Patterns and determinants of community antibiotic use in Australia: observational studies using large electronic health datasets

Abstract

Background: Antibiotic overuse in clinical practice is an important driver for antibiotic resistance. However, there is insufficient granularity regarding Australian community antibiotic prescribing patterns reported in the literature.

Aim and design: I conducted four population-based retrospective observational studies to understand the pattern of antibiotic use and the determinants in the Australian community. Chapter 2 examined the rate of antibiotic dispensing and microbiology testing in older adults by their comorbidities. Chapter 3 examined the adherence to guidelines in urinary tract infection episodes in general practice when antibiotics were prescribed. Chapter 4 compared the likelihood of antibiotic prescribing for upper respiratory tract infections in regular and after-hours general practice consultations. Chapter 5 examined the general practice-level broad- to narrow-spectrum antibiotic ratio and its association with patients’ antibiotic treatment non-response within the practice.

Methods: Chapter 2 was based on the 45 and Up Study, a large cohort study on older Australians linked to routinely collected health databases including antibiotic dispensing data. Chapter 3 to 5 used MedicineInsight, a national database of electronic health records from Australian general practices. Both descriptive analysis and multivariable modelling were used to identify determinants of antibiotic use.

Findings: 1) There was a discord between the high antibiotic dispensing rate and low microbiology testing rate among older people with chronic respiratory diseases, suggesting potential antibiotic overuse among those subgroups. 2) Some patient groups who are recommended to have routine urine testing in urinary tract infection episodes, e.g., patients aged <5 years, with recurrent urinary tract infections, or living in nursing homes, had a lower likelihood of testing than comparable patient groups. 3) After-hours consultations were associated with a higher likelihood of immediate antibiotic prescribing for upper respiratory tract infections in general practice. 4) The ratio of prescribing of broad- to narrow-spectrum antibiotics at the practice-level was a predictor for patients’ antibiotic treatment non-response in respiratory tract infection episodes, even if the patients had no previous individual-level antibiotic exposure.

Conclusion: These findings could provide implications for developing targeted antibiotic stewardship programs in the Australian community.