ICU collaborative achieves major reductions in hospital infections

Statewide program links QIO, hospital association, and hospitals

Hospitals participating in the Maryland Patient Safety Center’s (MPSC) Intensive Care Unit (ICU) Safety Culture Collaborative, a statewide effort to improve safety in intensive care units, are showing major improvements in the reduction of ventilator-associated pneumonias (VAPs) and catheter-related bloodstream infections (CR-BSIs). In fact, five hospitals achieved zero VAP episodes, and 10 hospitals achieved zero episodes of CR-BSIs.

"Since we started participating about 18 months ago, we haven’t had any line sepsis at all, and very few VAPs," reports Rosella Ganoudis, RN, CCRN, supervisor of critical care at Union Hospital of Cecil County, Elkton, MD. "At the very beginning of the program we had two, and maybe a total of four in 18 months, whereas before we had maybe four a month."

The MPSC brings together health care providers to study the causes of unsafe practices and put practical improvements in place to prevent errors. Designated in 2004 by the Maryland Healthcare Commission, the center’s vision is to make Maryland hospitals and nursing homes "the safest in the nation." MPSC is run jointly by the Maryland Hospital Association (MHA) and the Delmarva Foundation, the QIO (quality improvement organization) for Maryland and the District of Columbia. The MPSC’s ICU Safety Collaborative includes ICU teams from 37 hospitals that have implemented best practices to improve care in their intensive care units.

The purpose of the ICU Safety Collaborative, launched in September 2004 and concluded in October 2005, is to bring together multidisciplinary hospital teams and national improvement experts to achieve rapid and dramatic improvements in patients’ lives.

Hospital multidisciplinary teams attend three one-day workshops throughout the course of the collaborative.

Between workshops, teams test changes in their local environment and share results with other participants through e-methods and conference calls.

"The center is 20 months old now, and the reason we won the Eisenberg Award is because of the comprehensive nature of what we are doing," asserts William Minogue, MD, FACP, director of the center. "We’re very fortunate that the MHA and our QIO got together because they have great but different strengths. The MHA has had an education subsidiary for 30 years or more — the Maryland Education Institute, which runs conferences and courses — and Delmarva brings the know-how in collaboratives."

The MPSC actually has three key areas of focus, plus a research area, explains Margaret Toth, MD, chief quality officer for the Delmarva Foundation, which is based in Easton, MD. "One is near-miss reporting; the second is education and training; and the third is the ICU or collaborative arm," she explains. "But the collaborative is where the rubber meets the road — that’s the implementation arm."

Inside the collaborative model

Why is the collaborative approach so effective? "Traditionally, we have often known what the right things to do are, but they are much more difficult to put into practice," says Toth. "That’s not because people do not want to do them, but because the systems or the environment make them hard for one hospital or one quality manager at a time to achieve."

"The QIO has a staff of people who are expert in this," Minogue adds. "Their CEO ran IMPACT for Don Berwick at the IHI [Institute for Healthcare Improvement]. As wonderful as education is, we all go to conventions and hear bright ideas, go back to the same environment and the same culture, and not a hell of a lot happens. Education builds knowledge and a tools base, but a collaborative is a year-long journey."

For example, says Minogue, hospital CEOs must sign an agreement saying that certain of their team members (hospitalists, nurses, and so forth) will participate in the program for a specified number of hours a week, and that their facility will bear a specific data burden. They must cite deliverables such as reducing central line infections, reducing or eliminating VAPs, developing rapid response teams and reducing the incidence of deep-vein thrombosis and peptic ulcers.

"It’s the IHI model entirely," Minogue says. "There are three work sessions scattered over the year, and interaction between participants over a listserv and chats on a secure web portal. The first meeting involves national and local experts, which in our case included Peter Pronovost [a quality expert from Johns Hopkins]. The second meeting was a mixture of more information, but mostly questions about how participants were doing and reports from each hospital. The third meeting, at the end of the year-long journey, involves all hospitals reporting their results and impediments."

The three main workshop meetings "brought everyone together," says Toth. Each team brought physicians, nurses, respiratory therapists, and pharmacists, and they all talked about their practices.

"The times in between were called time action periods,’ during which we would have conference calls where we would talk about what was going on," she adds.

A number of tools were used, including the Agency for Healthcare Research and Quality’s safety culture questionnaire, which all ICUs had to fill out in the beginning and end of the year. "It has great questions," says Minogue. "One is really a zinger: If you or a loved one went into the ICU where you work, would you feel safe?’" He notes that even at prestigious institutions, only 45% initially said they would feel safe. "That’s now up to 75%-80% in units of all kinds," he observes.

Another program deliverable, he continues, is 50%+ improvement. "This takes doing rounds with true multidiscipline teams that include daily goals, for which everyone is in agreement. Around those agreed-upon goals are interventions and process measures that must be done 24/7’ — like elevating patients’ heads at 30 degrees to minimize the likelihood of VAP, or having no more amateurs’ putting in central lines. In the old days, interns did all that kind of stuff in on-the-job training. We also require a completely sterile situation and tight glucose control."

For each condition there is a "bundle" of these procedures to be followed, explains Toth. "The best way to make sure these happen is to pull together multidiscipline groups, round on patients every day, have a checklist and a roadmap, discuss what you want to do that day, and keep checking," she notes. "It turns out that having those bundles and having everyone on the same page really works."

Most of the protocols came either from IHI or Pronovost, she says. "We really didn’t make up anything new; we used procedures that had already been used and proved successful. We pulled that together into a change package, and basically said, If you want to improve, here are five things you need to work on.’"

It’s called a "culture collaborative," she emphasizes, because it not only focuses on what needs to be done and how but on who delivers the care. "The who’ is not administration, but doctors, nurses, dietitians, housekeepers," she emphasizes. "Our methodology brings all these people and their collective wisdom together."

Targeting the ICU

Why was the ICU selected as the collaborative’s first target? "We chose the ICU for a couple of reasons," says Toth. "First, we kept hearing from the hospitals that this was an area where they wanted to begin; we are very committed to being certain that the service and opportunities we provide meet their needs. Second, we had very strong evidence there were some fairly straightforward types of things that could be done in the ICU in terms of keeping patients from getting infected and dying from those infections, and we felt that those interventions, if pulled together in something like a collaborative, could be achieved fairly quickly."

The impact on a single ICU can be dramatic, as Ganoudis notes. "We’ve gotten to the point of knowing what a good data collection system is; before, we were not collecting the right stuff, but the collaborative gave us a way to collect the right data," she explains. "Our first month that we collected data, it said we gave appropriate care in 3% of our cases. We started initiating the program elements one at a time. The next month, we were up to 45%, then we moved into the 80s, and now it’s been in the 90s — our latest was between 97% and 99%."

When things do go wrong, she says, "We take that case to see if there are any common grounds on what happened, what might have happened, what could have been done differently," she says. "It is not done punitively but to learn what we have to do to correct things."

Multidisciplinary rounding has become part of the unit’s culture, says Ganoudis. "We look at things in a whole different light; our daily rounding, which includes pharmacy, nutrition, infection control, and nurses, involves getting a goal or plan for the day. We’ve started assessing the need and the ability to wean on every patient, because we won’t know if they can be weaned unless we do it. We have a pretty extensive oral care policy now, and it’s still being beefed up because the oral cavity is so full of germs."

The ICU nurses, she adds, have really taken the program seriously. "When they saw the first numbers they were flabbergasted," she notes. "Now, for example, every patient gets bathed twice a day." For isolation, they go into a room wearing a gown and gloves, she says. "I think that has made a really big difference."

The most important benefit of the collaborative, says Ganoudis, was being able to see the data. "We felt more of an ownership for the way we took care of patients," she explains. "It empowered staff to ask certain things, because nobody argued with them. In the past, they knew they needed to get this stuff but orders were hard to get. Now, it’s more of a routine; it’s a bundle, and more of a standing order."

Model is replicable

Toth says that the interplay between hospitals and interplay at the hospitals themselves during the multidiscipline rounding were key to the program’s success, along with the introduction of structured tools. "And the hospitals checked twice a day every day to see how they were doing and to make corrections," she adds. In addition, she says, the access to the web portal enabled all hospitals to share their experiences.

"In Maryland, every ICU was working on improving aspects of its care," she notes. "The majority worked with us, but some had been involved in ongoing collaboratives and continued to do so. The idea is that everyone is working on the same thing at the same time, so you have a great opportunity to learn from each other. The teams not only had a better multidiscipline approach in their own hospitals, but a great connection with peers on the outside as well; that’s what makes it work."

Minogue is convinced the Maryland statewide collaborative model is replicable. In fact, he says, "Michigan and New Jersey are getting similar results with Peter’s [Pronovost] help." Michigan, he notes, is concentrating almost exclusively on the ICU.

"Our program is comprehensive — education, data systems, special projects; that’s the key," he says. In fact, he notes, about a dozen states recently brought delegations from their hospital associations and QIOs to Maryland for a site visit. "We’re running a program on how this thing works; all of this is in the public domain," says Minogue.

He goes on to re-emphasize the fact that education alone will never engender permanent change. "Education is for cognitive improvement, but without culture change it’s not going to happen — or if it does, it’s not going to be permanent," he asserts. "In human systems, you tend to get a mediocre state; if you change that state, unless you are constantly applying energy, it will return to that mediocre state. This approach takes that return away — it’s permanent. Hospitals who once wanted a 2% infection rate now say they won’t settle for less than zero. If anything at all happens, they ask, What happened?’ That’s a sea change unlike anything I’ve seen in health care."

For more information, contact:

William F. Minogue MD, FACP, Director, Maryland Patient Safety Center, 6820 Deerpath Road, Elkridge, MD 21075-6234. Phone: (410) 540-9210. Fax: (410) 540-9139. E-mail:

Margaret Toth, MD, Chief Quality Officer, Delmarva Foundation, 9240 Centreville Road, Easton, MD 21601. Phone: (410) 712-7456. E-mail:

Rosella Ganoudis, RN, CCRN, Supervisor of Critical Care, Union Hospital of Cecil County, 106 Bow Street, Elkton, MD 21921. Phone: (410) 398-4000. E-mail: