Changing physician test ordering in a university hospital. An intervention of physician participation, explicit criteria, and feedback.

To decrease inappropriate test ordering by medical house staff in a university hospital, we examined the feasibility of an intervention that involved physicians in developing explicit criteria for ordering four specific tests and incorporated feedback of tests ordered. We implemented a time series design with measures at 12 and six months before, during, and three weeks after the intervention. During the intervention, orders for initial or admission chest roentgenograms decreased by 22% and repeated orders for routine urinalyses, chest roentgenograms, and leukocyte differential counts decreased by 23%, 30%, and 46%, respectively, compared with the six-month preintervention period. Orders for prothrombin time and/or partial thromboplastin time did not fall. After the intervention, most test ordering remained at the intervention level. These preliminary results suggest that this intervention may be effective and not overly costly.