Developing a "Sepsis Never Event" Measure for Our Hospital-wide Improvement Initiative

Background: Historical measure definitions for our sepsis initiative captured intent-to-treat without connection to final diagnosis or patient outcomes. Objectives: To develop a novel sepsis outcome measure capturing confirmed sepsis-related patient harm including a reliable process to identify cases fulfilling its criteria. Methods: Our sepsis leadership team reviewed existing sepsis measures. Through iterative review and consensus, we defined a measure prioritizing final diagnosis rather than intent-to-treat. We created an automated identification and manual chart review process to identify cases meeting criteria (Figs. 1, 2). Cases meeting criteria trigger in-depth case reviews in partnership with system-wide patient safety teams and unit leaders.Fig. 1.: Case review by our leadership team aimed to obtain consensus on the following questions to identify cases meeting criteria for our Sepsis Never Event measure.Fig. 2.: Steps for identification of Sepsis Never Events with automated identification, manual review, and leadership group consensus.Results: Sepsis Never Events are confirmed sepsis cases with infection-related organ dysfunction, antibiotic order-to-delivery >180 minutes, and subject matter consensus that antibiotic delay led to patient harm. Conclusions: Our novel measure targets zero harm for patients with sepsis and identifies deviant cases to facilitate in-depth review and continuous improvement.