Human-Machine Partnership: Data-Facilitated, Targeted CLABSI Reduction in the Face of a Global Pandemic

Executive Summary

From calendar year 2019 (CY19) to calendar year 2020 (CY20), as the world grappled with climbing COVID-19 cases, central line-associated blood stream infections (CLABSI) in the United States saw a 24% increase in the national standardized infection ratio (SIR), and a 22% increase in Pennsylvania.¹ At Jefferson Health, which comprised two divisions with six acute care hospitals on a common electronic medical record (EMR) at the time, CLABSI SIR increased by 3.98% over the same period. While this infection rate was lower than at national or state levels, even a modest increase in CLABSI SIR leads to higher patient morbidity, mortality, and longer hospital stays—threatening patient health and taxing hospital resources already strained by the novel coronavirus. Enmeshed in a global pandemic, handling the highest number of COVID-19 cases in the region, and staring down a February 2021 merger onto a single EMR that would see expansion to four divisions and eleven acute care hospitals, work was already in motion at Jefferson Health to address the issue of rising numbers of CLABSI cases. In Fall 2019, Jefferson Quality and Safety leadership identified five to six hot-button quality and safety measures for intensified focus and action planning: CLABSI, catheter-associated urinary tract infection (CAUTI), clostridioides difficile (C.Diff), restraint management – violent and non-violent, and venous thromboembolism) (VTE). The Jefferson Enterprise Analytics team then incorporated these measures into a new Epic Radar dashboard aimed directly at preventing and treating them, called the OnPoint Quality and Safety Huddle Dashboard – custom-built by the Jefferson team. The platform was designed to be immediately interpretable with big, easy-to-read meters, much like a car’s dashboard, and would provide all that the user needed to take corrective actions, at the push of a button. Released in a phased approach through CY20 and CY21, the OnPoint Quality and Safety Huddle Dashboard led to a marked decrease in CLABSI SIR in each division following implementation, culminating in a 28% reduction in CLABSI SIR for the health system as a whole in CY20 to CY21.   

Lessons learned include:

Avoid mission creep. Often, a tool like the OnPoint Quality and Safety Huddle Dashboard runs the risk of becoming ineffective, by virtue of trying to make it inclusive of every imaginable issue faced in the hospital. Opportunities for improvement then get lost in the noise of having too many measures, being packed too tightly, and the impact of, “if everything is important, then nothing is important.” At every stage of development, it is vital that the team remains focused on the key objectives and measures that will have the greatest impact.

Build in time to refine. Over the course to full deployment, the process measures informing central line care compliance were revisited more than once, with more nuance added each time to better reflect the reality of patient care.

Alignment is key. Discovery of a lack of full alignment, which impacted tracking ability, led to full enterprise alignment on standardized central line dressing change procedures and policy.

Education is critical to adoption. While real-time data is a critical element of the dashboard, it became clear during development/dashboard testing that more end user education would be needed to understand the difference in supporting report structure and timing, versus batch job timing for the report driving the dashboard. The slightest difference in time of report runs, which is inevitable, could wind up showing different numbers—both being accurate, dependent on the time of each run. Dashboard workshops were held virtually and, ultimately, tip sheet links were embedded directly on the dashboard, to always ensure immediate access to the most current version of education on the tool.