Recommendations for the implementation of a Patient Blood Management programme. Application to elective major orthopaedic surgery in adults

Introduction

Patient Blood Management (PBM) is a holistic approach to the management of blood as a resource for each, single patient; it is a multimodal strategy that is implemented through the use of a set of techniques that can be applied in individual cases. Indeed, the overall outcome resulting from the implementation of PBM cannot be fully appreciated and explained simply by summing the effects of the single strategies and techniques used, since these can only produce the expected optimal outcome if used in combination¹. PBM is, therefore, a patient-centred, multiprofessional, multidisciplinary and multimodal approach to the optimal management of anaemia and haemostasis (also during surgery), to limiting allogeneic transfusion needs in the peri-operative period, and to appropriate use of blood components and, when relevant, plasma-derived medicinal products². The concept of PBM is not centred on a specific pathology or procedure, nor on a specific discipline or sector of medicine, but is aimed at managing a resource, “the patient’s blood”, shifting attention from the blood component to the patient who, therefore, acquires a central and pre-eminent role^3,4.

PBM combines the dual purposes of improving the outcomes of patients and reducing costs, being based on the patient rather than on allogeneic blood as the resource. For this reason, PBM goes beyond the concept of appropriate use of blood components and plasma-derived medicinal products, since its purpose is to avoid or significantly reduce their use, managing, in good time, all the modifiable risk factors that can lead to a transfusion being required⁵. These aims can be achieved through the so-called “three pillars of PBM” (Table I)⁵, which are crucial for making the paradigmatic shift that characterises the innovative, patient-centred approach: (i) optimising the patient’s erythropoiesis; (ii) minimising bleeding; and (iii) optimising and exploiting an individual’s physiological reserve to tolerate anaemia⁵. Each of these three key points is a strategic response to clinical circumstances that can cause adverse outcomes and necessitate the use of allogeneic transfusion therapy, namely anaemia, blood loss and hypoxia, respectively.

Table I.

The three pillars of Patient Blood Management (modified from Hofmann A et al.⁵).

Period	Pillar 1	Pillar 2	Pillar 3
	Optimisation of erythropoiesis	Minimisation of blood loss	Optimisation of tolerance of anaemia
Pre-operative	Detect anaemia. Identify and treat underlying disorder(s) causing anaemia. Re-assess the patient, if necessary. Treat iron deficiency, sideropenic anaemia, anaemia of chronic disease and functional iron deficiency (the so-called “iron-restricted erythropoiesis”). Treat deficiencies in other haematinics.	Identify and manage bleeding risk. Minimise iatrogenic bleeding. Careful planning and preparation of the procedure. Pre-deposit, in very selected cases.	Assess and optimise the patient’s physiological reserve for tolerance of anaemia and risk factors. Compare estimated blood loss with the individual patient’s tolerable blood loss. Formulate patient-specific blood management plan that includes the blood-conservation techniques appropriate for the individual case. Adopt restrictive transfusion thresholds.
Intra-operative	6. Adequate planning of the operation after optimisation of erythropoiesis.	5. Meticulous haemostasis and surgical techniques. 6. Blood-sparing surgical techniques. 7. Blood-conserving anaesthetic techniques. 8. Autologous blood transfusion strategies. 9. Pharmacological strategies and haemostatic agents. 10. Point-of-care tests	5. Optimise cardiac output. 6. Optimise ventilation and oxygenation. 7. Adopt restrictive transfusion thresholds.
Post-operative	7. Stimulate erythropoiesis, if necessary. 8. Be aware of drug interactions that can cause/worsen post-operative anaemia.	11. Careful monitoring of the patient and management of post-operative bleeding. 12. Rapid warming/maintenance of normothermia (unless there is a specific indication for hypothermia). 13. Autologous blood transfusion, if appropriate. 14. Minimise iatrogenic bleeding. 15. Management of haemostasis and anticoagulation. 16. Prophylaxis of upper gastrointestinal tract bleeding. 17. Prophylaxis/treatment of infections.	8. Optimise tolerance of anaemia. 9. Maximise oxygen delivery. 10. Minimise oxygen consumption. 11. Adopt restrictive transfusion thresholds.

PBM is, therefore, intended to guarantee all patients a series of personalised programmes, based on surgical requirements and the characteristics of the patients themselves, with the dual purposes of using allogeneic transfusion support appropriately and reducing the need for this resource. For this reason, PBM requires multidisciplinary and multimodal strategies to systematically identify, evaluate and manage anaemia (boosting, if necessary, individual physiological reserves) and to avoid or minimise blood losses.

It seems necessary to produce specific national standards. In fact, in the USA, PBM is the object of attention from the Association for Advancing Transfusions and Cellular Therapies (formerly known as the American Association of Blood Banks - AABB) which recently published the first edition of “Standards for a Patient Blood Management Program” precisely with the aim of supplying healthcare structures with solid elements for the standardisation of procedures and activities for implementing and/or optimising a PBM programme. The Society for the Advancement of Blood Management (SABM), also in the USA, has published a second edition of “Administrative and Clinical Standards for Patient Blood Management Programs”⁶ and the Joint Commission has published seven parameters for measuring the performance of healthcare structures in the field of PBM⁷.

Methodology for producing the recommendations for the implementation of a Patient Blood Management programme

The methodology used to prepare the recommendation grades is based on that described by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group⁸. According to the GRADE system, recommendations are classified by grade, expressed in Arabic numbers (1, 2), according to their strength, and in letters (A, B, C), according to the quality and type of evidence derived from the studies.

In detail:

• - Grade 1: the authors are certain that the advantages for health clearly outweigh the disadvantages, in terms of both risk and financial cost. This is, therefore, a strong recommendation.

• - Grade 2: the authors are less certain, the balance between advantages and disadvantages is less clear.

This is, therefore, a weak recommendation.

As far as concerns the quality and type of evidence drawn from the studies on which the recommendations are based, these can be classified into three levels:

• - Grade A: high level. The evidence is derived from numerous, consistent randomised trials without major limitations. It is unlikely that future research will change the conclusions of these trials.

• - Grade B: moderate level. The evidence is derived from randomised clinical trials, but with major limitations (such as inconsistent results, wide confidence intervals, methodological weaknesses). Grade B is also attributed to recommendations deriving from strong evidence from observational studies or case series (for example, treatment effects or the demonstration of a dose-response effect). Subsequent research could modify the conclusions of these trials.

• - Grade C: low or very low level. The evidence is derived from the analysis of observational clinical studies with less consistent results, or from experts’ opinions/clinical experience. Future research is likely to change the conclusions reached.

In brief:

• - Grade 1A: strong recommendation based on scientific evidence of high quality.

• - Grade 1B: strong recommendation based on scientific evidence of moderate quality.

• - Grade 1C: strong recommendation based on scientific evidence of low quality.

• - Grade 2A: weak recommendation based on scientific evidence of high quality.

• - Grade 2B: weak recommendation based on scientific evidence of moderate quality.

• - Grade 2C: weak recommendation based on scientific evidence of low quality.

The clinical-organisational course of the patient undergoing elective major orthopaedic surgery

Pre-operative period and role of the pre-operative assessment

The pre-operative assessment is performed in order to obtain clinical and laboratory information supplementary to that acquired from the clinical history with the aim of⁹:

• - confirming, or not, the planned diagnostic-therapeutic pathway (clinical management);

• - identifying previously undetected disorders that must be treated before the operation or which make it necessary to modify the surgical or anaesthetic technique;

• - limiting possible adverse outcomes or increasing the benefits for patient by modifying, if necessary, the clinical pathway;

• - facilitating the evaluation of potential risks for the patient (including the possibility of informing the patient of a potentially increased risk);

• - predicting possible post-operative complications;

• - establishing the baseline reference parameters, which can be used for subsequent post-operative evaluations;

• - considering the value of performing screening not related to the operation.

A pre-operative evaluation is always necessary when anaesthesia is planned.

Selective use of pre-operative tests, based on the history, clinical examination, and type and invasiveness of the surgical and anaesthesiological procedures, helps the management of the patient.

In order to shorten the time spent in hospital and optimise the planning of elective surgery, the pre-operative evaluation is performed in an out-patient setting (pre-admission), at an appropriate time before the operation (on average, 30 days before), and includes any additional clinical-diagnostic investigations needed^10–12.

In all candidates for elective surgery, a multidisciplinary and multimodal approach must be used, based on an agreed programme of co-ordinated interventions, aimed at peri-operative management of the “patient’s blood” as a resource. These interventions start with the pre-operative optimisation of erythropoiesis, haemostasis and tolerance of anaemia.

Depending on the organisation in different hospitals or healthcare authorities, pre-operative management of the resource “patient’s blood” must guarantee a structured diagnostic-therapeutic pathway involving at least three specialists: a surgeon, an anaesthetist and a transfusion medicine specialist; these professionals collaborate in the setting of a multidisciplinary outpatient clinic (the Anaemia Clinic) which takes the role of case manager. The Anaemia Clinic must allow for, when necessary, the inclusion and collaboration of other medical staff, such as experts in haemostasis and thrombosis, haematologists, cardiologists or other specialists with expertise in identifying and treating, in a multidisciplinary and multimodal method, underlying disorders in candidates for elective surgery. Furthermore, the patient’s general practitioner should be kept informed and, when possible, involved.

The pre-operative process starts in the surgical outpatient clinic (Figure 1). During this appointment the surgeon, having defined the indication for surgery, the grading of the intervention and its level of priority, takes the patient’s history, with the purpose, among others, of detecting risk factors for bleeding. To this aim it can be useful to administer a questionnaire for an initial evaluation of bleeding disorders (Figure 2^12,13), which can be followed by appropriate pre-operative investigations.

Figure 1.

Pre-operative flow-chart for patients undergoing elective major orthopaedic surgery and included in a Patient Blood Management programme.

Figure 2.

Initial evaluation of bleeding disorders (modified from Liumbruno GM et al.¹², Nichols WL et al.¹³).

Subsequently the patient is sent to the pre-operative administration office which, in conformity with locally used procedures, enters the patient into a specific pre-operative course aimed at booking and performing the pre-operative investigations and anaesthesiological evaluation, as well as adding the patient to the waiting list (surgical register).

Once the prescribed investigations have been performed in a centre for pre-operative examinations, the patient is sent to the pre-operative anaesthesiology clinic where the anaesthetist examines the results of the pre-operative investigations and evaluates the patient’s clinical state; if the patient is considered fit for anaesthesia, the specialist defines the “risk class”¹⁴, obtains informed consent to the anaesthesia and prescribes the anaesthesiological pre-medication.

If the pre-operative investigations that have been carried out are not considered sufficient, the anaesthetist orders other investigations that will be performed, when possible, on the same day as ordered or, at any rate, as soon as possible (dedicated pre-operative pathway). Furthermore, the anaesthetist may consider that the patient is not fit for anaesthesia because of the presence of intercurrent or chronic disorders or anaemia, which require specific investigations and/or treatment. In this case, after consultation with the surgeon and/or other specialists of the Anaemia Clinic, the anaesthetist may decide to prescribe what is deemed useful, setting a new date for the anaesthesiological re-evaluation, through the centre for pre-operative investigations.

The specialists of the Anaemia Clinic (anaesthetist, surgeon, specialist in transfusion medicine, expert in haemostasis and thrombosis, haematologist, or other specialist with expertise in identifying and treating, in a multidisciplinary and multimodal method, underlying disorders in candidates for elective surgery), by adopting the strategies and techniques indicated in the three pillars of PBM (Table I)⁵, have the task of setting up the multidisciplinary programme of co-ordinated interventions, aimed at the peri-operative management of the resource “patient’s blood”. The three pillars of the PBM in the pre-operative period are:

1. Optimisation of erythropoiesis: detect anaemia; identify and treat its underlying causes; re-evaluate the patient, if necessary; treat iron deficiency and iron-deficiency anaemia, anaemia of chronic disease and functional iron deficiencies, so-called iron-restricted erythropoiesis; treat deficiencies of other haematinics.

2. Minimise blood losses: identify and manage bleeding risk, minimise iatrogenic bleeding, plan the procedure carefully and prepare well; organise pre-operative autologous blood donation in very selected cases.

3. Optimisation of the tolerance of anaemia: assess and optimise the patient’s physiological reserve to tolerate anaemia and risk factors; compare estimated blood loss with the individual patient’s tolerable blood loss; formulate a personalised blood management programme that includes patient-specific blood-conservation techniques; adopt restrictive blood transfusion thresholds.

Information about strategies included in local PBM programmes should be supplied to all patients, preferably before their admission to hospital, because this improves compliance with diagnostic-therapeutic care pathways adopted in the healthcare structure.

• - It is recommended that the patient’s pre-operative evaluation, aimed at detecting any anaemia and optimising erythropoiesis, at identifying and managing bleeding risk as well as assessing and optimising the patient’s personal physiological tolerance of anaemia and risk factors, be carried out at least 30 days before the planned date of the operation, in order to allow more detailed investigations and/or arrange appropriate treatment [1C].

• - It is recommended that a structured questionnaire, aimed at picking up risk factors for bleeding, is included when taking the clinical history of an adult candidate for elective (orthopaedic) surgery [1C].

• - It is recommended that all adult patients who are candidates for elective major orthopaedic surgery for which a multidisciplinary programme of co-ordinated interventions has been established involving the adoption of pharmacological and non-pharmacological techniques aimed at optimising erythropoiesis, minimising blood losses and optimising tolerance of anaemia, before giving consent to one or more of the above-mentioned treatments, receive detailed information on their clinical state and strategies to limit allogeneic transfusion needs included in the local PBM programme; explanatory material prepared ad hoc by the hospital may be used for this purpose [1C].

Once the pre-operative process has been concluded with the patient deemed fit for anaesthesia, the patient undergoes the pre-admission surgical assessment, with the purposes of: a) re-evaluating the grading of the operation to be performed; b) re-assessing the patient’s general and specific conditions; c) planning the procedure; d) filling in the clinical records and single treatment chart with the pre-operative prescriptions (for example: specific therapies to optimise erythropoiesis, antibiotic prophylaxis, anti-thrombotic prophylaxis); e) collecting the informed consent to the intervention.

Having reached this point of the diagnostic-therapeutic care pathway, the patient is ready to be admitted to hospital on the planned day.

Intra-operative period

Once the patient has completed the pre-operative preparation for anaesthesia and the operation, on the planned day the patient is admitted to the surgical ward, where he or she is greeted by a nurse. The nursing staff fill in the nursing records with the planned assessment forms (for example: pain detection, detection of pressure ulcers) and check that the patient has been suitably prepared with respect to the pre-operative instructions provided.

Subsequently, the surgeon (and anaesthetist, if necessary), having checked the correct planning of the operation after optimisation of erythropoiesis and the use of other strategies and techniques indicated in the pre-operative period in the three pillars of the PBM (Table I)⁵, examines, and, if necessary, updates the clinical documentation, the single treatment chart and the informed consent to the procedure; the same surgeon marks the site of surgery.

Before entering the operating theatre, the nurse checks the anaesthesiological instructions, administers and records the treatments prescribed (general, antibiotic, pre-anaesthesia), takes any blood samples required and informs the patient about the surgical procedure; the nurse also checks that the patient is not wearing any jewellery or removable prostheses, checks the patient’s personal hygiene (if necessary, inviting the patient to wash again before the operation), shaves the patient, when required, and cleans the area of skin involved by the operation with antiseptics.

The patient is transported to the induction/recovery room, where the surgical team (anaesthetist and surgeon) re-evaluates him or her, “signs in” the patient (using the operating theatre check-list), informs the patient, if collaborative, about the procedures that will be performed and introduces one or more devices for intravenous access. The nurse monitors the patient’s vital parameters and administers antibiotic prophylaxis; all the activities are recorded on the nursing form and on the single treatment chart

In the operating theatre the same team carries out the “time out” checks (operating theatre check-list), performs the operation and provides anaesthesia, optimises the macrocirculation, maintains homeostasis and takes samples for any intra-operative blood-chemistry tests.

In this stage the three pillars of the PBM involve the following.

1. Optimisation of erythropoiesis: check appropriate timing of the surgery after optimisation of erythropoiesis.

2. Minimisation of blood loss: ensure meticulous haemostasis and use appropriate surgical techniques; adopt blood-sparing strategies; use blood-conserving anaesthetic techniques; use autologous blood transfusion, if foreseen by the personalised PBM plan drawn up by the case manager of the Anaemia Clinic; use pharmacological strategies and haemostatic agents; use point-of-care (POC) tests.

3. Optimisation of tolerance of anaemia: optimise cardiac output; optimise ventilation and oxygenation; adopt restrictive transfusion thresholds.

At the end of the intervention, the surgical team carries out the “sign out” controls (operating theatre check-list) and the patient is then transferred to the induction/recovery room for post-operative observation. Here the patient is re-assessed and monitored until his or her transfer to the ward; the same team prescribes the post-operative controls and treatments that must be carried out in the ward in the first 24 hours following the operation and which also have the purpose of correctly implementing the PBM strategies included in the multidisciplinary programme of co-ordinated interventions.

Post-operative period

After the operation, the patient can be transferred to the surgical ward from which he or she came, or depending on needs and clinical conditions, can be admitted to an intensive or sub-intensive care facility.

The post-operative controls and treatments for the 24 hours following the surgical intervention are prescribed by the anaesthetist and surgeon during the period that the patient regains consciousness.

During the post-operative period, besides the management of any urgent or emergency clinical situations, the co-ordinated interventions foreseen by the multidisciplinary programme, aimed at implementing the techniques and strategies included in the three pillars of PBM and reported here, must be carried out.

1. Optimisation of erythropoiesis: stimulate erythropoiesis, if necessary; consider drug interactions that can cause or enhance post-operative anaemia.

2. Minimisation of blood losses: ensure careful monitoring of the patient and management of post-operative bleeding; guarantee fast rewarming/maintenance of normothermia (except when there is a specific indication for hypothermia); use autologous transfusion techniques if foreseen by the personalised PBM drawn up by the case manager of the Anaemia Clinic; minimise iatrogenic bleeding; manage haemostasis and anticoagulation; administer prophylaxis against upper gastrointestinal tract bleeding; give prophylaxis against infections and treat any that occur.

3. Optimisation of the tolerance of anaemia: optimise the tolerance of anaemia; maximise oxygen delivery; minimise oxygen consumption; adopt restrictive transfusion thresholds.

The above-mentioned strategies must be implemented throughout the post-operative period until the patient’s clinical conditions have normalised, which is when the feasibility of the patient’s discharge from hospital is evaluated. On discharge the ward doctor prepares the discharge letter and gives it to the patient. The discharge letter contains a description of the salient diagnostic-therapeutic interventions performed, information on the domiciliary management of the surgical wound and any drains, as well as the pharmacological treatment (medication reconciliation).

Healing of the surgical wound and correct compliance with the prescribed domiciliary drug treatment must be checked at each outpatient follow-up until conclusion of the post-discharge process (30 days if the patient has not had a prosthesis implanted, 1 year if prosthetic material has been implanted).

The three pillars of Patient Blood Management

Optimisation of erythropoiesis

Pre-operative period

Detection of anaemia

The prevalence of pre-operative anaemia in surgical patients varies greatly, ranging from 5% (among geriatric patients with hip fracture) to 75.8% (among patients with stage D colon cancer according to Dukes’ classification)¹⁵; among candidates for orthopaedic surgery associated with moderate to substantial peri-operative bleeding, such as elective joint replacement (of the hip or knee) or urgent surgery (of the hip), the prevalence of pre-operative anaemia ranges from 24±9% to 44±9%, respectively¹⁶. Epidemiological studies on anaemia in patients undergoing elective hip or knee replacement surgery showed that the anaemia is hypochromic and microcytic in between 23 and 70% of cases¹⁶. Figure 3 presents a simple algorithm for the diagnosis of iron-deficiency anaemia^17–19. Among the same patients the prevalence of vitamin B₁₂ deficiency is about 12% and that of folate deficiency about 3%^20,21. The other forms of anaemia are due to inflammatory disease, chronic kidney disease or an unknown cause^22,23.

Figure 3.

Algorithm for the diagnosis of iron-deficiency anaemia (modified from Muñoz M et al.¹⁷–¹⁹).

Hb: haemoglobin; MCH: mean corpuscular haemoglobin; The dashed arrow indicates the need to measure the ferritin level; *: The mean corpuscular volume (MCV) is a relatively late indicator of iron deficiency in patients without active bleeding; in the presence of a low MCV the differential diagnosis from thalassaemia must be made; the MCV can be normal in the presence of vitamin B12 or folate deficiency, post-haemorrhagic reticulocytosis, initial response to oral iron therapy, alcohol intake or myelodysplasia. ** Additional laboratory tests: reticulocyte count; creatinine; C-reactive protein. *** Additional laboratory tests to evaluate the iron deficiency: ratio between soluble transferrin receptor (sTfR) and logarithm of ferritin; hypochromic red blood cells; haemoglobin content in reticulocytes.

Pre-operative anaemia, even if mild, in patients who are candidates for major surgery (not cardiac) is independently associated with a higher risk of morbidity and mortality at 30 days²⁴. Furthermore, a recent, retrospective study demonstrated that, in the peri-operative period of patients undergoing heart surgery, the sum of two risk factors, that is anaemia (haematocrit <25%) and transfusion therapy with red cell concentrates, had a greater, and statistically significant, effect on post-operative morbidity and mortality²⁵. Anaemia is, therefore, a contraindication to performing elective surgery.

It is recommended that elective major surgery is not performed in patients who are found to have anaemia until the anaemia has been correctly classified and treated [1B].

Anaemia is defined according to threshold values of haemoglobin (Hb) indicated by the World Health Organization (WHO)²⁶: children up 5 years: 110 g/L; children between 5 and 12 years old: 115 g/L; children between 12 and 15 years: 120 g/L; pregnant women: 110 g/L; women who are not pregnant (aged 15 years old or more): 120 g/L; men (aged 15 years or more): 130 g/L.

It is recommended that the pre-operative assessment of the patient, aimed at detecting any anaemia and optimising erythropoiesis, is performed at least 30 days before the planned date of the operation, in order to enable more detailed diagnostic investigations and/or plan appropriate treatment [1C]^6,11,12.

It is recommended that, if a state of anaemia is detected, the subsequent laboratory tests are directed at identifying iron deficiency or other nutritional deficiencies (folic acid and/or vitamin B₁₂), chronic kidney disease and/or chronic inflammatory disorders [1C].

It is recommended that the detection and treatment of anaemia, and any further related clinical-diagnostic investigations, are included within a global PBM strategy become a standard of care for all candidates for elective surgery, especially if the risk of peri-operative bleeding is substantial [1C].

Stimulation of erythropoiesis

These recommendations concern the management of oral and intravenous iron therapy, given the paucity of studies on the use of other haematinics in the peri-operative setting²⁷.

Since the pre-operative Hb value is the main, independent risk factor for requiring transfusion support with packed red cells, it is recommended that any nutritional deficiencies (iron, vitamin B₁₂, folate), once detected, are treated with haematinics [1C]¹¹.

It is suggested that the target Hb value before elective major orthopaedic surgery is at least within the normal range according to the previously cited WHO criteria [2C]¹¹.

Oral iron therapy

In adult patients with iron-deficiency anaemia who are candidates for elective hip or knee replacement surgery, oral iron therapy, particularly if combined with restrictive transfusion protocols, has been found to be effective in the treatment of pre-operative anaemia, in limiting transfusion needs and, in some cases, also reducing the duration of time spent in hospital^28–30.

It is suggested that oral iron therapy is used for the treatment of pre-operative iron-deficiency and to minimise transfusion requirements in adult patients who are candidates for elective major orthopaedic surgery [2B].

Intravenous iron therapy

The pre-operative use of intravenous iron to treat adult patients with iron-deficiency anaemia who are candidates for major surgery, including orthopaedic surgery, has been shown to be effective in correcting anaemia and limiting transfusion requirements^31–33.

A recent randomised, multicentre study demonstrated the superiority (and safety) of infusion of iron carboxymaltose with respect to iron therapy per os in the correction of iron-deficiency anaemia in patients with a poor response to oral therapy³⁴.

It is suggested that intravenous iron therapy is used for the treatment of pre-operative iron-deficiency anaemia and to limit transfusion requirements in adult patients who are candidates for elective major orthopaedic surgery [2A].

Post-operative period

Detection of anaemia

Anaemia develops in 74% of patients during hospitalisation (including those admitted with normal Hb values) and, besides causing a substantial use of resources, in part because of the prolongation of the time spent in hospital, is associated with significant increases in mortality³⁵ and morbidity¹¹. The prevalence of anaemia in the post-operative period in patients who have undergone elective or urgent major orthopaedic surgery is much higher, being 51% and 87%, respectively¹⁶. Furthermore, anaemia, which can be present in as many as 90% of surgical patients¹⁵, although caused mainly by peri-operative bleeding, can be worsened by repeated withdrawal of blood for laboratory tests and by ineffective erythropoiesis due to the inflammatory response induced by the surgical procedure itself^36,37. This inflammation reduces the availability of iron for erythropoiesis through inhibition of its intestinal absorption and a reduction in hepcidin-mediated mobilisation of iron stores^23,38–40.

The best strategy to prevent post-operative anaemia and avoid transfusions is to detect and correct any anaemia present in the pre-operative period, when possible¹¹.

However, most adults in good health with normal baseline levels of Hb do not normally require transfusion therapy with packed red cells following a surgical intervention provided that the loss of blood during the intervention is less than 1,000 mL and the intravascular volume is maintained with crystalloids or colloids^12,41.

Stimulation of erythropoiesis

Oral iron therapy

As already mentioned, the ineffective erythropoiesis caused by the post-operative inflammatory response^36,37, which inhibits intestinal absorption of iron and reduces the mobilisation of iron stores, makes oral iron therapy unfeasible in the post-operative period^23,38–40.

Indeed, various randomised studies have demonstrated that oral iron therapy in patients undergoing elective or urgent orthopaedic surgery is not only not effective at correcting the post-operative anaemia and reducing transfusion requirements, but is also often associated with the development of adverse events^42–44.

Oral iron therapy is not recommended for the treatment of post-operative anaemia and minimisation of transfusion needs in patients undergoing elective or urgent major orthopaedic surgery [1B].

Intravenous iron therapy

Intravenous administration of iron in the post-operative period, in patients who have undergone lower limb arthroplasty^45,46 or correction of scoliosis⁴⁷, is effective at correcting anaemia and limiting transfusion needs.

It is suggested that intravenous iron therapy is used for the treatment of post-operative anaemia and to limit transfusion needs in patients undergoing elective major orthopaedic surgery [2C].

Short-term management in the peri-operative period

Stimulation of erythropoiesis

Iron therapy

In orthopaedic patients with hip fractures, the short-term use of intravenous iron and adoption of restrictive transfusion strategies were demonstrated to be effective in limiting transfusion requirements, particularly in non-anaemic patients^48–50; other studies, in anaemic patients, showed that intravenous iron was more effective if combined with recombinant human erythropoietin^51,52.

Similar results were obtained in patients undergoing knee replacement surgery⁵³. Recently a large, observational study in patients who underwent elective lower limb arthroplasty or urgent surgery for hip fracture further confirmed the above described results: short-term infusions of iron were found to be effective with or without concomitant treatment with erythropoietin⁵⁴.

Short-term treatment with intravenous iron is suggested in order to minimise transfusion requirements in adult patients undergoing elective major orthopaedic surgery who are at risk of severe anaemia in the post-operative period [2B].

Dose of iron therapy

Oral iron

A recently published randomised clinical trial in critically ill patients with iron-deficiency anaemia demonstrated that 65 mg of elemental iron/die per os are effective at reducing transfusion requirements.

A dose of 100 mg of elemental iron/die for 2–6 weeks prior to the surgical intervention is recommended for the treatment of pre-operative anaemia [1C]⁵⁵.

Intravenous iron

It is suggested that the dose of intravenous iron needed to replenish iron stores is calculated using Ganzoni’s formula^19,56: “total iron requirements (mg) = [desired Hb – actual Hb (mg/dL)] × weight (kg) × 0.24 + 500 mg (for iron stores)” [2C].

The therapeutic regimen to adopt varies depending on the formulation of iron used.

It is suggested administering 200 mg of elemental iron intravenously for every 500 mL of blood lost [2C]²⁷.

Safety of intravenous iron therapy

Numerous studies performed in thousands of patients in different clinical settings have demonstrated the safety of intravenous iron therapy^57,58. The European Medicines Agency (EMA) recently published “New recommendations to manage risk of allergic reactions with intravenous iron-containing medicines”⁵⁹. A year later, in 2014, Rampton et al. published guidelines for the management of hypersensitivity reactions to intravenous iron⁶⁰.

According to the Committee for Medicinal Products for Human Use (CHMP) of the EMA: a) intravenous iron-based medicines should be used when oral iron cannot be used or does not work, particularly in patients in dialysis, in the peri-operative period, or in the presence of disorders of gastrointestinal absorption; b) the benefits of intravenous iron outweigh its risks, provided that appropriate measures are taken to minimise the possibility of allergic reactions; c) the data on the risk of hypersensitivity come mainly from spontaneous post-marketing notifications and the total number of deaths or life-threatening events is low; d) these data cannot be used to pick up any differences in the safety profile of the various iron-containing preparations.

The CHMP of the EMA has also produced the following recommendations for healthcare staff, with the aim of improving patients’ safety.

1. Intravenous iron must only be administered when “staff trained to evaluate and manage anaphylactic and anaphylactoid reactions” and “resuscitation facilities” are immediately available.

2. A test dose is no longer recommended.

3. If a hypersensitivity reaction occurs, “healthcare professionals should immediately stop the iron administration and consider appropriate treatment for the hypersensitivity reaction”.

4. “Patients should be closely observed for signs and symptoms of hypersensitivity reactions during and for at least 30 minutes following each injection of an intravenous iron medicine.”

5. Intravenous iron is contraindicated in patients with hypersensitivity to the active substance or excipients or to other iron-containing products administered parenterally.

6. The risk of hypersensitivity is increased “in patients with known allergies or immune or inflammatory conditions and in patients with a history of severe asthma, eczema or other atopic allergy.”

7. “Intravenous iron products should not be used during pregnancy unless clearly necessary” and their use “should be confined to the second or third trimester, provided the benefits of treatment clearly outweigh the potential serious risks to the foetus such as anoxia and foetal distress.”

As far as concerns the safety of intravenous iron, Auerbach and Macdougall state that, based on all retrospective and prospective studies, provided that high molecular weight iron dextran is avoided, which is no longer on the market, all the other preparations are safe and probably much safer than most doctors perceive^58,61.

In contrast, as far as concerns the risk of infections the data currently available in the literature do not allow definitive conclusions to be drawn and, for this reason, it is suggested that intravenous iron therapy is avoided in patients with acute infections [2C]^27,57.

It is suggested that intravenous iron is not administered to patients with ferritin values >300–500 ng/mL and with a transferrin saturation >50% [2C]²⁷.

Use of erythropoietin in the peri-operative period

There are two possible strategies for the use of erythropoietin in the peri-operative period¹²: the erythropoietin can be administered to optimise autologous donation, in the very few cases in which autologous donation is indicated, or it can be used in patients who are candidates for elective surgery who cannot complete a predeposit programme, in the very few cases in which such a programme is indicated. However, the prescription of erythropoietin α, β and z is currently paid for by the National Health Service if used as a treatment to increase the amount of autologous blood in the setting of a pre-operative deposit programme, with the limitations set out in the technical data sheet. Erythropoietin α can also be charged to the National Health Service when prescribed to reduce allogeneic transfusions in adults patients who are candidates for elective major orthopaedic surgery considered to involve a high risk of complications requiring transfusion, for which a pre-operative autologous blood donation programme is not available.

Erythropoietin has been found to be effective in limiting transfusion requirements in candidates for lower limb elective joint replacement surgery^62–64, although the costs are unacceptable⁶⁵.

It is suggested that erythropoietin is administered to adult candidates for elective major orthopaedic surgery who undergo a pre-deposit programme in which the donation of at least three units of whole blood is planned or for which a pre-deposit programme is not available and it is expected that the blood loss will be greater than 1,000 mL [2B].

In order to avoid a “functional deficiency” of iron during treatment with erythropoietin, it is suggested that intravenous iron is administered [2B]^12,66–69.

Minimisation of blood loss

Pre-operative period

Pre-deposit

Pre-deposit autologous blood transfusion consists in collecting units of blood from the patient (pre-deposit), storing them (without fractionation) and using them exclusively for the patient-donor. It has been widely shown that pre-depositing blood increases the risk of requiring transfusion therapy, including allogeneic transfusions⁷⁰.

In adult candidates for elective major orthopaedic surgery, it is recommended that the practice of pre-deposit is limited to subjects with rare red blood cell groups or complex alloimmunisation, for whom it is impossible to find compatible blood components [1A]^12,69.

In any case, there is no need for autologous blood collection if the patient’s basal Hb is such that, considering peri-operative losses, a stable, post-operative Hb of 100 g/L or more can be expected.

Contraindications to the collection of autologous blood are:

• - Hb values lower than the threshold values indicated by the WHO to define anaemia [children up 5 years: 110 g/L; children between 5 and 12 years old: 115 g/L; children between 12 and 15 years: 120 g/L; pregnant women: 110 g/L; women who are not pregnant (aged 15 years old or more): 120 g/L; men (aged 15 years or more): 130 g/L]²⁶;

• - severe heart disease;

• - positivity for one of the following tests, the results of which must be known before starting the autologous pre-deposit programme: HBsAg, anti-HCV antibodies, anti-HIV 1–2 antibodies;

• - epilepsy;

• - ongoing bacteraemia.

However, even in the presence of criteria for exclusion from autologous blood collection, a patient can, exceptionally, be accepted if the indications are appropriate and there are specific, documented clinical circumstances motivating the recourse to autologous donation.

It is recommended that the interval between collecting one unit of autologous blood and the next is at least 7 days and that, in all cases, the last unit is collected at least 7 days before the planned operation [1C]^71–76.

Identifying and managing the risk of bleeding

The second pillar of PBM includes all the strategies to minimise bleeding and preserve the individual blood reserves.

These strategies are applied right from the pre-operative stage, through the definition of parameters to stratify bleeding risk, that is, a detailed history and thorough clinical examination. Much care must be given to the patient’s drug history, since the number of patients being treated with antiplatelet agents, anticoagulants, anti-inflammatory drugs and antidepressants or herbal products with antiplatelet effects is increasing continuously.

Recent guidelines (British, Australian and Italian)^12,77–79 recommend the use of structured questionnaires to reduce bleeding and preserve individual blood reserves through the identification of patients at risk of bleeding and for a potential quantification of the bleeding risk in patients with congenital clotting disorders⁸⁰.

In fact, in the context of pre-operative screening, a standardised questionnaire covering the patient’s clinical and drug history seems to be superior to an assessment based only on the results of common laboratory tests [activated partial thromboplastin time (aPTT), prothrombin time /International Normalized Ratio (PT/INR), platelet count]⁸¹.

The second important element is the physical examination of the patient, which is aimed at detecting any signs of cutaneous bleeding (petechiae, ecchymoses, haematomas) which could suggest the presence of liver disease, a congenital clotting disorder or platelet disorder⁸².

Although various guidelines recommend the use of standard laboratory tests (aPTT, PT, platelet count) in the pre-operative period to define the bleeding risk^77,79,83, a systematic review demonstrated that altered coagulation screening tests in the pre-operative period are not predictive of intra- or post-operative bleeding⁸⁴. However, given their low cost and for reasons of prudence, the guidelines of the Italian Society for the Study of Haemostasis and Thrombosis (SISET) suggest that these tests are performed⁷⁹.

Low plasma levels of fibrinogen seem to be predictive of an increased risk of intra-operative bleeding in heart surgery⁸⁵.

Results of clotting tests performed in the pre-operative period with point-of-care (POC) instruments do not predict bleeding during or after surgery^86,87, whereas monitoring haemostasis by POC instruments during the intra-operative phase is useful for assessing the causes of bleeding.

Although there are reports on possible correlations between some haemostasis-related genetic polymorphisms or mutations and an increased risk of surgical bleeding, at present it is not possible to assign a clear predictive value to genetic tests⁸⁸.

Platelet function can be assessed through the use of several analysers (PFA 100/200, CPA Impact-R, MEA Multiplate, PlateletWorks, VerifyNow), which have a varied sensitivity for the different antiplatelet drugs. However, given the variability and poor standardisation of tests to determine platelet function, the SISET guidelines do not recommend their routine use prior to surgery⁷⁹. However, the guidelines of the European Society of Anaesthesiology suggest evaluating platelet function in the case of a positive history of bleeding and in the case of known alterations in platelet function because of congenital disorders or medication use⁸⁸.

Correct pre-operative management of the bleeding risk in surgical patients involves appropriate interventions to acquired factors (drugs, diseases) or congenital conditions predisposing to bleeding.

It is recommended that a careful personal and family history is taken from the patient in order to pick up any bleeding risk, information about ongoing medication use or the consumption of any over-the-counter or herbal products, because this assessment is considered more informative of peri-operative bleeding risk than isolated evaluation of results of pre-operative screening coagulation tests [1C].

It is suggested that the platelet count, PT and aPTT are determined before every surgical intervention or invasive procedure that carries a risk of bleeding [2C].

In the case of a positive history of bleeding, it is suggested that an expert in haemostasis is consulted in order to diagnose possible bleeding syndromes [2C].

Minimising iatrogenic blood loss and planning the procedure

Management of antiplatelet treatment

It is known that cyclo-oxygenase-2-selective non-steroidal anti-inflammatory drugs are not responsible for increased bleeding during total knee or hip replacement^89–91 and, for this reason, it is not necessary that the use of these drugs is suspended before elective lower limb arthroplasty. In contrast, ibuprofen, diclofenac and indomethacin significantly increase blood loss during total knee replacement surgery⁹², and so the use of these drugs should be suspended.

Monotherapy with aspirin (ASA) or clopidogrel does not need to be stopped before urgent orthopaedic surgery nor is it necessary to delay surgery in patients taking these drugs^93,94.

According to a recent Italian intersociety consensus statement⁹⁵ ASA, if taken for primary prevention, should be suspended 7 days before elective arthroplasties whereas it should be stopped on admission to hospital in the case of fracture of the femoral neck. If it is being taken for secondary prevention (i.e., in a patient who has had a previous cardiovascular event), it should be continued in the peri-operative period at a dose of 75–100 mg/die.

It is suggested that cyclo-oxygenase-2-selective non-steroidal anti-inflammatory drugs are not suspended prior to elective lower limb joint replacement surgery [2B].

It is suggested that ibuprofen, diclofenac and indomethacin are suspended prior to elective lower limb joint replacement surgery [2B].

It is recommended that ASA monotherapy, being taken for secondary prevention, is not suspended before elective lower limb joint replacement surgery [1B].

Ticlopidine and clopidogrel belong to the family of thienopyridines, being first- and second-generation examples, respectively; both inhibit ADP-induced platelet activation by binding to the P2Y12 receptor. Prasugrel is a third-generation thienopyridine, which must be converted to its active metabolite before binding to the P2Y12 platelet receptor. Thienopyridines have much stronger antiplatelet activity than ASA.

Various studies have described peri-operative bleeding complications in association with the use of clopidogrel and the risk of bleeding can increase when clopidogrel is given together with ASA. At present, there are no data available on the use of prasugrel in the peri-operative period; the platelet inhibition induced by this drug lasts at least 7 days⁸⁸.

Ticagrelor, another antiplatelet agent, unlike the thienopyridines, has a direct effect on the P2Y12 receptor, without requiring biotransformation by the P450 cytochrome; it has a rapid onset of action and the platelet inhibition decreases to 10% after about 4.5 days⁹⁶.

Since clopidogrel and prasugrel are responsible for peri-operative bleeding, it is recommended that these drugs are suspended 5 and 7 days, respectively, before surgery, in cases of increased bleeding risk [1C].

It is suggested that ticagrelor is suspended 5 days pre-operatively [2C].

Patients with acute coronary syndrome or those undergoing angioplasty benefit from dual antiplatelet therapy with a combination of ASA and another platelet anti-aggregant drug (thienopyridine or ticagrelor), even if this increases the risk of bleeding complications⁹⁷.

In the case of urgent or emergency surgery, it is recommended that the decision regarding the continuation of treatment with antiplatelet drugs in the peri-operative period is the result of a multidisciplinary evaluation [1C].

It is suggested that urgent or emergency surgical interventions are performed, maintaining dual antiplatelet therapy (ASA/clopidogrel; ASA/prasugrel; ASA/ticagrelor) or at least ASA, when the bleeding risk is high [2C].

Oral dual antiplatelet therapy is necessary for patients with stents, in whom the suspension of one or both of the anti-aggregant drugs, particularly in the first months after the procedure, carries a significant risk of thrombosis of the stent, which is a potentially fatal event⁹⁸. The growing number of coronary artery revascularisation procedures carried out each year is inevitably leading to an increase in the number of patients with coronary artery stents who must undergo surgical interventions. The management of antiplatelet therapy in these patients is often arbitrary, despite it being a critical factor in the prevention of ischaemic-haemorrhagic complications.

It is recommended that elective orthopaedic surgery is not performed during the first 3 months after implantation of a metal stent or during the first 12 months after implantation of a drug-eluting stent [1C].

Management of anticoagulant therapy

Vitamin K antagonists (VKA) are used for the prophylaxis or treatment of thromboembolic events, particularly in patients with mechanical heart valves, atrial fibrillation or a past history of venous thromboembolism. It is not always necessary to suspend the use of these drugs and the decision depends on the type and site of the operation or invasive procedure^88,99,100.

In patients with a low/medium risk of thromboembolism, it is suggested that VKA treatment is suspended 5 days before the planned elective joint replacement surgery and bridging therapy is established [administering low molecular weight heparin (LMWH) at prophylactic doses] in the following way: last dose of VKA on day −5; first subcutaneous dose of LMWH once a day, starting on day −4 if the patient was being treated with acenocoumarol, and starting on day – 3 if the patient was being treated with warfarin [2C].

In patients at high risk of thromboembolism [i.e., with atrial fibrillation and a CHADS₂ (Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, prior Stroke or transient ischaemic attack or thromboembolism) score >2; with recurrent venous thromboembolism treated for less than 3 months; with a mechanical heart valve] it is recommended that bridging therapy (LMWH at therapeutic doses) is given in the following manner: last dose of the VKA on day −5; first subcutaneous dose of LMWH twice daily starting from day −4 if the patient was being treated with acenocoumarol, and starting from day −3 if the patient was being treated with warfarin [1C].

An Italian study demonstrated the efficacy and safety of using reduced therapeutic doses (65–70 IU/kg twice daily) instead of the classical bridging therapy with 100 IU/kg twice daily¹⁰⁰.

It is suggested that the last dose of LMWH is administered 12 hours before a planned operation and/or invasive procedure, except when a full dose of anticoagulant is being used, in which case an interval of 24 hours is suggested [2C].

The new oral anticoagulants

Dabigatran etexilate is an oral inhibitor of thrombin. It has a half-life of 14–17 hours, is eliminated mostly (80%) in the urine, does not interact with food and does not necessitate blood tests for monitoring. It is indicated for the prevention of venous thromboembolism in major orthopaedic surgery, for the prevention of emboli in patients with atrial fibrillation and in the treatment and secondary prevention of deep vein thrombosis and pulmonary embolism^101–105.

Rivaroxaban and apixaban are oral inhibitors of activated factor X (FXa). They have half-lives of 9–12 hours and 10–15 hours, respectively; rivaroxaban is eliminated mainly through the kidneys (although only half in an active form), while apixaban is eliminated partly through the hepatic route and partly in the urine. Both are indicated for the prevention of venous thromboembolism in major orthopaedic surgery, for the prevention of emboli in patients with atrial fibrillation and for secondary prevention of deep vein thrombosis and pulmonary embolism^106–113.

The management of new oral anticoagulant (NOAC) use during surgical interventions is based largely on the consensus of opinions of experts, although the experts have produced contrasting advice, particularly with regards to the indication for and duration of bridging therapy with parenteral anticoagulants, which is proposed by some experts also for patients being treated with NOAC. Indeed, this is the strategy suggested by the recent guidelines from the European Society of Anaesthesiology⁸⁸. Some scientific societies have suggested indications for suspending or not suspending treatment with NOAC¹¹⁴. Based on the pharmacokinetic and pharmacodynamics characteristics of NOAC, a temporary (short-term) suspension of these drugs is possible without requiring bridging therapy, which would expose the patient to a higher risk of bleeding, as demonstrated by recent registry data¹¹⁵. For this reason the following recommendations and suggestions reflect the literature data and recent guidelines of the European Heart Rhythm Association (EHRA), which do not include the use of bridging therapy^{114,116–118}.

Table II shows the stratification of bleeding risk in relationship to invasive procedures or surgical interventions: the bleeding risk is defined as “clinically unimportant”, “low” or “high”¹¹⁴.

Table II.

Classification of elective surgical interventions divided according to bleeding risk. (Modified from Heidbuchel H et al.¹¹⁴)

Interventions with a clinically unimportant risk of bleeding: - dental interventions: - extraction of 1 to 3 teeth - paradontal surgery - abscess incision - implant positioning - ocular interventions: - cataract or glaucoma - endoscopy without surgery: - superficial surgery - abscess incision - small dermatological excisions

Interventions with a low risk of bleeding: - endoscopy with biopsy - prostate or bladder biopsy - ablation for supraventricular tachycardia (including left-sided ablation via single trans-septal puncture)

Interventions with a high risk of bleeding: - complex ablation - spinal or epidural anaesthesia; lumbar puncture - thoracic surgery - abdominal surgery - major orthopaedic surgery - liver biopsy - transurethral resection of prostate - renal biopsy

It is suggested that NOAC (dabigatran, rivaroxaban, apixaban) are not suspended and that the operation is performed 12–24 hours (depending on whether the drug is administered, respectively, once or twice a day) after the last dose in the case of: dermatological surgery, dental procedures, gastroscopy and colonoscopy (without biopsies), ocular interventions (particularly of the anterior chamber, such as cataract surgery) and operations involving a clinically unimportant risk of bleeding (see Table II) [2C]¹¹⁴.

It is suggested that NOAC are suspended 24 hours before elective surgery that carries a low risk of bleeding, in patients with normal renal function [creatinine clearance (CrCl) ≥80 mL/minute] [2C].

It is suggested that NOAC are suspended 48 hours before elective surgery that carries a high risk of bleeding in patients with normal renal function (CrCl ≥80 mL/minute) [2C].

In patients with impaired renal function, the suspension of the NOAC should be graduated according to the type of drug and the CrCl, as indicated in Table III¹¹⁴.

Table III.

When to suspend treatment with new oral anticoagulants before surgery on the basis of the patient’s renal function (determined by creatinine clearance [CrCl]) and bleeding risk (low or high, see Table II) associated with the surgical procedure. (Modified from Heidbuchel H et al.¹¹⁴)

CrCl (mL/minute)	Bleeding risk associated with the surgical procedure
	New oral anticoagulants
	Dabigatran		Apixaban		Rivaroxaban
	Low	High	Low	High	Low	High
≥80	≥24 h	≥48 h	≥24 h	≥48 h	≥24 h	≥48 h
50–80	≥36 h	≥72 h	≥24 h	≥48 h	≥24 h	≥48 h
30–50	≥48 h	≥96 h	≥24 h	≥48 h	≥24 h	≥48 h
15–30	NI	NI	≥36 h	≥48 h	≥36 h	≥48 h
<15			NI

CrCl: creatinine clearance; h: hours; NI: use of drug not indicated.

It is suggested that rivaroxaban and apixaban are suspended 36 or 48 hours before surgery with a low or high bleeding risk, respectively, in patients with a CrCl between 15–30 mL/minute; that dabigatran is suspended 36 or 72 hours before surgery with a low or high bleeding risk, respectively, in patients with a CrCl between 50–80 mL/minute; and that dabigatran is suspended 48 or 96 hours before surgery with a low or high risk of bleeding respectively, in patients with a CrCl between 30–50 mL/minute [2C].

New oral anticoagulants and laboratory tests

The NOAC do not require routine monitoring through coagulation tests; in any case, global tests, such as the PT and aPTT, are not useful for quantifying the anticoagulant effect of these new drugs and other quantitative tests are not yet available for routine use in all hospitals. Furthermore, POC instruments to determine the INR should not be used in patients being treated with NOAC. However, in emergencies (severe haemorrhage, thrombotic events, urgent surgery, liver or kidney failure, suspected overdose or pharmacological interactions) it may be necessary to quantify and have a rough idea of the anticoagulant effect of NOAC through the available coagulation tests; in these cases it is extremely important to know precisely when the drug was taken with respect to the blood sample, since the maximum effect of NOAC occurs at the time of their peak plasma concentration, which is reached 1–3 hours after drug intake¹¹⁴. Thus, for a correct clinical interpretation of the laboratory data, it is essential to know the type of drug taken, the dose and the time the last dose was taken.

At present, there is still no definitive information on whether assays prior to taking the next dose (trough concentration) or at the time of the peak plasma concentration are better indicated. Analyses carried out in the USA by the Food and Drug Administration (FDA) on the phase III studies “Re-LY” and “ROCKET AF” showed that the trough drug level was correlated with clinical events. On the basis of this evidence, it has been proposed that controls should be performed before the patient takes the next dose of the drug (trough concentration).

Both first and second level clotting tests are variably affected by the different anticoagulant drugs. In general, it can be stated that the PT and chromogenic tests for assaying anti-FXa activity are more influenced by anti-FXa drugs (rivaroxaban, apixaban and edoxaban), whereas the aPTT, thrombin time, ecarin clotting time and ecarin chromogenic assay are more affected by drugs that inhibit thrombin (dabigatran).

Two coagulation tests for evaluating the anticoagulant effect of the direct inhibitors of thrombin, such as dabigatran, have so far been identified in the literature: the aPTT and the diluted thrombin time. Alterations of these tests can be indicative of an increased bleeding risk^119,120. While the aPTT only provides a qualitative assessment and there is great variability between the different reagents on the market, the diluted thrombin time is able to provide a quantitative evaluation of the level of dabigatran present in the circulation and, considering its relative simplicity, should be used.

The inhibitors of FXa, such as rivaroxaban, cause a prolongation of the PT although this provides only approximate information, given the wide variability of results depending on the type of reagent used¹²¹.

An anti-FXa chromogenic test, using calibrators that are readily available on the market, has recently been developed. It is, however, important to clarify that, at present, these new tests (diluted thrombin time, for dabigatran, and anti-FXa activity assay, for the inhibitors of FXa) can provide plain information on the absence of the anticoagulant in the circulation, while there is not a clear, defined relationship between the circulating level and risk of bleeding or thrombosis. This aspect must be studied in the coming years.

At present we know that the concentrations of the different oral anticoagulants can be measured using tests that are specific and sensitive for the individual compounds: these tests are simple, cheap and can be performed in all laboratories. We also know that information regarding the concentration of the anticoagulant drugs in the blood is useful for managing patients who are being assessed for particular clinical situations, such as urgent or elective surgery, invasive procedures, and haemorrhagic or thromboembolic complications. Furthermore, in the same way as for VKA, the results of these tests could be useful in the near future for determining threshold levels of anticoagulation above which surgery, invasive procedures and thrombolysis are contraindicated. Currently, one of the most important practical problems is the limited number of Italian laboratories able to carry out these specific tests; however, the growing use of NOAC and the consequent increase in the clinical needs of treated patients should lead to a progressively greater use of these tests in various different hospital settings.

No evidence-based recommendations can be made on the use of laboratory tests in the pre-operative evaluation of the anticoagulant effect of the NOAC.

The management of patients with comorbidities related to altered haemostasis

Patients with endocrine, metabolic or systemic diseases, such as FX deficiency in amyloidosis, can have altered haemostasis with a tendency to bleeding, similar to that in patients with congenital deficiencies of clotting factors^122,123. The treatment strategy for these coagulopathies is frequently unclear.

It is suggested that the treatment of patients with altered haemostasis related to systemic, metabolic or endocrine diseases is established through consultation with an expert in haemostasis and thrombosis [2C].

Other drugs, besides antiplatelet agents and anticoagulants, such as selective serotonin reuptake inhibitors, valproic acid and Ginkgo biloba, can interfere with haemostasis, predisposing to bleeding.

Selective serotonin reuptake inhibitors have been associated with an increased tendency to bleed, due to depletion of serotonin from platelets¹²⁴; however, the transfusion requirements of patients being treated with selective serotonin reuptake inhibitors who underwent major orthopaedic surgery or heart surgery were not increased^125,126.

In some cases valproic acid (an anti-epileptic drug) can lead to a reduction in the levels of some clotting factors [FVII, FVIII, FXIII, von Willebrand factor (vWF), fibrinogen], platelets, protein C and antithrombin¹²⁷; however, these changes did not cause bleeding complications^128,129.

As far as concerns the extract of Ginkgo biloba, a meta-analysis of 18 randomised, controlled trials did not show an increase in bleeding associated with daily, oral intake of this medicinal plant¹³⁰.

It is suggested that treatment with selective serotonin release inhibitors is not routinely suspended in patients undergoing surgery [2B].

When selective serotonin reuptake inhibitors are used together with antiplatelet agents, it is suggested that the patient is assessed individually to define the strategy to adopt prior to surgery [2C].

In the case of surgery, it is suggested considering, on an individual basis and with specialist advice, the suspension of treatment with valproic acid, because this drug can promote bleeding [2C].

It is recommended that the use of Ginko biloba extract is not suspended in the case of surgical interventions [1B].

Management of the patient with a congenital bleeding disorder

Defects of primary haemostasis

The most common congenital bleeding disorder is certainly that of hereditary deficiency of vWF, which has prevalence of 0.6–1.3% in the population^13,131. The disease is due to either a lack of vWF or dysfunction of the protein and is classified into three types: type 1: partial quantitative defect; type 2: qualitative defect, of which there are four variants: 2A, 2B, 2M, and 2N; and type 3: total lack of vWF. The acquired form of von Willebrand’s disease (vWD) is the result of autoimmune and/or neoplastic disorders, or has a drug-based aetiology¹³².

The bleeding tendency is a consequence of altered platelet adhesion because of a lack or dysfunction of vWF and/or reduced levels of FVIII^13,131,132.

Given that the laboratory diagnosis of vWD is complex, the laboratory tests must always be guided by the patient’s clinical history and physical examination. Additionally, there are questionnaires and bleeding scores that have been specifically designed to evaluate and quantify bleeding risk^133–135.

The prevention and management of bleeding in vWD is based on three treatment strategies: administration of desmopressin, which induces mobilisation of vWF from endothelial storage sites (stimulation therapy); administration of plasma-derived concentrates of vWF or vWF-rich FVIII (replacement therapy); administration of antifibrinolytics or platelet transfusions (haemostatic therapy). There are numerous national and international guidelines on this subject^{13,131,136–140}.

It is recommended that patients with vWD are managed in the pre-operative period in collaboration with an expert in haemostasis and thrombosis [1C].

It is recommended that criteria are used to define and evaluate bleeding risk in patients with vWD [1C].

It is recommended that patients with vWD who are to undergo minor surgery and are expected to have mild bleeding are given desmopressin, after a trial test, at doses explained in specific guidelinese (0.3 μg/kg diluted in 50 mL of saline solution and infused slowly, over more than 30 minutes) [1C]^136–140.

It is recommended that patients with vWD who are to undergo major surgery and are expected to have clinically relevant bleeding are given replacement therapy with plasma-derived vWF or vWF-rich FVIII, using the treatment regimens explained in the specific guidelines [1C].

In patients with vWD who are to undergo major surgery, it is suggested that antifibrinolytics, as an adjuvant to more specific treatments, and platelet transfusions are used only in the case of failure of other treatments [2C].

In the context of disorders of primary haemostasis, platelet dysfunction is a diagnostic challenge. There does not seem to be a correlation between the severity of bleeding, the degree of vWF/platelet dysfunction and diagnostic laboratory tests, so that disorders of platelet function constitute a risk factor for bleeding rather than an unequivocal cause.

The use of platelet function tests with the PFA 100/200 cannot be recommended in patients with defects of primary haemostasis because of the poor sensitivity of the tests as well as false positive and false negative results^141–143.

The most common and least severe platelet disorders respond well to desmopressin, which shortens the bleeding time, whether used as prophylaxis or for the treatment of bleeding^144–146. When used, the standard dose is 0.3 μg/kg, diluted in 50 mL of saline solution and infused slowly, over more than 30 minutes¹⁴⁴.

Antifibrinolytics are useful as an adjunctive therapy in platelet disorders; small bleeds, such as those occurring in dental surgery, can respond well to these agents, even when used alone. The use of antifibrinolytic agents for the treatment of hereditary platelet disorders is not, however, based on evidence, although tranexamic acid (TXA) can partially correct the effects of clopidogrel on primary haemostasis^{142,144,147,148}.

In patients with hereditary disorders of platelet function, it is suggested that desmopressin is used for the prevention and control of bleeding, and that TXA is used as an adjuvant [2C].

The best known congenital platelet disorders are Glanzmann’s thrombasthenia and Bernard-Soulier syndrome. These are characterised, respectively, by defective platelet aggregation and adhesion, due to alterations in platelet membrane glycoproteins (GP IIb/IIIa in Glanzmann’s thrombasthenia and GP Ib-IX in Bernard-Soulier syndrome).

In most patients with these disorders, mild mucocutaneous bleeding responds to treatment with antifibrinolytic agents. However, the use of recombinant activated factor VII (rFVIIa) is indicated for surgery and invasive procedures (including dental extractions, sometimes), even if the technical data sheet states that it is indicated only for those patients with Glanzmann’s thrombasthenia who have antibodies to GPIIb/IIIa and/or human leucocyte antigens or have become refractory to platelet transfusions^149,150.

When used, rFVIIa should be administered at a dose of 90 μg/kg, immediately before the intervention, repeated every 2 hours for the first 12 hours and, subsequently, every 3–4 hours until the bleeding risk has disappeared¹⁵¹.

However, there are no universally accepted doses of rFVIIa defined for the various different settings of use. rFVIIa is not indicated in other forms of platelet disorders, for which platelet transfusions are necessary.

It is recommended considering the use of rFVIIa in patients with Glanzmann’s thrombasthenia who are to undergo surgical interventions [1C].

In patients with Glanzmann’s thrombasthenia or Bernard-Soulier syndrome undergoing elective, major surgery, it is suggested that platelet concentrates are transfused when other therapeutic options, including rFVIIa, do not guarantee a therapeutic effect [2C]^142,145.

It is suggested that the first doses of platelets are administered immediately before the intervention and further doses given after it, depending on clinical need [2C].

In cases of urgent surgery, single-unit platelet concentrates can be given, despite the awareness of the high risk of alloimmunisation and the consequent, subsequent limitation of response to this treatment¹⁴⁵.

As far as concerns treatment, the hereditary platelet disorders are considered moderate platelet disorders and, in the absence of platelet dysfunction, should be treated on the basis of the platelet count. Guidelines on platelet transfusions suggest a threshold of 50×10⁹ platelets/L for major surgery or invasive procedures^152–155 [liver biopsy, laparotomy, diagnostic lumbar puncture, insertion of central venous catheters (recent North-American guidelines suggest a threshold of 20×10⁹ platelets/L for this last procedure)]¹⁵⁵ and a threshold of 100×10⁹ platelets/L for neurosurgical or ophthalmological interventions^152–154. However, there is still insufficient evidence to recommend a threshold for prophylactic transfusions in the peri-operative period in patients with hereditary platelet disorders.

It is recommended that platelet transfusion are not used routinely in patients with hereditary platelet disorders [1C].

In patients with hereditary platelet disorders no evidence-based recommendations can be made concerning the threshold to adopt in the peri-operative period for prophylactic transfusion therapy with platelet concentrates.

Defects of haemostasis related to clotting factor deficiencies

Congenital deficiency of FVIII and FIX in the plasma causes haemophilia A and haemophilia B, respectively.

The prevalence of haemophilia A in the population is 1:10,000, while that of haemophilia B is 1:60,000. The clinical manifestations of haemophilia are spontaneous bleeding, mainly within joints, and excessive bleeding in the case of trauma and/or surgical interventions. The severity of the bleeding is related to the extent of the factor deficiency. Both forms of haemophilia are classified as mild, moderate or severe, depending on the level of FVIII or FIX present.

In the severe forms of haemophilia, replacement therapy can lead to the development of antibodies against FVIII or FIX, known as “inhibitors”, a critical circumstance that requires various therapeutic strategies.

Anti-FVIII autoantibodies cause the bleeding disorder known as acquired haemophilia, which is rare condition characterised by a predisposition to potentially dangerous bleeding. This disorder is usually associated with cancer, autoimmune disorders, drugs or pregnancy.

The treatment of haemophilia is essentially replacement therapy with plasma-derived or recombinant concentrates of the deficient clotting factor (FVIII or FIX). Mild haemophilia A can also be treated with desmopressin and TXA, instead of replacement of the deficient factor with concentrates.

Despite the wide variability in the doses of concentrates used for the prophylaxis or treatment of bleeding, in the case of surgery, the World Federation of Haemophilia recommends that the levels of the deficient factor are 80–100% in the pre-operative period¹⁵⁶; it also recommends that the levels are maintained around 60–80% in the first 3 days after the operation, around 40–60% for the next 3 days and around 30–50% in the second week after surgery.

In the case of haemophilia B, the recommended factor levels are slightly lower: 60–80%, 40–60%, 30–50% and 20–40%, respectively^{140,156–162}.

In conclusion, the recommendations for the management of patients with haemophilia or other congenital bleeding disorders are the following.

Collaboration with an expert in haemostasis and thrombosis is recommended during the planning of the surgical intervention [1C].

It is recommended that adequate replacement therapy is given during the peri-operative period [1C].

It is suggested that guidelines on how to perform replacement therapy (target level of deficient factor and duration of treatment) are followed for patients who are candidates for elective surgery [2C].

With regards to replacement therapy in the peri-operative period, it is recommended that either recombinant or plasma-derived concentrate is used [1C].

In the presence of inhibitor, treatment with rFVIIa or activated prothrombin complex concentrate (PCC), i.e. a factor eight inhibitor bypassing agent (FEIBA), is suggested [2C].

A recent, prospective study carried out on 24 haemophilic patients undergoing orthopaedic surgery showed the presence of subclinical deep vein thromboses in over 10% of cases. On the basis of these results, antithrombotic prophylaxis could also be indicated in haemophilic patients, in individual cases. Many Haemophilia Centres in Europe use pharmacological antithrombotic prophylaxis after orthopaedic surgery¹⁶³.

For haemophiliacs undergoing major surgery, it is suggested that peri-operative thromboprophylaxis is individualised [2C].

Congenital deficiencies of clotting factors other than FVIII and FIX are very rare and have a prevalence of between 1:500,000 and 1:2,000,000¹⁶⁴; the prevalence of autosomal dominant FXI deficiency is 1:30,000, but the most common of these deficiencies is that of FVII.

The levels of evidence on treatment of these deficiencies are low (descriptive studies and experts’ opinions) and data on pre-operative prophylactic therapy are scarce.

In cases of major surgery in patients with FVII deficiency, the proposed threshold for replacement therapy with FVII concentrate is 10% of the normal plasma level^164–166. Above this level replacement therapy does not seem to be necessary, as demonstrated by a retrospective analysis of surgical procedures conducted without giving such therapy and during which the frequency of bleeding events was 15%¹⁶⁷.

rFVIIa is the treatment of choice for congenital FVII deficiency; if this is not available, a plasma-derived concentrate of FVII is preferable to PCC, because of the potential prothrombotic effect of the latter^140,164.

In congenital FVII deficiency, as an alternative to rFVIIa, it is suggested that a plasma-derived concentrate of FVII is used at the dose of 10–40 IU/kg [2C]¹⁶⁴.

In patient with congenital FVII deficiency who are to undergo major surgery, it is suggested using a dose of rFVIIa of 15–30 μg/kg every 4–6 hours, usually giving at least three doses [2B]¹⁶⁴.

In cases of congenital deficiency of fibrinogen or FXIII the British guidelines recommend using specific concentrates¹⁶⁴. Replacement therapy is indicated in hypofibrinogenaemia when the concentration of fibrinogen is <1 g/L or <20–30% of normal. However, the possible risk of thrombosis must be kept in mind when using plasma-derived concentrates.

Fresh-frozen plasma, preferably industry-produced and virus-inactivated, is the only therapeutic option in patients with congenital FV or FXI deficiency¹⁶⁴. There is a plasma-derived concentrate of FXI on the market, but it is not currently available in Italy. Patients with FXI deficiency and inhibitors have been treated with success, on the occasion of surgery, with low doses of rFVIIa (33–47 μg/kg)¹⁶⁸; contemporaneous administration of TXA has also been demonstrated to be effective in controlling bleeding¹⁶⁴.

PCC are the reference concentrates for deficiencies of FII and FX, for which specific concentrates are not available^164,169,170.

In patients with other clotting factor deficiencies, no evidence-based recommendations can be made on the peri-operative use of rFVIIa, desmopressin or TXA.

Management of patients with acquired platelet disorders

To reduce the risk of bleeding associated with major surgery or invasive procedures in patients with acquired platelet disorders, it is suggested that a prophylactic transfusion of platelet concentrates is given if the platelet count is below the threshold of 50×10⁹ platelets/L [2C]^152–155.

It is suggested that the first doses of platelets are given immediately before the operation and further doses after it, depending on the clinical need [2C].

Intra-operative period

Autologous transfusion techniques

Acute normovolaemic haemodilution

Acute normovolaemic haemodilution is a type of autologous transfusion introduced in the 1970s^171–173. It consists in removing at least three or four units of autologous blood, maintaining isovolaemia, immediately before elective surgery¹⁷⁴. Acute normovolaemic haemodilution is usually performed after induction of anaesthesia¹⁷⁵, just before the surgical incision¹⁷⁶. The circulating blood volume is maintained by infusing crystalloids, administering 2–3 mL for every mL of blood removed, or colloids, in a ratio of 1:1 with the volume of blood removed¹⁷⁴.

The efficacy of acute normovolaemic haemodilution in reducing the need for allogeneic red blood cell transfusion is, however, doubtful¹⁷⁷. Various studies, including prospective, randomised trials, showed that it could reduce recourse to allogeneic transfusion therapy in patients undergoing elective heart surgery, orthopaedic operations (knee replacement), abdominal, vascular, urological, maxillofacial, or liver surgery and in patients undergoing surgery because of burns^178–189. However, other studies did not find any substantial benefit or even found an increased need for allogeneic transfusions^190–192.

The lack of real benefits from this procedure and the increased risk of requiring transfusion therapy, including allogeneic transfusions, were demonstrated by some meta-analyses^193–196 and confirmed by a recent study evaluating the efficacy of autologous transfusion techniques, based on mathematical models⁷⁶. Other pharmacological strategies for conserving blood, such as the use of TXA, have been found to be more effective than acute normovolaemic haemodilution in limiting the use of allogeneic red blood cell transfusions^195,197. Finally, the reduction of overall bleeding is minimal, there is an increase in intra-operative bleeding (heart, liver and thoracic surgery), the relative risk of re-operation because of bleeding is increased and there is a lack of information on the safety of the procedure^195,198,199.

The routine use of acute normovolaemic haemodilution as a (single) technique to reduce allogeneic transfusion requirements is not recommended [1A].

Intra-operative blood recovery

Intra-operative blood recovery is a method of saving blood in which blood lost in the surgical field is re-used. This blood is aspirated and anticoagulated before passing into a collection reservoir and from there, through microaggregate filters of varied diameter, into the bowl of a specific cell separator, to be concentrated by centrifugation and then washed with physiological saline before being re-infused back into the patient²⁰⁰.

Intra-operative blood recovery is indicated in many types of elective and urgent surgery, when the blood loss is expected to be at least 1,000 mL or ≥20% of the patient’s circulatory volume^200,201.

The meta-analyses currently available show that intra-operative blood recovery significantly reduces the need for allogeneic red blood cell transfusions in patients undergoing elective surgery^202–204. A study published in 2013 evaluating the efficacy of autologous blood transfusion techniques, based on mathematical models, indicated that this is the most effective and efficient of such tecniques⁷⁶. However, a recent randomised, controlled trial, carried out in the specific setting of elective lower limb joint replacement surgery in patients with pre-operative Hb concentrations above 130 g/L, found that intra-operative blood recovery, like post-operative blood recovery, not only did not reduce transfusion needs, but increased costs²⁰⁵.

It is recommended that intra-operative blood recovery is used in major orthopaedic surgery, including vertebral column surgery^206–212, only in cases in which blood loss is expected to be at least 1,000 mL or in any case ≥20% of the patient’s circulatory volume despite adopting multimodal strategies, including the use of blood-conserving techniques (pharmacological, surgical and anaesthesiological), and taking into account both the characteristics of the individual patient (bleeding risk, multiple alloimmunisation) and the experience of the surgical and anaesthetic teams [1B]¹⁹⁹.

Blood-saving surgical techniques

The greater the expected blood loss, the stronger the advice to use multiple mehtods, appropriate for the clinical circumstances, to reduce bleeding. The use of appropriate combinations of methods has a synergistic effect on reducing blood losses. Meticulous haemostasis and surgical technique are essential. The haemostasis must be achieved through a combination of methods, starting from the surgical approach, which must have the least traumatic impact possible, with well-planned surgical exposure, arriving through the least vascularised tissue planes and with atraumatic management of the tissues²¹³. Furthermore, immediately before the surgical incision, a dose of local anaesthetic and adrenaline can be given at the incision site in order to obtain rapid, localised vasoconstriction. The duration of the intervention must be minimised and, in cases of complex procedures in which repeated surgery is necessary, particularly in patients with multiple trauma, surgery should be fractionated and conducted in stages²¹⁴.

In the intra-operative phase the surgeon has various instruments available to achieve perfect haemostasis, such as monopolar or bipolar electrocautery devices, ultrasound cautery probes and argon coagulators; these last two are not commonly used in orthopaedic surgery and are mainly reserved for very selected cases, in particular patients with cancer²¹⁵.

As far as concerns knee arthroplasty, although special surgical techniques and instruments, which avoid violation of both the tibial and femoral intramedullary canals and, therefore, minimise blood losses, were introduced only recently, their use has become consolidated²¹⁶. However, given the notable pre-operative organisation and greater consumption of resources associated with the sue of these “custom-fit” methods, they are reserved for selected cases; their routine use is also avoided because, at present, there are no studies demonstrating their long-term efficacy in terms of functional outcomes and, furthermore, a recent meta-analysis did not show that they produce an overall improvement in surgical efficiency or that they are cost-effective²¹⁷.

A survey by the American Association of Hip and Knee Surgeons found that a tourniquet is often used during total knee replacement: the tourniquet must be applied to the root of the limb to be operated only in certain stages of the surgical procedure and after complete “squeezing” of the limb²¹⁸. However, some studies have associated tourniquet use with complications such as cutaneous or muscular damage, rhabdomyolysis, neurological damage, post-operative stiffness, deep vein thrombosis and pulmonary embolism^219,220. Furthermore, a recent meta-analysis provided some evidence that the use of a tourniquet can increase post-operative complications and reduce the range of movements in the early post-operative period, that it does not reduce total blood loss significantly or the need for transfusion therapy, but only decreases intra-operative bleeding²²¹. In contrast, another meta-analysis showed that application of a tourniquet could reduce total blood losses, lower the incidence of transfusions and shorten the duration of the operation, although the same analysis indicated some possible disadvantages and complications associated with tourniquet use²²².

With the purpose of containing intra-operative bleeding effectively during elective arthroplasties, it is suggested that combinations of surgical techniques and instruments (considered appropriate for their synergistic effect on reducing blood losses also on the basis of the surgical team’s experience) are used in order to minimise trauma to tissues and vessels and promote local haemostasis, which can also be aided by local administration of vasoconstrictive drugs [2C].

Given the possible post-operative complications and the controversial effect on reducing transfusion needs, no evidence-based recommendations can be made concerning the routine use of a tourniquet as a blood-conserving technique in elective prosthetic joint surgery.

Infusion fluids

The correction of bleeding-induced hypovolaemia by infusion of crystalloids and/or colloids is a priority in the management of patients with acute or subacute bleeding and is the first alternative to allogeneic transfusion, because acute hypovolaemia is less well tolerated than anaemia²⁷. Adequate restoration of the circulatory volume and, consequently, the cardiac output, enables oxygen transport to the tissues to be maintained. However, the best fluid replacement strategy in patients with critical bleeding is still the subject of debate²²³.

Crystalloid and non-protein-containing colloid solutions are the treatment of first choice^199,224. Isotonic 0.9% saline solution, Ringer’s solution and other balanced saline solutions, such as Ringer’s lactate, are those most commonly used at the moment. They are cheap and do not alter haemostasis or renal function²⁷. However, the choice of fluids to administer intravenously during surgical procedures and in critically ill patients is dictated more by clinical practice than by solid evidence²²⁵. The debate on the choice of fluid has concentrated mainly on evaluating differences in outcomes depending on whether crystalloid or colloid solutions are used. Recently, however, there has been interest in differences in outcomes in relation to the chlorine content in crystalloid solutions. In fact, new notions on the conventional Starling’s model, regarding microvascular fluid exchange, indicate that the efficacy of colloids in restoring and maintaining depleted intravascular volume is only moderately greater than that of crystalloids. Various very recent randomised controlled trials have demonstrated that the modest improvements in certain physiological end-points, obtained in the short-term with colloids, do not translate into a better outcome for the individual patient. Furthermore, there is substantial evidence that some types of fluids (solutions of hydroxyethyl starch and of albumin) can even have a negative effect on some outcome markers when administered to selected populations of patients.

The most commonly used colloids are the protein-free ones, such as solutions of hydroxyethyl starch or gelatine, and solutions of human albumin (natural colloid)²⁷. Given their low molecular weight, gelatine solutions have a short intravascular half-life (2–3 hours) and a limited capacity to expand plasma (70–80%). Solutions of 6% hydroxyethyl starch have a longer half-life (6–8 hours) and a greater capacity to expand the plasms (80–120%) and are, therefore, the colloids most widely used at present for expanding blood volume; however, when used as volume expanders to correct acute hypovolaemia in patients with haemorrhagic shock, they can increase the risk of mortality and renal failure²²⁶. Furthermore, massive use of high molecular weight hydroxyethyl starch solutions can cause alterations of haemostasis characterised by impaired platelet function^88,199. Finally, following a review, performed for safety reasons, of the indications for the use of hydroxyethyl starch solutions, restrictions have been applied to the duration of treatment, which must be limited to the initial phase of volume restoration, for a maximum of 24 hours. Furthermore, the following monitoring is expected: a) continuous haemodynamic monitoring; b) electrolytes and fluid balance; c) renal function, for at least 90 days; d) clotting parameters, in the case of repeated administrations²²⁷.

Albumin 5% has the capacity to expand the plasma volume by 75% of the volume infused²⁷; its clinical effects, in terms of morbidity and mortality at 28 days, are the same as those of isotonic saline 0.9% when used to restore circulatory volume in patients in intensive care²²⁸. For this reason, solutions of albumin 5% should be used as second-line treatment when crystalloid or non-protein colloid solutions have already been used at maximum doses, without having produced an adequate clinical response, and when non-protein colloids are contraindicated¹⁹⁹.

In order to correct bleeding-induced hypovolaemia, as a pharmacological alternative to improve oxygen transport, it is recommended that crystalloid or protein-free colloid solutions are used as the treatment of first choice, with albumin 5% as second-line therapy, when crystalloid or non-protein colloid solutions have already been used at maximum doses, without having produced an adequate clinical response, and when non-protein colloids are contraindicated [1A].

It is recommended that hydroxyethyl starch solutions are not used to correct acute hypovolaemia in bleeding patients because of the increased risk of mortality and renal failure [1B].

It is recommended that high molecular weight hydroxyethyl starch solutions are not used, in order to avoid haemostatic alterations characterised by impaired platelet function [1B].

Patients with moderate bleeding

Patients with limited or moderate bleeding (<30% of circulatory volume or <1,300 mL), who are not at risk of further bleeding, can be treated with infusions of crystalloid solutions^27,199,229, reserving the non-protein colloids for subjects with haemodynamic instability^199,224.

With the purpose of limiting transfusion support in patients with moderate bleeding, it is recommended that the patient’s circulatory volume is first restored using crystalloid or non-protein colloid solutions [1C].

Patients with severe bleeding

Patients with severe bleeding (30–40% of circulatory volume) can initially be treated with crystalloid solutions^27,199,229. Small volumes of Ringer’s lactate can help to maintain a systolic pressure of 80–90 mmHg (controlled hypotension)²³⁰. However, after the initial use of moderate volumes of crystalloids, it is justifiable to add non-protein colloids or vasoactive drugs. Once the circulatory volume has been restored, the possible need for transfusion can be evaluated based on laboratory parameters and estimated blood losses^27,199.

With the purpose of limiting transfusion support in patients with severe bleeding, it is recommended that the patient’s circulatory volume is first restored using crystalloid or non-protein colloid solutions [1C].

Patients with life-threatening bleeding

Patients with life-threatening bleeding (>40% of circulatory volume) who do not respond to initial treatment with 2 L of fluids, have instable haemodynamics or are bleeding at a rate ≥50 mL/minute, require transfusion of red cell concentrates, which can be a potentially life-saving therapy^27,199.

Massive transfusion protocols based on the use of fixed blood component ratios, derived from experience in the military setting, are not currently supported by solid evidence and have important limitations caused by a survival bias^27,231. For this reason, until new evidence is available, the traditional therapeutic approach, based on restoring circulatory volume, monitoring haemostasis and other laboratory parameters and providing transfusion support with the appropriate blood component, should be considered valid for the majority of bleeding patients^27,199.

The use of POC instruments for global monitoring of haemostasis of whole blood can facilitate an individualised approach to the required transfusion support and, in some cases, enable the use of fresh-frozen plasma to be reduced^27,231.

For specific recommendations on the use of POC instruments, refer to the section dedicated to this subject.

With the purpose of limiting transfusion support in patients with life-threatening bleeding, it is suggested that the patient’s circulatory volume is initially expanded with crystalloid or non-protein colloid solutions, followed by transfusion therapy with blood components and, if necessary, with plasma-derived medicinal products [2B].

Dosing infusion fluids

The starting dose of crystalloids (preferably Ringer’s lactate solution), to infuse at the rate of 60–80 mL/kg/hour to maintain a systolic pressure of about 80–90 mmHg, is 3 mL per mL of blood loss^27,230.

The initial dose of colloids is 1 mL per mL of blood loss^27,230.

Haemostatic agents for topical use

The haemostatic agents for topical use include blood components, such as fibrin glue, and medical devices.

Fibrin glue is a blood component for topical use that has been exploited in surgery for more than 20 years¹⁹⁹; its main components are fibrinogen, FXIII, thrombin (and calcium chloride with or without antifibrinolytics) which can be applied contemporaneously or in succession on the surfaces to be treated. The application of fibrin glue reproduces, in situ, the final stage of the coagulation cascade mediating the activation of fibrinogen by thrombin¹⁹⁹. Given its haemostatic potential, this blood component for topical applications has been used with the aim of reducing allogeneic transfusion requirements in various surgical settings, despite its high cost. However, its efficacy in controlling bleeding was found to be greatest in orthopaedic surgery²³².

In contrast, medical devices act through a mechanical or physical effect. Examples include cellulose or gelatine granules which, when saturated with blood, swell until forming a mass which reduces bleeding by mechanical compression.

The various haemostatic agents for topical use have overlapping clinical indications. Some products are indicated to facilitate haemostasis (Floseal^®, Sivek^®, Tabotamp^®, and Curaspon Standard^®), others, including fibrin glue, are also indicated to seal tissues and support sutures (Beriplast^®, Quixil^®, Tachosil^®, Tisseel^®, Tissucol^®, Coseal^®, and Glubran^®)²³³. However, some issues have emerged from the analysis of the efficacy and safety of these products, mainly because of the lack of solid evidence derived from well-performed, randomised studies²³⁴.

It is suggested that fibrin glue is used to promote local haemostasis and as a complementary method to limit intra-operative use of blood, on the basis of local protocols that take into consideration the characteristics of the individual patient (bleeding risk, multiple alloimmunisation), the type of operation, the experience of the surgical and anaesthetic teams, as well as the possibility of combining this blood component with other blood-saving strategies, taking into account the cost-efficacy ratio [2B].

No evidence-based recommendations can be made on the routine use of haemostatic medical devices for topical use in elective orthopaedic surgery.

Blood-conserving anaesthesiological techniques

Various studies have demonstrated that neuroaxial anaesthesia (epidural and subarachnoid) can be associated with a significant reduction in bleeding, especially if used during orthopaedic operations,^235,236. The data are not, however, unanimous because a meta-analysis concerning patients undergoing total knee replacement did not show significant differences depending on whether neuroaxial or general anaesthesia was used²³⁷. Nevertheless, it does seem that neuroaxial anaesthesia can reduce transfusion requirements because of the relative systemic hypotension associated with the technique which is due to sympathetic nervous system blockade and a consequent reduction in venous tone. This effect is, however, variable and its magnitude is not always predictable. In fact, there is not a good correlation between the dose of anaesthetic administered and the effect on blood pressure, because of the variations in blood pressure induced by the different level of nervous blockade obtained²³⁸.

Besides neuroaxial anaesthesia, lumbar plexus blockade was also associated with a reduction of both intra-operative blood loss (22%, 310 mL compared to 617 mL) and total blood loss (45%, 712 mL compared to 1,074 mL), when used in patients undergoing prosthetic hip surgery^239,240.

As far as concerns general anaesthesia, the use of completely intravenous propofol-based anaesthesia is associated with a decrease in blood loss in spinal surgery²⁴¹.

However, studies comparing blood losses related to various different anaesthesiological techniques in orthopaedic surgery should be interpreted with care, since most were conducted in a period in which more liberal transfusion practices were adopted compared to those currently in use. Furthermore, it is possible that additional blood-sparing techniques (such as peri-operative blood recovery) can reduce or even annul the advantages of using neuroaxial techniques of anaesthesia. Nevertheless, despite these limitations, it can be expected that these latter may reduce surgical bleeding, at least in selected subgroups of patients.

As far as concerns body temperature, it is now well established that this is related to the degree of transfusion support. Hypothermia during surgical procedures is produced by a combination of different factors which participate in the loss of body heat (low temperature in the operating theatre, administration of fluids that have not been warmed, changes in the mechanisms of thermoregulation induced by the anaesthesia and evaporation from body cavities in the case of abdominal or thoracic surgery). A drop, even moderate, in body temperature is able to affect the physiological mechanisms of haemostasis, modifying platelet function and inhibiting the temperature-dependent enzyme reactions of coagulation. It has been shown that even mild hypothermia (a decrease of <1 °C in the body temperature) can increase blood losses by as much as 16%, with a related increase in the risk of receiving transfusion therapy (22%)^242–244.

The choice of anaesthetic technique in orthopaedic surgery must take into account the potential benefit of regional techniques in limiting transfusion requirements; anaesthetists should be aware that regional anaesthesia in spontaneously ventilated subjects may have advantages in terms of transfusion requirements (by decreasing blood loss).

It is suggested that loco-regional anaesthetic techniques are used, depending on the experience of the anaesthetic team, with the aim of contributing to limiting intra-operative blood loss [2C].

It is essential that body temperature is monitored during surgery.

It is recommended that hypothermia is prevented, and treated if it occurs, by pre-warming infusion fluids and warming the patient; the purpose is to limit intra-operative bleeding, as well as for obvious reasons of comfort [1C].

Pharmacological techniques and haemostatic agents

Drugs to reduce surgical bleeding and transfusion needs in surgery can be used both in prevention and correction of any disorders of haemostasis in the peri-operative period. Recently the indications for the use of individual clotting factor concentrates (fibrinogen, rFVIIa) or combinations (PCC) have been extended to surgical patients with life-threatening bleeding, although solid confirmation of the evidence is still required⁸⁸.

Recent data confirm the central role of fibrinogen in the formation of stable clots. In this regard, there are emerging data from both randomised and prospective cohort studies, carried out in various clinical settings, indicating that optimisation of coagulation through administration of fibrinogen concentrate is associated with reductions in peri-operative bleeding and transfusion support^245–247.

A recent meta-analysis showed the clinical and statistical heterogeneity of the trials included and highlighted that it was impossible to classify the trials as being “at low risk of bias”; the meta-analysis confirmed the statistically significant reduction in allogeneic transfusion support but did not show any effect on the other pre-defined outcomes, including mortality and bleeding²⁴⁸. However, an update of the abovementioned meta-analysis²⁴⁹ did find a statistically significant reduction of bleeding, mainly in elective heart surgery, while reiterating the lack of demonstrated efficacy on mortality and the need for caution when extrapolating results to other settings such as trauma and obstetrics. In fact, the increasing use of fibrinogen concentrates is supported mainly by observational studies with substantial methodological limitations which do not currently justify the routine administration of these concentrates.

The usefulness of determining the pre-operative concentration of plasma fibrinogen, according to the Clauss method, has been highlighted^250,251. The liver produces from 2 to 5 grams/day of fibrinogen and guarantees mean plasma levels between 2 and 4.5 g/L²⁵².

Fibrinogen concentrate has been demonstrated to have a better cost/benefit ratio in clinical use when the indication comes from monitoring coagulation through thromboelastographic methods^88,199,253.

In patients with congenital deficiency of fibrinogen, levels of fibrinogen in the blood below 1 g/L are considered critical^164,166,254.

In contrast, in patients with acquired hypofibrinogenaemia, the critical levels of fibrinogen, in the presence of massive bleeding, are more controversial and supported mainly by guidelines rather than evidence deriving from randomised studies^248,249,255. In fact, two guidelines published in 2013 indicate that trigger levels of fibrinogenaemia in the presence of massive haemorrhage are <1.5–2 g/L^88,256, while five earlier recommendations, published between 2006 and 2011, indicate levels <0.8–1 g/L^{199,257–260}. Similarly, recommended target levels of plasma fibrinogen are >1 g/L²⁶¹ or between 1.5 and 2 g/L²⁵⁹.

Plasma-derived fibrinogen concentrates have a consolidated and privileged role, with respect to fresh-frozen plasma and cryoprecipitate, in the treatment of bleeding related to a congenital deficiency of this clotting factor^164,166,254. The use of this plasma-derived medicine is also increasing in the setting of bleeding due to acquired deficiencies, in part because of the numerous advantages it has compared to the use of allogeneic blood components. In fact, fibrinogen concentrate is safer, not only with regards to infection risk, and there are no problems of ABO incompatibility with the recipient; furthermore, fibrinogen concentrate enables replacement therapy to be delivered with much smaller volumes (200 mL per 4 grams of fibrinogen) compared to fresh-frozen plasma or cryoprecipitate²⁵⁵.

Data on the use of fibrinogen in elective orthopaedic surgery are very limited. The effect of fibrinogen concentrate was analysed retrospectively in a cohort of nine paediatric patients with craniosynostosis²⁶² and prospectively in 66 patients treated with crystalloids or colloids during major orthopaedic surgery²⁶³. In both studies, the administration of fibrinogen had favourable effects on the dynamics of clot formation and on clot strength.

For specific recommendations on the use of fibrinogen, refer to the paragraph on POC instruments

In recent years there has been an increase in off-label use of rFVIIa for the treatment of critical bleeding in surgical patients^264,265. However, a systematic Cochrane review recommends that rFVIIa is used exclusively in the context of clinical trials²⁶⁶. When used, the suggested dose is 90–120 mg/kg, repeated if necessary, after correction of some abnormalities, such as hypofibrinogenaemia, thrombocytopenia, hypothermia, acidosis and hyperfibrinolysis⁸⁸.

For specific recommendations on the use of rFVIIa, refer to the paragraph on POC instruments

The main indication for the use of non-activated PCC is neutralisation of the anticoagulant effect in patients receiving oral treatment with VKA or NOAC, while activated PCC (FEIBA) are used in the treatment of patients with inhibitors of clotting factors and, recently, have also been suggested for use in patients being treated with NOAC. Thus, PCC, activated or not, are used both in cases of severe bleeding and in cases of urgent surgery. These plasma-derived medicinal products have, in fact, been demonstrated to be useful in preventing or treating bleeding in the presence of peri-operative coagulopathy⁸⁸.

For specific recommendations on the use of PCC refer to the section on monitoring patients and management of post-operative bleeding

Fibrinolysis is an important cause of bleeding in surgical patients. There are some types of surgery (total knee replacement with a tourniquet, cardiopulmonary bypass, liver transplantation) which are more frequently associated with fibrinolysis and, therefore, benefit more than others from the use of TXA. Systematic reviews of randomised controlled trials indicate that the use of TXA has a significant impact, reducing transfusion support, particularly in cardiac and orthopaedic surgery, and indicate that the use of this drug is effectively associated with reductions in both the number of patients transfused and in their transfusion requirements^88,267,268. As far as concerns the dose of TXA in major orthopaedic surgery, a meta-analysis showed that the reduction in the risk of receiving red blood cell transfusions is independent of the total dose of TXA administered²⁶⁹. Similarly, the timing of the administration of TXA, that is, at induction of anaesthesia rather than before release of the tourniquet, does not seem to affect the efficacy or safety of the treatment²⁶⁷.

It is suggested that intravenous TXA is administered in hip and knee replacement surgery (as well as in major surgery of the spinal column) [2A]^270–278.

The most frequently used doses of TXA are as follows: total hip or knee replacement surgery, a starting dose of 10–15 mg/kg before surgery, followed or not by an infusion of 1 mg/kg/hour for 4–6 hours or repetition of the initial dose in the post-operative period; surgery to the spine, a starting dose of 20–100 mg/kg, followed by an infusion of 10 mg/kg/hour for 4–6 hours²⁷.

Since there are reports of a state of hypercoagulability being induced in some patients (subjects with a past history of thromboembolic events, those aged over 60 years, females, those undergoing hip trauma surgery or cancer surgery), it is suggested that the risks and benefits of TXA are analysed carefully before its use [2C]²⁷⁹.

TXA applied topically, also in combination with intravenous administration²⁸⁰, has been found to be safe and effective in containing transfusion requirements in elective hip and knee surgery^281–290 and does not cause adverse biomechanical effects on the prostheses used²⁹¹.

It is suggested that TXA is used topically, also in combination with intravenously administered TXA, in patients undergoing elective hip or knee arthroplasty [2B].

It is suggested that TXA is used only by the topical route in patients undergoing elective hip or knee arthroplasty who have risk factors for hypercoagulability according to the pre-operative history [2B].

Point-of-care testing

Since the decision to transfuse red cell concentrates in the peri-operative period is based in part on the patient’s Hb concentration, the availability of rapid, reliable measurements of this parameter, with POC instruments, increases the safety of transfusion and optimises this therapeutic intervention^88,292.

It is recommended that POC instruments acquired and used measure the concentration of Hb and the haematocrit without needing dilution of whole blood in the pre-analytic phase [1C].

The POC instruments for monitoring haemostasis use whole blood²⁹² and can be employed not only in transfusion facilities and laboratories, but also in the Accident and Emergency Department and operating theatre, guaranteeing shorter performance times than those of standard laboratory tests⁸⁸.

The main POC instruments that can be used for global monitoring of haemostasis are the thromboelastograph (TEG) and the thromboelastometer (ROTEM). These rapidly produce results in both numerical and graphical form and are able to detect the anticoagulant effect of acidosis or hypothermia and hyperthermia, since they can be used in a range of body temperatures between 22 °C and 42 °C; furthermore, they can pick up and quantify thrombocytopenia, clotting factor deficiencies, the effect of heparin, hypofibrinogenaemia and hyperfibrinolysis²⁹³.

Although a recent systematic Cochrane review showed that the use of TEG and ROTEM to monitor haemostasis in patients with massive bleeding does not have significant advantages in terms of morbidity and mortality compared with the use of standard laboratory tests²⁹⁴, these POC instruments are widely used in elective surgery, in heart surgery and, in particular, in liver transplantation and in elective major orthopaedic surgery to guide replacement therapy with clotting factors (with plasma or fibrinogen concentrate)^{27,245,246,295–299}, or to monitor hyperfibrinolysis and to evaluate the possible use of TXA³⁰⁰.

One limitation of the TEG is its incapacity to evaluate platelet function adequately^27,301.

It is suggested that POC instruments (TEG and ROTEM) are used to monitor overall haemostasis with the purpose of guiding clotting factor replacement therapy and to limit the use of transfusion support with blood components in elective major orthopaedic surgery with a high risk of bleeding or in the presence of major bleeding [2A].

It is suggested that the concentration of fibrinogen is measured pre-operatively using the Clauss method [2C]^199,250.

In the presence of massive bleeding during elective major orthopaedic surgery and in association with the correction of the triggering cause, it is suggested that severe hypofibrinogenaemia (<1 g/L) which persists despite treatment with fresh-frozen plasma should be treated with fibrinogen concentrate (or, if not available, with cryoprecipitate) [2C].

In the same conditions and during massive transfusion, it is suggested that treatment with fibrinogen concentrate is considered when the level of fibrinogenaemia is <1.5 g/L, to prevent the fibrinogen level from falling below 1 g/L, the critical threshold for haemostasis [2C].

It is suggested that the administration of fibrinogen is privileged over that of fresh-frozen plasma (or cryoprecipitate) when there are contraindications to volume overloading [2C].

It is suggested that an initial dose of 25–50 mg/kg of fibrinogen concentrate is administered [2C].

rFVIIa is indicated for the treatment of patients with haemophilia with inhibitors, those with congenital deficiency of FVII and patients with Glanzmann’s thrombasthenia. However, it has also been widely used to control bleeding in off-label settings, such as trauma, heart surgery, liver surgery, post-partum and cerebral haemorrhage.

Although rFVIIa has been demonstrated to have variable effects on morbidity and transfusion requirements, a definitive effect on mortality has not been observed and, for this reason, there is not a clear indication for its use, particularly in elective surgery^{256,265,302–304}. Furthermore, there have been recent reports of side effects, especially arterial and venous thromboembolic events, which have limited or even contraindicated its use, as in the case of cerebral haemorrhage^304,305. Exceptionally, the use of rFVIIa can be considered in patients with life-threatening haemorrhage and after the failure of conventional haemostatic treatments²⁷.

A good outcome and effective response to the use of rFVIIa do, however, require the prior restoration and control of some conditions: adequate levels of platelets and plasma fibrinogen, almost physiological levels of pH and body temperature and correction of hypocalcaemia^88,256.

Off-label administration of rFVIIa (90 μg/kg) is suggested for the treatment of bleeding uncontrolled by conventional surgical or radiological treatment and/or in the case of failure of haemostatic therapy [2C].

Post-operative period

Autologous transfusion techniques

Post-operative blood recovery

Post-operative blood recovery consists in collecting the blood that the patient loses through surgical drains into a specific container and subsequently re-infusing it into the patient³⁰⁶. This procedure can make use of two systems: “no wash” and “wash”.

With the “no wash system” the blood is transferred from the container connected to the drains to the infusion set and is re-infused without undergoing further treatment. There is an integrated double filtration system with a first filter of 100–200 μ, for fibrin and macroaggregates, and a second filter of 40 μ for microaggregates. Anticoagulants are not necessary because this blood does not contain fibrinogen. This system involves simple, economic and easy-to-use methods.

The “wash system” involves the use of specific equipment which centrifuges the blood collected, eliminates the supernatant, washes the red blood cells and resuspends them in saline solution. This system is more demanding and staff require careful training in its use.

The main controversies regarding the use of unwashed blood concern the volume of blood that can be recovered, which is often less than a therapeutic dose, the quality and safety of the blood, because of the numerous contaminants that it contains, and the lack of adoption of transfusion triggers to decide when to transfuse the patient^307–310.

However, although intuitively washing could reduce the presence of contaminants in recovered blood and the possible complications in recipients associated with their administration³⁰⁸, it should be remembered that, at present, there are no direct comparisons between blood products obtained with the two methods and the outcomes associated with their use³⁰⁹. For this reason, further studies are needed to evaluate the safety of post-operative blood recovery techniques that do not include washing of the recovered blood³¹¹.

Post-operative blood recovery, both with and without washing, has varied efficacy in reducing the amount of transfusion support with allogeneic red blood cells both in elective joint prosthesis surgery and in spinal surgery^{212,312–316}.

Numerous meta-analyses have demonstrated the efficacy of TXA, also when applied topically²⁸⁹, in limiting post-operative bleeding in patients undergoing major orthopaedic surgery^{269–277,289,290,317,318}: TXA could, therefore, further reduce the role of post-operative blood recovery techniques.

Furthermore, a recent randomised controlled trial demonstrated, in patients undergoing elective lower limb arthroplasty with pre-operative Hb above 130 g/L, that not only were intra-operative and post-operative blood recovery unable to decrease transfusion requirements, but they actually increased costs²⁰⁵.

It is recommended that post-operative blood recovery is used in elective major orthopaedic surgery (hip arthroplasty, knee arthroplasty, spinal surgery) only in the case in which it is expected that the post-operative blood loss will be ≥10% of the patient’s circulatory volume despite implementing multimodal strategies, including integrated use of other blood-conserving techniques (pharmacological, surgical and anaesthesiological), also taking into account the characteristics of the individual patient (bleeding risk, multiple alloimmunisation) [1B].

It is suggested that the use of “washed” blood should be preferred (both in orthopaedic surgery and in other settings) [2B]³⁰⁶.

When using systems that do not include washing, it is suggested that the concentration of free Hb is determined before re-infusing the unwashed blood, in order to check that the degree of haemolysis is less than 0.8% of the red cell mass contained in the product transfused into the patient [2C]^{306,319–321}.

Control of body temperature and prevention of stress ulcers

For the reasons already set out, patients must maintain a body temperature suitable for optimal haemostatic function also in the post-operative period. In this period a patient’s body temperature may be low because of ineffective treatment of a possible intra-operative hypothermia, because of the prolongation of thermomodulatory effects of anaesthetics or because of admission in poorly heated rooms³²². It is, therefore, essential that the patient’s body temperature continues to be monitored also in the period following any surgical procedure and that all methods and strategies useful for preventing hypothermia (blankets, warming fluids administered intravenously, etc.) are implemented.

It is recommended that the patient’s body temperature is monitored also in the post-operative period, implementing all strategies aimed at preventing hypothermia [1C].

Classically, all patients subjected to a state of stress (including that related to a surgical intervention) are at risk of upper gastrointestinal tract bleeding. Decades of research have demonstrated the advantages of pharmacological prophylaxis of this complication; such prophylaxis lowers the incidence of gastrointestinal bleeding deriving from the development of stress-induced gastritis and is based on the use of various categories of drugs: proton pump inhibitors, antagonists of histamine H2 receptors and sucralfate. In critically ill patients, proton pump inhibitors seem to be more effective than the antagonists of histamine H2 receptors in preventing clinically important bleeding from the upper gastrointestinal tract. The robustness of this conclusion is, however, limited by the poor methodological quality of the studies available, by differences between them and by possible publication bias³²³. Recent observational studies do, however, suggest that bleeding from stress ulcers is very rare. Furthermore, the risk of bleeding does not seem to be modified significantly by the use of acid-suppressing therapy which could even cause a potential increase in the risk of pneumonia and infections by Clostridium difficilis because of excessive alkalinisation of the gastric secretions^324,325.

Because it is considered that the mechanism responsible for stress-induced gastritis is a decrease in blood flow in the mucosa, with consequent tissue ischaemia, it is thought that early recovery of central alimentation, which improves mucosal blood flow, can lower the risk of occult or clinically evident bleeding. However, the clinical studies that have evaluated the efficacy of this strategy compared with acid-suppressing drugs have yielded variable results and, so far, randomised controlled trials have not been performed³²⁶.

It is suggested that prophylaxis of stress-induced gastrointestinal tract ulcers is not given. This suggestion does not apply to patients in the intensive care unit and those already with local disorders for which prophylaxis is specifically indicated [2C].

Careful monitoring of the patient and management of post-operative bleeding

Anticoagulant treatment and pharmacological interactions

In the post-operative period, re-introduction of anticoagulant therapy or the start of anticoagulant prophylaxis, dictated by the need to protect the patient from the risk of thromboembolism compels an evaluation of bleeding risk.

There are two main mechanisms by which anticoagulant drugs are metabolised and eliminated from the body: the cytochrome system (CYP450) and the P-glycoprotein transport system³²⁷. All the drugs or substances that activate or inhibit these systems can have repercussions on the anticoagulant effect which will be inhibited or potentiated.

At present, few interactions between NOAC and food are known. Table IV³²⁷ reports the possible interactions between NOAC and other drugs; these interactions are based on the mechanisms of metabolism and it is important to know them because elderly patients are ever more frequently taking multiple drugs concurrently over the long term³²⁷.

Table IV.

Interactions between new oral anticoagulants and other drugs (modified from Pengo V et al.³²⁷).

	Dabigatran	Rivaroxaban	Apixaban
Glycoprotein-P inhibitors: amiodarone, phenothiazine, thioxanthenes, carboxylic acid, azole antifungal agents, verapamil, antimalarial drugs, cyclosporine	Yes	Yes	Yes
Glycoprotein-P inducers: dexamethasone, rifampicin, hypericum*	Yes	Yes	Yes
Cytochrome CYP3A4 inhibitors: phenothiazine, carboxylic acid, azole antifungal agents, verapamil, erythromycin, telithromycin, nefazodone, antimalarial drugs, cyclosporine, thioxanthenes	No	Yes	Yes
Cytochrome CYP3A4 inducers: carbamazepine, efavirenz, nevirapine, phenytoin, phenobarbital, rifabutin, rifapentin, hypericum*, alcohol, eucalyptol	No	Yes	Yes
Non-steroidal anti-inflammatory drugs: aspirin, naproxen, diclofenac	Yes	Yes	Yes
Anti-platelet drug: clopidogrel	Yes	Yes	Yes

^*or St. John’s wort, Hypericum perforatum.

Many of the numerous interactions between warfarin and other drugs, or food, are due to the former’s almost exclusive hepatic clearance, mediated by CYP450, in particular the isoform CYP2C9. In contrast, dabigatran seems to be metabolised in the plasma and liver without the intervention of CYP450, but is a substrate for P-glycoprotein and, for this reason, is susceptible to the effects of inhibitors or activators of P-glycoprotein. Rivaroxaban and apixaban are metabolised by the CYP3A4 isoform of CYP450 and are also substrates for P-glycoprotein³²⁸, thus being sensitive to inhibitors and inducers of both systems. However, the demonstrated pharmacological interactions only make dose adjustments of NOAC necessary in some cases. Other possible interactions are caused by the association with non-steroidal anti-inflammatory drugs and platelet anti-aggregants³²⁹.

Monitoring the patient and post-operative bleeding

In the post-operative period the patient may be subject to both thromboembolic and haemorrhagic complications.

In the post-operative period, it is recommended that antiplatelet therapy is restarted as early as possible, to prevent platelet activation and possible thrombotic complications [1C].

It is suggested that the first post-operative dose of clopidogrel or prasugrel is administered no more than 24 hours after the end of the operation [2C].

It is suggested that this first post-operative dose is not a loading dose [2C].

Although the use of transfusion targets, determined by coagulation tests or POC tests, are used to guide treatment with blood components, these tests are not useful for monitoring thrombotic risk^330–333. The evidence for post-operative use of platelet function tests (PFA-200, MEA) is very limited^334,335.

Some studies have shown that desmopressin can improve altered platelet function in volunteers treated with aspirin or clopidogrel^336,337; however, its use in acquired bleeding disorders is not supported by solid evidence.

Despite the lack of studies on the capacity of transfused platelets to overcome the effect of treatment with clopidogrel or prasugrel, the management of important bleeding in patients being treated with antiplatelet drugs is based on the transfusion of platelets³³⁸. Likewise, studies have not been performed on the efficacy of platelet transfusions in patients being treated with ticagrelor; the presence of this drug in the plasma makes the transfusion of platelets ineffective when the drug has been administered in the preceding 12 hours⁸⁸.

It is suggested that platelets are transfused in the case of intra-operative or post-operative bleeding, clearly related to the intake of antiplatelet agents (aspirin, clopidogrel, prasugrel) [2C].

Unfractionated heparin (UFH) and LMWH are used in bridging therapy, in peri-operative thromboprophylaxis, in the treatment of venous thromboembolism, in disseminated intravascular coagulation and for anticoagulation in haemodialysis and heart surgery.

The aPTT and measurement of anti-FXa activity are able to determine the anticoagulant effect of UFH and LMWH, respectively, even if these tests are only useful in particular circumstances in the context of surgery.

The anticoagulant effect of heparins can be corrected rapidly by the intravenous administration of protamine sulphate (1 mg of protamine neutralises 100 IU of UFH or 100 units of anti-FXa of LMWH)³³⁹; however, the correction induced by protamine is less efficient in the case of subcutaneous administration of UFH or LMWH and it may, therefore, be necessary to prolong the intravenous infusion of protamine or give a second dose.

It is recommended that bleeding associated with the intravenous administration of UFH is treated with an intravenous infusion of protamine sulphate at a dose of 1 mg per 100 IU of UFH administered in the preceding 2–3 hours [1A].

It is suggested that bleeding associated with subcutaneous administration of UFH that does not respond to intravenous protamine sulphate (1 mg per 100 IU of UFH) is treated with a continuous intravenous infusion of protamine sulphate at a dose guided by the aPTT [2C].

It is suggested that bleeding associated with subcutaneous administration of LMWH is treated with intravenous protamine sulphate at a dose of 1 mg per 100 units of anti-FXa of LMWH administered [2C].

If there is no response, it is suggested that a second dose of protamine sulphate is given (0.5 mg per 100 units of anti-FXa of LMWH administered) [2C].

Fondaparinux is a synthetic analogue of the polysaccharide sequence present in UFH and LMWH with a selective action on FXa, which is inactivated after binding to antithrombin. It is used in the prophylaxis and treatment of thromboembolic diseases and myocardial infarction. There is no antidote for fondaparinux; on the basis of in vitro studies, it has been proposed that rFVIIa is given to control severe bleeding caused by the administration of fondaparinux^339,340.

It is suggested that bleeding related to the subcutaneous administration of fondaparinux is treated with rFVIIa (off-label use) [2C].

Pharmacological VKA, such as acenocoumarol (sintrom) and warfarin (coumadin) and others not yet available in Italy, are used for the prophylaxis or treatment of ischaemic and thrombotic events in patients with mechanical prosthetic heart values, in atrial fibrillation and in thromboembolic disease. With an increase in the number of elderly subjects and of disorders associated with thromboembolism, treatment with VKA is becoming ever more widespread and thus the bleeding complications of these drugs are becoming more frequent.

The anticoagulant effect of VKA can be rapidly reversed by the prompt use of PCC^341–347, although there are no randomised controlled trials supporting their use in a setting other than haemophilia and the therapeutic dose has not been standardised because of the differences between the individual products²⁵⁶.

The therapeutic anticoagulant effect of VKA is evaluated by measuring the INR which, before a surgical interventions, should be 1.5 or less. To achieve such a level, the Federation of Centres for the Diagnosis of Thrombosis and Monitoring of Antithrombotic Therapy argues for rapid neutralisation of VKA by intravenous administration of vitamin K and PCC, at a dose regulated on the basis of the pre-operative value of the INR³⁴⁸.

Since the use of PCC exposes treated patients to an increased risk of arterial and venous thrombosis, thromboprophylaxis should be resumed as soon as the bleeding has been controlled.

In patients receiving oral anticoagulant treatment with VKA who must undergo urgent surgery or who develop bleeding complications in the post-operative period, it is recommended that the anticoagulant therapy is interrupted immediately, that 10–20 mg of vitamin K1 is administered as soon as possible by slow intravenous infusion (15–30 minutes) and that PCC is infused slowly at the following doses: 20 IU/kg, if the INR is <2; 30 IU/kg, if the INR is between 2 and 3.9; 40 IU/kg, if the INR is between 4 and 5.9; and 60 IU/kg, if the INR is >6 [1B].

In recent years, NOAC (dabigatran, rivaroxaban, apixaban) are becoming increasingly used for the prophylaxis and treatment of thromboembolic events. The current lack of valid laboratory tests able to measure their effective anticoagulant action makes therapeutic monitoring futile, whereas such monitoring is essential for VKA.

However, there is a specific test to measure the anti-FXa activity of the new anticoagulants with an anti-FXa effect (rivaroxaban, apixaban)³⁴⁹ and there is also a method to evaluate the inhibitory effect of the drugs with an anti-thrombin mechanism of action (dabigatran): the diluted thrombin time¹¹⁹. However, an accepted protocol for their peri-operative use has not yet been established.

No evidence-based recommendations can be made on the use of laboratory tests in the post-operative evaluation of the anticoagulant effect of NOAC.

Whereas there is an antidote for VKA, there is not a specific antidote to neutralise the anticoagulant effect of NOAC, whether their mechanism of action is anti-FXa or anti-thrombin³³⁹.

Clinical experience is very limited and the paucity of published studies does not allow specific recommendations to be made. However, it has recently been demonstrated, in healthy volunteers, that the effect of these drugs on the tests of inhibition of FXa can be quickly and completely corrected by high doses of PCC (50 IU/kg); in contrast, this same dose does not correct the prolongation of the aPTT and thrombin time induced by dabigatran³⁵⁰.

In the case of direct inhibitors of thrombin (dabigatran), the anticoagulant effect has been corrected by dialysis or the administration of activated concentrates of PCC (FEIBA), although the guidelines of the European Heart Rhythm Association suggest using PCC, as for other NOAC¹¹⁴. There is little evidence in favour of the use of rFVIIa^114,351,352.

In patients being treated or suspected of being treated with oral anti-FXa agents, such as rivaroxaban or apixaban, it is suggested that anti-FXa activity is determined, where this test is available [2C].

In the presence of important bleeding, it is suggested that the effect of rivaroxaban or apixaban is neutralised with high doses of PCC (25–50 IU/kg) or with activated PCC (50 U/kg, repeated if necessary) or rFVIIa (90 μg/kg) [2C].

In the presence of important bleeding in patients being treated with oral direct inhibitors of thrombin, such as dabigatran, it is suggested that high doses of PCC (25–50 IU/kg) are administered, or activated PCC (50 U/kg, repeated if necessary) or rFVIIa (90 μg/kg) [2B].

The more common and less severe platelet disorders respond well to desmopressin, which shortens the bleeding time, whether used as prophylaxis or for the treatment of bleeding^144–146. When used, the standard dose is 0.3 μg/kg, diluted in 50 mL of saline solution and infused slowly, over more than 30 minutes¹⁴⁴. The use of antifibrinolytic drugs in inherited platelet disorders is not evidence-based.

In patients with inherited disorders of platelet function, it is suggested that desmopressin is used for the prevention and control of post-operative bleeding, and that TXA is used as an adjuvant [2C].

rFVIIa is indicated for the treatment of bleeding only in patients with Glanzmann’s thrombasthenia, whereas it is not indicated for any other congenital or acquired platelet disorders. When used, the proposed dose is 90 μg/kg, repeated every 2 hours for 12 hours and, subsequently, every 3–4 hours until the bleeding stops¹⁴².

It is recommended that rFVIIa is used in the treatment of bleeding in patients with Glanzmann’s thrombasthenia [1C].

Patients with thrombocytopenia resemble subjects with modest platelet disorders as far as concerns the treatment of post-operative bleeding complications¹⁵³ and should be treated on the basis of the platelet count and concomitant clinical conditions^{152–155,306,353,354}.

In patients with platelet disorders with a platelet count between 50 and 100×10⁹/L, it is suggested that transfusion of platelet concentrates is considered if there is a high bleeding risk or a risk of bleeding in critical sites such as the eye or brain [2C]^150,152,306.

In the case of congenital functional platelet deficiencies, it is suggested that transfusion of platelet concentrates is considered, independently of the platelet count, in the presence of peri-operative bleeding not related to the surgery or to other clotting disorders [2C]^150,152,306.

Among the bleeding disorders due to a congenital or acquired deficiency of clotting factors, the lack of VWF is undoubtedly the most widespread and the treatment, in the case of post-operative bleeding, is as recommended below^{13,131,133,136,137}.

It is recommended that the post-operative management of the patient with vWD is conducted in collaboration with an expert in haemostasis and thrombosis [1C].

In the presence of mild, post-operative bleeding in a patient with vWD, it is recommended that desmopressin is used, after a trial test, at doses described in specific guidelines (0.3 μg/kg, diluted in 50 mL of saline solution and infused slowly, over more than 30 minutes) [1C]^136–140.

In vWD patients with severe post-operative bleeding it is recommended that replacement therapy with plasma-derived vWF or vWF-rich FVIII is given, according to the treatment regimens set out in specific guidelines [1C].

In vWD patients undergoing major surgery, it is suggested administering antifibrinolytics in the post-operative period, as an adjuvant to more specific treatment, and platelet transfusions, only in the case of failure of other treatments [2C].

It is recommended that the post-operative management of patients with haemophilia or other congenital bleeding disorders is conducted in collaboration with an expert in haemostasis and thrombosis [1C].

Personalised thromboprophylaxis is suggested in the post-operative period in haemophilic patients undergoing major surgery [2C]¹⁶³.

For patients with deficiencies of FVIII (haemophilia A) or FIX (haemophilia B) without inhibitors, the recommendations for post-operative management are reported below^{140,156,158,160}.

It is recommended that appropriate replacement therapy is given [1C].

It is suggested that published guidelines are followed when giving replacement therapy (target of the missing factor and duration of treatment) [2C].

For replacement treatment in the post-operative period, it is recommended that either recombinant or plasma-derived concentrates are used [1C].

The suggested treatment in patients with inhibitors is rFVIIa or activated PCC (FEIBA) [2C].

rFVIIa is the treatment of choice for congenital deficiency of FVII; if this is not available, plasma concentrate of FVII is preferred over PCC, because of the potential prothrombotic effect of these latter^140,164.

In the post-operative period in patients with congenital deficiency of FVII, it is suggested that plasma concentrate of FVII, at a dose of 10–40 IU/kg, is used as an alternative to rFVIIa [2C]¹⁶⁴.

In the post-operative period in patients with congenital deficiency of FVII undergoing major surgery, it is suggested that rFVIIa, at a dose of 15–30 μg/kg every 4–6 hours, is used, usually for a minimum of three doses [2B]¹⁶⁴.

In cases of congenital deficiency of fibrinogen and FXIII, the British guidelines recommend the use of specific concentrates¹⁶⁴. In hypofibrinogenaemia, replacement therapy is indicated for fibrinogen levels <1 g/L or <20–30% of normal. However, a possible risk of thrombosis must be kept in mind when plasma fibrinogen concentrates are used.

In the case of a congenital lack of FV and FXI, fresh-frozen plasma, preferably a industry-produced, virus-inactivated product, is the only therapeutic option, also in the post-operative period¹⁶⁴. Patients with FXI deficiency and inhibitors undergoing surgery can be treated successfully with low doses of rFVIIa (33–47 μg/kg)¹⁶⁸; the contemporary administration of TXA was found to be effective in controlling bleeding¹⁶⁴.

PCC are the reference concentrates for the treatment of FII and FX deficiencies, for which specific concentrates are not available^164,169,170.

In patients with other rare clotting factor deficiencies no evidence-based recommendations can be made on the peri-operative use of rFVIIa, desmopressin or TXA.

Management of the patient with acquired thrombocytopenia

It is suggested that the decision to transfuse platelet concentrates in the post-operative period is not based exclusively on a low platelet count, but also on the clinical conditions of the patient (in particular, body temperature >38.5 °C, plasma clotting disorders, recent bleeding, neurological deficits) [2C]^150,152,306.

Platelet transfusion is rarely indicated in the surgical patient with normal platelet function in whom the platelet count is greater than 100×10⁹/L; such a transfusion is suggested, however, when the count is less than 50×10⁹/L and there is excessive bleeding [2C]^150,152,306.

In the case of intermediate values (platelet count between 50×10⁹/L and 100×10⁹/L) it is suggested that platelet transfusions are considered in specific situations such as secondary platelet dysfunction or when there is a risk of bleeding in critical sites, such as the eye and brain [2C]^150,152,306.

In the case of an acquired deficit of platelet function (for example: anti-aggregant drugs, cardiopulmonary by-pass) it is suggested that platelet transfusions are administered, independently of the platelet count, in the presence of peri-operative bleeding not related to the surgery or to other clotting disorders [2C]^150,152,306.

In the case of acute disseminated intravascular coagulation, in the presence of major bleeding and thrombocytopenia, it is suggested that the platelet count is maintained around 50×10⁹/L [2C]³⁰⁶.

In patients with disseminated intravascular coagulation who are not bleeding, prophylactic transfusion of platelet concentrates is suggested in cases in which the thrombocytopenia and stratification of bleeding risk indicate a high probability of bleeding [2C]³⁰⁶.

It is suggested that prophylactic platelet transfusion is not used routinely when the thrombocytopenia is due to increased platelet destruction (heparin-induced thrombocytopenia, autoimmune thrombocytopenia, thrombotic thrombocytopenic purpura) because it is ineffective in these cases [2C]^152,306.

In anaemic, thrombocytopenic patients (platelet count ≤20×10⁹/L) without active bleeding, it is suggested that the haematocrit is increased to about 30% to reduce the bleeding risk [2B]³⁰⁶.

Optimising tolerance of anaemia

Pre-operative period

Evaluation and optimisation of the individual physiological reserve for tolerance of anaemia and risk factors

Since the decrease in availability of oxygen in an acute episode of anaemia is initially compensated for by an increase in cardiac output, the compensatory cardiac reserve could be used to determine the tolerance to anaemia. However, although evaluating cardio-respiratory functional reserve in patients who are candidates for surgery is part of daily clinical practice for anaesthetists, estimating tolerance of anaemia in an individual patient remains a challenge that has not been completely won. In fact, there are not yet methods or investigations that enable the tolerance of anaemia of a given patient to be predicted a priori³⁵⁵. There are, however, various, detailed guidelines to which the reader is referred^356–359.

In any case, it is worth remembering some key concepts here:

• - pre-operative evaluation of the cardiac reserve is performed primarily to identify those patients for whom cardiac stress related to the operation and to the peri-operative period could represent a particular risk for morbidity or mortality which would be associated with the risk already present due to the patient’s underlying condition. From this point of view, the tolerance of anaemia is deduced from the real state of compensation and the cardiac reserve, determined by various tests³⁵⁵.

• - There is little evidence from well-conducted, randomised trials and many recommendations are based mainly on experts’ opinions³⁵⁹.

• - In the last 5–10 years, various steps forward have been made. Greater emphasis is now given to stratification of clinical risk rather than to routine cardiac testing³⁶⁰.

• - The same considerations apply to the pre-operative evaluation of respiratory function and, given the increase in the current day population of overweight subjects, also to the metabolic evaluation³⁶¹.

In all patients potentially at risk of acute peri-operative anaemia, it is recommended that the cardio-respiratory reserve is evaluated during the pre-operative assessment, using specific diagnostic protocols and flow-charts based on the best, periodically updated indications [1C].

Adoption of restrictive transfusion thresholds

Most of the currently available guidelines recommend basing decisions on transfusion therapy with red blood cells on both the patient’s Hb values as well as symptoms of anaemia (chest pain, congestive heart failure, tachycardia not responsive to the administration of fluid or postural hypotension)^{12,199,306,362–366}. Although the search for physiological transfusion triggers that are most indicative of the state of tissue oxygenation and the presence of ischaemia continues to be a subject of research, at present there are no data from clinical studies that support the use of such measurements, which have, in any case, not yet been validated in clinical practice³⁶⁷. The Hb concentration does, therefore, remain the most widely used transfusion trigger, together with clinical assessment of the patient.

Following the Transfusion Requirements in Critical Care (TRICC) clinical trial³⁶⁸, published in 1999, various studies showed that adopting a restrictive transfusion threshold (Hb=70–80 g/L) guarantees patients’ safety while reducing transfusion use.

In particular, in orthopaedic patients, the recent transfusion trigger trial for Functional Outcomes in Cardiovascular patients Undergoing Surgical hip fracture repair (FOCUS) demonstrated, in elderly patients at high cardiovascular risk, that a liberal strategy (Hb transfusion threshold=100 g/L) did not reduce mortality, the rate of inability to walk autonomously at 60 days after surgery or even morbidity during the admission, in comparison with a restrictive strategy (Hb transfusion threshold >80 g/L or symptomatic anaemia)³⁶⁹. These results were recently further supported by a post-hoc analysis of a randomised study of patients undergoing lower limb prosthetic surgery³⁷⁰ and by a meta-analysis³⁷¹.

The greater efficacy of a restrictive transfusion threshold (Hb<70 g/L) compared to the liberal one (Hb<90 g/L) was also supported by a significantly lower mortality in patients with gastrointestinal bleeding³⁷².

Another randomised study demonstrated that a restrictive strategy (haematocrit >24%) and a liberal one (haematocrit >30%) were equally safe in patients undergoing heart surgery³⁷³.

On the other hand, a recent pilot study in patients with acute coronary syndrome showed that the liberal transfusion strategy was associated with a better outcome in this particular setting³⁷⁴.

It is recommended that a restrictive transfusion threshold (Hb 70–80 g/L) is adopted in all hospitalised, clinically stable patients [1A].

It is suggested that a restrictive transfusion threshold is adopted in hospitalised patients with previous cardiovascular disorders [2B].

It is suggested that transfusion therapy is considered in hospitalised patients with previous cardiovascular disease if they are symptomatic or have a Hb≤80 g/L [2B].

It is suggested that hospitalised patients with acute coronary syndrome who are haemodynamically stable undergo a careful clinical evaluation, aimed at determining a personalised transfusion threshold [2C].

Intra-operative period

Optimising cardiac output. Optimising ventilation and oxygenation

The availability of oxygen in the tissues is the product of the cardiac output and the arterial oxygen content. For this reason, from a physiological point of view, the availability of oxygen in tissues can be stabilised by manipulating the two factors involved.

There is currently more and better clinical evidence regarding optimisation of haemodynamic function (or goal-directed therapy [GDT])^375,376. However, two recent, randomised studies in surgical patients at high risk did not show a significant impact of GDT on outcome^377,378. The physiological rationale underlying GDT is that bringing a patient’s cardiovascular function to an optimal level, through a specific, codified therapeutic approach, ensures the best oxygen supply to tissues in that individual patient. GDT is, therefore, mainly based on the intravenous administration of fluids, but also cardio-active drugs, in doses such as to bring the individual patient to the part most efficient of his or her cardiac performance, as expressed by the Maestrini-Starling law^379,380. This objective can only be achieved if systems for haemodynamic monitoring are used that evaluate cardiac output rather than the so-called “filling pressure”, heart rate or mean systemic blood pressure. In fact, these last parameters have now been considered for some years as insensitive indicators of hypovolaemia or changes in cardiac output³⁸¹. Cardiac output has traditionally been evaluated by an invasively introduced Swan-Ganz catheter. More recently, less invasive, reliable monitoring instruments have become available (particularly in circumstances of totally assisted pulmonary ventilation)^382–384.

It has been amply demonstrated that the management of circulatory volume according to the principles of GDT also enables better dosing of fluid therapy, avoiding the complications related to the administration of too much or too little fluids, thus positively affecting the patients’ outcome^{375,385–388}.

Prompt intra-operative management of haemodynamics, according to the principles of GDT, is recommended [1B].

It is recommended that hypervolaemia and hypovolaemia are avoided [1B].

It is recommended privileging systems that monitor blood flow in order to guide fluid therapy [1B].

In cases in which the patient is undergoing mechanical ventilation, the intra-operative availability of oxygen in the tissues could also be improved by increasing the inspired fraction of oxygen (FiO₂), since this is able to increase the arterial partial pressure of oxygen proportionally. Furthermore, because of the great difference in the partial pressure of oxygen between blood and tissues, in conditions of marked anaemia, an increase in the amount of oxygen physically dissolved in the blood, which is obtained by ventilation with high FiO₂, can improve tissue oxygen availability. In this regard, it has been demonstrated in animal models that hyperoxic ventilation increases the rate of survival and tolerance of anaemia by guaranteeing tissue oxygenation even at very low levels of Hb^389–392. Thus, hyperoxic ventilation can be considered for patients with acute anaemia undergoing general anaesthesia. However, at present there is a lack of large clinical studies demonstrating that this method does actually improve tissue oxygenation during acute, severe anaemia. Furthermore, there are not yet controlled studies clarifying whether ventilation with high concentrations of oxygen can really improve survival in cases of acute anaemia that continues for prolonged periods. Finally, it should be remembered that the side effects of prolonged hyperoxic ventilation (arteriolar vasoconstriction, generation of oxygen radicals and formation of atelectasis) can outweigh the positive effects on oxygen transport and tissue oxygenation³⁵⁵. However, considering that the exposure is limited in time (intra-operative period) it is reasonable to expect that the damage known to occur with prolonged hyperoxia does not develop in surgical patients treated with this method³⁹³.

In the case of acute intra-operative anaemia in patients undergoing general anaesthesia and mechanical ventilation, it is suggested that hyperoxic pulmonary ventilation is administered (FiO₂=1). This technique, especially if used in combination with normovolaemic haemodilution, can be useful, at least to enable greater availability of tissue oxygen until the bleeding is controlled [2C].

Post-operative period

Optimising the tolerance of anaemia. Maximising oxygen delivery. Minimising oxygen consumption

The considerations on maximising oxygenation in the intra-operative period are essentially valid also in the post-operative period. However, it is worth emphasising that the techniques used necessitate monitoring and a level of knowledge and specific skills that strongly suggest that the post-operative management of patients at risk is performed in appropriately equipped healthcare settings (recovery room, intensive care unit, sub-intensive care unit)^394,395.

Indeed, one aspect worth highlighting is that post-operative complications, including those due to poor tolerance of anaemia, if not promptly identified and treated adequately, can have a substantial impact on both the patient and the whole healthcare system, since they can be potentially disastrous for the former and associated with prolonged admissions, with increased costs, for the latter. In this regard, recent data suggest that these costs could be due not only to the short-term management of the patient, but also to the long-term management, because of repeated hospital admissions for recurrences and the development of a state of chronic disease^396–398.

In the case of patients with a reduced physiological reserve, which may cause a peri-operative reduction in tolerance of anaemia, it is recommended that post-operative observation is planned for an appropriate period (the duration of which varies according to the attendant clinician) in settings of variable intensity of care depending on the patients’ requirements. In these structures it should be possible to monitor and, if necessary, support the physiological variables involved in the process of adaptation to anaemia [1C].

It has been known for many years that there is a relationship between post-operative pain, stress and oxygen consumption³⁹⁹. Furthermore, acute post-operative pain, if not adequately treated, causes a series of reflexes that alter physiological homeostasis, such as the secretion of catabolic hormones, with significant haemodynamic, renal and immune repercussions. Finally, post-operative pain causes important respiratory alterations, inducing a restrictive syndrome. In particular the activity of the diaphragm and other respiratory muscles is reduced after abdominal surgery. The logical consequences of this are a change in the ventilation/perfusion ratio and a greater probability of developing hypercapnia, hypoxia, atelectasis and inflammatory lung phenomena^400,401. Consequently, pain is nowadays considered a complication of surgery, rather than a simple symptom, and does, therefore, require appropriate treatment according to well-defined protocols^402,403.

Epidural analgesia is a consolidated technique that, for several years, has been considered the gold standard in the management of post-operative pain, particularly in the elderly and in patients with impaired respiratory function^404–407. However, recent data show that the benefits of this technique are not as great as thought and that the advantages with regard to decreasing the incidence of cardiovascular and respiratory complications are probably limited to high-risk patients undergoing abdominal or thoracic surgery who receive thoracic epidural analgesia only with local anaesthetics. Furthermore, there is recent evidence showing that less invasive, regional analgesic techniques (paravertebral block for thoracotomy, femoral block for hip and knee surgery, local infiltrations - possibly continuous - of the wound, local analgesic infiltration in lower limb surgery) are as effective as epidural analgesia. For this reason, the number of indications for epidural analgesia seems to be decreasing^408–413. In orthopaedic limb surgery, the use of continuous peripheral nerve blocks has been shown to be as effective as peridural analgesia^414–417.

In order to optimise the treatment of pain, it is recommended that the level of post-operative pain (at rest and not) is evaluated periodically, using an international scales: a numerical rating scale, a visual analogue scale or a verbal rating scale. This information must be recorded in the patient’s clinical records [1A].

Given their greater efficacy, it is recommended that epidural techniques of analgesia are used rather than systemic opioids [1A].

In limb surgery, it is recommended that continuous peripheral nerve blocks are used [1A].