In This Section      

Dome - Striving for consensus on quality

February 2012

Striving for consensus on quality

Date: February 9, 2012

Across Johns Hopkins Medicine, the Armstrong Institute fosters community-based fixes for stubborn patient safety problems.

Peter Pronovost
1 2 3
Armstrong Institute Director Peter Pronovost uses the mathematical model of a fractal, which features endlessly repeating patterns, to describe the potential power of clinical communities.

Seated around a conference table at Howard County General Hospital, 20 intensive-care clinicians and managers from across Johns Hopkins Medicine sip coffee, nibble grapes and ponder their first collective strike against several of the most pernicious and costly health care-acquired infections.

But first, “We need to consider what the critical quality elements are,” says Brad Winters, an intensivist and neuro-anesthesiologist at The Johns Hopkins Hospital.

Whether they target ventilator-associated pneumonia, catheter-associated urinary tract infections, or another condition, the clinicians can’t move forward until they agree upon a common goal. That’s a formidable challenge for practitioners hailing from five hospitals, with, at times, different sets of metrics for identifying, tracking and preventing health care-acquired infections, or HAIs.

It’s also a valuable opportunity for the ICU providers who have come together as a
“clinical community,” a model for improving patient safety that builds upon the shared wisdom of its members. Organized by discipline or medical setting, the ICU group is one of four fledgling clinical communities chartered by the Armstrong Institute for Patient Safety and Quality. The other groups revolve around hospitalist medicine (the practice of caring for acutely ill patients in hospital settings), the improvement of medication safety and the post-anesthesia care units (PACUs).

A thriving network of clinical communities is a central component of the Armstrong Institute’s strategy for advancing patient safety and excellent care, says Director Peter Pronovost. “Clinical communities are a way to tap into clinicians’ innate interest in improving quality,” he says.

As it takes shape, the clinical community concept will create an officially sanctioned, system-wide forum for “identifying and solving problems together, to learn together and share, and to ensure that patient safety measures are devised by clinicians, not imposed upon them,” Pronovost says. “We learned that locally developed interventions are wiser, more likely to be effective and more likely to be implemented compared to top down interventions.”

At every level of care at each affiliate, the clinical community model can be replicated with the Armstrong Institute’s central support, adds Pronovost. He likens the model to a fractal, a mathematical model that features endlessly repeating patterns. Correspondingly, clinical communities are designed to devise safety and quality improvements that are replicated at every level of an organization to create a “unified whole” care delivery system.

Pronovost seeks to use this organizational model to promote institutional autonomy in the face of mandates from federal regulatory groups. “By coming together as a community to determine what we want to be held accountable for and then backing up our convictions with sound science, we can find solutions that are more specific to Johns Hopkins Medicine.”

Before moving forward, however, a clinical community must coalesce. As recently as four months ago, many of the ICU personnel present at the Howard County General Hospital meeting were strangers to one another. I-Fong Sun Lehman, the Armstrong Institute’s administrator for clinical communities, guided the evening’s lively and at times testy conversation among medical directors, nurse managers, nurse practitioners, respiratory therapists and others. “The sense of community takes root just by getting them to know each other and become familiar with one another’s workplace culture and language,” Lehman says.

Her role provides a critical pillar of support to the communities. Just as any organism requires a source of energy to grow and multiply, clinical communities require a vertical core of support in order to develop, reproduce and sustain interventions across the multiple levels of a health care organization, Lehman says.

The vertical core also supplies the backbone for fruitful collaborations, Pronovost says. “Human factor engineers, psychologists, sociologists, biostatisticians, epidemiologists and clinicians may all see the world a little differently, but together they can assemble a rich picture that allows us to find interventions that are likely to work.”

A shared mission

As a pioneering researcher in the field of clinical communities, Mary Dixon-Woods, a medical sociologist at the University of Leicester in England, found that one size does not fit all when it comes to effective patient care. According to her research, and that of others, patient outcomes varied from hospital to hospital across Great Britain in spite of standard protocols and procedures. Woods theorized that the best strategies for quality care are cultivated on a local level through discussion, debate and ultimately consensus among peers motivated by a collective concern for patients’ welfare.

In a review of the Michigan Keystone Project, a successful initiative spearheaded by Pronovost and Chris Goeschel to reduce central line-associated bloodstream infections in ICUs across Michigan, Woods recognized the same strategies. At the time unfamiliar with the clinical community model, Pronovost and Goeschel, Armstrong director of strategic development and research initiatives, instinctively gravitated toward a peer-to-peer approach to infection prevention.

“In Michigan, we made progress because we used these communities to overcome potential turf battles and resistance among participants,” Pronovost notes. “Now we’re leveraging that wisdom across the JHM clinical system.”

Comprehensive Unit-based Safety Program (CUSP) initiatives and other patient safety and quality improvement projects already abound in the health system, and it is not the Armstrong Institute’s intention to replace such efforts, Lehman says. “We recognize local improvement efforts. We hope that each clinical group will openly communicate and consider collaborative opportunities with other safety efforts. Community members could leverage one unit’s successful project and pilot that across the health system.” 

 “I think we can all probably learn something from each other. It would be good to exchange ideas about what’s working for us, what kind of problems we’ve faced and how we did or didn’t succeed,” says clinical community participant Roy Brower, MICU director at Hopkins Hospital. During the discussion at Howard County General, though, Brower had questioned whether the clinical community concept may only reinforce the “top-down” approach that it’s meant to replace. “The way our hospitals are put together, I think the solutions are going to be different from one place to the next,” he said. 

Pointing to the numerous interventions, including HAI-prevention strategies, developed by his unit’s clinical practice committee, Brower says, “I think the Armstrong Institute should explicitly state that the best performance improvement initiatives are generated locally, and it should promote that idea and should provide the resources to help local groups to accomplish that.”

The science of safety

The clinical community concept has found additional acceptance as its fertile research possibilities become apparent. “Five years ago, doctors would never have embraced the idea of clinical communities,” Pronovost says. Now, the faculty is beginning to recognize their potential for research and publication in the nascent field of patient safety and quality, he says. “They see it as a legitimate science and as a path to success.”

To foster discipline and progress within each clinical community, the Armstrong Institute has formalized its expectations in a set of ground rules: Aspiring communities must draft a charter with goals, membership information, concrete plans for publishing findings or other accountability benchmarks to be approved by the Armstrong Institute. Also required are presentations before the new enterprise-wide Patient Safety and Quality Board.   

In a year, Pronovost foresees that Hopkins Medicine’s first four clinical communities will have improved patient outcomes, yielded impressive scholarly work, and reduced medical costs. He also envisions the launch of numerous new clinical communities devoted to solving problems specific to certain diseases or broader priorities, such as nurturing the practice of patient- and family- centered care.

In time, clinical communities also will elevate patient care by nurturing stronger relationships across the institution, Lehman says. As clinical communities advance patient safety and care, they will also enhance an initiative perceived by the newly reorganized enterprise, called Johns Hopkins Medicine 3.0, to unify the institution’s culture of patient care, proponents say. For instance, by developing a universal metric to define and track the rate of a particular HAI, the ICU group would also contribute to a more standardized care delivery system throughout the organization.

As the clinical community model drives institutional integration, it will also spark a new level of innovation, unfettered by academic silos, Pronovost says. “In this field, and in all of medicine, the breakthroughs are going to come from the intersections of disciplines, not from within them.”

—Stephanie Shapiro