156 A Nurse-Led Bedside Coaching Approach to Improve Sepsis Bundle Compliance at a Large Academic Trauma Center

Abstract

Sepsis is a leading cause of mortality in the United States, and the Centers for Medicare and Medicaid have developed a sepsis bundle quality metric focused on a series of timestamped process metrics in an attempt to reduce mortality. We seek to improve the CMS sepsis bundle compliance rate and reduce mortality in the emergency department through the implementation of a nurse-led bedside coaching program. We developed a pre- and post-intervention observational study to analyze sepsis bundle compliance and mortality rates following the implementation of a nurse-led bedside coaching program. Clinical nurse specialists were trained as sepsis coaches with the aim to respond to the bedside of patients with suspected sepsis. The sepsis coach utilized a bundle compliance checklist to direct and simplify the time dependent multi-step treatment process. Sepsis bundle compliance was tracked for 8 months, including the 4 months prior to implementation and 4 months after implementation of the nurse-led coaching role. Patient records with severe sepsis were audited and analyzed for compliance with sepsis bundle metrics. The primary outcomes measured were sepsis bundle compliance and sepsis-related mortality. During the 4 months prior to intervention, 204 patients were discharged from the hospital or died with a sepsis-related ICD-10 code. The cumulative mortality rate was 9.8% over this time period, and the baseline sepsis bundle compliance rate was 32% among a random subset of patients (N=77). The most missed elements were IV fluid (N=47, 61%) and lactate re-check levels (N=26, 34%). In the 4 months post-implementation of the nurse-led coach, 291 patients were discharged from the hospital or died with a sepsis-related ICD-10 code. The overall cumulative mortality rate over this time period was 10.9% and was not statistically different to the 4 months prior to implementation (p=0.7679). However, of 91 patients randomly chosen over this time period, the cumulative bundle compliance was 48.3% regardless of whether a coach was at the bedside (compared to a pre-implementation percent of 32.5%, p-value= 0.054). Surprisingly, the cumulative bundle compliance improved for both coached and non-coached patients suggesting an effect beyond direct coaching (68%, p-value = 0.0003 versus 45%, p-value =0.1155 respectively compared to the baseline 32.5%). Overall, sepsis bundle compliance in the post-intervention period was significantly higher than the pre-intervention period following the implementation of a nurse-led bedside coaching program. Compared to pre-intervention, we found a far greater improvement in sepsis bundle compliance in “coached” cases than “non-coached” cases, but the effect of the coaching program increased compliance in both cases. Changes in overall mortality were variable without a significant change pre- and post- implementation. While a larger multi-centered study may be necessary, our initial findings show that nurse-led bedside coaches can increase sepsis bundle compliance, but perhaps the bundle compliance itself does not improve overall mortality.