Trends, determinants and differences in antibiotic use in 68 residential aged care homes in Australia, 2014-2017: a longitudinal analysis of electronic health record data

• Published in: BMC health services research Vol. 20; no. 1; pp. 883 - 12
• Main Authors: Raban, M Z, Lind, K E, Day, R O, Gray, L, Georgiou, A, Westbrook, J I
• Format: Journal Article
• Language: English
• Published: England BioMed Central Ltd 18-09-2020; BioMed Central; BMC
• Subjects: Aged; Aged, 80 and over; Anti-bacterial agents; Anti-Bacterial Agents - therapeutic use; Antibiotics; Antimicrobial agents; Antimicrobial Stewardship - statistics & numerical data; Australia; Datasets; Dementia; Disease prevention; Drug administration; Drug resistance; Drug Utilization - statistics & numerical data; Elder care; Electronic Health Records; Electronic records; Female; Health services; Homes for the aged; Homes for the Aged - statistics & numerical data; Humans; Long-term care; Male; Management; Medical records; Nursing homes; Nursing Homes - statistics & numerical data; Residential care facilities; Residential facilities; Retrospective Studies; Staphylococcus infections; Statistics; Trends; Urinary Tract Infections - drug therapy; Australia; Pharmacoepidemiology; Nursing homes; Anti-bacterial agents; Homes for the aged; Electronic health records; Residential facilities; Long-term care
• ISSN: 1472-6963; 1472-6963
• DOI: 10.1186/s12913-020-05723-3

Internationally, point prevalence surveys are the main source of antibiotic use data in residential aged care (RAC). Our objective was to describe temporal trends in antibiotic use and antibiotics flagged for restricted use, resident characteristics associated with use, and variation in use by RAC home, using electronic health record data.
We conducted a retrospective cohort study of 9793 unique residents aged ≥65 years in 68 RAC homes between September 2014 and September 2017, using electronic health records. We modelled the primary outcome of days of antibiotic therapy /1000 resident days (DOT/1000 days), and secondary outcomes of number of courses/1000 days and the annual prevalence of antibiotic use. Antibiotic use was examined for all antibiotics and antibiotics on the World Health Organization's (WHO) Watch List (i.e. antibiotics flagged for restricted use).
In 2017, there were 85 DOT/1000 days (99% CI: 79, 92), 8.0 courses/1000 days (99% CI: 7.6, 8.5), and 63.4% (99% CI: 61.9, 65.0) of residents received at least one course of antibiotics. There were 7.7 DOT/1000 days (99% CI: 6.69, 8.77) of antibiotics on the WHO Watch List administered in 2017. Antibiotic use increased annually by 4.09 DOT/1000 days (99% CI: 1.18, 6.99) before adjusting for resident factors, and 3.12 DOT/1000 days (99% CI: - 0.05, 6.29) after adjustment. Annual prevalence of antibiotic use decreased from 68.4% (99% CI: 66.9, 69.9) in 2015 to 63.4% (99% CI: 61.9, 65.0) in 2017, suggesting fewer residents were on antibiotics, but using them for longer. Resident factors associated with higher use were increasing age; chronic respiratory disease; a history of urinary tract infections, and skin and soft tissue infections; but dementia was associated with lower use. RAC home level antibiotic use ranged between 44.0 to 169.2 DOT/1000 days in 2016. Adjusting for resident factors marginally reduced this range (42.6 to 155.5 DOT/1000 days).
Antibiotic course length and RAC homes with high use should be a focus of antimicrobial stewardship interventions. Practices in RAC homes with low use could inform interventions and warrant further investigation. This study provides a model for using electronic health records as a data source for antibiotic use surveillance in RAC.