Critical Path Network
Hospitals collaborate to reduce surgical infections
QIO-led program cuts rate 27% in one year
Fifty-six hospitals from 50 states as well as U.S. territories, collaborating to improve surgical care, significantly cut the rate of surgical infections for more than 35,000 patients in a yearlong, nationwide effort sponsored by the Centers for Medicare & Medicare Services (CMS) and led by Qualis Health, the quality improvement organization (QIO) for Washington, Alaska, and Idaho.
The 44 hospitals that provided data throughout the collaborative reduced their surgical-site infection rate by 27%. The results were published June 23 in The American Journal of Surgery.
Conducted in 2002-2003, the National Surgical Infection Prevention Collaborative also involved 43 QIOs working under contract to CMS.
A major cause of preventable morbidity and mortality in hospitals, surgical-site infections complicate an estimated 780,000 operations each year.
Research has shown that compared to similar-risk patients undergoing the same surgery, a patient who gets a surgical-site infection is twice as likely to die, five to six times more likely to require readmission, and likely to stay in the hospital twice as long.
The costs of these complications may range from $30,000 to $50,000 per major surgery.
The collaborative emphasized rapid testing of small changes in the work of surgical teams, then incorporating successful modifications into routine care.
Surgical teams from the national collaborative hospitals joined staff from state-based QIOs at a series of two-day learning sessions with Qualis Health over the course of a year.
Most of the teams came from large, urban hospitals, although some small, rural institutions participated as well.
Between sessions, the teams worked with their local QIOs and communicated frequently with each other to share information about implementing improvements, barriers encountered, and lessons learned.
Three processes to improve
All teams in the collaborative agreed to focus on improving performance on three processes that CMS uses as national quality measures:
1. administration of antibiotics within 60 minutes of surgical incision;
2. use of appropriate antibiotics;
3. discontinuation of antibiotics within 24 hours of the end of surgery.
Most of the teams also worked on improving performance on one or more of the following: control of glucose levels during surgery, avoiding hypothermia during surgery, use of supplemental oxygen during surgery and recovery, and clipping rather than shaving the surgical site.
Over the course of the collaborative, the median performance of participating hospital teams improved on all process measures. The overall infection rate fell more than a quarter, from 2.3% in the first three months of the collaborative to 1.7% in the last three months.
Hospitals participating in the collaborative began with a higher-than-average performance on this measure: a median 70% rate of administering antibiotics within 60 minutes of incision. By the end of the collaborative, median compliance had risen to 92%.
Recent research shows, for example, that patients receive antibiotics in the 60 minutes prior to surgical incision — a key technique for avoiding infections — only a little more than half the time.
Training in the adoption of successful interventions identified in the National Surgical Infection Prevention Collaborative subsequently was conducted over the past three years by QIOs in every state.
Multiple goals sought
There were multiple reasons for doing the collaborative, explains Jonathan Sugarman, MD, CEO of Qualis Health.
"The first was to kick off an effort to prevent surgical infections and to try this [collaborative] method at the national level," he continues. "But another was to provide QIOs the ability to learn how to implement the collaboratives."
This, Sugarman says, was a nontrivial undertaking. "It requires a lot of content knowledge, a lot of structure, involving what works and what does not in these groups," he explains.
"This involved training for QI professionals. There was additional time spent with the QIOs — they had some coached practices and asked questions," Sugarman notes.
"The goal was to develop in each state a hospital that could participate in a statewide collaborative that would be facilitated by them." Almost every hospital that was asked to participate did so, he says.
There are a couple of key requirements for setting up a collaborative, Sugarman continues.
"Since they are not in and of themselves research, you tend to focus on the implementation of evidence-based practices," he adds. "There has to be an actual set of changes that are known to be effective."
"It’s a fairly structured event," adds E. Patchen Dellinger, MD, professor and vice chairman of the department of surgery and chief of the division of general surgery at the University of Washington, Seattle, the lead facility in the cooperative.
"Our national collaborative was preceded by a two- to three-day meeting run by IHI [the Cambridge, MA-based Institute for Healthcare Improvement, credited with pioneering the collaborative model] to get everyone up to speed [on the collaborative process]. Qualis, the functional arm that ran it, was of course already very experienced in this process," he explains.
"At the same time, however, Qualis was teaching the other QIOs how to run collaboratives, with the goal of going back to their regions and running their own," Dellinger adds.
During the training session, there was much talk about how to begin change with small units, and then achieving "spread" — engaging more physicians, more patient groups, and so forth, he says.
"Once you’ve done that, you must hold your gains," Dellinger continues. "In my personal observation, change is difficult, but you can do it; and the IHI model helped us get going. Spread is more difficult, and holding gains is more difficult still."
The model for improvement
The collaborative’s efforts to improve delivery of antibiotics at the appropriate time offer perhaps the clearest example of the "model for improvement," which is based on beginning with tests of small changes. It also was one of the best-documented and best-studied aspects of the evidence on which the initiative was based, he notes.
There have been a number of papers on the subject over the past 10 years, culminating in one two years ago on about 34,000 Medicare patients across the nation, which showed appropriate antibiotic delivery was accomplished in the average hospital less than 60% of the time, Dellinger adds.
"That’s not because docs did not know what to do, or they didn’t give the right order; it happens because it is a very complicated process, and hospitals do not focus on the mechanisms of delivery happening in the same way every time," he points out.
When Dellinger’s collaborative focused on the issue, for example, the teams typically would include a nurse, a QI professional, sometimes a surgeon, and sometimes an anesthesiologist.
"They would diagram the process — what happens between when a surgeon orders the antibiotic and when it’s given," he explains. "They would document what could make this process go wrong. Then we’d talk to Dr. Smith,’ and tell him, Here’s what we think we need to happen to make things right.’"
At that point, Dr. Smith’s patients would be tracked for six weeks or so. If the change did not work, another change would be tried.
"We’d do a couple more tests, and if the process worked, then we’d take all four colorectal surgeons in our hospital," Dellinger continues. "Then we’d do it for every surgeon, and that’s how to achieve spread."
One of the keys strategies for many hospitals was giving responsibility to the anesthesiologist, since he is in a very good position to judge when it is the right time to give the antibiotic.
"This worked for a lot of hospitals but not all of them. Others decided to use the nurses in the OR; this emphasizes that the same solution does not work for every facility," he observes.
Ongoing contact with QIOs
Another key to success was the ongoing contact between the QIOs and participating facilities, Dellinger says. "There were three learning sessions over 12 months, as well as an Outcomes Congress," he reports.
"At each learning session, there were content experts lecturing, basically sharing what the evidence showed. There were also human factors experts — one from IHI and several from Qualis. They talked about how to do the small change tests and so forth."
After that, he notes, the meeting would break into small groups where hospitals would share their experiences. "We shared what worked and what didn’t, so there was a lot of cross-pollination," Dellinger says. "Also, there was an e-mail list to which all participants in the collaborative could write questions, and the content experts could answer and share those answers with the list. There were also conference calls at least once a month."
Anecdotally, he says, "The impression I had, as well as a number of people from Qualis and the teams themselves, was that having an active clinical physician, preferably a surgeon, involved in the team was also very important. They are the captain of the ship, so if a surgeon went back and sold a concept to a colleague, this was very helpful."
States reap the benefits
Following the national collaborative, QIOs and participating facilities went back to their own states to foster more regional efforts. "Qualis subsequently ran one for Washington, Idaho, and Oregon in which we got 30 to 40 hospitals to participate," Dellinger reports.
"The results were quite good, in fact, quite comparable to the national collaborative," adds Sugarman. "In many cases, those organizations active in the national collaborative continued as mentors."
QIOs in more than 30 states report hospitals taking part in the training have gone on to show significant improvement. For example, 26 hospitals participating in California increased the proportion of surgical patients receiving antibiotics within one hour of incision from 73.8% to 84.3%.
In Colorado, 16 hospitals increased the proportion receiving antibiotics within one hour of incision from 62% to 88%; in Maryland, 16 hospitals went from 72% to 91.9%.
Nineteen hospitals in New Mexico went from 47.6% to 68%; and in Texas, 42 hospitals went from 61% to 84%.