Effect of general practice characteristics and antibiotic prescribing on Escherichia coli antibiotic non-susceptibility in the West Midlands region of England: a 4 year ecological study.

Objectives To assess the effect of general practice characteristics and antibiotic prescribing on the number of non-susceptible Escherichia coli isolated from urine specimens submitted from community settings, we undertook an ecological study of the general practice population in the West Midlands. Methods Descriptive analysis and multilevel modelling of temporal trends in antibiotic prescribing and non-susceptibility of E. coli urine isolates to a range of antibiotics prescribed in the community over a 4 year period. Results Nine of the 16 antibiotic prescribing/non-susceptibility combinations demonstrated a significant statistical linear correlation with non-susceptibility either for prescribing in a quarter or for prescribing within the previous 12 months. The magnitude of the effect varied, from a 0.3% increase in the odds of non-susceptibility to ampicillin/amoxicillin (when prescribing ampicillin/amoxicillin) to a 6.3% increase in the odds of non-susceptibility to nitrofurantoin (when prescribing nitrofurantoin) for an increase of 50 DDDs per 1000 practice population within a quarter (equivalent to ∼10 courses of antibiotics). In all 16 models, single-handed general practices were shown to have a significant association with increased numbers of non-susceptible E. coli urine isolates (adjusted ORs 1.083-1.657). Increased prescribing of ampicillin/amoxicillin in winter periods was associated with increased non-susceptibility of E. coli isolated from urine specimens. Conclusions Small increases in antibiotic prescribing in individual general practices reduce the number of susceptible bacteria in the practice population. To maintain the effectiveness of available treatment, antibiotic stewardship should be encouraged and supported within each practice