Will transfusion errors due to human factors ever be eliminated?

To the Editor—The transfusion of ABO-mismatched blood usually results from patient misidentification and sample-labelling errors during phlebotomy. It is a common problem worldwide and causes more transfusion-related fatalities than does the transfusionmediated transmission of human immunodeficiency virus. Various approaches have been used to minimise human error during phlebotomy and blood transfusions; however, none have been convincingly and uniformly effective. Some hospitals require that a second specimen of blood be drawn at a later time to confirm the ABO blood grouping; otherwise, new patients will be given group O blood only. Some housestaff may draw two specimens at the same time and send the second sample later on, thus circumventing the need to redraw blood. Other hospitals require two staff members to identify the patient before blood sampling and to countercheck the labelling of the specimen. As a consequence, these procedures increase the workload of hospital nursing staff and can be a source of resentment. The use of the Blood-Loc mechanical barrier system (Novatek Medical Inc., Connecticut, United States) has been promoted in some centres to improve transfusion safety, but its acceptance has been limited because of the cumbersome procedures involved and the availability of other systems. The computerised bedside identification system (Ident-A Blood Identification System; Hollister Inc., Illinois, United States) is a promising method, but is expensive and requires at least one portable scanner in each hospital ward, as well as a sophisticated computer system.