Study objective: To measure diversity (or heterogeneity) of antibiotic use in a sample of hospitals in the United States and to assess an association with bacterial resistance.
Design: Observational cross-sectional study.
Setting: Forty-two general medical-surgical hospitals, mostly in the Eastern United States.
Data source: Administrative claims data and hospital antibiograms during 2003.
Measurements and main results: Antibiotic use was measured by defined daily dose (DDD), and diversity was assessed by using Simpson's Index and the Shannon-Weiner Index. Aggregated antibiograms (from 17 hospitals) were used to assess resistance. There were a total of 3,655,579 patient-days during 2003, with a mean +/- SD of 87,037 +/- 62,679 patient-days/hospital. Mean +/- SD antibiotic use was 704 +/- 244 DDD/1000 patient-days/hospital, and fluoroquinolones represented the largest single class. Mean +/- SD diversity by Simpson's Index was 0.861 +/- 0.022/hospital, and a strong correlation was noted between the two diversity measures. Multivariate analysis found no significant relationship between diversity and the proportion of resistant pathogens.
Conclusion: Diversity of aggregate antibiotic use can be measured, but additional investigations are needed to determine if it is a useful strategy to contain resistance.