Clinical Impact of Antimicrobial De-escalation in Critically Ill Patients: A Single-Center Cohort Study.

BACKGROUND Antimicrobial de-escalation (ADE) is recommended within antimicrobial stewardship programs; however, its safety and impact on multidrug-resistant organism (MDRO) emergence in critically ill patients remain uncertain.

METHODS We retrospectively studied adults (aged≥18 years) in a 12-bed intensive care unit (ICU) of a tertiary hospital in Tokyo, Japan, who received empirical antimicrobials for ≥72 h between January 2020 and December 2024. Patients were classified into ADE and No-ADE groups according to the European Society of Intensive Care Medicine/European Society of Clinical Microbiology and Infectious Diseases criteria. The primary outcome was 28-day mortality, whereas secondary outcomes were 90-day mortality and new MDRO detection. Multivariable and propensity score-matched analyses were performed.

RESULTS Among 960 patients, ADE occurred in 330 (34.4%). No significant differences were observed between ADE and No-ADE groups in 28-day mortality (11.8% vs. 10.8%) or 90-day mortality (17.3% vs. 17.5%). The proportion of patients with newly detected MDRO within 90 days was lower in the ADE group (5.5% vs. 9.2%). However, ADE was not independently associated with mortality. Although crude MDRO incidence was lower in the ADE group, multivariable analysis showed a statistically significant association between ADE and MDRO detection. ADE was associated with shorter empirical antimicrobial duration and ICU stay.

CONCLUSIONS ADE was not associated with increased mortality. Reduced empirical antimicrobial exposure and shorter ICU stay likely reflect stewardship-related processes rather than a direct causal effect.