Trends in penicillin and macrolide resistance among pneumococci in the UK and the Republic of Ireland in relation to antibiotic sales to pharmacies and dispensing doctors.

Abstract

It is widely believed that reducing antimicrobial usage should reduce resistance, although observational evidence is mixed. Pneumococci make ideal subjects to test this belief as they are widely surveyed and lack an animal reservoir. Accordingly, susceptibility data for pneumococci in the UK and Ireland were retrieved from the Health Protection Agency's LabBase/CoSurv system and from the European Antimicrobial Resistance Surveillance System (EARSS) and British Society for Antimicrobial Chemotherapy (BSAC) databases. The BSAC surveillance examines respiratory pneumococci; the other systems focus upon invasive organisms only, with the LabBase/CoSurv system being the most comprehensive, capturing data on most bacteraemias in England and Wales. National pharmacy sales data were obtained from the IMS Health MIDAS database and were modelled to the resistance data by logistic and linear regression analysis. All systems except for the BSAC respiratory surveillance data indicated that penicillin resistance has fallen significantly since 1999 in the UK, whereas macrolide resistance has been essentially stable, or has risen slightly. The data for Ireland were based on smaller sample sizes but suggested a fall in penicillin non-susceptibility from 1999 to 2004, with conflicting evidence for macrolide resistance. The recent decreasing trend in penicillin resistance is in contrast to a rising trend in England and Wales until (at least) 1997 and strongly rising macrolide resistance from 1989 to 1993. UK pharmacy sales of macrolides and oral beta-lactams fell by ca. 30% in the late 1990s following increased concern about resistance, before stabilising or rising weakly; sales in Ireland were stable or rose slightly in the study period. We conclude that falling penicillin resistance in pneumococci followed reduced sales of oral beta-lactams to pharmacies in the UK, but a similar fall in macrolide sales was not associated with any fall in resistance. Stabilisation or decline in penicillin resistance has occurred in Ireland despite stable or increasing oral beta-lactam sales.