This is a very difficlut question and there are many different opinions on the subject. When we encounter these patients, we sometimes ( depending the the frequency and amount of units transfused) , extend the work-up for up to 2 weeks as long as the DAT reaction is consistent. If possible we try to phenotype the patient and provide a phenotypically matched product, although many times this is not possible. Given that the body may respond to these antigens at any point and given the fact that we don't want to make our job more difficult down the road, it is prudent to do this. We have had patients as you describe that have developed three or four antibodies after multiple transfusions. Others do not develop any. This policy is typically made after some lengthy conversations with the blood bank director.