A colleague in Southern California reports that her hospital transfusion service is reviewing policies for indications and use of the 'Biological Crossmatch'.  The physicians at her facility are split on the benefits of this procedure.  Some physicians feel it is important to use a biological crossmatch for patients who require transfusion with incompatible blood because of a warm autoantibody, especially if that antibody cannot be 'adsorbed out' to ensure no underlying alloantibodies are present.  These same physicians prefer a biological crossmatch be done for RBC units that are incompatible because of an unidentified alloantibody, if transfusion is critical.  On the other hand, some physicians do not prefer a biological crossmatch since they can order that these transfusions be administered slowing with careful patient monitoring. In addition, these physicians sign a waiver indicating that they take responsibility. Physicians who sign such waivers have not had any patient experience a problem with gross hemolysis over decades following this practice.

Her question is: Is there any reason to have a procedure for using a biological crossmatch (ie., transfusing a small amount of blood, collecting a blood sample and testing for hemolysis, before administering the rest of the unit).  How many facilities use this?  Of those that use this approach, has it ever been of benefit in identifying a hemolytic reaction?  And of course, are there data to show that this procedure is effective?  If this procedure is used, are facilities checking for hemoglobinuria as well as hemoglobinemia? The bottom line is that she would like to learn what others do in regard to biological crossmatch.

The following responses have been received.

ADDENDA Aug. 11, 2004

1. A Blood Bank Fellow in Los Angeles comments that she was trained to rely on biologic crossmatches for patients with autoantibodies for whom all RBC units were incompatible and underlying alloantibodies could not yet be ruled out. Of note, she points out that if it were not for written instructions (SOPs) on the exact role to be played by the clinical personnel caring for the patient during a biologic crossmatch, it would have been difficult to ensure that clinicians and nurses on the wards fully understood this procedure. In order to assure clear lines of communication, the blood bank fellow would make a copy of the biologic crossmatch SOP and place it on the patient's chart for reference. The medical team taking care of the pateint would appreciate having the reference provided.

ADDENDA Aug. 12, 2004

2. A transfusion medicine physician in Arkansas sees no value in performing a biological (in vivo) crossmatch for those with autoantibodies. He states that in his opinion "The most important consideration faced by the blood bank faces in transfusing patients with autoantibodies is the detection of underlying alloantibodies. If time permits, the appropriate adsorption study should be done to rule in or rule out their presence. I see no added benefit to performing an in vivo crossmatch, and see danger in relying on this procedure in lieu of the appropriate serological evaluation (i.e. adsorption study). The only antibody that the in vivo crossmatch will likely identify is one capable of causing immediate intravascular hemolysis. If we provide ABO group-compatible blood, which we should always do, then it is almost inconceivable that we would ever detect intravascular hemolysis due to an underlying alloantibody." He adds that in his experience "These IgG antibodies just aren’t capable of behaving in this manner. So, immunologically speaking, the in vivo crossmatch has never made sense to me. More than this though, I’m concerned that the biological crossmatch gives everyone a false sense of security and may serve as an excuse to not fully work up a patient in the appropriate manner."

He concludes that in his 10+ years of practice, he has never seen or heard of a positive in vivo crossmatch for which transfusion was halted, but he has been involved in the transfusion of hundreds of patients with autoantibodies who did not have an exacerbation of their hemolysis or, at least, any evidence of associated intravascular hemolysis. At each of the institutions where he has practiced, he has eliminated the biological crossmatch, and suggests that others consider doing the same.

ADDENDA Aug. 16, 2004

3. A colleague in Iowa reports that they will use a biological crossmatch when a physician insists that a patient must be transfused with incompatible RBCs yet the significance of the crossmatch incompatibility is not certain. As an example, she shares details about a case that they experienced several years ago. An elderly lady was admitted with a 6 g/dL hgb, and a physician ordered 3 RBC units for transfusion. A 'three-cell' antibody screen was negative, and antiglobulin crossmatches were compatible; the 3 ordered RBC units were transfused without incident. The patient was discharged, only to be readmitted two weeks later again with a low hemoglobin level. This time her antibody screen was positive and all reagent red cells of several different antibody identification panels reacted at varying strengths in all phases of testing, with some hemolysis at 37 degree incubation. The patient's direct antiglobulin test was weakly positive microscopically. The physician was advised not to transfuse the patient until the blood bank could complete their workup and determine the new antibody specificity (or specificities), as the patient was stable. The physician insisted on transfusing the patient and did not want to wait. A biological (in vivo) crossmatch was performed, during which the patient developed red plasma (the phlebotomy was atraumatic) suggesting a positive test. A blood specimen was sent to a reference lab which identified anti-Vel.

ADDENDA Aug. 17, 2004

4. An internationally respected immunohematologist from Los Angeles agrees with the remarks by the physician in Arkansas. In the immunohematologist's opinion, the biological crossmatch will only show the presence of an antibody capable of causing complement mediated intravascular lysis. Outside of ABO, such antibodies (allo or auto) are very rare. He adds that the correspondent from Iowa describes such an example (anti-Vel), but believes that they will probably never encounter another such situation for the rest of their career, even if they have 20 years left. In his opinion, such rare findings have to be balanced with the false sense of security that physicians have after getting a "negative" biological crossmatch. Most physicians do not realize that the most powerful Rh antibodies and most powerful Kell and Kidd antibodies will not be detected by the biological crossmatch. The immunohematologist very much agrees with the Arkansas colleague that the use of the biologic crossmatch test may serve as an excuse for an inadequate investigation. This is particularly so in the case of AIHA where the process to exclude alloantibodies is tedious and time consuming. The biological crossmatch is no substitution for adsorption (especially autoadsorption) procedures. If time does not allow this approach, then antigen-matching the patient's phenotype and/or a simple dilution procedure are more efficient than the biological crossmatch. He adds that none of these approaches are perfect and underlying alloantibodies can be missed, but these are rarely of great clinical significance. As the predictive value of the biological crossmatch is so low, he has to agree with the conclusions of the colleague from Arkansas, but if it has to be performed, it is vital that the policies/SOP contain a statement concerning the message to the physician discussing the inadequacies of this test in relation to the interpretation of the results [e.g., that most clinically significant (even strong Rh, etc.), antibodies will not be detected]. He would be interested to know if the SOP mentioned by the Blood Bank Fellow contains appropriate warnings to the physician.

ADDENDA Sept. 22, 2004

5. A medical technologist who works as a transfusion service laboratory manager in Texas reports that her facility employs a version of an "in vivo" crossmatch for ALL patients for whom they have to transfuse a unit that is "incompatible". The procedure pre-dates both herself and her medical director; both are hoping to find someone who can point them to an original literature reference for a procedure that requires using a 10 mL aliquot of washed incompatible red cells (to remove any exogenous free hemoglobin) as the initial aliquot during the in vivo test. They obtain a pre-transfusion plasma hemoglobin test (a baseline sample). Ten minutes after the 10 mL aliquot is infused, another plasma hemoglobin sample is collected and tested. The baseline and the post-transfusion plasma hemoglobin levels are then compared to determine whether it is safe to transfuse the remainder of the unit.

Her questions:

1. Is anyone familiar with the procedure and where it might have originated?
2. Is the use of an "in vivo" crossmatch procedure in cases of warm AIHA &/or strong cold autoantibodies commonplace in other large medical facilities? (Her hospital has a large hem/onc practice, BMT, HPC transplant and solid organ transplant programs.)
3. Is there a more effective procedure or mechanism to help ensure the safety of the patient during a transfusion of a technically incompatible donor unit?

ADDENDA Nov. 2, 2006

6. A transfusion medicine physician and blood bank laboratory medical technologist report that their transfusion service is located in a 650 bed medical center in southwestern Virginia. They currently perform a "hemolysis check" or "biological crossmatch" every time they need to transfuse "least incompatible" blood, in addition to obtaining consent from the patient and their physician. They are considering dropping the biological crossmatch which consists of stopping the transfusion after 15 minutes, drawing a sample from the patient, and resuming the transfusion after visual inspection of the patient's plasma for gross hemolysis. Based on their experience, they doubt that this procedure will ever detect hemolysis of the type likely to be encountered in these kinds of transfusion episodes, which they imagine would involve mainly an extravascular mechanism mediated by IgG antibodies with destruction taking place over a time frame longer than 15 minutes. They wonder if other colleagues (besides those who have already responded at this discussion) would provide comments.

7. Editor's note: The discussions at page 66 at http://www.redcross.org/pubs/immuno/18_3.pdf and at page 154 at https://www.prepare.org/pubs/immuno/20_3_04.pdf are also germane to this discussion.