Review

Transfusion reactions: prevention, diagnosis, and treatment

Meghan Delaney, Silvano Wendel, Rachel S Bercovitz, Joan Cid, Claudia Cohn, Nancy M Dunbar, Torunn O Apelseth, Mark Popovsky,
Simon J Stanworth, Alan Tinmouth, Leo Van De Watering, Jonathan H Waters, Mark Yazer, Alyssa Ziman, for the Biomedical Excellence
for Safer Transfusion (BEST) Collaborative

Blood transfusion is one of the most common procedures in patients in hospital so it is imperative that clinicians are Lancet 2016; 388: 282536
knowledgeable about appropriate blood product administration, as well as the signs, symptoms, and management of Published Online
transfusion reactions. In this Review, we, an international panel, provide a synopsis of the pathophysiology, treatment, April 12, 2016
http://dx.doi.org/10.1016/
and management of each diagnostic category of transfusion reaction using evidence-based recommendations
S0140-6736(15)01313-6
whenever available.
Bloodworks NW, Seattle, WA,
USA (M Delaney DO); University
Introduction this Review do not have available published evidence of Washington, Department of
Blood transfusions are one of the most common (gure 1). In these instances, we provide recommendations Laboratory Medicine, Seattle,
WA, USA (M Delaney); Hospital
procedures for patients in the hospital and are associated and no grade is given. We provided some published
Sirio Libanes Blood Bank,
with substantial risks and cost; therefore health-care haemovigilance reports as references to guide the reader Sao Paulo, Brazil (S Wendel MD);
providers need to understand the hazards related to to additional content, but these are not used for evidence- BloodCenter of Wisconsin,
blood product administration.1 Although awareness is based recommendations. Diverse sources of data exist to Milwaukee, WI, USA
(R S Bercovitz MD); Department
increasing of the clinical ecacy of restrictive dene rates of specic transfusion reaction categories;
of Hemotherapy and
transfusion thresholds in some settingssuch that however, these reactions might be under-reported, and Hemostasis, CDB, IDIBAPS,
providers are being prompted to consider alternatives to aected by hospital factors and by the patients underlying Hospital Clnic, UB, Barcelona,
transfusion and make treatment decisions to avoid disease.3 Medication doses, when provided, are noted in Spain (J Cid MD); Department of
Laboratory Medicine and
unnecessary transfusionstransfusions are still an the appendix only. Transfusion might produce other
Pathology, University of
essential component of care in certain patient adverse eects, such as transfusion-related immuno- Minnesota, Minneapolis, MN,
populations.2 Transfusion reactions are the most modulation or viral infections, which are not usually USA (C Cohn MD); Department
frequent adverse event associated with the administration classied as transfusion reactions and, therefore, we do of Pathology and Department
of Medicine, Dartmouth-
of blood products, occurring in up to one in not include them in this Review. We also do not include
Hitchcock Medical Center,
100 transfusions (table 1). A transfusion reaction can studies on plasma derivatives. Lebanon, New Hampshire, USA
lead to severe discomfort for the patient and extra cost (N M Dunbar MD); Laboratory of
burden to the health-care system.35 Although rare, General management of transfusion reactions Clinical Biochemistry and
Department of Immunology
reactions can be fatal, with transfusion of about one in Transfusion reactions are usually reported to the physician and Transfusion Medicine,
200 000420 000 units associated with death.6 Given the by the nurse administering the blood product and often Haukeland University Hospital,
diversity of risks, clinicians should have accessible cause a change in vital signs or a new symptom.9 The Bergen, Norway
information about the nature, denitions, and algorithm summarises the initial clinical assessment of a (T O Apelseth MD); Haemonetics
Corporation, Braintree, MA,
management of transfusion-related adverse events. patient having a transfusion reaction (gure 2). Depending USA (M Popovsky MD); NHS
on the severity, the main treatment strategy for all reaction Blood and Transplant/Oxford
Review design and methods types is to stop the transfusion and keep the intravenous University Hospitals NHS Trust,
In this Review, we aim to provide a description of each line open with normal isotonic saline; start supportive John Radcliffe Hospital, Oxford,
UK (S J Stanworth FRCP);
clinical entity, as well as treatment and prevention care to address the patients cardiac, respiratory, and renal Radcliffe Department of
guidelines based on published work, whenever available, functions as necessary; and provide symptomatic therapy. Medicine, University of Oxford,
and expert advice. In the appendix we oer a detailed The blood product labelling and patient identication Oxford, UK (S J Stanworth);
guide for diagnostic, treatment, and management should be rechecked to conrm that the patient received Department of Medicine and
Department of Laboratory
principles in a single-page format for each category to their intended product and the reaction should be reported Medicine & Pathology,
provide a more extensive and detailed description that to the blood transfusion laboratory for additional testing.10 University of Ottawa, Ottawa,
could be used at the patients bedside. These universal procedures should be done in all ON, Canada (A Tinmouth MD);
We derived diagnostic categories for transfusion University of Ottawa Centre for
transfusion reactions, irrespective of the type of reaction.
Transfusion Research, Ottawa
reactions from denitions from the US National Hospital Research Institute,
Healthcare Safety Network (NHSN) haemovigilance Ottawa, ON, Canada
Search strategy and selection criteria
module.7 We graded evidence-based recommendations (A Tinmouth); Center for Clinical
Transfusion Research, Sanquin,
using the Chest8 grading system: grade 1A2C (table 2). A reference librarian identied studies through Mesh
Netherlands
Since many publications on this topic are uncontrolled keyword searches of the electronic databases of Cochrane (L Van De Watering MD);
case reports, case series, and retrospective cohort studies, Library and MEDLINE from Jan 1, 1940, to Dec 31, 2014, for Department of Anesthesiology
the evidence quality score reects the quality of the the diagnostic categories, blood components and & Bioengineering, University of
Pittsburgh & McGowan
literature. For example, there are few reports of transfusions, and adverse reactions. We included only articles Institute for Regenerative
transfusion reactions in the paediatric population. published in English. We excluded those focusing on plasma Medicine, Pittsburgh, PA, USA
Therefore, although evidence-based recommendations derivatives. Results are available on request. (Prof J H Waters MD); Division of
are the goal, some clinical situations that we discuss in Transfusion Medicine,

www.thelancet.com Vol 388 December 3, 2016 2825

 Review

Department of Pathology, Allergic and anaphylactic transfusion reactions suggests that transfusion can be restarted with the same
University of Pittsburgh, Allergic transfusion reactions occur during or within 4 h unit at a reduced rate under direct observation.14 The
Institute for Transfusion
Medicine, Pittsburgh, PA, USA
of transfusion with a blood component and are most transfusion must be discontinued if symptoms recur or if
(Prof M Yazer MD); and Division frequently associated with platelet transfusions (302 per additional symptoms appear beyond local cutaneous
of Transfusion Medicine, 100 000 platelet units).11 Symptoms are caused by manifestations.
Department of Pathology and mediators such as histamine, released on activation of Anaphylactic reactions (incidence eight per 100 000 units)
Laboratory Medicine, David
Geffen School of Medicine at
mast cells and basophils.12 Most allergic transfusion require prompt intramuscular administration of
UCLA, Los Angeles, CA, USA reactions are mild, with rash, pruritus, urticaria (hives), epinephrine (adrenaline; grade 1A).13,14 In addition to
(Prof A Ziman MD) and localised angio-oedema.7 The most severe reactions supportive measures, the following second-line drugs can
Correspondence to: are anaphylactic, characterised by a life-threatening be considered: H antihistamine (eg, chlorpheniramine,
Dr Meghan Delaney, systemic reaction, typically presenting as bronchospasm, diphenhydramine; grade 1C), bronchodilators
Bloodworks NW, Seattle,
WA 98115, USA
respiratory distress, and hypotension.7,13 ( adrenergic agonisteg, salbutamol solution;
[email protected] In mild allergic transfusion reactions (cutaneous grade 1C); glucocorticoid for intravenous administration
symptoms only), H antihistamine administration (eg, (eg, hydrocortisone or methylprednisolone; grade 1C); and
See Online for appendix diphenhydramine) should give symptomatic relief intravenous H antihistamine (eg, ranitidine; grade 1C).14
(grade 1A).1416 If symptoms resolve, then clinical experience Patients with a history of allergic transfusion reactions
should be monitored closely when receiving subsequent
Prevalence (per 100 000 units transfused) transfusions. There is no evidence to support routine
prophylaxis with antihistamines or glucocorticoids in
Allergic transfusion reaction 1122
patients with previous mild allergic transfusion reactions
Anaphylactic transfusion reaction 8
(grade 2C).17 Patients with moderate to severe allergic
Acute haemolytic transfusion reaction 2579
transfusion reactions should be counselled about their
Delayed haemolytic transfusion reaction 40
diagnosis and needs for future transfusion. In these
Delayed serological transfusion reaction 489757
patients, premedication with antihistamines (grade 2C),
Febrile non-haemolytic transfusion reaction 10003000
minimisation of the plasma content of the unit by removal
Hyperhaemolytic transfusion reaction Unknown
of excess supernatant (centrifugation or washing), or use of
Hypotensive transfusion reaction 1890
platelets stored in additive solutions reduces the incidence
Massive transfusion associated reactions (citrate, Unknown
or decreases the severity of future reactions (grade 1C).14,18,19
potassium, cold toxicity)
Use of corticosteroids as premedication has not been
Post-transfusion purpura Unknown
studied, but is used widely in our experience. For a patient
Septic transfusion reaction 00333 (product dependent)
with a history of an anaphylactic transfusion reaction,
Transfusion-associated circulatory overload 109 exclusion of serum protein deciency (eg, immunoglobulin
Transfusion-associated graft versus host disease Extremely rare (near 0%) with irradiation or pathogen A and haptoglobin) and other allergies might be warranted
reduction methods (grade 1C).14 In case of immunoglobulin A deciency with
Transfusion-associated necrotising enterocolitis Unknown anti-immunoglobulin A antibodies, but no history of an
Transfusion-related acute lung injury 0410 with mitigation (varies by component and anaphylactic reaction, use of immunoglobulin A-decient
post-implementation of risk mitigation strategies) or washed blood components can be undertaken; however,
Table 1: Rates of transfusion reactions
the supporting evidence is debated.14,20

Description Methodological quality of supporting evidence Implications

1A Strong recommendation, high quality evidence RCTs without important limitations or overwhelming Strong recommendation, can apply to most patients in most
evidence from observational studies circumstances without reservation
1B Strong recommendation, moderate quality RCTs with important limitations (inconsistent results, Strong recommendation, can apply to most patients in most
evidence methodological aws, indirect, or imprecise) or exceptionally circumstances without reservation
strong evidence from observational studies
1C Strong recommendation, low quality or very Observational studies or case series Strong recommendation but might change when higher quality
low quality evidence evidence becomes available
2A Weak recommendation, high quality evidence RCTs without important limitations or overwhelming Weak recommendation, best action might dier depending on
evidence from observational studies circumstances or patients or societal values
2B Weak recommendation, moderate quality RCTs with important limitations (inconsistent results, Weak recommendation; best action might dier depending on
evidence methodological aws, indirect, or imprecise) or exceptionally circumstances or patients or societal values
strong evidence from observational studies
2C Weak recommendation, low quality or very low Observational studies or case series Very weak recommendations; other alternatives might be equally
quality evidence reasonable

Used from Guyatt et al,8 with permission. RCT=randomised controlled trial.

Table 2: Evidence grading system by recommendation

2826 www.thelancet.com Vol 388 December 3, 2016

 Review

Acute haemolytic transfusion reactions

Acute haemolytic transfusion reactions can be either Initial search of Cochrane Library and PubMed
Allergic reactions (n=1569)
immune or non-immune. Immune-mediated acute Acute haemolytic reactions (n=4271)
haemolytic transfusion reactions result from infusion of Delayed serological reactions (n=16)
Febrile non-haemolytic reactions (n=53)
red blood cells that are incompatible with the patients Hyperhaemolysis or delayed haemolysis reactions (n=12)
anti-A, anti-B, or other red blood cell antibodies. Immune Hypotensive reactions (n=797)
acute haemolytic transfusion reactions are usually caused Citrate toxicity (n=63)
Cold and temperature toxicity (n=893)
by failure of patient identication at specimen collection Hyperkalaemic cardiac arrhythmia (n=11)
or transfusion, and less commonly by infusion of Post-transfusion purpura (n=784)
Acute gut injury reactions (n=119)
incompatible plasma, usually from an apheresis platelet Septic reactions (n=2869)
transfusion. In either setting, the antigenantibody Transfusion-associated graft versus host disease (n=1390)
interaction can lead to intravascular or extravascular Transfusion-associated circulatory overload (n=1038)
Transfusion-related acute lung injury (n=816)
haemolysis, presenting with sudden onset of fever or
chills (the most common [80%], and often the only
symptom), pain (from kidney capsular distension), Author review of reference list (each category)
Additional hand searches
hypotension, and dyspnoea. Other signs can include gross Section drafts written (main paper and appendix)
haemoglobinuria or haemoglobinaemia, disseminated
intravascular coagulation, acute renal failure, shock, and
death.7 Since fever and chills might be the only early signs, Review draft (two separate additional authors)

it is important to monitor patients during transfusions

and stop the transfusion immediately if there is any Final draft and grade of evidence
change in vital signs or the appearance of unexpected Review and conrmation by author group
symptoms.14 Health-care facilities should establish policies
that dene vital sign changes that should prompt Final manuscript with final citation list
evaluation of a suspected transfusion reaction.10 General reactions (n=13)
Allergic reactions (n=8)
Immune acute haemolytic transfusion reactions are Acute haemolytic reactions (n=8)
diagnosed on the basis of clinical ndings and Delayed serological reactions (n=2)
demonstration of serological incompatibility. Management Febrile non-haemolytic reactions (n=6)
Hyperhaemolysis or delayed haemolysis reactions (n=24)
is supportive. In severe reactions, cardiovascular, renal, Hypotensive reactions (n=9)
and respiratory support, and treatment for disseminated Massive transfusion (n=13)
(citrate toxicity, cold or temperature toxicity, hyperkalaemic cardiac
intravascular coagulation with bleeding, might be arrhythmia)
necessary.14 No evidence exists for the use of any specic Post-transfusion purpura (n=6)
intervention after an ABO-incompatible red blood cell Acute gut injury (n=4)
Septic reactions (n=9)
transfusion, although case reports highlight the use of red Transfusion-associated graft versus host disease (n=5)
blood cell or plasma exchange (grade 2C), intravenous Transfusion-associated circulatory overload (n=5)
Transfusion-related acute lung injury (n=9)
immunoglobulin (grade 2C), and complement-inhibiting
drugs (grade 2C).2124 Prevention relies on systems-based
Figure 1: Literature review and manuscript construction
practices and comprehensive training to ensure proper
patient identication at critical steps in the specimen
collection and transfusion processes (grade 1A).25,26 antibody titre subsequently decreases to levels
Non-immune acute haemolytic transfusion reactions undetectable by routine antibody detection testing. With
occur when red blood cells are haemolysed by factors standard laboratory techniques, 25% of red blood cell
other than antibodies, such as coadministration of red alloantibodies become undetectable over a median
blood cells with an incompatible crystalloid solution (eg, follow-up of 10 months after initial development, thus
5% dextrose solution), incorrect storage of blood, or use putting patients at risk for delayed transfusion reactions.30
of malfunctioning or non-validated administration Delayed haemolytic transfusion reactions are similar to
systems.27,28 Prevention requires close adherence to blood serological reactions with regard to mechanism and
handling and administration policies. timecourse. Delayed haemolytic transfusion reaction is
usually due to an anamnestic immune response when
Delayed haemolytic or delayed serological the recipient is unknowingly transfused with a red blood
transfusion reactions cell unit that expresses the cognate antigen.31 Re-exposure
The incidence of delayed haemolytic transfusion reaction to the foreign antigen causes a rise in red blood cell
is one per 2500 transfusions, but rises to 11% in patients antibody titres 24 h to 28 days after transfusion,
with sickle-cell disease.29 Patients at risk for delayed accompanied by either a fall or failure of haemoglobin
haemolytic or serological transfusion reactions include increment, rise in indirect bilirubin, or a positive direct
those with a history of red blood cell antibodies (through antiglobulin (Coombs) test; subsequent laboratory
pregnancy or transfusion exposure) in which the testing with elution studies usually demonstrates the

www.thelancet.com Vol 388 December 3, 2016 2827

 Review

All transfusions must be stopped when a patient is experiencing a reaction and assessed by a provider
Provide supportive therapy to support vital organ function (cardiac, pulmonary, renal)
For questions regarding transfusion reaction diagnosis or management, call the transfusion service, or other appropriate physician

Reaction Symptoms Interventions

Increase in temperature

Possible febrile non- Incremental increase <1C above Close observation, frequent vital signs
haemolytic reaction baseline and no other new symptoms If stable and no other new symptoms then continue with transfusion

Stop transfusion, keep intravenous line open, assess patient, check patient ID and unit ID
and compatibility
Possible bacterial Incremental increase 1C above Antipyretic drug
contamination baseline, or incremental increase Consider blood cultures (patient); empirical antibiotics if neutropenic
<1C with any other new symptoms Do not resume transfusion
(chills or rigors, hypotension, Strongly consider culturing blood product if 2C increase in temperature or if high clinical
Possible haemolysis nausea or vomiting) suspicion of sepsis
Notify blood transfusion laboratory; return unit (with administration set) plus
post-transfusion patient sample to blood transfusion laboratory

For consistently febrile patient due to underlying disease or treatment, when possible:
Avoid starting transfusion if patients temperature is increasing
Treat fever with antipyretic drug before starting transfusion
If incremental increase in temperature 1C above baseline treat as per above (stop and do not resume transfusion, cultures if indicated)
Notify blood transfusion laboratory, return unit (with administration set) plus post-transfusion patient sample to blood transfusion
laboratory

Allergic symptoms

Stop transfusion, keep intravenous line open, and assess patient

Antihistamines
Notify patient clinician and blood transfusion laboratory; sample not required
Urticaria Mild hives, rash, or skin itching only
If symptoms resolve, then can resume transfusion
If symptoms do not improve or worsen or recur then discontinue transfusion; return unit
(with administration set) to blood transfusion laboratory

Stop transfusion, keep intravenous line open, assess patient, check patient ID and unit ID
and compatibility
Hives, rash, itching, and or any other
Possible allergic Antihistamines
new symptoms (throat, eye, and
reaction Do not resume transfusion
tongue swelling, etc)
Notify blood transfusion laboratory; return unit (with administration set) plus
post-transfusion patient sample to blood transfusion laboratory

Respiratory symptoms

Stop transfusion, keep intravenous line open, assess patient, check patient ID and unit ID
Possible anaphylaxis, and patient compatibility
transfusion- Treat symptoms as indicated (adrenaline, antihistamines, steroids; oxygen and
associated circulatory Bronchospasm, dyspnoea, tachypnoea respiratory support, diuretics; uid, blood pressure, and renal support)
overload, septic and hypoxaemia, copious frothy Chest radiograph for presence of bilateral interstitial inltrate, if suggestive of
transfusion reaction, pink-tinged uid (from endotrachel tube) transfusion-related acute lung injury
or transfusion-related Blood cultures (patient and product), if high clinical suspicion of sepsis
acute lung injury Do not resume transfusion
Notify blood transfusion laboratory; return unit with administration set, plus
post-transfusion patient sample. Associated products can be quaratined

All other symptoms

Stop transfusion, keep intravenous line open, assess unit, check patient ID and unit ID
Possible anaphylaxis, and patient compatibility
Treat symptoms as indicated (adrenaline, antihistamines, steroids; oxygen and respiratory
Figure 2: Transfusion reaction haemolytic Chills, rigors, hypotension, nausea or
transfusion reaction, support, diuretics; uid, blood pressure, and renal support)
decision-tree vomiting, feeling of impending doom,
uid overload, or Blood cultures (patient and product) if high clinical suspicion of sepsis
Algorithm to guide assessment back or chest pain, intravenous site
transfusion-related Do not resume transfusion
and actions to take when a pain, cough, dyspnoea, hypoxia
acute lung injury Notify blood transfusion laboratory; return unit with administration set, plus
transfusion reaction is initially post-transfusion patient sample. Associated products can be quaratined
identied. Actions should go
from left to right.

2828 www.thelancet.com Vol 388 December 3, 2016

 Review

alloantibody. The most prominent clinical features of infection or haemolysis. Because febrile non-haemolytic
delayed haemolytic transfusion reactions include dark reactions are a diagnosis of exclusion, other important
urine or jaundice (4550%) followed by fever; chest, transfusion-related aetiologies must be ruled out with
abdominal or back pain; dyspnoea; chills; and post-reaction laboratory evaluation to detect haemolysis
hypertension.29,32,33 In patients with sickle-cell disease, (direct antiglobulin test and visual check for grossly
diagnosis might be delayed when only anaemia and haemolysed plasma). For patients who do not improve
jaundice are present if these symptoms are attributed to after cessation of transfusion or antipyretics, have a
veno-occlusive painful crisis. temperature increase of 2C or higher, or have clinical
Retrospective studies show that delayed serological signs of new bacterial infection, clinicians should
transfusion reactions are more common than are exclude a septic transfusion reaction; this is especially
haemolytic ones (066% vs 012%, respectively) in important after a platelet transfusion.50 When the
patients in hospital.34,35 Both share similar serological evaluation nds no other cause, such as an underlying
ndings, but patients with delayed serological reactions febrile illness, and testing for haemolysis is negative, a
do not have clinical signs or laboratory evidence of diagnosis of febrile non-haemolytic reaction can be
haemolysis. The antibodies most commonly responsible made. Antipyretic drugs and pethidine (meperidine) are
for both types of reactions are from the Rh, Kell, Duy, appropriate, although no studies have delineated their
Kidd, MNS, and Diego blood group systems.34 Less eectiveness.51
commonly, alloantibodies to low incidence antigens that Pre-storage leucocyte reduction can prevent febrile
are not detected by antibody detection screening tests can non-haemolytic reactions (grade 1A).52 Premedication
cause unrecognised haemolytic or serological transfusion with antipyretics does not decrease rate of reactions in
reactions. When there are signs of hemolysis, most patients and should be discouraged (grade 1A).48
retrospective crossmatching can be diagnostic.36,37 However, use of antipyretic drugs before transfusion for
Most patients do not require treatment other than patients who are persistently febrile due to underlying
additional transfusions to maintain desired haemoglobin. disease can enable transfusion completion in our
Red blood cell exchange transfusion to remove experience.53 The use of platelet additive solutions
incompatible red cells (grade 2C) or anti-CD20 in decreases the rate of reactions from 05% to 017%
combination with methylprednisolone have been (grade 1B).18
proposed for management of delayed haemolytic
transfusion reaction in patients with sickle-cell disease Hyperhaemolytic transfusion reactions
(grade 2C).38,39 Hyperhaemolytic transfusion reactions are rare,
Prevention is based on sensitive laboratory testing, life-threatening haemolytic transfusion reactions that
centralised medical records, and red blood cell unit typically occur in patients with haemoglobinopathies
selection.40 A central repository accessible across (1% to 19% of transfusions in patients with sickle-cell
health-care systems that includes patient red cell antibody disease), but can be seen in those with other disorders.5456
histories can inform the transfusing facility of previously Hyperhaemolytic transfusion reactions should be
identied antibodies, even if they are no longer detectable, suspected when the post-transfusion haemoglobin
and thereby ensure selection of compatible units for concentration is lower than the pre-transfusion
transfusion (grade 1B).4143 Prospective red cell antigen concentration. Signs include raised indirect bilirubin
matching can decrease alloimmunisation and thus the and lactate dehydrogenase and low concentrations
risk of subsequent delayed haemolytic transfusion of haptoglobin. A fall in absolute reticulocyte
reactions (grade 1A).44,45 Non-alloimmunised patients with count (decrease from baseline concentration) during
sickle-cell disease or thalassaemia should, at a minimum, haemolysis and a rise in reticulocyte count with recovery
receive red blood cells matched for Rh (D, C, c, E, e) and is a common nding. Hyperhaemolytic transfusion
K antigens; more highly antigen matched units should be reactions exist in acute and delayed forms. The acute
selected as is feasible (grade 1A).4447 form usually occurs less than 7 days after red blood cell
transfusion. Serological investigation of post-transfusion
Febrile non-haemolytic reactions samples might not show new or additional red blood cell
Febrile non-haemolytic reactions are common, occurring alloantibodies and direct antiglobulin test might be
in about 1% of transfusion episodes (13% per unit negative; furthermore, transfusion of antigen-negative
transfused).48 Febrile non-haemolytic reactions are crossmatch compatible units might not prevent this
caused by pro-inammatory cytokines or recipient reaction. In the delayed form, which usually occurs more
antibodies encountering donor antigen in the blood than 7 days after red blood cell transfusion, the direct
product.49 Reactions clinically present as a temperature antiglobulin test is positive and red blood cell
rise of 1C or higher, and can be accompanied by alloantibodies are identied in the post-transfusion
transient hypertension, chills, rigors, and discomfort. In sample.57 The diagnosis of acute hyperhaemolytic
the presence of fever, the transfusion must be stopped transfusion is challenging, and a high index of suspicion
immediately and the patient assessed closely for signs of is needed.

www.thelancet.com Vol 388 December 3, 2016 2829

 Review

Avoidance of further transfusions is a treatment Massive transfusion-associated reactions

recommendation for mild cases because this can worsen (citrate, potassium, cold toxicity)
haemolysis.57 However, if the patient presents with rapid Massive transfusion does not have a standard denition,
haemolysis and severe anaemia then transfusion might but can be described as a blood loss rate of 150 mL per
be needed. In such cases, intravenous immunoglobulin min, transfusion of 50% of a patients total blood
and corticosteroid (eg, methylprednisolone) are volume over 3 h, or more than ten units of red blood
recommended (grade 2C).58 For severe cases, additional cells in 24 h. Massive transfusion typically occurs in
intravenous immunoglobulin can be given, with uncontrolled haemorrhage, such as after trauma, but
consideration of associated risks such as renal toxicity, can also occur during surgical procedures, organ
changes in serological testing, and thromboembolic transplantation, or in non-bleeding patient undergoing
events (grade 2C).58 Rituximab and plasma exchange transfusion for sickle-cell disease or haemolytic disease
might be successful in severe cases (grade 2C).59 of the newborn.
Erythropoietin and eculizumab are not currently Reactions related to massive transfusion are
recommended due to insucient data showing ecacy.60 multifactorial, caused both by patient factors (eg, hepatic
The patient should be counselled about their diagnosis injury and shock) and by factors associated with
and the risk associated with future transfusions. transfusion of large volumes of blood products, including
sodium citrate (the anticoagulant used in stored blood
Hypotensive transfusion reactions products) and supernatant potassium, as well as the
Acute hypotensive transfusion reactions are uncommon infusion of large volumes of refrigerated products.71
and dened by an abrupt drop in systolic or diastolic When the patients metabolic ability to break down
blood pressure of more than 30 mm Hg within 15 min of citrate is exceeded, ionised calcium levels can drop,
the start of transfusion and resolving quickly (within resulting in tingling, paraesthesia, and changes in
10 min) once transfusion is stopped.61 Hypotension is the cardiac function, including alterations of cardiac
predominant manifestation; respiratory, gastrointestinal, depolarisation (prolonged QT interval) and blunting of
or mild allergic symptoms might also be present. left ventricular response (citrate toxicity).72,73 Management
These reactions are thought to occur with activation of constitutes administration of supplemental calcium,
the intrinsic contact activation pathway of the coagulation usually calcium gluconate or calcium citrate (grade 1A).74
cascade and generation of bradykinin and its active During storage of red blood cells, potassium
metabolite des-Arg9-bradykinin.62 Both kinins are potent concentration of the unit supernatant increases. The
vasodilators that cause facial ushing and a drop, often supernatant volume is about 2540% of the total unit
severe, in systolic and diastolic blood pressure, which in volume with a potassium concentration substantially
turn, triggers an increase in heart rate. These kinins also higher than that of normal human plasma.75 Transfusion-
produce slow contraction of the intestinal smooth muscle associated hyperkalaemic cardiac arrest has been
causing abdominal pain. reported after administration of large volumes or rapidly
Hypotensive reactions are more likely to occur in transfused red blood cells, particularly in children and
patients who have hypertension, are taking angiotensin- adolescents with hypovolaemia.76,77 Longer storage age
converting enzyme (ACE) inhibitors (since bradykinin and irradiation of the red blood cell product, rate and
metabolism is less ecient in the presence of an ACE volume of red blood cell transfusion, age and weight of
inhibitor), are being transfused blood products through a patient, and presence of comorbidities (hyperglycaemia,
negatively charged bedside leucocyte reduction lter, are hypocalcaemia, hypothermia, acidosis, and renal
undergoing apheresis, or are receiving platelets.6365 These insuciency) are risk factors for transfusion-associated
reactions have also been reported during cardiopulmonary hyperkalaemic cardiac arrest. Treatment of hyperkalaemia
bypass and radical prostatectomy.6668 Transfusion must be might include insulin, glucose, calcium gluconate, and
stopped and prompt clinical assessment and supportive furosemide. Given that most cases of transfusion-
therapy given; no specic treatment is indicated because associated hyperkalaemic cardiac arrest have been
the hypotension typically resolves once transfusion is reported in the perioperative setting, known risk factors
stopped. The same unit should not be restarted because should be considered before massive transfusion.76
symptoms might recur. Other transfusion reactions in Patients with low total blood volume who might receive a
which hypotension can be a sign, such as allergic, large volume of red blood cells in a short period should
haemolytic, septic reactions, transfusion-related acute be transfused at a maximum infusion rate of 05 mL/kg
lung injury, or anaphylaxis, must be excluded. No routine per min.75 Use of red blood cells units with less
preventative measures have been identied other than supernatant (washing or plasma-reduced), or fresh units
not using a bedside leucocyte reduction lter.67 If the (710 days old; grade 1B), and avoidance of red blood
patient is being treated with an ACE inhibitor and needs cell units that are irradiated more than 12 h before
continuing transfusion therapy, physicians should transfusion (grade 2C), might decrease the risk of
consider switching them to another class of transfusion-associated hyperkalaemic cardiac arrest.78
antihypertensive drug (grade 2C).69,70 Findings from small studies have shown use of an inline

2830 www.thelancet.com Vol 388 December 3, 2016

 Review

potassium lter before transfusion or ultraltration of are more eective than those from random donors in the
the priming volume before initiation of extracorporeal acute thrombocytopenic phase. Prevention of recurrence
life support to be eective (grade 1B).79,80 of post-transfusion purpura can include use of washed
Massive transfusion can also be associated with red blood cell units, or use of platelet and red blood cell
hypothermia (cold toxicity). Red blood cell and plasma units from HPA compatible donors or autologous
units (which are stored at refrigerated temperatures) can transfusion.88 Leucocyte reduced blood components are
lead to hypothermia when given rapidly and in large required (grade 2A).89 The clinical sta and patient
volumes.81 In severe hypothermia (<30C), cardiac should be advised on the risk of recurrence with future
conduction slows leading to cardiac arrest. Other eects transfusions and need for antigen-negative or washed
of hypothermia include slowing of temperature- blood products (grade 2C).
dependent enzymatic reactions, resulting in impaired
citrate and delayed drug metabolism, impairment of the Septic transfusion reactions
coagulation cascade, and reduction of platelet function Septic transfusion reactions usually present during or
resulting in coagulopathy. Hypothermia can be managed within 4 h of transfusion. Severe septic reactions occur in
with forced air warming devices (grade 1A), and, in about 58 00075 000 transfusions a year, although
extreme circumstances, warm peritoneal lavage or bacterial contamination of platelets is thought to be
cardiopulmonary bypass (grade 1A).82,83 much more common.9092 Fever, rigors, hypotension, and
Prospective monitoring and planning can prevent other signs associated with systemic inammatory
these reactions; ionised calcium concentrations should response syndrome are the most common presentation.
be measured regularly and supplemental calcium given Denitive diagnosis of transfusion-transmitted bacterial
as needed (grade 1B). Inline blood warming devices infection requires isolation of the same organism from
should be used to warm blood products rapidly to the blood product and patient, but can be presumed in a
normal body temperature during transfusion in the culture-negative patient with clinical sepsis if bacteria are
massive transfusion setting (grade 1A).82 isolated from the transfused unit.93
In a patient with new bacterial bloodstream infection
Post-transfusion purpura following transfusion, all units recently transfused
Post-transfusion purpura is a rare reaction dened as should be evaluated for bacterial contamination with
thrombocytopenia that develops 512 days after red blood Gram stain and culture.93 Bacterial cultures should be
cell or platelet transfusion. The clinical pattern consists taken from the patient and any indwelling lines before
of rapid onset of thrombocytopenia (platelet count can antibiotics are started.14 Broad-spectrum antibiotics such
fall from normal ranges to below 10 10 per L within as -lactams and aminoglycosides should be started
24 h), typically in a middle-aged or elderly woman with a empirically (grade 1A) with anti-Pseudomonas spp
recent history of red blood cell or platelet transfusion.6,84 coverage if a red blood cell unit is implicated.93
Other ndings might include widespread purpura, Procedures to reduce bacterial contamination of
bleeding from mucous membranes, and, in severe cases, blood products include donor screening and proper
intracranial haemorrhage and death.85 The transfusion skin disinfection before collection, sequestering the
precipitating the fall in platelet count causes a secondary, rst 1050 mL of donated blood (and skin plug) in a
or anamnestic, immune response, increasing antibody small pouch that is diverted away from the collected
titres directed against specic human platelet antigens blood, visual inspection of all units before issue, and
(HPA). Post-transfusion purpura usually aects HPA-1a- pre-transfusion bacterial surveillance of platelet units
negative individuals (phenotypic frequency up to 2% (grade 1B).91,94 Platelet units have the highest bacterial
depending on patient ethnic origin) who have previously contamination rate (one in 30005000 units) because
been alloimmunised by pregnancy; however, other HPA platelets are stored at room temperature;95 but many do
antigens might be implicated. In elderly patients, platelet not cause infection because they are removed from the
transfusions, multiple transfusions, and the presence of inventory due to positive surveillance, or transfused
comorbidities are risk factors.86 The mechanism of before bacterial growth has reached a clinically
destruction of the patients own antigen-negative signicant level.94
platelets remains unclear. Pathogen reduction systems use ultraviolet light to
Diagnosis is conrmed by the detection of platelet- crosslink nucleic acids (with or without amotosalen) to
specic alloantibodies. Management should be treat blood products and inactivate viruses, bacteria,
supportive. In untreated cases, thrombocytopenia usually and parasites.94,96,97 Prospective studies of pathogen
persists for 728 days, but can continue for longer. reduction systems for platelets show that their use is
Treatment with intravenous immunoglobulin (grade 1B), associated with lower septic event rates than transfusion
steroids, or plasma exchange is indicated (grade 2C).87 of conventionally prepared platelets.98 Since national
Platelet transfusion can be given, but is sometimes implementation of pathogen reduction systems in 2011
associated with poor increments; there is no evidence in Switzerland, septic transfusion reactions have
that platelet concentrates from antigen-negative donors decreased (grade 1A).98

www.thelancet.com Vol 388 December 3, 2016 2831

 Review

Transfusion-associated circulatory overload intrauterine transfusion are also at risk. Immuno-

Transfusion-associated circulatory overload is an under- competent patients are at risk when receiving cellular
recognised reaction, aecting about 18% of patients components from blood relatives or if being transfused
who are transfused99101 or occurring after about one in a donor population with little HLA diversity.106
in 9177 transfused components.102 There is no consensus The signs and symptoms of transfusion-associated
for diagnosing transfusion-associated circulatory graft versus host disease develop 510 days after
overload; the NHSN denition requires new onset, or transfusion and usually consist of an erythematous
acute exacerbation of three or more of the following, maculopapular rash, fever, abdominal pain, diarrhoea,
within 6 h of transfusion: respiratory distress, raised nausea, and vomiting. Laboratory tests show
brain natriuretic peptide (BNP or NT-pro-BNP), increased pancytopenia, abnormal liver function, and electrolyte
central venous pressure, left heart failure, positive uid disturbances. A skin biopsy from the aected area can
balance, or pulmonary oedema.7 These criteria are help with diagnosis; although not specic, typical
similar to the UK Serious Hazards of Transfusion features include interface lymphocytic inltrate with
(SHOT) National Haemoviligance scheme for diagnosis basil cell vacuolization.105 Full marrow aplasia, evident on
of transfusion-associated circulatory overload, although bone marrow biopsy, usually develops within 21 days of
with a shorter timeline (4 h) to development of signs and transfusion. Transfusion-associated graft versus host
symptoms.6 Transfusion-associated circulatory overload disease is nearly always fatal; death is usually attributable
is caused by an excessive quantity of transfused blood to infections.107
components or an excessive rate of transfusion (excessive Management is supportive. Transfusion-associated
is relative to each patient). An inammatory component graft versus host disease can be prevented by irradiating
might also exist.103 Risk factors include older age, renal cellular blood components with gamma-rays or x-rays, or
failure (especially if on dialysis), pre-existing uid by treating blood products with pathogen reduction
overload, cardiac dysfunction, administration of large technology to disrupt the residual lymphocytes ability to
volumes of blood products, and rapid administration proliferate (grade 1B).108 Leucocyte reduction is not
rate. The dierential diagnosis of transfusion-associated sucient for prevention; however, recent SHOT data
circulatory overload includes transfusion-related acute suggest a threshold eect for the number for T cells
lung injury, septic transfusion reaction, and acute needed to cause the reaction.3
haemolytic transfusion reaction.
Treatment of transfusion-associated circulatory overload Transfusion-associated necrotising enterocolitis
requires stopping transfusion and administering supple- Necrotising enterocolitis is common in preterm and
mental oxygen as needed. Administration of diuretics can very low birthweight neonate infants. The pathogenesis
be both diagnostic and therapeutic. At-risk patients should of transfusion-associated necrotising enterocolitis is
be identied (grade 2C) and given transfusions slowly unknown; some investigators have postulated a
over 34 h (grade 2C), with the smallest quantity of blood connection with transfusion but the literature is
products given (ie, one unit, divided into two components) dominated by retrospective case-control studies with
to achieve the clinical goal (grade 2C).100 For patients with a moderate risk of bias.109,110 Prospective studies are needed
history of transfusion-associated circulatory overload, the to assess the causality of any association between
benet of diuretics before or during the transfusion has necrotising enterocolitis and transfusion, and the place
not been studied, but might be logical in the of withholding feeds during transfusion, which could
haemodynamically stable patient. aect blood ow to the gastrointestinal tract.111,112

Transfusion-associated graft versus host disease Transfusion-related acute lung injury

Transfusion-associated graft versus host disease is an Transfusion-related acute lung injury is characterised by
extremely rare adverse event caused by transfusion the development of non-cardiogenic pulmonary oedema
of cellular components containing viable donor after transfusion. Although understanding of the
lymphocytes that recognise their new host as foreign pathogenesis has increased greatly in the past few
and engraft in the recipient.104 Transfusion with whole decades, it remains incompletely understood.92,93
blood, red blood cells, platelets, HLA-matched platelets, Cognate anti-HLA or anti-human neutrophil antigen
and granulocytes has been implicated.104,105 At risk are (anti-HNA) antibodies alone are enough to cause
severely immunodecient patients such as recipients of transfusion-related acute lung injury, but most cases are
haemopoietic stem cell transplantation (past and postulated to occur through a two event model. The rst
current) or patients with congenital immunodeciency event is a clinical disorder that causes activation of the
aecting T cells or Hodgkins lymphoma; those in need pulmonary endothelium, leading to the sequestration
of neonatal exchange transfusions; and patients taking and priming of neutrophils in the lung. Clinical risk
high-dose chemotherapy or radiotherapy, purine- factors that might function as the rst event include
analogue drugs, alemtuzumab, or anti-thymocyte high interleukin 8 concentrations, liver surgery, chronic
globulin for aplastic anaemia.105 Fetuses in need of alcohol abuse, shock, high peak airway pressure during

2832 www.thelancet.com Vol 388 December 3, 2016

 Review

mechanical ventilation, current smoking, and positive tidal volume ventilation and a restrictive uid strategy as
uid balance.113 The second event results from the in other causes of acute lung injury (grade 1A).117 A
blood product transfusion, which activates the restrictive transfusion strategy to avoid unnecessary
primed neutrophils causing endothelial damage and transfusions will also be preventive.
subsequently acute lung injury. This can result from
either passive transfer of antibodies (immune-mediated) Conclusion
or pro-inammatory mediators (non-immune mediated) In this Review we present the salient features and
in the transfused component. Since neutrophil management of the dierent diagnostic categories for
sequestration and activation is involved in development transfusion reactions. We recognise the highly variable
of transfusion-related acute lung injury, recipient pathophysiological mechanisms that underlie reactions,
factors including neutrophil number and function also as well as the diverse risk factors patients might have. In
probably play an important role. acute transfusion reactions prompt recognition and
Risk for immune-mediated lung injury after cessation of transfusion is crucial, as well as
transfusion varies by blood component. Risk has been communication with the transfusion service and
reduced substantially by strategies targeting donor laboratory. Correct diagnosis is essential to provide
selection and blood product collection (eg, use of male appropriate treatment and to ensure the safety of any
donors only for plasma and plasma used for suspension future transfusions. Many of the evidence-based recom-
of buy coat derived platelet pools, and screening of mendations are supported by weak recommendations
female apheresis platelet donors for HLA/HNA due to sparse publications. Prospective studies are needed
antibodies with retesting after pregnancies; grade 2C).114 in all populations; evidence is particularly sparse in
Available risk estimates per component transfused (after children and patients who have repeat transfusion needs.
full implementation of immune-mediated risk mitigation Contributors
strategies) are based on active reporting and might MD conceived of the Review, gathered co-authors, guided paper
underestimate risk (plasma 04 per 100 000 units, development, and wrote and edited the report. SW guided the paper
submission to journal, wrote and reviewed sections of the report, and
apheresis platelets one per 100 000 units, and red blood graded evidence. RSB, JC, CC, NMD, TOA, SJS, JHW, and MY wrote
cells 05 per 100 000 units).115 sections of paper, reviewed sections of the report, and graded evidence.
Available risk mitigation strategies do not address non- MP, AT, and LVDW guided paper development, reviewed the report, and
immune-mediated injury. Novel methods for risk provided expert review. AZ assisted with paper conception, guided paper
development, and wrote and edited the report.
reduction are currently under investigation. A technique
for pre-storage experimental ltration for red blood cell Declaration of interests
MP is an employee of Haemonetics Corporation, Braintree, MA. USA;
units,116 which removes antibodies, lipids, white blood a manufacturer of blood processing equipment. All other authors declare
cells, and platelets, and prevents transfusion-related no competing interests.
acute lung injury, is in development in an animal model. Acknowledgments
The clinical presentation of transfusion-related acute The authors are scientic or guest members of the Biomedical
lung injury includes dyspnoea, tachypnoea, and Excellence for Safer Transfusion (BEST) Collaborative, an international For the BEST Collaborative see
hypoxaemia, sometimes accompanied by rigors, research organisation that works collaboratively to improve http://bestcollaborative.org/
transfusion-related services through standardisation of analytic
tachycardia, fever, hypothermia, and hypotension or techniques, development of new procedures, systematic review of
hypertension.117 Copious frothy pink-tinged uid might evidence, and execution of clinical and laboratory studies. We thank
be seen in the endotracheal tube of mechanically Dana Matthews, Dee Townsend-McCall, and Karyn Brundige for
creating the rst transfusion reaction decision-tree and Rikke Ogawa for
ventilated patients, but this nding is non-specic.114
her assistance with the literature search.
Transient leucopenia might be noted.117 Bilateral
References
interstitial inltrates are present on chest radiograph 1 Pfuntner A, Wier LM, Stocks C. Most Frequent Procedures
but this nding is non-specic and dicult to Performed in U.S. Hospitals, 2011. Rockville: Agency for Healthcare
distinguish from overload oedema.118 Diagnosis is made Research and Quality, 2013.
2 Callum JL, Waters JH, Shaz BH, Sloan SR, Murphy MF. The AABB
on the basis of clinical and radiographic ndings in recommendations for the Choosing Wisely campaign of the
conjunction with a temporal association with American Board of Internal Medicine. Transfusion 2014; 54: 234452.
transfusion (typically within 6 h, although delayed cases 3 Bolton-Maggs PHB, ed, on behalf of the Serious Hazards of
presenting up to 72 h after transfusion have been Transfusion (SHOT) Steering Group. The 2013 Annual SHOT
Report. Manchester: SHOT, 2014.
described).119 Transfusion-related acute lung injury can 4 Riley W, Smalley B, Pulkrabek S, Clay ME, McCullough J.
be dicult to distinguish from oedema associated with Using lean techniques to dene the platelet (PLT) transfusion
heart failure, and other acute transfusion reactions with process and cost-eectiveness to evaluate PLT dose transfusion
strategies. Transfusion 2012; 52: 195767.
similar presentations (transfusion-associated circulatory 5 Ezidiegwu CN, Lauenstein KJ, Rosales LG, Kelly KC, Henry JB.
overload, septic transfusion reaction, and anaphylaxis) Febrile nonhemolytic transfusion reactions. Management by
should be excluded.117 premedication and cost implications in adult patients.
Arch Pathol Lab Med 2004; 128: 99195.
Management of transfusion-related acute lung injury is 6 Bolton-Maggs PH. Bullet points from SHOT: key messages and
supportive, with supplemental oxygen or mechanical recommendations from the Annual SHOT Report 2013.
ventilation given as needed, and application of restrictive Transfus Med 2014; 24: 197203.

www.thelancet.com Vol 388 December 3, 2016 2833

 Review

7 CDC. NHSN Biovigilance Component, Hemovigilance Module 30 Schonewille H, Haak HL, van Zijl AM. RBC antibody persistence.
Surveillance Protocol v2.1.3. Atlanta: Centers for Disease Control and Transfusion 2000; 40: 112731.
Prevention, 2014. 31 Unni N, Peddinghaus M, Tormey CA, Stack G. Record
8 Guyatt G, Gutterman D, Baumann MH, et al. Grading strength of fragmentation due to transfusion at multiple health care facilities:
recommendations and quality of evidence in clinical guidelines: a risk factor for delayed hemolytic transfusion reactions. Transfusion
report from an American College of Chest Physicians task force. 2014; 54: 98103.
Chest 2006; 129: 17481. 32 Rosse WF, Gallagher D, Kinney TR, et al, and the The Cooperative
9 Elliott M, Coventry A. Critical care: the eight vital signs of patient Study of Sickle Cell Disease. Transfusion and alloimmunization in
monitoring. Br J Nurs 2012; 21: 62125. sickle cell disease. Blood 1990; 76: 143137.
10 Fung M, Grossman BJ, Hillyer CD, Westho CM. Technical 33 de Montalembert M, Dumont MD, Heilbronner C, et al. Delayed
Manual, 18th edn. Glen Burnie, MD: AABB Press, 2014. hemolytic transfusion reaction in children with sickle cell disease.
11 Harvey AR, Basavaraju SV, Chung KW, Kuehnert MJ. Haematologica 2011; 96: 80107.
Transfusion-related adverse reactions reported to the National 34 Ness PM, Shirey RS, Thoman SK, Buck SA. The dierentiation of
Healthcare Safety Network Hemovigilance Module, United States, delayed serologic and delayed hemolytic transfusion reactions:
2010 to 2012. Transfusion 2014; published online Nov 5. DOI:10.1111/ incidence, long-term serologic ndings, and clinical signicance.
trf.12918. Transfusion 1990; 30: 68893.
12 Hirayama F. Current understanding of allergic transfusion 35 Vamvakas EC, Pineda AA, Reisner R, Santrach PJ, Moore SB.
reactions: incidence, pathogenesis, laboratory tests, prevention and The dierentiation of delayed hemolytic and delayed serologic
treatment. Br J Haematol 2013; 160: 43444. transfusion reactions: incidence and predictors of hemolysis.
13 Simons FE, Ardusso LR, Bil MB, et al, and the World Allergy Transfusion 1995; 35: 2632.
Organization. World Allergy Organization anaphylaxis guidelines: 36 Noizat-Pirenne F. Relevance of blood groups in transfusion of sickle
summary. J Allergy Clin Immunol 2011; 127: 587, e122. cell disease patients. C R Biol 2013; 336: 15258.
14 Tinegate H, Birchall J, Gray A, et al, and the BCSH Blood 37 Koshy R, Patel B, Harrison JS. Anti-Kpa-induced severe delayed
Transfusion Task Force. Guideline on the investigation and hemolytic transfusion reaction. Immunohematology 2009; 25: 4447.
management of acute transfusion reactions. Prepared by the BCSH 38 Tormey CA, Stack G. Limiting the extent of a delayed hemolytic
Blood Transfusion Task Force. Br J Haematol 2012; 159: 14353. transfusion reaction with automated red blood cell exchange.
15 Lin RY, Curry A, Pesola GR, et al. Improved outcomes in patients Arch Pathol Lab Med 2013; 137: 86164.
with acute allergic syndromes who are treated with combined H1 39 Noizat-Pirenne F, Habibi A, Mekontso-Dessap A, et al. The use of
and H2 antagonists. Ann Emerg Med 2000; 36: 46268. rituximab to prevent severe delayed haemolytic transfusion reaction
16 Runge JW, Martinez JC, Caravati EM, Williamson SG, Hartsell SC. in immunized patients with sickle cell disease. Vox Sang 2015;
Histamine antagonists in the treatment of acute allergic reactions. 108: 26267.
Ann Emerg Med 1992; 21: 23742. 40 Winters JL, Richa EM, Bryant SC, Tauscher CD, Bendix BJ,
17 Mart-Carvajal AJ, Sol I, Gonzlez LE, Leon de Gonzalez G, Stubbs JR. Polyethylene glycol antiglobulin tube versus gel
Rodriguez-Malagon N. Pharmacological interventions for the microcolumn: inuence on the incidence of delayed hemolytic
prevention of allergic and febrile non-haemolytic transfusion transfusion reactions and delayed serologic transfusion reactions.
reactions. Cochrane Database Syst Rev 2010; (6): CD007539. Transfusion 2010; 50: 144452.
18 Cohn CS, Stubbs J, Schwartz J, et al. A comparison of adverse 41 Harm SK, Yazer MH, Monis GF, Triulzi DJ, Aubuchon JP,
reaction rates for PAS C versus plasma platelet units. Transfusion Delaney M. A centralized recipient database enhances the serologic
2014; 54: 192734. safety of RBC transfusions for patients with sickle cell disease.
19 Tobian AA, Savage WJ, Tisch DJ, Thoman S, King KE, Ness PM. Am J Clin Pathol 2014; 141: 25661.
Prevention of allergic transfusion reactions to platelets and red 42 Delaney M, Dinwiddie S, Nester TN, Aubuchon JA.
blood cells through plasma reduction. Transfusion 2011; 51: 167683. The immunohematologic and patient safety benets of a
20 Sandler SG, Eder AF, Goldman M, Winters JL. The entity of centralized transfusion database. Transfusion 2013; 53: 77176.
immunoglobulin A-related anaphylactic transfusion reactions is not 43 Schwickerath V, Kowalski M, Menitove JE. Regional registry of
evidence based. Transfusion 2015; 55: 199204. patient alloantibodies: rst-year experience. Transfusion 2010;
21 Weinstock C, Mhle R, Dorn C, et al. Successful use of eculizumab 50: 146570.
for treatment of an acute hemolytic reaction after ABO-incompatible 44 Vichinsky EP, Luban NL, Wright E, et al, and the Stroke Prevention
red blood cell transfusion. Transfusion 2015; 55: 60510. Trail in Sickle Cell Anemia. Prospective RBC phenotype matching
22 Fredlund H, Bersus O, Bjrsell-Ostlilng E, Filbey D. in a stroke-prevention trial in sickle cell anemia: a multicenter
A retrospective study of acute plasma exchange in severe transfusion trial. Transfusion 2001; 41: 108692.
intravascular hemolysis. Eur J Haematol 1989; 43: 25961. 45 Lasalle-Williams M, Nuss R, Le T, et al. Extended red blood cell
23 Seager OA, Nesmith MA, Begelman KA, et al. Massive acute antigen matching for transfusions in sickle cell disease: a review of
hemodilution for incompatible blood reaction. JAMA 1974; a 14-year experience from a single center (CME). Transfusion 2011;
229: 79092. 51: 173239.
24 Kohan AI, Niborski RC, Rey JA, et al. High-dose intravenous 46 Chou ST, Jackson T, Vege S, Smith-Whitley K, Friedman DF,
immunoglobulin in non-ABO transfusion incompatibility. Westho CM. High prevalence of red blood cell alloimmunization
Vox Sang 1994; 67: 19598. in sickle cell disease despite transfusion from Rh-matched minority
25 Figueroa PI, Ziman A, Wheeler C, Gornbein J, Monson M, donors. Blood 2013; 122: 106271.
Calhoun L. Nearly two decades using the check-type to prevent ABO 47 Pujani M, Pahuja S, Dhingra B, Chandra J, Jain M.
incompatible transfusions: one institutions experience. Alloimmunisation in thalassaemics: a comparison between
Am J Clin Pathol 2006; 126: 42226. recipients of usual matched and partial better matched blood.
26 Goodnough LT, Viele M, Fontaine MJ, et al. Implementation of a An evaluation at a tertiary care centre in India. Blood Transfus 2014;
two-specimen requirement for verication of ABO/Rh for blood 12 (suppl 1): s10004.
transfusion. Transfusion 2009; 49: 132128. 48 Sanders RP, Maddirala SD, Geiger TL, et al. Premedication with
27 Strautz RL, Nelson JM, Meyer EA, Shulman IA. Compatibility of acetaminophen or diphenhydramine for transfusion with
ADSOL-stored red cells with intravenous solutions. leucoreduced blood products in children. Br J Haematol 2005;
Am J Emerg Med 1989; 7: 16264. 130: 78187.
28 McCullough J, Polesky HF, Nelson C, Ho T. Iatrogenic hemolysis: 49 Heddle NM. Pathophysiology of febrile nonhemolytic transfusion
a complication of blood warmed by a microwave device. reactions. Curr Opin Hematol 1999; 6: 42026.
Anesth Analg 1972; 51: 10206. 50 Jacobs MR, Smith D, Heaton WA, Zantek ND, Good CE, and the
29 Talano JA, Hillery CA, Gottschall JL, Baylerian DM, Scott JP. PGD Study Group. Detection of bacterial contamination in
Delayed hemolytic transfusion reaction/hyperhemolysis syndrome prestorage culture-negative apheresis platelets on day of issue with
in children with sickle cell disease. Pediatrics 2003; 111: e66165. the Pan Genera Detection test. Transfusion 2011; 51: 257382.

2834 www.thelancet.com Vol 388 December 3, 2016

 Review

51 Friedlander M, Noble WH. Meperidine to control shivering associated 74 Meikle A, Milne B. Management of prolonged QT interval during a
with platelet transfusion reaction. Can J Anaesth 1989; 36: 46062. massive transfusion: calcium, magnesium or both? Can J Anaesth
52 Heddle NM, Blajchman MA, Meyer RM, et al. A randomized 2000; 47: 79295.
controlled trial comparing the frequency of acute reactions to plasma- 75 Strauss RG. RBC storage and avoiding hyperkalemia from
removed platelets and prestorage WBC-reduced platelets. Transfusion transfusions to neonates & infants. Transfusion 2010; 50: 186265.
2002; 42: 55666. http://onlinelibrary.wiley.com/o/cochrane/ 76 Lee AC, Reduque LL, Luban NL, Ness PM, Anton B, Heitmiller ES.
clcentral/articles/825/CN-00389825/frame.html. (serial online). Transfusion-associated hyperkalemic cardiac arrest in pediatric
53 Kennedy LD, Case LD, Hurd DD, Cruz JM, Pomper GJ. patients receiving massive transfusion. Transfusion 2014; 54: 24454.
A prospective, randomized, double-blind controlled trial of 77 Sesok-Pizzini D, Pizzini MA. Hyperkalemic cardiac arrest in
acetaminophen and diphenhydramine pretransfusion medication pediatric patients undergoing massive transfusion: unplanned
versus placebo for the prevention of transfusion reactions. emergencies. Transfusion 2014; 54: 47.
Transfusion 2008; 48: 228591. 78 Weiskopf RB, Schnapp S, Rouine-Rapp K, Bostrom A, Toy P.
54 Diamond WJ, Brown FL Jr, Bitterman P, Klein HG, Davey RJ, Extracellular potassium concentrations in red blood cell
Winslow RM. Delayed hemolytic transfusion reaction presenting suspensions after irradiation and washing. Transfusion 2005;
as sickle-cell crisis. Ann Intern Med 1980; 93: 23134. 45: 1295301.
55 Garratty G. What do we mean by Hyperhaemolysis and what is 79 Inaba S, Nibu K, Takano H, et al. Potassium-adsorption lter for RBC
the cause? Transfus Med 2012; 22: 7779. transfusion: a phase III clinical trial. Transfusion 2000; 40: 146974.
56 Darabi K, Dzik S. Hyperhemolysis syndrome in anemia of chronic 80 Delaney M, Axdor-Dickey RL, Crockett GI, Falconer AL,
disease. Transfusion 2005; 45: 193033. Levario MJ, McMullan DM. Risk of extracorporeal life support
57 Win N, New H, Lee E, de la Fuente J. Hyperhemolysis syndrome in circuit-related hyperkalemia is reduced by prebypass ultraltration.
sickle cell disease: case report (recurrent episode) and literature Pediatr Crit Care Med 2013; 14: e26367.
review. Transfusion 2008; 48: 123138. 81 Boyan CP. Cold or warmed blood for massive transfusions.
58 Win N, Sinha S, Lee E, Mills W. Treatment with intravenous Ann Surg 1964; 160: 28286.
immunoglobulin and steroids may correct severe anemia in 82 Moola S, Lockwood C. Eectiveness of strategies for the
hyperhemolytic transfusion reactions: case report and literature management and/or prevention of hypothermia within the adult
review. Transfus Med Rev 2010; 24: 6467. perioperative environment. Int J Evid-Based Healthc 2011; 9: 33745.
59 Uhlmann EJ, Shenoy S, Goodnough LT. Successful treatment of 83 Leslie K, Sessler DI. Perioperative hypothermia in the high-risk
recurrent hyperhemolysis syndrome with immunosuppression surgical patient. Best Pract Res Clin Anaesthesiol 2003; 17: 48598.
and plasma-to-red blood cell exchange transfusion. Transfusion 84 Gonzalez CE, Pengetze YM. Post-transfusion purpura.
2014; 54: 38488. Curr Hematol Rep 2005; 4: 15459.
60 Gupta S, Fenves A, Nance ST, Sykes DB, Dzik WS. Hyperhemolysis 85 Kalish RI, Jacobs B. Post-transfusion purpura: initiation by
syndrome in a patient without a hemoglobinopathy, unresponsive leukocyte-poor red cells in a polytransfused woman. Vox Sang 1987;
to treatment with eculizumab. Transfusion 2015; 55: 62328. 53: 16972.
61 Whitaker BI. National Blood Collection and Utilization Survey 2011. 86 Menis M, Forshee RA, Anderson SA, et al. Posttransfusion purpura
Report number HHSP23320110008TC, OMB Number 0990-0313. occurrence and potential risk factors among the inpatient US
Bethesda: AABB, 2011. elderly, as recorded in large Medicare databases during 2011
62 Moreau ME, Thibault L, Dsormeaux A, et al. Generation of kinins through 2012. Transfusion 2015; 55: 28495.
during preparation and storage of whole blood-derived platelet 87 Mueller-Eckhardt C, Kiefel V. High-dose IgG for post-transfusion
concentrates. Transfusion 2007; 47: 41020. purpura-revisited. Blut 1988; 57: 16367.
63 Sweeney JD, Dupuis M, Mega AP. Hypotensive reactions to red 88 Win N, Peterkin MA, Watson WH. The therapeutic value of
cells ltered at the bedside, but not to those ltered before storage, HPA-1a-negative platelet transfusion in post-transfusion purpura
in patients taking ACE inhibitors. Transfusion 1998; 38: 41011, complicated by life-threatening haemorrhage. Vox Sang 1995;
author reply 41315. 69: 13839.
64 Takahashi TA, Abe H, Hosoda M, Nakai K, Sekiguchi S. Bradykinin 89 Williamson LM, Stainsby D, Jones H, et al. The impact of universal
generation during ltration of platelet concentrates with a white leukodepletion of the blood supply on hemovigilance reports of
cell-reduction lter. Transfusion 1995; 35: 967. posttransfusion purpura and transfusion-associated graft-versus-host
65 Pagano MB, Ness PM, Chajewski OS, King KE, Wu Y, Tobian AA. disease. Transfusion 2007; 47: 145567.
Hypotensive transfusion reactions in the era of prestorage 90 Funk MB, Lohmann A, Guenay S, et al. Transfusion-transmitted
leukoreduction. Transfusion 2015; 55: 166874. bacterial infections haemovigilance data of German blood
66 Arnold DM, Molinaro G, Warkentin TE, et al. Hypotensive establishments (1997-2010). Transfus Med Hemother 2011; 38: 26671.
transfusion reactions can occur with blood products that are 91 Robillard P, Delage G, Itaj NK, Goldman M. Use of hemovigilance
leukoreduced before storage. Transfusion 2004; 44: 136166. data to evaluate the eectiveness of diversion and bacterial
67 Cyr M, Eastlund T, Blais C Jr, Rouleau JL, Adam A. Bradykinin detection. Transfusion 2011; 51: 140511.
metabolism and hypotensive transfusion reactions. Transfusion 92 Lafeuillade B, Eb F, Ounnoughene N, et al. Residual risk and
2001; 41: 13650. retrospective analysis of transfusion-transmitted bacterial infection
68 Owen HG, Brecher ME. Atypical reactions associated with use of reported by the French National Hemovigilance Network from
angiotensin-converting enzyme inhibitors and apheresis. 2000 to 2008. Transfusion 2015; 55: 63646.
Transfusion 1994; 34: 89194. 93 Eder AF, Goldman M. How do I investigate septic transfusion
69 Kalra A, Palaniswamy C, Patel R, Kalra A, Selvaraj DR. reactions and blood donors with culture-positive platelet donations?
Acute hypotensive transfusion reaction with concomitant use of Transfusion 2011; 51: 166268.
angiotensin-converting enzyme inhibitors: a case report and review 94 Katus MC, Szczepiorkowski ZM, Dumont LJ, Dunbar NM. Safety of
of the literature. Am J Ther 2012; 19: e9094. platelet transfusion: past, present and future. Vox Sang 2014;
70 Doria C, Elia ES, Kang Y, et al. Acute hypotensive transfusion 107: 10313.
reaction during liver transplantation in a patient on angiotensin 95 Murphy WG, Coakley P. Testing platelet components for bacterial
converting enzyme inhibitors from low aminopeptidase P activity. contamination. Transfus Apher Sci 2011; 45: 6974.
Liver Transpl 2008; 14: 68487.
96 Lin L, Dikeman R, Molini B, et al. Photochemical treatment of
71 Sihler KC, Napolitano LM. Complications of massive transfusion. platelet concentrates with amotosalen and long-wavelength
Chest 2010; 137: 20920. ultraviolet light inactivates a broad spectrum of pathogenic bacteria.
72 Dzik WH, Kirkley SA. Citrate toxicity during massive blood Transfusion 2004; 44: 1496504.
transfusion. Transfus Med Rev 1988; 2: 7694. 97 Zavizion B, Serebryanik D, Chapman J, Alford B, Purmal A.
73 Olinger GN, Hottenrott C, Mulder DG, et al. Acute clinical Inactivation of Gram-negative and Gram-positive bacteria in red cell
hypocalcemic myocardial depression during rapid blood transfusion concentrates using INACTINE PEN110 chemistry. Vox Sang 2004;
and postoperative hemodialysis: a preventable complication. 87: 14349.
J Thorac Cardiovasc Surg 1976; 72: 50311.

www.thelancet.com Vol 388 December 3, 2016 2835

 Review

98 Amsler L, Jutzi M. Haemovigilance annual report 2014. 109 Kirpalani H, Zupancic JA. Do transfusions cause necrotizing
Berne: Swissmedic, 2015. enterocolitis? The complementary role of randomized trials and
99 Narick C, Triulzi DJ, Yazer MH. Transfusion-associated circulatory observational studies. Semin Perinatol 2012; 36: 26976.
overload after plasma transfusion. Transfusion 2012; 52: 16065. 110 Mohamed A, Shah PS. Transfusion associated necrotizing enterocolitis:
100 Andrzejewski C Jr, Casey MA, Popovsky MA. How we view and a meta-analysis of observational data. Pediatrics 2012; 129: 52940.
approach transfusion-associated circulatory overload: pathogenesis, 111 Kenton AB, Hegemier S, Smith EO, et al. Platelet transfusions in
diagnosis, management, mitigation, and prevention. Transfusion infants with necrotizing enterocolitis do not lower mortality but
2013; 53: 303747. may increase morbidity. J Perinatol 2005; 25: 17377.
101 Lieberman L, Maskens C, Cserti-Gazdewich C, et al. A retrospective 112 El-Dib M, Narang S, Lee E, Massaro AN, Aly H. Red blood cell
review of patient factors, transfusion practices, and outcomes in transfusion, feeding and necrotizing enterocolitis in preterm
patients with transfusion-associated circulatory overload. infants. J Perinatol 2011; 31: 18387.
Transfus Med Rev 2013; 27: 20612. 113 Toy P, Gajic O, Bacchetti P, et al, and the TRALI Study Group.
102 Piccin A, Cronin M, Brady R, Sweeney J, Marcheselli L, Lawlor E. Transfusion-related acute lung injury: incidence and risk factors.
Transfusion-associated circulatory overload in Ireland: a review of Blood 2012; 119: 175767.
cases reported to the National Haemovigilance Oce 2000 to 2010. 114 Sayah DM, Looney MR, Toy P. Transfusion reactions: newer
Transfusion 2015; 55: 122330. concepts on the pathophysiology, incidence, treatment, and
103 Andrzejewski C Jr, Popovsky MA, Stec TC, et al. Hemotherapy prevention of transfusion-related acute lung injury. Crit Care Clin
bedside biovigilance involving vital sign values and characteristics 2012; 28: 36372, v.
of patients with suspected transfusion reactions associated with 115 Shaz BH. Bye-bye TRALI: by understanding and innovation.
uid challenges: can some cases of transfusion-associated Blood 2014; 123: 337476.
circulatory overload have proinammatory aspects? Transfusion 116 Silliman CC, Kelher MR, Khan SY, et al. Experimental prestorage
2012; 52: 231020. ltration removes antibodies and decreases lipids in RBC
104 Anderson KC, Weinstein HJ. Transfusion-associated graft-versus-host supernatants mitigating TRALI in vivo. Blood 2014; 123: 348895.
disease. N Engl J Med 1990; 323: 31521. 117 Vlaar AP, Juermans NP. Transfusion-related acute lung injury:
105 Rhl H, Bein G, Sachs UJ. Transfusion-associated graft-versus-host a clinical review. Lancet 2013; 382: 98494.
disease. Transfus Med Rev 2009; 23: 6271. 118 Van der Linden P, Lambermont M, Dierick A, et al, for the Working
106 Ohto H, Yasuda H, Noguchi M, Abe R. Risk of transfusion-associated Group of the Superior Health Council. Recommendations in the
graft-versus-host disease as a result of directed donations from event of a suspected transfusion-related acute lung injury (TRALI).
relatives. Transfusion 1992; 32: 69193. Acta Clin Belg 2012; 67: 20108.
107 Treleaven J, Gennery A, Marsh J, et al. Guidelines on the use of 119 Toy P, Popovsky MA, Abraham E, et al, for the National Heart, Lung
irradiated blood components prepared by the British Committee for and Blood Institute Working Group on TRALI. Transfusion-related
Standards in Haematology blood transfusion task force. acute lung injury: denition and review. Crit Care Med 2005;
Br J Haematol 2011; 152: 3551. 33: 72126.
108 Marschner S, Fast LD, Baldwin WM 3rd, Slichter SJ, Goodrich RP.
White blood cell inactivation after treatment with riboavin and
ultraviolet light. Transfusion 2010; 50: 248998.

2836 www.thelancet.com Vol 388 December 3, 2016