[Effects of restrictions on use of vancomycin in a German university hospital]

An antibiotic order form intervention does not improve or reduce vancomycin use

OBJECTIVES

To determine whether a paper-based antibiotic ordering system is an effective antibiotic stewardship measure.

METHODS

An antibiotic order form (AOF) was introduced in July 2001 at a pediatric tertiary care hospital. Vancomycin courses prescribed before and after the AOF introduction were retrospectively reviewed based on Hospital Infection Control Practices Advisory Committee guidelines. The impact of the AOF on the appropriateness of vancomycin prescribing was evaluated in univariate and multivariable analyses that adjusted for other factors associated with appropriateness of vancomycin use. The density of vancomycin use after introduction of the AOF was also assessed.

RESULTS

Compliance with the AOF was poor (<50%) during the planned study period; therefore an additional 2 months of improved compliance (70-80%) were included. Rates of inappropriate vancomycin use increased during the study periods: 35% before AOF; 39% post-AOF; and 51% during the improved compliance period. On adjusted analysis, vancomycin utilization was significantly more inappropriate after introduction of the AOF. Vancomycin doses per 1000 patient days increased after introduction of the AOF.

CONCLUSIONS

Inappropriate vancomycin use and vancomycin use overall increased after the introduction of an AOF. An AOF intervention did not have its intended effect of improving and reducing vancomycin use.

Antibiotic use in German university hospitals 1998-2000 (Project INTERUNI-II)

In a pilot study we established a hospital antibiotic use database from which estimates for antibiotic consumption in German hospitals (expressed as defined daily doses per 100 patient days, DDD/100) can be retrieved for both benchmarking and antibiotic use and resistance research purposes. Data from eight university hospitals (1998-2000) showed a mean antibiotic use density of 60.1 DDD/100 in the surgical and 79.3 DDD/100 in the medical services. Antibiotic use was higher in the intensive care units (surgery: 146 DDD/100, medicine: 187 DDD/100) than in haematology-oncology services (110.8 DDD/100) or in other surgical (51.6 DDD/100) and medical (66 DDD/100) service areas. There were major differences in the use of specific antibacterial drug classes between service areas. The established database allows detailed analysis in antibacterial drug use and can be linked to bacterial resistance surveillance databases.

Clinical determinants for the recovery of fungal and mezlocillin-resistant pathogens from bile specimens

We conducted a retrospective analysis of all bile specimens obtained for routine cultures from January 1995 through December 1999 at our tertiary care hospital. Results of microbiologic testing were linked to clinical parameters gathered by means of chart review. A total of 722 isolates were cultured from 345 of 454 bile specimens obtained from 288 individual patients. Prior receipt of a >7-day course of antibiotics (odds ratio [OR], 5.7), extensive leukocytosis (leukocyte count, >20,000 cells/microL) on admission (OR, 7.8), endoscopic or percutaneous biliary manipulation during the previous 14 days (OR, 2.9), and treatment in an internal medicine ward (OR, 2.5) were independent factors significantly associated (Pless-than-or-eq, slant.05) with recovery of Candida species from bile specimens. Culture of mezlocillin-resistant bacteria from bile specimens was independently associated with the specimen having been obtained >1 week after admission (OR, 3.8), lack of history of endoscopic biliary drainage (OR, 3.2), and high serum aspartate aminotransferase levels (>72 U/L) on admission (OR, 2.6). Prospective studies are warranted to evaluate accordingly adjusted empiric therapies for biliary infections.