Regional variation in ampicillin and trimethoprim resistance in Escherichia coli in England from 1990 to 1997, in relation to antibacterial prescribing.

Over 200 hospitals in England report resistance data for bacteraemia and meningitis isolates to the Public Health Laboratory Service. We reviewed ampicillin and trimethoprim resistance rates from 1990 to 1997 for Escherichia coli, which is the species reported most frequently from these bacteraemias. Ampicillin resistance was relatively stable over time, but varied between Health Regions. The proportion of ampicillin-resistant E. coli in the East Anglia region remained </=42% in all years except one and that in the South Western region always remained <50%. At the other extreme, the proportions of ampicillin-resistant isolates in the Northern and Trent regions never fell below 59%. The prevalence of resistance to trimethoprim rose over time in most regions; again, however, the prevalence of resistant isolates was lowest in the East Anglia and South Western regions, whereas the highest resistance rates were reported from Mersey, NW Thames, NE Thames and North Western regions. These observations were related to data for community prescribing, which accounts for most ampicillin and trimethoprim use. Prescribing data for ampicillin and trimethoprim from 1987 to 1997 were obtained from the IMS-HEALTH Medical Data Index, and data for all antibacterial drugs between 1995 and 1997 from the Prescription Pricing Authority. Correlations between resistance rates and prescribing of specific antibiotics were weak, although there was some trend for regions with high total prescribing to have higher rates of ampicillin resistance. The South Western region was conspicuous both for low rates of resistance and low prescribing. Several factors may determine the lack of wider and more obvious relationships between resistance and prescribing. In particular, regions may be inappropriately large areas to test the relationship, isolates from bacteraemias may not be representative of those experiencing selection pressure in the community and the resistance data may have been distorted by nosocomial strains, although this seems unlikely with E. coli.