Abstract Antimicrobial resistance and inappropriate antibiotic use are urgent threats to global health and patient safety. Early antibiotic prescribing often occurs despite uncertain infection status, leading to unnecessary exposure and contributing to resistance. This article advocates for strengthening antimicrobial stewardship by integrating diagnostic stewardship and Bayesian reasoning to more explicitly consider disease probability in clinical decisions. In particular, probability thresholds should be included in clinical guidelines to inform testing decisions that may prompt antibiotic initiation. We provide several clinical examples and highlight the need for clinician education in Bayesian reasoning, underpinned by real-time decision support tools. Anchoring clinical practice in probability-based frameworks promises to refine antibiotic prescribing, enhance patient safety outcomes, and curb the progression of antimicrobial resistance.
Tsalik et al. (2026) studied this question.