Evidence and triggers for the transfusion of blood and blood

products
A. Shah,1 S. J. Stanworth2 and S. McKechnie3
1 Specialist Registrar, Adult Intensive Care Unit, 2 Consultant, Department of Haematology, 3
Consultant, Department
of Anaesthesia and Intensive Care, John Radcliffe Hospital, Oxford, UK
Summary
Allogeneic red cell transfusion is a commonly used treatment to improve the oxygen carrying
capacity of blood during the peri-operative period. Increasing arterial oxygen content by
increasing haemoglobin does not necessarily increase tissue oxygen delivery or uptake. Although
the evidence-base for red cell transfusion practice is incomplete, randomised studies across a
range of clinical settings, including surgery, consistently support the restrictive use of red cells,
with no evidence of benefit for maintaining patients at higher haemoglobin thresholds (liberal
strategy). A recent meta-analysis of 7593 patients concluded that a restrictive transfusion strategy
was associated with a reduced risk of healthcare-associated infections (pneumonia, mediastinitis,
wound infection, sepsis) when compared with a liberal transfusion strategy. The degree to which
the optimal haemoglobin concentration or transfusion trigger should be modified for patients
with additional specific risk factors (e.g. ischaemic heart disease), remains less clear and requires
further research. Although most clinical practice guidelines recommend restrictive
use of red cells, and many blood transfusion services have seen marked falls in overall usage of
red cells, the use of other blood components such as fresh frozen plasma, platelets, and
cryoprecipitate has risen. In clinical practice, administration of fresh frozen plasma is usually
guided by laboratory tests of coagulation, mainly prothrombin time, international normalised
ratio and activated partial thromboplastin time, but the predictive value of these tests to predict
bleeding is poor.

Introduction
This article focuses on the transfusion of red cells and other blood components (fresh frozen
plasma (FFP), platelets, cryoprecipitate and fibrinogen) in the management of stable non-
bleeding patients, with specific emphasis in the peri-operative setting. The management
of major haemorrhage (traumatic and non-traumatic) is covered elsewhere in this supplement. In
the main part of the article, we discuss the rationale, use and risks of red cell transfusion and in
the second part, we discuss the use of prophylactic blood components.

Methods
We searched for systematic reviews and randomized controlled trials using the Transfusion
Evidence Library (www.transfusionevidencelibrary.com) produced by the NHS Blood and
Transplant Systematic Reviews Initiative. We also reviewed recently published
guidelines and searched for related trials on www.clinicaltrials.gov and the International
Standard Randomised Controlled Trial Number (ISRCTN)
register.

Allogeneic red cell transfusion

Allogeneic red cell transfusion is a commonly used treatment to correct
anaemia and improve the oxygen
carrying capacity of blood during the peri-operative period and in critically ill
patients. Approximately, 85 million red cell units are transfused worldwide
annually [1]. Approximately, 2 million units of red cells were issued across
the UK in 2011 [2], with surgical patients receiving approximately 40% of
transfused allogeneic blood [3]. The last few years have seen a marked fall in
overall usage of red cells for transfusion. The reasons for this are
multifactorial, and include multiple national and local educational initiatives
(now often covered by the term patient blood management, described
below) aimed at increasing awareness of the risks of transfusion, improved
surgical techniques, and the need to consider alternatives, coupled with a
lack of evidence to support maintaining higher haemoglobin concentrations
by transfusion. Aerobic metabolism is dependent on adequate tissue oxygen
delivery D_ O2), which in turn is dependent on cardiac output (CO) and
arterial oxygen content (CaO2).
DO2 CO_CaO2
CO_1:39_Hb__SaO2 PaO2_0:003
Under normal conditions, the vast majority of oxygen is carried by haemoglobin (Hb). The
haemoglobin concentration is therefore critical in determining arterial oxygen content, with a fall
in Hb resulting in reduced arterial oxygen content and, unless compensated, a fall in DO2. Case
reports highlight survival with Hb as low as 20 g.l_1 [4] and experimental studies have shown
that young healthy adults can compensate for Hb as low as 40 g.l_1 without significant
deleterious effects if the circulating blood volume is maintained with fluids [5].
However, acutely unwell and elderly patients, especially those with comorbidities, are less likely
to tolerate such low levels. In the context of inadequate oxygen delivery, a clinician
must decide between increasing the cardiac output (fluids and/or inotropes) or increasing the
arterial oxygen content (correction of hypoxaemia and/or red cell transfusion). Global markers of
inadequate oxygen delivery, such as low central, or mixed, venous saturations and lactic acidosis,
may be useful but lack specificity. Haemoglobin concentration, in the absence of other specific
clinical and laboratory tests, remains the most widely used trigger for blood transfusion in
clinical practice.

Benefit vs risks of allogeneic red cell transfusion

The following points provide a summary of the issues to be considered when assessing the
potential benefit of managing anaemia with red cell transfusion:
1 Although the evidence-base for red cell transfusion practice is incomplete, randomised studies
consistently support the restrictive use of red cells in most settings, with no evidence of benefit
for maintaining patients at higher haemoglobin thresholds (liberal strategy).
2 The degree to which the optimal Hb or transfusion trigger should be modified for patients with
additional specific risk factors (e.g. ischaemic heart disease) remains less clear. It seems sensible
to modify the decision to transfuse in the presence of symptoms such as chest pain, heart failure,
or tachycardia unresponsive to fluid resuscitation [6].
3 Guidelines suggest there is no evidence to transfuse when the Hb is > 100 g. l_1.
4 Current UK practice allows for the storage of red cells to a maximum of 35 days [7]. It is well
recognised that biochemical and physiological changes occur in stored red cells but the clinical
consequences of red cell storage lesions remains unclear. Randomised studies assessing the
impact of the storage age of blood on patient outcome are ongoing.
5 Increasing arterial oxygen content by increasing Hb does not necessarily increase tissue
oxygen delivery or uptake. Transfused cells have altered rheological properties and increasing
Hb increases the haematocrit (Hct) and blood viscosity. This can reduce blood flow through the
microcirculation.

Red cell transfusions rarely increase oxygen uptake V_ O2) except in extreme situations where
oxygen uptake is directly dependent on oxygen delivery such as severe circulatory shock [8, 9].
Blood loss of up to 30% can be adequately treated with crystalloids or colloids [3].
Observational studies suggest that patients who receive allogeneic red cell transfusion are at
increased risk of mortality, ischaemic complications, delayed wound healing, and increased
length of stay [1012].
However, it is not clear whether this link is causal orassociative, as the findings of most
observational studies are confounded by the fact that sicker patients, with
associated worse clinical outcomes receive more blood. A number of randomised studies
published in recent years have started to inform the riskbenefit ratio of red cell transfusion.
These trials have typically randomly allocated patients to a restrictive or liberal transfusion
strategy. A Cochrane systematic review [13] identifying 19 such randomised studies involving
6264 patients across a variety of clinical settings, concluded that a Hb threshold (or transfusion
trigger) of 70 or 80 g.l_1 is associated with fewer red cell transfusions without adverse
association with either mortality, cardiac morbidity, functional recovery or hospital
length of stay. The authors do comment on the significant heterogeneity between studies in this
review [14]. The risks of allogeneic red cell transfusion include those common to all blood
components, such as transfusion- transmitted infection or errors in the processing and
administration, and those specific to red cells, such as storage age lesion. The Annual Serious
Hazards of Transfusion (SHOT) reports provide insight into the risks of transfusions in UK
practice. More than half of the cases reported were due to preventable mistakes, with the leading
error being incorrect blood component transfusion. Acute transfusion reactions
were the commonest cause of pathological and unpredictable incidents [2]. These risks are
summarised in the Table 1.
There is awareness of the potential risks of transfusion- associated immunomodulation, for
example in cancer recurrence [15]. A Cochrane systematic review found a moderate association
between colorectal cancer recurrence and allogeneic red cell transfusion an association that
increases when larger volumes of blood are given. This review concluded that red cell
transfusion should be restricted in patients undergoing colorectal surgery with curative intent
[16]. A very recent meta-analysis of 7593 patients concluded that a restrictive transfusion
strategy was associated with a reduced risk of healthcare-associated infections (pneumonia,
mediastinitis, wound infection, sepsis) when compared with a liberal transfusion strategy [17].

Peri-operative anaemia
The World Health Organization (WHO) defines anaemia as Hb < 130 g.l_1 in men, < 120 g.l_1
in nonpregnant women and < 110 g.l_1 in pregnant women [18]. Approximately, 1.26 billion
people worldwide are affected by anaemia [18]. The prevalence of pre-operative anaemia varies
widely from 5% to 76% [19], and has been linked to increased postoperative mortality,
morbidity, length of stay and decreased quality of life [1921]. Certain subgroups,
for example, those with colorectal cancer, have a higher prevalence of pre-operative anaemia
ranging from 39% (Dukes A) to 76% (Dukes D) [22]. Orthopaedic surgery is a major consumer
of red cells and two studies in the UK showed a prevalence of approximately 2030% in patients
undergoing primary/revision hip or knee arthroplasty [23, 24]. Figures on the prevalence of intra-
operative anaemia are sparse. Perhaps unsurprisingly, given the

effects of blood loss and haemodilution, postoperative anaemia appears more common than pre-
operative anaemia, with a reported range of 5187% in orthopaedic surgery [25]. In critical care,
anaemia affects approximately 6080% of patients, with the Hb dropping by about 5 g .l_1 per
day in most studies [26]. A large multicentre study [26] showed the mean Hb in patients admitted
to western European ICUs following elective surgery was 110 g.l_1. In clinical practice, anaemia
is the main predictor of red cell transfusion [27], and this is independently
associated with increased morbidity and mortality as mentioned before [1012, 28].
Anaemia arises mainly due to a reduction in red cell production, increased red cell destruction,
blood loss (acute or chronic), or a combination [20]. Preoperative anaemia in elective patients is
frequently chronic. It may be due to iron deficiency, chronic inflammation, and related to
malnourishment and/or alcohol abuse. Patients may also have vitamin B12 or folate deficiencies
[18, 29]. A retrospective observational study evaluating the characteristics of pre-operative
anaemia in patients undergoing elective orthopaedic joint arthroplasty found that 64% of patients
had a normochromic normocytic anaemia, 23% a hypochromic anaemia and 12% had other types
of anaemia [30]. Acute pre-operative anaemia is uncommon in clinically stable elective cases,
but is frequently associated with significant haemorrhage as occurs in obstetrics and following
trauma [20]. Intra-operative anaemia can be caused by acute blood loss and/or haemodilution as
a result of fluid administration. Postoperative anaemia also frequently results from blood loss
and/or haemodilution. In addition, surgery triggers an acute inflammatory response, which
induces a pattern of anaemia similar to that seen in critically ill patients reduced iron
availability due to hepcidin-mediated inhibition of iron absorption from the duodenum and
release from macrophages, diminished erythropoietin response, and impaired proliferation
and differentiation of erythroid progenitors in the bone marrow [23, 31, 32]. In the peri-operative
setting, patient blood management is a relatively recent concept based on three pillars (i)
detection and treatment of pre-operative anaemia; (ii) reduction in peri-operative blood loss;
(iii) harnessing the patient-specific physiological reserve of anaemia [10]. These strategies are
discussed elsewhere in this journal. However, even with these initiatives, allogeneic red cell
transfusion remains the commonest and (arguably) most readily available treatment
to increase Hb [20]. Pre-operative transfusion The detailed management of pre-operative
anaemia is discussed in detail elsewhere. Before elective surgery, where there is time (days to
weeks) to correct anaemia, iron, nutritional supplements and erythropoietin stimulating agents,
may all have a role. There is little role for these interventions in the acute correction of anaemia.
The WHOs Clinical Use of Blood Handbook states that there is rarely a justification for the use
of preoperative blood transfusion simply to facilitate elective surgery [33]. One study found the
primary predictor of pre-operative blood transfusion to be an Hb of < 100 g.l_1 [34]. The British
Committee for Standards in Haematology (BCSH) Guidelines state that there is no case for
transfusion back to a normal haemoglobin level either before or after surgery, and to avoid
transfusion when the Hb is above 10 g.dl_1 [35]. The need for transfusion in acute anaemia
secondary to blood loss should be based on estimation of lost circulating volume, risk of ongoing
bleeding, Hb, and patient susceptibility to complications of inadequate oxygenation. Patients
with sickle cell disease often receive preoperative red cell transfusions to dilute sickle cells and
preserve oxygen carrying capacity, as the peri-operative complication rate is high [36]. A recent
Cochrane systematic review indicated that conservative therapy (transfusing to maintain Hb of
100 g.l_1 or no transfusion), may be as effective as aggressive therapy (transfusing to decrease
the haemoglobin S level to < 30%) in some patient groups [37]. Intra- and postoperative
allogeneic transfusion Intra-operative goal-directed fluid therapy, to optimize stroke volume and
cardiac output, has become established practice in major elective surgery. However, most
algorithms for intra-operative goal-directed

fluid therapy are based on crystalloid and/or colloid administration. The role of red cell
transfusion in intra-operative goal-directed therapy has not been extensively studied. A recent
study investigating the effect of a perioperative cardiac output-guided haemodynamic algorithm
on outcomes in patients undergoing major gastrointestinal surgery (OPTIMISE) [29], included
maintaining an Hb of > 80 g.l_1 as part of the algorithm. Use of the algorithm did not improve a
composite outcome of complications and 30-day mortality compared to usual care in this study,
but inclusion of the OPTIMISE results in a meta-analysis did suggest a significant reduction in
complications with the intervention[38]. Early goal-directed therapy in severe sepsis and septic
shock has traditionally included targeting Hct > 30% (Hb 100 g.l_1) as part of a package of
formalized care within the first 6 h of presentation [39]. However, these recommendations are
largely drawn from one single-centre study [39, 40]. The generalisability of formalised goal-
directed therapy in sepsis has recently been tested in a North American multicentre study
(ProCESS) in which the transfusion trigger was Hct < 30% in the protocol-based goal-directed
therapy group compared to a transfusion trigger of Hb < 75 g.l_1 in the protocol-based standard
therapy group [41]. There was no difference in the primary (60- day mortality) or secondary
outcomes in this study. The results of similar studies from the UK (ProMISE) [42] and Australia
(ARISE) are awaited [43]. Several studies have compared different transfusion triggers in
surgical sub-specialties. These are summarized in Table 2. Trials regarding acute blood loss
secondary to trauma have been covered elsewhere in this series. The results from these studies
underpin guidelines for transfusion management. For example, The American Association of
Blood Banks guidelines [6] recommend that red cell transfusion should be considered in
postoperative surgical patients, if the Hb is < 80 g.l_1 or in the presence of symptoms such as
chest pain, heart failure or orthostatic hypotension or tachycardia unresponsive to fluid
resuscitation [6]. Stable adult critical care patients should be considered for red celltransfusion if
the Hb < 70 g.l_1.
Use of FFP, platelets, cryoprecipitate and fibrinogen concentrate
The use of FFP, platelets, cryoprecipitate and fibrinogen in the management of bleeding is
discussed elsewhere. This section focuses on the prophylactic use of these blood components.
Fresh frozen plasma Fresh frozen plasma is human donor plasma frozen within 8 h of collection.
After thawing, FFP contains almost normal levels of plasma coagulation factors, acute phase
reactants, immunoglobulins and albumin [44]. Current UK and European Union guidelines
require that only Factor VIII levels need to be quality controlled [45]. The usual recommended
dose of FFP is 15 ml.kg_1. In clinical practice, administration of FFP is usually guided by
laboratory tests of coagulation, mainly prothrombin time (PT), international normalised ratio
(INR) and activated partial thromboplastin time (APTT). These tests are based on the traditional
coagulation model consisting of intrinsic and extrinsic pathways. This model has now largely
been superseded by a more complex cell-based model, in which haemostasis occurs on multiple
cellular surfaces in three steps initiation, amplification and propagation [46]. A detailed review
on the limitations of these tests is beyond the scope of this article and can be
found elsewhere [47], but important points include:
PT and APTT vary in sensitivity for reduced levels of coagulation factors. Elevated PT/INR
results do not necessarily correlate with increased risk of bleeding [48].
Patients with an elevated INR (1.31.9) can still have adequate concentrations of coagulation
factors to support normal clotting [49].
These tests take time to perform which limits their usefulness in situations of acute
haemorrhage.
Results are dependent on the laboratory processes and reagents used. Thromboelastography
(TEG_, Haemonetics Corp, Braintree, MA, USA) and rotational thromboelastometry (ROTEM_,
TEM International, Munich, Germany) are point-of-care testing devices that provide a dynamic
display of the viscoelastic properties of blood

from the initial phase of fibrin formation to clot formation, retraction and eventually fibrinolysis
[50]. An important advantage of these point-of-care devices is the ability to provide rapid
assessment of coagulation status, which may be of particular value when managing patients with
bleeding (see other reviews in this supplement). These assays have been more extensively used
and evaluated in the setting of cardiac surgery and trauma, with some studies indicating that
algorithms incorporating their use reduce the need for blood transfusion. Evidence that the use of
these technologies improves clinical outcomes is less clear, and is the subject of ongoing review
by the National Institute for Health and Care Excellence (NICE). Approximately, 280 000 units
of FFP are issued annually in the UK to adults (93%) and children (7%) [51]. A recent systematic
review found no evidence of benefit for prophylactic or therapeutic FFP use across a
range of settings including use in liver disease, cardiac surgery, warfarin anticoagulation
reversal, thrombotic thrombocytopenic purpura, burns, shock and head injury [52]. However,
despite 80 randomised controlled trials on this topic, the methodological quality of many trials is
poor [52]. A large observational study of coagulopathy in ICU (ISOC-1) [53] illustrates the
clinical uncertainties surrounding the use of FFP. In this study of just under 2000 ICU
admissions, one third of patients had a prolonged PT and a third of these patients received
FFP. Around 50% of patients received FFP in the absence of evidence of clinical bleeding, and
many had only mild or moderate prolongation of PT. The dose of FFP administered was also
highly variable in this study. Given current evidence, a more appropriate plasma transfusion
strategy may be one that emphasizes the therapeutic use of FFP in bleeding patients. The
role, if any, of prophylactic plasma transfusion in nonbleeding patients with abnormalities of
standard coagulation tests, which have limited value in predicting bleeding, remains unclear.
Table 3 summarises current recommendations for the prophylactic use of FFP. Fresh frozen
plasma use has increased without clear evidence of benefit. In view of this, clinicians must
consider the risks associated with FFP transfusion. These include Transfusion-Related Acute
Lung Injury (TRALI), Transfusion-Associated Circulatory Overload and allergic/anaphylactic
reactions. Less common hazards include transmission of infection and haemolytic
and non-haemolytic reactions [54].
Platelets
Platelets play an important role in primary haemostasis following vascular or tissue injury.
Platelet activation occurs when the glycoprotein receptor found on platelet surfaces binds to von
Willebrand factor on endothelium. This results in secretion of several procoagulant
factors such as calcium, serotonin, ADP and fibrinogen, which in turn promote aggregation by
glycoprotein IIb/IIIa receptors to form a platelet plug. Abnormalities of platelets can be broadly
divided into quantitative (platelet number) and qualitative (platelet function). Common causes
during the perioperative period are listed in Table 4. Normal platelet count ranges between 150
and 450 9 109.l_1. Thrombocytopenia is commonly defined as a platelet count < 150 9 109.l_1
with mild thrombocytopenia being between 70 and 150 9 109.l_1 and severe thrombocytopenia
being < 50 9 109.l_1. Clinical guidelines typically use platelet number to guide platelet
transfusion; less regard is given to assessment of platelet function [54]. In the UK, platelets are
produced from centrifugation of whole blood or by apheresis [55]. Both methods generate the
same number of platelets and are considered therapeutically equivalent. A unit of platelets also
contains plasma (or platelet additive solutions) to maintain platelet function during storage.
Pooled platelet concentrates can be stored in a closed system at 22 C with continuous gentle
agitation for up to 5 days. One unit of apheresis platelets would be expected to increase the
platelet count in an average adult by 2040 9 109.l_1 [56]. Observational data from critical care
patients shows a median increment of 15 9 109.l_1 following one platelet transfusion
[57]. The use of platelets in the UK has steadily been increasing from approximately 210 000
U in 2001 to 260 000 U in 2011/12 [58]. Indications for platelet transfusion can be broadly
classified into:
1 Prophylactic:
Routine use in non-bleeding patients with
thrombocytopenia
Presence of risk factors for bleeding, e.g. sepsis, lines
are based on absolute platelet number. Much of the evidence for prophylactic transfusion in non-
bleeding patients has come from haematology patients with chemotherapy-associated
thrombocytopenia [59]. Randomised controlled trials in this cohort have shown that a transfusion
trigger of 10 9 109.l_1 confers no increased bleeding risk when compared to trigger
of 20 9 109.l_1 [59]. There is a lack of evidence to guide platelet transfusions to cover
invasive/surgical procedures, with guidelines based mainly on expert opinion [55]. There does
not appear to be a role for the prophylactic transfusion of platelets in patients on
cardiopulmonary bypass, with transfusion reserved for bleeding patients who have had a surgical
cause excluded[55]. The use of prophylactic platelet transfusion is not recommended in patients
with heparin-induced thrombocytopenia or idiopathic thrombocytopenic purpura and is
contraindicated in patients with thrombotic thrombocytopenic purpura [55]. The potential
benefits of platelet transfusion must be balanced against the risks. Non-haemolytic febrile
reactions and mild allergic reactions are common with an estimated incidence of 2% and 4%,
respectively. Anaphylaxis occurs rarely (1:20 000 to 1:50 000 of transfusions), but accounts for
approximately 40% of the serious adverse events reported. Platelets are more commonly
implicated in TRALI than red cells [60]. As with all blood components, there is a risk of
transfusion-transmitted infection. The rate of viral transmission human immunodeficiency
virus, hepatitis B, and hepatitis C appears extremely low. Transfusion- transmitted bacterial
infection has an estimated incidence of 1:10 000 platelet transfusions. The risk of bacterial
contamination is higher than for other blood products because, unlike other blood components,
platelets are processed and stored at room temperature. Bacterial screening has been introduced
to reduce the risk of transfusion-transmitted bacterial infection. Cryoprecipitate Current available
concentrated sources for fibrinogen replacement are cryoprecipitate or fibrinogen concentrate.
Cryoprecipitate is an allogeneic blood product prepared from human plasma. Cryoprecipitate is
derived from controlled thawing of FFP at 16 C to precipitate higher molecular weight proteins
such as Factor VIII, von Willebrands Factor, fibronectin, Factor XIII and fibrinogen [44]. In the
absence of a concentrated source, large volumes of FFP would be required for adequate
fibrinogen supplementation, thus exposing the patient to fluid overload and TRALI.
Cryoprecipitate use was first described in the 1960s for treatment of patients with factor VIII
deficiency. Now, the most common use of cryoprecipitate is fibrinogen replacement in patients
with acquired hypofibrinogenaemia and bleeding. Despite almost 50 years of use, evidence of
efficacy is very limited. Assessment of the predictive value of differing fibrinogen concentrations
has only been described in a few studies of major bleeding after trauma and surgery. It has been
reported in obstetric patients that a low plasma fibrinogen concentration
is associated with increased bleeding [61]. A Cochrane review evaluating the effectiveness of
fibrinogen concentrate for bleeding patients found six small trials in elective surgery, all of low
quality and underpowered for mortality although reporting a reduction in the incidence of
allogeneic transfusions (RR 0.47, 95% CI 0.320.72) [62]. One trial [63] assessed the efficacy of
fibrinogen concentrate therapy in bleeding patients undergoing radical cystectomy that
received intra-operative fluid replacement with colloid. The other [64] was a pilot randomised
controlled trial assessing prophylactic infusion of fibrinogen concentrate in patients before
cardiac surgery. Both studies were of small sample size, enrolling 20 patients each.
Cryoprecipitate should be reserved for use in patients with documented dysfibrinogenaemia or
hypofibrinogenaemia secondary to major haemorrhage, massive transfusion, disseminated
intravascular coagulopathy or before an invasive procedure. Guidelines recommend transfusion
of cryoprecipitate if the fibrinogen level is < 1.5 g.dl_1 with clinical evidence of bleeding,
although there is no high-level evidence to support any threshold [65]. As it is not a source of all
coagulation factors, it should not be used routinely as replacement therapy in patients with global
coagulopathy, e.g. liver disease. Approximately, 10 single bags of cryoprecipitate derived from
whole blood are needed to raise the plasma fibrinogen level by 1 g.dl_1 in an
adult [65]. A UK regional cross-sectional study showed that the primary indication for
cryoprecipitate transfusion was haemorrhage (81% of cases), followed by prophylaxis (17%)
[66]. Prophylactic use commonly occurred in the settings of liver failure, haematology/oncology
and critical care. The study also highlighted important issues such as the wide variation in the
doses of cryoprecipitate transfused and a lack of response of fibrinogen levels to cryoprecipitate.

Conclusion
Allogeneic red cell transfusion is a very common intervention in the peri-operative period. The
commonest trigger used to guide red cell transfusion remains the Hb. In most stable, non-
bleeding patients, a restrictive transfusion strategy with a typical transfusion trigger of 70 g.l_1,
appears to be safe and non-inferior to more liberal transfusion strategies. Larger, welldesigned
clinical trials are needed in high-risk patients such as those with acute coronary syndrome.
Although current guidelines frequently reference PT in recommending when to transfuse FFP to
correctcoagulopathy or reduce bleeding risk, there are uncertaintiesabout the value of standard
coagulation tests in predicting clinical bleeding risk and a lack of evidence of benefit for the
prophylactic or therapeutic administration of FFP in non-bleeding patients. Guidelines for the
prophylactic administration of platelets are largely based on platelet count, rather than function,
and are drawn from observational studies in haematological malignancy patients and expert
opinion. Cryoprecipitate transfusion should be considered in patients with bleeding, and based on
a fibrinogen concentration threshold of < 1.5 g.dl_1.

Competing interests
No external funding and no competing interests declared.