Prescription for Change

How to scale sepsis protocols across an entire health system

Clinicians know how sepsis should be treated. That isn’t the challenge. The hard part is building a care system that organizes all the moving pieces, ensuring the care team delivers the right treatment every time. Over the past year, we identified ten imperatives to standardize and expedite sepsis care, featuring tested strategies from organizations like Wake Forest Baptist Health and Inova Mount Vernon Hospital.

It's hard—but necessary—to scale best practice across hospitals

Hardwiring best practice in a single hospital—in the ED, ICU, and on the floor—is a tremendous achievement, resulting in lives saved. But, for most organizations, the work can’t stop there. This best practice needs to be replicated across the entire system, and multiple hospitals. Scaling a sepsis care standard isn’t easy. Systems have to establish a new care standard, contend with local resource and staffing limitations, and secure support for broad practice change among many different care team members at each site.

But it’s not impossible—in the past year, Carolinas HealthCare System scaled a best-in-class sepsis care model from one of their facilities across 14 other hospitals and satellite emergency departments in the system. Between April and July 2014, compared to the same months in 2013, their sepsis mortality rate reduced by 12%. I spoke with Dr. Alan Heffner, the medical director of the ICU at Carolinas HealthCare System’s Carolinas Medical Center and physician champion for the sepsis initiative, to learn from their success.

Advice from Carolinas HealthCare System

Carolinas started out like many other large health systems. Clinicians had varying views of sepsis presentations, and, too often, this led to underestimating the severity of illness and delaying essential treatments.

But, one hospital in the system was getting sepsis care consistently right, and had the outcomes to prove it. So, Dr. Heffner and other clinical leaders decided to use this model as the foundation from which to build a standard sepsis protocol that would be hardwired throughout the system—they branded it “Code Sepsis.” Here’s the advice they would give to others who also want to hardwire sepsis care across a large system:

Don't reinvent the wheel.
There’s a good chance “best practice” (or better practice) exists somewhere in your system. Compare outcomes data and interview key stakeholders to find it. The model you identify may not be directly applicable to other facilities “as is,” but it’s a good starting point.

Get broad input, then scale simultaneously.
Carolinas didn’t roll out Code Sepsis facility by facility—this would have allowed too much variation to creep in. Instead, they created a multidisciplinary sepsis collaborative, representing the entire system, to vet and approve Code Sepsis. The group reviewed the clinical evidence and protocols, but largely spent their time deciding what tools and processes they would need to build to support broad practice change, track progress, and hold clinicians accountable. These decisions reflected input from the entire system, and were pushed out to all facilities at once.

Build workflow tools to support consistent practice.
Broad practice change is hard, and telling the care team to follow a new sepsis protocol is not sufficient. That’s why the sepsis collaborative developed clearly defined standards, education, clinical tools, and order sets to make it easy for frontline clinicians to understand and use the Code Sepsis clinical pathway.

One successful resource was a simple Code Sepsis poster to remind clinicians how to assess a patient’s sepsis risk, and when and how to initiate Code Sepsis treatment. Another resource was a Code Sepsis aftercare audit tool, a feedback tool on bundle adherence that was disseminated to clinical team members within days of the patient encounter.

Provide dedicated, boots-on-the-ground support.
Carolinas has a full-time Sepsis Program Coordinator (an experienced ICU nurse) at each facility. Initially, the Coordinators’ focus was educating and training clinicians, and bringing frontline feedback to the collaborative.

Now that Code Sepsis is in place, the coordinators track adherence to the sepsis bundle, provide coaching to non-adherent care teams, and look for common challenges to address through system improvements. Acknowledging the costs associated with a dedicated role, the right coordinator can be an important catalyst for rollout at the local level.

Track and share data on progress.
Carolinas ensures data is meaningful by sharing two types of data with critical constituencies. First, they share hospital-level outcomes data, like mortality, length of stay, and cost per case, monthly with hospital leaders. All facilities are publicly compared, so there’s a healthy dose of peer pressure. The data are shared at monthly sepsis collaborative meetings and provide a platform to discuss real life cases and lessons learned.

Second, they share process-related adherence data with frontline clinicians. Sepsis Program Coordinators manually abstract data to complete the Code Sepsis audit tool for every patient to provide feedback to the clinical care team, with an emphasis on opportunities to improve care. The completed form is emailed to the relevant care team members and ICU/ED physician champions within days of the patient encounter. Manual data abstraction is a big time investment, but Carolinas HealthCare System finds timely, accurate and expert feedback is a compelling force for improvement that is worth the effort.

Get an executive sponsor.
A system-wide initiative won’t get off the ground without real support from leadership—both because of the signal value and resource requirements. At Carolinas, facility CMOs and medical directors have goals tied to sepsis performance. This ensures that Code Sepsis stays top-of-mind and that leaders are communicating its importance regularly to staff.

Results from 'Code Sepsis'

These strategies have paid off for Carolinas and their patients. In addition to the mortality reductions cited above, from Q4 2013 to Q2 2014, the mean hospital LOS for patients for whom Code Sepsis was activated was 1.1 days shorter than sepsis patients for whom Code Sepsis was not activated. This represents an estimated cost avoidance of over $259,000 based on variable hospital cost per day of $370.


Make sure you don't miss a post

Subscribe to updates from Prescription for Change and get all the lastest research and insights for CMOs delivered right to your inbox.