Modifying empiric antibiotic prescribing: experience with one strategy in a medical residency program

Initial experience with one strategy to modify selection of empiric antibiotic therapy by medical residents in presented. New orders for parenteral antibiotics written by residents before receiving culture results were randomly allocated to either a recommendation or a control group. Antibiotic orders from each group were compared with guidelines for antibiotic selection, and suggestions to change antibiotics were made only for orders from the recommendation group that did not match the guidelines. Recommendations for change appeared necessary for about 20% of the orders within each group. At follow-up, 78% of these orders had been changed as suggested in the recommendation group, whereas only 10% of the orders in the control group were changed (p less than 0.005). The mean difference in daily cost per order was a savings of $27.40 in the recommendation group and only $1.11 in the control group (p less than 0.01).