Improving Sepsis Outcomes in Hospitalized Children Within a Safety Net Hospital
Project Lead: Anna Duran, MD
Project AIM(s):
- To implement a sepsis screening huddle within 20 minutes when patients meet a clinical trigger based on vital signs, 60% of the time within 9 months of initiation
- Ensure patients who screen positive for sepsis receive antibiotics within 3 hours, 60% of the time within 9 months of QI project initiation
- Ensure that patients who screen positive for severe sepsis/septic shock receive antibiotics within 60 minutes, 60% of the time within 9 months of QI project initiation
Narrative Description
The intervention composed of implementing a sepsis trigger tool completed in all pediatric units, computer-based and one-on-one nursing education in the target inpatient unit (General Pediatric Unit), ongoing provider education, sepsis huddle script with appropriate health literacy level, language, and feedback from parent, and an electronic medical record sepsis order set including antibiotics, fluid resuscitation, diagnostic testing and relevant consults.
While the project got off to a slow start due to delays in IRB approval and data abstraction, most outcomes have been accomplished at this point. This project involved a very large working group with variable commitments, different baseline practices among multiple clinical areas, and other challenges. While this project was originally intended to address the inpatient pediatric population in a specific unit, it started in outpatient settings, including the Emergency and Urgent Care departments where workflows are different from inpatient settings. Some processes had to be changed slightly to fit into the workflow of a specific unit. In addition, it was determined that all inpatient settings should be included, which was done on a rolling scale. Challenges and delays with getting order sets and the electronic trigger tool live in the medical record, due to high workload for hospital Information Technology. These IT components were accomplished, and fine tuning continues. The occurrence of three epidemics (COVID, RSV and influenza) last winter led to extraordinarily high census and extreme nursing shortages. This slowed progress on our project. Afterward, the team regrouped and got back on track, although staffing shortages continue.
The sepsis screening huddle was implemented in spring 2022 in most units, although some units were a bit behind, particularly inpatient units as project details were addressed during the roll-out including some programming challenges with the alert. A standard practice for sepsis screening huddles was not in place prior to our QI efforts. In Q4 2022, a sepsis screening huddle was only reported for 2 patients within all inpatient pediatric units. We recognize that huddles were occurring but may not have always been documented. Documentation is being evaluated as patients are pulled according to key pieces of documentation. We have found poor/missing documentation, which leads to coding errors and challenges in data collection and quality improvement. Efforts to provide evaluation and feedback of documentation are ongoing.
Building the infrastructure to have a robust and consistent sepsis response system, with an automated trigger tool, was a gigantic lift. The project required extensive hospital support with weekly meetings including health information technology staff, children’s hospital leadership and nursing stakeholders. It would have been impossible without the support of our hospital quality outcomes office. The biggest rate limiting step was building the electronic trigger tool, then monitoring its use and making needed adjustments to decrease false positive alarms. The second biggest challenge was driving a culture change among residents and charge nurses. For pediatrics, we do not have a dedicated sepsis response team. It falls onto the shoulders of the charge nurses and residents who are already responsible for up to 32 patients on the general pediatric ward. Prioritizing the sepsis alert is an ongoing educational initiative. We also learned about and had to address pragmatic barriers. For example, we discovered that the unit charge nurses get hundreds of secure electronic alerts/messages in addition to sepsis screen alerts per shift. The sepsis alert comes across like any regular text and can easily get lost and/or lead to alarm fatigue. When we used the paper clinical trigger tool, which should circumvent the electronic trigger tool hiccups and also the text limitation, we found that nursing was not regularly using the tool, since it is not built into the regular workflow of nurses charting on the EMR. The clinical trigger tool is a paper tool that they have to remember to refer to and then act on.
Education is now planned as part of annual education for nursing staff. Having “champions” within clinical areas has been an effective method for disseminating information and reminding providers and staff of the processes expected when a sepsis trigger is activated, enacting lasting change in practice. We are in the process of identifying champions in every clinical area. This will serve to remind providers and staff of the processes expected when a sepsis trigger is activated and to hardwire practices
Diagnostic quality problem type, failure, or category (symptoms, observed problems, gaps in performance) addressed by the intervention
- Information gathering
- Information integration
- Information interpretation
- Establishing an explanation (diagnosis)
Root causes/causative factors addressed by the intervention
- Workflow (includes testing, follow-up, and referrals)
- Physical environmental or work system factors
- Health information sharing and accessibility via health IT
Setting of the diagnostic quality improvement intervention
- Ambulatory medical care setting (e.g., clinic, office, urgent care)