This is the authors' abstract. We don't add key points for this paper.
BACKGROUND: Patients with suspected bloodstream infection often receive broad-spectrum antibiotics with anaerobic activity in the absence of clinical indication for anaerobic coverage. Anti-anaerobic antibiotics have been linked to adverse clinical outcomes in other populations, potentially by depleting intestinal anaerobes. METHODS: We conducted a planned sub-study of the multisite BALANCE randomized controlled trial of antibiotic duration for bloodstream infection to assess the impact of anti-anaerobic antibiotics (receipt from three days pre-index culture to seven days post-index) on mortality and gut microbiome composition with metagenomic sequencing in patients without clinical indication for anaerobic coverage who survived to seven days post-index culture. The primary exposure was receipt of anti-anaerobic antibiotics from three days prior to the index culture to seven days post-index culture. RESULTS: Among the 2851 eligible patients included in our primary analysis, 2106 (74%) received anti-anaerobic antibiotics and 745 (26%) did not. After balancing measured potential confounders through inverse probability of treatment weighting, anti-anaerobic antibiotics were associated with higher 90-day mortality (OR = 1.41, 95% CI 1.03 to 1.92, p = 0.03) and depletion of gut anaerobe relative abundance (fixed effect estimate = -16.59, 95% CI -30.67 to -2.52, p = 0.02). Increased duration of anti-anaerobic antibiotics was associated with greater mortality risk and additional gut anaerobe depletion. CONCLUSIONS: Anti-anaerobic antibiotics are associated with increased mortality and gut microbiome disruption in patients with bloodstream infection. Minimizing exposure to anti-anaerobic antibiotics for bloodstream infection should be further explored in clinical trials as a potential treatment strategy to improve patient outcomes.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Armstrong et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: