Early Goal Directed Therapy Reduces Sepsis Complications and Mortality

PeaceHealth/St. Joseph Hospital
Bellingham, Washington, USA

Team

PeaceHealth/St. Joseph Hospital participates in IHI’s Learning and Innovation Community for Improving Outcomes for High-Risk and Critically Ill Patients. 

 
Faye Lindquist, RN, MN, Chief Nursing Officer
Donald Berry, MD, ICU Medical Director, Sepsis Team Physician Champion
Ralph Weiche, MD, ED Sepsis Team Physician Champion
Sandra Brooks, Critical Services Director
Diana Meyer, RN, MN, Director Clinical Education
Margie Campbell, RN, BSN, ICU Nurse Manager
Becky Stermer, RN, BSN, Critical Services Outcome Coordinator
Marilyn Bufton, PharmD, ICU Pharmacotherapist

 

Aim

Our aim was to reduce mortality associated with severe sepsis and septic shock by 25 percent within one year of implementing the Sepsis Resuscitation and Sepsis Management Bundles based on the IHI and Surviving Sepsis Campaign guidelines. To help us achieve this aim we set the following specific aims:

  • Increase compliance with the Sepsis Resuscitation Bundle to 60 percent by January 2009
  • Increase compliance with the Sepsis Management Bundle to 75 percent by January 2009

 

Measures

  • Sepsis Mortality
  • Sepsis Resuscitation Bundle Compliance
  • Sepsis Management Bundle Compliance

Changes
Changes Implemented:
  • Education at the time of implementation: We created a computer-based training (CBT) module and required that it be completed by ICU and ED nurses within three months of introducing the new order sets. A year later we included a sepsis review with case histories and outcomes data as part of our ICU Skills Day. We made the sepsis CBT available to our physicians and had a continuing medical education program (CME) with Dr. Nguyen from Loma Linda University during the first few months after implementation of the Sepsis Bundles.
  • Timely and detailed feedback to providers: With the addition of an ICU outcomes coordinator, we are able to identify gaps in processes and compliance and give timely feedback to providers. Sepsis outcomes are posted in the ED and ICU and reviewed during medical and nursing staff meetings. Individual feedback to physicians is given by Sepsis Physician Champions; feedback to other providers is given through the outcomes coordinator.
  • Close multidisciplinary partnership between the ED and the ICU: Examining existing processes and implementing changes to support early goal directed therapy for sepsis requires knowledge of the details of delivery of care. We were able to avoid or correct many misconceptions about how these interventions would work because we had bedside caregivers from each unit on our team.
  • Making it easier to do the right thing: Interventions for sepsis resuscitation are more complex than some of the other bundles we use. By creating a streamlined, chronologic resuscitation checklist, we simplified getting the most vital elements of resuscitation completed. We had the entire order set created in our electronic medical record so the unit secretaries did not have to hunt through multiple screens to enter each order. We asked Respiratory Therapy to take charge of asking for central venous oxygen saturation (ScvO2) samples because they are the ones who run them and added it to their electronic worklist.
  • Ensuring that resources are readily available: When we looked at why antibiotics were not being given in a timely way we discovered that there was a delay in getting them from the pharmacy to the bedside. We added the most frequently used broad-spectrum antibiotics to the ED Pyxis to ensure easy availability. We also created a sepsis admission kit, which contains everything (except antibiotics) needed for sepsis resuscitation.
  • Early identification for inpatients: We have improved early identification of inpatients with severe sepsis or septic shock through an electronic surveillance report delivered by email to the ICU outcomes coordinator along with twice daily housewide rounding by the ICU stat nurse. Coordination between the stat nurse and the outcome coordinator has resulted in early identification of patients who meet criteria for severe sepsis and are sometimes able to avoid a transfer to ICU by early intervention with fluid resuscitation and timely antibiotic administration.
 
 
Bundle Elements:
Sepsis Resuscitation Bundle [*Modified from IHI's Sepsis Resuscitation Bundle]
All elements to be achieved within six hours of presentation with systemic inflammatory response syndrome (SIRS) criteria and severe sepsis/septic shock criteria:
  1. Lactate drawn
  2. Blood cultures before antibiotics
  3. Broad spectrum antibiotics within three hours
  4. Intravenous fluid administration to achieve central venous pressure (CVP) of more than eight 
  5. Pressors to maintain mean arterial pressure (MAP) greater than 65 if blood pressure not responsive to initial fluid boluses
  6. Central venous oxygen saturation (ScvO2) greater than 70 percent if patient requires pressors to keep MAP greater than 65 or initial lactate greater than four
 
Sepsis Management Bundle [*Modified from IHI's Sepsis Management Bundle]
  1. Steroid administration per policy (or document reason why not given)
  2. Xigris administration per policy (or document reason why not given)
  3. Glucose control greater than 70 and less than 150
  4. Maintain median inspiratory plateau pressures (IPP) less than 30



Results

graph_sepsisresuscitationbundle.jpg

graph_sepsismanagmentbundle.jpg

graph_sepdidmortality.jpg

table_mortalityindividualfactors.jpg
 

 

Summary of Results / Lessons Learned / Next Steps

Summary of Results:
To address our overall goal of reducing morbidity and mortality related to severe sepsis and septic shock, a multidisciplinary team with representatives from both the ED and the ICU created order sets for sepsis resuscitation and management based on the IHI and Surviving Sepsis Campaign guidelines. We found that the partnership between the ICU and the ED, as well as ongoing surveillance and feedback to individual providers, improved our Sepsis Resuscitation Bundle compliance and overall sepsis mortality.
 
By using Sepsis Resuscitation and Sepsis Management Bundles we have reduced mortality related to severe sepsis and septic shock from 41 to 50 percent in the two months prior to implementation to an average of 17 percent after implementation. We continue to struggle to achieve all elements of each of the bundles, particularly measuring and achieving ScvO2 and CVP targets during the resuscitation phase and maintaining glucose control during the management stage. We achieved 63 percent compliance with the Sepsis Resuscitation Bundle in January 2009 but it decreased to 50 percent by March, primarily due to noncompliance with ScvO2 monitoring. Compliance with the Sepsis Management Bundle dropped precipitously in February from 73 percent to 29 percent, primarily due to poor glycemic control. A closer look at the data revealed that six out of 14 patients had hypoglycemic episodes and five of 14 had a failure to keep the median glucose less than 150. 
 
The Mortality by Individual Factors table above includes data from December 2007 to January 2009. It shows that higher rates of completed bundle components are associated with surviving sepsis and illustrates the struggle we have to achieve high levels across all components of the bundle.
 
Keys to the gains we have been able to achieve so far have been:
  • Close cooperation between ED and ICU multidisciplinary staff. It has been invaluable to have engaged physician champions from both the ED and ICU staff and to have the practical insights of both ED and ICU nurses.
  • Close scrutiny of the details of how the work is done. It is so important to involve bedside staff and go through the process in detail. ED nurses told us about the hold-up with getting antibiotics to the bedside. Unit secretaries explained to us how time consuming it was to enter each order, so we had the entire order set bundled in the electronic medical record. Physicians told us about the difficulty of inserting the ScvO2 central lines and the drawback of having to remove it when patients need an MRI. We changed to a ScvO2 probe, which is more easily inserted through a standard central line and can be removed without having to pull the central line. 
  • Timely feedback and ongoing education. 
  • ICU Stat nurses, who assist during resuscitation in the ED and help identify septic inpatients early.
 
Next Steps:
Our next steps are to implement a nurse-initiated sepsis resuscitation protocol for ED and ICU nurses or Stat nurses. Using a nurse-initiated protocol when a patient meets both SIRS and Severe Sepsis or Septic Shock criteria, an ICU or ED nurse would notify a physician and proceed without delay to draw labs, order some initial diagnostic tests (including lactate and blood cultures, CXR and EKG), and start fluid resuscitation.
 

Contact Information

Becky Stermer, RN, BSN
Critical Services Outcome Coordinator
PeaceHealth/St. Joseph Hospital
bstermer@peacehealth.org
 

Average Content Rating
(1 user)
Please login to rate or comment on this content.
User Comments