Abstract
BACKGROUND: Timely administration of antibiotics is critical in the management of infectious diseases, particularly in preventing progression to sepsis. Despite the urgency, the appropriate timing of antibiotic treatment, especially in non-septic infections, remains unclear. This study aimed to assess the association between antibiotic timing and progression to sepsis among patients admitted to the Emergency Department (ED) for suspected infection. METHODS: A retrospective cohort study utilized data from three tertiary-care hospital EDs between January 2021 and June 2023. Adult patients hospitalized for clinical infection were included. The primary outcome was sepsis development, while secondary outcomes included hospital mortality, Intensive Care Unit (ICU) admission, ICU length of stay (LOS), and hospital LOS. The main exposure was the duration from ED arrival to initial antibiotic administration. Multivariable logistic and negative binomial regression were employed to adjust for confounders and assess associations. RESULTS: The study included 1279 infected adult patients, with 20.5% developing sepsis and 10.3% admitted to the ICU, resulting in 3.8% in-hospital deaths. The median time from ED arrival to initial antibiotic administration was 123 min (interquartile range(IQR), 79-241 min). Although per-hour delays in antibiotic administration showed association with sepsis development (Adjusted Odds Ratio(aOR) (95% CI) 1.071(1.044-1.099); P < 0.001), the primary threshold analysis demonstrated that significant associations with sepsis (aOR (95% CI), 3.468 (2.131-5.623); P < 0.001), in-hospital mortality (aOR 2.487, 95% CI 1.083-5.440; P = 0.026), adverse clinical outcomes, and prolonged LOS emerged only when delays exceeded 12 h. These findings were also confirmed by sensitivity analyses. CONCLUSIONS: These hypothesis-generating findings suggest that, in non-severe infections, delays within 12 h might not be associated with significantly increased risks, potentially allowing time for diagnostic clarification without apparent harm. However, due to the observational nature of the study and potential biases, prospective studies are required to confirm these associations.