Decline of antimicrobial resistance in Pseudomonas aeruginosa bacteraemia following the COVID-19 pandemic: a longitudinal observational study

Abstract

Background: Understanding the effects of changes brought by the COVID-19 pandemic on antimicrobial resistance in P. aeruginosa (PA) is essential to inform clinical management.

Methods: This single-centre retrospective cohort study included adult inpatients with PA bacteraemia at the University Hospital Zurich between January 2014 and December 2023. The primary outcome was the association between the start of the COVID-19 pandemic and PA with multidrug resistance (MDR), defined as resistance to ≥3 of 5 antibiotic classes. We used logistic regression to adjust for age, sex and ICU treatment. Secondary outcomes included changes in resistance patterns, patient demographics and antimicrobial consumption.

Results: A total of 493 instances of PA bacteraemia in 333 patients were observed during the study period. The proportion of MDRPA declined from 21% (62/291) pre-pandemic to 9% (19/202) post-pandemic (adjusted OR 0.38, 95% CI 0.18-0.79, p = 0.01). The occurrence of MDRPA during hospitalization following an initial instance of non-MDRPA bacteraemia was rare and unlikely to happen earlier than after 2 weeks. After the start of the pandemic, we observed no MDRPA cases involving cardiovascular or pulmonary diseases and marked reductions in patients with burn injuries or organ transplants. Furthermore, ciprofloxacin and tobramycin use significantly decreased after the start of the pandemic. Overall in-hospital mortality among patients with MDRPA bacteraemia remained high (28%), with no substantial differences between time periods.

Conclusion: We observed a decline in MDRPA occurrence after the start of the COVID-19 pandemic, possibly driven by intensified infection control measures, shifts in antimicrobial use and changes in patient populations.

Figure 1.

Kaplan–Meier estimates of MDRPA bacteraemia incidence during hospital stay after the first non-MDRPA bacteraemia. The cumulative incidence curve illustrates the time to in-hospital development of MDRPA bacteraemia during the same hospitalization after a first non-MDRPA bacteraemia. The absence of events occurring within the first 2 weeks of hospitalization suggests that in-hospital acquisition of MDRPA is unlikely during this early period. Time is displayed in days since the initial non-MDRPA bacteraemia instance.

Figure 2.

(a) Number of first PA bacteraemia cases stratified by year. Total annual number of patients with first PA bacteraemia instance from 2014 to 2023, showing stable overall PA occurrence despite changes in MDRPA proportions over the study period. (b) Percentage of MDRPA in first bacteraemia cases stratified by year. Annual proportion of multidrug-resistant MDRPA among first PA bacteraemia instances from 2014 to 2023. A substantial decline in MDRPA proportions is observed after the start of the pandemic in 2020, with only 3% detected in 2023.

Figure 3.

Kaplan–Meier estimates of in-hospital death in patients with a first MDRPA bacteraemia before and after start of the pandemic. The Kaplan–Meier curve shows the cumulative probability of in-hospital mortality for patients with first MDRPA bacteraemia, stratified by the pre-pandemic (2014–2019) and pandemic/post-pandemic (2020–23) periods. Mortality estimates are displayed over time, with no substantial differences in mortality observed between the two periods (HR 1.57, 95% CI 0.43–5.67, P = 0.49; adjusted for sex, age and ICU treatment).

Figure 4.

Antimicrobial resistance before and after the start of the COVID-19 pandemic. Forest plot of OR and 95% CI adjusted for ICU treatment, sex and age (per 5 years) and taking clustering of multiple instances of bacteraemia per patient into account. ORs for the ICU population were not adjusted for ICU treatment.

Figure 5.

(a) Overall median antimicrobial days of therapy (DOT) per month before and after the start of the pandemic. Box plots showing the median monthly antimicrobial DOT for each substance before and after the start of the COVID-19 pandemic. Pip/Taz , piperacillin–tazobactam. (b) Median antimicrobial DOT per month before and after the start of the pandemic on the ICU only. Box plots showing the median monthly antimicrobial DOT for each substance before and after the start of the COVID-19 pandemic on the ICU. Pip/Taz, piperacillin–tazobactam.