Jasper Palmer, a patient transport worker at Albert Einstein Medical Center in Philadelphia, noticed that the gowns and gloves he and other staff wore while moving patients infected with methicillin-resistant Staphylococcus aureus (MRSA) were overwhelming the hospital's trash cans. The piles of discarded attire spilled out of disposal bins onto the floor, contaminating surrounding surfaces.
So Mr. Palmer devised his own method: He took off his gown, rolled it up into the size of a baseball, and pulled his gloves over it to contain it in a tight package.
It might have ended there—except for a nurse who noticed his approach and spread it around the hospital as part of an innovative MRSA-prevention project, which uses a strategy called positive deviance. Today, the “Palmer method” for de-gowning is a well-known technique at Einstein, and Mr. Palmer speaks at regional workshops on MRSA.
Positive deviance is a bottom-up, rather than top-down, approach to solving prevalent, seemingly intractable problems in an institution or a community. It's based on the observation that, in most communities, certain individuals or groups (positive deviants) have found better solutions to these problems than their peers have.
“Positive deviance uses a process of interviews to highlight these people's solutions and spread them throughout the community,” explained Curt Lindberg, chief learning and science officer at the Plexus Institute, a nonprofit organization that uses “complexity science” to tackle problems in health care organizations and other institutions.
“Rather than imposing externally defined best practices, as is common in many quality-improvement initiatives, it generates solutions from within. And because they come from within, these solutions are less likely to be rejected as foreign DNA—too impractical, too difficult, or inappropriate for a particular institution's culture,” said Dr. Lindberg, who has a doctorate in management.
Monique and the late Jerry Sternin, founders of the Positive Deviance Initiative, helped develop the positive deviance technique, which has been used around the world to generate solutions to complex health care problems, including childhood malnutrition, infant mortality and morbidity, poor pregnancy outcomes, and the spread of HIV. Only within the past several years, however, has it begun to get a foothold in U.S. hospitals.
In 2006, six U.S. hospitals participated in a pilot project, funded by the Robert Wood Johnson Foundation and led by Plexus, that used positive deviance to improve MRSA prevention. MRSA seemed like an ideal target for a positive deviance pilot. It's an enormous problem for hospitals, infecting some 95,000 patients each year and killing about 18,000 of them, according to the CDC.
The strategies for combating MRSA—strict hand hygiene, barrier methods on all colonized and infected patients, and active surveillance cultures—are well known, and yet the educational campaigns and best practice programs designed to promote them have largely failed.
“We knew that solutions like handwashing were things that you taught and retaught and remeasured and didn't get anywhere,” said Patricia Norstrand, RN, senior director of quality, risk, safety and infection control at Franklin Square Hospital in Baltimore, one of the pilot institutions. “Infections were on the rise, and the typical things we were doing just weren't working. We could teach from here until the cows come home and we weren't getting anywhere.”
So with guidance from Plexus, Franklin Square and the other pilot hospitals embarked on a series of meetings. These small-group discussions—hundreds of them at each institution—included staff from every level of the hospital, ranging from physicians and leadership to nurses, social workers, chaplains, custodians, unit secretaries, staff, and transport workers. They were asked such questions as
- What steps do you take to prevent transmission of MRSA?
- What prevents you from taking these steps all the time?
- Do you know of anyone who has overcome those barriers?
Staff who identified the positive deviants in their midst would then help spread the newly discovered practices. Formal and informal discussions among diverse groups of employees led to new practices and solutions. Some practices were adopted immediately, and leadership responded quickly when approvals were needed.
Owning the problem—and the solutions
“In one of our dialogues, residents were saying that they didn't wear isolation gowns because they couldn't fit them over their lab coats, which had pockets filled with tons of things. And they weren't taking their coats off to go into the isolation rooms, because we'd gotten rid of the isolation carts and started putting things in the pass-through cabinets so the floors weren't cluttered,” said Ms. Norstrand.
A nurse manager participating in that discussion went to a hardware store, bought hooks, and had them placed in isolation rooms so that the residents would have a place to hang their coats. As a result, gown usage improved from 9,000 gowns a quarter to 12,000 a quarter.
The experience highlights an important difference between positive deviance (PD) and traditional performance improvement efforts, according to Ms. Norstrand. “With performance improvement, I'd tell environmental services to put up hooks everywhere. With PD, the nurse manager told others. First, they didn't do anything about it. Then, they saw it was working on other floors, and took it upon themselves to get it done and marketed it to the residents. If I had put the hooks up, they wouldn't have been out there marketing. It's their process and their solution.”
Between 2006 and 2008, the pilot units at the six hospitals saw their MRSA rates decline by 73% overall. Two pilot units—Franklin Square's intensive care unit and the intensive care unit at the Billings Clinic in Billings, Mont.—reported hospital-acquired MRSA infection rates of zero for at least one quarter. The CDC analyzed data from three of the hospitals (Albert Einstein Medical Center, the Billings Clinic, and the University of Louisville Hospital) and found that MRSA rates dropped by 26% to 62% and that the proportion of S. aureus infections caused by methicillin-resistant bacteria also decreased. The CDC analysis was presented at the Society for Healthcare Epidemiology of America's 2009 annual meeting.
The project's results have now led more than 50 other hospitals across the country—as well as two in Colombia and six in Canada—to adopt positive deviance approaches to reducing MRSA infection rates.
The MRSA project is the largest positive deviance effort in health care in the U.S., but it's not the only one. In 2004, at Waterbury Hospital in Connecticut, nephrologist Anthony Cusano, MD, used the technique to improve medication reconciliation.
“Patients were leaving the hospital with bad instructions on their medication regimens. We all knew it, because we'd see them in the office later,” he said. “I'd designed some forms and we had a whole campaign to make it better, but it never really took hold among practitioners who weren't heavily involved in medications in their own practices.”
A small phone survey of discharged patients revealed the scope of the problem: Two days after discharge, 80% had problems with their medications. “Only 5% had forms filled out clearly defining their medication regimens,” said Dr. Cusano. “Their instructions were in medical lingo, like q6h instead of every six hours. We presented some of the stories, and everyone could see the problem.”
Dr. Cusano then expanded the phone survey by asking individual nurses, residents and attending physicians to make a couple of phone calls apiece to patients and report back. “When they talked to patients who had a good outcome, they'd find out what was done and identify the positively deviant practices that way,” he said. For example, one staffer who made calls came up with an idea for a special medication card that patients could carry with them at all times to write down their medication requirements.
A few months later, Dr. Cusano repeated the survey: This time the rate of medication problems declined to 30%, and the rate of form usage and instruction exchange went from 5% to 95%.
Five years after the original intervention, he repeated the survey again and found that those results had persisted. “I was really shocked. We had changed the culture of the institution,” he said.
Bringing PD to your hospital
So how can your hospital use positive deviance?
First, you have to target the appropriate problem. “PD works when the community thinks there's a significant enough problem to work on, and the nature of the problem has to be social and behavioral,” said Nancy Iversen, RN, director of patient safety and infection control at the Billings Clinic. If the problem is money or technology, positive deviance is probably not going to solve it.
Expertise is also helpful. “Our experience to date has been that hospitals need at least a modest amount of support to do this,” said Mr. Lindberg. “We've seen some instances where individuals in hospitals have tried to pick this up on their own, and they haven't gotten too far, because some of the methods are just so different than conventional approaches to quality improvement and hospital management.”
Even with expert help, hospitals should be prepared for some discomfort during implementation. Positive deviance requires a hospital to shake up its traditional hierarchy. Instead of rigid, agenda-driven, top-down scheduled meetings, the discussions are open to anyone who's interested, and the agenda is set by those who know the most about the problem: usually, front-line staff who have the most contact with patients.
“Most quality improvement projects are driven by middle management people who view themselves as experts,” said Jon Lloyd, MD, senior clinical advisor on MRSA prevention for Plexus. “With PD, those people are the facilitators. They go in as the catalysts with questions, not as the experts with solutions, because they don't know what the functional experts know. And with PD, you don't just identify solutions and have middle management implement them. You provide the front-line staff with the support necessary to act themselves and spread what they're doing around.”
It's an approach that flattens the traditional hierarchy of the hospital, said Ms. Iversen. “It elevates often-unheard voices to levels of prominence and gives power to those who traditionally don't have it—the people who touch patients most of the time. When people stop being asked, they stop speaking up, and they go about their day and say ‘Why bother?’ With PD, our staff felt that they were trusted to solve a very significant problem, and that encouraged them.”
After an infection rate of zero in the first quarter of 2008, the Billings Clinic has had a couple of MRSA infections in the ensuing months, according to Ms. Iversen. “We're not done,” she said. “But when I show the [infection statistics] to the team, they don't throw up their hands—because they own this. They have resolved that they're going to win.”