Has lockdown helped the UK achieve its antimicrobial stewardship targets? A retrospective cohort study using electronic health records

Talk Code: 
1E.3
Presenter: 
Rupert Payne
Co-authors: 
Rachel Denholm, Jessica James, Alastair D Hay
Author institutions: 
Centre for Academic Primary Care, Population Health Sciences, University of Bristol

Problem

Recent national primary care antimicrobial stewardship objectives have included reducing overall and broad-spectrum antibiotic prescribing. COVID-19 pandemic mitigation strategies have included two national lockdowns during which primary care has implemented remote consulting. The aim of this study was to investigate the impact of the first lockdown on the frequency of common infection diagnoses and antibiotic prescribing in the age-groups most affected: children and older adults.

Approach

A retrospective cohort analysis of patients aged 0–10 years and ≥65 years diagnosed with lower respiratory tract infection (LRTI), upper respiratory tract infection (URTI), pyelonephritis, other urinary tract infections (UTI), gastroenteritis or skin infection in primary care comparing weeks 11 to 22 in 2015 to 2019 with the same (lockdown weeks) in 2020 using data from the Clinical Practice Research Datalink. Mixed-effect regression models were used to explore changes in practice rates of infection diagnosis and proportion of infection episodes prescribed an antibiotic within three days of a diagnosis, before and during lockdown, with differences by patient and practice characteristics examined. Changes in antibiotic classes were investigated.

Findings

Practice rates of common infections decreased during lockdown across all subgroups in both age groups, by between 90% (IRR=0.10, 95% CI 0.09-0.11 for URTI in 0-10 year olds) and 42% (IRR=0.58, 0.55-0.62 for UTI in ≥65 year olds). Antibacterial prescriptions for LRTIs decreased in lockdown, compared to pre-lockdown for both 0-10s (OR=0.35; 95% CI 0.23-0.54) and ≥65s (OR=0.21; 0.18-0.24). Among the ≥65s, decreases in antibacterial prescribing were greatest among men, never smokers, and those with fewest comorbidities. Prescribing increased for URTI (OR=1.26; 1.12-1.40) among 0-10s, and among ≥65s for UTIs (OR=1.14; 1.02-1.26) and skin infections (OR=1.15; 1.02-1.30). Differences in prescribing were observed across ethnic groups for URTIs in ≥65s, with greatest reductions observed in non-white patients, but not for other infections. Increases in prescribing were observed for UTI and skin infections in those with no prior antibiotic prescribing in the last 12 months. In contrast, there were no clear subgroup effects within children for any infection. During lockdown a greater use of narrow spectrum and first-line recommended antibiotics was observed, for example reflected in reductions in the use of ciprofloxacin for UTIs.

Consequences

This study, the first to investigate changes in antibacterial prescribing by infection during the first national lockdown, indicates a sudden and substantial improvement in the overall quantity and quality of primary care antibiotic prescribing, potentially bringing forward the date by which national antimicrobial stewardship targets may be achieved. Whether these changes are a consequence of a reduction in face-to-face primary care consultations, community infection transmission and/or patient health seeking behaviour, requires further investigation.

Submitted by: 
Rupert Payne
Funding acknowledgement: