ABSTRACT
The Mayo Clinic (Rochester, Minn) Division of Transfusion Medicine evaluated the effect of methods of error detection, analysis, and prevention on the rate of errors occurring from 1982 through 1992. We defined an error as any deviation from the standard operating procedure. Twenty-four standard operating procedures were monitored for errors that related to donor processing, testing of donor blood, patient testing, and transfusion. The estimate of the overall error rate and 95% confidence interval fluctuated between 20 and 30 per 10(4) procedures. The transcription error rate declined from 21 to six per 10(4) procedures as a result of changes to systems using computer-generated labels and bar codes. We concluded that when errors are recognized, and appropriate system changes are made, these errors can be prevented.
