Evaluation of a tool to benchmark hospital antibiotic prescribing in the United Kingdom

Abstract

Objective: To investigate whether Defined Daily Dose/Finished Consultant Episode (DDD/FCE) ratio is sensitive to changes in prescribing patterns and could be used as a prescribing indicator in hospitals.

Method: The study comprised two phases. In phase 1 the weekly DDD/FCE ratios for two antibiotics were calculated and monitored in one acute NHS hospital before and after the implementation of prescribing guidelines, which impacted on the use of the antibiotics. In phase 2 data on the use of four antibiotics over one year was collected from fifty-eight medium to large acute hospitals in England, together with corresponding FCE data. DDD/FCE ratios for each antibiotic in each hospital were compared.

Main outcome measure: Whether the DDD/FCE ratio for two antibiotics changed in one hospital following the introduction of prescribing guidelines for these antibiotics. The variability in DDD/FCE ratio for two broad spectrum antibiotics compared to two narrow spectrum antibiotics across a number of acute hospitals in England.

Results: In phase 1 the DDD/FCE ratios for the two antibiotics were lower post implementation of the guidelines indicating that the ratio was sensitive to changes in prescribing. In phase 2 the median DDD/FCE ratios of the two broad spectrum antibiotics from all fifty eight hospitals were much higher (0.126, and 0.265) than for the two narrow spectrum antibiotics (0.048, and 0.021), indicating higher use of the broad spectrum antibiotics. Furthermore, the variation in prescribing between the hospitals, as indicated by the inter-quartile range about the median, was greater for the two broad spectrum agents (0.201 and 0.193), than for the narrow spectrum agents (0.06, and 0.042), as was expected.

Conclusion: The DDD/FCE ratio is sensitive to changes in prescribing and can reflect differences in the use of antibiotics between hospitals, after accounting for differences in activity. DDD/FCE ratio has the potential to be used to also account for differences in case mix between hospitals although further work is needed in this area.