CHANGING PHYSICIAN BEHAVIOR IN ORDERING DIGOXIN ASSAYS

OBJECTIVE: To assess the ability to modify physicians' use of serum di goxin assays in a sustained fashion through (1) an educational interve ntion by a clinical pharmacist, and (2) changes in the computerized me dical information system. DESIGN: A before/after methodology was used to compare test use by hospital staff physicians in two phases. Phase 1 was an educational intervention conducted by a clinical pharmacist w ith an 8-month follow-up. Phase 2 was a medical information system int ervention with a 12-month follow-up. PATIENTS: Adult inpatients from J uly 1990 through December 1993 who received either digoxin therapy or at least one serum digoxin assay. MAIN OUTCOME MEASURE: Digoxin assays per patient day while receiving digoxin (assays/digoxin day), in-hosp ital mortality, and length of stay were compared before and after impl ementation of the interventions. RESULTS: A total Of 9468 patients rec eived a digoxin and/or serum digoxin assay. Baseline use of serum digo xin assays was 0.178 assays/digoxin day. Following phase 1, the educat ional intervention, use declined 20.2% to 0.142 assays/digoxin day (p < 0.03). After phase 2, the implementation of changes in the medical i nformation system, digoxin assay use was maintained at 16.3% less than that at baseline (p < 0.03). Patient mortality was unaffected. CONCLU SIONS: A low-intensity educational intervention by a clinical pharmaci st supplemented by medical information system modification resulted in an important decrease in the use of digoxin assays. The change in phy sician behavior was sustained for more than 18 months. The model prese nted is not labor intensive, does not require continuous maintenance b y healthcare personnel for a sustained effect, and may be widely appli cable to healthcare providers.