Antibiotic prescribing in people with and without diabetes: a population-based matched cohort study

Aims: To characterize utilization of antibiotics, including those with higher antimicrobial resistance potential, among people with type 1 diabetes and type 2 diabetes in England. Methods: Cohort study in the Clinical Practice Research Datalink, including 524 285 patients with type 2 diabetes and 33 843 with type 1 diabetes alive on 1 January 2015, matched to two non-diabetes patients on age-sex-practice and followed up to 31 December 2019. Antibiotic counts and rates were calculated for UK AWaRe (Access, Watch, Reserve) antibiotic categories. Poisson regression estimated rate ratios (RRs) of antibiotic prescriptions for diabetes versus non-diabetes, adjusting for confounders including infection count, and further stratified by age, sex, ethnicity, and deprivation. Results: Higher antibiotic prescription rates were found in diabetes groups (type 1 diabetes RR = 2.21, type 2 diabetes RR = 1.57), which were only partially attenuated after full adjustment (type 1 diabetes RR = 1.56, 95% CI 1.55-1.57, type 2 diabetes RR = 1.14, 95% CI 1.13-1.14). Elevated estimates were seen across both Access and Watch AWaRe categories. Prescription rates were higher in females and White participants, and increased with increasing age bands and deprivation. Conclusions: People with diabetes are exposed to more courses of antibiotics even when controlling for infection count, including broader spectrum antibiotics which require monitoring and risk driving antimicrobial resistance.