Transfusion audits have a long history and they are required by many r
egulatory agencies. These audits have been touted as effective in redu
cing unnecessary transfusions as evidenced by many published articles
on the subject. Most of these studies, however, have one or more flaws
in their design including the use of historical controls, disregard o
f the Hawthorne effect, use of multiple interventions, and publication
bias. Studies differ in the baseline rate of inappropriate transfusio
ns and have different methods of measuring these rates. There is also
little data on how long the effect of implementing a transfusion audit
system may last. Transfusion audits appear most successful when there
is a high baseline rate of inappropriate transfusions and interventio
ns other than the audit itself are performed. Individual institutions
should critically evaluate whether or not their current system of tran
sfusion audits is useful. (C) 1998 Elsevier Science Ltd. All rights re
served.