LAUSR

Key Points Question Do existing administrative data methods commonly used for quality and research purposes adequately capture sepsis events in critically ill patients with traumatic injury? Findings In this cohort study that included 3194 adults, using automated clinical database (Third International Consensus Definitions for Sepsis and Septic Shock [Sepsis-3]) criteria identified significantly more cases of sepsis, with worse morbidity outcomes, compared with using the National Trauma Data Bank and administrative coding methods. Meaning The results of this study suggest that administrative classification methods misclassify and underestimate sepsis events among severely injured patients with trauma, suggesting that new analytic approaches are needed for this population.