Temporal and regional trends of antibiotic use in long-term aged care facilities across 39 countries, 1985-2019: systematic review and meta-analysis

Background Antibiotic misuse is a key contributor to antimicrobial
resistance and a concern in long-term aged care facilities (LTCFs). Our
objectives were to:  i) summarise key indicators of systemic
antibiotic use and appropriateness of use, and ii) examine temporal and
regional variations in antibiotic use, in LTCFs (PROSPERO registration
CRD42018107125). Methods & Findings Medline and EMBASE were
searched for studies published between 1990-2021 reporting antibiotic use
rates in LTCFs. Random effects meta-analysis provided pooled estimates of
antibiotic use rates (percentage of residents on an antibiotic on a single
day [point prevalence] and over 12 months [period prevalence]; percentage
of appropriate prescriptions). Meta-regression examined associations
between antibiotic use, year of measurement and region.  A total
of 90 articles representing 78 studies from 39 countries with data between
1985-2019 were included. Pooled estimates of point prevalence and 12-month
period prevalence were 5.2% (95% CI: 3.3-7.9; n=523,171) and 62.0% (95%
CI: 54.0-69.3; n=946,127), respectively. Point prevalence varied
significantly between regions (Q=224.1, df=7, p<0.001), and ranged
from 2.4% (95% CI: 1.7-2.7) in Eastern Europe to 9.0% in the British Isles
(95% CI: 7.6-10.5) and Northern Europe (95% CI: 7.7-10.5). Twelve-month
period prevalence varied significantly between region (Q=15.1, df-3,
p=0.002) and ranged from 53.9% (95% CI: 48.3-59.4) in the British Isles to
68.3% (95% CI: 63.6-72.7) in Australia. Meta-regression found no
association between year of measurement and antibiotic use prevalence. The
pooled estimate of the percentage of appropriate antibiotic prescriptions
was 28.5% (95% CI: 10.3-58.0; n=17,245) as assessed by the McGeer
criteria. Year of measurement was associated with decreasing
appropriateness of antibiotic use over time (OR: 0.78, 95% CI:
0.67-0.91).  The most frequently used antibiotic classes were
penicillins (n=44 studies), cephalosporins (n=36),
sulphonamides/trimethoprim (n=31), and quinolones (n=28). 
Conclusions Coordinated efforts focusing on LTCFs are required to address
antibiotic misuse in LTCFs. Our analysis provides overall baseline and
regional estimates for future monitoring of antibiotic use in LTCFs.