Formative Evaluation to Inform Implementation of Sepsis Bundles in Emergency Departments.

Abstract

Sepsis bundles are effective in reducing mortality for sepsis, the leading cause of admissions to intensive care units. However, little is known about factors that hinder and facilitate the delivery of sepsis bundles in emergency departments. Guided by the Exploration, Preparation, Implementation, and Sustainment (EPIS) Framework, the objective of the present study was to characterize determinants of implementing sepsis bundles in emergency departments in the United States. A formative evaluation guided by the EPIS was conducted with nurses, physicians, and pharmacists using qualitative interviews (N=66 interviews) and quantitative surveys (N=86 surveys). Qualitative interviews were analyzed using a matrix-guided rapid analytic approach. Quantitative surveys were analyzed descriptively. Scores on quantitative measures indicated that participants perceived evidence-based interventions positively and had moderately positive perceptions of implementation climate and organizational readiness for change. Qualitative results indicate barriers aligned with the EPIS constructs of outer context (eg, perceived rigidity related to Centers for Medicare & Medicaid Services bundle requirements), inner context (eg, alert fatigue), bridging factors (eg, staffing shortages), and innovation factors (eg, concerns about fluid overload). Participants reported facilitators including nurse-driven protocols (inner context) and bundles being straightforward (innovation factors). Findings highlight areas that warrant attention during sepsis bundle implementation efforts in emergency departments, including staff turnover, hesitancy in initiating bundles, competing or misaligned priorities related to bundle implementation, and developing hospital cultures that promote sepsis bundle compliance. These findings will be used to tailor implementation strategies that will support bundle compliance in a hybrid type II effectiveness-implementation trial.