intrauterine exchange transfusion.

An antibody titer may also be used to help differentiate immune anti-D from passively acquired anti-D from RhIG. The titer level in RhIG is rarely above 4.37 Performing a titer is one way to confirm the presence of antibodies belonging to the group collectively known as HTLA (high titer, low avidity). These high-frequency antibodies are observed at the AHG phase of testing with weakly positive reactions. These weak reactions persist through extensive dilutions (as high as 2048). 38 Examples of these antibodies include anti-Ch, -Rg, -Csa, -Yka, -Kna, -McCa, and -JMH. These antibodies are usually not clinically significant but may mask significant antibodies.

Providing Compatible Blood Products

The relative difficulty in providing compatible blood products is determined by the frequency of the antigen in the population and by the clinical significance of the antibody. If the antibody does not cause decreased survival of antigenpositive RBCs, then use of random blood products that are crossmatch-compatible is acceptable. Examples of such antibodies include anti-M, -N, -P1, -Lea, and -Leb.39 When the patient sample contains a clinically significant antibody or the patient has a past history of a clinically significant antibody, units for transfusion must be antigennegative. The crossmatch technique must demonstrate compatibility at the AHG phase.40 If the patient sample is plentiful, the technologist may choose to crossmatch units, then antigen-type those that are crossmatch-compatible. If the sample is limited or if the antibody is no longer detectable in the serum, units should be antigen-typed first, then crossmatched. Knowing the incidence of the antigen in the population is helpful when determining the number of units that must be antigen-typed to find a sufficient number to fill the crossmatch request. The number of units requested is divided by the frequency of antigen-negative individuals. For example, if a crossmatch for 2 units is received for a patient with anti-E, the calculation would be 2 (units requested)/0.70 (the frequency of E-negative individuals). The result is 2.8, meaning that 3 units would need to be typed for E in order to find 2 Enegative units. When multiple specificities are present, the frequencies of antigen negative are multiplied together. If Enegative, c-negative units were required in the above example, the calculation would be 2/(0.70 _ 0.20) _ 14.3. Fourteen or fifteen units would have to be antigen-typed for E and c in order to find two units that are negative for both antigens. In certain cases, knowing the ethnicity of the donor is helpful when selecting units for antigen testing because the frequency of some antigens varies between the races. For example, when searching for units that are Fyanegative, it may be prudent to screen black donors as 90 percent will be negative for Fya compared with only 34 percent of whites. It has been proposed by some transfusion medicine experts that certain populations, particularly sickle cell and betathalassemia patients, receive units that are phenotypically matched.41,42 These multiply transfused patients seem more likely to make alloantibody than the general population; if

they are immunized, it may be difficult to find compatible blood. When the number of antigens that must be negative makes it difficult to find suitable units, rare-donor registries may be consulted. These registries maintain lists of donors and may provide frozen units of rare phenotypes. Another approach is to transfuse units that are phenotypically matched for Rh antigens and K, as these antigens are the most immunogenic.43

Resolving Difficult Antibody Identification Problems

Case One: Multiple Antibodies
Suspect multiple antibodies when all or most of the screen and panel cells are positive but reactions are at different strengths or in different phases and the autocontrol is negative. When using inclusion techniques, no single antibody accounts for all of the reactions seen. Resolution may include performing a selected cell panel to exclude certain specificities. Several sets of cells may need to be tested in order to narrow the list of possible antibodies (see Fig. 1210). Enzyme techniques may allow for separation of antibodies. In Figure 1215, the specificity of the antibodies was unclear after the initial panel. After repeating the panel using ficin-treated cells, it appears that the antibodies present are anti-K and anti-Fyb. The enzyme treatment removed the Fyb antigens, allowing the anti-K to present clearly. Anti-C and anti-Jka were also not eliminated by the initial panel. However, one would expect these antibodies to demonstrate enhanced reactivity with the ficin-treated cells. Anti-Jka