Automatic identification of community-onset sepsis among patients attending an emergency department – a validation study

Abstract

Abstract Purpose To develop and validate a software application able to identify cases of community-onset sepsis based on Sepsis-3 criteria among patients visiting an emergency department.Methods An application, developed with the software QlikView®, accessed data from the electronic health record systems TakeCare® and CliniSoft®. It utilized indicators like blood culture data, antibiotic administration, and Sequential Organ Failure Assessment scores to detect sepsis cases likely to fulfil Sepis-3 criteria. The application was tested retrospectively against a cohort from a large city hospital in Stockholm during a period of two years. The performance of the application was compared to physician record review in a subset of cases identified by stratified random sampling.Results Among 229,195 emergency department visits leading to 60,213 hospital admissions, the application detected 7,027 cases of sepsis. Validation using physician record review of totally 426 cases as gold standard demonstrated a sensitivity, specificity, positive predictive value and negative predictive value of 94%, 98%, 92% and 99%, respectively. The lower respiratory tract was the most common site of infection.Conclusion This software application effectively identified community-onset sepsis patients using electronic health record data with high performance. It has the potential to improve sepsis case identification, enable further research on sepsis incidence and mortality, and support adherence to treatment guidelines as well as contributing to enhanced sepsis surveillance and the implementation of quality improvement interventions.