Early empirical antimicrobial therapy does not prevent sepsis development in critically ill surgical patients with suspected nosocomial infection: a retrospective analysis

Background: Early administration of antibiotics to septic shock patients decreases in-hospital mortality, but there is a lack of studies evaluating the role of early empirical antibiotics in surgical patients with suspected nosocomial infection without sepsis. Methods: Retrospective cohort of adult patients admitted to a surgical Intensive Care Unit in a tertiary hospital. We defined early empirical antibiotic group by the initiation of antibiotic therapy within 24h after infection’s suspicion, and conservative group by antibiotic therapy initiation 24 hours after infection’s suspicion or not prescribed within 14 days. The primary outcome was a composite of death, septic shock or sepsis within 14 days from the clinical suspicion of infection. Regression models were used to identify associations between factors and the primary outcome. Results: From 2007 patients admitted to intensive care unit, 341 surgical patients (71% trauma patients) with suspected nosocomial infection without sepsis and with no obvious source of infection were included in the cohort. Age, gender, traumatic brain injury, admission type (trauma vs. non-trauma), SAPS 3, SOFA, vasopressor use or rate of mechanical ventilation did not differ between early empirical antibiotic and conservative groups. In the conservative group, 57% of patients received antibiotics within 14 days. The composite primary outcome occurred in 41% of patients in the conservative group and 56% in the early empirical antimicrobial group, (p=0.007). The 14-day incidence of septic shock or mortality was similar in both groups. Multivariate analysis showed early antimicrobial therapy (OR 1.83 [95% CI 1.16-2.88] , p = 0.008), non-trauma admission (OR 2.32 [1.40-3.90], p = 0.001) and mechanical ventilation (OR 2.09 [1.31-3.35], p = 0.002) were associated with the primary outcome. Exploratory analysis including only patients with positive cultures also did not find any benefit of early empiric antibiotic therapy (OR 1.39 [0.78-2.49], p = 0.26) Conclusions: Early empiric antibiotic therapy does not decrease the incidence of sepsis, septic shock or death within 14 days in critically ill stable surgical patients with suspected infection but with no obvious source.