Inadequate empirical therapy and unnecessary broad-spectrum antibiotic use in community-onset bloodstream infection: A retrospective cohort study

OBJECTIVES: To evaluate empirical antimicrobial therapy (EAT) adequacy, risk factors for inadequate EAT (IEAT) and multidrug-resistant (MDR) infection, and unnecessary broad-spectrum use in emergency department (ED) patients with bloodstream infection (BSI). METHODS: Retrospective cohort study of adults with community-onset BSI at a tertiary-care ED (2022-2024). Adequate EAT (AEAT) was ≥1 administered agent with in vitro activity. Logistic regression identified factors associated with IEAT, unnecessary broad-spectrum therapy, and MDR infection. Cox regression assessed 7- and 30-day mortality. RESULTS: Among 1,716 BSI episodes, 464 (27%) received IEAT or no EAT. MDR organisms accounted for 370 cases (21.6%) and were associated with IEAT (aOR 7.40, 95% CI 5.55-9.93). Of 606 patients receiving broad-spectrum EAT, 490 (80.9%) regimens were unnecessary. Community-acquired infection was associated with unnecessary broad-spectrum therapy (aOR 3.96, 95% CI 1.65-11.79), and healthcare-associated infection with MDR infection (aOR 2.02, 95% CI 1.31-3.14). Seven-day mortality was associated with polymicrobial infection (aHR 2.14, 95% CI 1.39-3.30), long-term care facility residence (aHR 1.72, 95% CI 1.20-2.46), and vasopressor requirement (aHR 3.20, 95% CI 2.21-4.62). CONCLUSIONS: IEAT was associated with MDR pathogens, and unnecessary broad-spectrum use was common. Improved MDR risk stratification may optimise EAT by reducing inadequate coverage and avoidable broad-spectrum exposure in the ED.