Key points are not available for this paper at this time.
The recent focus on medical risk and financial cost has prompted a need for better guidelines for prescribing the transfusion of blood components. In 1987, to respond to the issues of quality transfusion practice and accurate evaluation, LDS Hospital (Salt Lake City, UT) began using a computerized, knowledge-based blood-ordering system. Each transfusion request was reviewed and flagged by the computer when it did not meet the criteria established by the medical staff. The study reviewed the use of red cells, platelets, and fresh-frozen plasma in 13,082 transfusion orders for 5847 consecutive patients from July 1, 1988, through June 30, 1989. The evaluation assessed, first, the adherence of physicians to computerized criteria and, second, their adherence to the quality of transfusion practice. A high percentage of the blood units ordered met the established criteria: 91.2 percent for the red cell transfusions, 72.9 percent for platelets, and 81.7 percent for fresh-frozen plasma. From the July 1, 1987, implementation date through June 1989, the mean hematocrit of persons being transfused dropped from 28.6 to 27.7 percent (0.29 = 0.28) (p less than 0.005) and the number of orders requiring review by the quality assurance department dropped from 100 to 14 percent; moreover, there was a true-exception rate of only 0.37 percent. The use of the computer system effected the implementation of the following measures: 1) identification of the indications and establishment of clear clinical and biologic parameters for every transfusion, and 2) measurement and improvement of institutional transfusion practice. These results demonstrated the efficacy of a computerized hospital information system in implementing continuous quality improvement for transfusion practice.(ABSTRACT TRUNCATED AT 250 WORDS)
Lepage et al. (Wed,) studied this question.