Feature Story

Riding up and down
on the ULR seesaw
Countries in which ULR is mandated as public policy

September 2003
Anne Paxton

SARS. West Nile virus. Chagas’ disease. Variant CJD. Bacterial contamination. Hospital error. Pick up the latest newsletters from the American Red Cross and America’s Blood Centers, and these are the issues drawing the most attention and concern. Universal leukoreduction is no longer among the top six.

Nevertheless, ULR remains a $300 million-plus controversy in blood services—one on which expert opinion is divided.

Few would have predicted that outcome four years ago. At that time, most leukoreduction was selective—that is, physicians ordered leukoreduced blood only for patients at high risk for clinical problems that develop because of the presence of leukocytes. But in 1998 the FDA’s Blood Products Advisory Committee agreed that all cellular blood components should be leukoreduced. In Canada and most of Europe, ULR became official policy.

The FDA, it appeared, was poised to make it mandatory, and the majority of U.S. blood centers opted to adopt ULR on their own, even though leukocyte reduction would add at least $30 to the cost of each unit of red cells.

When the FDA took no position, doubts grew about whether that cost was justified, and there was a gradual drop-off in blood centers that provided ULR.

It was a major policy reversal when, in July of this year, the American Red Cross announced it would return to offering both leukoreduced and nonleukoreduced red cells to hospitals.

The Red Cross was among the first to climb aboard the ULR bandwagon, and nearly all of its blood centers had converted. But the organization said a reversal was justified by the lack of an FDA mandate, lack of definitive clinical evidence, and requests from hospitals to allow ordering physicians to decide which patients require leukoreduced blood products.

Today, only about 70 to 75 percent of U.S. blood services leukoreduce all of their blood components (down from a high of about 90 percent). James P. AuBuchon, MD, vice chair of the CAP’s Transfusion Medicine Resource Committee and professor of pathology and medicine at Dartmouth-Hitchcock Medical Center, Lebanon, NH, thinks ULR’s heyday is over. “The importance of leukoreduction for specific patient groups has remained unchanged. However, the tide of universal application of this technology is definitely on the ebb,” he says.

The decision to convert one way or the other is too expensive to be made easily. “We’re in a period where blood suppliers have made large dollar commitments to moving into the arena of ULR,” says Arthur W. Bracey, MD, an advisor to the CAP Transfusion Medicine Resource Committee and director of the transfusion service at St. Luke’s Episcopal Hospital and the Texas Heart Institute, Houston. “Once you make the jump it’s very hard because there are a lot of costs associated with it. But now some of the science is beginning to evolve, and people are asking, Is this really necessary?”

Many see the Red Cross’ reversion to selective leukoreduction as economically driven. Neil Blumberg, MD, professor of pathology and laboratory medicine and director of the transfusion medicine section/ blood bank at University of Rochester Medical Center, NY, says, “The Red Cross provides some of the most expensive blood in the U.S., for all sorts of reasons, and people are threatening to go to other blood suppliers because of the expense.”

But others contend that the Red Cross’ embrace of ULR wasn’t that well founded in the first place. America’s Blood Centers, whose members provide roughly half of the nation’s blood supply, never went as far as the Red Cross on ULR. “From the beginning, we’ve been pro-community choice because some of our members adamantly believed it was right for the community” and others did not, says Jim MacPherson, ABC chief executive officer.

“We had to take a position consistent with what our members believe,” he adds. “That remains our position today, and I think we’ve been vindicated. The reason why the FDA never mandated ULR was there was never a consensus in the medical community about whether it should be standard or not, and the data are still controversial.”

In the United States, expert opinion is split over whether ULR is appropriate, says Laurence A. Sherman, MD, JD, emeritus professor of pathology at Northwestern University, Chicago, member of the CAP Council on Scientific Affairs, and former co-chair of the CAP’s Transfusion Medicine Resource Committee.“Some benefits have been advanced, some denied, and the quantitative benefit has varied even in the same institution,” he says.

But north of the border, where blood services have more than four years’ experience with universal leukoreduction, the take on ULR is different. “We follow very closely what happens in the U.S.,” says Graham D. Sher, MD, PhD, chief executive officer of Canadian Blood Services. “And if one looks at the literature and listens to the meetings, the prevailing sense is the evidence in favor is certainly not getting any stronger, and there’s possibly some evidence that some patient groups may not benefit.”

“However, we’ve just sponsored two very large studies in Canada on the benefits of nonleukoreduced versus leukoreduced blood, and although they are marginal, there are benefits for both pediatric and adult patients,” Dr. Sher adds.

Reported in the April 16 Journal of the American Medical Association (2003;289:1941–1949), the larger of the studies compared the health of 6,982 Canadian adults who received blood transfusions before the Canadian law requiring ULR with the health of 7,804 people who received transfusions afterward.

Lead investigator Paul C. Hébert, MD, MHSc, and his colleagues discovered that patients in the second group were less likely to die while in the hospital and less likely to develop a fever or require antibiotics, although the incidence of postoperative infections was the same.

Those findings did not surprise Edward L. Snyder, MD, professor and associate chair for clinical affairs with the Department of Laboratory Medicine at Yale University. At Yale-New Haven Hospital, the benefits of converting to ULR (which the entire state of Connecticut did in 2000), have been substantial.

Dr. Snyder maintains that the well-known difference in febrile reaction incidence has been downplayed too much. When the FDA backed down on mandating ULR, opponents of leukocyte reduction believed even more fervently that ULR wasn’t necessary, he says. “It was very cavalier of these opponents to assume the acceptability of febrile reactions with severe chills for someone else,” he says.

His hospital’s primary intent in converting from selective to universal leukoreduction was to decrease the incidence of febrile reactions and prevent as much as possible the occurrence of transfusion-related HLA, alloimmunization, and CMV transmission, he explains.

Since January 2000, when the conversion was complete, “we have seen a drop in HLA [human leukocyte antigen] antibody, we have not seen transmission of CMV by blood transfusion, and our incidence of febrile reactions to red cells and platelets dropped 45 percent and 95 percent respectively, compared to when we were only using 25 percent bedside leukoreduction,” he says. That means “we’ve gone to only half a dozen cases a year of febrile reactions to platelets, whereas we used to see six to 12 a month.”

Dr. Snyder, who notes he is a consultant to Pall Corp., a manufacturer of leukocyte filters, stresses that no one was recruited for the Yale study, which has been submitted for publication. “This was an evaluation of febrile transfusion reactions in patients reported to the blood bank by hospital staff, so there was minimal potential for bias,” he says. He adds that comparable results have been found in other studies.

While the significance of the Canadian study has been questioned by experts like Dr. AuBuchon, who co-wrote (with Howard L. Corwin, MD) an editorial appearing in the same issue of JAMA (2003; 289: 1993–1995), Dr. Snyder believes it is compelling evidence of ULR’s value: “Every time something comes out, it’s criticized as not being the quintessential paper. I’m not trying to claim ULR will reduce cancer, but I feel everyone should have the opportunity not to have a febrile reaction. We’re talking about patients with reactions that are not life-threatening—but they’re distressing.”

Dr. Blumberg says ULR has virtually no negative clinical side effects and believes the issue is fairly straightforward.

Infection, he stresses, is the most common costly and potentially lethal complication of surgery, which accounts for 50 percent of the blood used in the United States. “Whether LR of transfusions for surgical patients reduces postoperative infections, morbidity, and mortality is debated, but the data are quite clear,” he says. “Eight randomized trials have looked at it, and in six there is clear-cut evidence of a 50 percent reduction in postoperative infections. That means one out of every 10 patients who would have developed an infection with nonleukoreduced blood had that infection prevented.”

A before-and-after study such as Canada’s is not as powerful as a randomized trial, Dr. Blumberg admits. “But they found fairly strong evidence of reduced complications in newborns, fewer antibiotics, fever, and mortality. It doesn’t sound like a lot, but one percent means you only need to treat 100 patients to get savings and to actually prevent one in-hospital death. We wish we had antibiotics that were that good.”

“At the height of the HIV epidemic in San Francisco,” he continues, “one in 100 patients may have become infected with HIV before we had testing available. Here we’re talking about one in 100 who may get their life saved today. So if they’re correct, then ULR is the single most important advance in transfusion medicine since infectious disease testing.”

The meta-analyses of randomized trials cited by URL opponents have reached misleading conclusions, Dr. Blumberg believes, because they are methodologically unsound and scientifically invalid. “We’re in the process of performing a reanalysis of the literature that I think will convince many people they’ve been led down the wrong path.” The original studies are clear, he says, but they’ve been subjected to secondary analyses. “These analyses include nonexistent data from patients who received no transfusions whatever,” thus minimizing the likelihood of seeing benefits in the patients who were actually transfused with leukoreduced blood, he says.

Critics of the Canadian study, on the other hand, maintain that since it was retrospective, many of the improvements detected could be ascribed to factors unrelated to ULR. Dr. AuBuchon believes that a study at Massachusetts General Hospital, published last year in Transfusion (Dzik WH, Anderson JK, O’Neill EM, et al. 2002; 42: 1114– 1122), is the best designed and most significant study conducted so far.

“This is the first and to my knowledge only study to look at universal application of leukoreduction in a large general hospital, where it was applied to all patients who did not have a specific indication. And it clearly showed that ULR did not offer the benefits that had been proposed in other smaller studies,” he says.

Some papers reporting on studies conducted outside the United States have shown benefits, but their findings could not be replicated in this country, Dr. AuBuchon adds. “I wish I were smart enough to figure out why, but clearly one cannot take studies from one locale and apply them elsewhere.”

Although the Canadian study was well designed and thorough, Dr. Bracey warns that the problems in retrospective studies should not be discounted. “It’s important to remember that the care of these patients is not a static target. Hospitals are continually trying to intervene and reduce the number of negative outcomes, with measures ranging from the dressing on the wound to cleaning the patient’s mouth,” he says.

“So if you look at one time versus another, there are an incredible number of confounders, and you need a randomized, prospective study to overcome them.”

Working in a cardiovascular environment, Dr. Bracey is particularly suspicious of claims that large absolute reductions in mortality could be credited to blood policy. “Blood might play an important role, but that’s really making it the centerpiece,” he says. A randomized trial at his institution, which he presented at last year’s American Association of Blood Banks meeting, found no difference in mortality, length of stay, or incidence of infection for routine versus leukoreduced transfusions.

Blood service directors can be accused of a kind of doublespeak, he admits. “On one side, you say you will do everything you can, even in the absence of hard data, to make blood as safe as possible without any constraints in overall cost-containment or efficiency. On the other hand, you have to ask, If the data are not that convincing, should we really be committing resources knowing we have a limited ability to spend on health care? Years ago an expert government economist said the American public will demand we do everything we can, but I think that theory needs to be tested by the American public,” Dr. Bracey says.

“We’re still sort of shell-shocked from the aftermath of HIV, and I personally would never like to see such negative outcomes from transfusion, but are we overreacting? That’s the critical question.”

The CAP has been involved in an advocacy program to put the decision about leukoreduction back in the realm of the medical practitioner, Dr. Bracey says. “There’s been communication between the College and the FDA and other elements of government in that regard, so I think that’s an important effort.”

From each blood center’s standpoint, there is a tipping point in decisionmaking about ULR because of the logistics of inventory management. “If you have hospitals committed to 100 percent ULR,” MacPherson says, “and their use of blood is going to vary from week to week, how much do you have to have to make sure they’re all covered, while at the same time having enough unleukoreduced blood to meet the needs of other customers?”

“If only 15 percent want nonleukoreduced blood, you can probably meet most of their needs. But they say once you cross the threshold of 30 or 40 percent LR blood, then it becomes a nightmare,” he says. “You never really know. You’re always going to wind up sending something nobody wants, and it’s difficult to manage something with such a short shelf life as blood.”

From an inventory management standpoint, a universal approach is much easier, agrees Dr. Sher of Canadian Blood Services. “Your products are all standardized, you avoid the physician saying ‘I want product A’ when product B is the only one available, and no patients are likely being harmed,” he says.

Dr. Bracey is sympathetic to the logistics of maintaining separate inventories, but he notes that competition from non-ULR sources drove the Red Cross back to a mixed inventory in place of systems offering exclusively leukoreduced blood. “In North Carolina,” he says, “hospitals formed a consortium to collect their own blood, and what they’re offering is not 100 percent leukoreduced, so the Red Cross needed to make this move in order to compete in that market.”

As an example of another complication, he adds, at St. Luke’s “we’re using standard blood as our primary blood product, but we have select categories of patients where we’ve increased the amount of leukoreduced blood we use because our secondary supplier is 100 percent leukoreduced. If we were following the guidelines, we’d probably use 20 percent LR blood, but we’re currently using 35 percent to 40 percent. That’s because of the oversupply of our secondary supplier.”

ULR was adopted in Canada because of the many benefits it was thought to offer transfusion recipients, but it was adopted in Europe, primarily in the United Kingdom, not because of its reduction of febrile reactions, according to Dr. Snyder. “It was because of concerns over leukocytes’ role in the spread of vCJD [variant Creutzfeldt-Jakob disease],” he says. “Hearkening back to the HIV epidemic, some French government officials had actually been put in jail under very old statutes in French law for failing to act in the public interest.”

“Nobody wanted to be in power in a country when ‘AIDS II’ hit and they had done nothing to protect the people, but it seemed that experimental animals were less likely to get the disease if these white cells were missing. Now we know vCJD is apparently not readily spread through white cells. So ULR didn’t pan out, but they’ve kept it anyway because of a reluctance to go back.”

The situation in the United States is different. “The Canadian medical and political climates have been tumultuous ones over the last decade,” says Dr. AuBuchon, “and there are many forces that appear to be involved in the decisionmaking process that are not extant in this country or do not have the same power.”

The cost of ULR to Canada was considerable—a $35 million to $40 million addition to the national blood system budget in 1999. U.S. hospitals are much more sensitive to price increases, Dr. Sher notes. “Of course, the blood system in Canada is very different in funding and governance in that hospitals and patients do not pay. We get funding at a global level from provincial and territorial compartments, and we provide it at no cost to the hospital.”

“We haven’t seen the same sort of ‘pushback’ around costing here,” agrees Dana Devine, PhD, executive director of research and development for Canadian Blood Services in Vancouver, British Columbia, and a coauthor of the Hébert study.

“In my view, the real truth is at the moment we don’t know whether it’s been worth the cost,” she adds. Other medical interventions may be compared using a dollar-to-life scale, she says, “but that scale doesn’t seem to apply to the blood systems of the world. We make financial expenditures in blood systems that we probably wouldn’t make elsewhere.”

Dr. Snyder says conversion to ULR at Yale-New Haven Hospital cost about $400,000, and they were able to cost-shift some savings in CMV testing. “I can’t say we had substantial cost savings by not having to work up febrile reactions, but there were some savings. Also, there were savings from the quantity discount on blood products from the American Red Cross.”

“The added cost is somewhere in the $30 to $40 range for a unit of red cells,” he adds, “and you have to decide if a decrease of 94 percent in febrile reactions is worth spending that kind of money.”

Would the payback from other blood safety measures be much higher, as has been suggested? Dr. Snyder is not convinced: “Some people say it would be far better to develop a system of labeling blood at the bedside to make sure you have the right blood product to the right patient, but I don’t think you should put all your resources into bar-code systems,” which can be defeated or worked around.

While some have suggested, too, that ULR may be replaced by pathogen reduction, Dr. Snyder says, “the question is whether pathogen reduction is going to obviate the need for leukoreduction. I don’t think it will be because febrile reactions can still occur.”

Blood centers in this country have already spent at least $300 million converting voluntarily to ULR. Could the added cost of using leukoreduced blood continue to drop if more blood services embrace it? MacPherson is doubtful. As more filters are made and sold and competition among manufacturers has increased, the price has stabilized, but he suggests that achieving much more cost-efficiency might not be possible. “When you get down around the $10 to $12 level for the actual cost of the filter, plus labor, you’re not going to go a whole lot lower.”

But doubling the number of red cell inventories to eight can lead to less obvious cost increases, Dr. Blumberg says. Although the Red Cross will now make nonleukoreduced blood available in response to demands for cheaper blood, “that means many patients who should get it won’t because they’ve run out of one of the eight inventories, and people will be under the misguided apprehension that it won’t cause suffering of patients and will save money because the blood bank budget will be less. But the hospital will spend more.”

What makes cost-effectiveness determinations difficult, Dr. Blumberg notes, is that the costs of ULR show up in the blood bank expenses. “The benefits may be $2 or $3 or $4 million, but they will be in shortened length of stay, cost of antibiotics, ventilator time, or ICU time—not to mention that the patients get to go home for Christmas instead of to the mortuary. It’s in the clinical milieu that the savings occur.”

It’s difficult to convince hospital management. “I’ve shown our hospital administrators that when we implemented ULR, the savings significantly outstripped the costs,” he says. But at most institutions, “everybody has an ax hanging over their head. The bean-counter mentality that prevails among most administrators will make them tell the blood bank, ‘Get your budget back to where you were.’”

Medicare payment policy adds other complications. “Blood has not been in the Medicare/Medicaid market basket updates for the last five or six years,” MacPherson points out. “It was taken out in the 1990s when the cost of blood was relatively stable; it was below the CPI for health care.”

“But then because of nucleic acid testing, leukoreduction, other tests, and now the potential promise of pathogen reduction, we’ve been seeing huge increases in the last few years,” he says. ABC, the Red Cross, and AABB have pushed hard to get Congress to include blood in the market basket, and their efforts have paid off.

“In recent Medicare budget bills, Congress has added money to catch up with the price increases in blood, and in the next fiscal year blood will be in the market basket update, and they will be tracking the cost of blood,” MacPherson says.

But this move has to be put in perspective, he cautions. “Blood is only, at the most, four percent of the patient’s hospital bill. It’s really small potatoes, and if you increase the reimbursement by 10 or 30 percent, you’re still talking about very little amounts of money. The real question is, Is there sufficient reimbursement?, period.”

On this question as well as on ULR, MacPherson believes prospects for more studies are slim. “It’s very, very difficult because they have to be very long term, you need a large number of patients, and there are other higher priorities in this country. When you’re talking about real deaths and real threats, those take No. 1 priority.”

The political environment has shifted, Dr. Blumberg concedes. “Patients are not as worried about getting AIDS, so ULR looks like a political nonstarter.” But he urges, “It’s time to start arguing about the data in a way that’s constructive, because it’s extremely clear.”

Until there is a national consensus, Dr. Sherman would prefer a pragmatic approach. “We continue to scramble for an adequate blood supply,” he points out. “We do not have systematic, broad, ongoing data on blood availability, nor do we more than partially understand donor motivation and recruitment. This is an area where commitment of more resources is needed.”

He notes that leukoreduction was thought only a few decades ago to improve kidney graft survival, and the opposite turned out to be true. “We believe ongoing re-evaluation of any therapeutic modality is wise,” he says.

The greatest risks to transfusion recipients have been well documented, Dr. AuBuchon notes. “For recipients of red cells, the risk comes from mislabeling of samples at the time of collection of the specimen and transfusion of the unit to the wrong patient. For platelet recipients, it comes from bacterial contamination. And I’m pleased both the CAP and the AABB are now directing laboratories’ attention to these issues. The other major risk faced by any recipient of a plasma-containing component is transfusion-related acute lung injury, or TRALI, and as yet we do not have a good means of reducing this risk.”

All health care systems choose which interventions to fund, whether the choices are deliberate or not, Dr. Sherman stresses. “When I’ve gone to Canada and read complaints in the newspapers about the unavailability of ER and other services, I’d say Canadians have made a choice between certain kinds of services. When people here tell me they are being patient advocates, what they usually mean is they’re being advocates for transfused patients or diabetic patients”—possibly to the detriment of other types of patients, he says.

But the real question is, What other service would lose out if ULR became official policy? “Medicare is not going to increase the total budget if we went to ULR,” Dr. Sherman says.

“Ideally, if you went to the hospital with, say, a mild pneumonia, and you had no underlying cardiac or pulmonary problems, you’d want to be in an intensive-care unit just on the off-chance that a one-in-10,000 complication could occur that would not have been apparent on your admission. You could also say if a previously unsuspected anemia were present, you’d want to have all your blood leukoreduced, or that you only want blood from a blood drive you had at a convent.”

“That’s what everybody would want if you could have your druthers,” Dr. Sherman says. “But you can’t say that’s the best way to spend medical dollars. You have to decide at what point health care dollars would be better spent on other things that will benefit more people as a whole.”
Anne Paxton is a writer in Seattle.