December 2010
Feature Story

Anne Paxton

In the debate over whether it is better to transfuse more blood to cardiac surgery patients, or less, most transfusion medicine experts believe the jury is no longer out.

Substantial evidence has already been amassed that there are narrow benefits to red blood cell transfusions in patients without hemorrhage, and that reducing or avoiding transfusions in cardiac patients is associated with improved outcomes. The Society of Thoracic Surgeons (STS) and the Society of Cardiovascular Anesthesiologists in 2007 published clinical practice guidelines on perioperative blood transfusion, emphasizing blood conservation.

So the two studies of coronary artery bypass surgery patients published Oct. 13 in the Journal of the American Medical Association—a randomized, controlled trial conducted in Brazil and an observational study of 102,470 patients in the U.S.—are not really trailblazing. The Brazilian study found comparable patient outcomes whether conservative or liberal transfusion strategies were used; the U.S. study found no significant difference in mortality rates among hospitals despite highly variable transfusion rates. But the studies are fueling increased concern—and even impatience—with the lack of progress in reining in unnecessary transfusions. (See “New transfusion studies in brief”)

“What’s novel about the U.S. study is the lack of novelty,” says anesthesiologist Timothy Hannon, MD, MBA, president and founder of Strategic Blood Management, a consulting firm in Indianapolis. There’s a bit of new ground in the Brazilian study because it is a randomized trial, he says. “But it’s just another patient population demonstrating that there is no benefit of a liberal strategy of blood transfusing in patients having cardiac surgery.”

“Simply putting more articles out there is not changing clinical practice. So what is the missing ingredient?”

That the Society of Thoracic Surgeons has adopted clinical practice guidelines is unique, Dr. Hannon points out. “No other component society has ever done that. But here we are, and there hasn’t been any change in variability of transfusion practices since 1991. In spite of the fact that more and more evidence is out there, and that three years ago standards were put out, over the course of 20 years nothing has changed. And in fact, the numbers are even a little worse today than before.”

The Brazilian and U.S. studies are inverse sides of the same coin, says Lawrence T. Goodnough, MD, director of the transfusion service at Stanford University Medical Center and co-author, with Aryeh S. Shander, MD, of an editorial on the two new studies (JAMA. 2010;304:1610–1611). The Brazilian study indicated a difference in average hemoglobin concentration of about 1.5 g/dL between the liberal-transfusion-strategy group and the restrictive-strategy group. But a large percentage of all the patients received transfusions, he points out—78 percent in the “liberal” arm of the study and 47 percent in the “conservative” arm.

While the new U.S. study is the first in cardiothoracic surgery, it reconfirms or echoes what was demonstrated as far back as 1991 in a group of 12 academic medical centers, says Dr. Goodnough, who co-authored the 1991 article (Goodnough LT, et al. JAMA. 1991;265:86–90). “Taking a straightforward, homogeneous group of patients—in our case, first-time bypass patients who are so-called ‘good risk patients’—we demonstrated there was substantial variability in transfusion outcomes, that is, transfusion practices and blood conservation intervention with respect to red cell platelets.”

A similar finding was made in 1999 with the Stover cardiac surgery database, again demonstrating striking variability (Stover EP, et al. Anesthesiology. 1998;88:327–333). Another noteworthy article was Paul Hebert’s on critical care patients, published in the New England Journal of Medicine in 1999, Dr. Goodnough says (Hebert PC, et al. N Engl J Med. 1999;340:409–417). “That study found you can allow people’s hemoglobin concentrations to drift down to 7 to 9 g/dL, an average of 8.2 g/dL compared to a 10 to 12 range and average of 10.5.” But despite the array of research, “now we’re left in 2010 again with a persistent variability, even in the face of the 2007 guidelines from STS.”

The trials done so far do not compare transfusion to no transfusion, he adds. “They’re comparing transfusing more liberally to transfusing more conservatively. If 80 percent of patients receive blood on the one hand, while on the other hand only 50 percent receive blood, yet everyone seems to do about the same, then what do we know about the evidence for triggers? What the Brazil study does in the cardiac surgery population is reinforce the idea that you can really accept a more conservative transfusion trigger, because the patients had the same outcomes as the more aggressively transfused patients.”

The guidelines issued three years ago by the Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists are somewhat complicated. “The range of transfusion triggers is from 6 to 8 g/dL, so before you fix the lesion the guideline is 8 g/dL, during the procedure on the pump it’s 6 g/dL, and then post-op it’s 7 g/dL.” Lack of progress in complying with the specialty groups’ guidelines cannot be blamed on clin­ician lack of awareness, Dr. Goodnough says. “We did reference a study where close to 90 percent of treating physicians in the field were aware of the guidelines. So it looks like they are pretty well disseminated.”

Resistance to the guidelines is a more likely cause, he thinks, because many physicians go right ahead and transfuse blood, as they were taught. “They don’t buy into them. They’re aware, but there’s not necessarily full consensus in the field, even by the nonprofit organizations.”

“The guidelines were really designed to lead the field, as opposed to reflecting the status quo,” Dr. Goodnough says. “They may just be a step too far out in front of treating physicians, and that’s one reason why people don’t seem to be paying attention to them.”

At one time, cardiac surgery was known for bolder moves toward blood conservation. Dr. Goodnough points out that because of the large amounts of blood required and the volumes of surgery performed, and because of the concerns about blood risk and blood inventory, early in the 1970s cardiothoracic surgery led the field of blood conservation on the score of intra- and post-op salvage and reinfusion of autologous blood.

Efforts to improve on that record, though, have stalled in part because in the U.S., cardiothoracic surgery has never been regarded as elective surgery. “If you have a newly diagnosed cardiac lesion, the patient is rarely discharged and scheduled for surgery some weeks later, as in Europe. So there’s never the opportunity to have a conversation with patients about alternatives such as autologous blood donations.”

By contrast, with orthopedic surgery procedures such as joint replacement, patients have more say about their blood transfusions, and transfusion practices and blood conservation strategies have made tremendous progress in ways that got no traction at all in cardiothoracic surgery.

The inertia of entrenched institutional practice can also be a factor. As part of an effort to move its institutional practices forward, Stanford has just adopted a physician education tool on its physician order-entry system as a demonstration project. “It’s a pop-up that reminds the physician of our institutionwide recommendation that a hemoglobin of 7 g/dL be used as the transfusion trigger for patients in general, and for CV surgery or medical patients with ACS, the trigger would be 8 g/dL. It’s not a policy, but an educational intervention to remind physicians of the recommended transfusion triggers.” Stanford plans to follow up with quality data to see if the reminder makes a difference, he says. “And I think we need more of that kind of project.”

In any hospital, there can be tension between younger physicians, often armed with more recent knowledge, and older physicians with more ingrained transfusion practices. “Sometimes the younger physicians want to be more conservative with transfusions. On the other hand, I’ve heard it the other way around where a senior faculty member wants to follow the guideline, but in the post-op ICU the residents might decide to transfuse a patient simply so the nurse doesn’t wake them up with positive chest tube drainage or hemoglobin data in the middle of the night. So depending on the time of day, there’s an element where individual team members may subvert adherence to the guidelines.”

The situation with transfusions bears a resemblance to that with caesarean sections, says Dr. Goodnough, who has published research on postpartum hemorrhage. “There’s an agreement that the C-section rate is way too high even for first-time deliveries. The guidelines are out there. But maternal mortality today is probably not any better, and probably even worse, than it was 10 to 20 years ago. There are more elderly mothers, more multi-gestational pregnancies, more patients who want to be induced, and other logistical reasons why people are simply not interested in following the C-section guidelines.”

The research on transfusion practice is tending to be unanimous, however. “That’s why we called for the concept that like other indicators listed by the STS guidelines, such as beta blockers, why not also have blood transfusion as a quality indicator of outcomes? This would be consistent with what the Joint Commission is trying to do in recommending quality indicators in accreditation practices, and consistent with a Consumer Reports-style, transparent way of scoring programs to which patients and the public have access.”

As Dr. Hannon sees it, there is an important gap at fault, but not a lack of knowledge. “It’s simply that people aren’t picking up the tools.”

For example, the Society for the Advancement of Blood Management, which just promoted the fourth annual Blood Management Awareness week, reports that about 110 blood management programs exist at hospitals in the U.S.—but there are at least 5,900 other hospitals that don’t have programs. As an aside, he says, some of those 110 programs say they have a “blood management program” even if they have little more than a cell-saver machine.

The Joint Commission, which regards its position as something of a “bully pulpit” for blood management awareness in general, could shape up as an influential force, Dr. Hannon says. “They have certainly raised awareness of this. They have seven draft performance measures for blood management.” The first five of those measures are simply reiterations of existing standards with such standbys as informed consent and documentation of transfusion indications, he notes. But the last two relate to anemia management and blood product availability before surgery.

“I believe the Joint Commission’s first objective is simply getting hospitals to meet existing standards. People are more vigilant if their performance measure moves from the bottom of the clipboard to the top of the clipboard, which is what happens when it goes from a standard to a measurement. These seven measures don’t change the world. They can’t cover everything in blood management, but emphasizing them as measures raises the bar in blood management.”

In his view, the failure of the hospitals to comply with proven guidelines on transfusion reduction relates to broader “failed change” efforts across the board. Well-known change management guru John P. Kotter, Dr. Hannon says, has studied a large number of failed change efforts in Fortune 500 companies and identified eight root causes. As Kotter outlines in his book Leading Change, they are the failures to establish a sense of urgency, to have a guiding team, to develop a vision, to communicate the vision, to remove obstacles to the vision, to plan for and create short-term wins, and to anchor changes in the company’s culture. The final cause: declaring victory too soon.

“But my theory is that the ‘implementation gap’ is not lack of knowledge or tools; it’s lack of implementation skills,” Dr. Hannon says. “And that’s a relatively generic statement for blood management or anything in health care.” Evidence for that, he says, is General Electric’s use of the change acceleration process, which uses the change management model. It acknowledges there is a softer side of change that has to do with social dynamics.

“It’s not just laypeople who think it’s counterintuitive to administer fewer transfusions—I think it’s physicians and nurses. The basic fact is that transfusions have not been well studied in the past. They are largely a grandfathered therapy,” Dr. Hannon says. “It’s been in common use for over 100 years, but it’s really been only in the last decade that it’s been studied closely.”

When such close study does occur, “we come to two conclusions. One is that transfusions are not as beneficial as we previously thought, and the second is that they are more harmful than we previously thought, particularly when we look at what are called non-infectious risk transfusions. When you combine those findings, it doesn’t mean you don’t use transfusions, but you shift the risk-benefit calculation to being more conservative.”

“To my knowledge, there aren’t any medical schools that have incorporated blood component therapy as core curricula,” Dr. Hannon says. “That’s why hospitals end up playing catch-up in trying to educate young docs and old docs and young and old nurses on the risks and benefits.”

But he is optimistic about prospects for change. “As consultants who help institutions with compliance and improving safety within transfusion services and the OR, we get hired to help reduce costs. But the benefit is, oddly enough, quality and safety. And that’s the promise of blood management. We haven’t had the push for the economic side of blood management to this point, but I think that’s starting to change.”

“These two latest studies show again that if you transfuse more, you don’t get better outcomes,” says Susan D. Roseff, MD, medical director of the transfusion service at Virginia Commonwealth University Health System in Richmond and a member of the CAP Transfusion Medicine Resource Committee. “The question becomes: Why would you transfuse more?”

Blood reduction has been a priority at Dr. Roseff’s institution for some time. “More than five years ago, our cardiac surgeons found that by reducing the number of red blood cells they transfused, they could improve their outcomes.”

But such policies are still relatively rare. “Between in­stitutions, transfusion practice continues to be all over the board for every patient population,” she says. Despite the STS guidelines, there are still those who are resistant to change and will be critical of the available data. “They might argue that not all of the data is evidence-based, it was not collected in a randomized controlled trial, or the numbers weren’t large enough, or it’s just one institution, or there were design flaws, and so on.”

Dr. Roseff says the studies are viewed as controversial, as other changes in transfusion practice over the years seemed to be. Some doctors have told her, for example, “The data presented don’t match my patients, so the findings can’t be generalized.”

On the flip side, some practitioners have ethical concerns about changing their practice, she says, summing up their thoughts as: “If I have experience in a group of patients and they are doing well with what I do, then I change my practice and start doing something I’ve never done before, is that ethical treatment? If I believe that a hemoglobin trigger of such-and-such is appropriate, is it ethical practice for me not to abide by my own standard of practice?”

“This type of rationalization that allows them to remain in their comfort zone is a big problem,” Dr. Roseff says.

It’s a discussion that her transfusion service has had with the clinicians at her institution, who might say: “Who are you to tell me I should treat patients differently when they’re doing well? And what if they start not doing well?”

The cardiothoracic surgeons at Dr. Roseff’s institution have strictly adhered to their more conservative transfusion guidelines. “Our surgeons strive to transfuse on the basis of physiologic parameters, rather than just the hemoglobin level, and feel that perfusion is better at lower hemoglobin levels. So they definitely believe in the ‘less is more’ theory of transfusion for many of their patients.”

Partial credit goes to a cardiac surgeon and cardiac anesthesiologist who were particularly zealous about reducing transfusions and persuaded others. “After witnessing some bad outcomes they felt were directly attributable to transfusion,” she says, “including inflammatory reactions and severe hypotensive reactions, they ­began to gather data. When they saw their outcomes improve, they adopted their more restrictive transfusion practices as their standard of care.”

Another factor in VCU Medical Center’s success is the practice of looking at all strategies to reduce transfusion, including the use of intraoperative salvage. The hospital’s chief perfusionist is a part of the blood utilization committee and reports on trends in the different surgical subspecialties each quarter. “Having the ability to introduce new technologies, and to increase awareness of their use, impacts transfusion practice throughout the hospital,” Dr. Roseff says.

The hospital’s blood utilization committee writes letters to people who transfuse outside the accepted guidelines. “As unpopular as this is to some people, the letters do make people stop and think. This ‘administrative’ process really hits home and can be very effective.”

The cardiothoracic surgery program at VCU Medical Center has a rule that residents on call at night can’t order transfusions without the approval of the attending physician. It helps ensure adherence to that standard transfusion practice and gives the residents an opportunity to learn about transfusion. “This is especially important, since medical students may only get two or three hours of formal training in transfusion medicine and that’s it—they’re done.” In addition, the VCU transfusion medicine physicians take an active role in delivering ongoing educational sessions for house officers about transfusion medicine, “but that may not be true in a lot of institutions,” Dr. Roseff says.

Would a broad-based, randomized, controlled trial finally turn the tide in favor of fewer transfusions? Dr. Roseff thinks having additional trials that can be easily generalized might help convince the skeptics. “I still believe people have their own preconceived notions of what they accept and what they don’t.” Even at her hospital, where there is institutional support for a blood management program, “it can still be an uphill battle to try to alter transfusion practice.”

“You have to have the data, and you have to use it to say: Is there a way we can do this better? This isn’t just about less transfusion; it’s about appropriate and thoughtful transfusion. You have to keep challenging people and asking the important questions.” Just as clinicians decide which antibiotic to put a patient on and for how long, “the same decision process should be used for transfusion of blood products.” Eroding preconceptions takes multiple studies and multiple modalities, but eventually it will have an impact on thinking patterns, Dr. Roseff believes. “The more data that come out, the better it will be for us to convince clinicians there are ways to do more with less. I think that’s the bottom line.”

Public rankings of institutions in terms of transfusion practices could potentially play a role in improving practices, Dr. Goodnough thinks. As he notes, it has been well recognized in the scientific and lay literature that once cystic fibrosis programs for children were publicly ranked on quantifiable measures such as survival, “it did not lead to the demise of the programs as feared, but in contrast it re-galvanized people to see how they could do better. And everybody benefited from that.”

“We’re not recommending public rankings,” he says, “but it’s a fact of life that everybody is interested in ranking CVS programs within a state by outcomes. I have no argument with that. The idea of shopping for your cardiologist based on more objective measures is probably a healthy thing.”

Related Links

• New blood transfusion ­studies in brief