Published in the April 2007 issue of Today’s Hospitalist
In the world of quality improvement, two truisms are often cited to explain challenges that hospitals face when improving patient safety. The first is that organizing physicians is like herding cats. The second is that if you’ve seen one successful quality improvement (QI) program, you’ve seen one successful QI program, meaning that initiatives often can’t be reliably replicated.
Now, however, a new study has dispelled both of those notions by showing how a group of hospitals in Michigan eliminated catheter-based bloodstream infections. The project not only identified successful techniques to convince physicians and nurses to change their approach to preventing infections, but also duplicated that success in the ICUs of 100 hospitals.
In the study, which was published in the Dec. 28, 2006, issue of New England Journal of Medicine, researchers implemented a number of common-sense interventions to reduce central-line infections.
While the study’s strategies were fairly typical, their results were anything but. Over the 18 months of the study, the mean infection rate plummeted from 7.7 to 1.4 per 1,000 catheter days.
A big factor in researchers’ success was their ability to convince physicians and nurses to adhere to interventions that can dramatically “and fairly easily “make a dent in central-line infections. Chief among those strategies was the use of daily goals to improve communication among clinicians; a "no-exceptions" approach to adhering to evidence-based procedures; and, perhaps most importantly, a willingness to change medical culture when patient safety was at stake.
A new approach to infections
Most of the interventions used in the study to prevent catheter-related bloodstream infections are relatively common. Clinicians were urged to routinely wash and clean their hands with chlorhexidine, use full-barrier precautions when inserting catheters, and remove unneeded catheters in a timely manner.
Study participants also used standardized single-use central-line supply carts and avoided femoral artery catheter placement, which is associated with higher infection rates than subclavian vein insertion. And clinicians were required to use a checklist to document how well they were adhering to the safety practices implemented as part of the research.
None of these interventions is particularly new, so what made them so successful? Peter Pronovost, MD, PhD, the study’s principal investigator and an anesthesiologist and intensivist at Johns Hopkins Hospital in Baltimore, explains that the key was changing the culture regarding infections.
To help support that shift, researchers created physician-nurse leadership teams at each hospital. The teams not only provided educational support on safety improvement but worked with infection-control professionals to ensure ongoing surveillance.
To increase adherence to the study interventions, for example, clinicians who failed to comply with any of the prescribed safety practices in non-emergency situations were literally "stopped in the act" by team members. That held true whether the offending party was a nurse or a physician.
"In many places," Dr. Pronovost explains, "the nurses said, ‘If I tell Dr. Smith to stop and wash his hands, he’ll bite my head off.’ " To overcome that hurdle, researchers framed the team intervention not as a power or political struggle, but as a patient issue. "We got agreement by having everyone agree that preventable harm is untenable and that all patients deserve the benefit of evidence-based care."
The importance of culture change
Efforts to change hospital culture are so critical, says Dr. Pronovost, because physicians tend to dispute or dismiss the results of quality improvement.
In some instances, that skepticism stems from the fact that the scientific underpinnings of many quality improvement initiatives are weak. In other cases, physicians are wary of initiatives that seem more geared to "marketing" quality improvement than to improving the quality of care.
Even worse, he adds, health professionals frequently take the position that mistakes and practices that cause infections "are the cost of doing business," a mindset that prevents broad-scale safety improvements in the hospital.
"The huge culture change was getting clinicians to actually say that these mistakes are mostly preventable vs. inevitable," Dr. Pronovost says. The next step was to move them away from the mindset that central-line infections are to be expected in the very ill patients that populate ICUs.
Another mindset that needed to be changed: Because infection rates and their monitoring have historically been the purview of hospital infection-control staff and risk managers, clinicians often feel removed from infections that do occur.
Instead, participating in the study helped convince physicians and nurses that they, not infection-control personnel, "own" infection rates.
"The social support system was important," Dr. Pronovost explains. "We found that if you don’t get the culture change, the rest of this stuff doesn’t stick. This can’t come from your chief medical officer pushing you. There’s too much friction in the hospital, and there aren’t enough resources."
Sustained results
Research has shown that U.S. hospitals employ the simple safety practices used in the study only about 30% of the time, notes Dr. Pronovost, who is medical director of Johns Hopkins Hospital’s Center for Innovations in Quality Patient Care and a longtime leader in patient safety improvement.
Given that poor performance, Dr. Pronovost notes what is perhaps the most impressive result of the project: Central-line infections dropped significantly in all participating hospitals, regardless of the facilities’ type, size or staffing levels. Equally impressive, the reduction in infection rates was sustained over
the 18 months of the study.
Dr. Pronovost hypothesizes that the project’s impact was so long-lasting in part because of the study’s design and because of the support the participating hospitals received from federal grants.
The pilot study was funded primarily by the Agency for Healthcare Research and Quality and carried out with assistance from the Michigan Health and Hospital Association in partnership with Johns Hopkins researchers, who provided the technical component. Researchers from the University of Michigan in Ann Arbor co-authored the study.
The rapid, sustained decrease in infection rates shown by the study also sends an important message to naysayers, Dr. Pronovost adds: If a central support mechanism is in place, large-scale, multi-site safety improvement projects are doable.
"It’s that partnership “with the academic center serving up the technical piece and partnering with the hospital association and large group of providers to implement it and make the culture change “that is really magical," Dr. Pronovost says. "Clearly, to do this well far exceeds the resources that any single place, including Johns Hopkins, has available."
Think regionally, act locally
That raises the issue: With patient safety such a hot-button issue, why aren’t more statewide initiatives taking place, with buy-in from large stakeholders and adequate funding and support?
According to Dr. Pronovost, the same groups that collaborated on the Michigan ICU project are now working with the World Health Organization to implement similar initiatives in several countries. They are also exploring the possibility of launching similar projects in other states.
For hospitalists in individual hospitals, "There is great value in creating a social support system to help with this work," he says. Physicians can partner with their state hospital association, other hospitals in their system or other hospitals in their area, he suggests, "to divide up the work and learn from each other."
For initiatives that target only one hospital, Dr. Pronovost urges hospitalists to consider a narrower focus on reducing infection rates on the medical floors, for instance, or complying with core Joint Commission measures.
But even tailored initiatives won’t gain traction, he warns, unless cultural issues are tackled along with clinical ones. And practically speaking, Dr. Pronovost adds, hospitalists who are serious about working in quality improvement should receive some training in public health.
"That’s absolutely critical," he says, "because you need the skills of a public health degree to lead the technical pieces."
Bonnie Darves is a freelance writer specializing in health care. She is based in Lake Oswego, Ore.
