Blood transfusion in the ICU

This chapter deals with the indications, contraindications and complications of blood transfusion in the ICU. The contents and properties of packed red blood cells, the physiology of acute haemorrhage and the physiological responses to blood transfusion are detailed in other chapters.

The National Blood Authority of Australia has issued a series of practice guidelines which can be helpful in guiding the shaking bloodied hand of the ICU practitoner. The guidelines have been endorsed by our college, which means that at least some of the Australian intensivists must agree with them.

The following entries are a summary of these clinical practice guidelines, to help answer past paper SAQs such as Question 21 from the second paper of 2007 and Question 13 from the first paper of 2003.

Transfusion in Critical Care

The critical care module of the NBA guidelines makes the following recommendations for critically ill patients who are not hosing blood out of every orifice:

Follow a restrictive transfusion strategy.

  • Be guided by the whole clinical picture rather than the numerical Hb value
  • That said, some numerical values are important to remember:
    • Transfuse patients with a Hb under 70
    • Don't transfuse patients with a Hb over 90
    • For the 70-90 group, be guided by the clinical significance of anaemia.
  • For those with acute coronary syndromes, the target Hb should be 80, and probably one should not aim any higher than that.

Tolerate a mild coagulopathy and thrombocytopenia for routine procedures.

The NBA makes the following statements about the use of separated blood products:

  • Don't use FFP for patients with an INR under 2.0. Even if you plan to insert lines or do some other sort of invasive procedures, the evidence demonstrates that it is safe to do so.
  • Don't give people "routine" cryoprecipitate to maintain a given fibrinogen concentration
  • Give platelets for a platelet count under 20. Patients with a platelet count over 50 can safely be subjected to normal ICU plumbing procedures.

The NBA loves tranexamic acid.

  • Use it as early as possible (preferably within the first 3 hours) for any acute traumatic bleeding, or bleeding in general. Though since the HALT-IT trial (Roberts et al, 2020) you probably woldn't use it for GI bleeding.
  • The NBA uses CRASH-2 trial dosing recommendations (1g bolus, followed by a 1g infusion over 8 hours).

Evidence behind the restrictive blood transfusion policy

Much of the above has been guided by the 1999 TRICC study, which observed the effects of transfusing 838 ICU patients to different Hb levels (70-90 vs 100-120). Though they did not find any 30-day mortality difference between the two groups, there was a significant in-hospital mortality difference (22% vs 28%) which favoured the restrictive transfusion strategy.

Since 1999, many more RCTs have contributed data to the growing impression that keeping Hb over 70 in most patients is a superior approach in terms of mortality and morbidity. A 2013 JAMA article provides a helpful synopsis of these results, and summarizes a total of over 6000 patient cases in support of restrictive transfusion practice. The lower haemoglobin threshold was associated with no harm, and may be associated with some sort of survival benefit.

A critical appraisal of routine blood transfusion

A "routine transfusion" in this setting is the transfusion which has a "numeric trigger" as opposed to a clinical indication (i.e. the patient is asymptomatic).

Rationale for some sort of strategy

  • Blood products are a scarce resource (and very expensive outside of an Australia-like socialised health care system)
  • Blood products are not without risk; specifically, there is a real nonzero risk of contracting a disabling infectious disease, or developing a life-threatening reaction.
  • Benefits of transfusion of stored red cells are unclear in this "borderline" population
  • Risk of transfusion needs to be weighed against benefit

Arguments in support of routine transfusion to a Hb of >70

  • Transfusion in the developed world is impressively safe.
  • Clinical features of anaemia may be insidious in onset, and it is possible to survive at a dramatically decreased haemoglobin - the argument is that this may not be the best quality of care for critically ill patients.
  • In the critically ill patients, the adverse effects of anaemia may be obscured by the ongoing critical illness.
  • Anaemia increases myocardial oxygen demand and reduced myocardial oxygen supply.
  • There are certain groups of critically ill patients for whom "routine" transfusion to a certain Hb level appears to be well supported by evidemce. These groups include patients with ongoing myocardial ischaemia and those undergoing ECMO.

Arguments against the routine transfusion to a Hb of >70

  • Anaemia with a poor systemic or myocardial oxygen supply/demand relationship is unlikely to be clinically silent (and nobody is arguing against transfusion for clinically significant anaemia)
  • There has never been any mortality benefit show for this practice in the critically ill population (outside of the acute coronary syndrome population)
  • Transfusion of stored cells contributes cells with a diminished oxygen carrying capacity
  • There are situations where exposure to foreign cells is to be avoided at all costs on medical grounds (eg. in patients being worked up for a transplant)

Practical approach to a questionably indicated transfusion

  • Need for consent is unquestioned (if the patient is exsanguinating, different story)
  • Involving the patient, ideally
  • Involving the family, as surrogates- in which case, validity of their involvement should be considered (eg. are these next of kin, or work friends?)
  • Cultural background needs to be considered (eg. Jehovah's Witness)
  • Age needs to be considered (underage? Covered by duty of care? The family may have little recourse, but the patient who is transfused against their will may be excommunicated from the cult).
  • If no cultural or ethical barriers, consider the blood bank may not supply the blood without an indication. Consider that "the haemoglobin's low" is not an indication. Will you manufacture an indication and lie, or will you engage with them on a theoretical level, talking about oxygen carrying capacity of blood and whatnot?

Alternatives to transfusion

Spahn et al (2013) have done an excellent review of this for  Lancet. In summary:

  • Transfuse something not quite blood, or less risky blood
    • Autologous transfusion (eg. pre-banked blood) - though usually the JWs do not go for this option
    • Blood surrogates (largely experimental)
  • Avoidance of blood loss
    • Advanced haemostasis measures intraoperatively, eg. topical haemostatic agents
    • Use of low volume blood tubes
    • Rationalisation of blood tests
  • Support of effected (or supra-normal) erythropoiesis
    • Good nutrition
    • Adequate micronutrients
    • Adequate haematinic vitamins
    • Avoidance of bone marrow toxins
    • Iron infusion
    • Erythropoietin


Goodnough, Lawrence T., Jerrold H. Levy, and Michael F. Murphy. "Concepts of blood transfusion in adults." The Lancet 381.9880 (2013): 1845-1854.

Spahn, Donat R., and Lawrence T. Goodnough. "Alternatives to blood transfusion." The Lancet 381.9880 (2013): 1855-1865.

There is also a rescinded document from the NHMRC (2001) which has been used to guide practice: Clinical Practice Guidelines on the Use of Blood Components.

To some extent this document has been superceded by the Australian and New Zealand Society of Blood Transfusion GUIDELINES FOR THE ADMINISTRATION OF BLOOD PRODUCTS.

The Patient Blood Management Guidelines from the National Blood Authority of Australia is another series of documents worth looking at - it contains several important modules which have been reviewed and which act as successors to the 2001 NHMRC guidelines.

Treleaven, Jennie, et al. "Guidelines on the use of irradiated blood components prepared by the British Committee for Standards in Haematology blood transfusion task force." British Journal of Haematology 152.1 (2011): 35-51.

Aoun, Elie, et al. "Transfusion‐associated GVHD: 10 years’ experience at the American University of Beirut—Medical Center." Transfusion 43.12 (2003): 1672-1676.

Heddle, Nancy M., and Morris A. Blajchman. "The leukodepletion of cellular blood products in the prevention of HLA-alloimmunization and refractoriness to allogeneic platelet transfusions [editorial]." Blood 85.3 (1995): 603-606.

Sharma, R. R., and Neelam Marwaha. "Leukoreduced blood components: Advantages and strategies for its implementation in developing countries."Asian journal of transfusion science 4.1 (2010): 3.

Dzik, Walter H. "Leukoreduction of blood components." Current opinion in hematology 9.6 (2002): 521-526.

Corwin, Howard L., and James P. AuBuchon. "Is leukoreduction of blood components for everyone?." JAMA 289.15 (2003): 1993-1995.

Blajchman, M. A. "The clinical benefits of the leukoreduction of blood products."Journal of Trauma-Injury, Infection, and Critical Care 60.6 (2006): S83-S90.

Rosenbaum, Lizabeth, et al. "The reintroduction of nonleukoreduced blood: would patients and clinicians agree?." Transfusion 51.12 (2011): 2739-2743.

Bilgin, Y. M., L. M. van de Watering, and A. Brand. "Clinical effects of leucoreduction of blood transfusions." Neth J Med 69.10 (2011): 441-450.

Australian Red Cross - Blood Service Policy on "The Age of Red Cells"

Hess, John R. "Red cell changes during storage.Transfusion and Apheresis Science 43.1 (2010): 51-59.

Bennett-Guerrero, Elliott, et al. "Evolution of adverse changes in stored RBCs."Proceedings of the National Academy of Sciences 104.43 (2007): 17063-17068.

Hébert, Paul C., et al. "A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care." New England Journal of Medicine340.6 (1999): 409-417.

Carson, Jeffrey L., Paul A. Carless, and Paul C. Hébert. "Outcomes using lower vs higher hemoglobin thresholds for red blood cell transfusion." Jama 309.1 (2013): 83-84.

Lelubre, C., J. L. Vincent, and F. S. Taccone. "Red blood cell transfusion strategies in critically ill patients: lessons from recent randomized clinical studies." Minerva anestesiologica (2016).

Spahn, Donat R., and Lawrence T. Goodnough. "Alternatives to blood transfusion." The Lancet 381.9880 (2013): 1855-1865.

Roberts, Ian, et al. "Effects of a high-dose 24-h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial." The Lancet 395.10241 (2020): 1927-1936.