Assessment for Implementation Methods in Sepsis

Sepsis is the leading cause of admission to intensive care units (ICUs) in the U.S., and the leading cause of respiratory failure and death in ICUs. The majority of ICU patients with sepsis or septic shock have either respiratory infection as the source of sepsis, or have respiratory failure requiring mechanical ventilation. In recognition of the burden of sepsis in the U.S., sepsis "bundles" were introduced to facilitate guideline implementation in clinical practice (known as the 3-Hour bundle). Since the introduction of sepsis bundles, multiple observational studies have demonstrated a consistent, strong association between implementation of sepsis bundles and improved survival. These data led to the New York State (NYS) Sepsis initiative, which demonstrated a significant association between adherence with sepsis bundles and improved survival, and the Centers for Medicare and Medicaid Services (CMS) mandated public reporting of sepsis measures (SEP-1). Analysis of the NYS database has revealed that completion of the 3-Hour bundle in patients with respiratory failure was associated with an 8.6% absolute reduction in mortality (18.5 RRR). For those patients who completed the 3-Hour bundle within 1 hour, the mortality reduction was even higher, 9.8% (RRR 21.7%). In 2018, the Hour-1 bundle was published to underscore the need for urgency in the treatment of septic patients. We believe that the primary beneficial effect of both the Hour-1 and 3-Hour bundle is in patients with respiratory failure. It is not known if implementation of the Hour-1 bundle reduces mortality more than the 3 Hour bundle. Although adherence with the 3-Hour bundle (SEP-1) is mandated by CMS, compliance is moderate (60%), suggesting an active implementation process for the 3-hour bundle is necessary to compare the Hour-1 bundle to the 3-Hour bundle. The current proposal is a pragmatic, cluster-randomized clinical trial using a hybrid type 2 effectiveness-implementation approach to evaluate mortality and respiratory failure-based outcomes and bundle adherence, in emergency room patients with sepsis. We will compare the hour-1 bundle to the 3-hour bundle. The outcome measures include hospital mortality, hospital length of stay, ventilator-free days, and incidence of respiratory failure. The effectiveness of a clinical intervention (Hour-1 bundle) is implemented using a rigorous implementation strategy (the Exploration, adoption/Preparation, Implementation, Sustainment - EPIS - multi-level conceptual model) for both the 1- and 3-Hour bundles. We will also evaluate a possible precision-based approach in this study. Routine clinical information available at hospital presentation will identify 4 discrete, sepsis phenotypes and we hypothesize that 2 of these identify patients who are significantly more likely to benefit from the 1-Hour bundle in future studies. Our team has extensive experience conducting multi-center trials in sepsis guided by effective models of implementation science. We have developed automated screening tools to identify patients with sepsis and have guided state-wide implementation of sepsis performance measures. We have also used automated EHR methods to screen for and enroll septic patients into these studies. This novel, hybrid type 2 effectiveness-implementation approach addresses key gaps to facilitate the implementation of evidence-based strategies to improve patient outcomes from sepsis and acute respiratory failure.