FDA 2011 guidance (I don't know what AABB last standards say about this) states "If ABO typing discrepancies exist, you should not rely on a computer crossmatch. This is particularly important if there is mixed field red cell reactivity, missing serum reactivity, or apparent change in blood type following hematopoietic stem cell transplantation. Under those circumstances, your procedures should provide for compatibility testing using serologic crossmatch techniques" ... But what if this discrepancy is easily explained by a recent non-isogroup transfusion... Would you still reject computer crossmatch? For what reason? How do you manage these situations in your lab?