Patients are more likely to die from an accident on the expressway traveling to a hospital than from a blood transfusion once they're admitted, said Don Siegel, MD, PhD, during his Internal Medicine Meeting 2015 session on transfusion medicine.
“You need to put these things in perspective when you're discussing transfusions' risks with patients,” said Dr. Siegel, who is a professor of pathology and laboratory medicine at the University of Pennsylvania in Philadelphia.
Rare as they are, though, transfusion reactions do still occur, and he offered tips on how hospitalists can identify and treat their patients' negative responses to transfused blood.
Acute hemolytic transfusion reactions are fairly rare, but the consequences can be severe. They occur when patients are given blood that is not a match to their blood type. The cause is usually a clerical mistake, rather than a laboratory one, according to Dr. Siegel. “Either the blood that was collected initially from the patient for the type and screen was mislabeled—it was for a different patient—or when it came time to actually getting the bag, you transfused it to the wrong patient,” he said.
When this happens, patients can display any of a number of symptoms, including fever, chills, respiratory distress, hyper- or hypotension, and pain. “If a patient is having 1 of these symptoms, the transfusion should be stopped immediately, and you do an immediate clerical check just to make sure it wasn't the wrong unit for the patient,” said Dr. Siegel.
The transfusion bag and a sample of the patient's blood should be sent back to the blood bank for investigation, along with a sample for urinalysis to the appropriate laboratory, and the patient should be treated based on symptoms, which often primarily affect the kidneys.
A patient who develops a fever during transfusion is more likely to be having a febrile nonhemolytic transfusion reaction, which is a response to dead and dying white cells or cytokines that were released from white cells in transfused blood during storage. “Blood is universally leukoreduced these days, so the incidence of these febrile transfusion reactions has diminished. However, they still do happen,” said Dr. Siegel. Often the febrile response is related to the patient's underlying disease and just coincident with the transfusion.
Because fever can be a sign of either type of reaction, the transfusion should be stopped and a workup should be completed if a patient's temperature increases by 1 °C during a transfusion.
Patients can also have allergic reactions to transfusions. “If a patient happens to have a peanut allergy, and the blood donor had a Reese's Peanut Butter Cup on their way to the Red Cross, there can be peanut antigens in the donor's blood,” Dr. Siegel said.
If the allergy is severe, the reaction can be fatal, and blood donations are not screened for such issues. “It's just bad luck if it happens. Any blood component can cause an allergic reaction,” he said.
Someone could also be more generally allergic to donor blood. “The patient may be IgA deficient and have anti-IgA antibodies,” said Dr. Siegel. Testing for this allergy requires looking not only for the overall IgA level and antibodies in the patient but also deficiencies of only certain patient IgA subtypes. “It can be a little complicated to work up the patient for IgA deficiency. If the patient does have that, then the blood bank will need to order products for your patient that are collected from donors who are IgA deficient,” he said.
Having cellular products washed free of donor IgA with saline before transfusion is another option, but it is time-consuming and plasma products cannot be washed. Minor allergic reactions should be treated with an antihistamine, while anaphylactoid reactions can be treated with epinephrine, then prednisone (although transfusionists should follow their institutions' transfusion policies for dealing with such reactions, Dr. Siegel noted).
Bacterial contamination of blood products can cause very serious reactions, including sepsis. These are most often associated with platelet transfusions because platelets need to be stored at room temperature, as opposed to red cells and plasma which are stored refrigerated and frozen, respectively. New requirements for culturing donor platelets or employing other methods to reduce or detect bacterial contamination have dramatically reduced the incidence of these reactions, which had been the top cause of mortality from transfusion, Dr. Siegel said.
With better procedures in place to reduce the incidence of acute hemolytic transfusion reactions and the bacterial contamination of blood products, the transfusion reaction that should be highest on the physician's radar is transfusion-related acute lung injury (TRALI). “It's the leading cause of transfusion-related mortality reported to the FDA,” said Dr. Siegel. “It's a transfusion reaction that is not really anyone's fault. There's no practical way to assure that it won't happen in a given patient.”
The causes of TRALI are not fully understood, but it generally occurs when donor plasma has human leukocyte antigen (HLA) antibodies that happen to match the HLA type of the transfused patient. It appears that some other stressor combines to trigger a reaction that damages the lungs, Dr. Siegel said.
Women who have been pregnant are much more likely to have formed HLA antibodies, so blood collection facilities have instituted policies of only using male donors' plasma for the production of fresh frozen plasma. Additionally, cases of TRALI are tracked back to the donor, who is then excluded from donation in the future.
These techniques help prevent future cases, “but it's not going to do your particular patient any good,” noted Dr. Siegel. Almost all patients with TRALI require oxygen support, but 80% recover within 48 to 96 hours. In addition to oxygen, patients should be treated with respiratory and volume support and pressors if needed to raise blood pressure.
TRALI should not be treated with diuretics, but it is, in some cases of misdiagnosis. “One of the confusing things is distinguishing this from volume overload from the transfusion: transfusion-associated circulatory overload (TACO),” said Dr. Siegel. “The signs and symptoms are often very difficult to tell whether it's TRALI or TACO, and in fact both can occur at the same time in the patient.”
Clues to differentiating the conditions include that hypertension is more constant in TACO and transient in TRALI. Also, TACO patients' pulmonary edema is more likely to rapidly improve with diuresis. Some think that B-type natriuretic peptide testing can be helpful, but that's controversial, according to Dr. Siegel.
One consequence of transfusion that clinicians and patients should put into perspective is that of viral transmission by transfusion. “The incidence nowadays of HIV or hepatitis C being transmitted by transfusion is very, very low. It's probably less than 1 in 3 million,” said Dr. Siegel. Hepatitis B is transmitted in about 1 out of 1 million to 1.5 million transfusions, he estimated. “There are nucleic acid-based tests that have really closed down the time window of where the donor could be infected with the virus but not have mounted antibodies to the virus that would be detected,” he said. However, there are emerging transfusion-transmitted pathogens for which blood is not currently tested, including dengue virus and the Babesia parasite.
While physicians can rarely do anything to help their patients avoid the most common transfusion reactions, Dr. Siegel did have some tips on optimizing transfusion-related care. He urged his audience to avoid these common errors.
Ordering the wrong dose of platelets. “There are 2 different kinds of platelet products, and some hospitals have both kinds,” said Dr. Siegel. Whole blood-derived platelets are called “a unit of platelets” and are typically dispensed in a pack of 4, 5, or 6 units. “When you order 6 units of platelets, you'll get a pool of them together in 1 bag and it should be a sufficient dose to raise a patient's platelet count as much as 30,000 per microliter or more,” he said.
The other type is apheresis-derived platelets. These platelets offer the advantages of coming from a single donor and being less likely to be contaminated with bacteria. However, the terminology of dosing can be confusing. “A bag of apheresis platelets, by the FDA labeling, is also called a unit of platelets, but it's really equivalent to a pack of 6 units of whole blood-derived platelets in terms of number of platelets,” said Dr. Siegel.
Thus, ordering platelets by the unit can create confusion about how much is needed. “What we've done at our hospital is just not gotten involved in that and we have clinicians order platelets by the dose. We let them leave it up to us in the blood bank whether to fill the dose with a unit of apheresis platelets or a pool of 6 whole blood-derived platelets,” he said.
Ordering plasma to correct a high international normalized ratio (INR). “The INR of fresh frozen plasma is not 1.0 as one might assume. It's not necessarily 1.1 or 1.2 or 1.3,” said Dr. Siegel. Plasma can have an INR as high as 1.6. “We often get orders for plasma for a patient who has an INR of 1.4 or 1.5 and they want to fix it. In fact, if you gave plasma, it might even make the INR worse,” he said.
Assuming leukoreduction protects immunocompromised patients. Leukoreduction removes enough white cells to reduce the incidence of febrile transfusion reactions, HLA antigen sensitization to the patients, and transmission of cytomegalovirus infection, but will not render cellular blood products safe from causing transfusion-associated graft-versus-host disease. Immunosuppressed patients require additional precautions. “It's not good enough to use just leukoreduced blood. The blood needs to be irradiated, which is something you may need to specifically order from your transfusion service, or in our institution, we have a policy to irradiate all red cell and platelet components so that we don't have to rely on the clinician appropriately ordering irradiated components,” said Dr. Siegel.
Being impatient with the blood typing and screening process. “Sometimes blood is available in 20 minutes or half an hour and other times it seems to take an hour or 2 or more. You can't understand: What is the big deal?” said Dr. Siegel. Blood typing itself is a pretty straightforward process—about 10 to 20 minutes, he reported.
But the next step is the antibody screen. If the patient's blood is negative for antibodies to minor blood group antigens, then a match for ABO and Rh can be found and blood is sent to the bedside quickly. However, if the antibody screen is positive, the process can be much lengthier. “All bets are off if the antibody screen is positive, because then you have to identify the specific antibody or antibodies,” said Dr. Siegel. Once the antibodies are identified and blood lacking the corresponding antigens is found, the donor blood will also have to be physically cross-matched with the patient's sample. The whole process takes some time for blood bank staff to complete. “It's not done by robot or machine,” noted Dr. Siegel.