Abstract
Background/Objectives: Antimicrobial resistance is an escalating public health threat. Despite the low prevalence of community-acquired bacterial coinfection (CABC) in patients presenting with COVID-19, antibiotics are frequently prescribed. We evaluated antibiotic prescribing for suspected CABC in adults admitted with COVID-19 to a metropolitan referral hospital in South Australia and the impact of a targeted, locally developed educational intervention. Methods: We conducted a closed-loop retrospective audit of consecutive adult COVID-19 non-intensive care unit admissions to the General Medicine Unit over a six-month period (5 February-6 August 2024), totalling 126 episodes after exclusions. Antibiotic prescribing rates were compared before and after implementation of targeted education to admitting clinicians. This was delivered via guideline-focused presentations at unit meetings, emails and posters in key clinical areas detailing the consensus recommendation, and brief electronic reminder messages. Prescribing appropriateness was inferred from adherence to the consensus recommendation, which included predefined criteria relating to total leukocyte count, neutrophilia, and radiological findings. Statistical comparisons were performed using appropriate statistical tests for categorical and continuous variables. Results: Following the intervention, antibiotic prescribing decreased from 26% to 22% of admissions (p = 0.57), and prescribing among patients who did not meet the specified criteria fell from 16% to 6% (p = 0.30). The proportion of prescriptions meeting the consensus criteria increased from 62% to 80% (p = 0.43). Conclusions: Baseline antibiotic use was substantially lower than commonly published rates but still exceeded expected CABC prevalence. Targeted education and simple, locally adaptable guidance may support improved antimicrobial stewardship in COVID-19 admissions.