Explaining variation in antibiotic prescribing between general practices in the UK.

Area of Science:

• General Practice
• Infectious Diseases
• Health Services Research

Background:

• Antibiotic prescribing rates exhibit considerable variation across primary care practices in England.
• Prescribers often attribute high prescribing rates to individual patient case mix, including comorbidities.

Purpose of the Study:

• To investigate the extent to which patient comorbidities explain the observed variation in antibiotic prescribing rates among English primary care practices.
• To identify other key factors influencing antibiotic prescribing variations.

Main Methods:

• Utilized primary care consultation and antibiotic prescribing data from The Health Improvement Network (THIN) database for 2013.
• Employed Boosted Regression Trees (BRTs) and Negative Binomial Regression (NBR) models to assess associations between predictors and prescribing rates.
• Included variables such as infection-related consultation rates, comorbidity prevalence, corticosteroid/immunosuppressive drug use, and demographics.

Main Results:

• The median antibiotic prescribing rate was 65.6 per 100 patients across 348 practices.
• Consultation rates (53.5%) and steroid/immunosuppressive drug use (31.6%) were the most influential factors in the BRT model, with comorbidities explaining only 12.2%.
• Models incorporating all variables explained 57% of the deviance, compared to 21% for models considering only comorbidities, age, and gender.

Conclusions:

• Variation in patient comorbidities does not account for the majority of practice-level differences in antibiotic prescribing.
• High rates of respiratory tract infection consultations and corticosteroid prescribing are significant contributors to prescribing variation.
• These findings suggest focusing on consultation patterns and steroid prescribing may be key to reducing antibiotic use in primary care.