Characteristics and outcomes of patients assessed for suspected sepsis the emergency department

Background Structured sepsis screening and management tools are recommended. Evaluation of their effectiveness is limited by spectrum bias, with apparent improvements in outcomes due to capturing a broader population with overall lower baseline risk. The aim was to determine the true incidence of sepsis and septic shock and outcomes among patients assessed for suspected sepsis in an adult tertiary hospital Emergency Department (ED). Methods A retrospective observational study identified patients with suspected sepsis admitted via the Emergency Department (ED) during the calendar years 2019 and 2020. Patients were subsequently classified into four groups: infection, sepsis, septic shock and non-infection using standardised objective criteria. Baseline characteristics, processes of care and clinical outcomes were compared between groups. Results Five hundred and ninety-five patients were assessed for suspected sepsis in the ED. These were subsequently classified as 211 (34%) with infection, 266 (46%) with sepsis, 58 (10%) with septic shock, and 60 (10%) with an alternative diagnosis. Rates of organ support, intensive care admission and mortality varied significantly between groups. Conclusions An ED sepsis screening program identifies patients with a range of clinical phenotypes, with 44% of patients misclassified. Measuring the effectiveness of sepsis recognition tools based on those assessed for suspected sepsis will generate misleading results. Standardised confirmation of sepsis is required to adjust for baseline risk and generate credible results to drive quality improvements, and inform policy-making and resource allocation.