An exploratory study of antibiotic prescribing in the CPRD Aurum database

Date of ISAC Approval: 
29/05/2019
Lay Summary: 
Unnecessary antibiotic prescribing in primary care is receiving a lot of attention at present because of growing concern about the problem of antibiotic resistant infections. Our previous research using CPRD shows that there are high antibiotic prescribing rates in the UK. The most common reasons for GPs giving antibiotics are for respiratory infections like coughs colds and sore throats, as well as urine infections and skin conditions. GPs record no definite reason for the antibiotic prescription in about half of all antibiotic prescriptions. Our previous reserach used the CPRD Gold database. This study aims to find out whether similar patterns of antibiotic use are also observed in CPRD Aurum data. We propose to analyse records for patients who have been prescribed antibiotics during 2017. A sample of patients will be taken from the denominator file to give about 100,000 patients of whom about 60,000 may have been prescribed antibiotics. We will estimate antibiotic prescribing rates by age and gender. We will evaluate antibiotic prescribing rates separately for respiratory infections like coughs colds and sore throats, as well as urine infections and skin conditions. We will find out how often GPs do not code the reasons for antibiotic prescriptions we will also evaluate repeat prescriptions. We will compare the results from the CPRD Aurum database with those from CPRD Gold. This will enable us to become familiar with the CPRD Aurum data and to see whether patterns of antibiotic prescribing are similarly recorded in the GOLD and Aurum databases. This will enable us to plan future studies.
Technical Summary: 
This proposal is for a descriptive study of antibiotic prescribing in the CPRD Aurum database. Antibiotic prescribing in primary care is receiving a lot of attention at present because of growing concern about the problem of antimicrobial drug resistance. We have reported several studies of antibiotic utilisation based on analysis of data derived from the Vision practice-system from the CPRD GOLD database, others have reported similar results using the THIN database. The studies show that there are high antibiotic prescribing rates in the UK. The most common indications for antibiotic prescribing are for respiratory, genito-urinary and skin conditions. However, about half of antibiotic prescriptions may not have coded indications recorded, with either no codes, non-specific codes or repeat prescriptions recorded. This study aims to find out whether similar patterns of antibiotic drug use are also observed in CPRD Aurum data. A sample of patients will be taken from the denominator file to give about 100,000 patients of whom about 60,000 may have been prescribed antibiotics in any year. We will estimate antibiotic prescribing rates by age and gender. We will evaluate antibiotic prescribing rates separately for respiratory, genito-urinary and skin conditions. We will evaluate uncoded antibiotic prescriptions including no codes recorded, non-specific codes recorded and recorded repeat prescriptions. We will compare the results with estimates form the CPRD Aurum database. This will enable us to gain basic familiarity with the CPRD Aurum data and to see whether patterns of antibiotic prescribing are similarly recorded in the GOLD and Aurum databases. This will enable us to plan future studies.
Health Outcomes to be Measured: 
Antibiotic prescriptions: using codes that represent all drugs in section 5.1 of the British National Formulary, excluding drugs for TB and Leprosy and methenamine.
Collaborators: 

Professor Martin Gulliford - Chief Investigator - King's College London
Professor Martin Gulliford - Corresponding Applicant - King's College London