Anaemia should be fully investigated and treated before elective surgery.
Blood transfusion should be used to correct anaemia only as a last resort or in the emergency setting.
Abnormal results of haematological investigations without a clear aetiology should be discussed early with a haematologist in order to avoid delay in surgery.
Patients with a known haematological diagnosis should be discussed with a haematologist early in the pre-assessment period.
Patients taking long-term anticoagulation treatment need to be assessed for the need for perioperative anticoagulant bridging.
A multidisciplinary approach is recommended with patients who refuse the transfusion of allogeneic blood. Clear documentation of decisions and the patient’s wishes is essential.
Abnormal Full Blood Count
A full blood count (FBC) is an automated evaluation of the cells circulating in the bloodstream. It broadly assesses three fractions: erythrocytes (RBCs), leucocytes and platelets. Abnormally low or high counts can be a manifestation of disease.
Anaemia is a condition in which the number of red cells and therefore haemoglobin concentration (Hb) is reduced. It can be broadly categorised into microcytic (low MCV – mean cell volume), normocytic (normal MCV) and macrocytic (high MCV) anaemias. By World Health Organization criteria, anaemia is defined as Hb<120g/L in a female and <130g/L in a male. Preoperative anaemia is associated with increased post-operative morbidity and mortality (Dunne et al., 2002). Patient blood management (PBM) is a patient-centred and multidisciplinary approach to manage anaemia, minimise iatrogenic blood loss and harness tolerance to anaemia in an effort to improve patient outcome. PBM is becoming an important entity in preoperative assessment. Preoperative anaemia should be detected, investigated and treated. For a diagram presenting investigations and management of preoperative anaemia please refer to Figure 15.1.
Figure 15.1 Investigations and management of preoperative anaemia.
Iron deficiency may be a result of inadequate dietary intake, malabsorption or blood loss. As stores diminish, RBCs become microcytic and then hypochromic (pale due to a low concentration of Hb) and lastly the Hb falls. It can normally be diagnosed by measuring serum ferritin, which is a marker of iron stores. In inflammatory states or in patients with liver disease, ferritin can be falsely raised and full iron-binding studies may be required for diagnosis. Treatment should be with iron replacement therapy unless there is symptomatic anaemia or surgery cannot be delayed, in which case blood transfusion may be necessary. Iron replacement is normally given orally. Hb should rise by 20g/L in 3 weeks; however, if the patient is intolerant, fails to respond or more rapid correction is needed, then parenteral iron should be considered. The rise in Hb following IV iron and oral iron is similar at 12 weeks (Goddard et al., 2011); however, the improvement in Hb is faster with IV iron and so faster response will be seen by 6 weeks (Van Wyck et al., 2009). In the majority of patients with iron-deficiency anaemia, there is no place for erythropoietic stimulating agents, and their use should be discussed with a haematologist on a case-by-case basis.
Thalassaemias are a group of genetic disorders characterised by abnormal haemoglobin synthesis; alpha and beta-thalassaemias are the most common. Clinically significant thalassaemia is normally identified at birth or in early childhood and is therefore unlikely to be diagnosed in the pre-assessment clinic. Thalassaemia should be considered in patients with microcytosis with or without anaemia in whom iron stores are replete. Diagnosis is with haemoglobin electrophoresis. Most thalassaemia patients will not have an anaemia which precludes surgery.
Vitamin B12 and folate deficiency cause megaloblastic anaemia as both are needed for maturation of red cells. As well as macrocytosis and anaemia, severe deficiency can cause a pancytopenia along with laboratory features of haemolysis: hyperbilirubinaemia and elevated lactate dehydrogenase. Replacement therapy usually leads to a rapid normalisation of the Hb; blood transfusion is rarely indicated.
Drugs associated with a macrocytic anaemia include folate antagonists such as methotrexate and anticonvulsants such as valproate.
Reticulocytes (immature red cells) are larger than mature RBCs. Therefore, if their numbers increase there will be a macrocytosis. A reticulocytosis may result from rapid destruction of RBC (haemolysis), or recovery from anaemia after a haematinic deficiency or from haemorrhage. Blood film examination can help to clarify the aetiology of the haemolysis and a direct antiglobulin test (DAT) may be positive in autoimmune haemolysis. If haemolysis is suspected, a haematologist should be contacted.
Macrocytic anaemia can be seen in myelodysplastic syndrome (MDS) or myeloma, especially if it is associated with other cytopenias. Blood film examination may be helpful, but a bone marrow biopsy is usually needed to make a formal diagnosis. Surgery does not necessarily need to be delayed if a haematological condition is suspected; however, this decision should be made in conjunction with a haematologist.
Other causes of macrocytosis include liver dysfunction, hypothyroidism, excess alcohol and pregnancy.
While there are many causes for this, a normocytic anaemia is classically due to anaemia of chronic disease (ACD). It occurs in inflammatory, infective and malignant disease. There is often an inadequate erythropoietin (EPO) response and a reduced ability of the body to utilise iron stores (Weiss and Goodnough, 2005). Typically ferritin is high/normal, transferrin reduced/normal and total iron binding capacity reduced/normal. Reticulocyte count and serum EPO level may be inappropriately low for the degree of anaemia. Treatment of ACD should involve managing the underlying condition. It may include intravenous iron infusions and erythropoietin-stimulating agents. If correction of anaemia is urgent, then transfusion should be considered. The investigation and management of a normocytic anaemia should be guided by haematologists.
Patient Blood Management
The quickest way of normalising anaemia is through blood transfusion. However, this is an expensive resource with each red cell unit currently costing around £120; in addition, allogeneic blood transfusion may have serious adverse effects, e.g. transfusion associated overload and transfusion reactions. Anaemia in preoperative patients is associated with increased mortality, morbidity and length of stay (Shander et al., 2012). At the same time, there is increasing evidence that the correction of preoperative anaemia with red cell transfusion may contribute to this increase in morbidity and mortality (Shander et al., 2012). As a result, there is a clear need for alternative approaches to managing anaemia.
Patient blood management (PBM) is a multidisciplinary, evidence-based and individualised approach to optimising the care of patients who may require a blood transfusion; this focusses on measures for blood avoidance with appropriate transfusion triggers and the correct use of blood when it is needed.
In surgical patients, measures include preoperative optimisation of anaemia and haemostasis, intraoperative methods such as cell salvage and post-operative management of anaemia and haemostasis. It is therefore vital that the identification of anaemia is made early to ensure sufficient time to optimise the patient before surgery and to formulate an appropriate perioperative plan.
Polycythaemia is a high concentration of red cells in the blood. The British Committee for Standards in Haematology (BCSH) defines polycythaemia as a persistently raised haematocrit (Hct) >0.48 in women and >0.52 in men (McMullin et al., 2005).
An elevated Hct is commonly due to sampling technique or a reduction in circulating plasma volume (i.e. dehydration); this does not represent a true increase in red cell mass.
Genuine polycythaemia warrants further investigation; certainly Hct >0.6 in a man and >0.56 in a woman requires urgent assessment. Polycythaemia can be primary or secondary.
Primary polycythaemia (polycythaemia vera) is a myeloproliferative neoplasm (MPN) due to defective red cell precursors in the bone marrow, which leads to overproduction of RBCs. In primary polycythaemia, there may be an associated rise in platelets and/or white blood cells.
Secondary polycythaemia is due to an increase in EPO levels. This may be physiological, for example, in patients with chronic pulmonary disease. Non-physiological causes include EPO-secreting renal tumours and anabolic steroid use.
There is an association between elevated Hct and thrombosis (McMullin et al., 2005). In addition, an uncontrolled Hct in primary polycythaemia has been associated with increased morbidity and mortality in surgical patients (Wasserman and Gilbert, 1964). Therefore, a haematologist should decide whether venesection is necessary. The target Hct should be informed by the underlying cause of the polycythaemia and the risk of perioperative bleeding.
Thrombocytosis is defined as a platelet count above 450 × 109/L (Harrison et al., 2010).
It is commonly secondary to infection, inflammation, malignancy, iron deficiency or hyposplenism. Once a reactive thrombocytosis is excluded, investigations should aim to distinguish between the different causes of clonal thrombocytosis/MPN; the most common include polycythaemia vera, essential thrombocythaemia, myelofibrosis and chronic myeloid leukaemia.
Clonal thrombocytosis increases the risk of thrombosis, and this risk increases with age. Paradoxically, in extreme thrombocytosis (>1000 × 109/L) there can be an increased risk of bleeding (Bleeker and Hogan, 2011). The risk of thrombotic and bleeding complications with reactive thrombocytosis is low, even in the setting of extreme thrombocytosis (Buss, Stuart and Lipscomb, 1985).
A reactive thrombocytosis does not typically require preoperative intervention. Clonal thrombocytosis may require treatment preoperatively; this depends on the platelet count, the patient’s age and other risk factors for thrombosis and bleeding. A haematology opinion is recommended.
Thrombocytopenia is defined as a platelet count below the normal range for the normal population (approximately 150 × 109/L). However, further investigation is not usually necessary for an isolated thrombocytopenia unless the level is <100 × 109/L.
A low platelet count must be confirmed by blood film examination. FBC samples contain EDTA to prevent the blood coagulating before analysis, but EDTA can occasionally induce platelet clumping, causing a falsely low result. Examination of the blood film will demonstrate large platelet clumps and a repeat sample in a citrate tube will yield the true result.
Thrombocytopenia has many possible causes. Splenomegaly can cause pooling of platelets, leading to a reduction in circulating numbers. Reduced platelet production could be due to a genetic disorder, medications causing bone marrow suppression, alcohol abuse, myelodysplastic syndrome (MDS) or leukaemia. A number of conditions can cause increased breakdown of platelets and therefore lead to thrombocytopenia; these include disseminated intravascular coagulation (DIC) and immune destruction. Management should include a full history and examination with laboratory investigations of coagulation (including d-dimer level and fibrinogen), haematinic screen, liver biochemistry and a blood film.
Treatment depends on the aetiology of the thrombocytopenia and may include corticosteroids, intravenous immunoglobulin or platelet transfusions. Different invasive procedures require different minimum platelet levels; if a platelet transfusion is needed, administration should be as close to the procedure as possible. Involvement of a haematologist is paramount.
Leucocytosis and Leucopenia
Most causes of neutrophilia are reactive (i.e. secondary to infection, malignancy or inflammation). However, if it is persistent and no secondary cause is apparent, MPN should be considered. In MPN, unless the count is very high or other FBC indices are abnormal, then it would be unusual to delay surgery. However, discussion with a haematologist is encouraged.
Neutropenia is defined as a count of <1.5 × 109/L; it is a common finding (Gibson and Berliner, 2014). Benign ethnic neutropenia is thought to be the most common cause and affects 25–50 per cent of black Africans (Haddy, Rana and Castro, 1999). Neutropenia can also be secondary to autoimmune disease and medications. MDS and leukaemia are rarer causes and are often associated with other FBC abnormalities.
Neutropenia may result in post-operative infection and poor wound healing. Discussion with a haematologist is encouraged.
The number of circulating lymphocytes varies with age and is higher in children than adults; counts > 4 × 109/L in adults and >9 × 109/L in infants are abnormal (Macintyre and Linch, 1998). Causes include viral infections, smoking and lymphoproliferative disorders. In most cases, surgery need not be delayed; however, a haematologist should be consulted.
Lymphopenia (<1.5 × 109/L in adults and age-dependent in children) is a common finding, especially in elderly patients. The most common cause is a transient fall following a recent viral infection. Other causes include HIV, autoimmune disease and medications (e.g. chemotherapy or corticosteroids). Unless there are recurrent infections or other cytopenias, further investigations in the preoperative period are not necessary (Brass, McKay and Scott, 2014).
Abnormal Coagulation Profile
Routine coagulation testing is frequently performed in the preoperative setting with the assumption that it will identify patients who have an increased risk of bleeding. However, the prothrombin time (PT) and activated partial thromboplastin time (APTT) were developed as screening tests to identify patients with haemophilia or other coagulation factor deficiencies; they were not designed to determine the bleeding risk of patients with no personal or family history of bleeding. Unselected coagulation testing is a poor predictor of bleeding (Chee et al., 2008). Indiscriminate use of these tests can lead to further unnecessary testing and delays to surgery.
Patients undergoing surgery should have a full bleeding history taken, focussing on previous haemostatic challenges, family history, current medication and illnesses associated with coagulopathy such as liver disease. While recent publications discourage routine coagulation testing in patients with a negative bleeding history (Chee et al., 2008; Van Veen, Spahn and Makris, 2011), data are limited about operations associated with a high risk of morbidity and mortality from bleeding complications (e.g. neurosurgery). It is therefore essential to follow local and national guidance on preoperative coagulation testing.
Isolated Prothrombin Time Prolongation
This is usually due to therapy with coumarins (e.g. warfarin) or vitamin K deficiency. Other causes include liver disease, direct oral anticoagulants (DOACs) and rarely factor VII deficiency. While deficiency of factors X, V, II and fibrinogen can prolong the PT, these factors are involved in the common pathway of the coagulation cascade and deficiency would usually also prolong the APTT.
When a patient is taking a coumarin anticoagulant, the PT result (in seconds) will vary according to the sensitivity of the laboratory reagent (thromboplastin) used. The international normalised ratio (INR) was devised to standardise the results between laboratories, so that INR results are comparable. Each thromboplastin is assigned an International Sensitivity Index (ISI) by the manufacturer. The INR is calculated by dividing the patient’s PT by the control PT (which is the logarithmic mean normal PT of the population) raised to the power of the ISI. The INR is valid only for assessing the anticoagulant effect of coumarins and should not be used to describe any other prolongation of the PT.
Isolated Activated Prothrombin Time Prolongation
Common causes include unfractionated heparin (UFH), lupus anticoagulant (LA), low molecular weight heparin (LMWH) and the DOACs. Deficiency in coagulation factors VIII, IX and XI will prolong the APTT; von Willebrand disease (VWD) can be associated with a low factor VIII level and therefore prolong the APTT. Rarely, contact factor deficiencies can also cause an isolated prolongation. These include factor XII, high molecular weight kininogen (HMWK) and prekallikrein (PK). Deficiencies of contact factors are not associated with a bleeding tendency.
Prothrombin Time and Activated Prothrombin Time Prolongation
Common causes include vitamin K deficiency, liver disease, DIC and DOACs. Warfarin can also cause a prolonged APTT; however, it is only mildly prolonged in comparison to the PT or INR. Deficiencies of factors X, V, II and fibrinogen can prolong both tests.
Prolonged Thrombin Time
Thrombin Time ( TT) is commonly prolonged by anticoagulants, including heparins and dabigatran. Hypofibrinogenaemia or dysfibrinogenaemia will prolong the TT. Hypoalbuminaemia causes an in vitro prolongation of the TT. TT can also be affected by elevated levels of fibrin degradation products, e.g. D-dimer.
There is a markedly prolonged TT. Depending on the sensitivity of reagents used in the laboratory, the APTT is typically prolonged and the PT is usually within the normal range.
Investigative Approach to the Abnormal Coagulation Screen
If an obvious cause for the abnormal screen is not apparent, then further investigations will be necessary.
It is important to differentiate between a true factor deficiency and an inhibitor; mixing studies can help. For a 50:50 mix, patient plasma is mixed with pooled normal plasma in a 1:1 ratio; the APTT is then repeated with the mixed sample. In theory, when there is a factor deficiency, then the APTT should shorten to the time of the control plasma. If an inhibitor is present (either a specific coagulation factor inhibitor or an anti-phospholipid antibody, e.g. LA), then there should be no shortening of the APTT. However, in practice, a haematologist should be involved in the interpretation of mixing tests.
In the event of a full correction, coagulation factors VIII, IX, XI and XII should be tested; the results will need to be interpreted by a haematologist. If the factor assays are normal, then HMWK and PK can be assessed, although this is not usually necessary.
If an inhibitor is suspected, then a test for LA should be performed. The result will need to be interpreted by a haematologist. If LA is confirmed, then in the absence of bleeding symptoms further investigations are not normally necessary. If LA is not found, then specialist tests will be needed and should be performed under the guidance of a haematologist.
Prothrombin Time and/or Activated Prothrombin Time Prolongation
In a patient with a prolonged PT and or APTT, a therapeutic trial of vitamin K should be given to diagnose and treat vitamin K deficiency. If this does not correct the prolongation, then further investigations will be necessary.
TT and fibrinogen should be tested. A prolonged PT and APTT with a low fibrinogen and prolonged TT is most commonly associated with liver dysfunction, although hypofibrinogenaemia/dysfibrinogenaemia and DIC can have a similar laboratory profile. A D-dimer assay is useful in diagnosing DIC.
Mixing studies can be considered; LA typically only causes a prolonged APTT, although some PT reagents may be sensitive.
If liver disease is not present and the TT and fibrinogen assays are normal, then coagulation factor assays should be performed. Factor VII deficiency is the most likely diagnosis if there is isolated prolongation of PT. If both the PT and APTT are affected, then factors II, V and X should be assessed.
If factor assays are contemplated, then a haematologist should be consulted prior to the ordering and in the interpretation of the tests. An initial approach to the abnormal coagulation profile is shown in the Figure 15.2.
Figure 15.2 An initial management of the abnormal coagulation screen.
Management of Abnormal Coagulation
When a cause for the abnormal coagulation profile is found, the input of a haematologist will usually be required.
LA and contact factor deficiencies do not increase the risk of bleeding and normalisation of the coagulation screen is not required. They can, however, affect accurate perioperative monitoring of anticoagulation (e.g. in cardiothoracic surgery), and so consultation with a haematologist is recommended.
Any patient found to have an inherited bleeding disorder will need to have a perioperative plan prepared in consultation with a haematologist. It used to be thought that patients with liver disease and a prolonged PT and or APTT were ‘auto-anticoagulated’ and at increased risk of bleeding. However, the liver synthesises anticoagulant as well as procoagulant factors (conventional coagulation tests, e.g. PT and APTT are insensitive to natural anticoagulants); therefore, liver disease patients are not ‘auto-anticoagulated’ and may even be at increased risk of thrombosis (Søgaard et al., 2009). The practice of using plasma transfusions to correct abnormal coagulation test results is questionable. There is little evidence to support their use in liver patients (Tripodi and Mannucci, 2011), but some guidelines do recommend them (Hanje and Patel, 2007). The use of plasma products in these patients should be discussed with a haematologist.
Abnormal Antibody Screen
A positive antibody screen indicates that an antibody in the patient’s plasma has bound to one or more of the blood group antigens on the screening cells. When this happens, further testing is required to identify the antibody/antibodies.
An alloantibody binds to a blood group antigen which the patient lacks; patients may develop more than one alloantibody. These develop from prior exposure to blood products or pregnancy; antibodies can also be naturally occurring (e.g. anti-A and anti-B). Patients who receive recurrent blood transfusions (e.g. those with sickle cell disease) are particularly at risk. In most instances, patients should be transfused only with RBCs which lack the antigen to which their antibody binds.
An autoantibody binds to the patient’s own red cell antigen(s). Often these antibodies are pan-reactive (i.e. they also bind to blood group antigens of non-self red cells). They are most commonly seen in autoimmune haemolytic anaemia.
When a patient is already known to have an alloantibody, it is essential to alert the blood bank. Over time, antibody titres can fall below the level of detection, resulting in negative screening tests; this can lead to the transfusion of incompatible units and subsequent transfusion reactions.
If blood is required for surgery, the blood bank will arrange for compatible units to be made available. However, identifying these units can be technically challenging with complex serological tests requiring further patient samples. Communication between the clinical team and the haematologist/blood bank is essential to ensure that the correct tests are performed, that adequate stock is available and that transfusion is appropriate (to prevent waste of an expensive and limited resource).
Perioperative Anticoagulant Bridging
The perioperative management of patients receiving long-term anticoagulation is a common but challenging problem: 25 per cent of patients taking anticoagulant therapy require temporary cessation within 2 years (Healey et al., 2012). While continuing anticoagulation is associated with an increased risk of perioperative bleeding, discontinuation increases the thrombotic risk. Vitamin K antagonists, such as warfarin, have a long half-life so the drug will need to be stopped several days before surgery. In such instances, bridging anticoagulation (short-term use of an anticoagulant with a short half-life such as parenteral heparin) can reduce the risk of thrombosis (Spyropoulos, 2010). DOACs, with their short half-lives, have the potential to simplify perioperative anticoagulant management; however, there are few published data on this.
When to Stop Anticoagulation
Patients undergoing procedures with a low risk of bleeding (e.g. tooth extractions, cataract extraction, endoscopy without biopsy) do not require interruption of anticoagulation (Douketis, Johnson and Turpie, 2004).
For procedures associated with major bleeding (e.g. abdominal surgery), or procedures in which even localised bleeding can cause significant morbidity (e.g. spinal anaesthesia), anticoagulation needs to be interrupted.
When Bridging Is Indicated
Unless the patient is at low risk of peri-procedural thrombosis, then bridging therapy will be required.
Mechanical Heart Valves
The thrombotic risk of mechanical valves varies: prosthetic mitral valves confer a higher risk than aortic valves; older designs (e.g. caged-ball) confer higher risk than the newer bileaflet designs (Pibarot and Dumesnil, 2009). It has been suggested that patients at low risk for thrombosis (e.g. bileaflet aortic valve prosthesis without atrial fibrillation or other risk factors for stroke) do not need bridging (Douketis et al., 2012). However, this has not been prospectively validated and the default position should be that bridging is required for all patients with mechanical valves.
The CHA2DS2VASc score and its predecessor the CHADS2 score are validated scoring systems for the risk of stroke in non-valvular AF. They can also predict thrombotic and mortality outcomes in the perioperative period (see Chapter 9 of this volume; Van Diepen et al., 2014). The American College of Chest Physicians (ACCP) stratify patients’ perioperative thrombotic risk (low, medium or high) according to their CHADS2 score (Douketis et al., 2012):
Low: CHADS2 score 0–2 (without previous stroke or TIA)
Moderate: CHADS2 score 3 or 4
High: CHADS2 score 5 or 6, recent (< 3 months) stroke/TIA, rheumatic valvular heart disease
The ACCP recommends that patients in the low-risk group do not require perioperative bridging. However, assessment of thrombotic risk should be individualised and other factors considered (e.g. active cancer).
The risk of venous thromboembolism (VTE) recurrence is dependent upon factors including active malignancy, congenital thrombophilias, antiphospholipid syndrome and most importantly whether the index event was provoked or spontaneous. The guidance on when preoperative bridging is required varies. What follows is a suggested risk-stratification model based on the current guidance (Douketis et al., 2012; Keeling et al., 2011; Lip and Douketis, 2014):
Low: Previous VTE >3 months ago (in absence of active malignancy) with no other risk factors.
Moderate: Active malignancy; VTE between 4 and 12 weeks earlier; recurrent VTE; non-severe thrombophilia (heterozygosity for Factor V Leiden mutation or prothrombin gene mutation).
High: VTE within previous 4 weeks; severe thrombophilia (e.g. Protein C deficiency, antithrombin deficiency or combined abnormalities).
Bridging is recommended for those in the moderate and high groups, but not for low-risk individuals. Local guidance should be followed and if there is any uncertainty, discussion with a haematologist is advised.