of change of impedance can be used to estimate cardiac output.

• Pulse contour analysis. Analyses the arterial waveform to calculate cardiac output by using an estimated compliance of the arterial tree to calculate the stroke volume needed to increase the blood pressure by a given amount. Accuracy limited by non-linear vascular compliance, damped waveforms and aortic wall pathology.

Transoesophageal echocardiography (TOE)

Increasing use in critical care. Uses the physical principle that sound is reflected from tissue interfaces. Piezoelectric crystals transmit and receive acoustic signals to build a two-dimensional image. TOE provides better cardiac views than transthoracic echocardiography, particularly of valves, contractility and filling status.

TOE is of particular use in the assessment of:

• left ventricular systolic function

• left ventricular filling

• wall motion abnormalities

• cardiac tamponade

• valve function

• chest trauma (limited views of ascending aorta).

TOE has shown a poor relationship between PCWP and left ventricular end-diastolic volume, particularly in septic patients and those on high doses of inotropes. In these patients, mean pulmonary capillary pressure may be a more accurate measure.

Bibliography

American Society of Anesthesiologists Task Force on Pulmonary Artery Catheterization. Practice guidelines for pulmonary artery catheterization. Anesthesiology. 2003;99:988-1014.

Antonelli M., Levy M., Andrews P.J., et al. Hemodynamic monitoring in shock and implications for management. International Consensus Conference. Intensive Care Med. 2007;33:575-590.

Dawson S., Jhanji S., Pears J.R.M. Cardiac output monitoring: basic science and clinical application. Anaesthesia. 2008;63:172-181.

Jhanji S., Dawson J., Pearse R.M. Cardiac output monitoring: basic science and clinical application. Anaesthesia. 2008;63:172-181.

Klein A.A., Snell A., Nashef S.A., et al. The impact of intra-operative transoesophageal echocardiography on cardiac surgical practice. Anaesthesia. 2009;64:947-952.

Moppett I., Shajar M. Transoesophageal echocardiography. BJA CEPD Rev. 2001;1:72-75.

Van Lieshout J.J., Wesseling K.H. Continuous cardiac output by pulse contour analysis? Br J Anaesth. 2001;86:467-469.

Fluid and Electrolyte Balance

The neonate has a greater proportion of body water and in a different distribution than the adult (Fig. 7.5). More fluid is distributed within the extracellular compartment (interstitial and plasma volume) compared with the adult, resulting in a larger volume of distribution for water-soluble drugs. A large proportion of interstitial fluid is excreted within the first few weeks after birth and adult levels are attained by adolescence.

Figure 7.5 Body water distribution in the neonate and adult.

A 70 kg male has about 42 kg of water distributed through three body compartments.

• extracellular fluid – 20% (plasma volume 4%, interstitial fluid 16%)

• intracellular fluid – 40%

Plasma volume expansion is least effective with fluids that are distributed throughout all body compartments and most effective with those that remain within the intravascular compartment. Therefore:

• 1000 mL 5% dextrose (plasma, interstitial and intracellular compartments) expands plasma by 1000 × 4/60 = 67 mL

• 1000 mL normal saline (plasma and interstitial compartments) expands plasma by 1000 × 4/20 = 200 mL

• 1000 mL colloid (plasma only) expands plasma by 1000 × 4/4 = 1000 mL.

Homeostatic control

A rise in serum osmolality stimulates osmoreceptors in the anterior hypothalamus to release ADH from the supraoptic nuclei. ADH release is also stimulated by volume receptors in the left atrium and carotid sinus baroreceptors in response to hypovolaemia.

ADH increases permeability of the distal collecting ducts, resulting in more fluid reabsorption and a reduction in the osmolality of body fluids.

Aldosterone is produced by the adrenal cortex in response to ACTH, renin, hyponatraemia or hyperkalaemia. It acts on the distal tubules and collecting ducts to increase Na⁺ resorption and K⁺ excretion.

Renin is secreted by the juxtaglomerular apparatus in response to a fall in extracellular volume or BP and acts on angiotensinogen to form angiotensin I which is converted to angiotensin II by passage through the lungs. Vasoconstriction produced by angiotensin II produces a marked rise in both systolic and diastolic blood pressures.

Normal fluid requirements

Fluid and electrolyte requirements per 24 h

• Water: 35 mL.kg⁻¹

• Na⁺: 2 mmol.kg⁻¹

• K⁺: 1 mmol.kg⁻¹

Loss

• Insensible loss (skin and lungs) 500–1000 mL.24 h^−1.

• Urine (minimum) 300 mL.24 h^−1.

• Faeces 500 mL.24 h⁻¹.

Assessment of hydration

Thirst, skin turgor (loss of turgor at >10% fluid deficit), hydration of mucous membranes, core–peripheral temperature gradient, pulse rate and volume, postural hypotension (>20% fluid deficit), urine output and osmolality and fluid balance.

Postoperative fluid requirements

Intravenous fluids administered perioperatively during minor gynaecological surgery reduce morbidity, particularly nausea and dizziness. However, blood coagulation appears to be accelerated by haemodilution with saline, and in patients undergoing elective abdominal surgery, the incidence of DVT was four times greater than in the fluid-restricted group (Janvrin et al 1980).

Despite this latter study, it is generally agreed that the advantages of perioperative fluids outweigh any disadvantages. Hartmann’s 15 mL.kg⁻¹.h⁻¹ has been suggested for major surgery; a rate shown to improve postoperative renal function. Septic patients or those with lung trauma have raised extravascular lung water, and lesser rates may be necessary to avoid pulmonary oedema. In addition, give blood to maintain Hb >8.5–9.0 g.dL⁻¹.

Surgical stress causes release of ADH, renin and aldosterone, resulting in sodium and water retention, potassium excretion and an inability to excrete a hypotonic urine. Postoperative catabolism increases the minimum metabolic demand for water from 20 to 30 mL.kg⁻¹ per day, i.e. 2000 mL.day⁻¹. The addition of 100 g.day⁻¹ of glucose reduces nitrogen loss by up to 60%. Therefore, give maintenance fluids of 2000 mL 5% dextrose.24 h⁻¹ postoperatively with 30 mmol KCl added to each 1 L bag to provide daily K⁺ requirements. Hidden losses are difficult to judge so titrate fluids according to urine output (>0.5 mL.kg⁻¹.h⁻¹). Sodium retention is greatly reduced by 48 h so then add Na⁺ to maintenance fluids and reduce KCl supplements.

Albumin

Single polypeptide of 585 amino acids. Synthesized in the endoplasmic reticulum of hepatocytes at 9–12 g/day but can increase 2–3 times in states of maximum synthesis. Stimulus to production is colloid osmotic pressure, osmolality of the extravascular liver space, insulin, thyroxine and cortisol. Catabolized by vascular endothelium. 5% of albumin is removed from the intravascular space per hour. Clinical properties of albumin include:

• Binding and transport – strong negative charge binds Ca²⁺ (40%), Cu²⁺, thyroxine, bilirubin and amino acids; also binds warfarin, phenytoin, NSAIDs and digoxin

• Maintenance of colloid osmotic pressure (COP) – contributes to 80% of COP

• Free radical scavenging – thiol groups scavenge reactive oxygen and nitrogen species

• Platelet inhibition and antithrombotic effects

• Effects on vascular permeability – may bind within the subendothelium to alter capillary membrane permeability.

Serum albumin decreases due to dilutional effects with crystalloid/colloid solutions, redistribution due to altered capillary permeability (five-fold increase during sepsis), decreased synthesis in septic patients, and increased loss from kidney or gut.

Correlation between COP and serum albumin is poor. Therefore, oedema associated with hypoalbuminaemia is not necessarily related and may be related more to lymphatic dysfunction. The acute-phase response is initially associated with a decrease in albumin synthesis, possibly due to IL-6-mediated inhibition of synthesis. A later hypermetabolic phase results in increased albumin synthesis.

Benefits of correcting hypoalbuminaemia are unclear. A prospective randomized study of 475 ICU patients comparing albumin and gelatin solutions failed to show any benefit (Stockwell et al 1992). In 70 children with burns, albumin supplementation failed to improve morbidity or mortality (Greenhalgh et al 1995). In septic patients, albumin infusions will only increase COP for a relatively short period. Increased capillary permeability results in >60% of albumin leaving the intravascular compartment within 4 h, potentially worsening oedema.

A controversial systematic review by the Cochrane Group of 23 randomized controlled trials found that the risk of death was 6% greater in the group treated with albumin compared with those receiving crystalloids or no treatment (Cochrane Injuries Group 1998). The Committee on Safety of Medicines now advises doctors to restrict the use of, and take special care when using, human albumin, but states that there is ‘insufficient evidence of harm to warrant withdrawal of albumin’. Hypoalbuminaemia in itself is not an appropriate indication. Risks of hypervolaemia and cardiovascular overload warrant monitoring in patients receiving albumin.

Intravenous fluids

Starches

Amylopectin linked with hydroxyethyl groups in a glucose moiety making a polymer similar to glycogen. Large range of molecular weights.

Hespan is 6% hetastarch; t_1/2 24 h because of the hydroxyethyl–glucose bond.

Voluven is similar to Hespan but with a lower molecular weight. Allergic reactions are less likely than with Voluven compared with other colloids.

Human albumin solution

Derived from human plasma by fractionation. Heat sterilized with minimal risk of infective transmission.

Colloid versus crystalloid controversy (Table 7.4)

There is some evidence to suggest that crystalloid resuscitation is associated with a lower mortality in trauma patients.

Table 7.4 Comparison of crystalloids and colloids

	Advantages	Disadvantages
Crystalloid	Cheap	Larger volumes needed
	Replaces extravascular loss	Small ↑ in plasma volume
	Increased GFR	Peripheral and pulmonary oedema
	Minimal effect on clot quality
Colloid	Smaller volumes needed	Risk of anaphylaxis
	Prolonged ↑ in plasma volume	Relatively expensive
	Reduced peripheral oedema	Coagulopathy
		Poor clot quality

British Consensus Guidelines on Intravenous Fluid Therapy for Adult Surgical Patients

Powell-Tuck et al 2008

Summary and recommendations

Food and fluids should be provided orally or enterally and i.v. infusions discontinued as soon as possible. The effects of surgical and metabolic stress on the renin-angiotensin–aldosterone system and on vasopressin should be understood. Nutrition should be assessed and cautiously maintained. The oedematous patient should be managed with particular care, in order to achieve successful negative sodium and water balance.

1. Because of the risk of inducing hyperchloraemic acidosis in routine practice, when crystalloid resuscitation or replacement is indicated, balanced salt solutions, e.g. Ringer’s lactate/acetate or Hartmann’s solution should replace 0.9% saline, except in cases of hypochloraemia, e.g. from vomiting or gastric drainage.

2. Solutions such as 4%/0.18% dextrose/saline and 5% dextrose are important sources of free water for maintenance, but should be used with caution as excessive amounts may cause dangerous hyponatraemia, especially in children and the elderly. These solutions are not appropriate for resuscitation or replacement therapy except in conditions of significant free water deficit, e.g. diabetes insipidus.

3. To meet maintenance requirements, adult patients should receive sodium 50–100 mmol/day, potassium 40–80 mmol/day in 1.5–2.5 L of water by the oral, enteral or parenteral route (or a combination of routes). Additional amounts should only be given to correct deficit or continuing losses. Careful monitoring should be undertaken using clinical examination, fluid balance charts, and regular weighing when possible.

Preoperative fluid management

4. In patients without disorders of gastric emptying undergoing elective surgery clear non-particulate oral fluids should not be withheld for more than 2 hours prior to the induction of anaesthesia.

5. In the absence of disorders of gastric emptying or diabetes, preoperative administration of carbohydrate rich beverages 2–3 h before induction of anaesthesia may improve patient well being and facilitate recovery from surgery. It should be considered in the routine preoperative preparation for elective surgery.

6. Routine use of preoperative mechanical bowel preparation is not beneficial and may complicate intra and postoperative management of fluid and electrolyte balance. Its use should therefore be avoided whenever possible.

7. Where mechanical bowel preparation is used, fluid and electrolyte derangements commonly occur and should be corrected by simultaneous intravenous fluid therapy with Hartmann’s or Ringer–Lactate/acetate type solutions.

8. Excessive losses from gastric aspiration/vomiting should be treated preoperatively with an appropriate crystalloid solution which includes an appropriate potassium supplement. Hypochloraemia is an indication for the use of 0.9% saline, with sufficient additions of potassium and care not to produce sodium overload.

9. Losses from diarrhoea/ileostomy/small bowel fistula/ileus/obstruction should be replaced volume for volume with Hartmann’s or Ringer-Lactate/acetate type solutions. Saline depletion, for example due to excessive diuretic exposure, is best managed with a balanced electrolyte solution such as Hartmann’s.

10. In high-risk surgical patients, preoperative treatment with intravenous fluid and inotropes should be aimed at achieving predetermined goals for cardiac output and oxygen delivery as this may improve survival.

11. Although currently logistically difficult in many centres, preoperative or operative hypovolaemia should be diagnosed by flow-based measurements wherever possible. The clinical context should also be taken into account, as this will provide an important indication of whether hypovolaemia is possible or likely. When direct flow measurements are not possible, hypovolaemia will be diagnosed clinically on the basis of pulse, peripheral perfusion and capillary refill, venous (JVP/CVP) pressure and Glasgow Coma Scale, together with acid–base and lactate measurements. A low urine output can be misleading and needs to be interpreted in the context of the patient’s cardiovascular parameters above.

12. Hypovolaemia due predominantly to blood loss should be treated with either a balanced crystalloid solution or a suitable colloid until packed red cells are available. Hypovolaemia due to severe inflammation such as infection, peritonitis, pancreatitis or burns should be treated with either a suitable colloid or a balanced crystalloid. In either clinical scenario, care must be taken to administer sufficient balanced crystalloid and colloid to normalize haemodynamic parameters and

13. Minimize overload. The ability of critically ill patients to excrete excess sodium and water is compromised, placing them at risk of severe interstitial oedema. The administration of large volumes of colloid without sufficient free water (e.g. 5% dextrose) may precipitate a hyperoncotic state.

14. When the diagnosis of hypovolaemia is in doubt and the CVP is not raised, the response to a bolus infusion of 200 mL of a suitable colloid or crystalloid should be tested. The response should be assessed using the patient’s cardiac output and stroke volume measured by flow-based technology if available. Alternatively, the clinical response may be monitored by measurement/estimation of the pulse, capillary refill, CVP and blood pressure before and 15 min after receiving the infusion. This procedure should be repeated until there is no further increase in stroke volume and improvement in the clinical parameters.

Intraoperative fluid management

15. In patients undergoing some forms of orthopaedic and abdominal surgery, intraoperative treatment with intravenous fluid to achieve an optimal value of stroke volume should be used where possible as this may reduce postoperative complication rates and duration of hospital stay.

16. Patients undergoing non-elective major abdominal or orthopaedic surgery should receive intravenous fluid to achieve an optimal value of stroke volume during and for the first 8 hours after surgery. This may be supplemented by a low dose dopexamine infusion.

Postoperative fluid and nutritional management

17. Details of fluids administered must be clearly recorded and easily accessible.

18. When patients leave theatre for the ward, HDU or ICU their volume status should be assessed. The volume and type of fluids given perioperatively should be reviewed and compared with fluid losses in theatre including urine and insensible losses.

19. In patients who are euvolaemic and haemodynamically stable a return to oral fluid administration should be achieved as soon as possible.

20. In patients requiring continuing i.v. maintenance fluids, these should be sodium poor and of low enough volume until the patient has returned their sodium and fluid balance over the perioperative period to zero. When this has been achieved the i.v. fluid volume and content should be those required for daily maintenance and replacement of any on-going additional losses.

21. The haemodynamic and fluid status of those patients who fail to excrete their perioperative sodium load, and especially whose urine sodium concentration is <20 mmol/L, should be reviewed.

22. In high-risk patients undergoing major abdominal surgery, postoperative treatment with intravenous fluid and low dose dopexamine should be considered, in order to achieve a predetermined value for systemic oxygen delivery, as this may reduce postoperative complication rates and duration of hospital stay.

23. In patients who are oedematous, hypovolaemia if present must be treated, followed by a gradual persistent negative sodium and water balance based on urine sodium concentration or excretion. Plasma potassium concentration should be monitored and where necessary potassium intake adjusted.

24. Nutritionally depleted patients need cautious refeeding orally, enterally or parenterally, with feeds supplemented in potassium, phosphate and thiamine. Generally, and particularly if oedema is present, these feeds should be reduced in water and sodium. Though refeeding syndrome is a risk, improved nutrition will help to restore normal partitioning of sodium, potassium and water between intra and extra-cellular spaces.

25. Surgical patients should be nutritionally screened, and NICE guidelines for perioperative nutritional support adhered to. Care should be taken to mitigate risks of the re-feeding syndrome.

Fluid management in acute kidney injury (AKI)

26. Based on current evidence, higher molecular weight hydroxyethyl starch (hetastarch and pentastarch MW ≥200 kDa) should be avoided in patients with severe sepsis due to an increased risk of AKI.

27. Higher molecular weight hydroxyethyl starch (hetastarch and pentastarch MW ≥200 kDa) should be avoided in brain-dead kidney donors due to reports of osmotic nephrosis-like lesions.

28. Balanced electrolyte solutions containing potassium can be used cautiously in patients with AKI closely monitored on HDU or ICU in preference to 0.9% saline. If free water is required, 5% dextrose or dextrose saline should be used. Patients developing hyperkalaemia or progressive AKI should be switched to non potassium containing crystalloid solutions such as 0.45% saline or 4%/0.18 dextrose/saline.

29. In patients with AKI, fluid balance must be closely observed and fluid overload avoided. In patients who show signs of refractory fluid overload, renal replacement therapy should be considered early to mobilize interstitial oedema and correct extracellular electrolyte and acid–base abnormalities.

30. Patients at risk of developing AKI secondary to rhabdomyolysis must receive aggressive fluid resuscitation with an isotonic crystalloid solution to correct hypovolaemia. There is insufficient evidence to recommend the specific composition of the crystalloid.

Bibliography

Boldt J. The balanced concept of fluid resuscitation. Br J Anaesth. 2007;99:312-315.

Chappell D., Hofmann-Kiefer K., Conzen P., et al. A rational approach to perioperative fluid management. Anesthesiology. 2008;109:723-740.

Choi P.T.L., Yip G., Quinonez L.G., et al. Crystalloids vs colloids in fluid resuscitation: a systematic review. Crit Care Med. 1999;27:200-210.

Cochrane Injuries Group. Human albumin administration in critically injured patients: systematic review of randomised controlled trials. BMJ. 1998;317:235-240.

Greenhalgh D.G., Housinger T.A., Kagan R.J., et al. Maintenance of serum albumin levels in pediatric burn patients: a prospective, randomized trial. J Trauma. 1995;39:67-74.

Grocott M.P.W., Mythen M., Gan T.J. Perioperative fluid management and clinical outcome in adults. Anesth Analg. 2005;100:1093-1106.

Janvrin S.B., Davies G., Greenhalgh R.M. Postoperative deep vein thrombosis caused by intravenous fluids during surgery. Br J Surg. 1980;67:690-693.

Powell-Tuck J., Gosling P., Lobo D.N., et al. British Consensus Guidelines on Intravenous Fluid Therapy for Adult Surgical Patients, 2008 www.ics.ac.uk/intensive_care_professional/standards_and_guidelines/british_consensus_guidelines_on_intravenous_fluid_therapy_for_adult_surgical_patients__giftasup__2008.

Rassam S.S., Counsell D.J. Perioperative fluid therapy. Contin Edu Anaesth, Crit Care Pain. 2005;5:161-165.

Stockwell M.A., Soni N., Riley B. Colloid solutions in the critically ill. A randomised comparison of albumin and polygeline. 1. Outcome and duration of stay in the intensive care unit. Anaesthesia. 1992;47:3-6.

Strandvik G.F. Hypertonic saline in critical care: a review of the literature and guidelines for use in hypotensive states and raised intracranial pressure. Anaesthesia. 2009;64:990-1003.

Methicillin Resistant Staphylococcus Aureus (MRSA)

Increasing resistance to penicillin-based antibiotics since their introduction in the 1950s. Due to ß-lactamase enzyme and penicillin binding protein. MRSA first recognized in Europe in the 1960s. Rates of infection greatest in ICU > surgical wards > general wards > community.

Routes of transmission are hands, environment and colonized patients. Control spread through hand washing, cleaning, screening of patients and eradication.

Risk factors for infection are:

• MRSA colonization (60% critically ill patients will develop infection).

• Length of ICU and hospital stay (odds ratio = 2.5 for hospital stay of >2 weeks).

• Severity of illness and intensity of care.

• Intravascular devices. Nine-fold increase in MRSA infection in patients with intravascular catheters.

• Antibiotic therapy – some studies have shown that number of antibiotics used and duration of therapy directly increases risk of MRSA infection.

Mandatory reporting of all MRSA bacteraemias to the Department of Health since 2001. Mortality greater with MRSA than methicillin-sensitive Staph aureus (MSSA). NCEPOD 2001 showed that 1.8% of surgical deaths were associated with MRSA infection.

Glycopeptides (vancomycin, teicoplanin) are still the mainstay of treatment. Vancomycin-resistant strains may be sensitive to linezolid and quinupristin/dalfopristin.

Bibliography

Hardy K.J., Hawkey P.M., Gao F., et al. Methicillin resistant Staphylococcus aureus in the critically ill. Br J Anaesth. 2004;92:121-130.

Nitric Oxide

In 1987, nitric oxide (NO) was identified as an endothelium-derived relaxing factor. It is a free radical acting as a local transcellular messenger through binding to transition metals within enzymes such as guanylate cyclase. About 1 mM of endogenous nitric oxide is synthesized per day. The synthesis and actions of NO are shown in Figure 7.6. NO is involved in:

• regulation of smooth muscle tone

• antimicrobial defence by its free radical action

• platelet and neutrophil adherence and aggregation

• peripheral and central neurotransmission

• regulation of smooth muscle proliferation.

Figure 7.6 Synthesis and action of nitric oxide (NO).

Cardiovascular effects

Endothelial nitric oxide synthase (NOS) produces nitric oxide in response to changes in blood velocity (shear stress). NO in turn causes smooth muscle relaxation by activating cGMP which modulates calcium concentration and therefore vascular smooth muscle cell tone to match blood vessel calibre to flow. NO synthesis is also increased by acetylcholine, bradykinin, hypoxia and α-adrenergic stimulation, although the degree of vasodilation induced by these factors is uncertain. Excess NO produced in the presence of ischaemia may contribute to ischaemic damage through its free radical damage to myocardium. Inhibitors of NO synthesis may protect against myocardial reperfusion injury. Inadequate NO production increases platelet aggregation and adhesion.

Pharmacological nitrates (nitroprusside, isosorbide mononitrate, nitrites) are converted to NO within smooth muscle cells, increasing cGMP production. Sulphydryl (-SH) groups are required to form NO, but excessive nitrate therapy depletes –SH, leading to therapeutic tolerance.

Hypertension

Basal production of NO continuously vasodilates the peripheral circulation. Basal NO release is greater in arteries than in veins. Untreated hypertension in humans is associated with reduced NO synthesis. l-arginine supplementation alone does not restore NO levels to normal, suggesting that other mechanisms in addition to substrate depletion are involved. Excess end-products of glycosylation and oxidized lipoproteins in atherosclerosis have been proposed to inactivate nitric oxide synthetase.

Synthetic vasodilators

Organic nitrates such as GTN and nitroprusside undergo enzymatic reduction, releasing NO within vascular smooth muscles. Small vessels, e.g. coronary arteries, lack the capacity for this metabolism and only large capacitance vessels and coronary vessels >100 μm dilate in response to clinical doses of nitroglycerine. This selective vasodilation is less likely to cause coronary steal and contributes to the anti-anginal effects of nitrates. Tolerance to organic nitrates occurs due to a reduction in the biotransformation of the drug.

Respiratory effects

Normal pulmonary vascular tone is very low and exogenous NO has little effect on pulmonary vascular resistance. However, in disease states where pulmonary vascular tone is increased (pulmonary hypertension, ARDS, chronic respiratory failure), NO selectively vasodilates pulmonary vasculature around functioning alveolar units. This local vasodilation diverts blood towards functioning alveoli, reduces intrapulmonary shunting and may improve oxygenation. Significant systemic vasodilation does not occur because NO is rapidly inactivated by binding to haemoglobin.

Although NO improves arterial oxygen tension, no studies yet show improvement in mortality in adults with ARDS. Treatment of persistent pulmonary hypertension of the newborn appears promising and may avoid the need to use extracorporeal membrane oxygenation. In infants with hypoxic respiratory failure, NO appears as effective as ECMO in reducing mortality but is cheaper and simpler to use. Normal dose range is 0.5–20 ppm.

Neuronal NO

NO acts as a neurotransmitter in both the central and peripheral nervous systems, where its release is stimulated by excitatory amino acids. Neural functions of NO include peripheral non-adrenergic non-cholinergic transmission, neurotransmission in contractile and secretory tissue, synaptic plasticity, learning and memory.

Toxicity of NO

NO forms several toxic products, including nitrogen dioxide and nitric and nitrous acids. NO also reacts with free radicals to form peroxynitrite (ONOO²⁻), which is highly cytotoxic. The clinical significance of these effects is uncertain. In the circulation, NO combines with iron to form methaemoglobin, which may be a problem with high levels of inhaled NO.

NO and sepsis

Inflammatory mediators such as TNF, cytokines and interferon induce nitric oxide synthase in endothelial cells, parenchymal cells and macrophages, increasing NO production. This may be the mechanism by which sepsis results in vasodilation. Some studies have shown that inhibition of NO synthesis with l-NMMA prevents these haemodynamic changes. Nitric oxide synthase inhibition may be a potential route for the treatment of sepsis. A large randomized controlled trial of a specific nitric oxide synthase antagonist in human septic shock was stopped because of an increased mortality rate in the treatment group and a large study of ketoconazole (which has anti-nitric oxide properties) showed no difference in mortality in ARDS.

Bibliography

Griffiths M.J., Evans T.W. Inhaled nitric oxide therapy in adults. N Engl J Med. 2005;353:2683-2695.

Young D. Nitric oxide. BJA CEPD Rev. 2002;2:161-164.

Non-Invasive Ventilation

Equipment

Requires ventilators that can deliver low resistance, high flow rates. Most popular machines deliver bilevel positive airway pressure (BiPAP). Simplest form is CPAP only. High levels of support are PEEP + pressure/volume support. Pressure-controlled modes have the advantage over volume-control in that they can compensate for air leaks. Three main types of mask:

Caples S.M., Gay P.C. Noninvasive positive pressure ventilation in the intensive care unit. Crit Care Med. 2005;33:2651-2658.

Hill N.S., Brennan J., Garpestad E., et al. Noninvasive ventilation in acute respiratory failure. Crit Care Med. 2007;35:2402-2407.

Nutrition

Malnutrition

Malnutrition is common in hospital patients. Preoperative nutritional support can improve nutritional status but may only improve morbidity and mortality in severely malnourished patients. This support must be maintained for at least 7 days preoperatively to show any benefit. Prospective studies have demonstrated a benefit for postoperative nutritional support.

Effects of malnutrition

CVS. Decreased HR, CO and CVP

Respiratory. Decreased inspiratory force and FVC; weaning more difficult.

GI. Decreased gut motility, gut atrophy and increased gut permeability to intestinal bacteria.

Other. Decreased metabolic rate. Increased susceptibility to infection, poor wound healing, muscle weakness resulting in immobility, oedema and impaired organ function. Increased morbidity and mortality.

Starvation and stress result in different rates of malnutrition and changes in catabolic pathways (Table 7.5).

Table 7.5 Effect of starvation and stress on energy expenditure and malnutrition

	Starvation	Stress
Energy expenditure	+	+++
R:Q	0.7	0.8–0.85
Malnutrition	+	+++
Rate of malnutrition	+	+++

R:Q

R:Q = CO₂ output/O₂ consumption

• CHO = 1.0

• Protein = 0.83

• Fat = 0.71.

Nutritional support

Indications for nutritional support

• Albumin <30 g/L

• Marked weight loss, muscle wasting and oedema

• Dietary history showing decreased intake for >1 week

• Medical and surgical disorders likely to result in malnutrition

• Postoperative starvation for >10 days.

Composition

Optimal combination of water, carbohydrate, protein, fat, vitamins and trace elements. Calories as glucose or fat; protein as amino acids.

Calorie:nitrogen ratio of 200 calories:1 g N₂, decreasing to 100 calories:1 g N₂ in catabolic and septic patients.

Monitor electrolytes daily, monitor blood sugar every 48 h, monitor liver function tests weekly.

Caloric content

• CHO – 4 kcal/g

• Protein – 4 kcal/g

• Fat – 9 kcal/g.

Enteral feeding

Requires at least 25 cm ileum. Paralytic ileus only affects the stomach and colon. Small bowel motility and absorption often remain normal and therefore bowel sounds and flatus are not required to start enteral feeding. Early enteral feeding through a nasojejunal tube or feeding jejunostomy may prevent paralytic ileus.

Bibliography

Bistrian B.R., McCowen K.C. Nutritional and metabolic support in the adult intensive care unit: key controversies. Crit Care Med. 2006;34:1525-1531.

Napolitano L., Cresci G., Martindale R.G., et al. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine and American Society for Parenteral and Enteral Nutrition: Executive Summary. Crit Care Med. 2009;37:1757-1761.

Oxygen Transport

Physiology

Oxygen cascade (PO₂) (kPa)

• Dry atmospheric gas 21.1

• Inspired tracheal gas 19.8

• Alveolar gas 14.7

• Arterial blood 13.3

• Mixed venous blood 5.3.

Table 7.6 Oxygen content of arterial and mixed venous blood

	Oxygen content (mL/dL)
	Arterial	Mixed venous
Total	20.0	15.0
Attached to Hb	19.7	14.9
Dissolved	0.3	0.1
	2.0 (100% O₂)
Saturation	97%	73%

Oxygen content

where:

C_aO₂ = arterial oxygen content

(1.3 × Hb × S_aO₂) – i.e. 1 g Hb, when completely saturated, binds 1.3 mL O₂

(0.003 × PaO₂) = amount of O₂ dissolved in plasma = 0.003 mL.mmHg⁻¹.

Oxygen delivery

where:

DO₂ = oxygen delivery

= cardiac output.

Oxygen uptake

VO₂ = × (C_aO₂ − C_VO₂) (Fick equation)

where:

VO₂ = oxygen uptake

C_vO₂ = venous oxygen content.

Oxygen extraction ratio

The oxygen extraction ratio (O₂ER) is the fractional uptake of oxygen from the capillary bed.

Normal response

Decreased blood flow results in increased O₂ extraction, i.e. drop in cardiac index is balanced by increased (S_aO₂ – S_vO₂). Thus VO₂ remains unchanged.

All vascular beds can increase O₂ extraction if flow drops, except coronary circulation and diaphragm. Therefore it is necessary to maintain cardiac output in patients with coronary artery disease because DO₂ is flow dependent.

Response in critically ill patients

Oxygen extraction from capillary beds may not increase where necessary and oxygen uptake (VO₂) becomes flow-dependent (Fig. 7.7). Therefore it is imperative to maintain cardiac output to maintain oxygen supply to the tissues.

Figure 7.7 Oxygen delivery (Do₂) versus uptake (Vo₂) in critically ill patients.

Mixed venous oxygen

Under normal conditions, venous oxygen levels will vary directly with changes in cardiac output. This is the rationale for using mixed venous (pulmonary artery) oxygen saturation to monitor changes in cardiac output:

Causes of low mixed venous oxygen

• Hypoxaemia

• Increased metabolic rate

• Low cardiac output

• Anaemia.

Causes of high mixed venous oxygen

• Decreased O₂ uptake, e.g. hypothermia, cell poisoning

• Left-to-right shunt

• Inappropriately high cardiac output, e.g. excessive use of inotropes/vasodilators.

S_vO₂ usually reflects cardiac output, but in seriously ill patients unable to mount a compensatory response to low blood flow, S_vO₂ will change little in response to changes in cardiac output. Thus, these patients show little correlation between venous oxygen and cardiac index.

Lactic acid

When the metabolic rate exceeds the rate of oxygen supply, tissues will switch to anaerobic metabolism and produce lactic acid. Therefore, the serum lactate concentration reflects the balance between VO₂ and the metabolic demand for oxygen.

Lactate is the end-product of anaerobic glycolysis, but it is also produced under normal aerobic conditions. The lactate anion is cleared by the liver and used for gluconeogenesis. Renal clearance becomes significant once levels reach 6–7 mEq/L. Hepatic failure probably contributes little to raised lactate levels in critically ill patients. A normal serum lactate does not exclude tissue ischaemia.

Raised lactate levels not associated with organ ischaemia are found with bacterial pneumonia, thiamine deficiency, generalized seizures, respiratory alkalosis and generalized trauma.

Bibliography

Morris C.G., Low J. Metabolic acidosis in the critically ill: part 1. Classification and pathophysiology. Anaesthesia. 2008;63:294-301.

Morris C.G., Low J. Metabolic acidosis in the critically ill: part 2. Causes and treatment. Anaesthesia. 2008;63:396-411.

Renal Replacement Therapy

In patients with severe acute renal failure, renal excretory function is lost. Resolution may take several weeks during which catabolism is marked, producing increased amounts of waste products. Requirements for intravascular space arise from intravenous drugs, fluids and feeding.

Renal replacement therapy therefore aims to remove excess water and remove unwanted solutes.

Normal kidney

• filters 180 L/day

• filters solutes weighing <60 000 Da

• reabsorbs 99% filtrate.

Urea is the waste product of protein metabolism (60 Da). Creatinine is the waste product of muscle metabolism (113 Da).

Immediate management of oliguria/acute renal failure

• Assess and treat any circulatory impairment

• Assess and treat any respiratory impairment

• Treat any acute manifestations if acute renal failure (salt and water retention, hypertension, hyperkalaemia, acidosis)

• Exclude urinary tract obstruction or infection

• Establish underlying cause and begin treatment as soon as possible

• Ensure patient is not taking nephrotoxic drugs.

History

Intermittent haemofiltration was introduced in the 1950s and remained the only form of renal replacement therapy until the introduction of intermittent haemodialysis in 1967. Technical problems delayed the introduction of continuous techniques until the 1980s with continuous arteriovenous haemofiltration (CAVH) followed by continuous arteriovenous haemodiafiltration (CAVHD).

Filters

• Clear molecules <60 000 Da

• Filter surface area approx. 70 m²

• Low resistance to enable high flow.

Principles

Water removal

Water is removed by a process called ultrafiltration. This process is similar to that performed by the glomerulus in the kidney. It requires a hydrostatic driving force to overcome oncotic pressure and move water across a semi-permeable membrane.

In the artificial kidney, this can be achieved by:

• applying a positive pressure (patient’s blood pressure/pump pressure) across a semi-permeable membrane

• using a hyperosmolar solution (e.g. peritoneal dialysis)

• applying a negative pressure to the dialysate side of the membrane (e.g. haemodialysis).

Solute removal

This can be achieved by either diffusion or ultrafiltration.

Diffusion across a semi-permeable membrane. This is movement of solutes across a semi-permeable membrane from an area of high concentration to one of low concentration. A concentration gradient is therefore always necessary for diffusion to occur. Molecules of a smaller MW will move across the membrane more readily than those with a larger MW (Fig. 7.8). A semi-permeable membrane has a defined pore size; any molecule exceeding this will not be able to pass through.

Figure 7.8 Diffusion across a semi-permeable membrane.

This is the principle utilized for dialysis.

Ultrafiltration (convective transport) is the bulk movement of water molecules containing permeable solutes through a semi-permeable membrane (Fig. 7.9). Water molecules are small and can pass through all semi-permeable membranes. The driving force for ultrafiltration can be either an osmotic gradient or hydrostatic pressure. Haemofiltration is based on the principle of ultrafiltration.

Figure 7.9 Ultrafiltration through a semi-permeable membrane.

The volume of filtrate removed can be in excess of 2 L/h, and to maintain CVS stability, fluid must be replaced concurrently. The fluid used as replacement should be isotonic and should aim to replace the solutes lost as filtrate that would otherwise be selectively reabsorbed by the ‘normal’ kidney.

Types of filtration circuit

Continuous arteriovenous haemofiltration (CAVH) (Fig. 7.10a)

Blood flows from an artery (A) into the filter to produce an ultrafiltrate before return to the body via a vein (V). Used as an intermittent technique, this was the earliest method of renal replacement therapy. A systolic BP >90 mmHg is adequate for a driving pressure.

Figure 7.10 Types of filtration circuit. (A) Continuous arteriovenous haemofiltration (CAVH). (B) Continuous venovenous haemofiltration (CVVH). (C) Continuous arteriovenous haemodiafiltration (CAVHD). (D) Continuous venovenous haemodiafiltration (CVVHD).

Continuous venovenous haemofiltration (CVVH) (Fig. 7.10b)

Continuous arteriovenous haemodiafiltration (CAVHD) (Fig. 7.10c)

The circuit is the same as that for CAVH, but ultrafiltrate is added as a countercurrent across the membrane to generate diffusive and convective clearance. This produces an ultradiafiltrate (UDF).

Continuous venovenous haemodiafiltration (CVVHD) (Fig. 7.10d)

This is the same circuit as for CVVH but with the addition of dialysate to increase clearance of waste products.

Intermittent versus continuous filtration

Fluid removal in intermittent dialysis has to occur over a short period of time, but if it is too rapid or too large a volume, hypotension will result. A longer duration of dialysis at a lesser ultrafiltration rate is more appropriate for ITU patients, who are often haemodynamically unstable.

Intermittent dialysis is less effective than continuous dialysis because waste products and water in the interstitial compartment do not have time to equilibrate with blood. Following a short period of dialysis, urea and creatinine levels are initially low, but will rise relatively quickly as these waste products equilibrate with the lower levels in the blood.

Intermittent, large-volume dialysis also risks a disequilibrium syndrome in which a relatively high intracellular concentration of urea and creatinine following dialysis causes an osmotic effect, drawing water into cells.

Techniques using dialysate fluid reduce the chance of electrolyte imbalance as serum electrolyte levels equilibrate with the dialysate fluid; for example, if the dialysate fluid’s K⁺ is 4.6 mmol.L⁻¹ and the serum K⁺ falls below that level, K⁺ will diffuse across the membrane until the level reaches 4.6 mmol.L⁻¹. Equally, the reverse would happen if the patient’s level was greater than that of the dialysate.

Replacement fluids

Bicarbonate ions are lost by filtration, resulting in a metabolic acidosis unless replaced.

Bicarbonate ions are not present in most replacement solutions because:

• bicarbonate ions are converted to carbonates which dissociate, with resultant loss of CO₂ by diffusion through the container wall

• mixtures of bicarbonate and calcium precipitate to form calcium carbonate.

Lactate is used as a bicarbonate substitute, but depends upon adequate hepatic metabolism to convert it to new bicarbonate ions via the TCA cycle. In patients with hepatic dysfunction or lactic acidosis, bicarbonate containing replacement fluid (without calcium) or lactate-free replacement fluid (where sodium bicarbonate is infused independently of the circuit) must be used as an alternative.

Anticoagulation

Contact of blood with plastic surfaces and filter activates clotting cascade. Anticoagulation is therefore required to prevent thrombus deposition in the circuit. Usually with heparin (5–10 u.kg⁻¹, aiming to keep APTR 2.0–2.5. Prostacyclin (2.5–10 ng.kg⁻¹.min⁻¹) may be used as an alternative to prevent heparin-induced thrombocytopenia, but it can cause hypotension.

Many ITU patients have coagulopathies and do not require further anticoagulation.

Potential problems

• Arterial/venous access

• Infection

• Bleeding

• Heat loss

• Emboli – air/thrombus

• Disconnection

• Intravascular fluid depletion

• Drug/colloid clearance.

Drug clearance

The effect of renal failure and renal replacement therapy on drug levels is variable and difficult to predict. Many factors affect drug levels in these patients, such as whether the drug is normally eliminated entirely or partly by renal excretion and the effect of plasma protein levels (often low) on drug binding. For many drugs with minor or no side-effects, no modification to dose is required. For more toxic drugs with a low therapeutic index, monitoring of drug levels is required. The loading dose should be the same as that used in a patient with normal renal function, but subsequent doses should be given according to drug levels. It is important to try to avoid nephrotoxic drugs in patients with renal failure, or if they are used, care should be taken to prevent doses reaching nephrotoxic levels.

Elimination of aminoglycosides, vancomycin, aminophylline and digoxin is greatly reduced in patients with renal failure; they require dosing based on drug levels to prevent toxicity.

Bibliography

Hall N.A., Fox A.J. Renal replacement therapies in critical care. Contin Edu Anaesth, Crit Care Pain. 2006;6:197-202.

Joannidis M. Continuous renal replacement therapy in sepsis and multisystem organ failure. Semin Dial. 2009;22:160-164.

Mallick N.P., Gokal R. Haemodialysis. Lancet. 1999;353:737-742.

Short A., Cumming A. ABC of intensive care. Renal support. BMJ. 1999;319:41-44.

Scoring Systems

Problems of predictive scoring systems

Physiological reserve. Prior health is an important predictor of outcome in acute illness. Best correlation with age.

Selection bias. Errors in predictive power if patients used to create the database are different to those being evaluated.

Lead-time bias. Patients delayed in their arrival to ITU may fare worse than those admitted earlier, despite having the same disease.

APACHE (acute physiology and chronic health evaluation)

Designed to predict outcome in groups of ITU patients, such as risk of death, ICU length of stay, type and amount of therapy, and nursing intensity. Standardized mortality rate (actual versus predicted mortality rate) can also be used to assess unit performance. Daily APACHE risk predictions are a precise measure of the patient’s condition and may aid in patient treatment decisions. An increasing score is associated with increasing risk of death.

APACHE I was developed in 1981 based on 34 variables, age and previous health. Criticisms that it considered unmeasured variables as normal and that it involved too many variables led to the development of APACHE II in 1985.

APACHE II was simplified to 12 physiological variables, age and chronic health evaluation and one of 34 admission diagnoses. Its main criticisms are failure to compensate for lead-time bias and the ability to select only one diagnostic criterion. This led to poor performance when applied to trauma victims, patients receiving TPN, severely ill postoperative patients, patients with myocardial infarction and those with congestive cardiac failure.

APACHE III was developed in 1991 and is based on multiple regression values from a database of approximately 300 000 patients. The APACHE III system consists of three scores:

• age – maximum score 24 points

• chronic health evaluation – maximum score 23 points

• acute physiology score – 17 variables using the worst value in 24 h, giving a maximum score of 252 points. The variables are:

Mean blood pressure	Respiratory rate	Temperature
Pulse	Glasgow Coma Scale	Urine output
Haematocrit	White cell count	Blood pH
PaO₂	P_aCO₂	Serum sodium
Serum albumin	Serum bilirubin	Serum glucose
Serum creatinine	Blood urea nitrogen