Objective: Antibiotic de-escalation (ADE) is a key antimicrobial stewardship quality indicator. We aimed to evaluate the rate and patterns of ADE among clinically stable patients with bloodstream infection.

Methods: We analysed secondary data from two prospective multicentre cohort studies, BLOOMY and BLOOMY-PREDICT. ADE was assessed among patients eligible for safe ADE on day 5 after index blood culture. Narrowing of antibiotic spectrum was determined by a ranking based on WHO AWaRe-classification. Risk factors for not performing ADE (non-ADE) were studied using multivariable logistic regression.

Results: In total, 937 of 3824 study patients (24.50%) were eligible, of which 218 (23.27%) did not have an option for de-escalation based on antimicrobial susceptibility testing. Of 719 patients in which ADE was feasible, only 406 (56.47%) received ADE. Empiric monotherapy (OR [95% CI] = 5.68 [3.77-8.56], P < 0.001), Gram-negative pathogen (OR 2.27 [1.60-3.24], P < 0.001), healthcare-associated infection (OR 1.55 [1.01-2.39], P = 0.046), and hospital-acquisition (OR 1.63 [1.02-2.61], P = 0.042) were identified as independent factors associated with non-ADE. Conversely, ICU treatment on day 0 (OR 0.60 [0.38-0.95], P = 0.029) was independently associated with de-escalation, alongside a strong study centre effect (OR 0.26 [0.15-0.44], P < 0.001). Further, in-hospital mortality was not associated with ADE (46/406, 11.33% vs. 29/313, 9.27%, P = 0.369). Low ADE rates in patients with urogenital focus and unnecessary carbapenem use were issues of particular concern in our cohort.

Conclusions: Available opportunities for ADE were frequently missed in our setting, especially in Gram-negative bloodstream infection. Antimicrobial stewardship efforts should therefore be strengthened to promote ADE.

Keywords: Antibiotic de-escalation; Antibiotic streamlining; Antibiotic therapy; Antimicrobial stewardship; Bloodstream infection; Rational antibiotic use.