Background: There is a clear association between heavy antimicrobial consumption within a population and the frequent recovery of resistant bacteria, but whether a reduction in antimicrobial use can reverse this process is less clear. We investigated the effect of a national restriction of sulphonamide prescribing in the UK on the prevalence of sulphonamide resistance in Escherichia coli.
Methods: Consecutive clinical isolates of E coli were collected at the Royal London Hospital in 1991 and 1999. These collections, each of more than 350 isolates, were compared. Minimum inhibitory concentrations of sulphamethoxazole and eight other antimicrobials were determined. The presence and locations of sulphonamide-resistance genes were examined by PCR, plasmid extraction, Southern hybridisation, and transconjugation.
Findings: Despite a huge decrease in sulphonamide prescriptions (from 3,208,000 [corrected] prescriptions per year in 1991 to 77,000 [corrected] in 1999), the frequency of resistance remained high in 1999 (165/359 [46.0%] vs 143/360 [39.7%] in 1991; difference 6.2% [95% CI 20.9 to 13.3]). Integron-borne sulI was present in 16.4% of isolates in 1991 and 17.5% in 1999. The prevalence of sulII increased from 26.7% in 1991 to 36.5% in 1999 (difference 9.8% [3.1 to 16.5] p=0.0046). SulII was located on large plasmids, at least some of which were conjugative multiresistance determinants.
Interpretation: These results show that a huge decrease in antibiotic prescribing does not necessarily reduce resistance within a useful time. The main reason seems to be the genetic linkage of the index resistance to other resistance determinants.