When should a hospitalized patient be transfused?

Observational studies have raised concerns by linking morbidity and mortality to red blood cell use. Among 1,958 surgical patients who refused blood transfusion on religious grounds, there was an increase in mortality when hemoglobin levels were <6.0 g/dL. Hemoglobin levels higher than 7.0 g/dL showed no increased mortality.11 A recent comprehensive review included 272,596 surgical, trauma, and ICU patients in 45 observational studies. The review included studies with end points, including mortality, infections, multiorgan dysfunction syndrome, and acute respiratory distress syndrome, and concluded transfusions are associated with a higher risk of morbidity and mortality.12 (see Figure 1, p. 20)

Higher rates of infection associated with transfusions occurred in patients with post-operative trauma, acute injuries, gastrointestinal cancer undergoing surgery, coronary bypass surgery, hip surgery, burns, critical illness, and patients requiring ventilation. (see Figure 2, p. 21)12 The increased infection risk likely is due to the transient depression of the immune system induced by red blood cell transfusion. Prolonged hospital stays in postoperative colorectal surgery patients and ICU patients have been associated with transfusions.13

^{click for large version}

Figure 1: Association between blood transfusion and the risk of death (odds ratio [OR ] and 95% confidence

A meta-analysis of the few randomized controlled trials favors the restrictive use of red blood cells. The preponderance of the evidence comes from the Transfusion Requirements in Critical Care (TRICC) trial.14 This randomized control trial in critically ill medical and surgical patients demonstrated a restrictive strategy (transfusion trigger of <7.0 g/dL) and was as effective as a liberal transfusion strategy (transfusion trigger <10.0 g/dL). (see Figure 3, p. 22) Indeed, patients in the restrictive arm of the trial, who were less ill and under age 55 had a lower mortality rate than those who were transfused liberally.15 To date, there are no hospital-based randomized control trials that evaluate outcomes of anemic non-ICU medical patients.

This evidence has created a growing consensus that a restrictive use of blood results in improved patient outcomes. In patients without cardiovascular disease the evidence suggests most patients tolerate a hemoglobin level of 7.0 g/dL.5

Cardiac Patients

Experimental and clinical evidence suggests patients with cardiovascular disease are less tolerant of anemia. Patients with coronary disease are more likely to have adverse outcomes than those without coronary disease, if they do not have a red blood cell transfusion.11,16

^{click for large version}

Figure 2: Association between blood transfusion and the risk of infectious complications (odds ratio [OR ])

The myocardium has a higher oxygen extraction ratio compared to the tissue oxygen extraction ratio, making it more sensitive to anemia.17,18 The presence of cardiac disease may require a higher threshold to transfuse blood; however, the precise recommended threshold remains controversial. A restrictive red blood cell transfusion strategy (maintaining the hemoglobin between 7.0 g/dL and 9.0 g/dL) appeared to be safe in most critically ill patients with cardiovascular disease.14

The data is more conflicting for patients with an acute coronary syndrome (ACS). Some studies have found increased mortality and another concluded ACS decreased with red blood cell use.19-21 Further research is needed to determine when red blood cells should be given to patients with coronary disease.

Gastrointestinal Bleeding

The decision to transfuse for gastrointestinal (GI) bleeding takes into account the site and etiology of the bleeding, availability of treatments, and risk of continued bleeding. Once the blood loss is controlled, a decision must be made on how to treat the anemia. Currently, no studies have looked at outcomes for patients who did and did not receive blood for an acute or chronic GI bleed.