ED Sepsis Alert: Improving Sepsis Care in the NMH Emergency Department
Joshua Zimmerman, MD, Colin McCloskey, MD, Kory Gebhardt, MD
Severe sepsis and septic shock are clinical entities encountered frequently in the ED. Mortality related to severe sepsis is directly related to the time to antibiotic and fluid resuscitation. The most recent guidelines established by the Surviving Sepsis Campaign have been adopted by the Center for Medicaid Services (CMS) as a benchmark measure in quality as of October 2015. This bundle includes broad-spectrum antibiotic therapy after blood cultures, measurement of a serum lactate level and intravenous crystalloid resuscitation of 30 milliliters/kilogram.
Our project aims to improve the early identification of severe sepsis and subsequently the time to completion of the three-hour bundle in the Northwestern Memorial Hospital Emergency Department by incorporating serum lactate levels into the existing triage process. Our project began with an analysis of our current ED based sepsis care. We analyzed whether a lactate measure was obtained, and how this impacted antibiotic administration times as well as the ESI triage score assigned to all patients with a final diagnosis of severe sepsis or septic shock. We found that, regardless of ESI category, our patients were consistently receiving antibiotics over 3 hours from initial presentation. We further found that this was even more pronounced for lower ESI categories and for patients with low lactate levels.
After this initial analysis, we implemented our intervention., which consisted of incorporating a serum lactate measurement into the triage process. If a serum lactate level greater than 2 was obtained on any patient with signs or symptoms of an infectious process, they were treated as potentially septic and elevated to an ESI 1 triage category. This was done independent of vital signs and other factors that would otherwise alter their ESI category. A cutoff of 2 was selected given this was the level that was selected in the CMS criteria.
We analyzed some preliminary data approximately seven weeks post-intervention, and found 2.5 times more serum lactate levels were being sent from triage than prior to intervention. Comparing the total serum lactates sent during the first 6 hours of ED stay, pre-intervention, 1,495 measures were sent, while post-intervention 1,818 were sent. Additionally, the median time to measurement decreased by 5 minutes on all qualifying patients, which includes all those with a serum lactate >2. Our results, to date, are limited and have not demonstrated a significant decrease in time to antibiotics. However, the early results are promising. There has been a significant increase in measurement of serum lactate levels, which was one goal of our project. We are hopeful that with further data and post-intervention analysis, our intervention will continue to show greater compliance with the 3-hour CMS bundle and eventually improvement in time to antibiotic therapy.