Back in the 80's, had a cardiac patient undergoing CABG and received 8 units B NEG RBC during surgery. Three days later, a new specimen was received that typed mixed field AB POS with a reverse type group A. Supervisor retrieved the chemistry specimen from 3 days prior and it typed as A POS. Turns out the phlebotomist drew the patient in the other bed and this was long before second sample requirements .... Patient was a little "oozy" post-operatively but survived the experience.
Same hospital, elderly female patient typed as O NEG. Blood was crossmatched and issued during the weekly computer downtime. As I went through the stack of units that had been issued updating the computer records, got a major flag as at computer issue - the unit was A NEG and that was just the first of 2 units that had been issued during this 2 hour downtime. Investigation revealed that 2 A NEG RBC were placed in the O NEG inventory, were crossmatched using immediate-spin and were compatible, and out the door they went. The patient was fine.