Transfusion Thresholds in ICU

Dr Swapnil Pawar November 3, 2019 2779 1 5

share close

Dr. Jose Chacko & Dr. Swapnil Pawar

Anaemia is a common condition associated with critically ill patients admitted to an intensive care unit (ICU). The World Health Organization (WHO) defines anaemia as a haemoglobin (Hb) level less than 13 g/dL(milligram per decilitre) in men and less than 12 g/dL in women. Various studies have shown that approximately two-thirds of critical patients admitted to an ICU have a haemoglobin concentration of less than 12 g/dl on the day of admission, and 97% of the patients become anemic after a week in ICU.

The cause of anemia in critical illness is complex and often multifactorial.  Primary mechanisms include:

 a. loss of red blood cells (RBCs) due to phlebotomy/ frequent blood sampling and bleeding from the surgical site, trauma, venous access site, or gastrointestinal bleed.

 b. decreased production of RBCs due to suppression of bone marrow secondary to inflammatory cytokines, drugs, functional or absolute erythropoietin deficiency due to renal dysfunction.

c. nutritional (iron, folic acid, vitamin B 12) deficiency.

d. increased destruction of RBCs (hemolysis) or RBC precursors in the bone marrow due to toxins and drugs.

RBC transfusions remain an important treatment for improving tissue oxygenation in the setting of inadequate oxygen-carrying capacity. However, cautious use of allogeneic pRBC transfusions is warranted given that the purported benefits of increased oxygen delivery may not always outweigh the potential harms due to risks of volume overload, transfusion reactions, acute bacterial infection, and chronic infectious disease transmission. In addition, the 75 million units of pRBCs collected worldwide per annum are costly and limited in supply, given the need for administrative systems to screen donors, process whole blood, and store the collected products. 

The management of anaemia and its impact on critically ill patients has been a matter of debate.

Why should we worry about anemia in ICU patients?

Oxygen bound to hemoglobin is carried by RBCs to the tissues; any drop in the hemoglobin level may potentially impair oxygen delivery. Compensatory mechanisms lead to an increase in cardiac output, resulting in increased oxygen delivery; besides, tissues respond with an increase in oxygen extraction. Healthy young adults may tolerate hemoglobin levels of up to 4 gm/dl in the absence of hypovolemia. However, in critically ill patients and in those with underlying comorbidities, compensatory mechanisms may be inefficient. The hemoglobin level at which compensatory mechanisms fail, leading to impaired oxygen consumption may be considered to the “critical” level at which RBC transfusions may be required.

what’s the rationale of using different cut-offs to treat anemia in critically ill patients?

In a clinical setting, it is difficult to ascertain the critical hemoglobin level below which oxygen delivery may be significantly compromised. We go by cut-off levels largely because hemoglobin levels are easy to measure and contribute significantly to global oxygen delivery. However, the hemoglobin level may not be efficacious in determining organ-specific oxygen delivery, for instance, to the ischemic myocardium. Besides, we do not have a reliably reproducible technique to measure the degree of cellular hypoxia. The critical level of oxygen delivery may also be variable among different subgroups of patients and different organ systems. Considering these practical difficulties that we encounter in clinical practice, arbitrary thresholds have been proposed for prophylactic RBC transfusion among critically ill patients. 

A recent meta-analysis published in Critical Care Medicine suggests that the safety of restrictive transfusion thresholds differs for critically ill patients versus perioperative patients. In critical care patients, a restrictive transfusion strategy (transfusion trigger ~7–8 g/dL in the majority of studies) significantly reduced mortality, stroke/TIA, transfusion reactions, allogeneic blood exposure, and hospital length of stay. In contrast, for perioperative patients, current evidence suggests that a restrictive transfusion strategy may increase the risk of mortality, although the overall results were not statistically significant and lack of power prevents definitive conclusions.

In sum, these numbers suggest that for every 1,000 critical care patients treated with a restrictive transfusion strategy instead of a liberal strategy, a total of 31 (95% CI, 6–55) fewer patients would die at 30 days, 13 (95% CI, 1–21) fewer patients would experience stroke/TIA, 16 (95% CI, 7–22) fewer would experience transfusion reactions, 663 (95% CI, 429–771) fewer would be exposed to allogeneic RBC transfusion, and more than 1,000 hospital days would be saved. In contrast, in perioperative patients, the evidence for mortality is less definitive due to lower power.

Let’s talk about GI bleed patients. What’s the evidence out there in favor of or against transfusing at Hb <7gm/dl vs Hb < 9 gm/ dl

Several observational and controlled studies had previously suggested that a restrictive strategy did not worsen or may even improve survival in upper GI bleed. In a more recent Spanish, single-center study, patients with active acute upper GI bleed were randomized to a restrictive trigger of 7 gm/dl vs. a liberal strategy of 9 gm/dl. About 50% of the bleed was due to peptic ulcer, followed by the variceal bleed. The primary outcome of the all-cause mortality at 45 days, was significantly lower in the restrictive strategy group; rebleeding was less with the restrictive strategy; adverse events were also less with a restrictive strategy. On subgroup analysis, the probability of survival was significantly higher in the subgroup of patients with cirrhosis and Child-Pugh class A or B disease (hazard ratio, 0.30; 95% CI, 0.11 to 0.85), but not in those with cirrhosis and Child-Pugh class C disease (hazard ratio, 1.04; 95% CI, 0.45 to 2.37). Among patients with peptic ulcer bleed, the survival was not significantly different between groups. 

What about septic shock patients? What’s the evidence?

Rivers et al. in their landmark trial on early goal-directed therapy (EGDT) evaluated the efficacy of a “bundled” approach in patients with severe sepsis and septic shock. One of the interventions in the bundle was RBC transfusion to a target of 10 gm/dl if the central venous oxygen saturation remained less than 70%. Patients who received early goal-directed therapy (EGDT) had significantly better clinical outcomes. This led to the widespread use of a liberal transfusion strategy during the early phase of resuscitation in patients with septic shock. However, as this approach included a combination of several interventions, the effectiveness of individual components of the bundle was difficult to discern. 

The ProCESS, ARISE, and ProMISe trials compared EGDT to protocolized or usual care. The control arms of these studies used a lower hemoglobin threshold, with reduced transfusion requirements; however, there was no significant difference in clinical outcomes in patients who received protocolized or usual care. These trials suggested that transfusion of RBCs aiming for higher hemoglobin thresholds may not improve outcomes during the early phase of sepsis resuscitation. The Transfusion Requirements in Septic Shock (TRISS) trial provided more robust evidence supporting a restrictive strategy. This study included 998 patients with septic shock who were randomized to receive leukoreduced RBC transfusions at thresholds of 7 and 9 gm/dl in the restrictive and liberal arms, respectively. During ICU stay, the restrictive group received a median of one unit of RBCs (interquartile range: 0–3), compared to four (interquartile range: 2–7) in the liberal group. The primary endpoint, the 90-d mortality, did not differ between groups. Furthermore, there was no difference in the incidence of ischemic events, severe adverse reactions, or the requirement for life support. In light of these studies, there is robust evidence to support a restrictive transfusion strategy in patients with septic shock.

And lastly cardiothoracic patients – what’s the evidence

Anemia is generally considered to be associated with poor outcomes following cardiac surgery. However, acute kidney injury, infections, cardiac, respiratory, and neurological complications have been shown to be more common in patients who are transfused after cardiac surgery compared to those who do not receive transfusions in the postoperative period. The Transfusion Requirements After Cardiac Surgery (TRACS) study (2010) randomized 502 patients who underwent cardiac surgery under cardiopulmonary bypass; RBCs were transfused postoperatively to maintain a target hematocrit level of 30 vs. 24%.

The composite primary outcome included 30-day all-cause mortality, acute respiratory distress syndrome, cardiogenic shock, and acute kidney injury requiring renal replacement therapy. The primary endpoint did not differ between groups; besides, the number of units transfused was found to be an independent predictor of death or the development of clinical complications at 30 days. 

In a multicentric study involving 17 centers in the United Kingdom, 2007 post-operative cardiac surgical patients were randomized to receive transfusions at a threshold hemoglobin level of 9 gm/dl vs. 7.5 gm/dl. Patients were evaluated for a composite primary outcome including serious infection, or an acute ischemic event including stroke, acute myocardial infarction, gut ischemia, or acute kidney injury up to 3 months post-randomization. Patients randomized to the liberal group had nearly twice as many transfusions as those who were randomized to the restrictive group. The composite primary outcome was not significantly different between groups. The 90-d mortality, a secondary outcome, was higher with the restrictive strategy. Regardless of the transfusion strategy, the number of units transfused was an independent predictor of death or clinical complications at 30 days.

In a more recent, larger randomized controlled trial of cardiac surgical patients, a restrictive strategy (transfusion trigger 7.5 gm/dl beginning with the induction of anesthesia) was compared to a more liberal strategy (transfusion trigger  <9.5 gm/dl in theatre or intensive care ICU or <8.5 gm/dl in the ward) (Mazer et al., 2017) The composite primary outcome included all-cause mortality, acute myocardial infarction, stroke, or acute kidney injury requiring dialysis at hospital discharge or by day 28. A restrictive strategy was found to be non-inferior to a more liberal strategy in this study. Based on the available evidence, a restrictive strategy with the maintenance of hemoglobin around 8 gm/dl appears to be safe in patients undergoing cardiac surgery. 

These studies provide fairly robust evidence that compared to conventionally followed triggers, a lower threshold of around 8 gm/dl may be appropriate following cardiac surgery. 

Thanks, Dr. Chacko. There is another large RCT called FOCUS, which compared Hb < 8gm/dl vs Hb> 10gm/dl in patients who underwent surgery for hip fracture and who had a history of risk factors for cardiovascular diseases and found that liberal transfusion strategy did not reduce rates of death or inability to walk independently at 60-day follow up in elderly patients at high risk of cardiovascular risk.

However, there are two recent systematic reviews in Cardiac surgery and in perioperative transfusion practice ( including cardiac surgery) reported higher mortality with restrictive transfusion thresholds. A systematic review by Prof Tim Walsh in Critical Care 2017 found only 11 blood transfusion threshold RCTs that included patients with co-existing cardiovascular diseases. They also reported no evidence of a difference in 30-day mortality, however, they reported an increased risk of ACS in patients with co-existing CVDs who were randomised to restrictive strategies. 

Most of the guidelines in different countries recommend not to transfuse if Hb > 10. They recommend transfuse if Hb < 8 in CVD..but there is a gray zone between 8-10. Audits of blood transfusion practices in the UK have shown that 20% of blood product usage is outside guideline recommendations. 

What we need to understand is that there is a spectrum of severity of both CVD and critical illness, and it follows that the balance of risks and benefits of transfusion may change along the spectrum.


Hébert PC, Martin C, Yetisir E. A Multicenter, Randomized, Controlled Clinical Trial of Transfusion Requirements in Critical Care. The New England Journal of Medicine. 1999:9.

Holst LB, Haase N, Wetterslev J, Wernerman J, Guttormsen AB, Karlsson S, et al. Lower versus Higher Hemoglobin Threshold for Transfusion in Septic Shock. N Engl J Med. 2014;371:1381-1391. doi:10.1056/NEJMoa1406617

Villanueva C, Colomo A, Bosch A, et al. Transfusion Strategies for Acute Upper Gastrointestinal Bleeding. N Engl J Med. 2013;368(1):11-21. doi:10.1056/NEJMoa1211801

Hajjar LA, Vincent JL, Galas FR, Nakamura RE, Silva CM, Santos MH, et al. The TRACS Randomized Controlled Trial. JAMA. 2010;304:1559-1567. doi: 10.1001/jama.2010.1446.

Murphy GJ, Pike K, Rogers CA, et al. Liberal or Restrictive Transfusion after Cardiac Surgery. N Engl J Med. 2015;372(11):997-1008. doi:10.1056/NEJMoa1403612

Mazer CD, Whitlock RP, Fergusson DA, et al. Restrictive or Liberal Red-Cell Transfusion for Cardiac Surgery. N Engl J Med. 2017;377(22):2133-2144. doi:10.1056/NEJMoa1711818

Rate it
Previous episode
CRASH 3 – A Fairy Tale of Victory of Tranexamic acid & Dr Utako Okamoto
eCritCare Podcast
share playlist_add
  • 2111
  • 1


CRASH 3 – A Fairy Tale of Victory of Tranexamic acid & Dr Utako Okamoto

Dr Swapnil Pawar October 26, 2019

Effects of tranexamic acid on death, disability, vascular occlusive events and other morbidities in patients with acute traumatic brain injury (CRASH-3): a randomised, placebo-controlled trial Intro –  As we know, […]

Read more trending_flat

Similar episodes