I have seen a vast improvement in physicians ordering unnecessary crossmatches by having the crossmatch to transfusion statistics, with a goal c:t ratio, reviewed at the Transfusion committee meeting. The hospital provides an anonymous list of physicians (they know who they are, but no one else knows who they are) so they can compare the data (competition). There are still a few physicians who order unnecessarily. We know who they are, and we set up units with long exp dates.