When you begin to look at blood transfusions as liquid tissue transplants, you realize that this routine procedure bears real risks and should be handled accordingly.
For years I have given a lecture concerning the safety of blood transfusions. In it I point out the rare likelihood of catching some nasty viral disease, the fact that ABO incompatibility issues resulting in hemolytic transfusion reactions are also rare and always the result of human error, and that the litany of more mild reactions are very infrequent.
In short, the lecture detailed the disconnect between the apparent safety of blood transfusions and the big fuss that hospitals make regarding the consent procedure for giving blood. By making such a formal consent procedure for getting patient’s permission to receive blood, hospitals send the message that getting blood must be dangerous.
As it turns out, however, I may have been overlooking a subset of real risks associated with blood transfusions. When blood administration is considered a liquid tissue transplant there is a whole range of new considerations that come into play. Think of it – you are giving living tissue from one person to another (and each unit represents a different source of the transplant). And, consistent with organ or tissue transplants, there is a whole procedure to determine if the tissue is compatible with that of the recipient – the typing and crossmatching done in the lab.
The literature on this aspect of blood transfusions is fascinating. For example, below is a meta-analysis comparing more restrictive transfusion thresholds (<7gm/ dL) to more liberal ones. The findings are remarkable. Three randomized trials involving a total of 2,364 patients found that essentially all outcomes were better with a restrictive vs. a liberal transfusion policy.
Specifically the restrictive policy was associated with a reduction in inpatient, 30-day and total mortality (risk ratios [RRs] 0.74, 0.77 and 0.80, respectively), as well as in the development of acute coronary syndrome (RR 0.44), pulmonary edema (RR 0.48), rebleeding (RR 0.64) and bacterial infection (RR 0.86). In these three trials, transfusions were given to 55% of patients randomized to the restrictive strategy vs. 94% of those randomized to liberal man- agement. The number needed to treat with a restrictive rather than a liberal policy to prevent one death was 33. Given the milions of transfusions administered in the U.S. every year, the number of preventable deaths is considerable (the Red Cross says that every 2 seconds someone needs blood in the U.S. and there are about 30 million blood component transfusions each year and the average transfusion is three units).
In 16 other studies (4,572 patients) in which a less restrictive transfusion strategy (hemoglobin trigger 7.5-10.0g/dL) was compared with a liberal strategy, there was a reduction in exposure to blood transfusion (RR 0.60), but no statistical effect on mortality, ACS, pulmonary edema or bacterial infection (so the goal needs to be a trigger of less than 7.5 gm/dL).
IMPACT OF MORE RESTRICTIVE BLOOD TRANSFUSION STRATEGIES ON CLINICAL OUTCOMES: A META-ANALYSIS AND SYSTEMATIC REVIEW
Saltpeter, S.R., et al, Am J Med 127:124, 2014
BACKGROUND: It has traditionally been assumed that administration of red blood cells, to reverse anemia, improves outcomes, but there is no definitive evidence to support this claim. Recent trials have suggested that limiting transfusions by lowering the hemoglobin transfusion trigger might be beneficial.
METHODS: The authors, from Stanford and Brown Universities, performed a meta-analysis of randomized controlled trials comparing outcomes after RBC transfusion using a restrictive vs. a liberal hemoglobin trigger.
RESULTS: In three trials (2,364 adult or pediatric critical care patients or patients with upper gastrointestinal bleeding) comparing a very restrictive policy (hemoglobin trigger of less than 7g/dL) with a more liberal policy, the restrictive policy was associated with a reduction in inpatient, 30-day and total mortality (risk ratios [ORs] 0.74, 0.77 and 0.80,respectively), as well as in the development of acute coronary syndrome (ACS) (RR 0.44), pulmonary edema (OR 0.48), rebleeding (RR 0.64) and bacterial infection (RR 0.86). In these three trials, transfusions were given to 55% of patients randomized to the restrictive strategy vs. 94% of those randomized to liberal management. In 16 other studies (4,572 patients) in which a less restrictive transfusion strategy (hemoglobin trigger 7.5-10.0g/dL) was compared with a liberal strategy, there was a reduction in exposure to blood transfusion (RR 0.60), but no statistical effect on mortality, ACS, pulmonary edema or bacterial infection.
CONCLUSIONS: In three studies of a very restrictive transfusion strategy, using a hemoglobin below 7g/dL as the trigger for RBC transfusion led to improved patient outcomes.
for reprints) (PMID: 24331453)
Copyright 2014 by Emergency Medical Abstracts – All Rights Reserved 6/14 - #6
The assumption can always be made that the sicker patients were the ones to receive transfusion while the less sick did not get them (or as much) and this may have resulted in the worse outcomes, however, the three trials cited above were randomized which supposedly would eliminate this obvious source of bias. Another advantage to a more restrictive transfusion policy is that there was a 40% reduction in the number of patients receiving blood with an average of 2 units less per person.
The authors of the above analysis note 11 papers citing worse outcomes in patients with anemia due to illness or bleeding, compared with simple supportive measures, such as hydration. In addition, seven observational studies are cited that consistently demonstrate that transfusions are associated with an increased risk of ad- verse events after controlling for potential confounding variables, even when using a restrictive transfusion strategy.
Four studies are cited that make the point that the increased risk seems to be directly proportional to the amount of blood transfused and the length of storage of the transfused red cells. One theory is that there may be an inflammatory response to the infused blood product.
Blood is transfused with the intent of increasing the oxygen carrying potential and thus make more oxygen available to vital organs. Three studies are cited that found no significant improvement in oxygen delivery when compared with supportive care, despite an increase in blood oxygen content. It is felt this problem is the result of increased blood viscosity associated with transfusions along with a loss of red cell function as the result of preservation and storage. The authors assert, based on four studies, that normovolemic anemia with hemoglobin levels of 5-6gm/dL is well tolerated in cardiovascular and critical illness and may have beneficial hemodynamic effects.
The authors conclude that “so far, there is little trial evidence that blood transfusions significantly improve oxygen delivery or clinical outcomes in any setting or with any nadir hemoglobin level.”
What does the American Society of Anesthesiologists have to say about restrictive blood transfusions? As one of their five items in the Choosing Wisely initiative the anesthesiologists say:
Don’t administer packed red blood cells to a young, healthy patient without ongoing blood loss and a hemoglobin of 6gm/ dL unless symptomatic or hemodynamically unstable.
• The hemoglobin transfusion threshold used in multiple studies has varied from 6.0 to 10.0 g/dL.
• The optimal hemoglobin/hematocrit criterion for transfusion remains controversial in several clinical settings.
• Nevertheless, compared with higher hemoglobin thresholds, a lower hemoglobin threshold is associated with fewer red blood cell units transfused without adverse associations with mortality, cardiac morbidity, functional recovery or length of hospital stay.
• Hospital mortality remains lower in patients randomized to a lower hemoglobin threshold for transfusion versus those randomized to a higher hemoglobin threshold.
• The decision to transfuse should be based on a combination of both clinical and hemodynamic parameters.
What does the Society of Hospital Medicine – Adult Hospital Medicine have to say about blood transfusions?
Here is their Choosing Wisely recommendation.
Avoid transfusions of red blood cells for arbitrary hemoglobin or hematocrit thresholds and in the absence of symptoms of active coronary disease, heart failure or stroke.
The AABB (formerly the American Association of Blood Banks) recommends adhering to a restrictive transfusion strategy (7 to 8 g/dL) in hospitalized, stable patients.
The AABB suggests that transfusion decisions be influenced by symptoms as well as hemoglobin concentration.
According to a National Institutes of Health Consensus Conference, no single criterion should be used as an indication for red cell component therapy. Instead, multiple factors related to the patient’s clinical status and oxygen delivery should be considered.
Listed below are some additional papers from the EMA database that don’t put transfusions in a very positive light.
RED BLOOD CELL TRANSFUSION IS ASSOCIATED WITH INCREASED REBLEEDING IN PATIENTS WITH NONVARICEAL UPPER GASTROINTESTINAL BLEEDING
Restellini, S., et al, Aliment Pharmacol Ther 37(3):316, February 2013
BACKGROUND: Evidence is scarce for guiding effective use of red blood cell (RBC) transfusions in patients with nonvariceal upper gastrointestinal (UGI) bleeding.
METHODS: These multinational, principally European, authors conducted a retrospective study of a random sample of 1,677 patients undergoing endoscopy at 18 Canadian hospitals for nonvariceal UGI bleeding to assess the effects of red blood cell (RBC) transfusion on rates of rebleeding and mortality.
RESULTS: Most patients were male (62%) (mean age 66.2). The initial hemoglobin level was 9.7g/dl. Transfusion of RBCs was documented for just over half the patients (54%). The baseline hemoglobin was 8.2g/dl in patients who were transfused vs. 11.5g/dl in those who were not. The presence of shock on presentation was 43.5% in the former group vs. 18.4% in the latter. On univariate analysis, rebleeding occurred in 23.6% of the transfused patients vs. 11.3% of those who were not transfused (p<0.01) and all-cause 30-day mortality was higher in the former group (6.8% vs. 3.7%, P=0.005). On multivariate analysis that adjusted for potential confounders, RBC transfusion within 24 hours of presentation was associated with a significantly increased risk of rebleeding (odds ratio [OR] 1.8, p=0.005), but was not a significant predictor of 30-day mortality (OR 1.0).
– no reprints) Copyright 2013 by Emergency Medical Abstracts – All Rights Reserved 12/13 - #36
RED BLOOD CELL TRANSFUSION: A CLINICAL PRACTICE GUIDELINE FROM THE AABB
Carson, J.L., et al, Ann Intern Med 157(1):49, July 3, 2012
BACKGROUND: High-quality evidence on the benefits and harms of transfusion of red blood cells (RBCs) is limited, and there is substantial variability in transfusion practices.
METHODS: The authors report on a literature-based guideline for RBC transfusion developed by an expert panel convened by the AABB (formerly the American Association of Blood Banks). The guideline pertains to hospitalized hemodynamically stable patients.
RESULTS: There is high-quality evidence to support adherence to a restrictive rather than a liberal transfusion policy, with consideration of transfusion at a hemoglobin of 7g/ dL or lower in adult and pediatric ICU patients, and 8g/dL or lower (or development of symptoms such as chest pain, orthostatic hypotension, tachycardia unresponsive to fluid resuscitation or congestive heart failure [CHF]) in postoperative patients. There is moderate quality evidence to support consideration of transfusion of RBCs at a hemoglobin of 8g/ dL or lower, or development of symptoms, in patients with preexisting cardiovascular disease. Based on very low quality evidence, the panel cannot recommend for or against a liberal or restrictive transfusion policy in hospitalized hemodynamically stable patients with acute coronary syndrome (ACS). Finally, based on low-quality evidence the panel suggests that decisions regarding transfusion be influenced by both symptoms and hemoglobin concentrations.
for reprints) Copyright 2013 by Emergency Medical Abstracts – All Rights Reserved 8/13 - #8
TRANSFUSION STRATEGIES FOR ACUTE UPPER GASTROINTESTINAL BLEEDING
Villanueva, c., et al, n Engl J Med 368(1):11, January 3, 2013
BACKGROUND: There is some evidence to suggest that a restrictive transfusion strategy might be preferable to a liberal strategy in some situations.
METHODS: In this prospective Spanish study, 889 adults with acute upper gastrointestinal (GI) bleeding without massive exsanguinating hemorrhage were randomized to a restrictive or liberal transfusion strategy. The hemoglobin threshold for transfusion of red blood cells was 7g/dL in the restrictive strategy group (post-transfusion target 7-9g/dL) and 9g/dL in the liberal strategy group (post-transfusion target 9-11g/dL).
RESULTS: Baseline characteristics were similar in the two study groups; the hemoglobin on presentation was 9.6g/dL in the restrictive strategy group and 9.4g/dL in the liberal strategy group. When compared with the liberal strategy group, the restrictive group was significantly more likely to receive no transfusion (51% vs. 14%, p<0.001), received fewer transfused units (1.5 vs. 3.7, p<0.001), and had a lower rate of mortality at 45 days (5% vs. 9%, p=0.02, adjusted hazard ratio [HR] 0.55) and further bleeding (10% vs. 16%, p=0.01, HR 0.68). Patients randomized to a restrictive transfusion strategy had a shorter hospital stay (9.6 vs. 11.5 days, p=0.01) and a lower complication rate (40% vs. 48%, p=0.02, HR 0.73). Specifically, the restrictive strategy group had a significantly lower rate of transfusion reactions (3% vs. 9%, p=0.001) and cardiac complications (11% vs. 16%, p=0.04) than the liberal strategy group, but there were no significant differences between the groups in pulmonary complications, acute kidney injury, stroke or TIA, or bacterial infections.
for reprints) Copyright 2013 by Emergency Medical Abstracts – All Rights Reserved 7/13 - #12
When blood is considered a liquid tissue transplant then the fact that each unit comes from a different person can be seen to potentially worsen the effect produced by blood – particularly on the immune system. Here are some additional facts that suggest that blood should be given with particular caution:
• Patients receiving blood during cardiac surgery have almost twice the mortality of those who do not. The national incidence of transfusions as- sociated with bypass surgery ranges between 10% at some hospitals to and over 90% at others reflecting a lack of adequate concern regarding the use of blood in many hospitals.
• In a 2006 study of 11,963 coronary bypasses (Koch., C., Critical Care Medicine, 34(6), 1608-1616) the transfusion of red blood cells was associated with a greater risk of:
- ̊Renal failure (2.96 times)
- ̊Prolonged time on a ventilator (1.79 times)
- ̊Mortality (1.77 times)
- Serious infection (1.76 times)
- ̊Cardiac events (1.55 times)
- ̊Neurologic events (1.37 times)
• Another study by Koch (NEJM, 358: 1229, March 20, 2008) comparing 2,872 cardiac surgery patients receiving blood within two weeks of donation with 3,130 receiving blood donated more than two weeks prior to surgery found:
- In hospital mortality was higher (2.8% vs 1.7%) as was the duration of ventilatory support, the rate of renal failure, sepsis and multiorgan failure.
- Survival at one year was more common in those receiving the newer blood (92.6% vs 89%).
- The risk of infection after orthopedic surgery in patients who have received blood is 1.5-3.5 times greater than those who have not received blood
- Blood transfusions have been associated with an increased risk of postoperative infection, ventilator-acquired pneumonia, central line sepsis, increased ICU and hospital length of stay and short and long term mortality rates
- A study by Taylor, R.W., et al (Crit Care Med, 30 (10):2249, 2002) of 1,717 medical and surgical ICU pa- tients: ̊ Found a nosocomial infection rate of 15.4% in those transfused (416 patients) vs. 2.9% in those not transfused (1,301 patients).
- For each unit of blood transfused the infection rate increased by 50%.
- The mortality rate of the transfused patients was 24.3% vs 10.2% for those not transfused. This was not, however, a randomized study and transfused patient mortality may have been related to the underlying need for trans- fusions. This is harder to say for the rates of nosocomial infection. Given a dose-response risk of infection, it suggests that the admonition to give at least two units when a transfusion is ordered is challengable.
- • Transfusion-related immunomodulation has been well documented in association with blood transfusions for many years. It is reflected in a 7-10 fold increase in post-operative infections.
- In addition, increased cancer recurrence rates have been associated with blood transfusions and an increased five-year mortality rate.
Bottom line – it appears that the real risks of giving blood are higher than we appreciated and emergency physicians need to carefully weigh the benefits of giving blood with the substantial risks oullined above.
Richard Bukata, MD is Editor of Emergency Medical Abstracts (www.ccme.org)