Skip Navigation
Print This Page
Share this page: More

Quality Update - Improvements That Last

Summer 2010

Improvements That Last

By: Peter Pronovost
Date: July 23, 2010

Johns Hopkins' patient safety guru shares his tips on designing quality improvement projects that have an enduring impact.

Peter Pronovost

Too often, successes in safety and quality improvement are short-lived: A project is launched with great enthusiasm. Early results show promise. Perhaps the leaders write a paper on it. And a year later it’s dead in the water.

There are many reasons for this. Sometimes the project is a pilot, and when it’s over the group loses the funding, resources or internal support required to keep it going. In other cases the project may feel like extra work for staff, who forget about the effort once the next initiative rolls along.    

Very few safety and quality improvement projects actually work in the first place, and even fewer sustain success. But it is possible. This year, a Hopkins-led team of researchers published a paper in the British Medical Journal about a long-running effort to reduce central catheter-related bloodstream infections in intensive care units across Michigan. By following simple steps, such as handwashing before catheter insertion, they not only saw a dramatic initial reduction in infections, but after three years they had maintained it. Infections remained 70 percent below the levels before the project.

This paper was one of the first to show that safety improvements can stick, and we’re working to replicate that result in other states and countries. I believe that what has made this project work long-term should apply to other safety efforts. Here are some ideas for what’s behind this success:

Make it easy. Health care workers are always being told what to do by their hospital, accrediting agencies and others. So if you’re thinking about adding to that list, help your staff to comply. For instance, we didn’t just ask caregivers in Michigan to follow basic steps for catheter insertion; we made it easier by introducing a central-line cart that contained all of the equipment required to place a line while maintaining sterility. Proper line placement became part of the routine, not something extraneous to their regular work.

Keep data manageable. Don’t track more data than your group will be able to handle. When we were designing this project, some participants wanted to collect dozens of data elements, such as the type of organisms infecting patients. But that wasn’t feasible in the long term. We decided to collect only what we needed to calculate infection rates—the number of infections divided by total catheter-days.

Monitor and report performance. Feed data back to staff so they can track their progress and take ownership of the effort. In many hospitals, frontline staff still don’t know their units’ infection rates. That data is guarded by infection control departments, while frontline caregivers should own it.

Create accountability. While hospitals have well-developed nursing leadership structures, physician leadership is much less robust. In our program, hospital CEOs assigned a physician leader in each area where catheters were inserted to be responsible for the infections in that area. This created a chain of accountability for the providers placing catheters.   

Change culture. I can’t emphasize it enough: You can’t sustain improvement without changing workplace culture. In Michigan, we fostered mutually respecting relationships so that doctors, nurses, administrators and others can work toward common goals—for instance, by reminding one another to follow infection control measures. Perhaps the largest culture change was convincing clinicians and leaders that they actually could eliminate infections.

Those of us who work in patient safety spend a lot of time asking ourselves if a potential solution will work. But just as often, we need to ask ourselves, How do we make it last? 

Related Content

© The Johns Hopkins University, The Johns Hopkins Hospital, and Johns Hopkins Health System. All rights reserved.