For decades, bloodstream infection risk was believed to be part of the trade-off of central lines: they provide life-saving access that wasn’t possible prior to their invention, but the process is risky.
Central lines provide access directly to the major blood vessels of the body. They allow treatments and monitoring that cannot be delivered through smaller peripheral vessels. Direct access for treatment, however, is also direct access for any contaminants. Bloodstream infections, known as CLABSIs when they are connected to a central line, are particularly devastating for critically ill patients. Epidemiologists began to identify reliable methods of preventing CLABSIs relatively early on, but implementing them would prove challenging.
By 2005 hospital leaders knew what prevented CLABSI but they didn’t know how to make sure those things were reliably done every time by the thousands of caregivers doing them, often under emergent circumstances. Teaching caregivers was not enough. Even teaching them why and how those things ought to be done was not enough. Caregivers had to do all the prevention measures, every time. Observational data was telling us that well-intentioned care providers were only hitting every best-practice mark about 65% of the time.
The Michigan Keystone ICU Project
Then came the Michigan Keystone ICU Project led by Dr. Peter Pronovost. The project was a state-wide initiative launched in 2003 by the Michigan Health & Hospital Association and Johns Hopkins University with the ambitious goal of virtually eliminating central line infections in Michigan ICUs and changing the industry’s understanding of acceptable risk.
It is widely remembered for its use of best practice checklists that were easily observable and kits that made it quick and easy to gather the materials for maximal sterile barrier precautions at the time of insertion. Two key components of the project “education” and “executive support”, are often mentioned but not elaborated upon.
In reality, the project approached the issue of improving practice reliability from two angles simultaneously. Checklists and supply kits provided clarity and the physical materials to follow the prevention steps while education and executive support fostered a cultural shift towards high reliability in using best practices.
More than just “staff education”, the work included both RN and MD leads who received education, had every other week check-ins with peers and leaders, and attended statewide peer conferences twice annually to further their knowledge and strengthen their social ties to other safety leaders. Participating hospital executives committed to ensuring that the physician and nurse leaders could devote 20% of their time to the project. That protected time was critical, allowing leads to support change at the bedside through real time coaching, reinforcing expectations and removing barriers.
The goal of the project was reliability in practice: clinicians needed to do every preventative step, every time to reduce CLABSI to near zero rates. Although the steps were widely known, some degree of deviation was culturally accepted when the circumstances of patient care threw up roadblocks.
Perpetually pressed for time, doctors and nurses might proceed without a missing safety supply rather than hunting one down, or rush contact time for an antiseptic rather than delay caring for the next critical patient. Individual small deviations from best practice each add only a small increment of risk and the hospital environment supplies a never-ending stream of competing priorities.
This investment of time and resources into staff education, particularly in those staff who volunteer to lead cultural change, is a powerful signal from organization leaders showing what they value.
Nurse and physician time are incredibly valuable resources for hospital operations. By dedicating clinician time to CLABSI prevention executives not only provided clinician leads with time to carry out improvement work, but also demonstrated in ways that were visible at all levels of care that the initiative was important enough to dedicate scarce resources to.
And it worked. CLABSI rates in Michigan ICUs fell to near zero in the first 3 months, with a sustained reduction of 66% over the next 18 months. Moreover, the project demonstrated that reliability in the use of best practices is possible, even across the thousands of care providers working around the clock in over 100 Michigan ICUs. A nation of hospitals was on notice: infection rates that previously represented acceptable risk were now evidence that improvement work was needed.
CLABSI Prevention Today
Today CLABSI prevention checklists are found in every hospital and insertion kits are available from every vendor. The legacy of Keystone is a shift in understanding what is possible in prevention yet not all aspects of the Keystone project are as easy to reproduce. The investment in cultural change, that costly and conspicuous commitment to finding leaders in the clinical space and providing them with time, education, and the support of colleagues, can’t be purchased at scale. It requires focus and follow-through from executive leaders in each hospital.
The difference between what Keystone project participants invested and what I have observed during my own years in infection prevention is striking. Education and executive support are boiled down to a 1-hour webinar, a 15-minute in-service, and a note on the huddle board. Executives seek champions for prevention work but provide no time or resources.
Keystone taught us the value of standardization and observable, repeatable practices and changed the health care industry’s baseline for acceptable risk. The project demanded that bedside clinicians put prevention first, even when it cost precious time needed elsewhere. It also demanded executives make a similar investment: designating clinician time to do the necessary work of implementation even though the cost is high. More than 20 years after Keystone we have learned to expect clinicians to do the hard things required for prevention. I’m not sure we learned to expect the same from the organizations asking them to do it.