This website provides data on common infections and antimicrobial prescribing in primary care in the United Kindom. The project is based on analysis of data from the general practice research database. The work was funded by a grant from the Department of Health and is a collaboration between the University College London Centre for Infectious Disease Epidemiology and the Health Protection Agency.

Practices

GPRD practices are roughly representative of practices in the United Kingdom. Analyses are based on 179 general practices with continuous data over the time period mid 1990 to mid 2000. The population structure (age and gender) of these 179 practices is similar to the UK population.

Code lists

Code lists for all infections were developed using Read and Oxmis Codes. Codes were grouped into the following categories:- Upper respiratory tract infection , lower respiratory tract infection , sore throat , otitis media , sinusitis , skin and soft tissue infections , urinary tract infections , genital tract infections , gastrointestinal infections , eye infections , Non specific symptoms and signs of infection , other infections and non infectious conditions that may result in an antimicrobial prescription .

Code lists for possible adverse outcomes of infection were also generated (mastoiditis, cholesteatoma, recurrent/chronic otitis media, perforated ear drum, quinsy, acute rheumatic fever, scarlet fever, meningitis/intracranial sepsis, pneumonia/intrathoracic sepsis).

Code lists for antimicrobials (including antibacterials, antivirals, antifungals, antihelmintics, antiprotozoals, insecticides for topical use, antimycobacterials) were developed - systemic drugs were classified according to the ATC classification. In order to ensure similar compounds were grouped together topical preparations were classified into the same group as their systemic equivalent.

Linking consultations and prescribing behaviour

In order to allow linkage of antimicrobials with the condition for which they were prescribed analyses were confined to consulations where there was a single infectious diagnosis on the same day SDSD. Based on data from 50 practices this SDSD methodology allows confident lnkage of the diagnosis and prescription and includes 80% of consultations where an infection and/or an antimicrobial was prescribed but excludes consultations where antibiotics are prescribed with no indication recorded (14%) and consultations where there are either multiple antimicrobials or multiple infections (6%). The proportion of antimicrobials prescribed without a recorded indication varied by drug category. Time trends for infection consultation rates and for condition-specific prescribing rates use SDSD methodology. Time trends in overal antimicrobial prescribing utilise all prescriptions whether or not a diagnosis was recorded.

Validating results against MSGP4

For more information, contact Andrew Hayward, a.hayward@pcps.ucl.ac.uk or Irene Petersen, i.petersen@pcps.ucl.ac.uk