Evaluation and management of hospital antibiotic use

Abstract

Antimicrobials are unique drugs in that they target "infectious" or "transferable"
diseases. There is considerable evidence linking increasing antimicrobial use with
increasing resistance. Resistant bacteria do not know the boundaries, either between countries or within a society between hospital and primary care. Inappropriate prescribing of antimicrobials in hospitals therefore has consequences for whole communities and problems may spread both nationally and internationally. The gathering of reliable measurements of antibiotic use in hospitals employing standardised methods is essential to building an evidence base and highlighting inconsistencies at national and international levels.

In this study, after data processing, validating and record linkage, a method for
electronic conversion of drug supply data to the ATC/DDD classification and for
longitudinal analysis was established for Tayside and then for a set of European
hospitals. Time series analysis and interrupted time series analysis were described and used for longitudinal surveillance and interventional study of antimicrobial use. This thesis explores issues concerning the evolution and management of hospital antimicrobial use using a wide range of methods. A series of drug utilisation research studies were implemented as the basis of research methods that, in combination of previously described methods, provided novel studies.

No single measure can currently capture all of the aspects of hospital antibiotic use. However, a combination of detailed, point prevalence data from individual patients with longitudinal analysis of total consumption can provide meaningful data for comparison between hospitals and for analysis of the relationship between use and outcome. Additionally, there is a need to apply standard processes and novel methods to produce more meaningful surveillances.

Longitudinal and point prevalence surveillances together with an explanation of
variations in hospital characteristics are used to produce a set of coherent measurements of hospital antimicrobial use.

Administrative data for longitudinal surveys requires continuous quality control.
Whereas drug utilisation researchers and clinicians should target a set of indicators for interventional studies, large studies at national or international level need central data processing by country to identify targets for evaluation and for interventional studies. Support from experts in other fields is needed to address any shortcomings that may be experienced during continuous antibiotic drug utilisation monitoring at national and international levels.

Date of Award	2010
Original language	English
Awarding Institution	University of Dundee
Supervisor	Peter Davey (Supervisor)