The Orthopaedic Surgery Transfusion Haemoglobin European Overview (OSTHEO) study prospectively analysed blood management data from 3996 patients undergoing elective orthopaedic surgery in Europe in 1999.
26 Data were collected from 225 centres in France, Germany, Greece, Italy, the Netherlands, and Spain. Overall, 69% of patients received transfusion, including 35% who received autologous transfusions only and 25% who received allogeneic blood transfusions (ABTs) only.
Around 45% of major orthopaedic surgery patients receive perioperative RBC transfusion.
5 Orthopaedic surgery is the leading surgical indication for transfusion, accounting for ~10% of all RBC units transfused.
27,28 Patients with preoperative anaemia are more likely to receive perioperative blood transfusions in the setting of orthopaedic,
4,9,10,14,19,26,29,30 and other types of surgery,
17,30 than non-anaemic patients. For example, in 1142 patients undergoing elective hip or knee surgery in the UK in 2000–1, 42.0% of anaemic patients received a transfusion compared with 21.3% of patients overall, and a preoperative Hb level of <11.0 g dl
−1 was strongly and independently associated with transfusion (odds ratio 13.92;
P<0.001).
10 In a prospective, observational Austrian study, the main predictors for allogeneic RBC transfusions were preoperative and lowest Hb and surgical blood loss.
30The use of ABTs to correct anaemia in surgical patients is not supported by evidence of benefit. Transfusions increase Hb levels in critically ill patients, but they do not appear to reliably improve tissue oxygenation.
31,32 ABTs are associated with a number of risk factors that can be divided into transfusion-transmitted infectious complications (e.g. HIV and hepatitis), immunological complications [e.g. immunomodulation resulting in postoperative infection, sepsis, antibody-mediated alloimmunization, graft-
vs-host disease (GVHD), haemolytic transfusion reactions, and allergic reactions], transfusion-related acute lung injury (TRALI; which has both immunological and non-immunological properties), and non-infectious non-immunological complications (e.g. acute lung injury, transfusion errors, non-haemolytic and haemolytic reactions, circulatory overload, and metabolic disturbances).
22,23Cohort studies have documented an increased risk of infectious complications, fluid overload, delay in wound healing, and prolonged hospital stay in transfused patients undergoing orthopaedic surgery.
29,33–36 Adverse transfusion outcomes have also been documented in critically ill patients, trauma patients, and cardiac or colorectal surgery.
37Observational studies have found higher postoperative mortality rates in transfused patients undergoing cardiac surgery compared with those who were not transfused, even after adjustment for other risk factors.
38 Randomized controlled trials (RCTs) in patients undergoing cardiac surgery have shown that white blood cell (WBC)-reduced blood products carry a lower mortality risk than non-WBC-reduced products.
23 However, this has not been observed after other types of surgery, and this issue is not resolved.
23TRALI, acute transfusion reactions, haemolytic transfusion reactions, transfusion-transmitted infections, circulatory overload, and (in the UK) transfusion-associated GVHD are the main causes of death and morbidity related to ABT.
39,40 TRALI has a presentation similar to acute respiratory distress syndrome, which has also been linked to transfusion.
41 Different models have been proposed for the pathogenesis of TRALI.
42,43 An antibody-mediated immune reaction, whereby antibodies in donated plasma activate recipient neutrophils within the lung, has been proposed.
43 Alternatively, the ‘two-event’ model proposes that factors present in transfused stored blood products trigger TRALI in patients in whom endothelial activation is already present due to an initial proinflammatory event, for example, infection, surgery, or trauma.
42 Transfusion-related immunomodulation proposes that deleterious proinflammatory effects result from a complex interplay between transfusion effects, genetic factors, intercurrent illnesses, and inflammatory mediators and effector cells.
44However, a causal relationship between transfusion and these serious complications is not established. Much of the evidence for these complications comes from retrospective, uncontrolled, non-randomized, observational cohort studies that may be subject to bias and confounding factors, and predominantly demonstrate correlations rather than causal relationships. Patients who are less well relative to healthy patients are more likely to receive transfusion. Thus, transfusion may be a surrogate for other causes of poor outcome. However, the association between allogeneic RBC transfusions and adverse outcome has been shown so consistently over the last three decades, that causation is highly likely.
45,46The storage of donated RBCs has enabled a more efficient use of blood supplies. However, allogeneic blood undergoes changes during prolonged storage (storage lesions), which may affect patient outcome. These changes include the accumulation of proinflammatory metabolic and breakdown products (e.g. lysophospholipids) that may be linked to TRALI, changes in cell shape, acidosis, membrane loss, haemolysis, increased rigidity, and stronger endothelial attachment.
44,47 Some studies have correlated the use of older, stored RBCs with an increased risk of complications and mortality in cardiac surgery and trauma patients.
48,49 However, the impact of storage on post-surgical outcome is a matter of debate.
44 The evidence supporting this association comes from observational studies that could be subject to bias and confounding factors. Some studies have reported that the storage age of RBCs was not independently associated with poorer outcomes,
50 while a recent meta-analysis concluded that the available data do not support an association between older RBCs and increased morbidity or mortality.
51The risks of transfusion are acknowledged by the American Association of Blood Banks, American Blood Commission, and American Red Cross.
52 The 2009 International Consensus Conference on Transfusion and Outcomes (ICCTO; Phoenix, AZ, USA) concluded that, based on available evidence and considering a number of scenarios in which transfusions are commonly given, ABT is not likely to improve patient outcome in most scenarios (but may even cause harm). The impact of ABT on patient outcomes remains uncertain and in need of further investigation, providing a rationale for more judicious use of ABT.
53,54Additional disadvantages with the use of transfusions relate to scarcity and cost of blood. The supply of blood for transfusions is limited by an ageing population and increasingly restrictive screening criteria. The chain of supply can also be overwhelmed by disaster events, potentially leading to blood shortages at the local level. While more blood may be donated in response to disaster events, the daily supply margin is limited and days during which there is a sudden increase in the use can disrupt the blood supply. While this may be a problem in developing nations, it may still affect first-world countries with established infrastructure.
22 The increasing scarcity of blood and measures to reduce the risks of infection transmission have increased the direct costs of transfusion.
55 The true costs of transfusion services are likely to have been underestimated owing to their complexity. A detailed study of the costs associated with all activities involved in providing transfusion services in surgical patients in four hospitals in the USA, Austria, and Switzerland recently estimated expenditures ranging from $1.62 to $6.03 million per hospital. The total costs were 3- to 5-fold higher than blood product acquisition costs alone.
56 Indirect costs have previously associated with the legal ramifications of contaminated blood supplies, and also the personal costs to affected donors and patients.
56–58Despite the risks, limited benefits, costs, and scarcity, a substantial proportion of RBC transfusions given today may not be appropriate or justified according to the report of the ICCTO panel and several reports on variable transfusion practices in otherwise comparable populations.
53 These findings support a more cautious transfusion strategy than the current system. A retrospective study in two US hospitals found a significant correlation between inadequate or suboptimal documentation and failure to justify transfusion, with 73% of inadequately documented transfusions not meeting hospital guidelines.
59 Inappropriate transfusion practices may endanger patients and waste resources. A recent RCT demonstrated that a liberal transfusion strategy (a higher Hb threshold of 10 g dl
−1) did not improve outcomes (death, inability to walk independently at a 60 day follow-up and in-hospital mordibity) compared with a restrictive strategy (Hb threshold of <8 g dl
−1), in a population of high-risk patients after hip-fracture surgery.
60Patient blood management
The clinical, logistic, and economic disadvantages of RBC transfusion have prompted recommendations for its restriction, particularly unnecessarily transfused stored blood
54—and an interest in new approaches.
5,23,24,37 The concept of PBM (or ‘blood conservation’) has been developed to promote ‘the appropriate provision and use of blood, its components and derivatives, and strategies to reduce or avoid the need for a blood transfusion’.
61 However, the concept has been developed with more emphasis on preventative measures and improving patient outcome. PBM relies on three key strategies to achieve its goals: optimize the patient's own RBC mass, minimize blood loss, and harness and optimize physiological tolerance of anaemia.
24,27 Therefore, PBM requires a multidisciplinary, multimodal, individualized strategy for avoiding and controlling blood loss, and to systematically identify, evaluate, and manage anaemia (Fig. ).
62
Perioperative measures Before operation, PBM involves a careful assessment of bleeding risk and anaemia well in advance of surgery (e.g. 30 days) to allow full evaluation and correction of anaemia. This has been specifically recommended in patients undergoing orthopaedic surgery.
63 The patient's own blood should be conserved by restricting blood drawn for tests and by restricting the use of antiplatelet and anticoagulant agents to situations where these drugs are indicated. Pharmacological and mechanical venous thromboembolism prophylaxis measures are both widely used in patients undergoing THR and TKR, with recommendations of better compliance with established evidence-based guidelines.
64 Autologous preoperative donation was once promoted to decrease the need for ABT. However, predonated blood is also subject to storage lesion, and is labour intensive, expensive, and inefficient, with almost half of predonated autologous units not used.
29 Some patients may not be able to predonate blood due to comorbidities and current or potential anaemia.
3 There are possible deleterious effects of blood storage for units predonated weeks ahead of surgery.
24It has been suggested that Hb levels before elective orthopaedic surgery should be within the normal range defined by the World Health Organization (WHO; ≥12 g dl
−1 in women; ≥13 g dl
−1 in men).
61 Iron supplementation should be used to correct iron deficiency. Oral iron is effective in reducing the need for transfusion before orthopaedic surgery.
65,66 Erythropoiesis-stimulating agents (ESAs) may need to be added in patients with preoperative anaemia.
66 Data from prospective, non-randomized case series suggest that i.v. iron can correct iron deficiency anaemia before elective orthopaedic surgery
67 and reduce rates of ABT, postoperative infection, and mortality in patients with hip fracture, compared with historical controls.
68,69 Another case series found that the perioperative use of i.v. iron in conjunction with ESA therapy in anaemic patients with hip fracture (Hb levels of <13 g dl
−1) reduced the proportion of patients requiring transfusion, the number of units transfused, and the rate of postoperative infections compared with a parallel control group. There was no difference in the 30 day mortality or the mean duration of hospitalization.
70 Previous recommendations that patients with preoperative anaemia due to iron deficiency or chronic disease may receive preoperative treatment with i.v. or oral iron, depending on the timing of surgery, the patient's tolerance of oral iron, and iron status,
71,72 are based on low-to-moderate-quality evidence. Further RCTs of i.v. iron are currently underway.
ESA therapy may be used within PBM strategies. ESA increases Hb and reduces the need for transfusion in patients undergoing orthopaedic and cardiac surgeries.
3,5,24,62,73 It has been suggested that ESA should be used in anaemic patients in whom iron deficiency anaemia has been ruled out or corrected.
63 Iron supplementation plus ESA corrected normocytic anaemia due to chronic disease in preoperative orthopaedic surgery patients;
10 this can lead to increased transfusion avoidance.
10,74 Patients should receive iron supplementation (preferably i.v.) throughout the use of ESA to avoid functional iron deficiency.
63,73 A recent study evaluated a blood conservation protocol involving a restrictive transfusion trigger (Hb <8 g dl
−1) and perioperative administration of i.v. iron with (
n=115) or without (
n=81) ESA in patients undergoing hip fracture repair.
75 Patients who received ESA therapy together with i.v. iron had a significantly lower allogeneic transfusion rate (60%
vs 42%,
P=0.013) and higher postoperative Hb than those who received iron alone, but there was no difference in postoperative complications or 30 day mortality rate. The combination of a restrictive Hb trigger, i.v. iron, and ESA reduced the use of ABT in patients (
n=139) undergoing TKR surgery.
76 In this study, patients who received unwashed shed blood after operation (if their preoperative Hb was <13.0 g dl
−1), in addition to ESA and i.v. iron, showed a significantly reduced hospital stay compared with those who received i.v. iron and ESA alone (
P<0.05). A more conservative use has been advocated because of concerns about side-effects of ESAs. After regulatory changes in the light of safety concerns, the use of ESA has significantly decreased in cancer centres, but no effect was seen on transfusions.
77 These changes in usage may impact upon individual PBM strategies, as alternatives to ESA become important.
Intraoperative measures to prevent blood loss include patient positioning and the use of electrocautery, tourniquets, vasoconstrictors, and topical or systemic (e.g. tranexamic acid epsilon aminocaproic acid, aprotinin, and desmopressin) and local haemostatic agents.
22,24,25,78–80 Acute normovolaemic haemodilution, where blood is collected of the operation for potential transfusion after operation and replaced by a crystalloid or colloid solution, can be used to lessen the loss of RBCs and clotting factors during bleeding. Reviews of the use of colloids and crystalloids have shown that while there is no difference in overall patient survival between the two forms, specific products are best suited to certain situations.
81,82 Autologous blood cell salvage is particularly useful for procedures involving massive blood loss and in patients who object to the use of ABT. Postoperative measures include close monitoring of bleeding and anaemia and the continuing various measures described above.
24,27 Individual blood conservation approaches may each save ~1–2 units of blood, and when multiple approaches are used together, 2 units of blood can usually be saved.
24,25 Implementation and effectiveness of PBM programmes A cluster randomization study in Canadian hospitals demonstrated that a multicomponent blood conservation algorithm reduced the use of ABT and increased ESA use in orthopaedic surgery patients.
83 Similarly, in France, a blood conservation algorithm introduced in one orthopaedic unit in 2005 changed local practice, reducing the overall use of transfusions by 56% and wastage of autologous blood units by 50% (
P=0.002), and increasing the rate of ESA usage (from 6.6 to 17.3%;
P<0.05). These changes were associated with a 50% reduction in hospital costs that offset the costs associated with increased ESA usage, resulting in no significant change in overall costs.
84There is increasing awareness of the need to integrate PBM within routine surgical care. PBM has been successfully implemented in some centres in the USA.
25 The Government of Western Australia has implemented PBM state-wide as the standard of care.
80 The Australian Red Cross has also issued guidance on measures to reduce the need for ABT, including in surgical patients.
85 
This article has been 
