Haematological problems are common in the ICU. Most are related to blood loss, the need for large volume blood transfusions and the development of coagulation disorders. Bone marrow failure and immunosuppression are also problems in some patients.

Anaemia is common in critically ill patients and often necessitates repeated blood transfusion. Anaemia may be the direct result of an underlying disease process, but more commonly is multifactorial. Common factors that may contribute to anaemia are listed in Box 10.1.

Where possible, the underlying causes of anaemia should be addressed. Blood loss, including that from blood sampling, should be minimized.

There has been significant debate about the optimum haemoglobin level for critically ill patients and the level at which transfusions should be instituted (the so-called transfusion threshold). The following are key considerations:

Oxygen carriage and delivery. The oxygen content of blood is given by Hb × SaO₂ × 1.34. Raising haemoglobin is an effective way of improving oxygen content and delivery. (See Oxygen delivery and oxygen consumption, p. 68.)

Myocardial function. Myocardial ischaemia and diastolic dysfunction may occur in the stressed heart when the haematocrit falls below 0.18. In the presence of coronary artery disease, the threshold is 0.24 or higher.

Rheology. In vitro and probably in vivo, blood viscosity is reduced as haematocrit falls below 0.24 and rises above 0.3. This may have implications for perfusion of the microcirculation, particularly in critical illness and following vascular surgery. Recent evidence suggests that both very low and very high haematocrits are associated with impaired tissue perfusion. Haematocrit can be estimated from Hb (g / dL) × 3/100.

Immunology. There is some evidence that transfusion induces a degree of immunosuppression, particularly massive transfusion. This is important following any major surgery, and especially so where surgery has been performed for malignant disease.

Finite risks associated with each individual blood product transfused (see below).

Recent work suggests that restrictive transfusion strategies produce the best outcome in critically ill patients, and that the optimal level in most cases is Hb of 8–10 g / dL. Transfusion thresholds of around 7 g / dL are reasonable in most patients, with an aim to raise the haemoglobin to no more than 10 g / dL. The rate of transfusion will depend on the clinical circumstances, but for many ‘routine transfusions’ the blood can be given slowly over a number of hours to avoid rapid volume loading effects.

The blood volume of a patient will vary considerably with size and age, but is approximately 80 mL/kg, lean body mass.

In the UK, blood is donated by unpaid volunteers, who undergo general health screening. Whole blood is collected into a citrate-based anticoagulant solution (chelating calcium to prevent clotting) and then further separated to yield individual blood components such as platelets, fresh frozen plasma (FFP) and cryoprecipitate.

All donations are serologically tested for HIV-1 and HIV-2, hepatitis B and C, syphilis and cytomegalovirus (CMV). CMV-free blood components are used for immunosuppressed patients and those under 1 year of age.

Recently, potential transmission of new variant Creutzfeldt–Jakob disease (vCJD) has become a concern. It is likely that in the near future screening of donors for vCJD will become available. Currently in the UK all blood products have white cells removed (leucodepletion) as a precaution against vCJD transmission (white cell count < 5 × 10⁶). Continuing concerns regarding the potential for carriage and transmission of vCJD by the UK blood donor pool has resulted in some plasma products being sourced from outside the UK, principally from the USA.

Despite these and other measures, there remain risks associated with the use of all blood products. Although the most commonly reported major transfusion-related injury results from the wrong blood being transfused into the wrong patient, even when errors such as this are eliminated, there are finite risks associated with infection, immunological and idiosyncratic reactions. Therefore, you should not undertake transfusion of blood products lightly. In many cases, conservative management may be more appropriate. (See Risks and complications of blood transfusion, p. 252.)

The following component blood products are available.

Red cells

Whole blood was traditionally used for the replacement of blood loss. Whole blood is now rarely available. Red cell concentrates are usually used. The plasma component of whole blood is removed by the National Blood Transfusion Service (removes clotting factors and albumin) and the red cells are suspended in an additive solution prior to their issue to hospital transfusion laboratories. The two commonly used additive solutions are:

citrate, phosphate, dextrose and adenosine (CPDA)

sodium chloride, adenosine, glucose and mannitol (SAGM).

The blood products potentially available for red cell replacement are shown in Table 10.1.

TABLE 10.1 Red cell products

Fresh frozen plasma

Fresh frozen plasma (FFP) may be from a single donor or recovered from pooled donors. The volume of units provided therefore varies from 150 to 500 mL. FFP contains both labile and stable factors, including albumin, gamma globulin, fibrinogen and factor VIII. Usually 2–4 units are given when required for coagulopathy (prolonged PT). Plasma products should be ABO compatible.

Recent concerns regarding virus transmission have resulted in treated plasma products becoming available. There are currently two: methylene blue treated and solvent detergent treated. The characteristics of available plasma products are compared in Table 10.2.

TABLE 10.2 Fresh frozen plasma products

Recent reports of procoagulant complications with solvent detergent-treated FFP are thought to be due to relative protein S deficiency. This, together with the problems associated with pooled donors (1000 donors per batch), may limit its widespread use. However, the field is moving forward all the time. There is much current interest in production of a universal (ABO independent) pathogen inactivated plasma that would be safe for use in all blood groups, thereby eliminating the risk of ABO incompatibility and major transfusion reactions.

As with all blood products, the need for fresh frozen plasma should be critically assessed in every case. Although it is well established that fresh frozen plasma corrects coagulopathy and has a self-evident effect on the bleeding, there are no randomized controlled trials showing clinical benefit from the use of fresh frozen plasma. Nevertheless, solvent detergent and methylene blue treated plasma remain licensed in Europe and are an attractive option for patients needing massive or repeated transfusion. (See Major haemorrhage, p. 251.)

Cryoprecipitate

Cryoprecipitate is provided as one to six single donations per pack, suspended in 10–20 mL plasma. It contains fibrinogen and factor VIII. It is used to correct coagulopathy where fibrinogen levels are depleted. Six units generally raise fibrinogen levels by approximately 1 g/L.

Platelets

Platelets may be single donor, or pooled (five or six donors). Six units typically raise the platelet count by approximately 10–20 × 10⁹/L. Platelets should ideally, but not necessarily, be ABO compatible. Platelet transfusion should only be used when absolutely necessary. Apart from all of the other risks of blood transfusion, platelets are kept at room temperature and can rarely give rise to overwhelming bacterial infection due to infusion of infected product. Although this is very rare, it is catastrophic, and is often fatal.

The complications of blood transfusion are discussed below. The biggest cause of major ABO incompatibility reactions is human error. Most commonly these result from failure to follow approved procedures. The following notes are applicable to all blood products.

Requesting blood products

Ensure that you follow national / locally agreed procedures and guidelines.

Ensure blood samples are carefully labelled and that the accompanying request forms are accurately completed.

Ensure that blood products are appropriately prescribed.

Check recipient identity

Before administering any blood product, it is essential that the product is matched to the intended recipient. The product label always states the nature of the contents (e.g. whole blood, FFP), its storage temperature, expiry date and time, ABO and RhD grouping, donation or batch number and details of the patient against whom it has been cross-matched. These details must be verified against the wrist band of the patient who is the intended recipient.

When possible, ask the patient to confirm his or her identity and that the details on the identification band are correct.

Ensure that the patient’s name and identification number on the wristband match those on the intended blood product.

Ensure that the ABO blood group, rhesus blood group and unit identification number are all correct.

In some hospitals, a form accompanies blood products, identifying the units issued; this is intended as a record to be placed in the patient’s notes. These forms are not intended to be used as part of the checking procedure, and do not help to ensure that the correct unit of blood is given to the correct patient. The only acceptable checking process is to confirm that the patient details on the product label and those on the patient’s wrist band are the same.

Administration

Blood products are stored under carefully controlled conditions in the blood bank. Once removed from controlled storage they must be used within set time frames as indicated on the pack, typically as shown in Table 10.3.

TABLE 10.3 Storage and use of blood products

Blood should ideally be given through a large cannula (minimum 18 gauge) to avoid haemolysis. Standard giving sets with 170 μm filters are adequate. Microaggregate filters are not necessary and must never be used when giving platelets. If blood is given through the same set as other fluids, calcium-containing solutions should be avoided. Consider the use of a blood warmer.

A significant risk in managing major haemorrhage is the non-availability of appropriate blood when needed. Clear communication, sending blood samples to the laboratory in a timely fashion and prioritizing preparations (calling help, preparing blood warmers, cell savers and rapid infusion devices) all reduce the risk of disaster.

Where possible, control the source of the bleeding, for example by direct pressure.

Ensure adequate vascular access: at least 2 × 14-gauge peripheral lines or single large-bore cannula such as 8.5-Fr introducer sheath. This need not necessarily be inserted into a central vein; a peripheral vein may well be easier to cannulate in an emergency, and just as adequate.

Continue background or maintenance fluids to provide free water, glucose and electrolyte requirements.

Commence initial volume replacement with a crystalloid or simple colloid such as modified gelatin (e.g. Gelofusine or Haemaccel).

Continue with packed cells to maintain a haematocrit of 0.26–0.32.

After 5 units of blood, consider changing to whole blood if available and / or giving FFP to minimize the effects of dilutional coagulopathy.

Recheck FBC, U&Es, clotting and thromboelastogram (TEG) if available.

Ensure adequate treatment of coagulopathy. In particular, keep the ionized calcium above 0.85 mmol / L. Maintain normothermia with active warming of the patient if necessary.

After 10 units, recheck clotting. Consider cryoprecipitate, platelets and further FFP (see Coagulopathy below).

The successful management of major haemorrhage usually depends on surgical control of the bleeding. Therefore as soon as significant haemorrhage is identified or suspected inform your consultant and the senior members of the relevant clinical team directly. Notifying the most junior member of other teams, and waiting for him / her to make a decision and refer the matter to more senior members of the team, is likely to waste valuable time. If available, consider the use of a cell saver, to reduce the need for ‘bank’ blood.

Complications of blood transfusion include fluid overload, hypothermia, hypocalcaemia, acidosis and dilutional coagulopathy. ARDS and multiple organ failure are also considered to be complications of massive transfusion. Bacterial contamination of blood occurs rarely and is usually fatal (platelet transfusion carries the greatest risk because of the need to store at room temperature).

Recent figures for risk of transmission of viral infection suggest the risk of HIV transmission is around 0.14 per million donor exposures and the risk of hepatitis B and C are a little higher 1.7 and 0.8 per million respectively.

Acute transfusion reactions are relatively uncommon. They include:

haemolytic (75% due to ABO incompatibility)

anaphylactic (antibodies to IgA in IgA-deficient patients)

febrile white cell reactions (antibodies to leucocyte antigens)

transfusion-related acute lung injury (antibodies to leucocyte antigens)

urticaria (1–2% of transfusions).

Severe haemolytic reactions due to ABO incompatibility are rare and usually result from the wrong blood being given. Febrile reactions are less common with leucocyte-depleted blood but may still occur.

Jehovah’s Witnesses have strong religious views regarding the acceptability of blood and blood products. Wherever possible, you should clarify the individual’s wishes and religious views, which should be respected. Failure to do so may constitute an assault. The Jehovah’s Witness patient who requires intensive care may of course be unable to express their wishes and to give or withhold informed consent. In the case of adults, advice may be sought from next of kin or the Jehovah’s Witness hospital liaison committee. (See Ethical and legal issues, p. 29, and Death and different cultural views, p. 440.)

The following notes are broad guidelines only, and may not be completely acceptable to all individuals:

Blood (red cells, whole blood), FFP, platelets may not be given to Jehovah’s Witnesses under any circumstances.

Predonated blood is generally not acceptable.

Albumin and cryoprecipitate is accepted by some (but not all) Jehovah’s Witnesses.

Factor concentrates: concentrates of specific factors (for example, factor XI, IX and VII) are generally accepted. Seek the advice of your hospital liaison committee.

Extracorporeal circuits. The majority of Jehovah’s Witnesses accept blood that has been passed through an extracorporeal circuit. This allows for cardiac surgery (cardiopulmonary bypass) and for renal dialysis. Intraoperative cell salvage is generally acceptable provided the blood is reinfused immediately, at the time of surgery, and not several hours later on the ICU.

An individual contract / management plan may be prepared before treatment to clarify treatment options, when the patient is well enough to do this.

Management

Patients with an anticipated major haemorrhage should receive supplementation of iron and other haematinics preoperatively.

Erythropoietin (EPO) may be prescribed to stimulate red cell production. Side-effects are uncommon, but include hypertension, headache and stroke. Not all patients respond.

Minimize any potential blood loss during surgery,

Consider aggressive hypervolaemic haemodilution to reduce the haematocrit of any blood lost during surgery. Haemoglobin concentrations as low as 5g / dL may be tolerated (see below).

Consider techniques such as elective hypotension to further reduce blood loss.

Avoid unnecessary blood sampling.

Jehovah’s Witnesses who develop a coagulation disorder pose a special problem. The use of antifibrinolytic drugs (such as aprotinin) is generally acceptable and should be considered early on. If there is evidence of endogenous heparins (as evidenced by a prolonged APTT), consider giving protamine (see below). Management of the coagulopathy will depend on which clotting factors (if any) the individual is prepared to accept.

Much useful information has come from the management of such patients, with the realization that otherwise reasonably fit patients can survive for long periods with extremely low haemoglobin concentrations. Case reports cite survival down to Hb concentrations of 1–2 g / dL. However, for the older frailer patient with significant comorbidity, the reality is somewhat different and most patients who have a sustained haemoglobin below 5 g / dL will fail to improve and die days / weeks later from multiple organ failure.

To understand coagulation disorders you need to understand the mechanisms by which haemostasis is normally achieved. Following injury to a blood vessel, a series of events is initiated:

Vasoconstriction reduces blood flow in the damaged vessel.

Factor VII and platelets adhere to exposed tissue factor (TF) in the vessel wall and release a number of mediators, including ADP. More platelets are attracted, which rapidly form a temporary haemostatic platelet plug.

The coagulation cascade is triggered, which results in the conversion of fibrinogen to fibrin. Cross-linking of fibrin molecules results in conversion of the primary platelet plug to an organized clot.

The classical coagulation cascade is shown in Fig. 10.1. This represents the situation as present in vitro. While it is useful for understanding the serine protease cascade, and for working out in the laboratory the nature of coagulation disorders it does not represent the situation in vivo.

Fig. 10.1 Coagulation cascade. Coagulation pathway as classically described, showing intrinsic, extrinsic and common (shown in bold type) pathways. In-vivo binding of factor VII to exposed tissue factor may be the primary initiating mechanism.

In vivo, it is the interaction between factor VII, platelets and the vessel wall that is pivotal to the coagulation process.

Factor VII forms a complex with tissue factor (TF) on the surface of cells at the site of injury, including platelets. This complex then initiates the coagulation cascade by activating factors X and IX, generating the so-called ‘thrombin burst’ and accelerating clot formation. This is shown in Fig. 10.2.

Fig. 10.2 In-vivo clotting mechanism. Note pivotal role of factor Vllla, platelets and generation of thrombin burst leading to formation of fibrin plug.

The pivotal role of factor VII has led to the development of activated factor VII (VIIa) as a proposed therapy for uncontrolled bleeding. (See Activated factor VII, p. 261.)

There are also mechanisms within the body to prevent clot formation in healthy vessels and to dissolve established clots. These fibrinolytic pathways are shown in Fig. 10.3.

Fig. 10.3 Fibrinolytic pathways. Plasmin is a non-specific proteolytic enzyme that degrades factors V and VIII, fibrinogen and fibrin, and also inhibits both the coagulation cascade and the conversion of fibrinogen to fibrin. α₂-Antiplasmin and α₂-macroglobulin inhibit plasmin.

Under normal circumstances, therefore, there is a constant balance maintained between procoagulant mechanisms and anticoagulant mechanisms. If this balance becomes disturbed, bleeding or thrombosis may result.

The term coagulopathy is generally used in respect of those disorders of haemostasis that produce a bleeding tendency. (Prothrombotic disorders are considered below.)

Causes

Coagulopathy may result from failure of clot formation, failure of clot stabilization or excessive activation of fibrinolysis. Often more than one process is involved, and the early involvement of a haematologist is advisable. Typical causes of coagulopathy are shown in Box 10.2.

Box 10.2 Typical causes of coagulopathy

Congenital	Acquired
Haemophilia A (factor VIII)	Acquired / functional factor deficiency
Haemophilia B (factor IX)	Dilutional coagulopathy
Von Willebrand’s disease	Thrombocytopenia
Other factor deficiencies	Sepsis
	Hypothermia
	Hepatic dysfunction
	Vitamin K deficiency / malabsorption
	Renal failure
	Drugs

In intensive care, acquired causes of coagulopathy are much more common than congenital causes. Many factors may contribute; these are listed below.

Sepsis may produce bone marrow suppression (thrombocytopenia) and triggers inflammatory cascades, which activate both coagulation and fibrinolysis.

Reduced gastrointestinal absorption of fat-soluble vitamins (A, D, E, K) leads to reduced manufacture of vitamin K-dependent factors (II, VII, IX, X, protein C).

Reduced hepatic reticuloendothelial function permits increased circulating levels of endogenous heparinoids (potentiating antithrombin III and inhibiting factors V, X).

In renal failure uraemia impairs platelet function.

Massive transfusion and vigorous fluid loading can result in ‘dilutional coagulopathy’.

Citrate anticoagulants may persist in the circulation for some time, chelating calcium and potentially reducing cardiac contractility. Additionally, some synthetic colloids may impair platelet function (dextrans, hetastarch in particular).

Activated fibrinolysis leads to consumptive coagulopathy.

Effects of heparin given for DVT treatment / prophylaxis, as anticoagulation for extracorporeal circuits, during cardiac / vascular surgery etc.

Patients who require therapeutic anticoagulation for underlying conditions, for example, patients with a mechanical prosthetic heart valve or previous DVT / PE requiring long-term anticoagulation, pose a particular challenge, since continued anticoagulation may lead to risks of bleeding. This may require a balanced judgement about the relative risks and benefits of continuing anticoagulation. Seek appropriate advice, both from your consultant and from a haematologist with an interest in anticoagulation.

Investigations

Basic investigations include platelet count, prothrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen, and fibrinogen breakdown products (FDPs / D dimers). If a specific factor deficiency is considered likely, then individual factor assay may be appropriate. Seek haematological advice. Normal ranges are shown in Table 10.4.

TABLE 10.4 Coagulation tests

	Normal range	Significance
Platelets	150–450 × 109/L	See thrombocytopenia below
PT	12–14 s (INR = PT/control; normal INR = 1)	Extrinsic and common pathway Marker of hepatic dysfunction / vitamin K deficiency Used to monitor warfarin therapy
APTT	30–40 s	Intrinsic and common pathway Used to monitor heparin therapy
TT	10–12 s	Tests conversion of fibrinogen to fibrin Prolonged by heparin/FDPs/D dimers
Fibrinogen	>2 g/L	Reduced in dilutional coagulopathy, liver failure, fibrinolysis (DIC)
D dimers	<0.2 g/L	Increased in presence of fibrinolysis (DIC)

Thromboelastography (TEG)

While the in-vitro tests listed above can be useful diagnostically to determine the likely cause of a coagulopathy, they test individual aspects of the coagulation process rather than reflect the overall process of clot formation. The thromboelastograph can be used to provide a dynamic test of coagulation and fibrinolysis. This can be used to help identify the need for FFP, cryoprecipitate, platelets or antifibrinolytic therapy. An increasing number of hospitals have access to TEG devices in theatre and some critical care units. If available, you should familiarize yourself with the device and interpretation of the information that it gives you before you have to use it in an emergency situation. There is now a body of evidence to suggest that the management of coagulopathy based on interpretation thromboelastogram and related technologies not only helps in the management of haemostasis, but also reduces the quantities of clotting products required. A typical TEG trace is shown in Fig. 10.4.

Fig. 10.4 Thromboelastogram (TEG) from a normal healthy (non-pregnant) adult.

Management of coagulopathy

In general, coagulation abnormalities should only be treated if there is active bleeding or when the potential consequences of bleeding may be disastrous.

Ensure that the patient is adequately resuscitated. Oxygen, i.v. access and adequate volume or blood replacement. Correct hypothermia.

The commonest cause of bleeding in the postoperative patient is failure of surgical haemostasis. Surgical causes of bleeding must be excluded. Seek surgical advice.

Other underlying causes of bleeding and coagulopathy should be addressed. Do not forget inherited causes, e.g. the haemophilias, although these are rare.

Perform basic investigations of haemostasis, as above. The diagnosis should be clear from a combination of history, examination and the results of these tests.

If the APTT is prolonged, suspect heparinoids or other inhibitors. The laboratory may be able to repeat the APTT in the presence of a heparinase to help distinguish this.

If heparin is present, but you do not know how much has been given, give protamine 50 mg then repeat the APTT. If you know the dose of heparin given, then protamine 1 mg per 100 units of heparin given should provide adequate reversal.

If the PT and APTT are prolonged in the absence of exogenous anticoagulants give FFP, which should be administered in 2–4 unit aliquots until PT falls below 20 s. Additionally, if fibrinogen depletion is marked, consider cryoprecipitate 6 units initially.

If the platelet count is < 80 × 10⁹/L, give 4 units of platelets, although more may be required if the count is lower or where the response to transfusion is limited, for example in DIC. In renal failure, platelet function may be abnormal even if platelet numbers are adequate. Desmopressin (DDAVP) 20 μg as a one-off bolus releases peripheral stores of factor VIII:Rag. This increases platelet ‘stickiness’ and thereby improves function.

Check the ionized calcium: if below 0.85 mmol/L, this may contribute to coagulopathy. Give 2.5–10 mmol calcium chloride slowly. (See Hypocalcaemia, p. 207.)

In malabsorption states and liver disease, vitamin K 10 mg can partially correct clotting disorders.

Prothrombin complex concentrate

Patients who have significantly raised INR from absolute or relative warfarin overdose or other causes with significant bleeding can have the coagulation defect rapidly reversed by infusion of prothrombin complex (Beriplex). This is expensive and may carry the risk of overcorrection and a prothombotic tendency. However, it has the attraction of a low infusion volume, and viral safety. Moreover, it does not carry the risk of transfusion-related acute lung injury sometimes encountered with the use of fresh frozen plasma. Although outside current guidelines, prothrombin complex is increasingly used in a number of situations where traditionally fresh frozen plasma has been the mainstay. Seek advice from the Haematology Department.

Activated factor VII (VIIa)

There has been a great deal of interest recently in the role of activated factor VII in the management of coagulopathy and uncontrolled bleeding. Activated factor VII binds to exposed tissue factor on damaged endothelial surfaces and activates the coagulation process, leading to localized fibrin production and clot formation. There is increasing evidence that VIIa is effective in reducing bleeding when other measures have failed. (See Normal haemostatic mechanisms, p. 255.)

Thrombocytopenia is a common finding in critically ill patients. Common causes are shown in Box 10.3.

The normal range for platelets is 150–400 × 10⁹/L. However, a normal platelet count does not necessarily imply normal platelet function. Furthermore, patients with hypersplenism may exhibit a reduced platelet count, but with relatively well-preserved platelet function. Consider a functional test (for example, TEG).

In general terms, significant bleeding secondary to thrombocytopenia is uncommon unless the count is very low. Therefore, do not give prophylactic platelets unless the platelet count is less than 20 × 10⁹/L. Below this level, there is a risk of spontaneous intracranial haemorrhage. In the presence of active bleeding, it is reasonable to give platelets in order to keep the platelet count above 80–100 × 10⁹/L. In all cases, attention should be paid to addressing the underlying cause of the thrombocytopenia.

Heparin-induced thrombocytopenia (HIT)

Heparin-induced thrombocytopenia is an immune-mediated phenomenon which occurs in up to 5% of patients receiving unfractionated heparin and up to 1% of patients receiving low molecular weight heparin. Antibodies to the heparin–platelet complex lead to platelet activation and release of procoagulant mediators, resulting in both thrombocytopenia and major venous and arterial thrombosis. The latter may cause catastrophic limb ischaemia.

Consider the diagnosis if the platelet count falls by more than 50% after exposure to heparin (usually occurs within 4 days), or if thrombosis occurs despite heparinization and / or thrombocytopenia. Thrombosis may be manifest as deep vein thrombosis, pulmonary embolism, arterial thrombosis, thrombotic stroke, or myocardial infarction.

Send blood for HIT screen (discuss with haematology service).

Discontinue all heparin, including unfractionated and low molecular weight heparin.

Avoid platelet transfusion; bleeding is uncommon.

Consider anticoagulation if necessary with an alternative agent e.g. lepirudin (see BNF).

TTP–HUS

In thrombotic thrombocytopenic purpura (TTP), thrombocytopenia is associated with thrombosis (typically cerebrovascular). In the haemolytic uraemia syndrome (HUS), thrombocytopenia is associated with renal dysfunction.

In adults, TTP and HUS are considered to be different presentations of the same underlying pathological processes. Deficiencies of plasma protease (usually acquired but can be congenital) result in an abnormally large von Willebrand molecule, which activates platelets and causes (microvascular) thrombosis. Depending on the site of this, the clinical picture may be predominantly of renal failure (HUS) or neurological deficit (TTP).

Management is primarily supportive. Do not give platelets. This is associated with increased intravascular thrombosis and a poorer outcome. Fresh frozen plasma (to replace plasma protease) and / or plasma exchange may be of benefit.

Disseminated intravascular coagulation (DIC) is a complex process arising as a result of generalized activation of the inflammatory cascade. It involves activation of clotting within the microvasculature, with consequent tissue damage. There is a consumptive coagulopathy, where normal clotting fails to take place because of depletion of circulating factors. The process is generally accompanied by activated fibrinolysis, with clot instability. The breakdown products of fibrinogen (D-dimer) and fibrin (FDPs) are in themselves anticoagulant, thus adding an extra level of complexity.

The management of DIC presents a challenge, and early involvement of a haematologist is essential. The principles of management revolve around treating the underlying cause and adequate replacement therapy with FFP, cryoprecipitate and platelets. Some units employ antithrombin III in the treatment of DIC. Seek advice. There is no role for heparinization.

Purpura are small purple spots in the skin resulting from microvascular haemorrhage. Although these are not always pathological, e.g. ‘senile purpura’ is secondary to vessel fragility in old age, they should raise suspicion of underlying pathology. Typically, purpura are associated with thrombocytopenia (see above), but they are also seen in other conditions in which there is inflammation or damage to the microvasculature. Important causes are shown in Box 10.4.

A purpuric rash may also be seen distal to a point of venous obstruction or mechanical constriction. For example, it may be seen in the head and neck following hanging or strangulation.

A number of factors predispose to thrombosis in ICU patients (Table 10.5).

TABLE 10.5 Factors which predispose to venous thrombosis

Some patients are at particular at risk of thrombosis. These include the pregnant, the obese and those with carcinomatosis, pelvic / hip injuries or surgery, myeloproliferative disease, systemic lupus erythematosus (lupus anticoagulant) and conditions such as TTP–HUS (see TTP–HUS, p. 263).

There are also some familial disorders that lead to increased risk of thrombosis, including protein C deficiency, protein S deficiency and antithrombin III deficiency. If these are suspected, advice on investigation and management should be sought from a haematologist.

Venous thrombosis

Despite predisposing factors being common, clinically obvious venous thrombosis is relatively unusual in ICU patients. The incidence of venous thrombosis detected by ultrasound or venography is, however, much higher, and therefore all patients should receive prophylactic anticoagulation unless contraindicated. (See DVT prophylaxis, p. 63.)

Peripheral venous thrombosis may present clinically as a tense swollen painful limb. This may be either a lower or upper limb, the latter is usually related to subclavian vein cannulation. Unilateral jugular thrombus is usually clinically unrecognized. Central venous thrombosis of the IVC or SVC presents with swelling of the affected area, limbs, trunk or head and neck. Collateral venous drainage may be apparent (distended superficial veins).

The diagnosis of DVT can usually be made with compression and Doppler ultrasound, although other imaging (venography, contrast CT) is required to identify more central clot. The risk of DVT progressing to pulmonary embolism is significant.

Arterial thrombosis / embolization

Arterial thrombosis / embolization commonly follows vascular surgery in arteriopathies, but can also be seen following trauma or accompanying other prothrombotic conditions such as heparin-induced thrombocytopenia (HIT) (see p. 263). Peripheral small vessel occlusion (e.g. trash foot) is common after aortic surgery, but may also accompany other conditions such as bacterial endocarditis and mycotic aneurysm. This usually improves over time. Embolization of a peripheral (limb) artery results in a pale, weak, numb and pulseless limb with loss of Doppler signals. Central arterial thrombosis of the distal aorta also occurs rarely (saddle embolus).

Management of thrombosis

Treat the underlying cause where appropriate (e.g. plasma exchange for TTP).

Arterial thrombosis will usually require urgent embolectomy and / or surgical exploration.

Local infusion of thrombolytic agents may occasionally be an alternative.

Systemic anticoagulation is usually required either with therapeuatic doses of subcutaneous low molecular weight heparin, e.g. Enoxaprin 150 units / kg / day, or with unfractionated intravenous heparin, typically 5000 unit loading dose followed by an infusion of 18 units / kg / 24 h adjusted according to APTT.

In cases of thrombosis associated with heparin-induced thrombocytopenia, the newer drugs lepirudin and fondaparinux avoid exacerbating HIT and can be administered parenterally, but care is required in appropriate dosing as excessive anticoagulation can be problematic. Unlike heparin or warfarin, they cannot easily be reversed. Seek advice from a haematologist on drug selection, dosage and monitoring.

Local or systemic thrombolysis can be used for venous thrombosis, but is usually reserved for more extensive central clot and pulmonary embolism. Patients with recurrent DVTs / PE may benefit from a temporary (up to 3 weeks) or permanent IVC filter. Do not insert femoral vein catheters if such devices are present. (See Pulmonary embolism, p. 107.)

All critically ill patients on the intensive care unit should be considered to have some degree of altered immune function (immunocompromise). The causes are multifactorial as shown in Box 10.5.

Some specific patient groups are particularly at risk of relative immunocompromise. These include those with alcoholic liver disease and those suffering from malignancy. Such immunocompromise may be difficult to diagnose, as bone marrow function and other measures of response to infection may be apparently normal.

Patients with significant immunocompromise are, however, increasingly common in the ICU. Immune deficiency may be inherited (e.g. severe combined immune deficiency, SCID) or acquired. Most commonly, it is seen in patients with depressed bone marrow function, either as a result of an underlying disease process or as a result of treatment (e.g. following chemotherapy or immunosuppressive treatment following transplantation). Typical causes of significant immunocompromise are shown in Box 10.6.

Most severely immunocompromised patients referred to ICU have an established underlying diagnosis. Referral is usually precipitated either by respiratory failure, febrile neutropenia or sepsis syndrome. Often the time course is short, and the precipitating events may be unclear.

Management

The principles of management are largely the same as for the immunocompetent patient. Resuscitation and stabilization are the initial priorities.

Cross-infection with potentially resistant organisms is a major problem in intensive care, and can be disastrous in the immune-compromised patient. Patients should be barrier nursed in side rooms, ideally with positive pressure airflow, to protect them from further risk of infection. (See Infection control, p. 19.)

Some patients (e.g. those with haematological malignancy or on long-term chemotherapy) will have dedicated long-term venous access (Hickman line, Portacath, etc.). These can be used for resuscitation purposes; however, avoid accessing them for general purposes, to reduce the risk of infection in the catheter. It is often prudent to site a separate central venous catheter for short-term general use.

Baseline blood count including WBC count may give some clue as to the nature and severity of the immunocompromise (neutrophil / lymphocyte count). Temperature and C-reactive protein are useful markers of infection in neutropenic patients. Once the patient is adequately resuscitated, sources of sepsis must be sought. Clinical history and examination may give a strong clue to this. (See Pneumonia in immunocompromised patients, p. 145.)

Obvious sites of sepsis should be cultured and any abscesses drained.

Long term venous catheters are a common site of catheter-related sepsis. The infection may be obvious externally (swab entry site), or there may be no external signs of infection despite infected thrombophlebitis and / or catheter associated clot. Consider removal.

Send blood cultures. Remember previous multiple broad-spectrum antibiotic therapy may mask growth, predispose to resistant organisms, and further increase the risk of fungal infection.

Send urine and stool cultures.

Send serology for viral infection (CMV, herpes, EBV, etc.).

Send Candida and Aspergillus antigen tests.

Discuss likely or potentially unusual pathogens (e.g. protozoa) with the duty microbiologist and parent team. Suitable culture media should be used (e.g. bottles with antibiotic-binding resins). Alert the microbiology laboratory to this, so they can culture for unusual organisms. Antibiotic therapy is guided by culture results. In the first instance broad-spectrum antibiotics are usually required. Typical regimens are:

Piperacillin / tazobactam and tobramcycin as first-line agents.

Imipenem and vancomycin as second-line agents.

Fluconazole may be added as prophylaxis against fungi.

Co-trimoxazole may be added as prophylaxis against PCP.

Haematological malignancy

Patients with haematological malignancy are a tremendous diagnostic and therapeutic challenge in the ICU. A frequent problem is neutropenia following marrow ablation and transplantation. Recovery of WBC count may occasionally be hastened by administration of granulocyte–colony stimulating factor (GCSF).

The combination of haematological malignancy and requirement for IPPV carries a high mortality, especially where the pneumonia remains undiagnosed (patients with proven PCP often survive). When renal failure is added to this constellation, the mortality is even higher. As a result of this, some people are reluctant to admit patients with haematological malignancy to intensive care. Blanket policies of this sort are not appropriate and each patient should be considered on their merits. Those who are referred early and those who can be managed on non-invasive forms of respiratory support can do well.

AIDS

Patients known to have AIDS are usually on antiretroviral therapy. If tolerated, highly active antiretroviral therapy (HAART) is associated with maintenance of near normal health for many decades.

Patients with AIDS who develop PCP pneumonia or other intercurrent illness can have a reasonably good prognosis provided multiple organ failure does not supervene. CD4 count is a valuable guide to the likelihood of response to therapy and may give some indication as how advanced the patient is in their disease process. It may also be helpful to consider their viral load, as this is also a useful indicator of disease activity and progression, as well as the risk of infectivity to others. Seek specialist advice.

In a recent study, patients with AIDS showed a similar intensive care unit survival to those without AIDS. Factors predictive of survival included the APACHE score at admission and the CD4 count. Interestingly, HAART therapy does not appear to be a predictor.

Patients newly presenting with a disease characteristic of an immunosuppressed state, and who are subsequently diagnosed as HIV positive during the course of critical care admission, would not normally be commenced on antiretroviral therapy until after discharge from the intensive care unit, or recovery from the acute infective process. This is because acute institution of antiretroviral agents is unlikely significantly to affect the prognosis during the acute illness.

Organ transplant recipients

Solid organ transplant recipients no longer require special protection in the ICU during their perioperative course. The same is probably true should they develop opportunistic infections, surgical sepsis or pneumonia, as their immunosuppression is less severe than those with true immunocompromised states. If necessary, immunosuppression can be withdrawn and patients managed on steroids during an acute illness. Maintenance immunosuppressive therapy can then be reintroduced following recovery or in the event of rejection. Seek advice from the transplant team.