This is a renewal of an ongoing project (HS05300) designed to improve the clinical accuracy and cost-effectiveness of blood product transfusions. In the first phase, a large-scale audit of 16 hospitals revealed a high proportion of red cell (RBC) and fresh frozen plasma (FFP) transfusions which did not meet criteria of clinical appropriateness. A computer-based audit instrument was designed and used in a stratified random sample of hospitals to scan large numbers of transfusions for appropriateness with efficiency. A variety of algorithms for determining which episodes are flagged can thus be modeled before selected records are submitted for implicit physician review. Also in phase one, a survey of physician beliefs and attitudes about blood product use was created and administered. These components laid the groundwork for a comprehensive educational intervention which will be implemented in a controlled trial in the current project. Medical and surgical services in teaching and community hospitals will be randomized into experimental and control groups. After assessment of baseline practices and beliefs, experimental services will be given a comprehensive intervention combining the following components: face-to-face education, feedback, print-based materials, and order-form reminders. A modified crossover design with time-series analysis will be applied; ongoing audit procedures will be used to measure program effects. Benefits and costs will be calculated in both clinical and economic terms. In addition to providing insight into optimal management of and increasingly constrained blood supply, the research will shed empirical and theoretical light on improving clinical decisionmaking regard other acute-care technologies as well.
