Unit-based teams get results at Penn

Teams helped health system reach multiple goals

It's kind of like that old ad for Reese's Peanut Butter Cups: Peanut butter is great, chocolate is great, but imagine what can happen if they get mixed together. That's what happened when the chief nursing officer and chief medical officer of the University of Pennsylvania Health System got together to work on quality improvement projects. "We had an old quality model that used champions and experts. Sometimes they worked with physicians, sometimes with nurses," says Jeffrey I. Rohrbach, MSN, project manager for quality and safety at the University of Pennsylvania Health System. "But what if they worked together?"

Creating unit-based leadership teams consisting of a physician, a nurse, and a quality leader led to changes in the way every subsequent quality project was done and spurred projects that have had a far-reaching positive effect on patient care. In the old system, Rohrbach says, working to reduce bloodstream infections might be an imperative for leadership, but it was not even known at the bedside. "Trickle down did not work," he says. But by putting the onus on individual units, there is only so far the word has to go before it reaches the bedside, and the route that each triad takes to achieve the goal can take into account the particular needs of a unit: Cutting readmissions for oncology patients might require different actions from cutting readmissions for the transplant unit. "What works on one might not even be applicable on another, so looking for a global solution to a local problem doesn't work."

Under the new system, not only can those local problems have local solutions, but Rohrbach notes that the people working at the bedside are more comfortable sharing their own ideas with people they know well and who understand the particulars of the work environment. Team members exchange ideas in both formal and informal settings, sharing information both within and across units. Often, units work together on projects, as well, he says.

Each team consists of a physician leader, a nurse leader, and one of four quality/safety project managers. As the program has evolved, other disciplines have asked to participate — pharmacists are often included, as are other clinical specialists. Piloted in just five units, there are now 18 teams responsible for 22 units, including five intensive care units and a women's health service. One team covers two general medicine units, and another covers three oncology units. The surgical team has a slightly different make up — two physician co-leaders. No member of the team outranks any other, says Rohrbach, or has a voice that carries more weight than his or her counterparts.

The roles are different, though, with the physician responsible for communicating about the project to other physicians; the nurse manager in charge of implementing projects and communicating with frontline staff; and the leader gathering data, creating action plans, and project management. Because the quality leaders work across units, Rohrbach says they can often suggest that units work together on particular issues.

The teams meet weekly to look at trend reports and determine if anything needs particular attention. There is also informal communication between team members to discuss ongoing projects or issues that arise between meetings. Every month, groups of unit teams meet with the chief medical officer and chief nursing executive to present information on existing projects or ask for assistance.

There is also communication across the various hospitals in the health system with teams from one facility often hosting teams from another to talk about their efforts and accomplishments. Further cross-pollination happens through the chief nursing executive and chief medical officer councils involving senior clinical leadership from around the health system.

The program hasn't been without hiccups, Rohrbach notes. "We had a lot of pull from the nurses, who were very interested in it, but still have to do some pushing with the physicians. Some were interested and have become champions." Others, however, were coerced to participate through use of a stipend paid for participation on the teams. "We are buying their time and their work, and they are held accountable for that. They have to create baseline goals, ways to measure their success, and a list of projects to work on."

What has provided more pull than the stipends has been the astounding success the triads have achieved. When they started with five units in 2007, Penn Health concentrated on rounding with teams and central line-associated bloodstream infections (CLABSI). "We chose to work with units who had the strongest physician buy-in and the units with the highest infection rates," he says. Before the pilot had run its course, news of its success was spreading and other units were asking to participate. Now in its fifth year, every inpatient unit at the three-system hospitals has the leadership groups. Next up is expanding it to the emergency department and the cardiac catheterization lab. In the future, other system parts such as skilled nursing may be included, Rohrbach says.

If I had it to do over…

One thing that he wishes he'd done differently was to make sure the physician role is structured. "They were given a stipend; they were told to participate. But how to do that was not laid out," says Rohrbach. "So there were varying degrees of participation. Some teams met weekly. Some did not. Some would say, 'Oh, there's nothing to discuss this week,' and skip a meeting."

And while some physicians did not do much more than attend meetings, others took on additional work, went to faculty meetings to talk about initiatives and generally proved themselves superstars. "Now we have the work plans, the goals they have to submit, and quarterly meetings with the associate chief medical officer. If they do not meet their goals, they are dropped, and we start looking for physicians who are more interested in what we do."

They also created a rotating weekly agenda to help keep all the teams on track. Every week there is something specific to talk about. In one group, Rohrbach explains, it might be patient satisfaction and incident reports in week one. A quality champion might bring data to go over on week two. The following week might be a look at all the infections that there had been in the last month. "You have to create something to talk about."

Lastly, he counsels against limiting numbers. While the troika of nurse/physician/quality manager works, it doesn't include others that may have important input like pharmacists, discharge planners, or social workers. "We made it exclusive initially, which limited effectiveness, and it excluded people who were interested and wanted to help."

The results have been impressive. "When we started, our chief medical officer wanted to eliminate CLABSI and everyone laughed out loud." Now, a bloodstream infection is such a rarity that they count cases, not rates, and every single one merits a close look. One unit has been infection-free since January 2008, Rohrbach notes.

Indeed, the first indication that the unit-based leadership model was worth expanding came through a look at CLABSI rates. All units lowered their rates thanks to a facilitywide effort. But the units with the leadership teams? They saw even fewer infections. The total decrease in CLABSI rates since the inception of the program is from 6.75% in 2007, to 0.5% in 2009. Days since the last infection ranged between 300 and 1,210 days as of May 31, 2011; four units have gone more than 1,000 days without an infection.

Other benefits of the program include a decline in pressure ulcers of more than 20% in medical/surgical units and nearly 16% in critical care units between 2008 and 2009. The overall rate as of mid-2011 has fallen 40% since the unit teams started.

Urinary tract infections have declined by 30%, with some units going longer than a year since the last infection; ventilator-associated pneumonia has been absent for between 95 and 650 days on five units; medication reconciliation has improved from 55.3% to 79.6% of patients having a reconciliation within 24 hours of admission; errors and near misses are reported more often; patient satisfaction has improved; and providers think there is better teamwork and communication.

The work continues, Rohrbach says, with the latest project being attention to 30-day unplanned readmissions. Every patient who is discharged is checked to see if he or she was in the hospital within the previous 30 days. If so, Rohrbach says, "we do a mini-root cause analysis to drill down and look at why."

Other facilities are taking notice. Recently a Penn team flew to the University of Nebraska to talk about what they created. Word is that they will soon implement something similar. The system is also spreading the word locally to hospitals that come calling to see what they did. And they continue to present their findings at a variety of forums.

For more information on this topic contact Jeffrey I. Rohrbach, MSN, Project Manager for Quality and Safety, University of Pennsylvania Health System. Telephone: (267) 303-3225.