Background.Recent concerns that evidence on the appropriate use of antibiotics is not having an impact on prescribing trends are based on UK prescribing data relating to 1980-1991. The aim of this paper is to determine trends in antibiotic prescribing from 1993 to 1997 and link antibiotic prescriptions to diagnostic categories.

Method.A retrospective analysis of antibiotic prescriptions linked to diagnostic codes was carried out using the West Midlands General Practice Research Database.

Results.The prescribing rate for antibiotics fell from 963 prescriptions/1000 patients in 1993 to 807 prescriptions/1000 patients in 1997 (p<0.001). The proportion of antibiotic prescribing for respiratory conditions fell from 65 per cent in 1993 to 59 per cent in 1997 (p<0.001). The main decreases in antibiotic prescribing are accounted for by non-specific lower respiratory tract infections (-22 prescriptions/1000 patients), non-specific upper respiratory tract infections (-21/1000 patients) and throat infections (-20/1000 patients). There was increased prescribing for non-respiratory miscellaneous conditions (+6 prescriptions/1000 patients).

Conclusion.Overall antibiotic prescribing declined by 16 per cent between 1993 and 1997, primarily for respiratory conditions. These results of the current study are in marked contrast to an earlier review, which found an increase of 46 per cent between 1980 and 1991 in England. The level of antibiotic prescribing for conditions which may not be bacterial in origin is still high and there is scope for further reductions in antibiotic prescribing. This study highlights the need for regular epidemiological data to inform the debate on antibiotic prescribing.