Timely administration of effective antibiotics is important in sepsis management. Source-targeted antibiotics are believed to be most effective, but source identification could cause time delays. Objectives: First, to describe the accuracy/time delays of a diagnostic work-up and the association with time to antibiotics in septic emergency department (ED) patients. Second, to assess the fraction in which source-targeted antibiotics could have been administered solely on the basis of patient history and physical examination.Methods
Secondary analysis of the prospective observational study on septic ED patients was carried out. The time to test result availability was associated with time to antibiotics. The accuracy of the suspected source of infection in the ED was assessed. For patients with pneumosepsis, urosepsis, and abdominal sepsis, combinations of signs and symptoms were assessed to achieve a maximal positive predictive value for the sepsis source, identifying a subset of patients in whom source-targeted antibiotics could be administered without waiting for diagnostic test results.Results
The time to antibiotics increased by 18 (95% confidence interval: 12–24) min/h delay in test result availability (n=323). In 38–79% of patients, antibiotics were administered after additional tests, whereas the ED diagnosis was correct in 68–85% of patients. The maximal positive predictive value of signs and symptoms was 0.87 for patients with pneumosepsis and urosepsis and 0.75 for those with abdominal sepsis. Use of signs and symptoms would have led to correct ED diagnosis in 33% of patients.Conclusion
Diagnostic tests are associated with delayed administration of antibiotics to septic ED patients while increasing the diagnostic accuracy to only 68–85%. In one-third of septic ED patients, the choice of antibiotics could have been accurately determined solely on the basis of patient history and physical examination.