Policing blood use pays off for Allina Laboratories

Anne Ford

March 2013—A January study of almost 1,000 patients with acute gastrointestinal bleeding has found that restrictive blood transfusion strategies produce better patient outcomes. The study, “Transfusion strategies for acute upper gastrointestinal bleeding,” discovered that patients with severe acute upper GI bleeding who received blood transfusions when their hemoglobin levels fell below 7 g/dL, rather than 9 g/dL, had higher probabilities of survival at six weeks, as well as reduced rates of further bleeding and fewer adverse events (Villanueva C, et al. N Engl J Med. 2013;368:11–21).

“Not cutting edge but new and important,” Lauren Anthony, MD, medical director of Allina Health Medical Laboratories, Minneapolis, says of the research. “They looked at patients who were actively GI bleeding, and that hasn’t been addressed in the literature before.” And it adds to the body of evidence that patients do just as well, if not better, under restrictive rather than liberal transfusion strategies.

Despite this having been known for years now, physicians have been strangely slow to embrace them. “People don’t really get educated about this, because it doesn’t really fall into any clinical disciplines,” Dr. Anthony says.

“It’s a counterintuitive physiologic principle to most of us,” says David Tierney, MD, a hospitalist and the assistant program director for the internal medicine residency program at Abbott Northwestern in Minneapolis, one of the Allina Health System’s 11 hospitals, which are located in Minnesota and Wisconsin. “We think of red blood cells as being good things only and carrying oxygen. And so the concept that adding more of them to someone who has a partially full ‘tank’ is deleterious—that innately doesn’t make sense to the physiology we’re taught as physicians.”

And then there’s the power of tradition. “For a long time, the existing dogma of transfusion was always giving at least two units, and transfusing at a hemoglobin threshold of 10 [g/dL],” says Josh Martini, MD, an anesthesiologist at Allina Health’s Mercy Hospital in Minneapolis and chair of the hospital’s transfusion committee. “I was taught that as a medical student back in 1996 on my first medicine rotation. It has nothing to do with any evidence or any scientific backing. I was a resident when the original TRICC [Transfusion Requirements in Critical Care] trial came out. I remember I was on the orthopedic surgery service, and I felt like I made headway if I could get the orthopedic surgeon I worked with to transfuse at 9.5 instead of 10. That was my big goal.”

Against the twin foes of intuition and tradition, how can laboratories go about getting physicians to embrace restrictive transfusion strategies? A few physicians become convinced on their own after undergoing personal experiences like that of Charles Terzian, MD, a hospitalist at Allina Health’s United Hospital in St. Paul, Minn. As a trauma patient at a (non-Allina) hospital, Dr. Terzian received a blood transfusion in 1995. Two years later, “I got letters saying, ‘We might have transfused you bad blood,’” he recalls. “Luckily, all my serological testing was negative. But just going through that experience was enough to reinforce my position that blood transfusions can save lives, granted, but they can also cause problems.”

Others change from mere supporters of restrictive blood transfusion strategies to active proponents after seeing the dire effects an unnecessary transfusion can have on a patient. Dr. Martini recalls a patient in her 50s who came in with a GI bleed: “She wasn’t able to receive any blood products, because she had multiple antibodies and we couldn’t find blood for her. She ended up dying. When I looked through her records, it was evident that the only exposure she had to any blood products was a total knee arthroplasty a few years before. The most likely explanation for her development of antibodies was from this prior transfusion. When I looked at the records, the transfusion was inappropriate. I was limited in saving this woman because I couldn’t get appropriate blood products. You think: ‘One person’s decision to not follow the evidence contributed to this.’”

As powerful as experiences like these can be, it takes more than a few isolated proponents to affect how and under what conditions blood is transfused, especially in a system as large as Allina Health. In the last two years, Dr. Anthony has led a massive, systematic effort to get physicians in the 10 Allina Health hospitals where blood transfusions are performed to understand and embrace appropriately restrictive transfusion protocols. And it’s worked: Red blood cell use has dropped significantly. So significantly, in fact, Dr. Anthony calls Allina’s continuing efforts in this area “a game of inches.”

Before Allina Health began to consider systemwide transfusion education in 2010, “we certainly weren’t doing any worse [in this regard] than any major hospital system, but we had a large area for improvement,” Dr. Martini says. One of his colleagues, Dirck Rilla, CCP, remembers going into a meeting with surgeons a couple of years previously and attempting to give a process improvement talk that focused on blood conservation. “I was happily escorted out,” says Rilla, director of perfusion operations for the entire Allina Health System.

That all began to change when the system’s chief clinical officer, Penny Ann Wheeler, MD, decided to champion a transfusion education effort at Allina after hearing about the issue at an AHA meeting. And the then newly hired Dr. Anthony, who had recently left a hospital with a successful blood management program (Bronson Methodist, Kalamazoo, Mich.), was the right person to lead it. The first step: To get a clear picture of Allina hospitals’ blood use. It wasn’t pretty.

Utilization review by an external consultant through the Red Cross showed “we had overtransfusion of about 25 to 40 percent, depending on DRG code and patient mix,” Dr. Anthony says. And blood use varied dramatically by hospital. Rilla says, “In the cardiovascular arena, one hospital had an RBC transfusion rate of about 67 percent, and across town it was closer to just under 30.”

The second step: To assess physician readiness. “The lab can serve a leadership role, but it can’t implement it without physician champions who order blood,” Dr. Anthony says. “I mean, the lab just can’t do it, because the lab doesn’t order the blood. So I sent out to all the medical staff a physician readiness survey. It was basically: ‘Do you observe a lot of variation in transfusion practices when you’re on the floors? Do you think there’s a need for more transfusion education? What is the typical hemoglobin level you use to decide to give a transfusion?’ Questions like that. The responses showed us that most people definitely agreed that they were hearing about it in their professional societies and knew it was something there was an opportunity for.” The survey also asked whether respondents would be willing to become physician champions for restrictive transfusion strategies across Allina Health’s hospitals.

Many said yes. Even better, among those who said yes were several physicians in high-blood-use specialties. Says Dr. Tierney: “It’s hard for me as an internist to convince a cardiovascular surgeon that he shouldn’t give a postoperative heart patient a unit of blood when his hemoglobin’s 7.5. But for a cardiovascular surgeon to convince his partner is a much more effective practice.”

Once the physician champions had been identified, they gave Dr. Anthony surprising feedback. In their view, the transfusion effort should be rolled out systemwide, rather than piloted at one hospital and introduced only gradually at the others. “Some of the high-level champions, like a cardiovascular surgeon and a liver surgeon, said, ‘This is only going to work if you do it at the system level, because we have an EMR,’” she says. “They said, ‘You can’t be making changes to order sets and transfusion guidelines if you don’t include all the hospitals as you move forward; it’s not going to work.’ And it’s because of them that this has worked.”

Subsequently, Dr. Anthony visited each hospital’s medical leadership committees to discuss what a blood management program would entail, why it should be done, and how it would get started—“just starting the discussion, getting people used to the idea,” she says. In a further effort to “get a lot of people engaged and enthusiastic,” she invited blood management pioneer Aryeh Shander, MD, chief of anesthesiology, critical care medicine, and hyperbaric medicine at Englewood Hospital and Medical Center, NJ, to deliver to Allina medical staff talks about plasma and red blood cell transfusions.

As a newcomer to Allina Health, Dr. Anthony had to learn her way around hospital performance improvement in a large integrated system. Her guide: clinical nurse specialist Jodi Hartwig, who showed her how to structure the project and plug it into Allina Health’s performance improvement model. “Jodi explained how initiatives are rolled out in the system and said I needed to configure blood management the same way.”

With the help of a consultant, Dr. Anthony and her colleagues identified five areas of focus: revision of transfusion guidelines and order entry; cardiovascular surgery blood conservation; minimization of iatrogenic blood loss; preoperative anemia management; and awareness/education. “At that point, we knew what we needed to do, and we formed a systemwide transfusion care council that has all 10 hospitals represented,” she says. “We made sure everyone had the opportunity to participate in work groups, and through the care council, we brought the work group decisions back to the main group, and they were communicated to all 10 hospitals.”

For instance, they moved the hemoglobin threshold in their guidelines from eight to seven for stable, hospitalized patients. “When we implemented it, everyone was fully aware and had signed off on it, so it ended up being very successful. We didn’t have these small hospitals saying, ‘You changed something on us, and we had no input.’”

It helped that many of the working groups had people from more than one hospital. “Because we’re such a large system, Allina has excellent audio- and videoconferencing systems,” Dr. Anthony says. “We’re always sharing over the Web in live meetings. It’s such a part of our culture.”

As for the smaller hospitals, “we had a couple that didn’t want to have any part of blood management at first,” she adds. “They didn’t have time, or they didn’t think it was a big deal. We brought this up to Dr. Wheeler and said, ‘Look, we must have physician champions at all the hospitals. We can’t have people opting out.’ So she made sure we had all 10 hospitals represented, whether they wanted to be there or not. Most did, but there were a couple latecomers.”

By July 2012, the care council had succeeded in revising and launching the transfusion electronic order set to make it more difficult for physicians to order two units of blood at a time, since “single-unit transfusions are best in non-bleeding patients,” Dr. Anthony says.

That revised order set, says pathologist Brenda Katz, MD, medical director of Allina Health’s central laboratory, is “much more instructive to clinicians in terms of things like demanding a reason why they’re transfusing. And now one unit comes up automatically, instead of two. It used to be that they could choose whatever they wanted.” Hospitalist Dr. Tierney, who served as Abbott Northwestern’s physician lead for the order-set modifications, agrees that the order-set changes have been a powerful tool in altering physician behavior, in part because “we’ve tried to incorporate a lot of evidence into the ordering process,” he says. “That’s ultimately what helps physician practice change. Without the evidence, people just get pissed off because they get frustrated not being able to order something in the way they always have.”