The nonvalue of retrospective peer comparison feedback in containing hospital antibiotic costs

Antibiotics have accounted for an increasing percentage of hospital pharmacy charges. Recently, an inexpensive method, automated peer comparison feedback, has been developed to influence physician use of resources. The documented success of several implementations of this strategy led to a one-year experiment to influence hospital antibiotic utilization. Each month, attending physicians in the top 50 percentiles for expenditure were notified of their status in relation to their peers. Expenditures by feedback and control groups were compared to determine whether feedback would result in reduced expenditures by individuals, or whether there would be a generalized reduction in expenditure by the entire group (Hawthorne effect). Over the year, no significant reduction in expenditure was noted. However, some important utilization patterns were identified. Although more surgical patients received antibiotics than did nonsurgical patients, surgical antibiotic costs were less. Surgical therapy was typically of shorter duration and involved the use of less expensive antibiotics. Multiple-antibiotic prescribing was less frequent on surgical services. Thirty percent of attending physicians were responsible for 80 percent of all antibiotic costs; 60 percent of those in this top group were members of the medical cohort. In conclusion: (1) As implemented in the current study, automated peer comparison feedback was not an effective method for reducing antibiotic utilization; (2) Differences in prescribing patterns between services may dictate the best strategies for improving antibiotic utilization; (3) More attention should be directed toward the relatively small "reference group" of physicians responsible for most hospital antibiotic prescribing.