Understanding the contribution of influenza vaccines to antibiotic prescribing in the UK population.

Vaccination should reduce antibiotic prescribing (AP), the driver of antimicrobial resistance. Looking for reductions in Randomised Controlled Trials of vaccines is problematic as they are usually powered for large effect sizes of specific end-points. An alternative approach is to undertake observational studies using large routine data sources. The problem with this approach is ensuring major confounding is identified and accounted for when linking vaccine status and outcome, and demonstrating a causal relationship. We will conduct an observational study using the Clinical Practice Research Datalink (2007-2019) and The Health Improvement Network (THIN) (2007-2019) from the United Kingdom. Our primary aim is to measure AM prescribing in adults over 65 and children under 5 by frequency of influenza vaccination. We will use three complimentary approaches and different methods to adjust for confounding: (1) A self-controlled case series comparing annual AM prescribing within individuals by annual receipt of influenza vaccine and vaccine/pathogen match; (2) A cohort study balancing comparator populations using propensity scoring to control for major determinants of health care utilisation and underlying health status; (3) An interrupted time series analyses investigating paediatric influenza vaccine uptake (September 2013) and AM prescribing.